Fractal topology-based macro-mechanical superstructure and its decentralized load scheduling method

Abstract

The application discloses a macro-mechanical superstructure based on fractal topology and a decentralized load scheduling method thereof, and aims to solve the problems of rigid lever locking and spherical lap dislocation. The structure comprises fractal units with self-similarity, and the fractal units are connected through linkage transmission assemblies; the linkage transmission assembly comprises a posture synchronous gear transmission lever assembly, which is used for compensating space interference and forcibly synchronously conducting displacement; and through the elongated guide shaft and the deep hole guide structure, forced coaxial alignment after unloading is realized. The fractal topology force flow ripple conduction mechanism of the application can be mapped to the fields of AI computing power scheduling and 6G radio frequency, and has high industrial generality.

Description

Technical Field

[0001] This invention belongs to the field of mechanical metamaterials and vehicle collision avoidance and industrial vibration reduction technology, specifically involving a macroscopic mechanical superstructure based on fractal topology and its decentralized load scheduling method. Background Technology

[0002] Local overload and multi-node force flow interference are persistent problems in mechanical structures under complex loading conditions. Recent research in fractal topology and mechanical metamaterials has shown that self-similar structures can significantly improve stress distribution. The applicant previously filed a patent application entitled "A Load Scheduling Architecture Based on Fractal Topology," which disclosed a load scheduling infrastructure based on fractal topology. This infrastructure achieves multi-level load splitting and self-centering through large and small fractal units and hemispherical concave-convex connections. However, during subsequent in-depth research and engineering simulations, the applicant discovered the following two inherent physical interference defects in this infrastructure: 1. Rigid lever motion lock-up: The prior application used a fixed rigid three-pronged lever with no axial sliding capability. When the three ends of the three-pronged lever synchronously retract towards the fulcrum due to the load, the rigid lever cannot provide lateral clearance space due to the objective difference in the movement stroke of the three points in space, resulting in physical lock-up of the mechanism and failure of the synchronous attitude adjustment function of the multi-unit. 2. Failure of Hemispherical Overlap Joint Alignment: In the previous application, the small and large units relied solely on the spherical contact between the hemispherical protrusion and the hemispherical recess for overlap and self-centering, lacking a forced guiding and limiting structure. When the load is removed and the small unit is unloaded and reset, radial displacement easily occurs at the spherical overlap. Under renewed pressure, the concave-convex structure cannot be accurately aligned, resulting in uneven load distribution and overlap interference. Furthermore, the applicant discovered in subsequent analysis that the prior application's description of the triggering logic was not sufficiently clear. In fact, the activation principle of the fractal unit should be: regardless of whether the small unit's sensing point is triggered individually, partially, or all at once, the large unit will participate in the coordinated load distribution; when the three sensing points of the small unit are simultaneously compressed and reach a preset torque balance lock-up state, the small unit is in a hard-locked state because the lever cannot generate internal deflection; under continuous external load, the compression spring of the large unit below the small unit provides a downward stroke margin, and the small unit as a whole compresses the compression spring, thereby generating an overall vertical translation. This downward compression and translation process itself completes the physical decoupling of the torque balance dead point, and through this action, the force flow is forcibly transmitted to the large unit and adjacent fractal units, realizing the release of the force flow. Existing technologies can usually only alleviate the above defects by adjusting the size ratio of the lever and optimizing the fit tolerance of the sphere, but neither can fundamentally eliminate the physical interference of these two types of coupling at the geometric kinematic level. Summary of the Invention

[0003] The technical problem to be solved by this invention: This invention aims to solve the technical problems of rigid lever motion lock-up and spherical overlap alignment failure in existing fractal topology load scheduling architectures, and to provide a macroscopic mechanical superstructure with more accurate triggering logic. Technical solution: To address the aforementioned technical problems, this invention, based on the entire basic architecture disclosed in the prior applications (i.e., multi-level large and small fractal units, three-pronged linkage levers, spring assemblies, and hemispherical concave-convex connections), makes substantial improvements to the physical structure, as follows: 1. Add a posture synchronization gear transmission lever assembly: Based on the transmission gear pair that can slide laterally, the rigid synchronization plate and the return spring, a linkage transmission assembly with coordinated gear rotation displacement, lateral sliding displacement, axial extension displacement and angular deflection displacement is constructed. 2. Add an extended guide shaft and a deep hole guide structure at the fulcrum: The central tie rod of the large unit compression spring assembly is extended into a thin guide shaft, and a suitable coaxial depth guide hole is machined at the fulcrum of the small unit three-pronged lever, using the spring tie rod itself to form a forced coaxial guide mechanism. 3. Correction and clarification of triggering logic: In the fractal unit of the present invention, regardless of whether the sensing point of the small unit is triggered individually, partially, or all of them, the large unit participates in the coordinated current splitting; only when the three points are triggered synchronously, the large unit is forced to participate in the decoupling of the three-point lever torque balance. 4. Improve the transmission logic of bidirectional working conditions: When the small unit is subjected to synchronous upward load from three adjacent sensing points, the small unit undergoes overall rigid body upward translation. During this upward translation process, the elastic energy storage element of the small unit does not participate in load absorption and diversion, but only through the rigid body upward translation of the small unit, it is forcibly transmitted to the adjacent fractal units in the same layer through the linkage transmission component, so as to maintain the path continuity of the topological force flow ripple in the same layer.

