Method and system for constructing a unary, binary, ternary autopoietic system, and storage medium

Abstract

The application discloses a kind of unary binary three-body self-generation system construction method, system and storage medium, belong to complex system engineering management and control, industrial adaptive control, distributed intelligent system cross technical field.The application takes double-layer operation identity as the only core convergence target of whole life cycle, takes the closed-loop recursive isomorphism constraint of three big irreducible function primitives of information field state C, energy mechanism E, material structure M as ontology identity core guarantee, is synchronously run through self-reference and he points dual-mode whole life cycle, endogenously constructs space-time dimension and five-dimensional holographic coherence state tensor, combined with gradient safety control mechanism, realizes the self-generation, self-maintenance, self-repair, self-evolution of system.The application can be adapted to electronic and non-electronic entity system in whole scene, completely solves the pain point that traditional system ontology drifts for a long time, evolution out of control, significantly improves the reliability, adaptability and autonomous evolution ability of complex system.

Description

Technical Field

[0001] This invention belongs to the cross-technical fields of complex system engineering management and control, industrial adaptive control, and distributed intelligent systems. Specifically, it relates to a method for constructing a univariate two-state three-body self-generating system, a system, electronic equipment and storage medium, and computer program products. Background Technology

[0002] The theory of autopoietic systems was initially used to describe living systems, which are characterized by closed organization and open structure. Later, it was used to describe the core essence of complex systems: "self-generation, self-maintenance, self-repair, and self-evolution." It can explain the evolutionary laws of objectively existing systems, from quantum mechanics, atoms, and life to ecology and civilization, and provides underlying theoretical support for breaking through the architectural bottlenecks of existing adaptive control and intelligent systems. Existing technologies, based on automatic control, machine learning, and large language models, have formed technical solutions involving closed-loop state acquisition, state evaluation, rule updating, and iterative cycles. However, these solutions still have the following shortcomings in engineering applications: First, existing solutions cannot reconcile the core contradiction between "endogenous rule evolution" and "ontology survival security." Existing technologies fall into two extremes: one type of solution solidifies the system objective into a single, unadjustable rigid constraint, completely locking the system's endogenous evolution space. Essentially, it still belongs to adaptive control based on preset rules and is not a true self-generating system. The other type of solution allows rules to explore adaptive operation, which is prone to problems such as damage to the system ontology and steady-state collapse during rule evolution, violating the core essence of self-generating systems—"self-sustaining and self-surviving"—and failing to meet the high reliability requirements of industrial scenarios. Second, existing solutions lack a unified underlying architecture for self-generating systems. Various adaptive intelligent systems generally adopt a reductionist architecture of "preset rules + feedback correction" and a Newtonian external spacetime container benchmark, which is prone to time delays and cannot achieve true endogenous evolution. Third, existing solutions are highly dependent on specific scenarios, have fragmented architectures, lack universal and adaptable construction methods, and lack a unified and reproducible engineering implementation chain, severely limiting their application in non-electronic and non-computing scenarios such as industrial hydraulics and mechanical transmission. Summary of the Invention

[0003] To address the aforementioned core deficiencies of existing technologies, this invention proposes a complete method for constructing self-generated systems based on the original monistic two-state three-body self-generated system theory. By adopting a monistic two-state three-body architecture, it fundamentally solves the core pain points of the engineering implementation of self-generated system theory, providing a system ontology for the future engineering development of self-generated intelligent agents and swarm intelligence.

[0004] The aforementioned unary two-state three-body self-generating system refers to a dynamic whole that continuously and recursively generates and maintains its own stable evolution by taking the identity of two-layer operations as the convergence goal, the dynamic and stable existence of the unary ontology as the operational boundary, the self-referential and other-referential dual modes as the endogenous driving force, and the closed-loop recursive coupling of three irreducible functional primitives—information field state C, energy mechanism E, and material structure M—as the minimum operational link. In this application, the three functional primitives—information field state C, energy mechanism E, and material structure M—correspond to the three elements of the unary two-state three-body self-generating system: information field state, energy mechanism, and material structure. The attributes and characteristics of information field state, energy mechanism, material structure, and their relationships with information, energy, and matter are defined as follows:

[0005] Information field state C: The sum of all relationships, rules, constraints and potential possibilities in the system. It is the core guiding unit of the system, responsible for rule formation and constraint control. Its core characteristics are global diffusion, non-local correlation and potential orientation. Information is the sum of the patterns, order and meanings presented when the information field state of the system participates in coherence. Information attributes are manifested as different information characteristics under different observation operations in different system states.

[0006] Energy mechanism E: The dynamic sum of all dynamic processes, metabolic cycles and free flow networks in the system. It is the core transmission unit of the system, responsible for power diversion and driving execution. Its core characteristics are process, transformation and dynamics. Energy is the sum of the work potential, activity intensity and metabolic rate and other attributes that the system's energy mechanism exhibits when it participates in coherence. Energy attributes manifest different energy characteristics under different system states and different observation operations.

[0007] Material structure M: The organized sum of all physical carriers, actuators, spatial configurations and topological connections in the system. It is the core carrier unit of the system, responsible for physical execution and feedback acquisition. Its core characteristics are locality, embodiment and spatial configuration. Matter is the sum of the extension, form, texture and composition attributes of the system's material structure when it participates in coherence. The material attributes manifest different material characteristics under different system states and different observation operations.

[0008] Firstly, this invention provides a method for constructing a univariate two-state three-body self-generating system.

[0009] This invention provides a method for constructing a monistic two-state three-body self-generating system that is applicable to all scenarios, reproducible, and verifiable. The method is executed by an electronic computing device with sensing, computing, storage, and execution capabilities, and is used to construct self-generating capabilities for any target system with self-organizing potential, achieving self-generation, self-maintenance, self-repair, and self-evolution of the target system. The method uses the identity of two-layer operations as the unique core convergence objective throughout the entire lifecycle of the target system, the monistic ontology survival index as the core quantitative basis, the synchronous operation of self-indicating and other-indicating modes throughout their entire lifecycle as the endogenous evolutionary driving force, and the closed-loop recursive endogenous generation of three irreducible functional primitives—information field state C, energy mechanism E, and matter structure M—as the unique minimum operational link. The specific technical solution includes the following eight continuously executed steps, the entire process of which is shown in Figure 1:

[0010] Step S1: Bootstrapping and Isomorphic Mapping. This step aims to address the core technical challenges of existing technologies, such as the reliance on externally preset convergence targets for closed-loop system bootstrapping, lack of system ontology anchoring, inability to generate initial operating parameters endogenously, and the susceptibility to ontology shift and behavioral instability during long-term iterations. The key to this step is to anchor the rigid core layer operations of the target system's univariate ontology and the dynamic evolution layer operations to construct the core architecture of a univariate two-state three-body self-generating system with CEM closed-loop recursive isomorphic constraints.

[0011] Execution timing: System first power-on and initialization phase.

[0012] Core basis: underlying physical or rule-based security constraints and initial state parameters of the target system.

[0013] Operation content: (1) Collect the initial state parameters and insurmountable underlying security constraints of the target system, anchor them as the rigid core layer operation identity of the self-generated system, set them to a read-only state throughout the entire life cycle, prohibit any evolution or modification operations, and solidify the system's security bottom line; (2) Construct the core architecture of the self-generated system with CEM closed-loop recursion, and generate isomorphic rules endogenously through CEM closed-loop recursion within the architecture. Within the physical or rule range of the target system, endogenously generate the initial rule set of the information field state C primitive, the energy mechanism E primitive, and the material structure M primitive corresponding to the operation identity of the dynamic evolution layer of the self-generated system, and endogenously initialize the mapping weight matrix. With initial bias parameters Complete the architecture and binding of the self-generated system uni-ontology, determine the core operating carrier of the system, and the CEM closed-loop mapping principle is shown in Figure 2; (3) Combine the initial CEM primitive with the basic dimension of the endogenous spatiotemporal phase space to construct the initial five-dimensional holographic coherent state tensor. (3) Calculate the initial global coherence: Since the initial state is a perfect steady state, the initial global coherence is directly obtained. , solidify as the only steady-state reference for the entire process; (5) initially build the endogenous spatiotemporal phase dimension, define the initial parameters of the three dimensions of time, space and phase, and lay the dimensional foundation for subsequent primitive reshaping and tensor generation.

[0014] Output: Rigid core layer constraints, initial CEM primitives, initial mapping weights and bias parameters, initial five-dimensional tensor Initial global coherence The initial endogenous spatiotemporal phase dimension provides a safety baseline, steady-state benchmark, and initial operating framework for the entire process. All subsequent steps are carried out based on this. Without the initial benchmark, there is no basis for subsequent adaptive judgment.

[0015] The core logic of this step is as follows: without pre-setting external task objectives or artificially fixed operating order, only the insurmountable underlying security constraints are locked when the bootstrap starts. The underlying security constraints are only a mapping of the objective physical boundary and legal security requirements of the target system. No artificially preset operating rules, convergence states and evolution directions are set, so as to ensure the endogenous nature of the system's evolution direction as much as possible. After the bootstrap is completed, when the second closed-loop operation starts, the endogenous univariate ontology survival index is the only endogenous convergence target.

[0016] All endogenous generation operations in this step are executed by the processor of the electronic computing device. The initial state parameters collected and mapped, as well as the generated rules, vectors, and tensors, are stored in non-volatile computer-readable storage media to ensure the traceability and immutability of the initial state.

[0017] In this application, the closed-loop recursive endogenous generative isomorphism of CEM must simultaneously satisfy three core constraints, none of which can be omitted: (1) Dimensional closure constraint, including the tensor dimensions of C primitives, E primitives, and M primitives being completely closed, ensuring that the closed-loop endogenous generative mapping can be iterated infinitely without dimensional breaks; (2) Topological isomorphism constraint, the composite closed-loop endogenous generative mapping must satisfy , to ensure that the system always maintains the identity of the entity during the dynamic evolution process, and to solve the problem of entity offset in long-term iteration from the root; (3) global stability constraint, the Lipschitz constant of the composite closed-loop endogenous generation mapping is strictly less than 1, and the constraint is achieved by the spectrum normalization operation, ensuring that the system converges to the only global asymptotically stable fixed point that satisfies the isomorphic constraint, and completely avoiding the engineering risks of system divergence and oscillation.

[0018] Specifically, the CEM closed-loop recursive mapping refers to the mathematical rules, dimensional closure constraints, and convergence requirements of the information-guided mapping, energy-driven mapping, and structural feedback mapping among the self-generated system information field state C primitive, energy mechanism E primitive, and material structure M primitive constructed in this invention, satisfying... This ensures that the monistic ontology of the self-generated system remains intact throughout its entire lifecycle. Among these, the benchmark mapping... Including an immutable core and a variable shell, this application divides it into a rigid core layer and a dynamic evolution layer, corresponding to the organizational closure and structural openness of the self-generated system, respectively. It then implements this concept through a two-layer operational identity approach, combining rigid and dynamic operational identity. The quantitative expression for this two-layer operational identity is the unique operational identity fixed point throughout the system's entire lifecycle.

[0019] in For rigid operational identity, that is, the rigid core layer operational identity of the target system, the underlying security constraints that cannot be modified in the information field C primitive, the core rules for the survival of the unary ontology, the mathematical form and dimensional closure constraints of the CEM closed-loop recursive mapping are locked in a read-only state throughout the entire life cycle.

[0020] For dynamic operational identity, i.e., the operational identity of the target system's dynamic evolution layer, the corresponding endogenously evolving set of scenario adaptation rules in the information field C primitive, and the dynamic weights and bias parameters of the CEM mapping, are determined by the univariate ontology persistence index. Real-time representation.

[0021] The CEM closed-loop recursive endogenous generative isomorphism rules include:

[0022] (1) The closed-loop endogenous generation and flow rule is a two-way dialectical coupled closed loop of forward recursive generation and reverse recursive constraint of C primitive, E primitive, and M primitive. The basic flow path is: In the formula, , , These are the state vectors of the information field state C, the energy mechanism E, and the matter structure M, respectively. Information-guided mapping for the endogenous generation of E primitives from C primitives. The energy-driven mapping for the endogenous generation of M primitives from E primitives. The structure feedback mapping for the endogenous generation of C primitives from M primitives is a Lipschitz continuous mapping. The CEM closed-loop recursion is the unique minimum operational link for the endogenous generation of isomorphisms within the self-generated system, which is different from the eight-step closed-loop operational link of the self-generated system construction method in this invention.

[0023] (2) The closed-loop recursive isomorphism constraint is the core guarantee of the identity of the univariate ontology of the self-generated system, satisfying the formula: In the formula, For function composition operators, the composition order strictly corresponds to the closed-loop endogenous generation flow path; They are topologically isomorphic and equivalent; It is an identity mapping on C primitives, that is, a reference mapping that does not change the C primitive ontology.

[0024] (3) The closed-loop self-consistency deviation rate is within a reasonable range. The calculation formula is: In the formula, For the first The C-ary state vector of the iteration period. for Norm, To prevent small positive numbers from being divided by zero, the range of values ​​is: Closed-loop self-consistency deviation rate Used to verify the degree of isomorphism constraint satisfaction and system convergence state, the initial value of the preset compliance threshold is generally set to 0. ,when At that time, the closed-loop recursive endogenous generation isomorphic self-consistency of CEM is determined to be qualified. CEM recursive coupling self-consistency matching degree. Deviation rate with closed-loop self-consistency Negative correlation, can be indirectly obtained: .

[0025] (4) Lipschitz constant of the endogenous generative mapping of the composite closed loop The parameterized form of the single-step endogenous generation mapping satisfies the formula: In the formula, , , This is the weight matrix for the corresponding mapping. , , The bias parameters corresponding to the mapping are all optimizable parameters of the dynamic evolution layer; spectral normalization is performed on the three weight matrices in each iteration cycle, and their maximum singular values ​​are restricted to less than 1 to ensure the Lipschitz constant of the composite closed-loop endogenous generative mapping. ;

[0026] The CEM three-element composite closed-loop recursive generation isomorphic mapping formula is: in, For the current iteration period The learnable Lipschitz continuous mapping matrix at each time step can be incrementally updated according to the system's operating state. The specific learning algorithm can adopt conventional optimization methods in this field, such as gradient descent, Adam, etc., and this invention does not limit it. , , The three primitives have equal feature dimensions, satisfying the hard constraint. During the fractal process of a self-generated system, the isomorphic mapping matrix of the subsystem maintains strict topological isomorphism with that of the parent system. The subsystem also inherits the spectral normalization constraint of the parent system, and performs spectral normalization operation on its own isomorphic mapping matrix in each iteration cycle.

[0027] An optional implementation method is a dual-mapping construction mechanism for non-electronic entity systems. For non-electronic, non-computational entity target systems, such as mechanical transmission systems, fluid pipeline networks, heat exchange systems, chemical reaction vessels, and building structures, this step uses a dual-mapping mechanism of "electronic equivalent port mapping + computational tensor isomorphic mapping" to complete the isomorphic construction of the non-electronic system and incorporate it into the CEM closed-loop recursive management system. The core idea of ​​this mechanism is to abstract the input and output behavior of physical entities into equivalent electronic port models, and then map the port parameters into computable tensors aligned with the C and E primitive dimensions, thereby enabling the non-electronic system to participate in closed-loop recursive operation like a self-generated system. It consists of the following two steps, S101 and S102:

[0028] S101 Electronic Equivalent Port Mapping. The core objective of this step is to standardize and encapsulate the physical interaction interface of a non-electronic entity target system into a controllable drive equivalent port and a state feedback equivalent port that are fully compatible with electronic computing systems, thereby achieving interface unification from the physical world to the digital world.

[0029] (1) System body boundary and safety constraint calibration. First, a full-dimensional engineering analysis of the target system is performed to accurately calibrate the following irreducible core elements:

[0030] Irreducible body boundary: Defines the smallest independent functional unit of the system. The boundary division principle is "indivisible internally and independently controllable externally". For example, the body boundary of a hydraulic cylinder system is the cylinder body, piston, seals, and oil inlet / outlet ports; the body boundary of a chemical reactor is the reactor body, agitator, heating / cooling jacket, and material inlet / outlet ports.

[0031] Underlying rigid safety constraints: Identify insurmountable physical limits and safety red lines during system operation, including: maximum / minimum input thresholds (e.g., maximum pressure, maximum flow rate, maximum torque), maximum / minimum output thresholds (e.g., maximum displacement, maximum temperature), environmental tolerance limits (e.g., temperature range, vibration amplitude), and failure critical conditions. All constraints will synchronously anchor the rigid operational identity within the two-layer operational identity framework. .

[0032] Controllable input dimension: Enumerate all external control variables that the system can accept, with each control variable corresponding to an independent controllable drive equivalent port. For example: pressure / flow rate of hydraulic system, force / torque of mechanical system, thermal power of thermal system, and reactant feed rate of chemical system.

[0033] Observable output dimensions: Enumerate all measurable state variables of the system, with each state variable corresponding to an independent state feedback equivalent port. Examples include: displacement, velocity, acceleration, temperature, pressure, flow rate, concentration, and vibration spectrum.

[0034] (2) Port static characteristic calibration. Through offline testing, consulting the manufacturer's technical manual, or finite element simulation, establish the static input-output mapping relationship for each equivalent port:

[0035] Static calibration of the drive port: Establish a mathematical model between the control input and the steady-state output of the system. For linear systems, a linear function is used for fitting. For nonlinear systems, piecewise linearization or polynomial fitting is used. Example: Input voltage of an electronically controlled proportional valve. (0-10V) and output flow rate The relationship (0-100L / min) is as follows The intercept of 0.2 indicates that there is still a slight leakage at zero voltage.

[0036] Feedback port static calibration: Establishing the conversion relationship between physical quantities and standard electrical signals (0-10V, 4-20mA, digital quantity). Example: Pressure sensor output current. (4-20mA) and pressure The relationship between (0-10MPa) is as follows: .

[0037] (3) Port dynamic characteristic identification. By applying standard excitation signals (step signal, sinusoidal sweep signal, pseudo-random binary signal), the dynamic response characteristics of the system are identified, and the transfer function or state-space model is established:

[0038] First-order inertial systems: applicable to slowly varying systems such as temperature and pressure, with steady-state gain as the identification parameter. and time constant The transfer function is .

[0039] Second-order oscillatory systems: applicable to systems involving mechanical motion, fluid fluctuations, etc., with the natural frequency as the identifying parameter. Damping ratio The transfer function is .

