Industrial internet device running state evaluation method and system

Abstract

The application discloses an industrial internet equipment operation state evaluation method and system, and particularly relates to the field of operation state risk evaluation, which comprises the following steps: collecting operation state parameters in industrial equipment, marking the operation state parameters as operation state main variables, and recording the source equipment, measurement interface, sampling mode and update cycle of each operation state main variable; for each operation state main variable, extracting external input environment variables relied on by the operation state main variable, structurally registering the external input environment variables as state perception dependent items, and uniformly marking each state perception dependent item as an environment reference variable. By constructing the structural binding relationship between the operation state main variable and the environment reference variable, and forming a state structure binding chain atlas, the evaluation system can identify whether the reference structure is incorrectly inherited, drifted or replaced, thereby capturing the implicit judgment distortion problem caused by the semantic coordinate system error layer, and solving the core defect that the traditional method still produces systematic risk under the normal numerical condition.

Description

Technical Field

[0001] This invention relates to the field of operational status risk assessment technology, and more specifically, to a method and system for assessing the operational status of industrial internet devices. Background Technology

[0002] In the risk assessment of the operational status of industrial Internet devices, traditional methods generally rely on a combination of internal device status and external data feedback for judgment, and their core is based on threshold analysis and logical matching of operating parameters.

[0003] However, this risk assessment system contains a crucial premise that is widely overlooked: the "meaning" of all operating states is defined under a specific environmental reference system. Once this reference mechanism is misaligned, for example, due to incorrect sensor access, parameter inheritance failure, or upstream and downstream configuration drift, the system may use an incorrect environmental mapping source without its knowledge. In this case, although all operating state judgments may appear normal in terms of numerical and logical aspects, they are actually completed in an incorrect "semantic coordinate system," thus creating a system-wide judgment bias.

[0004] These types of problems do not manifest as faults or anomalies in the traditional sense and cannot be detected by standard mechanisms such as check codes, redundant calculations, or logical analysis. They are systemic hidden dangers at the infrastructure level. Summary of the Invention

[0005] To overcome the aforementioned deficiencies in the prior art, embodiments of the present invention provide an industrial internet device operation status assessment method and system. By constructing a structural binding relationship between the main variables of the operation status and environmental reference variables, and forming a state structure binding chain graph, the assessment system can identify whether the reference structure has been incorrectly inherited, drifted, or replaced, thereby capturing the implicit judgment distortion problem caused by semantic coordinate system misalignment, thus solving the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a method for evaluating the operating status of industrial internet devices, comprising:

[0007] After collecting the operating status parameters of industrial equipment, they are identified as operating status master variables, and the source device, measurement interface, sampling method and update cycle of each operating status master variable are recorded.

[0008] For each master variable of the operating state, extract the external input environment variables that it depends on, register the external input environment variables as state-aware dependencies, and uniformly identify each state-aware dependency as an environment reference variable; record the sensor path, measurement physical unit, sampling response delay and source configuration strategy content of each environment reference variable;

[0009] Establish a mapping relationship between runtime status master variables and environment reference variables, and encapsulate the dependency methods, unit logic, model version and triggering conditions in this mapping relationship into a binding parameter set;

[0010] Based on the runtime state master variables, environment reference variables, and binding parameter sets, a state structure binding chain graph is constructed, which serves as the initial structural input for subsequent semantic analysis and reference verification.

[0011] A reference structure consistency verification mechanism is established based on the state structure binding chain graph. This mechanism is used to identify whether the environment reference structure called during the evaluation process of the running state maintains version consistency and path stability.

[0012] In a preferred embodiment, the reference structure consistency verification mechanism includes:

[0013] All environmental reference variables in the state structure binding chain graph constructed in the first part are transformed into reference structures. The reference structures must contain information fields such as unit resolution method, boundary strategy, mapping structure and configuration source. Based on the current evaluation time, all reference structures called by the main variables of the running state are read and a reference structure call snapshot matrix is ​​generated. The reference structure call snapshot matrix is ​​used for subsequent structure consistency comparison.

[0014] For each pair of reference structures in the snapshot matrix, a structural comparison is performed. If there are version differences, path drifts, or structural breaks during the structural comparison, the call path is marked as a reference binding inconsistency path.

[0015] The runtime master variables with inconsistent reference binding paths are marked as semantically unstable nodes, and they are annotated in the state structure binding chain graph.

[0016] In a preferred embodiment, control behavior data and physical feedback data, starting from the main variable of the operating state, are extracted from the historical operation log of the equipment to construct an operating behavior data sequence group.

[0017] For each operational behavior data sequence in the operational behavior data sequence group, based on the state master variable change event recorded in the sequence, the control signal response record, feedback variable change process, and triggered response behavior result accompanying the change event are extracted and used as path nodes respectively. The nodes are then connected according to the actual occurrence time order of the events in the sequence to construct the state semantic behavior evolution path corresponding to the sequence.