[0004] Beneficial effects: 1. This invention completely eliminates motion lock-up interference of rigid levers when three points are synchronously retracted by attitude synchronization gear transmission lever assembly, ensuring that the fractal network is free from jamming during multi-unit synchronous attitude adjustment. 2. This invention utilizes a guide shaft formed by extending a spring rod and a deep hole structure at the fulcrum to force the large and small units to achieve radial and axial coaxial alignment during the unloading and reset phase, without relying on spherical self-positioning, thereby completely eliminating the hidden danger of overlapping misalignment failure. 3. This invention clarifies the traffic splitting and coordination mechanism from single-point triggering to three-point triggering, improves the same-layer communication guarantee logic under the upward posture, and expands the working condition adaptability of the architecture. 4. This architecture is not only suitable for vehicle collision avoidance and industrial shock absorption, but can also be applied across fields through mapping, possessing extremely high industrial versatility and strategic value. Attached Figure Description

[0005] Figure 1 : This is a top-down view of the macroscopic hexagonal topological array of the present invention. Figure 2 : This is a planar projection of the macroscopic hexagonal topological array of the present invention. Figure 3 This is a partial detailed framework diagram of the macroscopic hexagonal topological array of the present invention. Figure 4 This is a frame diagram of the small unit assembly of the present invention, with detailed part numbers marked in the diagram. Figure 5 : This is a perspective view of the small unit assembly of the present invention, with detailed part numbers marked in the figure. Figure 6 This is an internal cross-sectional view of the small unit spring assembly of the present invention. The figure shows the spring cylinder assembly that provides tension and the spring cylinder assembly that provides pressure, both of which are marked with detailed part numbers. Figure 7 This is a framework diagram of the large unit assembly of the present invention, with detailed part numbers marked in the diagram. Figure 8 This is a perspective view of the main unit assembly of the present invention, with detailed part numbers marked in the figure. Figure 9 This is an internal cross-sectional view of the large unit spring assembly of the present invention. The spring cylinder assembly that provides pressure and the spring cylinder assembly that provides tension are shown in the figure, and both are marked with detailed part numbers. Figure 10 This is a schematic diagram of a fractal basic unit structure consisting of three small units and one large unit. Figure 11 : A schematic diagram of the structure for mounting a telescopic transmission lever assembly on a fractal base unit. Figure 12 :for Figure 11 The perspective view shown has detailed part numbers marked on it. Figure 13 : This is an exploded view of the three-position synchronous gear transmission lever assembly of the present invention. Figure 14 : This is an exploded view of the lever and assembly of the present invention, excluding the fulcrum. Figure 15 : This is a part numbering diagram of the lever assembly of the hinged small unit of the present invention. Figure 16 : This is a partial part numbering diagram of the two assemblies of the hinged size unit of the present invention. Figure 17: This is a diagram showing the reference numerals for the lever assembly of the hinged large unit of this invention. Figure 18 This is a schematic diagram of the gear sleeve structure of the present invention, showing the large and small double-hole plates, the large and small snap rings, and the large and small gear variable diameter sleeves. Figure 19 This is a schematic diagram showing the disassembled state of the large unit and three small units of the present invention; the base of the three small units is connected to a fulcrum, and the fulcrum is provided with a hole, which is inserted through the pressure spring of the large unit to reflect the forced guiding and cooperating relationship. Statement of Drawings: Some standard mechanical parts (such as springs, threaded connectors, etc.) in the drawings of this application are drawn using simplified methods conventional in the art; for components such as flexible limiting chain links, functional schematic drawings are used to show their structural form and connection relationship. Those skilled in the art can clearly identify the structural form of each part and its assembly relationship based on the textual description and the corresponding reference numerals of each component in this specification; therefore, the disclosure of the drawings in this invention is sufficient. Detailed Implementation