[0040] Nonlinear compensation: For systems with strong nonlinearity such as dead zone, saturation, and hysteresis, a nonlinear modeling method based on neural networks is used for compensation to ensure that the linearity error of the equivalent port is ≤5% over the entire operating range.

[0041] (4) Standardized Equivalent Port Encapsulation. The calibrated drive ports and feedback ports are physically connected via industrial-grade data acquisition and control hardware (PLC, embedded controller, data acquisition card, edge computing gateway), and encapsulated in the software layer as standardized objects with a unified interface. Each port object contains the following standardized attributes and methods:

[0042] Attributes: Unique port ID, port type (driver / feedback), physical quantity unit, measurement range, calibration parameters, safety constraints, real-time value, timestamp;

[0043] Methods: Read port value, write port value, parameter configuration, security check, and fault diagnosis.

[0044] S102 Computational Tensor Isomorphism Mapping. The core objective of this step is to map the parameters of the physical ports into computable state vectors that are fully aligned with the C and E primitive dimensions of the self-generated system through orthogonal transformation, thereby achieving tensor space isomorphism between the physical and digital systems.

[0045] (1) Alignment of physical parameter dimensions. Let the standard feature dimensions of the self-generated system C and E primitives be . (Supports 16 / 64 / 128 / 256 / 512 dimensional standard settings). First, organize the real-time parameters of all equivalent ports into a physical parameter vector. ,in This is the sum of the number of drive ports and the number of feedback ports.

[0046] when Time: Principal Component Analysis (PCA) algorithm is used for dimensionality reduction, retaining the previous... Each principal component has an information loss rate of ≤5%.

[0047] when Time: Expand the vector using zero-padding. The filling position corresponds to the unused semantic dimension in the C / E primitive.

[0048] (2) Orthogonal transformation tensor mapping. Design column orthogonal transformation matrices. ,satisfy (Ensuring the transformation is reversible and information is lossless), map the physical parameter vector to the C primitive state vector: Similarly, define the transformation matrix. The mapping yields the E primitive state vector. The initialization of the transformation matrix adopts a hybrid strategy of "prioritizing physical semantics + data-driven optimization":

[0049] Based on the physical semantics, each physical quantity is mapped to the dimension in the C / E primitive that has the closest semantics, with a weight of 1, and the weights of the other dimensions are set to 0.

[0050] Based on operational data optimization, 1000 sets of input and output data were collected in the initial stage of system operation, and the transformation matrix was optimized through linear regression to minimize the mapping error.

[0051] (3) CEM closed-loop recursive link access. The mapped C and E primitive state vectors, together with the M primitive composed of physical actuator states, form a complete CEM triplet, which is then accessed through the standard eight-step closed-loop recursive operation link constructed by the self-generated system described in this invention.

[0052] Step S2: Closed-loop state acquisition and self-consistency deviation calculation. This step is the sensing entry point for a single iteration. Its core function is to acquire the current operating state of the system, quantify the self-consistency of the CEM closed-loop recursion, and provide basic data for subsequent sustainability assessment.

[0053] Execution timing: The first perception stage that begins in each iteration.

[0054] Core basis: Real-time feedback data from the system and real-time operating status of CEM primitives.

[0055] Operation content: (1) Collect real-time status data, external environment interaction data, and operation feedback signals of the three CEM primitives in the current iteration cycle through the system feedback interface to ensure full data coverage and no omissions; (2) Calculate the self-consistent deviation rate according to the CEM closed-loop recursive logic. (2) Quantify the degree of closed-loop coupling deviation of the three basic elements; (3) Derive the self-consistent matching degree of CEM recursive coupling from the self-consistent deviation rate. This reflects the degree of matching between the current operating state of the system and the core convergence target.

[0056] Output: Self-consistency deviation rate CEM recursive coupling self-consistent matching degree Real-time state datasets provide raw data for unary ontology persistence calculations and are the foundational sensing element for subsequent state assessment and security management.

[0057] All data collected in this step is stored in a circular buffer with a depth of no less than 100 iteration cycles to meet the needs of subsequent time series fitting and historical backtracking.

[0058] Step S3: Real-time assessment of the univariate ontology's sustainability. This step is the core decision-making step in a single iteration. Its core function is to quantify the ontology's sustainability status of the target system, intrinsically define the system's operational boundaries, and provide a unique intrinsic basis for security control and bimodal weight adjustment.

[0059] Execution timing: After status acquisition and before security control.

[0060] Core basis: Self-consistency deviation rate of S2 output CEM recursive coupling self-consistent matching degree Real-time status data.

[0061] Operation content: (1) Calculate the current period's univariate ontology sustainability index (2) Retrieve the univariate ontology persistence index Calculate the moving average using historical data from the past 50 periods. With sliding standard deviation Endogenous generation of dynamic high threshold Dynamic low threshold (3) Endogenously define the core survival red line and the peripheral capability boundary of the system, and delineate the ontological security zone; (4) Determine the current univariate ontological survival index. The location of the safe zone provides clear numerical data for subsequent safety management.

[0062] The full lifecycle runtime sequence of the univariate ontology sustainability index is shown in Figure 6. Its operation process is deeply bound to the four stages of the system's full lifecycle: initial anchoring, steady-state refinement, innovative breakthrough, and new steady-state anchoring. It endogenously defines the core sustainability boundary and innovative exploration space of the system, and achieves the synergistic unity of system steady-state maintenance and endogenous innovation.

[0063] Output result: Core survival red line, dynamic threshold , 1. Result of determining the safe zone of the main body.

[0064] Correlation between the preceding and following stages: Only output indicators related to the survival of the ontology, which is the sole basis for pre-control in S4 and an auxiliary basis for post-processing in S7, distinguishing the functional boundaries between single ontology indicators and globally coupled indicators.

[0065] One implementation method provides a quantitative expression for the univariate ontology persistence index: Where the hard constraint is ,and , The minimum guarantee weight for rigid operational consistency cannot be exceeded throughout the entire life cycle;

[0066] The , as the ontological persistence term, quantifies the degree of matching between the material structure M primitive and the safety benchmark configuration; the aforementioned The safety baseline configuration under the rigid operational identity constraint of the system is determined by the principal component center vector of the initial 1000-cycle steady-state data, and is adjusted only with the rigid constraint throughout the entire life cycle; For the tensor Frobenius norm;

[0067] The This refers to the environmental adaptability item, which quantifies the adaptability of the target system to the external environment; the aforementioned The ideal configuration is adaptively estimated based on environmental interaction data from the past 50 cycles and is updated in real time as the environment changes.

[0068] The The endogenous adaptive weights are used both to calculate the univariate ontology persistence index and to dynamically determine the weights of self-indicating and other-indicating modes in the self-generated system. The update formula is as follows: in This is the base weight for the ontology's persistence item, with a default value of 0.2. These are adaptive coefficients, with default values ​​of 0.5 and 0.6 respectively; The sliding standard deviation of the ontological duration term in the past T endogenous time quantum;

[0069] In this application, self-referential modes refer to the set of behaviors of a self-generated system directed inward to maintain its own operational identity. That is, through recursive generation between information field state C, energy mechanism E, and material structure M, it achieves self-preservation of ontological rules, self-maintenance of steady state, and self-repair of structure, and is the self-generated steady-state maintenance driving force of the system. Other-referential modes refer to the set of behaviors of a system directed outward to adapt to the external environment and explore new evolutionary paths. That is, through multi-dimensional and multi-level coherent transformations with systems at the same level and across levels outside the system and their basic elements, it achieves environmental perception, interaction adaptation, strategy generation, evolutionary exploration, and rule updating, and is the self-generated adaptive evolutionary driving force of the system.

[0070] Preferred self-referenced modal weights He refers to modal weights This enables the intrinsic binding of bimodal weights and sustainability metrics, eliminating the need for manually pre-setting weight allocation rules.

[0071] One implementation method is an endogenous generation rule for the core survival red line and dynamic threshold:

[0072] (1) Core survival red line formula: The core survival red line can only be adjusted upwards, not downwards. It is the ultimate safety bottom line that cannot be breached throughout the entire life cycle of the system. Any action that touches the red line will trigger the highest level of circuit breaker.

[0073] (2) Formula for peripheral capacity boundary: The peripheral capability boundary is the current safety redundancy of the system. The larger the value, the further the system is from the core red line, and the greater the space for innovation that can be explored.

[0074] (3) Dynamic threshold formula: in: for The moving average of the past 50 iterations. This represents the moving standard deviation for the corresponding period. The threshold coefficients are preferably initialized to 0.5 and 2.0, respectively. The endogenous fine-tuning rule is as follows: In the formula This represents the false alarm trigger rate over the past 100 periods. The steady-state recovery success rate over the past 100 cycles is represented by values ​​in the range [0,1].

[0075] Step S4: Gradient Security Management. This step serves as a safety fallback for each iteration. Its core function is to perform tiered security interventions based on the ontology's current state, ensuring ontology security throughout the entire system evolution process and addressing the passive limitation of existing technologies that prioritize "instability first, then protection."

[0076] Execution timing: Before rule generation and primitive update, when there are no system modifications.

[0077] Core basis: Only S3 output Core survival red line and dynamic threshold.

[0078] Operational procedures: Optimize the three-tiered control system. Level 1 Early Warning and Control: Slightly tighten the S5 candidate rule adoption threshold to limit the exploration range of other modalities, implementing mild preventative constraints without affecting the normal synchronous operation of the two modalities; Level 2 intervention and pre-control: The modification permissions for the core rule segments of the C primitive in S5 are locked, allowing only minor adjustments to non-core parameters. The steady-state maintenance weights of self-referential modes are strengthened to prevent further degradation of the ontology state. The permissions for updating mapping parameters are frozen. Three-level circuit breaker pre-control is implemented. The generation and updating of candidate rules for the S5 other-reference modal is directly suspended, and only the self-reference modal is retained to maintain the existing steady state, thus preventing dangerous rules from entering the subsequent update process from the source.

[0079] Output results: Dual-modal operation permissions, rule update constraints, and pre-control level instructions define insurmountable permission boundaries for the next step of rule generation and dual-modal weight adjustment. This is "preemptive risk prevention" without any actual repair or circuit breaker actions. It only performs permission control and forms a seamless connection with S7's post-event handling. The execution flow is shown in Figure 4.

[0080] Priority mapping rules: when but At that time, with As the core basis for control, a mild warning is triggered; when At that time, it directly triggers a Level 1 warning, unaffected by... Range constraints ensure timely handling of anomalies in closed-loop self-consistency.

[0081] Step S5: Bimodal Dynamic Weight Adjustment and Endogenous Pre-verification of Candidate Rules. This step is the core execution link for constructing the co-evolution of self-indicating and other-indicating modes in the self-generated system. Through the endogenous coupling of the three primitives CEM and the weight adjustment of self-indicating and other-indicating modes, the rule update of the C primitive, the synchronous adaptation of the E primitive, and the synchronous reshaping of the M primitive are realized. This achieves a synchronous balance between "steady-state maintenance and innovative evolution" in the self-generated system, which is the core inventive point that distinguishes this invention from the prior art.

[0082] Execution timing: After permission constraints are determined, but before primitive reshaping.

[0083] Core basis: S4 output pre-control permissions and rigid core layer constraints.

[0084] Operation content: (1) According to the pre-control level of S4, the dual-mode operation weight is endogenously adjusted; (2) The self-indicating mode continuously executes the closed-loop recursion of the existing C primitive rules, anchors the existing dynamic steady-state order of the system, and ensures the basic operation of the system; (3) The other-indicating mode generates candidate rules endogenously within the pre-control authority, and executes the rigid core layer similarity verification and the non-decreasing continuity verification in sequence, eliminates unqualified rules, and selects the only optimal compliant candidate rule; (4) Simultaneously receive the rule verification parameters of S7 reverse optimization, optimize the candidate rule generation logic this time, and realize the initial self-evolution.

[0085] Output results: Optimal compliance candidate rules and dual-modal operation weight parameters provide a legal basis for S6 primitive reshaping and tensor generation, inheriting S4 pre-control and connecting with S6 execution. The dual-modal operation logic is shown in Figure 3.

[0086] Specifically, the core function of the self-referential mode is to absolutely maintain the stability of the identity of rigid operations, while maintaining the existing steady-state order of the identity of dynamic operations, continuously executing the closed-loop recursion of the current C primitive rule, and anchoring the identity of the system ontology through the execution link of the target system; the core function of the other-referential mode is to perform endogenous optimization of the identity of dynamic operations without breaking the constraint of the identity of rigid operations, synchronously running the endogenously generated candidate rule attractor and verifying its survival potential.

[0087] One implementation method is a bimodal dynamic weighting adaptation rule:

[0088] (1) Self-referenced modal weights His modal weights Satisfy the global dynamic normalization requirements: The dual-modal operation runs synchronously in each iteration cycle under normal operating conditions, without mutual exclusion pause logic; the rule update function of the other modality is temporarily paused only when the third-level circuit breaker is triggered.

[0089] (2) The formula for endogenous adjustment of weights is: in The sigmoid activation function is used. The steepness coefficient is determined by the endogenous timescale of the target system; The minimum operating weight for the dual-mode operation is set to 0.05 to ensure synchronous operation of the dual-mode operation throughout its entire lifecycle.

[0090] (3) When hour, When the value approaches 0.95, the target system is maintained in a steady state dominated by the self-explaining mode, while the other-explaining modes operate synchronously in a low-weight exploratory state.

[0091] (4) When hour, Automatic improvement enhances the target system's ability to explore rules and restore order in other modalities.

[0092] Preferably, the rule update intensity is endogenously controlled: this refers to the rule update intensity of the modality. Using a continuously differentiable endogenous function, with its value strictly limited to the range of 0 to 0.2, the formula is as follows: During engineering implementation, upper and lower limits are applied to the input of the exponential term, with an upper limit of 0 and a lower limit of -20, to avoid the risk of numerical overflow under extreme conditions; the upper limit of 0.2 is determined through 10,000 Monte Carlo closed-loop evolution tests to ensure that single-cycle rule changes do not cause system oscillations.

[0093] The final execution formula for rule updates is: in For the current rule tensor, The optimal candidate rule attractor is selected; in steady state, only candidate attractors with higher sustainability than the current rule are adopted to ensure that the sustainability of the target system is non-decreasing.

[0094] One implementation method is to perform pre-implementation compliance verification for candidate rules: this is a predictive admission verification before the rules are implemented, specifically including the following steps:

[0095] S501 Rigid Red Line Verification: Verify that the segment corresponding to the rigid operation identity in the candidate rule has a cosine similarity ≥ 0.999 with the baseline rule of full lifecycle read-only locking, and remove invalid candidate rules that break the red line;

[0096] S502 Endogenous Threshold Generation: Self-referential mode weights based on the current period Endogenous generation rule adoption threshold ;

[0097] S503 Non-decreasing Survival Verification: Based on the rule change-survival change mapping model of the system's endogenous fitting, it calculates the predicted survival index for the next N periods corresponding to the candidate rule through a multi-step look-ahead prediction formula, only when... and When the candidate rules are retained, N is set to 5 by default, corresponding to the prediction window for the next 5 endogenous time quantum.

[0098] S504 Optimal Candidate Selection: For candidate rules that pass the verification, perform multi-objective non-dominated sorting. The optimization objective priority is: maximize the prediction sustainability index → ​​minimize the rule change magnitude → minimize the prediction self-consistency bias rate, and select the unique legal optimal candidate rule.

[0099] Step S6: Co-evolution and reshaping of the three CEM primitives and generation of endogenous spatiotemporal phase dimension. With the rigid core layer's operational uniformity as an absolute constraint, the C primitive rules are updated to compliance based on the optimal candidate rules selected in S5. Simultaneously, the energy-driven mechanism of the E primitive is adapted endogenously, and the structural state vector of the M primitive is reshaped, realizing the co-evolution of the three CEM primitives. Simultaneously, based on the coupling state of the three CEM primitives, the three dimensions of time, space, and phase are generated endogenously, constructing an endogenous spatiotemporal phase dimension deeply bound to the target system's operating state. Finally, based on the updated CEM primitives and the endogenous spatiotemporal phase dimension, a five-dimensional holographic coherent state tensor is generated. Calculate global coherence And the coherence of each dimension; all update operations are completed through the closed-loop execution link of the target system, and it is absolutely forbidden to modify the uniformity of rigid core layer operations.

[0100] Execution timing: After the compliance rules are determined, iterate on the core execution process.

[0101] Core basis: S5 output compliance candidate rules and rigid operational identity constraints.

[0102] Operation content: (1) Strictly follow the rigid operation identity read-only constraint, complete the C primitive rule update based on the compliant candidate rules, simultaneously adapt the E primitive and reshape the M primitive, realize the co-evolution of the three primitives of CEM, and never touch the forbidden zone of rigid operation identity; (2) Based on the updated CEM primitive, simultaneously update the time quantum, spatial grid point, and interphase coupling channel parameters of the endogenous spatiotemporal phase dimension, so that the endogenous spatiotemporal and the system state are matched in real time; (3) Generate the current period five-dimensional holographic coherent state tensor according to the five-dimensional tensor generation formula. (4) Substitute into the global coherence formula to calculate the global coherence of the current period. By combining the coherence of different dimensions, we can obtain the core indicators of the global coupling state of the system.

[0103] Output: Updated CEM primitives, new endogenous spatiotemporal phase dimension parameters, and five-dimensional holographic coherent state tensor. Global / dimension coherence provides a core basis for the next step of adaptive verification and hierarchical processing.

[0104] One implementation method features an endogenous spatiotemporal phase dimension endogenous rule: the endogenous time, space, and phase dimensions are entirely generated endogenously from the coupling state of the three basic CEM units, without any externally preset spatiotemporal container. The specific endogenous rule is as follows:

[0105] Endogenous time dimension: anchoring the energy mechanism E primitive, the time taken to complete a complete closed-loop recursive mapping with the three primitives of CEM is the minimum endogenous time quantum, the iteration period is an integer multiple of the endogenous time quantum, and the length of the time dimension is endogenously determined by the number of iterations in the entire life cycle of the system;

[0106] Endogenous spatial dimension: anchored material structure M primitive, corresponding to discrete grid points in the physical or regular space of M primitive, the grid point resolution is intrinsically determined by the state sampling accuracy of M primitive, and the boundary of the spatial dimension is intrinsically determined by the safe physical or regular boundary of M primitive;

[0107] Endogenous interphase dimension: Anchored information field state C primitive, corresponding to the coupling feature channel of the three primitives of CEM, is the coupling medium connecting the endogenous time dimension and the endogenous spatial dimension. The number of dimensions is endogenously determined by the sum of the feature dimensions of the three primitives of CEM.