[0018] In a preferred embodiment, for each path structure in the state semantic behavior evolution path, according to the path mode of starting from the state main variable change node, extending through the control signal and response behavior node to the feedback variable node, it is verified segment by segment whether a complete closed loop structure is formed in terms of temporal logic, response matching and causal coherence.

[0019] For the running state master variables identified as not forming a complete closed-loop structure in the state semantic behavior evolution path closure verification, the set of running state master variables that have been identified as semantic reference unstable nodes in the reference structure consistency check step is retrieved. The state semantic behavior evolution path corresponding to the running state master variable is located, and the path intersection between the path and the paths involved in the set of semantic reference unstable nodes is calculated in turn. If the path intersection result is not empty, the closure failure of the running state master variable is defined as being attributed to the inconsistency of the reference structure binding or the path drift phenomenon on which it depends. The running state master variable is identified as a semantic structure closure loop break node, and this node is submitted as input to the state semantic backtracking comparison process for subsequent identification of the reference path drift root cause of the semantic structure break.

[0020] In a preferred embodiment, for the runtime master variable identified as a semantic structure closed loop break node, the structural binding evolution process of the environmental reference variable on which it depends is tracked, and the change records of binding path, parsing logic, calling channel and configuration version at different time segments are extracted to generate a reference structure calling evolution trajectory diagram.

[0021] In the reference structure call evolution trajectory diagram, behavioral differences are identified for each structure evolution path. It is analyzed whether there are drift behaviors such as abrupt changes in structure definition, binding path drift, unit conversion logic breakage, and version switching anomalies. The time node and configuration source address corresponding to each type of drift behavior are recorded.

[0022] In a preferred embodiment, all identified structural drift behaviors are mapped back to the state-structure binding chain graph one by one according to the established structural mapping relationship between the main variables of the operating state and the environmental reference variables. During the mapping process, starting from each structural drift node, the path-level diffusion process triggered by it in the state-structure binding chain graph is traced to form a set of paths that propagate sequentially from the drift node to its associated nodes. Based on this, a staggered mapping diffusion chain is constructed. The diffusion chain is used to accurately describe the impact range and propagation direction of structural drift behavior on the operating state evaluation path.

[0023] State credibility structure annotation is performed on the path nodes covered in the misaligned mapping diffusion chain. These path nodes are identified as semantically aware dead zone nodes, and the annotation results are submitted as input to the state credibility hierarchical structure generation process for the final evaluation of the state credibility level.

[0024] In a preferred embodiment, for each operating state master variable, based on the established state structure binding chain graph, reference structure consistency verification results, state semantic behavior evolution path analysis results, and structure drift backtracking annotation results, the reference structure consistency score, behavior closed-loop stability score, and structure drift sensitivity score are comprehensively calculated to generate a three-dimensional credibility index set for the master variable.

[0025] The three-dimensional credibility index set of all operating state main variables is summarized and mapped into a multi-dimensional credibility matrix map; weighted scoring is performed based on the index distribution of each operating state main variable in the map to generate corresponding credibility level hierarchical labels.

[0026] In a preferred embodiment, for the main variable of the running state whose weighted score result in the three-dimensional credibility index set is lower than the preset credibility threshold, the control signal freezing, running state evaluation process interruption and path substitution switching operations are implemented based on its structural annotation status in the semantic perception dead zone node set and its corresponding path-level coverage in the misaligned mapping diffusion chain, so as to prevent the main variable of the running state from entering the subsequent stage of the running state evaluation process in the case of semantic undeterminability.

[0027] The credibility level hierarchical labels generated by weighted scoring using a three-dimensional credibility index set are used as structured feedback information and input into the operation status scheduling logic control flow. This is used to trigger the revision of semantic structure modeling rules, the adjustment of reference structure configuration parameters, and the dynamic updating of operation status risk assessment strategies. The feedback input guides the control flow to output the credibility calibration results and risk response instructions for the main variables of the operation status, forming an operation status risk assessment control closed loop based on semantic reference consistency and behavioral closed loop stability.

[0028] An industrial internet device operation status assessment system includes a master variable identification module, a reference dependency registration module, a mapping and binding construction module, a binding chain graph construction module, and a reference consistency verification module;

[0029] The main variable identification module is used to identify the operating status parameters of industrial equipment as operating status main variables after they are collected, and to record the source device, measurement interface, sampling method and update cycle of each operating status main variable;

[0030] The reference dependency registration module is used to extract the external input environment variables that each master variable of the running state depends on, register the external input environment variables as state-aware dependencies, and uniformly identify each state-aware dependency as an environment reference variable; for each environment reference variable, the sensor path to which it belongs, the physical unit of measurement, the sampling response delay and the source configuration strategy content are recorded.