[0006] This application claims priority to a patent application filed on June 18, 2026, entitled 'A Load Scheduling Architecture Based on Fractal Topology,' the entire technical content of which is incorporated herein by reference. The basic fractal topology architecture in this application enjoys the aforementioned priority date; improvements such as newly added gear linkage components, guide shaft structures, and vertical rotational degrees of freedom in this application do not enjoy the aforementioned priority date, and are based on the filing date of this application. Terminology Definition To facilitate understanding of the technical solution of this invention, the following definitions are made for some of the core terms appearing in the text: Load sensing point: refers to the structural part located at the end of the power transmission arm of the fractal unit, which is used to directly bear external loads and trigger the linkage transmission component to generate a diversion or decoupling response. Linkage transmission assembly: refers to a mechanical connection assembly disposed between adjacent fractal units, used to forcibly and synchronously transmit the displacement generated by the load on the fractal unit to other fractal units. Its specific forms include, but are not limited to, sleeve-type telescopic rods and attitude synchronization gear transmission lever assemblies. Decentralized topological force flow ripple: refers to the fractal units forming a topological network through linkage transmission components. Without the intervention of a central controller, based on the independent threshold judgment of each node, the single point or local displacement signal is transmitted outward step by step, forming a radial mechanical response similar to the diffusion of water waves. Torque balance lock-up: This refers to a situation where the transmission arm of the fractal unit is subjected to specific external pressure (e.g., simultaneous pressure at three points), and the torques of each arm cancel each other out, causing the lever mechanism to be unable to generate internal relative deflection, thus placing it in a mechanically "hard-locked" state. Releasing this lock-up depends on the overall displacement generated by the forced intervention of the larger unit. Fractal hierarchy: refers to the multiple scale levels (such as small units, large units, etc.) that constitute the fractal unit. The fractal levels are connected by spring assemblies and guide structures, and the topological connections between adjacent nodes within the layer are achieved through linkage transmission assemblies. Example 1: Based on a prior application of a basic fractal architecture Reference Figures 1 to 9 The basic architecture of this invention includes small units and large units. Reference Figure 4 and Figure 5 The small unit is equipped with a small unit compression spring assembly (1), a small unit tension spring assembly (2), a small unit long limiting chain link (3), a small unit short limiting chain link (4), a small unit threaded fastening pin (5), a small unit threaded fastening pin screw, (5-1) a small unit three-pronged linkage lever (6), a small unit three-pronged lever fulcrum with axial centering function (7), and a small unit three-pronged mounting base (8). Reference Figure 7 and Figure 8 The large unit is equipped with a large unit compression spring assembly (9), a large unit tension spring assembly (10), a large unit long limiting chain link (11), a large unit short limiting chain link (12), a large unit threaded fastening pin (13), a large unit threaded fastening pin screw (13-1), a large unit three-pronged linkage lever (14), a large unit three-pronged lever fulcrum with axial centering function (15), and a large unit three-pronged mounting base (16). The small and large units are connected by overlapping hemispherical protrusions and recesses to achieve basic force flow transmission and attitude self-centering, and the attitude constraint of the mechanism under load is maintained by flexible limiting of long and short chain links. This basic architecture constructs a multi-level fractal topology network. In addition, as a reset and constraint guarantee for this basic fractal architecture, the threaded fastening pin (5) of the small unit and the threaded fastening pin (13) of the large unit are designed to have equal axial lengths to ensure that the long / short chain links connected to them have a unified travel limit and constraint reference when reset. like Figure 4 , Figure 5 , Figure 7 and Figure 8As shown, the flexible limiting chains (3, 4, 11, 12) are configured such that both ends of the chain are respectively sleeved or threaded onto the corresponding pins (5, 13), forming axial positioning and anti-disengagement constraints at the connection. Under load, the spring cylinder is compressed, and the chain is relatively relaxed; however, during the unloading and reset phase, due to the geometric constraint that "the shortest distance between two points is a straight line," the relaxed chain is forcibly straightened. Combined with the geometric interference and guidance of the intermediate contact point, this ensures that the chain accurately returns to the initial tensioned position, effectively preventing the spring cylinder from shifting or collapsing due to lateral movement of the chain. Furthermore, such as Figure 7 , Figure 8 As shown, after the three-pronged lever fulcrum (15) and the three-pronged mounting base (16) of the large unit are assembled, a coaxial central through hole is provided in the middle. This central through hole is adapted to the extended thin guide shaft (13), so that the large unit of the present invention is not only a load scheduling unit, but also serves as an "interface mother" for the next level of fractal expansion. Through this reserved hole, the structure clearly provides physical guidance and assembly basis for the hierarchical stacking and nesting of subsequent larger fractal units, fully demonstrating that the fractal structure of the present invention has the possibility of progressive expansion at the hierarchical scale. As an alternative to this basic fractal architecture, the prior application also describes the following variations: 1. Two-point lever linear array (N=2): Two load sensing points are arranged collinearly at 180° and connected by a sleeve-type telescopic rod with axial sliding fit, forming a one-dimensional linear load transmission network. Under load, the sleeve absorbs the displacement difference between the two ends through axial expansion and contraction, so that the force flow is transmitted one-dimensionally only in a straight line, forming the simplest linear implementation of fractal topology. Under this two-point linear arrangement, if it is necessary to simplify the structure and eliminate the unloading idle stroke, the aforementioned 'bidirectional tension and compression spring' can also be used to replace the spring cylinder in the modified example. Those skilled in the art will understand that the aforementioned two-point linear array constitutes the one-dimensional basic form of the fractal topology; the two-point array can also be further optimized by applying conventional anti-jamming deformation examples (such as bidirectional tension / compression springs) described in this application, based on the sleeve telescopic rod, according to actual working conditions, to improve its reset effect. These are all known equivalent implementations of this basic fractal architecture. 2. Replacement of single-unit compression spring with bidirectional spring: The single tension spring inside the sleeve-type compression spring tension assembly at each lower end is replaced with a bidirectional tension-compression spring. One end of this bidirectional tension-compression spring is hinged to the lower end of the three-pronged lever, and the other end is hinged to the base. It can withstand both compressive and tensile loads simultaneously, thus eliminating the need for the original short-stroke flexible limiting chain. Those skilled in the art can make adaptive selections in the above-described various embodiments according to actual load conditions. These variations are all known equivalent implementations of the basic fractal architecture of the present invention. Example 2: Attitude Synchronization Gear Transmission Lever Assembly Reference Figures 13 to 18 To solve the problem of rigid lever movement locking, this embodiment adds a linkage transmission component, the core of which is the attitude synchronization gear transmission lever assembly. The assembly includes: a small unit transmission gear (22) with an extended guide rod, a large unit transmission gear (33) with an extended guide rod, a gear pin sleeve fixing double hole plate (24, 34), a gear pin sleeve (25, 35), a snap ring (27, 36), a posture synchronization return spring (30, 32), a threaded head pin (26, 39), and locking screws (26-1, 39-1) for locking the threaded head pin. Reference Figure 12 The figure clearly shows the macroscopic assembly relationship of the attitude synchronization gear transmission lever assembly of the present invention. Specifically, 17 is the small unit gear transmission synchronization fulcrum, 18 is the integrated large and small unit gear transmission synchronization fulcrum, 19 is the small unit long lever arm telescopic guide rod, 20 is the small unit long lever arm telescopic guide rod with space avoidance function, 21 is the large unit long lever arm telescopic guide rod, and 40 is the small unit sleeve-type short lever arm. These components are connected to the core gear assembly through assembly, together forming the topological transmission backbone of the decentralized force flow ripples. The assembly process is as follows: Step 1: Gear assembly pre-assembly and pivot installation First, mesh the tooth surfaces of the transmission gears (22, 33); then insert the gear pin sleeves (25, 35) into the axial holes of the gears; then fit the gear pin sleeve fixing double hole plates (24, 34) onto both ends of the sleeve; finally, insert the snap rings (27, 36) into the corresponding