[0108] The preferred method is the construction and computation of a five-dimensional holographic coherent state tensor space: based on the temporal, spatial, and phase dimensions endogenously constructed by the three primitives of CEM, a five-dimensional holographic coherent state tensor space adapted to the system's self-sustaining, self-repairing, and self-evolution requirements is constructed simultaneously, realizing the high-order non-additive coupling of the three primitives of CEM, and providing core support for the evaluation of the sustainability of a univariate ontology and anomaly localization.

[0109] The basic structure of the five-dimensional holographic coherent tensor space is as follows: Each dimension is intrinsically determined and dynamically adjusted by the self-generated system. The specific rules for intrinsic determination are as follows:

[0110] For spatial dimension group, corresponding to the three-dimensional spatial dimension built endogenously by the system, it is deeply related to the interphase dimension. The number of grid points is endogenously determined by the system based on the univariate ontology sustainability index and spatial resolution requirements. The grid point division strictly satisfies Shannon sampling theorem.

[0111] For time window dimension group, corresponding to the time dimension of the system's endogenous construction, the sequence length is based on the smallest endogenous time quantum of a complete closed loop recursion of CEM, and is dynamically determined by the system based on the iteration cycle and the univariate ontology persistence index.

[0112] For feature channel dimension groups, corresponding to the alternating dimensional features of the three primitives of CEM recursively coupled. The dimension size is the sum of the feature dimensions of the three primitives. The dimension arrangement corresponds one-to-one with the feature structure of the three primitives. There is no forced flattening requirement. It dynamically adapts with the endogenous evolution of the primitives.

[0113] In one implementation, the coupling generation formula for the five-dimensional holographic coherent state tensor is as follows: In the formula, : These are the five-dimensional tensors of the C, E, and M primitives after dimension alignment. They are generated endogenously in real time during the step (co-generation and reshaping of the three primitives CEM). Their dimensions are completely matched with the five-dimensional holographic coherent tensor space, and the original topological structure of the primitives is completely preserved.

[0114] Linear coupling weights: Default value The weights can be intrinsically and adaptively adjusted by the system based on the univariate ontology sustainability index—when the system's steady state is insufficient, the weights of the C primitives are increased. Strengthen the ability to maintain steady state; when the system is in a stable state, appropriately increase the weights of E and M primitives to enhance the ability to innovate and evolve.

[0115] Higher-order coupling weights: The self-generated system adaptively adjusts itself based on the univariate ontology's sustainability index, with the specific adjustment rules being as follows: (in , (This is an indicator of the current univariate ontology's sustainability; the more robust the system's steady state, the better.) Higher values ​​strengthen the higher-order non-additive coupling of the three primitives, enhancing the system's self-evolution capability; when the system is in a warning, intervention, or circuit breaker state, Automatically adjusted to near 0 to prioritize the steady-state convergence of linear coupling;

[0116] Ternary higher-order coupling terms: To adapt to different computing power scenarios in the self-generated system (such as industrial-grade high-computing-power devices and embedded low-computing-power edge nodes), two dynamically switchable implementation methods are supported. The switching logic is intrinsically determined by the system based on real-time computing power resources and the unary ontology's persistence requirements. Neither method changes the core definition and coupling logic of the five-dimensional tensor.

[0117] Method 1: High-precision full coupling, the formula is as follows The original topology of the CEM primitives is preserved without flattening. For tensor outer product operations, The result of the operation is a ninth-order tensor, which fully preserves all the coupling information of the three primitives in the spatial, feature and temporal dimensions, and is suitable for the high-order feature extraction requirements in the self-evolution process of the system. For tensor shrinking operations, the shrinking rules are intrinsically determined by the self-generated system. The ninth-order tensor obtained by the outer product is summed and normalized along the specified dimensions, and the final output dimension is the same as the tensor of the second order. A perfectly identical five-dimensional tensor;

[0118] In this invention, the preferred shrinkage rule based on the steady-state requirements of the self-generated system is: for the output tensor in (corresponding to respectively) The elements at that location satisfy all spatial pairs Feature alignment Time alignment Summing the indices and dividing by the normalization factor. (Alignment combination number), ensuring that the shrinking result accurately reflects the coupling strength of the three primitives in the corresponding dimensions, is defined point by point as follows: The summation covers all conditions that satisfy spatial coordinate alignment. Feature channel alignment Time step alignment index combinations, The normalization process is used to determine the total number of alignment combinations mentioned above, ensuring that the amplitude of the coupling result is within the steady-state range that the system can recognize, thus supporting subsequent coherence calculation and steady-state verification.

[0119] Method 2: Edge-end engineering adaptation coupling, the formula is as follows This is a simplified engineering approach for scenarios with limited computing power; This is a tensor vectorization operator, used solely for simplification in engineering operations. It flattens the five-dimensional tensors of the C, E, and M primitives into one-dimensional vectors, with vector dimensions of... It does not change the characteristic information of the primitive; The Kronecker product (vector outer product) is a product of three one-dimensional vectors to obtain a high-dimensional one-dimensional vector, which simplifies the computational complexity of high-order coupling and is adapted to the computing power of low-computing devices. The dimension reshaping operation reshapes the one-dimensional vector obtained from the Kronecker product into a new dimension. The five-dimensional tensor ensures that its dimensions are completely matched with the five-dimensional holographic coherent state tensor space, and can be directly used for subsequent coherence calculation and steady-state assessment.

[0120] As a preferred implementation method, the switching mechanism between Method 2 and Method 1 is as follows: The system intrinsically selects the coupling method based on real-time computing power resources, the unary ontology's sustainability index, and real-time requirements. When the unary ontology's sustainability index is ≥0.9 (sufficient in steady state) and the computing power is sufficient, it automatically switches to Method 1 to improve coupling accuracy. When the computing power is insufficient or the real-time requirements are high (such as during the anomaly handling phase), it automatically switches to Method 2 to ensure efficient system operation. The switching process does not affect the system's steady state.

[0121] As a preferred embodiment, the calculation of global coherence and multi-dimensional coherence is also included, and the formula for calculating global coherence is as follows: The dimensional coherence includes at least one of spatial coherence, feature coherence, and temporal coherence, and is used to locate the specific dimension in which the system anomaly occurs and trigger corresponding adaptive adjustment or self-repair actions. The current five-dimensional holographic coherent state tensor encompasses coupled information across all dimensions of space, features, and time. It is endogenously generated in this step and reflects the current self-evolutionary state of the system in real time. During the initialization of the system in step S1 (bootstrapping and isomorphic mapping), the baseline five-dimensional holographic coherent state tensor generated under zero input and steady-state conditions serves as the benchmark for steady-state evaluation of the system. It is not arbitrarily modified throughout the entire life cycle and can only be moderately calibrated through inverse self-optimization (step S7). The tensor Frobenius norm is defined as the square root of the sum of the squares of all elements of the tensor, i.e. It is used to quantize the overall magnitude of the tensor to ensure the accuracy of global coherence calculation;

[0122] Steady-state determination rule: when When the system is in a steady state, it is determined that the system as a whole is in a steady state; when When, a Level 1 warning is triggered; when At the same time, combined with the single entity's sustainability indicator, a second-level intervention or a third-level circuit breaker is triggered.

[0123] One possible preferred implementation involves calculating the coherence across dimensions. This involves real-time monitoring of the system status from three core dimensions—space, time, and interphase features—as well as a joint dimension of space and features. This allows for further pinpointing the specific dimension and location of anomalies, providing a precise basis for the system's next step of intrinsic self-repair. Its role in this method is as follows:

[0124] (1) Anomaly localization: When the global coherence is abnormal, the specific dimension in which the anomaly occurs is located by the abnormal distribution of the coherence in different dimensions (e.g., spatial coherence anomaly is a spatial layer problem, and feature coherence anomaly is a CEM primitive feature problem).

[0125] (2) Self-repair adaptation: Based on the dimension of the anomaly, the system generates targeted repair rules (such as optimizing the spatial grid distribution for spatial anomalies and correcting the corresponding primitive features for feature anomalies), thereby improving the accuracy and efficiency of self-repair.

[0126] (3) Reverse optimization: The time-series change data of the coherence of each dimension is synchronously entered into the system's reverse self-optimization model (step S7) to optimize the five-dimensional tensor coupling parameters and coherence threshold, thereby improving the system's ability to maintain steady state and predict anomalies.

[0127] (4) Aggregation application: The coherence of the subdimensions can be further aggregated into scalars (such as average value, maximum value) for comparison with preset thresholds, triggering adaptive adjustment or abnormal handling actions to adapt to the control requirements of different levels of the system.

[0128] Step S7: Closed-loop self-consistency verification, steady-state anchoring, and bidirectional self-optimization. Based on the co-evolved CEM primitive states, endogenous spatiotemporal phase space parameters, and global coherence. The system performs a four-layer progressive closed-loop self-consistency steady-state convergence test based on the coherence of each dimension, thus completing the endogenous steady-state anchoring. If all tests pass, the target system steady-state is deemed valid for this iteration, and the updated CEM primitive state and endogenous spatiotemporal phase space parameters are locked. Simultaneously, the pre-compliance verification model of step S5 is optimized in reverse. If any test fails, the corresponding level of anomaly handling mechanism is triggered, including endogenous self-healing closed loop, steady-state rollback, and graded safety intervention.

[0129] Execution timing: After primitive reshaping and tensor generation are completed, iterate through the closing stage.

[0130] Core basis: S6 output +S3 Update .

[0131] Operation details:

[0132] (1) Closed-loop self-consistency test. Rigid core final check, single-cycle convergence check, dynamic self-consistency check, and full-cycle steady-state check are performed sequentially. The threshold for each layer of checks is determined by… Adaptive adjustment The lower the threshold, the stricter the requirements, to prevent hidden instability.

[0133] Level 1: Rigid Core Layer Compliance Verification. Using the rigid, read-only, and consistent operation throughout its entire lifecycle, anchored in step S1, as the sole verification benchmark, verify whether the CEM primitive update in this iteration touches upon or modifies the rigid core layer rules. If the verification fails: all updates from this iteration are discarded, reverting to the safe, stable state before the update, while simultaneously triggering a level 3 circuit breaker and the generation of an immutable audit log. If the verification passes: proceed to the next verification stage.

[0134] Second level: Single-cycle convergence test. A spectral normalization test is performed on the weight matrices of the three closed-loop mappings of the CEM after this iteration to verify the Lipschitz constant of the composite closed-loop mapping. This ensures that the closed-loop mapping of this iteration meets the single-cycle global convergence requirement and avoids system iteration divergence; if the test fails: discard all the updated content of this iteration, revert to the safe steady state before the update, and trigger the second-level intervention mechanism; if the test passes: proceed to the next stage of testing.

[0135] Third level: Dynamic evolution layer self-consistency check. Based on the CEM primitive states after passing the check, the closed-loop self-consistency deviation rate of this iteration is recalculated. The test verifies the steady-state validity and self-consistency of the closed-loop mapping after the dynamic operation identity update; if the test fails, the endogenous self-repairing closed-loop process is triggered; if the test passes, the test proceeds to the next stage.

[0136] Level 4: Full-cycle steady-state convergence verification. Based on the self-consistency deviation rate calculated in this iteration, the moving average of the self-consistency deviation rate with a sliding window length of 10 iteration cycles is calculated to verify whether it meets the upper bound requirement of steady-state offset ≤0.05, ensuring the global asymptotic stability of the system in infinite recursive iterations and avoiding ontological drift and steady-state divergence in long-term operation; if the verification fails: the weight of the self-explicit mode is automatically increased to above 0.9 to strengthen steady-state maintenance until the moving average returns to the compliant range; if the verification passes: the system enters the endogenous steady-state anchoring stage.

[0137] By passing the above four layers of checks, the endogenous steady-state anchoring is completed. If all checks pass, the steady-state of the target system in this iteration is determined to be valid, and the updated CEM primitive state and endogenous spatiotemporal phase space parameters are officially locked. If any check fails, the endogenous self-healing closed loop or the fallback safety steady-state rollback mechanism is triggered to ensure the continued safety of the system.

[0138] (2) Abnormal handling mechanism. The following three-level abnormal handling mechanism is preferred.

[0139] Level 1 Early Warning Response: or The CEM coupling bias is gently corrected, and the rule threshold for the next round is slightly tightened without large-scale repair actions.

[0140] Level II intervention: or Initiate a special self-healing mechanism, generate repair rules for other-pointed modes, strengthen the weights of self-pointed modes, and specifically correct the system's incoherence problem;

[0141] Level 3 circuit breaker handling: or The circuit breaker will be triggered immediately, suspending all updates of the other-pointing mode, reverting to the steady-state state of the previous cycle, and prioritizing the survival of the original entity.

[0142] (3) Two-way self-optimization. All closed-loop self-consistency tests of the entire iteration process have passed, and endogenous steady-state anchoring has been completed. The C-element rule update data, unary ontology persistence index change, closed-loop self-consistency deviation rate time-series data, and global coherence convergence data of this iteration are entered into the system's online training set. The regression learning parameters of the rule change-persistence change mapping model in step S5 are updated, the model's feature fitting coefficients are corrected, and the model's prediction accuracy for persistence changes after rule execution is improved. Bayesian linear regression can be used to address overfitting issues with small samples and quantify prediction uncertainty. When the sample size is less than 100, Bayesian updating is used; when the sample size is greater than or equal to 100, traditional gradient descent updating is switched to balance prediction accuracy and computational efficiency. Simultaneously, based on the steady-state convergence effect of this iteration, the threshold coefficients of gradient-based safety control in step S4 and the basic value of rule adoption threshold in step S5 are adaptively fine-tuned to achieve endogenous adaptation between control logic and system operating state.

[0143] In this iteration, some checks failed. Based on the types and magnitudes of the deviations that failed the checks, as well as the system state change data, the pre-compliance verification rules of step S5 were automatically corrected. The rule admission thresholds for the corresponding risk scenarios were tightened, the risk feature weights of the rule change-persistence change mapping model were optimized, and the filtering capability for high-risk rules was strengthened to avoid the recurrence of similar prediction deviations and checks failing events. At the same time, all the data that failed the checks were stored in the system feature library for the pre-risk identification of subsequent candidate rules, thereby improving the system's ability to predict endogenous risks.

[0144] Output results: steady-state locking instruction, rollback execution instruction, reverse optimization parameter set, and iteration validity judgment result, completing the closed loop of this iteration. At the same time, it reversely empowers the preceding execution stage, forming a complete self-evolving closed loop of "prevention-execution-acceptance-iterative optimization" with the pre-control of step S4.

[0145] Specifically, based on the above-mentioned full-process test results, the endogenous system achieves refined adaptation for steady-state anchoring: by analyzing the temporal variation characteristics of the closed-loop self-consistency deviation rate, the fluctuation amplitude of the univariate ontology sustainability index, and the convergence level of the global coherence, the system endogenously identifies the current steady-state quality and ontology stability, and automatically pre-adjusts the initial weights of the bimodal modes and the upper limit of the rule update intensity for the next iteration. The greater the system steady-state deviation and the more obvious the sustainability fluctuation, the higher the initial weight of the self-exponential mode and the lower the upper limit of the rule update intensity, thus prioritizing the enhancement of the system's steady-state maintenance capability. The better the system's steady-state convergence effect and the more consistently stable the sustainability index at a high level, the more appropriately the initial weight of the other-exponential mode can be increased and the upper limit of the rule update intensity can be appropriately relaxed, thus preserving sufficient space for endogenous evolution and innovative exploration for the system.

[0146] Once all verification steps are passed and the endogenous steady-state anchoring is completed, the target system steady-state of this iteration is deemed valid. The updated CEM primitive state, endogenous spatiotemporal phase space parameters, five-dimensional holographic coherent state tensor, and global coherence benchmark are officially locked as the initial state for the next cycle of iteration. If any step fails verification and cannot be restored to a compliant steady state through endogenous self-healing closed loop, a fallback safe steady-state rollback is executed. The valid steady-state that passed the full process verification of the previous cycle is used as the initial state for the next iteration cycle, ensuring the continued safety of the system throughout the process.

[0147] In a preferred embodiment, the intrinsic self-healing closed-loop process mainly includes:

[0148] (1) Endogenous damage identification: Based on the temporal mutation characteristics of self-consistency deviation rate, the continuous downward trend of the univariate ontology survival index, and the compliance verification results of the rigid core layer, the type, location and degree of steady-state damage of the system are endogenously identified, without the need for manual preset of fault type and fault threshold.

[0149] (2) Endogenous generation of repair rules: Based on the damage identification results, under the absolute constraint of rigid operation identity, exclusive repair rule attractors are generated endogenously through other-find modalities, rather than calling manually preset fault handling procedures. The repair rules are strictly limited to the non-core rule segments corresponding to the damage and do not touch the core operating framework of the system.

[0150] (3) Closed-loop simulation and verification of repair effect: Based on the system's endogenous CEM closed-loop mapping model and historical steady-state benchmark data, multi-step positive closed-loop simulation is performed on the generated repair rules to simulate the full-cycle operation state of the system after the repair rules are executed. Only the effective candidate repair rules that can stably return to the compliance range after simulation are retained, such as the system self-consistency deviation rate and the univariate ontology sustainability index.

[0151] (4) Steady-state repair and secondary verification: Execute the optimal repair rules, simultaneously complete the repair and adaptation of the three basic CEM components, and re-execute the four-layer progressive closed-loop self-consistent verification of this step; if all verifications pass, the endogenous self-repair is completed and the repaired system steady state is locked; if the compliance requirements still cannot be met after repair, the safety steady-state rollback is executed as a fallback, and an unalterable repair audit log is generated at the same time.

[0152] As a preferred implementation, the reverse optimization process simultaneously performs innovation effectiveness assessment and consolidation. Specifically, the value of this iteration's rule update is quantified using an innovation effectiveness assessment formula: [Formula omitted for brevity]. in The default weights are 0.5, 0.3, and 0.2 respectively. This represents the difference between the mean of the survival over M consecutive periods after the implementation of the innovation rule and the value before the innovation. The sliding standard deviation of the sustainability indicator after the implementation of the innovative rules. To improve the success rate of adapting innovative rules across multiple scenarios; only when Furthermore, if the constraints are met for three consecutive verification cycles, the innovative rules will be incorporated into the system steady-state evaluation system to complete the endogenous anchoring of the new steady state, and the system's core capability benchmark will be updated.