[0031] The mapping and binding building module is used to establish the mapping relationship between runtime state master variables and environment reference variables, and to encapsulate the dependency method, unit logic, model version and triggering conditions in the mapping relationship into a binding parameter set;

[0032] The binding chain graph construction module constructs a state structure binding chain graph based on the runtime state master variable, environment reference variable, and binding parameter set. The state structure binding chain graph serves as the initial structural input for subsequent semantic analysis and reference verification.

[0033] The reference consistency verification module establishes a reference structure consistency verification mechanism based on the state structure binding chain graph. The reference structure consistency verification mechanism is used to identify whether the environment reference structure called during the evaluation process of the running state maintains version consistency and path stability.

[0034] The technical effects and advantages of this invention are as follows:

[0035] 1. By constructing a structural binding relationship between the main variables of the running state and the environmental reference variables, and forming a state structure binding chain graph, the evaluation system can identify whether the reference structure has been incorrectly inherited, drifted or replaced, thereby capturing the implicit judgment distortion caused by the misalignment of the semantic coordinate system, and solving the core defect of traditional methods that still produce systemic risks under normal numerical conditions;

[0036] 2. Construct a set of reference structures based on the binding chain graph, generate a snapshot matrix in each evaluation period, perform pairwise structure comparison, identify version inconsistencies, path drift and structural breakage, and mark the main variable of the problem as semantic reference instability node, so that the risk of structural anomalies can be captured and presented in a concrete way at an early stage.

[0037] 3. Extract the control behavior and feedback variable change sequence from historical logs, reconstruct the state semantic behavior evolution path, and construct a causal node structure diagram according to the order of event occurrence. This enables closed-loop logic verification of each state master variable at the physical execution layer, effectively determining whether there are response chain breaks or feedback misordering phenomena.

[0038] 4. Perform retrospective analysis on the semantic closed-loop fracture main variables, extract the structural evolution trajectory of the environmental reference variables on which they depend, identify structural drift nodes in unit conversion, path pointing or version inheritance, and construct a staggered mapping diffusion chain based on this, further identifying its influence radius and path-level ripple range on the overall evaluation chain.

[0039] 5. A three-dimensional credibility index set is constructed around the consistency of reference structure, the stability of behavioral closed loop, and the sensitivity of structural drift. This set is then integrated into a credibility scoring matrix. Combined with a density estimation function, a dynamic labeling layering mechanism with non-fixed thresholds is implemented, assigning a semantic stability level to each main variable and enhancing the sensitivity to fluctuations in the evaluation context. Attached Figure Description

[0040] Figure 1 The flowchart for state structure modeling and reference consistency verification of the present invention is shown.

[0041] Figure 2 This is a flowchart illustrating the state behavior evolution path construction and closed-loop verification process of the present invention.

[0042] Figure 3 This is a flowchart of the structural drift tracking and misaligned mapping diffusion chain construction process of the present invention.

[0043] Figure 4 This is a flowchart illustrating the closed-loop process for credibility assessment and risk control of the present invention.

[0044] Figure 5 This is a flowchart of the method steps of the present invention.

[0045] Figure 6 This is a system module diagram of the present invention. Detailed Implementation

[0046] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0047] Refer to the instruction manual appendix Figure 1-6 An embodiment of the present invention provides a method for evaluating the operational status of industrial internet devices, comprising:

[0048] After collecting the operating status parameters of industrial equipment, they are identified as operating status master variables, and the source device, measurement interface, sampling method and update cycle of each operating status master variable are recorded.

[0049] For each master variable of the operating state, extract the external input environment variables that it depends on, register the external input environment variables as state-aware dependencies, and uniformly identify each state-aware dependency as an environment reference variable; record the sensor path, measurement physical unit, sampling response delay and source configuration strategy content of each environment reference variable;

[0050] Establish a mapping relationship between runtime status master variables and environment reference variables, and encapsulate the dependency methods, unit logic, model version and triggering conditions in this mapping relationship into a binding parameter set;

[0051] Based on the runtime state master variables, environment reference variables, and binding parameter sets, a state structure binding chain graph is constructed, which serves as the initial structural input for subsequent semantic analysis and reference verification.

[0052] A reference structure consistency verification mechanism is established based on the state structure binding chain graph. The reference structure consistency verification mechanism is used to identify whether the environment reference structure called by the runtime state during the evaluation process maintains version consistency and path stability.

[0053] Furthermore, it needs to be clarified that in constructing the state structure binding chain graph, for each runtime state master variable, the environmental reference variable it depends on not only needs to establish its mapping relationship, but also needs to clarify the semantic mapping role played by each reference variable in data acquisition, parameter parsing, and configuration strategies. Therefore, the state structure binding chain graph should include the dependency path number, dependency granularity level, and version inheritance chain record for each master variable-reference variable path, so that in the subsequent semantic consistency analysis process, it can be traced back whether the judgment context corresponding to each state master variable originates from a unique and stable reference path structure. In practical applications, the structure of the above chain graph should support the characteristics of multi-path convergence, multi-version connectivity, and node traceability, ensuring that the graph is not only a static input, but also a semantic source graph for dynamic reference structure verification.