snap ring grooves of the sleeve to lock the double hole plates, forming a rigid synchronization module. The rigid synchronization module is placed between the gear transmission synchronization fulcrums (17, 18), the pin holes on the fulcrums are aligned with the holes on the sleeve, the threaded pins (26, 39) are inserted, and the locking screws (26-1, 39-1) are screwed into both ends for axial locking. Step 2: Assemble the guide seat and spring limiter. On the gear extension guide rod (22, 33) that has extended to the fulcrum, first insert the inner guide limit seat (29, 37); then put on the attitude synchronization reset spring (30, 32); then screw the outer guide thread limit sleeve (23, 31) into the threaded section of the extension rod, compress the spring to the preset length and lock the limit. Step 3: Connecting the lever arm and locking it in both directions Select the corresponding lever arm according to the corresponding fractal unit level; among them, the small unit long lever arm (19), the small unit long lever arm (20) with avoidance function, or the small unit short lever arm (40) correspond to the small unit level; the large unit long lever arm (21) corresponds to the large unit level. Fit the corresponding lever arm into or connect it to the end of the extension guide rod, and align the hole on the lever arm with the threaded hole on the inner guide limit seat; then screw the locking screw (28, 38) through the lever arm hole into the threaded hole of the inner guide limit seat and lock it, forming a bidirectional limiting and guiding constraint on the spring. Its working mechanism is as follows: Under load, the transmission gears (22, 33) are assembled onto the support pin through a gear pin sleeve, and a clearance is reserved between the sleeve and the pin for the gear to slide laterally along the pin. Simultaneously, the lateral installation distance between the four pin holes on the gear transmission synchronization fulcrum (17, 18) is greater than the lateral width of the gear assembly, providing sufficient physical clearance for the aforementioned lateral sliding. When the three-pronged linkage lever moves under load, the transmission gears not only undergo rotational displacement to synchronize with adjacent gears, but also undergo lateral sliding to compensate for horizontal interference; simultaneously, an axial sliding fit is provided between the telescopic guide rod and the extension rod to compensate for length changes, and its end hinge point has a rotational degree of freedom to compensate for angular deflection. During this process, the attitude synchronization return springs (30, 32) are compressed, and under the dual constraint of the inner and outer guide limit seats (29 / 37 and 23 / 31), it is ensured that the springs deform only along the axis of the guide rod without wobbling. The rotational displacement and lateral sliding displacement of the gear, as well as the axial extension and angular deflection displacement of the telescopic guide rod, work together to forcefully and synchronously transmit the force flow to the adjacent fractal units. After the load is removed, it is precisely due to the precise constraint of the above-mentioned double limiting structure that the return springs (30, 32) can smoothly rebound along the predetermined axis, driving the assembly to accurately return to the initial preload position and eliminating the free travel and wobble during the reset process. As an optional anti-lock variant of this linkage transmission component under extreme counter-torque interference, the fulcrum structure of the gear double-hole plate assembly is provided with a mounting hole, which is configured as a stepped hole structure. Two circular plates with pin holes are embedded in the stepped hole, and the pins pass through the pin holes of the two circular plates and are locked by fastening screws. This constitutes a degree of freedom that allows the assembly to generate passive deflection in the vertical section with the midpoint of the line connecting the centers of the two pins as the axis of rotation. When the input attitude signals at both ends of the gear double-hole plate assembly are consistent, the normal meshing transmission of the gear pair constitutes the dominant transmission path of attitude rigid synchronization, and the vertical rotational degree of freedom does not substantially restrict or interfere with this dominant transmission path. When the input attitude signals at both ends conflict (e.g., one end is pressed down while the other end is lifted in the opposite direction externally), the gear pair will instantly lock rigidly due to the hard contact of the tooth surface, losing its rotational synchronization capability. At this time, the entire assembly transforms into a rigid linkage arm component. The vertical rotational degree of freedom provides a deflection fulcrum for this rigid arm, allowing the force flow to cross the gear lock-up area in a rigid lever transmission manner (its transmission output attitude is opposite in phase to the input end), forcibly transmitting it to the adjacent fractal unit, thus ensuring that the force flow ripples are not interrupted even under extreme signal conflict conditions. When implementing this, those skilled in the art can add conventional reset guide and limit structures to this passive deflection degree of freedom according to the actual assembly conditions and load conditions to maintain its initial reference and prevent excessive deflection. Example 3: Extended guide shaft and deep hole guide structure Reference Figure 9 , Figure 19 To address the issue of hemispherical overlap misalignment in the prior application, this embodiment extends the central tie rod (9-3) of the large unit compression spring assembly (9) into an extended thin guide shaft with axial positioning function. This extended thin guide shaft (composed of the extended portion of 9-3) passes sequentially through the triangular mounting base (8) of the small unit and then into the triangular lever fulcrum (7) of the small unit. The triangular lever fulcrum (7) of the small unit is machined with a coaxial depth guide hole adapted to the extended thin guide shaft. When the external load is removed and the small and large units are unloaded and reset, the extended thin guide shaft inserts into the depth guide hole of the small unit, forcing the small and large units to achieve radial and axial coaxial alignment. This eliminates the need for passive spherical self-positioning via the hemispherical boss and recess, thus completely eliminating off-center load interference caused by overlap misalignment. It should be noted that in actual implementation, the present invention currently only includes a two-layer fractal structure of small unit and large unit; the central through hole reserved on the three-pronged lever fulcrum (15) and the three-pronged mounting base (16) of the large unit is similar in form to the corresponding structure on the small unit, but the reserved hole exists only as an interface to reflect the possibility of progressive expansion of the fractal architecture of the present invention in terms of hierarchical scale, and is not a necessary component of the current two-layer structure. Example 4: Elastic Energy Storage Materials Furthermore, the elastic energy storage element in the smallest unit of this invention is not limited to the helical spring and sleeve combination described in this embodiment. In a preferred alternative of this invention, the elastic energy storage element or energy-absorbing medium can be a long-lasting elastic medium (such as a gas spring, hydraulic damper, rubber elastomer, magnetic spring, electromagnetic actuator, or programmable force feedback actuator), or a disposable energy-absorbing medium (such as foamed metal, honeycomb structure, crushable tube, brittle shell, or phase change material). Those skilled in the art can make adaptive substitutions among the above media according to the expected load strength and energy absorption requirements. Example 5: Mapping in the fields of computer resource scheduling, artificial intelligence computing, and digital twin simulation The decentralized force flow ripple transmission mechanism of this invention, along with the logic of "large unit forced participation in three-point torque decoupling" and "small unit rigid body lifting to maintain same-layer ripples", is not limited to the diversion of mechanical loads in physical space. In the field of computer and AI resource scheduling: physical fractal units are mapped to "computing nodes" in a distributed cluster, load sensing points are mapped to "node computing load," and linkage transmission components are mapped to "task migration channels." When a single node is overloaded, local load sharing is triggered; when three nodes are simultaneously overloaded, forming a "computing deadlock," the corresponding logic of the larger unit forces participation in "three-point torque balance decoupling," transforming the deadlock pressure into a global state synchronization switch and breaking the computing deadlock. Conversely, when a node is triggered by an external signal to perform a "lift-up" action (such as a state maintenance command), the node undergoes an overall logical displacement, forcibly reporting the task state to adjacent nodes at the same layer, ensuring that the connection of the distributed topology network is not interrupted. Through this mapping, the system can complete cascading migration without central scheduling. In the field of digital twin simulation scheduling: fractal units are mapped to "simulation subdomains or process nodes" in the digital twin model, load sensing points are mapped to "simulation computing power consumption or real-time state changes," and linkage transmission components are mapped to "simulation data flow channels." When a single point of simulation computing power overload triggers local shunting, or when three points of synchronization reach a threshold triggering forced decoupling of large units, the digital twin system can automatically cascade simulation tasks to adjacent subdomains like