[0153] Step S8: Iterative Cycle and Steady-State Relay. This step is a closed-loop iterative process throughout the entire lifecycle. Its core function is to define the boundaries and rules of the system's iterative cycle, solidify the effective steady state, connect the preceding and following iterative cycles, and realize the continuous endogenous evolution of the system throughout its entire lifecycle. It is a core closed-loop step that ensures the continuous operation of the self-generating system.

[0154] Execution timing: After S7 is completed and the current iteration is closed, before starting the next iteration, it is executed once per iteration cycle.

[0155] Core execution hardware: processor, timing control module, non-volatile memory, and real-time clock module of electronic devices.

[0156] Key criteria: S7 output steady-state locking instructions, iteration validity determination results, reverse optimization parameters, and audit logs.

[0157] Core operational content:

[0158] (1) Anchoring of the initial state of the next cycle: The updated CEM primitives, new endogenous spatiotemporal phase space parameters, and five-dimensional holographic coherent state tensors that have passed the S7 test and completed steady-state locking are then anchored. The global coherence Ξ(t) is formally solidified as the initial state of the next iteration cycle, covering the initial benchmark of the previous cycle, and realizing the relay transmission of steady state.

[0159] (2) Data hierarchical retention and cleanup: Implement hierarchical management of data throughout the entire cycle, permanently retain rigid core layer constraints, historical steady-state benchmark data, and tamper-proof audit logs, and store them in non-volatile read-only or read-write memory; cyclically retain historical running data and reverse optimization model parameters from the past 100 iteration cycles and store them in a circular buffer; temporarily clean up temporary computation data, invalid candidate rule data, and intermediate parameters that have not passed the verification in this iteration to release computational resources;

[0160] (3) Iteration cycle timing calibration: Based on the endogenous time quantum, the timing of the next iteration cycle is calibrated to ensure that the iteration cycle is completely matched with the CEM closed-loop recursion time and avoid timing drift; when the endogenous time quantum of the system changes, the iteration cycle is adjusted synchronously to ensure the consistency of the endogenous time dimension; based on the time window dimension of the five-dimensional holographic coherent state tensor, a continuous temporal coherence index is constructed to ensure the continuity of the system's temporal state before and after the iteration cycle adjustment and avoid the ontological state drift caused by the cycle adjustment.

[0161] (4) Iterative jump trigger: Automatically triggers the jump instruction, jumps to the S2 closed-loop state acquisition and self-consistent deviation calculation step, restarts the entire process of "state acquisition-evaluation-control-evolution-reshaping-verification", and forms an infinite loop of self-generated operation closed loop; under the three-level circuit breaker trigger state, the iterative jump can only point to the safe mode of the S1 bootstrap startup step, only performs the steady state recovery of the body, and does not perform any rule update and evolution.

[0162] The core constraints of this step are: only define the jump rules and initial state baseline for the iterative loop, do not repeat the specific functions of S2-S7, and ensure that all functional steps of the next iteration are executed independently by the corresponding steps after the jump, thus avoiding the uniqueness and rigor of the temporal logic and avoiding logical overlap and repeated execution; only solidify the valid steady state that has passed the full process verification of S7, and strictly prohibit the use of the state that has not passed the verification or has not been repaired as the initial state of the next cycle, and must revert to the valid safe steady state of the previous cycle to ensure the survival and security of the system itself; iterative jumps can only point to the S2 step, and it is strictly prohibited to skip the perception, evaluation, and control links of S2-S4 and directly enter the rule evolution and primitive update links, thus eliminating unconstrained rule modifications from the process perspective.

[0163] The output of this step includes: the valid initial state for the next iteration cycle, the iteration start command, the hierarchically retained runtime data, and the timing calibration parameters. These outputs ensure a seamless transition between the current and next iteration. All update operations are completed through the execution chain of the target system, dynamically updating the operational consistency of the dynamic evolution layer and prohibiting modification of the operational consistency of the rigid core layer. This guarantees the system's continuous self-sustaining, self-repairing, and self-evolving capabilities, forming a complete self-generating closed loop.

[0164] One alternative implementation is a multi-level master-slave nested collaboration and global convergence. This is used to achieve hierarchical collaborative self-generation across multiple systems, suitable for multi-level master-slave architecture scenarios such as smart factories, production line management, and large equipment clusters. It is a scenario-based extension of the core method without changing the core method's 8-step closed-loop operation process.

[0165] Core execution entities: electronic computing devices of the upper-layer master system and the lower-layer slave system, both of which run the core construction method of this invention.

[0166] Core operational content:

[0167] (1) Hierarchical interface definition: Three types of interaction interfaces are defined. The interface data format strictly matches the CEM primitive dimension to ensure the consistency of data interaction between the master and slave systems.

[0168] (2) Expose interfaces to the upper layer: The lower layer exposes its own unary ontology persistence index, C primitive core rules, and global coherence data to the upper layer main system from the system.

[0169] (3) Downward receiving interface: The upper-layer main system sends rigid collaborative constraints, hierarchical security boundaries, and global collaborative objectives to the lower-layer slave systems;

[0170] (4) Peer-to-peer interaction interface: The interface for state data interaction and collaborative rule synchronization between systems at the same level.

[0171] (5) Cross-level coherence control: The self-consistency compatibility of C primitive rules between upper and lower level systems is quantified by cross-level coherence index. The calculation formula is as follows: In the formula, The C primitive rule feature vector of the upper-level main system, The lower-level C primitive rule feature vector is the feature vector of the system. A higher value indicates stronger self-consistency between the upper and lower level rules.

[0172] Preferred, conflict-endogenous modification rule: when When a conflict occurs, a rule conflict correction process is automatically triggered. The correction process strictly follows the dual priority principle of "priority to the survival of the lower-level entity and coordination of the upper-level global goal."

[0173] The lower layer first checks whether the rule conflict breaks through the rigid core layer operation consistency. If it does, it directly rejects the upper layer rule and sends a conflict warning to the upper layer main system.

[0174] If the rigid constraints of the lower layer are not broken, the lower layer generates conflict correction rules from the system through the other-pointed modality, and adapts to the global collaborative goal of the upper layer without breaking the bottom line of its own existence.

[0175] Based on the conflict data fed back from the lower layer, the upper-level main system endogenously optimizes the global collaboration rules, reduces cross-level conflicts, and achieves endogenous self-consistency of the global rules.

[0176] Based on non-cooperative game theory, a global order endogenous convergence mechanism is constructed for a multi-level system. This mechanism eliminates the need for manually pre-setting global cooperation rules, enabling the self-generation of cooperation within the multi-level system. The core solution is as follows:

[0177] Preferably, the game payoff function is defined as follows: The achievement of the global collaborative goal of the upper-level main system and the non-decreasing persistence of the lower-level system's univariate ontology are defined as the two core payoff goals. In the formula, For the i-th lower-level player, the game payoff from the system is... ,and Ensuring the absolute priority of the survival of the underlying entity; As a univariate ontology persistence index for the lower-level system, The degree to which overall collaborative goals are achieved;

[0178] Preferably, the candidate collaborative rule game screening is as follows: the upper-level main system collects all candidate collaborative rules generated endogenously from the lower-level system, performs non-dominated sorting based on the game payoff function, and retains only the Pareto optimal rules that simultaneously satisfy "the non-decreasing existence of the lower-level system ontology" and "the positive gain of the upper-level global collaborative goal", and sends them down to the lower-level system for execution.

[0179] Preferred global convergence rigidity judgment rule: A multi-level nested system is considered globally convergent if and only if all of the following conditions are met simultaneously: the univariate ontology persistence index of the upper-level master system and all lower-level slave systems meets the steady-state requirement ( The mean cross-level coherence of all upper and lower layer pairings is ≥0.9; the above two conditions remain stable for 10 consecutive iterations without continuous fluctuations.

[0180] An optional implementation is a multi-node peer-to-peer collaborative architecture, used to achieve decentralized multi-node distributed collaborative self-generation. This architecture is suitable for peer-to-peer scenarios such as AGV clusters, distributed monitoring and control terminals, and decentralized production lines. It is a scenario-based extension of the core method without changing the 8-step closed-loop process of the core method. Multiple peer-to-peer devices running this method establish encrypted peer-to-peer connections through a trusted communication link, periodically synchronizing the CEM capability matrix and unary ontology sustainability index of each node, and performing cross-node CEM closed-loop recursive mapping and collaborative self-generation. The formula for calculating the coherence of peer-to-peer collaboration is: In the formula, N is the total number of nodes. , Let C be the feature vector of any two peer nodes. The global convergence criterion is: global convergence is determined if and only if the unary ontology persistence index of all nodes meets the steady-state requirement, and the peer-to-peer collaborative coherence is ≥0.9, and the above requirements are met for 10 consecutive iterations. For abnormal scenarios such as node failure and communication interruption in peer-to-peer clusters, a fully intrinsic cluster self-healing mechanism is constructed, requiring no manual intervention or central node scheduling. The core solution is as follows:

[0181] (1) Intrinsic identification of node faults: The nodes in the cluster are intrinsically identified through periodic heartbeat synchronization and survival index verification. The identification rule is: a node that has not fed back data for 3 consecutive synchronization cycles or whose univariate ontology survival index is lower than the core survival red line is judged as a fault node.

[0182] (2) Intrinsic Reconstruction of Cluster Topology: After identifying the faulty node, the remaining normal nodes in the cluster automatically reconstruct the collaborative topology, intrinsically adjust the collaborative rules and task allocation, and complete the collaborative tasks of the faulty node without breaking the bottom line of the survival of each node, so as to realize the intrinsic self-healing of the cluster function.

[0183] (3) Fault node re-entry mechanism: After the fault node recovers to normal, it synchronizes the current collaborative rules and global steady-state benchmark of the cluster through the peer interaction interface. After the cluster nodes collectively verify and pass the verification, it automatically re-enters the cluster without manual configuration, realizing the inherent elastic expansion of the cluster.

[0184] Secondly, this invention provides a univariate dual-state three-body self-generating system.

[0185] The univariate dual-state three-body self-generated system provided in this aspect fully runs the construction method described in the first aspect above, including a two-layer architecture of hardware carrier layer and functional implementation layer. All functional modules correspond one-to-one with the method steps in the first aspect above, ensuring the complete and unbiased implementation of the method.

[0186] Specifically, the hardware carrier layer. This layer provides the physical foundation and interaction platform for the system; all hardware components are industrial-grade compliant devices, including:

[0187] Processor unit: including but not limited to ARM Cortex series MCU / MPU, x86 industrial processor, FPGA programmable logic device, DSP digital signal processor, used to execute all computation, control, and logic processing steps of the method;

[0188] Storage units include: non-volatile read-only memory (ROM / NOR Flash) for storing audit logs that ensure the uniformity of rigid core layer operations and prevent tampering; non-volatile read-write memory (NAND Flash / EEPROM / Solid State Drive) for storing historical running data, model parameters, and program code; and volatile memory (DDR / SDRAM) for real-time computation and temporary data caching.

[0189] Sensing and interaction unit: includes sensor array, encoder, transmitter, industrial bus interface, digital or analog input interface, used to collect status data and environmental interaction data of the target system, corresponding to the status acquisition function of method step S2;

[0190] Execution drive unit: includes servo driver, motor, hydraulic valve, pneumatic component, relay, digital or analog output interface, used to execute control commands updated by CEM primitives to realize the physical action output of the system;

[0191] Communication unit: includes industrial Ethernet interface, CAN bus interface, 5G / 4G wireless communication interface, and fiber optic communication interface, used for data interaction between systems in multi-level collaborative and distributed collaborative scenarios.

[0192] Specifically, the functional implementation layer. The functional implementation layer adopts a modular, highly cohesive, and loosely coupled architecture. Each functional module corresponds one-to-one with a method step, and the data flow between modules perfectly matches the temporal logic of the methods, including:

[0193] Bootstrapping and isomorphic mapping module: corresponding to method step S1, used to complete system power-on initialization, rigid core layer constraint anchoring, CEM primitive endogenous initialization, dual mapping construction of non-electronic entity system, and initial five-dimensional tensor and benchmark coherence anchoring;

[0194] Data acquisition and preprocessing module: corresponding to step S2 of the method, used to complete the acquisition of full-dimensional state data, calculation of closed-loop self-consistency deviation rate, calculation of C-EM recursive coupling self-consistency matching degree, normalization of weight matrix spectrum, filtering and caching of acquired data;

[0195] The sustainability assessment and threshold generation module corresponds to step S3 of the method and is used to complete the calculation of the sustainability index of the unary ontology, the delineation of the core sustainability red line, the endogenous generation of dynamic high and low thresholds, the determination of the safe interval, and the sliding statistical analysis of time series data.

[0196] Gradient security control module: corresponding to step S4 of the method, used to complete hierarchical security pre-control, dual-modal operation permission definition, rule update constraint generation, and early warning and circuit breaker command output;

[0197] Bimodal Parallel Execution Module: Corresponding to method step S5, it is used to complete the endogenous adjustment of bimodal running weights, the steady-state maintenance of self-indicating modes, the endogenous generation of candidate rules for other-indicating modes, the four-layer pre-compliance verification, the calculation of rule update intensity, the online gradient descent update of mapping parameters, and the online update of the rule change-persistence mapping model;

[0198] Primitive Reshaping and Evolution Space Construction Block: Corresponding to method step S6, used to complete the co-occurrence update of the three primitives of CEM, the synchronous update of the endogenous spatiotemporal phase space, the generation of the five-dimensional holographic coherent state tensor, and the calculation of global coherence and multi-dimensional coherence;

[0199] Closed-loop verification and self-optimization module: corresponding to method step S7, used to complete four-layer progressive closed-loop self-consistent verification, endogenous steady-state anchoring, endogenous self-repairing closed-loop execution, reverse self-optimization of preceding modules, and generation of tamper-proof audit logs;

[0200] The loop iteration and collaborative interaction module corresponds to method step S8 and the multi-level / distributed collaborative implementation method. It is used to complete the interaction control of iteration cycle timing calibration, initial state anchoring of the next cycle, iteration jump triggering, multi-level master-slave nested collaboration, and distributed peer-to-peer collaboration.

[0201] The system's audit logs adopt a hash chain storage structure, with each log containing the hash value of the previous log, ensuring that the operational data is tamper-proof and traceable throughout its entire lifecycle, in accordance with the industrial control system information security standard GB / T22239-2019 "Information Security Technology Network Security Level Protection Basic Requirements"; the system's rigid core layer constraints adopt hardware-level read-only locking, which cannot be modified by software, ensuring that the system's underlying security boundary cannot be breached from the hardware level.

[0202] Thirdly, the present invention provides an electronic device.

[0203] This aspect provides an electronic device for realizing the construction of a self-generated system, including one or more processors and one or more memories; the one or more memories are coupled to the one or more processors via a bus, and the one or more memories are used to store computer program code, the computer program code including computer-executable instructions; when the computer-executable instructions are executed on the one or more processors, the electronic device causes the electronic device to perform all steps S1-S8 of the unary two-state three-body self-generated system construction method described in the first aspect above, as well as all operations of all preferred embodiments.

[0204] Specifically, the electronic devices include industrial motion controllers, PLC programmable logic controllers, embedded microcontroller systems, industrial computers, servo drives, intelligent robot controllers, distributed measurement and control terminals, edge computing gateways, intelligent production line master controllers, and all other industrial-grade electronic devices with computing, sensing, and execution capabilities.

[0205] In a preferred embodiment, the electronic device further includes: a hardware security encryption chip for hardware-level read-only locking of rigid core layer constraints, hash encryption and anti-tampering of audit logs; a real-time clock module for timing calibration and timestamp generation of iteration cycles; and an industrial-grade isolation interface for electrical isolation of sensors, actuators and communication buses, thereby improving the device's anti-interference capability and operational stability.

[0206] Fourthly, the present invention provides a computer-readable storage medium.

[0207] The non-volatile, non-transient computer-readable storage medium provided in this aspect includes computer-readable instructions that, when executed on a processor of an electronic device, cause the electronic device to perform all steps S1-S8 of the unary two-state three-body self-generated system construction method described in the first aspect, as well as all operations of all preferred embodiments.

[0208] Specifically, the specific implementation forms of the computer-readable storage medium include, but are not limited to: non-volatile and non-transient storage media such as Flash memory, EEPROM electrically erasable read-only memory, Mask ROM mask read-only memory, PROM programmable read-only memory, hard disk drive, solid disk drive, optical disk, and magneto-optical disk, but explicitly exclude transient transmission media such as carrier waves and signals.

[0209] Fifthly, the present invention provides a computer program product.

[0210] This aspect provides a computer program product including computer-readable instructions that, when executed on a processor of an electronic device, cause the electronic device to perform all steps S1-S8 of the unary two-state three-body self-generated system construction method described in the first aspect, as well as all operations of all preferred embodiments.

[0211] Specifically, the computer program product can be distributed via storage medium or provided via network download. Its operation depends on the electronic device described in the third aspect and the computer-readable storage medium described in the fourth aspect, and does not depart from the hardware carrier and core technical solution of the present invention.

[0212] This invention addresses four core fundamental deficiencies in existing technologies and proposes a complete technical solution. Compared with the closest existing technologies, it represents a generational technological breakthrough and offers significant industrial-grade benefits, as detailed below:

[0213] (1) This invention innovatively proposes a self-generated system construction method based on a univariate two-state three-body architecture. Based on the self-generated system CEM closed-loop recursive generation isomorphism, it innovatively proposes three ontological dimensions of endogenous construction time, space, and phase, and uses a five-dimensional holographic coherent state tensor space for engineering implementation. This realizes the complete endogenous generation of system evolution benchmark, operation rules, and evolution direction, and completely gets rid of the strong dependence of existing technologies on externally preset spatiotemporal containers, rule bases, and convergence targets. It realizes the core essence of self-generation and self-evolution as defined by the self-generated system theory.