[0054] The reference structure consistency verification mechanism includes:

[0055] All environmental reference variables in the state structure binding chain graph constructed in the first part are transformed into reference structures. The reference structures must contain information fields such as unit resolution method, boundary strategy, mapping structure and configuration source. Based on the current evaluation time, all reference structures called by the main variables of the running state are read and a reference structure call snapshot matrix is ​​generated. The reference structure call snapshot matrix is ​​used for subsequent structure consistency comparison.

[0056] For each pair of reference structures in the snapshot matrix, a structural comparison is performed. If there are version differences, path drifts, or structural breaks during the structural comparison, the call path is marked as a reference binding inconsistency path.

[0057] The runtime master variables with inconsistent reference binding paths are marked as semantically unstable nodes, and they are annotated in the state structure binding chain graph.

[0058] Extract control behavior data and physical feedback data, starting from the main variable of the operating status, from the historical operation log of the equipment, and construct a sequence group of operating behavior data;

[0059] For each operational behavior data sequence in the operational behavior data sequence group, based on the state master variable change event recorded in the sequence, the control signal response record, feedback variable change process, and triggered response behavior result accompanying the change event are extracted and used as path nodes respectively. The nodes are then connected according to the actual occurrence time order of the events in the sequence to construct the state semantic behavior evolution path corresponding to the sequence.

[0060] For each path structure in the state semantic behavior evolution path, according to its path mode of starting from the state main variable change node, extending through the control signal and response behavior node to the feedback variable node, verify whether a complete closed loop structure is formed in terms of temporal logic, response matching and causal coherence.

[0061] For the running state master variables identified as not forming a complete closed-loop structure in the state semantic behavior evolution path closure verification, the set of running state master variables that have been identified as semantic reference unstable nodes in the reference structure consistency check step is retrieved. The state semantic behavior evolution path corresponding to the running state master variable is located, and the path intersection between the path and the paths involved in the set of semantic reference unstable nodes is calculated in turn. If the path intersection result is not empty, the closure failure of the running state master variable is defined as being attributed to the inconsistency of the reference structure binding or the path drift phenomenon on which it depends. The running state master variable is identified as a semantic structure closure loop break node, and this node is submitted as input to the state semantic backtracking comparison process for subsequent identification of the reference path drift root cause of the semantic structure break.

[0062] For the runtime master variables identified as semantic structure closed loop break nodes, the structural binding evolution process of the environmental reference variables they depend on is tracked, and the change records of binding paths, parsing logic, calling channels and configuration versions at different time segments are extracted to generate a reference structure call evolution trajectory diagram.

[0063] In the reference structure call evolution trajectory diagram, behavioral differences are identified for each structure evolution path. It is analyzed whether there are drift behaviors such as abrupt changes in structure definition, binding path drift, unit conversion logic breakage, and version switching anomalies. The time node and configuration source address corresponding to each type of drift behavior are recorded.

[0064] All identified structural drift behaviors are mapped back to the state-structure binding chain graph one by one according to the established structural mapping relationship between the main variables of the operating state and the environmental reference variables. During the mapping process, starting from each structural drift node, the path-level diffusion process triggered by it in the state-structure binding chain graph is traced to form a set of paths that propagate sequentially from the drift node to its associated nodes. Based on this, a staggered mapping diffusion chain is constructed. The diffusion chain is used to accurately describe the impact range and propagation direction of structural drift behavior on the operating state assessment path.

[0065] State credibility structure annotation is performed on the path nodes covered in the misaligned mapping diffusion chain. These path nodes are identified as semantically aware dead zone nodes, and the annotation results are submitted as input to the state credibility hierarchical structure generation process for the final evaluation of the state credibility level.

[0066] For each operating state master variable, based on the established state structure binding chain graph, reference structure consistency verification results, state semantic behavior evolution path analysis results, and structure drift backtracking annotation results, the reference structure consistency score, behavior closed-loop stability score, and structure drift sensitivity score are comprehensively calculated to generate a three-dimensional credibility index set for the master variable.