ripples in mechanics without relying on a central host. Furthermore, when the overall system state "rises," it forces synchronization of all subdomain states, ensuring that the digital twin model maintains high fidelity operation with extremely low latency. Example 6: Other Cross-Domain Extension Implementation Methods The underlying mapping of the decentralized ripple logic of the present invention is not limited to the above-mentioned fields. As an extended embodiment of the present invention, this architecture is also applicable to the following scenarios: 1. Mapping Implementation Method in Integrated Circuit (IC) Design: Physical fractal units are mapped to "multi-core processing units or logic blocks" within the chip; load sensing points are mapped to "core data / instruction flow congestion status, bus occupancy, or local power consumption / temperature status"; and linkage components are mapped to "communication buses or data paths of the chip's on-chip network (NoC)". When three cores or logic blocks simultaneously experience instruction congestion, bus deadlock, or local hotspot overload, the chip's underlying large-unit logic (equivalent to the on-chip network's routing arbitration mechanism) is triggered, forcing a global data path synchronization reset or power allocation, breaking the multi-core "computing power deadlock". In the "lift-up" scenario, where only inter-core signal synchronization needs to be maintained without large power scheduling, the "lifted" core unit forcibly sends synchronization status signals to adjacent cores, ensuring uninterrupted topology connectivity of the on-chip network (NoC), achieving low-latency decentralized signal synchronization without global clock intervention. 2. Printed Circuit Board (PCB) Design: Physical units are mapped to "power / signal routing nodes" on the PCB, load sensing points are mapped to "local current, voltage drop, and thermal status," and linkage components are mapped to "copper traces, vias, and planar layers." When three critical power supply nodes or heat dissipation areas experience overload simultaneously, the large-unit logic intervenes, triggering a reset of the global power distribution network or a reconfiguration of the heat dissipation path. If only signal synchronization needs to be maintained, the node transmits a forced output synchronization clock signal to adjacent nodes, ensuring the connectivity of the entire PCB topology network. In the 3.6G wireless RF field: fractal units are mapped to "RF antenna array nodes," load sensing points are mapped to "RF channel power or signal overload states," and linkage transmission components are mapped to "RF matching paths." When the three antenna nodes synchronously reach the RF overload threshold, the system triggers a global RF matching path synchronous reset based on large-unit logic. In the "uplift" condition where only beamforming is required and there is no high-power overload, the nodes achieve decentralized wireless beamforming adaptive tuning by forcibly outputting phase / frequency synchronization signals without causing signal interruption. 4. Hardware-in-the-Loop (HiL) Simulation Testing: Physical fractal units are mapped to "real hardware nodes" (such as ECUs or actuators) in the HiL bench. When three nodes experience real-time computation overload simultaneously, the large unit logic is forcibly intervened, forcibly migrating the massive computational tasks to globally idle nodes. Under the "lift-up" condition of state synchronization maintenance, the nodes only act as relays of synchronization signals to ensure that the data flow of the hardware test network is not interrupted. Example 7: Mapping in the Field of Microscopic Material Structure Fractal units are mapped to "grain units" within the material, and sensing points are mapped to "grain boundary stress / strain." When stress waves propagate through the material, decentralized dissipation is achieved through the coordinated displacement of the microstructure. When three grain boundaries simultaneously reach yield stress deadlock, the large unit logic (corresponding to the next level of lattice arrangement) forces decoupling, breaking the deadlock and prompting the lattice to rearrange to disperse the stress. In cases where only waveform transmission is required without energy absorption, the grains, through minute coordinated displacement, act like a rigid body lifting, ensuring the continuous propagation of stress waves within the same layer of grain network without fracture. Example 8: Computer-readable storage medium and decentralized load simulation method. Referring to claim 10, the present invention also provides a computer-readable storage medium storing computer-executable instructions. When the instructions are executed by a processor, a load scheduling simulation method based on fractal topology is implemented. In this simulation method, the fractal topology of the physical structure can be mapped to a multi-scale computational model. As a non-limiting preferred embodiment, the smallest scale unit can be mapped to a fine-grid computational domain (or micro-computational node), the large unit can be mapped to a coarse-grid computational domain (or macro-computational node) surrounding the fine-grid, and the physical linkage transmission components can be mapped to boundary condition transmission channels between adjacent computational domains or nodes. The simulation execution steps under the above mapping model are as follows: 1. Fractal unit model construction: In the simulation environment, the fractal physical topology structure described in claim 1 is mapped as a virtual topology cluster containing multiple distributed computing nodes; 3. Load monitoring and threshold determination: Map the load sensing points in the physical structure to the "real-time computing power load or simulation state variable" of each computing node, and set the shunting trigger threshold; 4. Event-driven load balancing response: When the load of any computing node reaches a preset threshold, the simulation software executes local load balancing logic to migrate part of the simulation tasks of that node to idle nodes; when three or more adjacent nodes reach the load threshold simultaneously, a global load balancing mode switch is executed to simulate the "large unit forced participation in three-point torque balance decoupling" in the physical structure, thereby realizing the overall reset of the simulation task state and load redistribution. 4. Decentralized ripple migration: The transmission mechanism of the physical linkage components is mapped as a "data bus or task migration channel between distributed nodes"; when a global switch is triggered, the simulation task, like a physical "force flow ripple", is forced to migrate outward step by step along the adjacent virtual topology network until the computing load of the entire simulation system reaches equilibrium. This simulation method employs a fully event-driven mechanism during computer simulation, eliminating the need for a central scheduler. In digital twin and AI computing power simulation scenarios, it can significantly reduce simulation latency caused by global queuing, achieving extremely low latency and decentralized response equivalent to physical structures. Those skilled in the art should understand that the above-described mesh mapping method is merely one implementation example of this simulation method in terms of discretization of a specific computational domain. Any implementation that can be mapped to distributed computing nodes using the fractal topology structure described in claim 1 and execute the decentralized scheduling logic of threshold determination, global switching, and cascading migration described above belongs to the equivalent simulation implementation of this embodiment. Example 9: Description of Load Scheduling Conditions Working principle overview The essence of the transmission of topological force flow ripples in this invention lies in the following: Under external load, the large and small unit spring assemblies of the fractal unit first undergo elastic deformation, which then drives the three-pronged linkage lever to generate a spatial attitude offset. This offset attitude constitutes the physical "load state signal" in the decentralized network. The linkage transmission assembly (i.e., the attitude synchronization gear transmission lever assembly) strives to ensure, to the greatest extent possible, that the attitude signal avoids rigid lock-up during the step-by-step relay transmission process, thereby maintaining the continuous and smooth diffusion of the decentralized topological force flow ripples. This is achieved through the multi-degree-of-freedom spatial coordination within the assembly (including the rotation of the gear pair, the axial extension and retraction of the sleeve, the lateral micro-clear clearance sliding of the pin shaft, and the horizontal rotational avoidance degree of freedom at the fulcrum configured to cope with possible signal conflicts or extreme interference). Throughout this process, regardless of whether the fractal unit is subjected to single-point, two-point, or three-point pressure, the large unit consistently participates in load distribution and attitude coordination. The core function of the large unit lies in its own mechanical structure and decoupling mechanism, which, during the transmission of topological force flow signals, minimizes the risk of rigid system lock-up caused by load concentration, multi-point interference, or drastic attitude changes, thereby ensuring the overall robustness of the force flow network. Operating Condition 1: Single-point pressure response (local flow division) When one end of a single three-pronged lever unit bears an external load, in the original "sleeve-type compression spring tension assembly" configuration, the spring assembly below that end does not participate in the work, while the other four spring assemblies (including the spring above the compressed end and the springs above and below the other two ends) work together to partially divert and absorb energy. If a variant configuration of