[0214] (2) This invention innovatively proposes an innovative driving mode that synchronously operates in both self-referential and other-referential modes throughout the entire lifecycle. By endogenously adjusting the weights of the two modes through the univariate ontology sustainability index, it achieves true self-generation of innovative strategies and rules and endogenous self-evolution of the system ontology, completely breaking the technical bias of existing technologies that "either lock the evolution or become unstable and out of control." Test data shows that the load disturbance recovery time of the solution of this invention is reduced from 200ms in the existing technology to 28ms, and the dynamic response capability is improved by 614%; the failure rate of the system during continuous 720 hours of variable load operation is increased from 82% in the existing technology to 100%, and the reliability is significantly improved.

[0215] (3) Based on the univariate ontology sustainability index, this invention intrinsically defines the core sustainability red line and dynamic operation boundary, and constructs a full-link gradient security control mechanism of "pre-control - in-process constraint - post-acceptance - iterative optimization". This achieves the intrinsic integration of control logic and system architecture, and completely solves the problems of easy failure of external control and ontology collapse during evolution in existing technologies. At the same time, through spectrum normalization constraints, four-layer progressive closed-loop verification, and Lipschitz constant constraints, a strict convergence guarantee system is established, realizing the global asymptotic stability of the time-varying system with endogenous rule evolution. Data shows that the security rule breakthrough rate of the solution of this invention is 0, and the ontology drift rate is ≤0.1% after 180 consecutive days of operation, which is far better than the ontology drift rate of more than 8% in existing technologies, and fully meets the stringent requirements of stable operation throughout the entire life cycle in industrial scenarios.

[0216] (4) Online incremental updates of CEM mapping parameters were implemented, and strict execution boundary constraints were established, with updates performed only in the steady-state phase and complete freezing during the fault phase. This not only enabled the system to adapt to time-varying operating conditions and hardware aging, but also eliminated the risk of system instability caused by parameter oscillations at the source. Under the condition of 10% hardware aging, the steady-state accuracy of this solution can still be maintained above 98%, while the steady-state accuracy of existing technologies drops to below 65%, achieving a qualitative breakthrough in operating condition adaptability.

[0217] (5) Realize the engineering implementation of self-generated systems in all scenarios. This invention breaks through the carrier limitation by using a dual mapping construction mechanism of "electronic equivalent port mapping + computational tensor isomorphic mapping". It realizes a set of methods that can be adapted to all target systems in all scenarios, such as electronic computing systems, industrial hydraulic systems, embedded low computing power systems, and non-electronic physical systems, and completely solves the core defect of existing technologies that can only be adapted to specific scenarios and cannot be universal. Attached Figure Description

[0218] Figure 1 is a flowchart of the entire process steps of the method for constructing a univariate two-state three-body self-generated system in the embodiment;

[0219] Figure 2 is a block diagram illustrating the principle of closed-loop recursive mapping of the three basic CEM elements in the embodiment;

[0220] Figure 3 is a logic block diagram of the co-evolution and weight adjustment of self-pointing mode and other-pointing mode in the embodiment;

[0221] Figure 4 is a flowchart illustrating the execution of the gradient-based security control in this embodiment;

[0222] Figure 5 is a block diagram illustrating the principle of the multi-level master-slave nested collaborative mechanism in this embodiment;

[0223] Figure 6 is a runtime sequence diagram of the entire lifecycle of the unary ontology persistence index in the embodiment. Detailed Implementation

[0224] The present invention will be further described in detail below with reference to specific embodiments. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of protection of the present invention. All reproducible implementations, conventional variations, and scenario adaptations completed by those skilled in the art based on the core technical solutions disclosed in this invention without creative effort are within the scope of protection of this invention. The following three differentiated embodiments verify the engineering implementation effects of three typical scenarios: single-node electronic industrial systems, non-electronic physical systems, and multi-node distributed collaborative systems.

[0225] Example 1: Industrial Single-Axis Servo Motor System

[0226] The execution unit in this embodiment is an industrial-grade motion controller (ARM Cortex-A9 quad-core 1GHz, 2GB DDR3, 8GB eMMC), equipped with a 0.75kW permanent magnet synchronous servo motor, a 17-bit absolute encoder, and current / temperature / vibration sensors. The control cycle is 1ms (intrinsic time quantum). It complies with the standard GB / T 16439-2016 "General Technical Conditions for AC Servo Systems" and is suitable for high-precision industrial servo control scenarios such as CNC machine tools and electronic manufacturing equipment.

[0227] First, define a fixed point for the identity of the two-level operation. The unique core convergence objective throughout the entire lifecycle is:

[0228] Rigid operation identity Rated speed 3000rpm, rated torque 2.3N・m, maximum position error ±0.05mm, maximum winding temperature 130℃, maximum current 1.2 times the rated value, the mathematical form and dimensional closure constraint of the CEM closed-loop recursive mapping are written to the controller FPGA-level read-only memory partition, and any evolutionary modification is prohibited throughout the entire life cycle.

[0229] Dynamic operation identity The position loop adaptive parameters, velocity loop feedforward rules, load disturbance compensation rules, and acceleration / deceleration curve optimization rules correspond to the evolvable rule segments of the information field state C primitive, which are updated endogenously by the system.

[0230] CEM primitive dimension closure constraint: Strictly follows the closed-loop recursive mapping rule to ensure complete dimension closure. In this embodiment, the dimension value is: rule feature dimension. Energy mapping interface dimension Degrees of freedom of material structure ,Right now: Closed-loop dimension verification: The link is completely closed, with no dimensional gaps, providing the only mathematical premise for calculating the self-consistent deviation rate and the endogenous generation of the five-dimensional tensor.

[0231] S1 Bootstrapping and Isomorphic Mapping. This step is the sole entry point for the self-generation of the self-generated system. It strictly adheres to the core rule of "only locking the underlying security constraints, without pre-setting the final convergence target or fixed operating order," completing the full endogenous self-generation of the system ontology. Specifically, it executes as follows:

[0232] Hardware self-test and initial data acquisition: At 1ms intervals, continuously acquire 1000 cycles of 6-dimensional core features including motor position, speed, current, temperature, encoder status, and bus voltage to construct... Initial feature matrix;

[0233] Rigid core layer anchoring: ensuring rigid operational uniformity Write to the controller's read-only storage partition, and lock it to read-only status for the entire lifecycle;

[0234] Intrinsic initialization of baseline parameters:

[0235] Safety benchmark Principal component center vectors of the initial 1000 periods of steady-state data, 6-dimensional unit normalized vectors, adjusted only with rigid constraints throughout the entire lifecycle;

[0236] Environmental benchmarks : The ideal state vector under initial no-load conditions, and The initial values ​​are consistent, and subsequent values ​​are updated in real time based on environmental interaction data.

[0237] CEM primitive endogenous generation: Based on principal component analysis (cumulative variance contribution rate of 99.2%), the initial primitives and mapping parameters are generated endogenously.

[0238] Initial C primitive rule set (6th order identity matrix);

[0239] Initial E-elemental energy driving parameters ;

[0240] Initial M-based primitive structure state vector ;

[0241] Initialize the mapping weight matrix Bias parameters All of these are optimizable parameters of the dynamic evolution layer;

[0242] Endogenous spatiotemporal phase dimension is self-generated: entirely generated endogenously from the coupling state of CEM primitives, without any external pre-defined spatiotemporal container.

[0243] Endogenous time dimension: anchoring the energy mechanism E primitive, the minimum endogenous time quantum is the time taken by the three primitives of CEM to complete a complete closed-loop recursive mapping of 1ms, and the iteration period is 1 times the endogenous time quantum;

[0244] Endogenous spatial dimension: anchored material structure M primitive, corresponding to discrete grid points of 6-dimensional states such as motor position and speed. The grid point resolution is intrinsically determined by the encoder sampling accuracy, and the spatial boundary is intrinsically determined by the safe physical boundary of the motor's rated stroke.

[0245] Endogenous interphase dimension: Anchored information field state C primitive, corresponding to the coupled feature channels of the three primitives of CEM, the number of dimensions is It is endogenously determined by the sum of the three basic characteristic dimensions;

[0246] Endogenous Construction of Five-Dimensional Holographic Coherent Tensor Space: Tensor Structure is Endogenously Determined by Endogenous Spatiotemporal Phase Dimension. Among them: spatial dimension group (M primitive 6-dimensional endogenous derivation), feature channel dimension group (Corresponding to endogenous alternating dimensions), time window dimension group (Corresponding to 4 endogenous time quanta);

[0247] Reference tensor and coherence anchoring: Based on the coupling generation formula of claim 18, higher-order coupling weights are taken. Linearly coupled weights The initial reference tensor is generated using a high-precision full-coupling method. : Among them, the ternary higher-order coupling term Initial global coherence Dimensional coherence It is solidified into a unique steady-state reference throughout the entire life cycle.

[0248] This step outputs: rigid core layer constraints, initial CEM primitives, initial mapping weights and bias parameters, endogenous spatiotemporal phase dimension, and initial five-dimensional tensor. Initial global coherence All data is stored on a non-volatile computer-readable storage medium, completing the self-generation of the self-generated system.

[0249] S2 Closed-loop state acquisition and self-consistent deviation calculation. This step is the sensing entry point for a single iteration, and its core execution is as follows:

[0250] Full-dimensional status acquisition: Within a 1ms cycle, the real-time status of the three CEM primitives, external load feedback, and ambient temperature data of the current iteration cycle are collected through the feedback interface and stored in a circular buffer with a depth of not less than 100 cycles.

[0251] Weight matrix spectral normalization: Initial weight matrix for this period , , ,calculate The maximum singular value is 1.004, and the execution spectrum is normalized. Ensure that the maximum singular value of all weight matrices is less than 1, and the Lipschitz constant of the composite closed-loop mapping. This satisfies the requirement of global asymptotic stability.

[0252] Self-consistency deviation rate calculation: Take the smallest positive number after removing zero. Substitute into the formula: The preset compliance threshold is 0.05. The closed-loop self-consistency is deemed satisfactory; isomorphic constraint loss function. This provides an objective function for subsequent parameter optimization. The determination of the self-consistent deviation rate compliance threshold of 0.05 is based on the following: Monte Carlo simulation testing of 10,000 industrial servo systems in closed loop operation showed that when δ(t) ≤ 0.05, the system steady-state control error ≤ ±0.02mm, meeting industrial control accuracy requirements; when δ(t) > 0.05, the control error deviation rate increased to 87%, therefore, 0.05 was determined as the compliance threshold. C-EM recursive coupling self-consistent matching degree calculation: The degree of matching between the quantified dynamic evolution layer and the operational identity fixed point.

[0253] Output of this step: Self-consistency deviation rate C-EM recursive coupling self-consistent matching degree Real-time CEM status dataset and spectral normalized weight matrix.

[0254] S3 Real-time assessment of the sustainability of a single ontology. This step is the core decision-making process for a single iteration, and its core execution includes:

[0255] Computational ontology persistence term: quantifies the degree of matching between the material structure M primitive and the safety baseline configuration.

[0256] Environment adaptability: Quantify the adaptability of the target system to the external environment.

[0257] Endogenous adaptive weight calculation: taking the basic weights of ontology persistence terms Adaptive coefficient , Minimum guarantee weight for core identity The moving standard deviation of the ontology term over the past 5 periods Substitute into the formula:

[0258] satisfy The constraint cannot be broken throughout its entire lifecycle. Minimum guarantee weight for core identity. The basis for this determination is that, through 720 hours of continuous variable load operation testing, when... When the system drift rate is ≤0.1%, the system drift rate is ≤0.1%. At that time, the body drift rate increased exponentially, reaching a maximum of 12.7%, so 0.2 was determined as the minimum weight that cannot be broken throughout the entire life cycle.

[0259] Simultaneously complete the intrinsic binding calculation of bimodal weights. , .

[0260] Calculation of the sustainability index of a single ontology:

[0261] Endogenous generation of core survival red lines and dynamic thresholds: The default value for the univariate ontology persistence index in the initial perfect steady state of the system is: Substituting, we get:

[0262] Dynamic threshold: The moving average of the univariate ontology sustainability index over the past 50 periods. Sliding standard deviation The initial values ​​of the threshold coefficients are respectively Substitute into the formula:

[0263] Safe zone determination: Current ,satisfy Therefore, the system is determined to be in the first-level warning zone.

[0264] Output of this step: Core survival red line Dynamic threshold / , safe interval determination results, endogenous adaptive weights and bimodal weights.

[0265] S4 graded security control. This step is a pre-emptive control phase, strictly based on the S3 judgment results to implement tiered intervention. No system modifications are made; only permission boundaries are defined. The core execution is: Current... This triggers a Level 1 early warning and control mechanism: the rule adoption threshold of S5 is raised to 0.95, only candidate attractors with significantly higher persistence than the current rules are adopted, steady-state constraints are strengthened, dual-mode normal synchronous operation is maintained, and no parameter update permissions are frozen.

[0266] Load change adaptation: When the load changes by a step of +50%, When the value drops to 0.88, a level 2 intervention is triggered, locking the modification permission for the C primitive core rule segment, allowing only minor adjustments to non-core parameters, and freezing the permission to update mapping parameters;

[0267] Extreme operating condition adaptation: When the winding temperature exceeds 130℃, If the circuit breaker is triggered, the third-level circuit breaker will be immediately cut off from the update channel of the other modal rules, switch to the safe steady-state mode dominated by the self modal, and generate an unalterable audit log.

[0268] This step outputs: dual-modal operation permissions, rule update constraints, and pre-control level instructions.

[0269] S5 Bimodal Dynamic Weight Adjustment and Endogenous Generation and Verification of Candidate Rules. This step is the core of the system's endogenous evolution and strictly follows the permission constraints of S4. Core execution:

[0270] Bimodal dynamic weight adjustment: Take the minimum operating weight in the bimodal mode. Steepness coefficient (Determined by a 1ms endogenous timescale), sigmoid activation function Substitute into the formula: satisfy It operates synchronously in both modes throughout the entire lifecycle, without mutual exclusion or pause logic;

[0271] Self-referential mode (weight 0.92): Continuously executes the closed-loop recursion of the current C primitive rule, anchors the existing dynamic steady-state order of the system, and ensures the stable operation of the 50.00mm positioning closed loop;

[0272] Pre-compliance verification of candidate rules: This refers to the modal endogenous generation of 8 sets of candidate rules, the strict execution of four layers of pre-verification according to priority, and the retention of only the best compliant rule.

[0273] S501 Rigid Red Line Verification: Verify the cosine similarity between the rigid operation identity segment of the candidate rule and the baseline rule, eliminate 2 invalid rules with similarity < 0.999, leaving 6 groups;

[0274] S502 Endogenous Threshold Generation: ;

[0275] S503 Non-decreasing Duration Verification: Based on the rule change-duration change mapping model, predict the duration of the forecast for the next 5 periods and eliminate 3 groups. According to the rules, there are 3 groups remaining;

[0276] S504 Optimal Candidate Selection: Perform multi-objective non-dominated sorting on the rules that pass the verification, with the optimization priority being "maximize prediction persistence → minimize rule change magnitude → minimize prediction self-consistency bias rate", and select the unique optimal compliance candidate rule with velocity feedforward compensation.

[0277] Rule update strength calculation: Substitute into the formula and input the limit to the exponent term. To avoid numerical overflow: Substitute values ​​and calculate step by step: Exponent term: Exponential function value: Take the minimum value: Final result: Validate constraints: It meets the safety upper limit requirements.

[0278] This step updates the rules using gradient descent with mapped parameters. The outputs of this step are: the optimal compliance candidate rule, the bimodal operating weights, and the rule update strength. The updated mapping weights and bias parameters complete the endogenous self-evolution of system rules.

[0279] S6 CEM: Symbiotic Reshaping of the Three Primitive Components and Generation of Endogenous Spatiotemporal Phase Dimensions. This step is the execution phase of the system's symbiotic evolution, with the rigid core layer as an absolute constraint. Core execution:

[0280] CEM Three-Element Symbiotic Reshaping: Based on the optimal candidate rules selected by S5, the C element is updated in compliance with the formula, the E element is adapted simultaneously, and the M element is reshaped, without touching the rigid core layer throughout the process.

[0281] Endogenous spatiotemporal phase dimension synchronous update: Based on the updated CEM primitive coupling state, the endogenous time window dimension, spatial grid resolution, and number of interphase coupling channels are synchronously adjusted to achieve deep binding between endogenous spatiotemporal and system operating state;

[0282] Five-dimensional holographic coherent state tensor generation (corresponding to claim 18): taking , A high-precision full coupling method is used to generate the five-dimensional tensor of the current period. Dimensions and initial baseline completely consistent;

[0283] Coherence calculation:

[0284] Global coherence: It meets the steady-state requirement of ≥0.9;

[0285] Dimensional coherence: Spatial coherence Feature coherence Temporal coherence This provides a precise basis for subsequent anomaly location and adaptive adjustment.

[0286] This step outputs: updated CEM primitives, new endogenous spatiotemporal phase dimension parameters, and a five-dimensional holographic coherent state tensor. Global / sub-dimensional coherence.

[0287] S7 Closed-loop self-consistency check, steady-state anchoring, and bidirectional self-optimization. This step is the iterative closure and self-optimization stage, and its core execution is as follows:

[0288] A four-layer progressive closed-loop self-consistent check is implemented strictly according to priority, and failure of any one item triggers the corresponding action:

[0289] S701 Rigid Core Layer Final Inspection: Verify that the cosine similarity between the evolved C primitive rigid segment and the read-only baseline rule is 0.9999 ≥ 0.9999, then pass; otherwise, directly revert to the previous cycle's safe steady state and trigger a level 3 circuit breaker.

[0290] S702 Single-cycle convergence verification: Verifying the Lipschitz constant of the updated CEM composite closed-loop mapping. If it passes, it will revert to the safe steady state of the previous cycle, triggering a level-two intervention.

[0291] S703 Dynamic Layer Self-Consistency Verification: Verifying the Actual Self-Consistency Deviation Rate After Evolution And actual sustainability indicators The previous cycle value was 0.983, which passed; failure would trigger an endogenous self-repairing closed loop.

[0292] S704 Full-cycle steady-state convergence verification: Verify that the moving average of the self-consistency deviation rate over 10 cycles is 0.022≤0.05, the moving standard deviation of the persistence index is 0.002≤0.01, and the global coherence is ≥0.9. If the verification fails, increase the weight of the self-explained mode and freeze the update amplitude of the other-explained mode.