[0067] The three-dimensional credibility index set of all operating state main variables is summarized and mapped into a multi-dimensional credibility matrix map; weighted scoring is performed based on the index distribution of each operating state main variable in the map to generate corresponding credibility level hierarchical labels;

[0068] For each running state main variable V i Construct its credibility index set C(V) i )= <C1(V i ),C2(V i ),C3(V i The three indicators correspond to: C1(V) i C2(V) represents the reference structure consistency score; i C3(V) represents the behavioral closed-loop stability score; i () represents the structural drift sensitivity score;

[0069] First indicator C1(V) i ):

[0070] By measuring the main variable V of the running state i The degree of consistency at the structural level among the multiple reference structures on which it depends; if there are version drifts, differences in unit interpretation, or path divergences in the definition logic of the reference structures, it is considered a decrease in consistency;

[0071] The reference structure is defined as a five-element structure:

[0072] in This refers to the unit analytic function, which is used to map from the original measurement unit to the system's judgment unit, such as "kg / cm". 2 →kPa” mapping structure; For boundary constraint functions, boundary constraint functions are used to define the boundary strategies of the allowed range of parameters, such as hard limiting and gradual transition; For reference mapping logic structures, such as segment functions, multilinear or nonlinear mapping formulas; This indicates the configuration source fingerprint, which is used to identify the configuration source hash of this structure, such as a configuration file number or a remote distribution ID; The call path topology structure describes how data is transmitted from the collection point to the judgment logic, and it includes routing hierarchy and caching strategy.

[0073] Establish structural difference functions between any two reference structures. Constructing a structural difference metric:

[0074]

[0075] in This indicates the difference in nonlinear mappings between unit functions (such as the overlap of defined segments);

[0076] Structural differences between boundary strategies;

[0077] The difference in nested structure represents the logic of the mapping function;

[0078] Indicates the Hamming distance for configuring fingerprint;

[0079] Indicates the layer offset length between path topologies;

[0080] Where φ(·) is the normalization function, used to map the structural difference to the interval [0, 1);

[0081] The final reference structure consistency score is expressed as follows: The maximum structural difference value is used as the consistency gap, and the result after deducting the difference is the consistency score. The closer the value in the formula is to 1, the more consistent all the reference structures are.

[0082] Second indicator C2(V) i ):

[0083] C2(V i The main variable V is used to measure the running state.i In the semantic behavior path triggered in the device, whether a complete causal closed loop is formed and can be verified by physical feedback or control response;

[0084] Construct a state semantic behavior evolution path graph:

[0085] Where N i A set of nodes, the set of nodes includes For nodes that are in a state change state To control signal nodes, In response to the behavior node, For physical feedback nodes; E i The set of edges connects the above nodes and has temporal labels and causal attributes;

[0086] Determine the number of paths that form a closed loop:

[0087] Where γ is the path sequence, from δ(γ) = 1 indicates that the path satisfies the time monotonicity, response rationality, and feedback alignment; otherwise, it is 0.

[0088] Calculate the closed-loop stability score of the behavior:

[0089] in This indicates the total number of paths extracted. This represents the cumulative incomplete score caused by lost responses, out-of-order feedback, and incomplete logic across all broken paths; C2(V) i The higher the score, the more stable and verifiable the response chain of the behavioral logic structure.

[0090] The third indicator C3(V) i ): Evaluate V i The scope and intensity of the influence of the reference structure on other main variables after structural changes occur during operation;

[0091] Calculation of structural disturbance intensity per unit time:

[0092]

[0093] in This indicates that at time point t, the main variable V represents the running state. i The kth reference structure bound The torsional strength value of the structure obtained during structural stability analysis; μ represents the q-th component of the structure. qΨ(·) represents the structural transformation measure of the q-th dimension, such as the change of boundary function using piecewise dissimilarity; Ψ(·) represents the aggregate function of structural perturbation, with trend accumulation weighting. This distance is used to measure the intensity of change of the qth structural component between two time points t-1 and t. This distance is a non-Euclidean distance metric specific to the structural dimension.

[0094] Calculate the diffusion effect of structural disturbances in the structural bonded chains:

[0095]

[0096] in This indicates that at time point t, the running state is determined by the main variable V. i The bound reference structure The total impact intensity of the structural drift behavior that occurs on the system after it spreads to other master variable nodes in the entire state structure binding chain graph; Indicates and A set of master variables with binding relationships; d i,j ρ(d) represents the path depth in the binding chain graph; ρ(d) is the distance decay function; λ(V) represents the path depth in the binding chain graph. j This indicates a strong dependency factor of the reference structure on which the main variable depends;

[0097] Structural drift sensitivity score:

[0098] Where Θ(·) is a nonlinear response function. The higher the value in the formula, the wider the diffusion and the greater the influence. The final score is 1-Θ(·), which reflects the stability of "resistance to structural disturbance".

[0099] Constructing the credibility graph matrix M cred :

[0100]

[0101] The credibility graph matrix collects the three-dimensional credibility indices of all n operating status main variables into a unified structured matrix; each row of the credibility graph matrix represents a main variable, and each column represents its corresponding three credibility dimensions, which are used for subsequent weighted scoring and labeling.