bidirectional tension and compression springs is used, the bidirectional tension and compression spring below that end will also actively participate in compression or tension to absorb displacement. Operating Condition 2: Two-point pressure response (partial flow division) When both ends of the three-pronged lever unit are simultaneously subjected to external loads, in the original configuration, the spring assemblies below the two compressed ends do not participate in the work, while the remaining spring assemblies work together to absorb energy. In the variant configuration using bidirectional tension and compression springs, the bidirectional tension and compression springs below the two compressed ends will also actively participate in the work (compression state), working together with other cooperating springs to complete the partial flow diversion. Operating Condition 3: Three-point pressure and dead-point decoupling (forced intervention of large unit) When the three endpoints of the three-pronged lever unit simultaneously reach the preset compression threshold, the lever arms at the three endpoints form a "three-point torque balance dead point." In this physical state, the forces at the three endpoints of the three-pronged lever cancel each other out, the lever as a whole no longer undergoes internal relative deflection, the mechanism is in a hard-locked state, and the three-pronged support is forced to undergo overall rigid body translation. Under this condition, if the original configuration is used, the bottom spring assembly will not move due to the deadlock travel limitation; if a bidirectional tension / compression spring variant is used, the bottom bidirectional tension / compression spring will also not participate in the work due to displacement jamming. This dead point state is forcibly decoupled by the large unit participating in the three-point torque balance, breaking the torque deadlock through the overall action of the large unit, releasing the force flow and activating the subsequent topological ripple transmission. Operating Condition 4: Cooperative Flow Diversion Characteristics of Large Units Regardless of whether the measurement point of the fractal element is under single-point, two-point, or three-point compression, the large element will synchronously participate in the coordinated load diversion. The difference lies in the magnitude of the force flow participating in the diversion under different trigger points. Since each large element is itself an independent structure for load diversion, its participation degree dynamically and adaptively changes with the cumulative amount of locally triggered loads. Condition 5: Lifting of three adjacent points and maintaining communication on the same floor When a small cell in the same layer of the topology array is simultaneously subjected to an upward load attitude signal by three adjacent small cells, this upward signal is forcibly transmitted synchronously via the attitude synchronization gear transmission lever assembly. This forces the small cell to undergo a rigid body upward translation without internal spring compression and energy absorption. This upward translation, as a purely physical "attitude synchronization signal," is forcibly transmitted to other adjacent small cells in the same layer, thereby keeping the force flow ripple path in the same layer of the topology network completely unobstructed and preventing force flow interruption due to the node's failure to participate in energy absorption. Example 10: Extended Implementation of Multi-Lever Arm and Multi-Point Topological Arrangement The number of load sensing points and transmission arms in the fractal unit of the present invention is not limited to a three-point structure. As one topological extension implementation, the fractal unit can be arranged with N arms (where N≥2). In practical applications: 1. When N=2, the two points are collinearly arranged and connected by a sleeve-type telescopic rod, forming a one-dimensional linear transmission topology. 2. When N=3, the three points are arranged non-collinearly at 120°, forming a two-dimensional planar conductive topology; 3. When N≥4, more complex high-dimensional topological cooperative networks can be formed by adapting non-orthogonal arrangements with different angles (such as four-point X-shaped arrangement, six-point honeycomb arrangement, etc.). Regardless of which lever arm arrangement method is adopted, as long as there are no orthogonal collinear paths between the lever arms, and a collaborative transmission network is formed through the linkage transmission component or force transmission component of the present invention (the specific implementation method can refer to the sleeve-type telescopic rod in Embodiment 1, or the attitude synchronization gear transmission lever assembly in Embodiment 2), its decentralized force flow ripple transmission mechanism is completely consistent with the three-point structure of the present invention in terms of underlying topological logic, and should be regarded as an equivalent extended implementation of the present invention. Example 11: Supplementary Implementation Method for Topology Array Closure As an independent physical implementation method supplemented herein to ensure the integrity of priority, the multiple fractal units in this invention can be connected end-to-end or stacked through linkage transmission components or force transmission components (the specific implementation of which can refer to the sleeve-type telescopic rod in Embodiment 1, or the attitude synchronization gear transmission lever assembly in Embodiment 2), forming a force flow loop without physical boundaries or a repetitive arrangement, specifically including: 1. Two-point one-dimensional loop closure: In an arrangement of N=2, two fractal units are connected end to end through force transmission components to form a closed loop force flow transmission circuit; 2. Two-dimensional surface closure (boundary force flow coupling): Fractal units are arranged into a double or multi-layer structure. The force transmission arm at the end of each fractal unit extends outward and is connected to the corresponding unit endpoint on the other side through a force transmission connector (such as a rigid link) of a certain length, so as to realize cross-surface force flow coupling and feedback in two-dimensional space. 3. Three-dimensional spherical closure: Based on the curvature requirements of the spherical arrangement, by adjusting the installation height of the force transmission components on the unit support points, each fractal unit is made to fit onto the curved or spherical surface, forming a spherical array closure; 4. Polyhedral Closure (Three-Dimensional Volumetric Closure): Fractal units are arranged on the faces of regular or irregular polyhedra (such as rectangles, hexagonal prisms, octahedrons, or other complex spatial polyhedra), interconnected through corner nodes or boundary force transmission components to form a three-dimensional force flow loop throughout space. During implementation, the installation height of each force transmission component at the unit support point can be adjusted to match the spatial angles and node spacing of the polyhedron without changing the physical length of the force transmission components themselves. The above-mentioned various physical closure schemes can effectively eliminate the phenomena of force flow interception and stress reflection at the edge of the array by topological force flow ripples, so that the decentralized force flow ripples of the present invention form a closed-loop dissipation without bounds in the fully closed topological network. Example 12: Implementation of a Single-Layer Shielded Delocked Topology Network As a topological variant independent of fractal hierarchy, the macroscopic mechanical superstructure of the present invention can also be configured as a single-layer decentralized topological network, that is, omitting the fractal hierarchy of the large unit, and only forming a topological array by connecting multiple small units end to end in the plane. In this single-layer topology variant, to fundamentally prevent the network from entering a physical lock-up state due to three-point torque imbalance, one of the load sensing points of each small unit is set to a shielded state. The shielded state means that the spring assembly corresponding to the shielded load sensing point is physically configured not to directly contact the external load or participate in energy absorption response, so that when the small unit is subjected to external force, it can only bear the external load through the remaining unshielded load sensing points. Through this "single-point shielding" setup, when external impact loads act on this single-layer network, each small unit can at most trigger a localized flow-dividing state of "two-point pressure," failing to reach the physical condition of "three-point torque balance lock-up." This variant, while retaining the decentralized force flow ripple transmission network, minimizes the risk of multi-node deadlock with its minimalist structure and passive shielding mechanism, maintaining the continuous and uninterrupted operation of the topology network without the need for large unit intervention. Those skilled in the art should understand that the specific physical implementation of the "shielding state" described in Embodiment Twelve above is not limited to the conventional structural forms listed in this embodiment. Any equivalent alternative means (including but not limited to physical shielding, physical removal, isolation structures, stiffness adjustment, stroke limitation, or mechanical locking, etc.) that enables the spring assembly corresponding to a load sensing point in the fractal unit to be configured not to participate in energy absorption response or directly bear external load when an external load is applied, thereby preventing the fractal unit from forming a three-point torque balance lock, should be considered as equivalent embodiments of the present invention and fall within the protection scope of the present invention.