[0293] Engineering applications of multidimensional coherence:

[0294] Precise anomaly localization and targeted self-repair: When the encoder is subjected to electromagnetic interference, The coherence dropped to 0.82, below the warning threshold of 0.9. Coherence in other dimensions remained stable. The intrinsic precise positioning anomaly occurred in the spatial dimension, triggering a level one warning. A sliding filter window was added to the encoder sampling data. After two cycles... The value was restored to 0.978, with no positional deviations throughout the entire process; when the current loop parameters drifted, When the value drops to 0.83 < 0.85, the warning threshold is reached. The endogenous localization anomaly occurs in the interphase dimension, triggering an endogenous self-repairing closed loop, and the steady state is restored within 3 cycles.

[0295] Steady-state adaptive optimization: The feature coherence over 20 consecutive periods is a weakness in terms of dimensionality. The number of feature channels in the interphase dimension is endogenously adjusted from 18 to 20. The mapping weights are optimized with the goal of maximizing feature coherence. After optimization, the feature coherence is improved to 0.982, and the position following error is reduced by 21.9%.

[0296] Closed-loop verification supplementary judgment: The coherence of each dimension is included in the full-cycle steady-state verification, requiring the lowest value of a single item to be ≥0.85 and without a continuous downward trend, filling the hidden instability loopholes that cannot be covered by the global indicators;

[0297] Reverse self-optimization data support: Incorporating multi-dimensional coherence time series data into the S5 pre-validation model, adding multi-dimensional coherence prediction constraints for candidate rules, and increasing the invalid rule filtering rate from 82% to 99.2%;

[0298] Complete execution of the intrinsic self-healing closed loop: For abnormal current loop deviations caused by winding temperature rise, a complete four-step self-healing process is executed:

[0299] Step 1 Endogenous damage identification: Based on the sudden change in self-consistency deviation rate and the decreasing trend of feature coherence, the endogenous damage type is identified as current loop parameter drift, and the damage degree is mild, without the need for manual preset fault database;

[0300] Step 2: Internal generation of repair rules: This refers to the internal generation of exclusive repair rules for modalities, which only optimize the corresponding mapping parameters and do not touch the core rules;

[0301] Step 3: Closed-loop deduction of repair effect: Forward deduction verifies that the repair rules can bring the system back to steady state, and only the effective repair rules are retained;

[0302] Step 4 Steady-state repair and secondary verification: Execute the repair rules, restore steady state within 3 cycles, pass all secondary verifications, and generate an immutable audit log;

[0303] Two-way self-optimization and innovation effectiveness evaluation:

[0304] After all validations pass, update the S5 rule change-persistence change mapping model weights to improve prediction accuracy;

[0305] Innovation effectiveness assessment: weighting , , Substitute into the formula: Calculated After three consecutive cycles of successful verification, the innovative achievements will be incorporated into the system steady-state evaluation system.

[0306] If the validation fails, the rule adoption threshold of S5 will be automatically tightened, the mapping model will be corrected, and the same type of deviation will be avoided from recurring.

[0307] This step outputs: steady-state locking instruction, rollback execution instruction, reverse optimization parameter set, iteration validity judgment result, and tamper-proof audit log, completing the system's self-repair, self-adaptation, and self-optimization.

[0308] S8 Iterative Loop and Steady-State Relay. This step is a closed-loop iterative process, with the core execution being:

[0309] Anchoring the initial state for the next cycle: The CEM primitives that have passed the S7 check and completed steady-state locking, the endogenous spatiotemporal phase dimension parameters, the five-dimensional tensor, and the global coherence are solidified as the initial state for the next iteration cycle;

[0310] Data tiered retention: Permanently retain rigid core layer constraints and audit logs; cyclically retain runtime data from the past 100 cycles; clear temporary computation data;

[0311] Iterative cycle timing calibration: Based on 1ms endogenous time quantum, calibrate the timing of the next cycle to avoid timing drift;

[0312] Iteration jump trigger: After 1ms, it automatically jumps back to S2 and starts the next round of iteration, forming a continuous self-generating closed loop.

[0313] A self-generated industrial single-axis servo motor system with a univariate, two-state, three-body architecture is constructed simultaneously. A two-layer architecture of hardware carrier layer plus function implementation layer is adopted, and all functional modules correspond one-to-one with the steps of the above S1-S8 method to ensure the completeness and unbiased implementation of the method.

[0314] The hardware layer provides the physical foundation and interaction platform for the system. All hardware components are industrial-grade compliant devices, and the specific configuration is as follows:

[0315] Processor unit: It adopts an industrial-grade motion controller and is equipped with an ARM Cortex-A9 quad-core 1GHz processor + FPGA coprocessor to execute all the calculation, control and logic processing steps of the method. The FPGA is used to implement hardware-level locking of the rigid core layer.

[0316] Storage units: Non-volatile read-only memory (FPGA built-in ROM) is used to store rigid operational identity constraints and immutable audit log root hash; Non-volatile read-write memory (8GB eMMC) is used to store historical running data, model parameters, and program code; Volatile memory (2GB DDR3) is used for real-time computing and temporary data caching.

[0317] Sensing and Interaction Unit: Encoder interface supports 17-bit absolute encoders with a 1MHz sampling frequency; Analog Input Interface: 4 channels of 16-bit AI for acquiring current, temperature, and vibration signals; Digital Input Interface: 16 channels of DI for acquiring limit switch and emergency stop signals; Industrial Ethernet Interface: EtherCAT for communication with host computer and other devices.

[0318] Execution drive unit: 0.75kW servo driver, supporting full closed-loop control of current loop, speed loop, and position loop; Digital output interface: 16 DO channels for controlling relays and audible and visual alarms.

[0319] Communication unit: Integrated EtherCAT industrial Ethernet interface, supports Modbus TCP and CANopen protocols, and can be expanded with a 5G wireless communication interface for remote monitoring.

[0320] The functional implementation layer adopts a modular, highly cohesive, and loosely coupled architecture design. Each functional module corresponds one-to-one with the method steps in Example 1, and the data flow between modules is completely matched with the temporal logic of the method.

[0321] The self-bootstrapping and isomorphic mapping module completes system power-on initialization, rigid core layer constraint anchoring, CEM primitive intrinsic initialization, and initial five-dimensional tensor and baseline coherence anchoring. After this module is completed, the system automatically enters a steady-state operation without manual intervention.

[0322] Data Acquisition and Preprocessing Module. This module completes full-dimensional state data acquisition, closed-loop self-consistent deviation rate calculation, CEM recursive coupling self-consistent matching degree calculation, weight matrix spectrum normalization processing, and sliding window filtering and circular buffer caching of the acquired data. This module runs on a 1ms cycle and serves as the system's sensing entry point.

[0323] The sustainability assessment and threshold generation module completes the calculation of the univariate ontology sustainability index, the delineation of the core sustainability red line, the endogenous generation of dynamic high and low thresholds, the determination of the safe interval, and the sliding statistical analysis of time-series data. The sustainability index output by this module is the sole basis for bimodal weight adjustment and rule selection.

[0324] The tiered security management module performs tiered security pre-control, defines dual-modal operation permissions, generates rule update constraints, and outputs early warning and circuit breaker commands. This module strictly adheres to rigid operational consistency constraints, ensuring that the safety red line will never be breached under any circumstances.

[0325] The bimodal parallel execution module performs the following: endogenous adjustment of bimodal running weights, steady-state maintenance of self-exponential modes, endogenous generation of candidate rules for other-exponential modes, four-layer pre-compliance verification, rule update strength calculation, online gradient descent update of mapping parameters, and online update of the rule change-persistence mapping model. This module is the core of the system's endogenous evolution.

[0326] The module for reshaping and constructing the evolution space of the CEM (Cybernetic Evolutionary Model) completes the symbiotic update of the three major CEM primitives, the synchronous update of the endogenous spatiotemporal phase space, the generation of the five-dimensional holographic coherent state tensor, and the calculation of global coherence and multi-dimensional coherence. This module ensures that the CEM closed loop remains self-consistent throughout the system's evolution.

[0327] Closed-loop verification and self-optimization module. This module completes four-layer progressive closed-loop self-consistent verification, endogenous steady-state anchoring, endogenous self-healing closed-loop execution, reverse self-optimization of preceding modules, and generation of tamper-proof audit logs. This module is the closing stage of the iteration process, ensuring that system evolution always moves towards improving survivability.

[0328] The iterative and collaborative interaction module completes iteration cycle timing calibration, initial state anchoring for the next cycle, iteration jump triggering, and multi-axis collaborative and distributed collaborative management. This module is responsible for maintaining the continuous closed-loop operation of the system.

[0329] Immutable audit logs: Employing a hash chain storage structure, a log block is generated every 100 control cycles, and each log block contains the hash value of the previous log block. All logs are stored in a read-only partition of the eMMC, making them impossible to modify or delete via software, and complying with the industrial control system information security standard GB / T 22239-2019.

[0330] Hardware-level rigid constraint locking: All rigid operation consistency constraints (rated speed 3000rpm, maximum winding temperature 130℃, etc.) are written to the FPGA's hardware-level read-only memory partition, which cannot be modified by any software method, ensuring that the system's underlying security boundary cannot be broken from the hardware level.

[0331] The univariate two-state three-body self-generated system constructed based on the above method was continuously operated for 720 hours on an electronic manufacturing equipment production line, which fully verified the effectiveness of the univariate two-state three-body self-generated system construction method and system of the present invention. The core technical effects are detailed in Table 1.

[0332] Table 1 Comparison of Self-Generation Capabilities of Industrial Single-Axis Servo Motor Systems

[0333] Load disturbance recovery time 28ms 200ms 614% Average position following error ±0.032mm ±0.12mm 73.3% Success rate of adaptation to unknown scenarios 99.5% 12% 729% Security rule violation rate 0 3.2% 100% Long-term operating body drift rate ≤0.1% ≤8.5% 98.8% Anomaly location and self-repair time ≤3ms ≥20ms 85%

[0334] Example 2: Non-electronic physical dual mapping construction of industrial hydraulic pump station system

[0335] This embodiment uses a general-purpose hydraulic pump station system in a factory as the implementation object. The core purpose is to verify the engineering feasibility of the "electronic equivalent port mapping + computational tensor isomorphic mapping" dual mapping mechanism, providing engineering verification for the subsequent construction of a complete self-generated system. The basic process is to first modify the hardware by adding sensors, then anchor the self-generated system's operational consistency index, perform electronic equivalent port mapping and computational tensor isomorphic mapping, and finally connect to the eight-step closed-loop operation link of the self-generated system.

[0336] Basic Equipment Parameters. As this is a non-electronic device, it requires the addition of sensors and other hardware to configure equivalent ports. Details of the modifications are shown in Table 2. Equipment Model: General-purpose vane pump hydraulic station (service life: 5 years); Rated Pressure: 16MPa; Rated Flow Rate: 100L / min; Drive Motor: 22kW three-phase asynchronous motor; Existing Control: Manual relief valve pressure regulation + thermal relay overload protection; Native Capabilities: No programmable controller, no sensor feedback, no automatic adjustment capability.

[0337] Table 2 Hardware Addition List

[0338] pressure sensor 0-25MPa, 4-20mA output Pump outlet main oil circuit 1 Temperature sensor PT100, -20~120℃ Middle of the fuel tank 1 Ultrasonic liquid level sensor 0-500mm, 4-20mA output fuel tank sidewall 1 proportional relief valve 4-20mA control, 0-25MPa Replace the original manual relief valve 1 Current transformer 0-50A, 4-20mA output Motor input line 1 Turbine flow meter 0-150L / min, 4-20mA output Main oil circuit 1 PLC controller Siemens S7-1200 (1214C DC / DC / DC) + SM1231 AI4×16bit Inside the control cabinet 1 Touchscreen (optional) KTP700 Basic Control cabinet panel 1

[0339] All sensors and actuators are connected to the PLC analog module via shielded cables, using a 4-20mA industrial standard signal, providing strong anti-interference capabilities. The 16-dimensional simplified computing architecture can operate stably on a standard S7-1200 PLC with a 100ms control cycle. Testing showed that a single complete closed-loop calculation takes an average of approximately 45ms, with a CPU utilization of approximately 65%, leaving 55ms available for communication and other tasks.

[0340] Dual-layer operational consistency anchoring. The unified entity of this system is the self-generated system core running in the PLC, with the hydraulic pump station being its M-element physical extension. Operational consistency adopts a dual-layer structure of rigid core + dynamic evolution, ensuring that the system can adapt to environmental changes while maintaining its essential nature. The following inviolable safety constraints are fixed in the PLC's read-only memory area and cannot be modified throughout its entire lifecycle, achieving triple redundancy protection of hardware + software + high-speed interrupt, as detailed in Table 3.

[0341] Table 3. List of Rigid Operational Identity Constraints

[0342] System maximum pressure ≤16MPa pressure sensor Level 3 fuse trips, shutting off the motor. Pressure switch hardwiring + PLC software + high-speed interrupt Maximum oil temperature ≤65℃ Temperature sensor Level 2 intervention, flow restriction operation PLC software + thermal relay backup Maximum motor current ≤45A Current transformer Level 1 warning, reduce load PLC software + circuit breaker backup Minimum fuel level in fuel tank ≥200mm Liquid level sensor Level 3 circuit breaker, preventing startup. Float switch hardwiring + PLC software Pressure shock change rate ≤5MPa / ms pressure sensor Secondary intervention, closing the proportional valve PLC high-speed analog signal acquisition interrupt, response time ≤1ms

[0343] Definition and Constraints of Dynamic Operational Identity. Dynamic operational identity is a set of rules that allow for endogenous adjustments to a monistic ontology to adapt to environmental changes, equipment aging, and unknown failures, while maintaining a rigid core. All dynamic adjustments must undergo a four-level process—rigid redline verification, endogenous threshold generation, non-decreasing sustainability verification, and optimal candidate selection—before taking effect. In this embodiment, dynamic operational identity specifically includes the following endogenously evolving content:

[0344] The basic control parameter set includes the proportional coefficient Kp, integral coefficient Ki, and derivative coefficient Kd of the PID controller, used for closed-loop control of pressure and flow. The system continuously optimizes these parameters through self-evolution capabilities to adapt to equipment wear and changes in operating conditions.

[0345] Sensor compensation coefficient set: This includes drift compensation coefficients and nonlinearity correction coefficients for sensors such as pressure, flow, and temperature. The system learns and corrects sensor biases online through multi-sensor cross-validation.

[0346] The fault self-healing rule set includes rules for sensor drift compensation, valve jamming and vibration compensation, and leakage adaptive compensation. These rules are not pre-written, but are discovered and solidified autonomously by the system through intrinsic exploration.

[0347] The load adaptive rule set includes rules for adjusting pressure setpoints, flow allocation, and energy consumption optimization under different loads. The system automatically adjusts its control strategy based on real-time load changes to achieve optimal energy efficiency.

[0348] Environmental adaptation rule set: This includes compensation rules for hydraulic oil viscosity based on oil temperature, and compensation rules for system response speed based on ambient temperature. The system intrinsically generates environmental adaptation strategies through long-term operational data accumulation.

[0349] Evolutionary constraints: The adjustment range of all dynamic parameters must not exceed the threshold specified by the rigid operation identity; any newly generated rule must be verified for at least 100 control cycles, and the isomorphic constraint deviation rate must be ≤0.03 before it can be solidified; when dynamic adjustment causes the isomorphic constraint deviation rate to exceed 0.05, the system automatically reverts to the previous stable version of the rule set; the similarity between dynamic operation identity and rigid operation identity is defined as: the cosine similarity between the C primitive state vector corresponding to the dynamic rule set and the rigid operation identity benchmark vector, requiring a similarity ≥0.999. This similarity is used to evaluate the overall consistency between the rule set and the rigid core, and the isomorphic constraint deviation rate is used to evaluate the single-cycle closed-loop self-consistency. The two complement each other and jointly ensure system stability.

[0350] The electronic equivalent port mapping process includes core element calibration such as the body boundary, static characteristic calibration, dynamic characteristic identification, and standardized port encapsulation.

[0351] Ontology boundary and core element calibration

[0352] Irreducible body boundaries: hydraulic pump, motor, proportional relief valve, oil tank, main oil circuit pipeline and inlet / outlet ports.

[0353] Controllable input dimension: 1 drive port - proportional relief valve control current (4-20mA).

[0354] Observable output dimensions: 5 feedback ports – system pressure, output flow rate, oil temperature, motor current, and oil tank level.

[0355] Static characteristic calibration: Through offline testing, an accurate static mathematical model is established for each port.

[0356] Drive port (proportional relief valve) calibration: Test conditions were motor rated speed and all loads closed. The test yielded the following relationship between input current and output pressure: 0 MPa for 4 mA input current, 4.2 MPa for 8 mA, 8.1 MPa for 12 mA, 12.0 MPa for 16 mA, and 16.2 MPa for 20 mA. Linear regression yielded the following relationship between pressure and current: The linearity error is ≤0.8% FS.

[0357] Feedback port unified calibration:

[0358] Pressure sensor: (4mA corresponds to 0MPa, 20mA corresponds to 25MPa).

[0359] Flow meter: Nonlinear error ≤1% FS.

[0360] Current transformer: .

[0361] Liquid level sensor: .

[0362] Temperature sensor: PT100 three-wire system, linearized internally by the PLC module.

[0363] Dynamic characteristic identification: Identifying the system transfer function through step response testing.

[0364] Pressure control system: (First-order inertia plus pure delay, , ).

[0365] Temperature control system: (Large inertial element, time constant 120s).

[0366] Standardized port encapsulation involves uniformly encapsulating all ports within the PLC using software. Each port object includes the following standardized attributes: unique port identifier, port type (drive / feedback), raw signal range, engineering range, calibration gain, calibration offset, dynamic time constant, current real-time value, and safety upper and lower limits. All ports utilize unified read / write interface functions, achieving decoupling from the core of the upper-level self-generated system.

[0367] Instantiate the above 1 driver port and 5 feedback ports, for a total of 6 standard port objects, with all parameters corresponding one-to-one with the calibration results above.

[0368] The computational tensor isomorphism mapping steps include dimension alignment, orthogonal transformation tensor mapping, and CEM closed-loop recursive access.

[0369] Dimensional alignment and orthogonal transformation. To simplify engineering implementation and adapt to the computing power of ordinary PLCs, this embodiment adopts a 16-dimensional standard simplified architecture (the actual system can be expanded to 64 / 128 dimensions). The 5-dimensional physical parameter vector... Mapped to a 16-dimensional C-ary state vector .