[0102] Generate a comprehensive credibility score Ξ(V) i ):

[0103]

[0104] Among them Ξ(V) i ) is the main variable V representing the running state. i The overall credibility score; C1(V) i)·C2(V i C3 (V) is used to emphasize that the consistency of the reference structure and the stability of the behavioral closed loop are the main basis for scoring; i C3(V) represents the structural drift sensitivity score. i A higher value of ) indicates greater susceptibility to structural disturbances; 1-C3(V i ) represents inverse stability quantization, used to construct the exponential suppression factor; κ is the exponential scaling factor, used to control the nonlinear enhancement rate of drift effects; exp(-κ·(1-C3(V)) i Γ(·) is an exponential suppression function, which causes a significant drop in score when the risk of structural drift is high; Γ(·) is a nonlinear normalization mapping function, which restricts the overall score to the open interval (0,1);

[0105] The scoring reflects three points: if C1 or C2 approaches zero, it indicates extremely poor structural consistency or closed-loop stability, resulting in a very low score; if C3 approaches 1, it indicates extremely high drift sensitivity, and the score is quickly suppressed by exponential factors; if all three are excellent, the final score is close to 1.

[0106] Trust level labels are generated by combining non-fixed thresholds with density self-partitioning. Construct a rating set rating set The set of credibility scores for all main variables of running status serves as the input data space for label partitioning;

[0107] Construct the density estimation function:

[0108] Where K h (Ξ) represents the kernel density function constructed in the score space; The choice of kernel function includes Epanechnikov, triangular kernel, double exponential kernel, etc.; h is the bandwidth, which controls the smoothness of local density estimation; n is the total number of running state main variables; j is the subscript variable for summation, indicating that the j-th running state main variable is currently being traversed; Ξ is the point where the density is currently being estimated; Ξ(V j ) represents the credibility score of the j-th main variable representing the running state;

[0109] Construct a trust level label function:

[0110]

[0111] Where α1 and α2 represent K h The local minimum point of (Ξ) is adaptively determined, that is, the boundary is divided at the density trough; the label can be named: low, medium, high reliability, or used as an index of the instruction type for controlling the engine.

[0112] For the main variable of the running state whose weighted score in the three-dimensional credibility index set is lower than the preset credibility threshold, the control signal freezing, running state evaluation process interruption and path substitution switching operations are implemented based on its structural annotation status in the semantic perception dead zone node set and its corresponding path-level coverage in the misaligned mapping diffusion chain, in order to prevent the main variable of the running state from entering the subsequent stage of the running state evaluation process in the case of semantic indeterminacy.

[0113] The credibility level hierarchical labels generated by weighted scoring using a three-dimensional credibility index set are used as structured feedback information and input into the operation status scheduling logic control flow. This is used to trigger the revision of semantic structure modeling rules, the adjustment of reference structure configuration parameters, and the dynamic update of operation status risk assessment strategies. The feedback input guides the control flow to output the credibility calibration results and risk response instructions for the main variables of the operation status, forming an operation status risk assessment control closed loop based on semantic reference consistency and behavioral closed loop stability.

[0114] Furthermore, it should be noted that when implementing closed-loop control for risk assessment based on a set of credibility indicators, for operational status main variables whose credibility indicator values ​​are below a set threshold, control measures should not be triggered solely based on the scoring results. Instead, the risk level of potential misjudgment on the overall operational assessment process should be determined by considering the main variable's position weight in the state structure binding chain graph, path coverage density, and propagation radius in the staggered mapping diffusion chain. If the main variable has a structural role that provides semantic reference to multiple other key nodes, the risk response scope needs to be expanded by simultaneously issuing freeze commands or parameter rollback operations to its downstream path nodes. This ensures that the risk assessment strategy can effectively suppress the spread of misjudgment and maintain the controllability and steady-state judgment capability of the entire assessment chain before the semantically undeterminable region is isolated.

[0115] An industrial internet device operation status assessment system includes a master variable identification module, a reference dependency registration module, a mapping and binding construction module, a binding chain graph construction module, and a reference consistency verification module;

[0116] The main variable identification module is used to identify the operating status parameters of industrial equipment as operating status main variables after they are collected, and to record the source device, measurement interface, sampling method and update cycle of each operating status main variable;

[0117] The reference dependency registration module is used to extract the external input environment variables that each master variable of the running state depends on, register the external input environment variables as state-aware dependencies, and uniformly identify each state-aware dependency as an environment reference variable; for each environment reference variable, the sensor path to which it belongs, the physical unit of measurement, the sampling response delay and the source configuration strategy content are recorded.

[0118] The mapping and binding building module is used to establish the mapping relationship between runtime state master variables and environment reference variables, and to encapsulate the dependency method, unit logic, model version and triggering conditions in the mapping relationship into a binding parameter set;

[0119] The binding chain graph construction module constructs a state structure binding chain graph based on the runtime state master variable, environment reference variable, and binding parameter set. The state structure binding chain graph serves as the initial structural input for subsequent semantic analysis and reference verification.

[0120] The reference consistency verification module establishes a reference structure consistency verification mechanism based on the state structure binding chain graph. The reference structure consistency verification mechanism is used to identify whether the environment reference structure called during the evaluation process of the running state maintains version consistency and path stability.