Claims

1. A macroscopic mechanical superstructure based on fractal topology, characterized in that, It includes at least one fractal unit; the fractal unit has self-similarity and contains a hierarchical structure of at least two scales, including small units and large units located at fractal levels adjacent to the small units; the large units provide the hierarchical expansion basis for the fractal structure of the fractal unit and are configured to forcibly participate in force flow to unlock when the torque balance lock occurs in the transmission arm of the fractal unit; both the small units and the large units are provided with load sensing points; when any one or more of the load sensing points sense a load exceeding a preset threshold, a diversion response is activated through the linkage transmission component; the small units are connected to each other, and the large units are connected to each other through the linkage transmission component within the corresponding level; The aforementioned linkage transmission component is used to form a topological transmission network between fractal nodes within a corresponding level; the linkage transmission component is configured to: when the small unit or the large unit is subjected to load and generates displacement, force the displacement to be synchronously transmitted to the adjacent small unit or the adjacent large unit through the linkage transmission component; the linkage transmission component releases or compensates for interference caused by displacement transmission conflicts or changes in spatial angle through one or more of axial extension, lateral sliding, or spatial attitude compensation, thereby activating decentralized topological force flow ripples; wherein, each fractal unit and its internal sub-units independently perform threshold determination and working mode switching, without the need for a central controller, which is a decentralized self-organizing behavior.