[0370] The first 5 dimensions of physical semantics of the C primitive:

[0371] Dimension 1: Pressure Deviation (the difference between target pressure and actual pressure, normalized to [-1,1])

[0372] Dimension 2: Rate of change of flow (the difference between the current flow and the flow of the previous period, normalized to [-1,1])

[0373] Third dimension: Temperature normalized value (0 for 20℃, 1.0 for 100℃)

[0374] 4th dimension: Normalized value of motor current (0A corresponds to 0, 45A corresponds to 1.0)

[0375] 5th dimension: Normalized liquid level value (0mm corresponds to 0, 500mm corresponds to 1.0)

[0376] Dimensions 6-16: Reserved dimensions, initially set to zero, activated during runtime through endogenous learning.

[0377] Methods for normalizing physical quantities:

[0378] Pressure deviation:

[0379] Rate of change in flow:

[0380] temperature:

[0381] Current:

[0382] Liquid level:

[0383] Initial orthogonal transformation matrix:

[0384] C primitive state vector calculation:

[0385] Similarly, the E primitive state vector Through the same structure The mapping is obtained.

[0386] Transformation matrix update rule: Gradient descent is used, with the objective of minimizing the isomorphic constraint deviation rate. Matrix elements are updated every 1000 control cycles (100 seconds), and the learning rate is... The updated formula is: in The gradient of the isomorphic constraint deviation rate with respect to the transformation matrix is ​​approximated in the PLC using the finite difference method.

[0387] CEM closed-loop recursive access. The mapped C and E primitive state vectors, together with the M primitive composed of physical actuator states, form a complete CEM triplet, which is then accessed through the eight-step closed-loop operation link constructed by the self-generated system described in this invention. This embodiment will not elaborate further.

[0388] Example of self-resolved troubleshooting for proportional relief valve port jamming

[0389] This embodiment does not pre-define any fault modes, characteristic frequencies, or solutions for "valve jamming." The system has no prior knowledge that "jitter signals can resolve valve jamming." By monitoring indicators such as the isomorphic constraint deviation rate in real time, it determines that the system has an anomaly, but cannot identify the specific fault type. Since the fault type cannot be identified, the system automatically switches to the other-finite mode and initiates a global random exploration. The exploration space is limited to all feasible operations of the proportional relief valve control current, specifically including:

[0390] Base current adjustment: ±10% (relative to the current value);

[0391] High-frequency jitter superposition: frequency randomly selected from 10-100Hz, amplitude randomly selected from 0.1-0.5mA;

[0392] Intermittent flow interruption: Interruption duration 0.1-1 second, interval 1-5 seconds;

[0393] Safety constraints during the exploration process: When executing any candidate action, if the system pressure is below 8MPa or above 15MPa, the current candidate action is immediately interrupted and marked as invalid to avoid exploration leading to dangerous conditions;

[0394] Each candidate action is executed for 100 control cycles (10 seconds), and the change in isomorphic constraint deviation rate is recorded; exploration timeout protection is set: if no valid solution is found after 2000 consecutive control cycles (200 seconds), an audible and visual alarm is triggered and the system returns to safe operation mode;

[0395] After approximately 500 control cycles (50 seconds) of random exploration, the system discovered that the action of "superimposing a 15Hz high-frequency jitter signal with an amplitude of 0.2mA on the basic control current" could continuously reduce the isomorphic constraint deviation rate to below 0.03 and reduce the pressure fluctuation from ±1.2MPa to ±0.18MPa. This action was determined to be an effective solution.

[0396] The system updates the solution (overlaying a 15Hz high-frequency jitter signal, with the amplitude dynamically adjusted according to the deviation rate) to the fault self-healing rule set in the C primitive, increments the rule version number, and synchronously updates the dynamic operation consistency. If the same type of anomaly occurs again in the future, the rule can be directly called to achieve rapid self-healing.

[0397] Example 3: Multi-node peer-to-peer collaborative self-generating system for mining AGV clusters

[0398] This embodiment implements an explosion-proof AGV cluster (5 peer nodes) in a coal mine. The core hardware of a single AGV is an STM32H743 explosion-proof controller with a 480MHz Cortex-M7 core, 1MB SRAM, and 2MB Flash. It runs on bare metal without an operating system, with an iteration cycle of 10ms (endogenous time quantum). It complies with the standards GB / T 30029-2013 "General Technical Conditions for Automated Guided Vehicles" and GB 3836.1-2010 "Explosive Atmospheres Part 1: General Requirements for Equipment".

[0399] This embodiment demonstrates the collaborative self-generation of a cluster in a coal mine under high gas, strong dust, multiple obstacles, and unstable communication scenarios, verifying the global convergence, collaborative self-evolution, fault self-repair, and environmental adaptation capabilities of the distributed system.

[0400] Single-node rigid operation identity Maximum driving speed 1m / s, maximum acceleration 0.5m / s², minimum obstacle avoidance distance 30cm, emergency stop response time ≤50ms, explosion-proof safety constraints, CEM closed-loop mapping rules, Flash read-only partition lock;

[0401] Single-node dimensional constraints: In low-computing-power scenarios, minimum-dimensional closure constraints are used. ,Right now The single-cycle full-process execution time is ≤2ms, which is fully compatible with low computing power explosion-proof hardware in wells;

[0402] Peer-to-peer architecture: Five AGVs establish encrypted peer-to-peer connections through an underground LoRa trusted communication link. There is no central master node. All nodes run the self-generated system construction method of this invention and periodically synchronize the CEM capability matrix, unary ontology sustainability index, and C primitive rule feature vector.

[0403] Core competency verification

[0404] Cluster self-generation and global convergence: After all nodes complete the bootstrapping process, a single-node self-generated system is built internally, and global convergence of the cluster is achieved through peer-to-peer collaborative coherence calculation. in The coherence of peer-to-peer coordination was calculated. All nodes The cluster remained stable for 10 consecutive cycles, indicating global convergence and self-organization without intervention from a central node.

[0405] Cluster Adaptive Capability: Dust in underground tunnels can cause a decrease in the accuracy of lidar ranging. The system can locate spatial dimension anomalies by using multi-dimensional coherence and intrinsically adjust the SLAM grid resolution and obstacle avoidance rules. The cluster can maintain stable movement throughout the process without collisions or stalls.

[0406] Cluster self-healing capability: If a single AGV encoder fails, the node triggers an intrinsic self-healing closed loop. At the same time, it synchronizes the cluster rules through the peer-to-peer collaborative link. The remaining nodes intrinsically adjust their driving paths and collaborative rules. There is no congestion or task interruption in the cluster. After the faulty node recovers, it automatically re-enters the cluster. The global convergence recovery time is ≤100ms.

[0407] Cluster self-evolution capability: During 72 hours of continuous underground transportation operations, the cluster operates in dual modes synchronously, endogenously optimizing path planning rules and collaborative obstacle avoidance rules, thereby improving transportation efficiency by 42%, reducing energy consumption by 27%, and not breaking the explosion-proof safety constraints throughout the entire process.

[0408] The above three embodiments are intended to illustrate the full-scene adaptability, engineering feasibility, and significant beneficial effects of the present invention. Those skilled in the art can make non-creative modifications and scene adaptations to the embodiments based on the core principles of the present invention, all of which fall within the protection scope of the present invention.

[0409] The operational identity described in this application is the only core anchor point for maintaining the identifiability of the self-generated system throughout its entire life cycle and achieving convergent evolution. Any multi-fixed-point architecture is either essentially a sub-target split of a single fixed point or will lead to the splitting and divergence of the system ontology, making it impossible to achieve true self-generation. If such a multi-anchor-point architecture can achieve steady-state self-generation, it can necessarily be classified into a single global operational identity, which is an equivalent variation of this invention and falls within the scope of protection.

[0410] The self-referential and other-referential bimodalities described in this application belong to the unique minimal complete dual set of the evolutionary dynamics of a self-generated system. Any claimed two-sided, two-image, binary, or other similar architectures are essentially terminological substitutions of the self-referential-other-referential bimodalities of this invention, and their core functions are completely consistent with the bimodalities of this invention, thus constituting equivalent features. Any claimed third or higher independent states are essentially sub-interval hierarchies, weight refinements, and scene splits of the aforementioned bimodalities, lacking an evolutionary dynamic core that is independent of the bimodalities and cannot be merged, thus failing to break through the functional boundaries of the bimodalities of this invention. Such polymorphic architectures are non-substantial refinements and variations of the bimodalities of this invention, constituting equivalent features and falling within the protection scope of this invention.

[0411] The CEM three-body primitives described in this application are functional primitives classified according to the irreducibility of the core functions of a self-generated system, rather than by the number of physical structures or sub-units. Regardless of the number of physical sub-units or the hierarchical division of the substructure of the target system, as long as its irreducible core functions can completely correspond to the three categories of information guidance, energy drive, and material carrying, it falls within the coverage of the three-body primitives of this invention. Multi-body architectures named four-body, five-body, six-body, eight-body, nine-body, or other N-body architectures, regardless of their terminology or structural division methods, must have irreducible core functional primitives that can be completely divided into three categories: the first category is information guidance functions that define system rules, constraints, convergence targets, and potential possibilities, which essentially correspond to the information field state C primitive of this invention; the second category is energy drive functions that realize system power transmission, rule implementation, and state transformation, which essentially correspond to the energy mechanism E primitive of this invention; and the third category is material carrying functions that constitute the physical carrier, boundary, and topological structure of the system, which essentially correspond to the material structure M primitive of this invention.

[0412] The core of the closed-loop recursion described in this application is the cyclic coupling of the CEM three-body primitives. Regardless of its nesting depth, the number of sub-loops, or the recursive iteration steps, as long as the core is an endogenous cycle of "information rules → energy drive → material carrier → information rule update", it falls within the scope of the recursive closed loop of this invention. Any multi-level, multi-nested recursive architecture is essentially a fractal extension of the recursive closed loop of this invention, and is an equivalent feature, falling within the scope of protection.

[0413] The following solutions are all equivalent solutions of this invention and fall within the protection scope of this invention: Solutions that perform terminology replacement, sub-function decomposition, and multi-body merging and naming on the CEM three-body primitives of this invention, with core functions completely consistent with the three-body primitives of this invention; solutions that perform terminology replacement, sub-state hierarchicalization, and polymorphic decomposition on the self-referential-other-referential bimodality of this invention, with core functions completely consistent with the bimodality of this invention; solutions that perform terminology replacement and sub-target decomposition on the operational identity fixed point of this invention, with the core being a unique convergence anchor point throughout the system's entire lifecycle; solutions that perform hierarchical nesting, multi-step decomposition, and sub-loop refinement on the recursive closed loop of this invention, with the core being CEM cyclic coupling; and other solutions that, compared with the technical features of the claims of this invention, employ essentially the same means to achieve essentially the same functions and effects, and which can be conceived by those skilled in the art without creative effort.

[0414] The core inventive contribution of this application is the discovery and verification that "unary operation identity fixed point + self-referential-other-referential bimodal evolutionary dynamics + information field state C - energy mechanism E - matter structure M three-body irreducible primitive + recursive closed loop" is the unique minimum complete meta-architecture of all self-generating systems capable of self-generation, self-maintenance, self-repair, and self-evolution. Any scheme of "unary two-state three-body + recursive closed loop", regardless of how the terminology is replaced or how the structure is refined, falls within the scope of the claims of this application; any scheme of "unary N-body M-state + recursive closed loop", as long as it can achieve the core technical effect of the self-generating system, is a technical feature of this application and falls within the protection scope of this invention.

Claims

1. A method for constructing a univariate, two-state, three-body self-generating system, the method comprising the steps of closed-loop state acquisition, state evaluation, rule updating, and iterative looping, executed by an electronic computing device with sensing, computing, storage, and execution capabilities, used to construct the self-generating capability of a target system, realizing the self-generation, self-maintenance, self-repair, and self-evolution of the target system, characterized in that, The method takes the identity of two-layer operations as the sole core convergence objective for the entire lifecycle of the target system, uses the univariate ontology survival index as the core quantitative basis, takes the synchronous operation of self-indicating and other-indicating modes throughout the entire lifecycle as the endogenous evolutionary driving force, and takes the closed-loop recursive endogenous generation of three irreducible functional primitives—information field state C, energy mechanism E, and material structure M—as the sole minimum operational link. It includes the following continuously executed steps: S1 Bootstrapping and Isomorphic Mapping: Collect the initial state parameters and underlying physical or rule-based security constraints of the target system, anchor the underlying security constraints as rigid core-layer operational identity, lock them in a read-only state throughout the entire lifecycle, and prohibit any evolutionary modifications; construct the self-generated system core architecture of CEM closed-loop recursion, and generate isomorphic rules through the CEM closed-loop recursive endogenous generation within the architecture. Within the physical or rule-based scope of the target system, endogenously generate the initial rule set of the information field state C primitive, the energy mechanism E primitive, and the material structure M corresponding to the operational identity of the dynamic evolution layer. The initial state vector, mapping weights, and bias parameters of the primitives, along with the endogenous time, space, and phase dimensions, complete the unary ontology anchoring of the target system. This step does not pre-set external task objectives or manually fixed operational order, but only locks the aforementioned insurmountable underlying security constraints. All endogenous generation operations in this step are executed by the processor of an electronic computing device, and the generated rules, vectors, and tensors are stored in computer-readable storage media. S2 Closed-loop state acquisition and self-consistency deviation calculation: Through the feedback interface of the target system, the real-time state, environment, and interactive feedback data of the three primitives of the CEM in the current iteration cycle are collected, and the self-consistency deviation rate of the CEM closed-loop recursion is calculated. Based on self-consistency deviation rate The self-consistent matching degree of CEM recursive coupling with which it is quantitatively negatively correlated was obtained. S3 Real-time assessment of the persistence of a univariate ontology: based on the self-consistency deviation rate C-EM recursive coupling self-consistent matching degree Calculate the univariate ontology persistence index based on the collected state data. ; Based on historical time-series data of the unary ontology persistence index, a dynamic high threshold is endogenously generated. With dynamic low threshold Simultaneously, it internally defines the core survival red line and peripheral capability boundaries of the system; S4 Gradient security control: Based on the univariate ontology survival index, core survival red line and dynamic threshold, it performs hierarchical security intervention, generates bimodal operation permissions and weight constraint rules, and applies them to subsequent bimodal evolution stages; S5 Bimodal weight adjustment and candidate rule internal pre-verification: Based on the bimodal operation permissions and weight constraint rules, it internally and dynamically adjusts the operation weights of self-indicating mode and other-indicating mode; Under normal operating conditions, the two modes run synchronously in each iteration cycle without mutual exclusion pause logic; only when the third-level circuit breaker is triggered, the rule generation and update function of the other-pointing mode is temporarily paused; the self-pointing mode continues to execute the closed-loop recursion of the current C primitive rule, anchoring the existing dynamic steady-state order of the system; He pointed out that the modal synchronous operation generates candidate rule attractors endogenously and performs pre-compliance verification of candidate rules to select the unique and optimal candidate rule; S6 CEM three basic elements symbiotic reshaping and endogenous spatiotemporal phase dimension generation: with the identity of the rigid core layer operation as an absolute constraint, together with the identity of the dynamic evolution layer operation and the optimal candidate rule selected in step S5, the compliance update of the C element rule is completed, the E element energy mechanism is adapted endogenously at the same time, and the M element material structure state vector is reshaped at the same time to realize the symbiotic evolution of the three basic elements of CEM; at the same time, based on the recursive coupling state of the three basic elements of CEM, the three dimensions of time, space and phase are generated endogenously to construct the endogenous spatiotemporal phase dimension deeply bound to the target system operation state; Based on the updated CEM primitives and the endogenous spatiotemporal phase dimension, a five-dimensional holographic coherent state tensor is generated. Calculate global coherence And multidimensional coherence; S7 closed-loop self-consistency test, steady-state anchoring and bidirectional self-optimization: based on the CEM primitive state after co-evolution, endogenous spatiotemporal phase dimension parameters, and global coherence. And the coherence of each dimension, perform closed-loop self-consistency steady-state convergence test, and complete the endogenous steady-state anchoring; If all tests pass, the target system is determined to be in a stable state in this iteration, and the recursive coupling state of the CEM primitives and the endogenous spatiotemporal phase space parameters are locked after this update. At the same time, the pre-verification model in step S5 is optimized in reverse. If any test fails, the corresponding level of abnormal handling mechanism is triggered, including endogenous self-healing closed loop, steady-state rollback, and graded safety intervention. S8 Iterative Cycle and Steady-State Relay: The steady-state effective CEM primitive state and endogenous spatiotemporal phase dimension parameters locked in step S7 are used as the initial state for the next iteration cycle. Jump back to step S2 to start the next iteration and synchronously repeat the entire execution chain of steps S2-S7. All update operations are completed through the execution chain of the target system, and it is forbidden to modify the uniformity of rigid core layer operations.

2. The method for constructing a univariate two-state three-body self-generated system according to claim 1, characterized in that, The quantification expression for the identity of the two-layer operation is a unique operation identity fixed point throughout the entire lifecycle: in For rigid operational identity, that is, the rigid core layer operational identity of the target system, the underlying security constraints that cannot be modified in the information field C primitive, the core rules for the survival of the unary ontology, the mathematical form and dimensional closure constraints of the CEM closed-loop recursive mapping are locked in a read-only state throughout the entire life cycle. For dynamic operational identity, i.e., the operational identity of the target system's dynamic evolution layer, the corresponding endogenously evolving set of scenario adaptation rules in the information field C primitive, and the dynamic weights and bias parameters of the CEM mapping, are determined by the univariate ontology persistence index. Real-time representation.