[0121] It is important to explain the overall nature of this solution. In traditional operational status assessment, the evaluation system relies on the numerical performance of operational status parameters and combines external feedback data for logical analysis and risk decision-making. However, this method implicitly assumes a key premise: the "meaning" of all status parameters is defined based on a specific environmental reference system. Once this reference structure drifts, becomes misaligned, has inheritance errors, or path substitutions, the system may make semantically incorrect judgments based on perfectly normal numerical values. Therefore, to address the extremely hidden and dangerous system-level hazard of "misaligned status judgment context" in industrial systems, this solution introduces a structured verification mechanism for the consistency of operational status semantic references and a tiered reliability assessment model.

[0122] This scheme establishes a structural binding relationship between the runtime state master variable and its dependent environmental reference variable, and further constructs a "state structure binding chain graph" as the input basis for semantic judgment. In this structure, not only are the source devices, sampling methods and update frequencies of the runtime state parameters recorded, but also key information such as the collection path, unit logic, parsing delay and configuration source of the associated environmental reference variables are tracked. The dependency relationship is concretized by binding parameter sets to ensure that the "semantic coordinate system" of each runtime state master variable can be explicitly represented and verified.

[0123] At the implementation level, the first step of this method is to construct a state structure binding chain graph to achieve full structural registration between internal device variables and external reference semantics. Then, a reference structure consistency verification mechanism is established to extract and compare the structure snapshots of all reference structures called at the current evaluation moment, identify version conflicts, path drifts or structural breaks, and mark the discovered abnormal paths as "semantic reference instability nodes". The identification of these nodes provides basic semantic instability prompts for subsequent evaluations.

[0124] Based on this, the method introduces the semantic behavior evolution path analysis of the running state. By extracting the control behavior data and feedback data triggered by each running state master variable in the historical running log, the causal path is reconstructed and it is determined whether a complete closed loop structure is formed. If a structural break is found during the closed loop verification process, and the break path intersects with the path of the "semantic reference instability node", it is determined that the semantic closed loop failure of the master variable is caused by the disorder of the reference binding structure, marked as the "semantic structure closed loop break node", and submitted to the structural drift backtracking comparison mechanism.

[0125] The structural drift backtracking mechanism further tracks the environmental reference variables on which these main variables depend, analyzes the structural binding changes at different time segments, examines whether there are drift behaviors such as abrupt changes in unit definition, path changes, and version switching anomalies, and draws a reference structure call evolution trajectory diagram. In the graph, all drift behavior nodes are projected back into the state structure binding chain graph according to the structural mapping path, forming a staggered mapping diffusion chain, identifying the scope of its impact propagation, and marking path-level risk areas as "semantic awareness dead zone nodes".

[0126] Based on the above analysis results, this scheme introduces a three-dimensional credibility index modeling process. Semantic stability scores are calculated for each operational state main variable from three dimensions: "reference structure consistency," "behavioral closed-loop stability," and "structural drift sensitivity." The three-dimensional scores of all main variables are aggregated to construct a "state credibility multidimensional matrix map," and dynamic weighted scoring is performed based on the distribution to form credibility level labels. Unlike the traditional fixed threshold stratification method, this method uses a score distribution analysis based on kernel density function to automatically identify distribution troughs as level division boundaries, achieving a more adaptive credibility label division.

[0127] Finally, if a main variable whose credibility index score is lower than the set threshold is also structurally labeled in a semantic perception dead zone node and has a large path-level propagation radius in the misaligned mapping diffusion chain, a risk response mechanism will be triggered. This mechanism includes freezing control signals, interrupting the evaluation process, or switching paths to prevent the main variable from continuing to enter the operational status evaluation process in a semantically undecidable state. At the same time, the credibility level hierarchical label is input as structured feedback information into the operational status scheduling control logic to guide the revision of semantic structure modeling rules, the dynamic adjustment of reference structure configuration strategies, and the real-time update of operational status risk assessment strategies. This constitutes a state credibility control closed loop based on semantic reference consistency and behavioral response closed loop.

[0128] In summary, the reason why this solution adopts a complete process from structural binding chain graph, reference consistency verification, behavioral closed-loop construction, structural drift backtracking, three-dimensional credibility assessment to risk control closed loop is that traditional assessment methods are difficult to cover the stability of the "semantic support structure" of operational status judgment. This structural instability is precisely the most difficult to detect, the most prone to misjudgment, and the most risky hidden fault root cause. By constructing this process-oriented structural reasoning system, this solution achieves comprehensive modeling, real-time verification, and credible hierarchical control of the semantic basis of operational status judgment, ensuring the accuracy, reliability, and controllability of the assessment logic in complex industrial contexts from the source.