2. The macroscopic mechanical superstructure according to claim 1, characterized in that, When any one, two, or three of the multiple load sensing points simultaneously sense a load exceeding a preset threshold, the fractal unit transmits the load to the lower-level large unit through the linkage transmission component, and the large unit participates in the flow diversion and coordination; and when the three points are triggered, the large unit participates in completing the three-point torque balance decoupling and force flow diversion.

3. The macroscopic mechanical superstructure according to claim 1, characterized in that, The linkage transmission assembly includes an attitude synchronization gear transmission lever assembly; the attitude synchronization gear transmission lever assembly includes a transmission gear with an extended guide rod, a gear pin sleeve, a rigid synchronization plate, a telescopic guide rod, a return spring, an inner guide limit seat, and an outer guide thread limit sleeve; the transmission gear is mounted on a support pin shaft via the gear pin sleeve, and a clearance is reserved between the sleeve and the pin shaft for the gear to slide laterally along the pin shaft; the rigid synchronization plate rigidly locks the pin sleeves of adjacent transmission gears to ensure that the gears are coplanar and synchronized; the telescopic guide rod and the extended rod are provided with an axial sliding fit, and the transmission gear generates... During displacement, the axial sliding compensates for the length change; the end hinge point of the telescopic guide rod and the extension rod has a rotational degree of freedom to compensate for the angular deflection caused by the movement of the three-pronged linkage arm; the return spring is sleeved on the extension guide rod and is axially limited and guided by the inner guide limit seat and the outer guide thread limit sleeve; the rotational displacement and lateral sliding displacement of the transmission gear, as well as the axial telescopic displacement and angular deflection displacement of the telescopic guide rod, work together to force the force flow to be synchronously transmitted to the adjacent fractal unit when the three-pronged linkage arm is displaced, and the return spring provides the restoring force after the load is removed.

4. The macroscopic mechanical superstructure according to claim 1, characterized in that, The large unit compression spring assembly of the fractal unit is provided with a central pull rod with an extended thin guide shaft; the small unit's three-pronged linkage lever fulcrum is provided with a coaxial depth guide hole adapted to the extended thin guide shaft; when the small unit and the large unit are unloaded and reset, the extended thin guide shaft is inserted into the depth guide hole, forcing the large and small units to achieve radial and axial coaxial alignment, eliminating overlapping misalignment interference.

5. The macroscopic mechanical superstructure according to claim 1, characterized in that, When some of the multiple load sensing points simultaneously detect that the load exceeds a preset threshold, the fractal unit performs partial flow splitting; when all of the multiple load sensing points simultaneously detect that the load exceeds the preset threshold, the fractal unit performs flow splitting mode switching, which is done by the fractal unit moving or switching its state synchronously as a whole, and sending trigger signals to adjacent units through all load sensing points at the same time to activate topological force flow ripples, so that the load propagates outwards along the fractal hierarchy level by level.

6. The macroscopic mechanical superstructure according to claim 1, characterized in that, The topological connection methods between multiple fractal units include one-dimensional ring closure, two-dimensional surface closure, or three-dimensional volumetric closure, to eliminate the force flow interruption of the topological force flow ripples at the edge of the array.

7. The macroscopic mechanical superstructure according to claim 1, characterized in that, The fractal unit has N power transmission arms, where N≥2; each power transmission arm has a load receiving part at its end, so that both the small and large units of the fractal unit have the load receiving part, and each load receiving part is associated with an elastic energy storage element or an energy absorption medium; when N=2, the power transmission arms are arranged collinearly to form a one-dimensional linear conduction structure; when N=3, the power transmission arms are arranged non-collinearly at 120° to form a two-dimensional planar conduction structure; when N is any positive integer greater than or equal to 4, the power transmission arms are arranged non-orthogonally to form a multi-dimensional topological cooperative network.

8. The macroscopic mechanical superstructure according to claim 1, characterized in that, The fractal unit is applied in the fields of computer resource scheduling, artificial intelligence computing scheduling, or digital twin simulation scheduling; the load sensing point corresponds to the resource usage status of the computer system, the computing load status of the artificial intelligence model, or the real-time simulation status of the digital twin model; the local splitting corresponds to the scheduling of local computing resources or AI inference / simulation tasks; the overall motion or state synchronization switching corresponds to the synchronization switching of scheduling strategies or AI training / simulation tasks; the ripple propagation corresponds to the cascading migration of computing load or AI / digital twin computing tasks among distributed nodes.

9. The macroscopic mechanical superstructure according to claim 1, characterized in that, The fractal unit is applied to the field of microstructure materials; the load sensing point corresponds to the local stress or strain inside the material; the local flow distribution corresponds to the local dislocation or deformation of the microstructure unit; the overall motion or state synchronization switching corresponds to the cooperative displacement of the microstructure unit; and the ripple propagation corresponds to the stepwise diffusion of stress waves in the material.

10. A computer-readable storage medium storing computer-executable instructions, characterized in that, When the instruction is executed by the processor, the load scheduling simulation method based on fractal topology as described in claim 1 is implemented. The method is applied to the load simulation of artificial intelligence computing simulation or digital twin system. The method includes: mapping fractal units to distributed computing nodes, mapping load sensing points to the computing power load state of nodes, and mapping linkage transmission components to data transmission and task migration channels between nodes; performing threshold determination, partial splitting, and splitting mode switching steps through the mapped computing nodes; when the computing node reaches the splitting switching condition according to the computing power load state, forcibly migrating the computing task to an adjacent computing node through the data transmission and task migration channel to activate decentralized computing power ripple propagation.

Images