3. The method for constructing a univariate two-state three-body self-generated system according to claim 1, characterized in that, The CEM closed-loop recursive endogenous generation isomorphic rules include: (1) closed-loop endogenous generation flow rules, which are bidirectional dialectical coupling closed loops of forward recursive generation and reverse recursive constraints of C primitives, E primitives, and M primitives, with the basic flow path being: In the formula, , , These are the state vectors of the information field state C, the energy mechanism E, and the matter structure M, respectively. Information-guided mapping for the endogenous generation of E primitives from C primitives. The energy-driven mapping for the endogenous generation of M primitives from E primitives. The structure feedback mapping for the endogenous generation of C primitives from M primitives, all three mappings are Lipschitz continuous mappings; (2) the closed-loop recursive isomorphism constraint is the core guarantee of the identity of the univariate ontology of the self-generated system, satisfying the formula: In the formula, For function composition operators, the composition order strictly corresponds to the closed-loop endogenous generation flow path; They are topologically isomorphic and equivalent; (2) The identity mapping on the C primitive, that is, the reference mapping that does not change the C primitive ontology; (3) The parameterized form of the single-step endogenous generation mapping, corresponding to the endogenous generation mapping weights and bias parameters in S1: In the formula, , , This is the weight matrix for the corresponding mapping. , , The bias parameters corresponding to the mapping are all optimizable parameters of the dynamic evolution layer; spectral normalization is performed on the three weight matrices in each iteration cycle, and their maximum singular values ​​are restricted to less than 1 to ensure the Lipschitz constant of the composite closed-loop endogenous generative mapping. (4) The closed-loop self-consistency deviation rate is not higher than the preset compliance threshold. The closed-loop self-consistency deviation rate is used to verify the degree of isomorphic constraint satisfaction and the system convergence state. The calculation formula is: In the formula, For the first The C-based primitive state vector of the iteration period. for Norm, To prevent small positive numbers from being divided by zero, the range of values ​​is: The preset compliance threshold is initially set to 0.

05. At that time, the closed-loop self-consistency is deemed satisfactory; the isomorphic constraint loss function is defined as follows: , which serves as the objective function for optimizing the mapping parameters.

4. The method for constructing a univariate two-state three-body self-generated system according to claim 1, characterized in that, The endogenous time, space, and phase dimensions are entirely generated endogenously from the recursive coupling states of the three fundamental units of the self-generated system CEM, without any externally pre-defined spacetime container. Specific endogenous rules include: Endogenous time dimension: anchored to the energy mechanism E fundamental unit, the minimum endogenous time quantum is the time taken for the three fundamental units of CEM to complete one full closed-loop recursive mapping, the iteration period is an integer multiple of the endogenous time quantum, and the length of the time dimension is endogenously determined by the number of iterations throughout the system's entire lifecycle; Endogenous space dimension: anchored to the material structure M fundamental unit, corresponding to the discrete grid points of the physical or regular space of the M fundamental unit, the grid resolution is endogenously determined by the state sampling accuracy of the M fundamental unit, and the boundary of the space dimension is endogenously determined by the safe physical or regular boundary of the M fundamental unit; Endogenous phase dimension: anchored to the information field state C fundamental unit, corresponding to the coupling feature channel of the three fundamental units of CEM, serving as the coupling medium connecting the endogenous time dimension and the endogenous space dimension, the number of dimensions is endogenously determined by the sum of the feature dimensions of the three fundamental units of CEM.

5. The method for constructing a univariate two-state three-body self-generated system according to claim 1, characterized in that, In step S2, the CEM recursive coupling self-consistent matching degree The calculation formula is: The CEM recursive coupling self-consistent matching degree is used to quantify the degree of matching between the dynamic evolution layer of the target system and the fixed point of operational identity.

6. The method for constructing a univariate two-state three-body self-generated system according to claim 1, characterized in that, In step S3, the unary ontology persistence index The quantization expression is: Where the hard constraint is ,and , The minimum guarantee weight for core identity cannot be exceeded throughout the entire lifecycle; , as the ontological persistence term, quantifies the degree of matching between the material structure M primitive and the safety benchmark configuration; the aforementioned This refers to the environmental adaptability item, which quantifies the adaptability of the target system to the external environment; the aforementioned For endogenous adaptive weights, the update formula is: in The basic weight for the ontology's persistence item; These are adaptive coefficients; The sliding standard deviation of the ontology duration term over the past T endogenous time quantum terms; self-referential mode weights. He refers to modal weights This enables the endogenous binding of dual-modal weights and sustainability indicators, providing a benchmark constraint for subsequent dynamic adjustments.

7. The method for constructing a univariate two-state three-body self-generated system according to claim 1, characterized in that, In step S3, the core sustainability red line and dynamic high threshold are generated endogenously based on the univariate ontology sustainability index. With dynamic low threshold The generation rules include: (1) the core survival red line formula: The core survival red line can only be adjusted upwards and cannot be adjusted downwards, which is the ultimate safety bottom line that cannot be broken throughout the entire life cycle of the system; (2) Dynamic threshold formula: in: for The moving average of the past 50 iterations. This represents the moving standard deviation for the corresponding period. , The threshold coefficients are initially set to 0.5 and 2.0, and can be finely adjusted based on the historical steady-state recovery success rate and false trigger rate.

8. The method for constructing a univariate two-state three-body self-generated system according to claim 1, characterized in that, In step S4, the specific rules for the gradient-based security control include: a. Level 1 warning: when When a critical phase transition state of the system is detected, an early warning is triggered, the rule adoption threshold is raised to 0.95, and only candidate attractors with significantly higher persistence than the current rule are adopted, thus strengthening steady-state constraints; b. Secondary intervention: when If the steady-state target is not reached after 5 consecutive iterations, the modification permission of the core rule segment of the C primitive is locked, only the parameter fine-tuning of non-core rules is allowed, the self-referential mode dominates the steady-state recovery, and the permission to update the mapping parameters is frozen; c. Three-level circuit breaker: when If the underlying security constraints of rigid operation homogeneity are breached or there is a risk to the survival of the ontology, the rule update channel of the other-pointing modality is immediately cut off, the system switches to the security steady state mode dominated by the self-pointing modality, the preset security maintenance procedure is executed, and an unalterable audit log is generated. All mapping parameters and rule updates are frozen with full permissions.

9. The method for constructing a univariate two-state three-body self-generated system according to claim 1, characterized in that, In step S5, the dynamic weight adjustment rule for the bimodal mode includes: a. self-referenced modal weights. His modal weights Satisfy the global dynamic normalization requirements: The bimodal operation runs synchronously in each iteration cycle under normal operating conditions, without mutual exclusion pause logic; the rule update function of the other modality is temporarily paused only when the third-level circuit breaker is triggered; b. The weight endogenous adjustment formula is: in The sigmoid activation function is used. The steepness coefficient is determined by the endogenous timescale of the target system; To minimize the operational weight of the dual-mode operation, a value of 0.05 is preferred to ensure synchronous operation of the dual-mode operation throughout its entire lifecycle; c. When hour, Approaching 0.95, the target system is maintained in a steady state primarily by the self-exponential mode, while the other-exponential modes operate synchronously in a low-weight exploratory state; d. When hour, Automatic improvement enhances the target system's ability to explore rules and restore order in other modalities.

10. The method for constructing a univariate two-state three-body self-generated system according to claim 1, characterized in that, Step S5 also includes the rule update strength of the other modal. Using a continuously differentiable endogenous function, with its value strictly limited to the range of 0 to 0.2, the formula is as follows: During project implementation, upper and lower limits are applied to the input of the exponent term, with an upper limit of 0 and a lower limit of -20, to avoid the risk of numerical overflow under extreme operating conditions. The final execution formula for rule updates is: in For the current rule tensor, The optimal candidate rule attractor is selected; in steady state, only candidate attractors with higher sustainability than the current rule are adopted to ensure that the sustainability of the target system is non-decreasing.

11. The method for constructing a univariate two-state three-body self-generated system according to claim 1, characterized in that, In step S5, the pre-compliance verification of the candidate rules for the execution of the self-referential modality is a predictive admission verification before the rule is implemented. Specifically, it includes the following steps: S501 Rigid Red Line Verification: Verify that the cosine similarity between the segments corresponding to the rigid core layer operations in the candidate rules and the baseline rule with full lifecycle read-only locking is ≥0.999, and eliminate invalid candidate rules that violate the red line; S502 Endogenous Threshold Generation: Based on the self-referential modal weights of the current period... Endogenous generation rule adoption threshold S503 Non-decreasing Survival Verification: Based on the system's endogenously fitted rule change-survival change mapping model, the predicted survival index for the next N periods corresponding to the candidate rule is calculated through a multi-step look-ahead prediction formula. Only when... and Retain candidate rules; S504 Optimal candidate screening: For candidate rules that pass the verification, perform multi-objective non-dominated sorting, with the optimization objective priority as follows: maximize the prediction sustainability index → ​​minimize the rule change magnitude → minimize the prediction self-consistency bias rate, and screen out the unique legal optimal candidate rule.

12. The method for constructing a univariate two-state three-body self-generated system according to claim 1, characterized in that, In step S7, the endogenous self-repairing closed-loop process includes the following steps: (1) Endogenous damage identification: Based on the temporal mutation characteristics of the self-consistency deviation rate, the continuous downward trend of the univariate ontology sustainability index, and the compliance verification results of the rigid core layer, the type and degree of steady-state damage of the system are endogenously identified without the need for manual preset fault types and fault thresholds; (2) Endogenous generation of repair rules: Based on the damage identification results, under the absolute constraint of rigid operation identity, a dedicated repair rule attractor is generated endogenously through the other-find modality, instead of calling the manually preset fault handling program; (3) Closed-loop simulation and verification of repair effect: Based on the system's endogenous CEM closed-loop mapping model and historical steady-state benchmark data, a multi-step positive closed-loop simulation is performed on the generated repair rules to simulate the full-cycle operation state of the system after the repair rules are executed, and only the effective candidate repair rules whose self-consistency deviation rate and univariate ontology sustainability index can stably return to the compliance range after the simulation are retained; (4) Steady-state repair and secondary verification: The optimal repair rule is executed, the repair adaptation of the three basic CEM elements is completed simultaneously, and step S7 is re-executed. The closed-loop self-consistency steady-state convergence test is performed; if all tests pass, the endogenous self-repair is completed and the steady state of the repaired system is locked; if the compliance requirements still cannot be met after repair, a fallback safe steady-state rollback is performed, and an immutable repair audit log is generated.

13. The method for constructing a univariate two-state three-body self-generated system according to claim 1, characterized in that, In step S7, the closed-loop self-consistency steady-state convergence test is an empirical acceptance test after the rule is implemented. It involves four progressive closed-loop self-consistency steady-state convergence tests, specifically including: S701 Rigid core layer check: verifying that the cosine similarity between the evolved C primitive rigid core layer segment and the read-only baseline rule is ≥0.9999. If it fails, it directly reverts to the previous cycle's safe steady state, triggering a level 3 circuit breaker; S702 Single-cycle convergence test: verifying the Lipschitz constant of the evolved CEM composite closed-loop mapping. To ensure the system's global asymptotic stability, if the test fails, it reverts to the previous cycle's safe steady state, triggering a second-level intervention; S703 Dynamic Layer Self-Consistency Test: Verifies the actual self-consistency deviation rate after evolution. Furthermore, the actual sustainability index must be greater than or equal to the sustainability index of the previous cycle; otherwise, an endogenous self-repairing closed loop is triggered. S704 Full-Cycle Steady-State Convergence Test: Using the system's endogenous time quantum as the smallest unit, verify that the moving average of the self-consistency deviation rate over 10 cycles is ≤0.05, the moving standard deviation of the sustainability index is ≤0.01, and the global coherence... If it fails, the weight of the self-referenced modality is increased, and the update magnitude and mapping parameter update of the other-referenced modality are frozen.

14. The method for constructing a univariate two-state three-body self-generated system according to claim 1, characterized in that, In step S7, the specific rules for the reverse optimization step S5's pre-compliance verification model are as follows: If all verifications in this iteration pass, the learnable parameters of the rule change-persistence change mapping model are updated to improve prediction accuracy, and effective innovation results are solidified based on the innovation effectiveness evaluation results; if the verification in this iteration fails, the rule adoption threshold in step S5 is automatically tightened, the mapping model is corrected, and the recurrence of similar prediction biases is avoided; the innovation effectiveness evaluation formula is: in ; The difference between the mean survival rate after the innovation is implemented and the mean survival rate before the innovation. To achieve the sliding standard deviation of the innovation's persistence after implementation, To improve the success rate of adaptation in multiple scenarios; only when Furthermore, if the constraints are met for more than three consecutive verification cycles, the innovative achievements will be incorporated into the system steady-state evaluation system.

15. The method for constructing a univariate two-state three-body self-generated system according to claim 1, characterized in that, Step S1 also includes isomorphic mapping for non-electronic, non-computational entity target systems through a dual-mapping construction mechanism. The specific execution rules are as follows: S101 Electronic equivalent port mapping: The irreducible ontology boundary, underlying security constraints, controllable input dimension, and observable output dimension of the non-electronic, non-computational target system are calibrated and encapsulated as controllable driving equivalent ports and state feedback equivalent ports through static characteristic calibration and dynamic characteristic identification; S102 Computational tensor isomorphic mapping: The parameters of the equivalent ports are mapped to computable state vectors that are fully aligned with the C and E primitive dimensions through dimension alignment orthogonal transformation, and incorporated into the CEM closed-loop recursive coupling link and convergence control system; After the non-electronic entity target system is constructed, its underlying security constraints are synchronously anchored to the identity of the two-layer operation, and its state vector is connected to the unary ontology persistence evaluation process in real time.

16. The method for constructing a univariate two-state three-body self-generated system according to claim 1, characterized in that, It also includes multi-level master-slave nested collaboration and global convergence steps: data interaction between the target system and the upper and lower level master-slave systems is realized through interfaces; the global order convergence of the multi-level nested system is realized through a game-theoretic selection mechanism, completing the collaborative self-generation of multiple systems; the interfaces include three types: upward exposure interfaces, downward receiving interfaces, and peer-level interaction interfaces, and the interface data format strictly matches the CEM primitive dimension; the self-consistency compatibility of the upper and lower level C primitive rules is quantified by the cross-level coherence index, and the calculation formula for cross-level coherence is: In the formula The C primitive rule feature vectors are the upper-layer master system and the lower-layer slave system, respectively. When the coherence is less than 0.8, the rule conflict correction process is triggered. The global convergence judgment rule is: if and only if the unary ontology survival index of the upper-layer master system and all lower-layer slave systems meets the steady-state requirement, and the average cross-level coherence between the upper and lower layers is ≥0.9, and the above requirements are met for more than 10 consecutive iterations, it is judged as global convergence.

17. The method for constructing a univariate two-state three-body self-generated system according to claim 1, characterized in that, It also includes the temporal, spatial, and phase dimensions intrinsically constructed based on the three fundamental CEM units, simultaneously constructing a five-dimensional holographic coherent tensor space adapted to the system's self-sustaining, self-repairing, and self-evolutionary requirements. This achieves high-order non-additive coupling of the three fundamental CEM units, providing core support for unary ontology persistence assessment and anomaly localization. The basic structure of the five-dimensional holographic coherent tensor space is... Each dimension is intrinsically determined and dynamically adjusted by the self-generated system. The specific rules for intrinsic determination are as follows: For spatial dimension group, corresponding to the three-dimensional spatial dimension built endogenously by the system, it is deeply related to the interphase dimension. The number of grid points is endogenously determined by the system based on the univariate ontology sustainability index and spatial resolution requirements. The grid point division strictly satisfies Shannon sampling theorem. For time window dimension group, corresponding to the time dimension of the system's endogenous construction, the sequence length is based on the smallest endogenous time quantum of a complete closed loop recursion of CEM, and is dynamically determined by the system based on the iteration cycle and the univariate ontology persistence index. The feature channel dimension group corresponds to the interphase dimensional features of the recursively coupled three primitives of CEM. The dimension size is the sum of the feature dimensions of the three primitives. The dimensional arrangement corresponds one-to-one with the feature structure of the three primitives, with no forced flattening requirement, and dynamically adapts with the endogenous evolution of the primitives. The coupling generation formula of the five-dimensional holographic coherent state tensor is: in , The higher-order coupling weights are intrinsically and adaptively adjusted by the system based on the unary ontology persistence index. For ternary higher-order coupling terms, two intrinsically switchable implementation methods are supported: Method 1: High-precision full coupling, the formula is... Method 1: Retain the original topology of CEM primitives without flattening; Method 2: Engineering-adaptive coupling at the edge, the formula is... This provides a simplified engineering approach for scenarios with limited computing power; it also includes calculation steps for global coherence and multi-dimensional coherence, with the global coherence calculation formula being... The dimensional coherence includes at least one of spatial coherence, feature coherence, and temporal coherence, and is used to locate the specific dimension in which the system anomaly occurs and trigger corresponding adaptive adjustment or self-repair actions.

18. The method for constructing a univariate two-state three-body self-generated system according to claim 1, characterized in that, It also includes a multi-node peer-to-peer collaborative extension step: multiple peer-to-peer devices running this method establish encrypted peer-to-peer connections through a trusted communication link, periodically synchronize the CEM capability matrix and unary ontology sustainability index of each node, and perform cross-node CEM closed-loop recursive mapping and collaborative self-generation; the formula for calculating the coherence of peer-to-peer collaboration is: In the formula, N is the total number of nodes. Let C be the feature vector of any two peer nodes; the global convergence criterion is: if and only if the unary ontology persistence index of all nodes meets the steady-state requirement, and the peer-level collaborative coherence is ≥0.9, and the above requirements are met for 10 consecutive iterations, it is determined to be globally converged.

19. A univariate two-state three-body self-generating system, characterized in that, The system operates the construction method described in any one of claims 1 to 18, including a hardware carrier layer and a function implementation layer; the hardware carrier layer includes a processor, memory, communication interface, sensor and actuator, or an electronic device or computing system with computing, storage, communication and sensing execution capabilities, providing a physical interaction basis for the system; the function implementation layer includes functional modules corresponding one-to-one with the method steps of claim 1, realizing the self-generated capability construction and operation of the target system.

20. An electronic device, characterized in that, include: One or more processors and one or more memories; the one or more memories are coupled to the one or more processors, and the one or more memories are used to store computer program code, the computer program code including computer instructions; when the computer instructions are executed on the processor, the electronic device causes the electronic device to perform the univariate two-state three-body self-generated system construction method as described in any one of claims 1 to 18.

21. A non-volatile, non-transient computer-readable storage medium, characterized in that, It includes computer-readable instructions that, when executed on an electronic device, cause the electronic device to perform the unary two-state three-body self-generated system construction method as described in any one of claims 1 to 18.

22. A computer program product, characterized in that, It includes computer-readable instructions that, when executed on an electronic device, cause the electronic device to perform the unary two-state three-body self-generated system construction method as described in any one of claims 1 to 18.

Images