[0129] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An industrial internet device running state evaluation method, comprising: After collecting the running state parameters in the industrial equipment, identify them as running state main variables, and record the source equipment, measurement interface, sampling method and update cycle of each running state main variable; Characterized in that: For each running state main variable, extract its dependent external input environment variable, register the external input environment variable as a state perception dependency, and uniformly identify each state perception dependency as an environment reference variable; record the sensor path, measurement physical unit, sampling response delay and source configuration strategy content of each environment reference variable; Establish the mapping relationship between the running state main variable and the environment reference variable, and encapsulate the dependency mode, unit logic, model version and trigger condition in the mapping relationship as a binding parameter set; Based on the running state main variable, the environment reference variable and the binding parameter set, a state structure binding chain graph is constructed, which is used as the initial structure input for subsequent semantic analysis and reference checking; Based on the state structure binding chain graph, a reference structure consistency checking mechanism is established, which is used to identify whether the environment reference structure called by the running state in the evaluation process remains version consistent and path stable; The reference structure consistency checking mechanism includes: All environment reference variables in the state structure binding chain graph are converted into reference structure bodies, which need to include unit analysis method, boundary strategy, mapping structure and configuration source information fields; based on the current evaluation time, read all reference structure bodies called by the running state main variables, and generate a reference structure call snapshot matrix, which is used for subsequent structure consistency comparison; Each pair of reference structure bodies in the reference structure call snapshot matrix is compared, and if there is a version difference, path drift or structure break, the call path corresponding to the compared reference structure body pair in the reference structure call snapshot matrix is marked as a reference binding inconsistency path; Mark the running state main variables with reference binding inconsistency paths as semantic reference instability nodes, and annotate them in the state structure binding chain graph; Extract the control behavior data and physical feedback data starting from the running state main variables from the device historical running log to construct a running behavior data sequence group; For each running behavior data sequence in the running behavior data sequence group, according to the state main variable change event recorded in the sequence, extract the control signal response record, feedback variable change process and response behavior result triggered by the change event as path nodes, and connect the nodes of the state main variable change event, the nodes of the control signal response record, the nodes of the feedback variable change process and the nodes of the response behavior result in the sequence according to the actual occurrence time sequence of the events in the sequence, to construct a state semantic behavior evolution path corresponding to the sequence.

2. The industrial internet device running state evaluation method of claim 1, characterized in that: For each path structure in the state semantic behavior evolution path, verify whether a complete closed loop structure is formed in terms of timing logic, response matching and causal consistency according to the path way starting from the state main variable change node, extending through the control signal and response behavior node to the feedback variable node; For the running state main variable identified as not forming a complete closed loop structure in the state semantic behavior evolution path closed loop verification, call the set of running state main variables identified as semantic reference instability nodes in the reference structure consistency checking link, locate the state semantic behavior evolution path corresponding to the running state main variable, and calculate the path intersection between the path and the path involved by the set of semantic reference instability nodes in turn; if the path intersection result is not empty, define that the closed loop failure of the running state main variable is due to the inconsistent reference structure binding or path drift phenomenon it depends on, identify the running state main variable as a semantic structure closed loop fracture node, and submit the semantic structure closed loop fracture node as input to the state semantic backtracking comparison process for subsequent identification of the reference path drift root cause causing the semantic structure fracture.

3. The industrial internet device running state evaluation method according to claim 2, characterized in that: For the running state main variable identified as a semantic structure closed loop fracture node, track the structure binding evolution process of the environmental reference variable it depends on, extract the change records of the binding path, analysis logic, calling channel and configuration version at different time segments, and generate a reference structure calling evolution track diagram; In the reference structure calling evolution track diagram, identify the behavior difference of each structure evolution path, analyze whether there is a drift behavior such as structure definition way mutation, binding path drift, unit conversion logic fracture, version switching anomaly, and record the time node and configuration source address corresponding to each type of drift behavior.

4. The industrial internet device running state evaluation method according to claim 3, characterized in that: Map all identified structure drift behaviors back to the state structure binding chain graph according to the structure mapping relationship between the running state main variable and the environmental reference variable; in the mapping process, take each structure drift node as the starting point, track the path level diffusion process it triggers in the state structure binding chain graph, form a path set that propagates from the drift node to its associated nodes in turn, and build a cross-layer mapping diffusion chain according to this, which is used to accurately describe the impact range and propagation direction of the structure drift behavior on the running state evaluation path; Perform state credibility structure labeling on the path nodes covered in the cross-layer mapping diffusion chain, identify these path nodes as semantic perception dead zone nodes, and submit their labeling results as input to the state credibility hierarchical structure generation process for the final evaluation of the running state credibility level.

5. The industrial internet device running state evaluation method according to claim 4, characterized in that: For each operating state main variable, based on the established state structure binding chain map, reference structure consistency checking results, state semantic behavior evolution path analysis results, and structure drift backtracking annotation results, the reference structure consistency score, behavior closed loop stability score, and structure drift sensitivity score are calculated to generate a three-dimensional credibility index set for the main variable; The three-dimensional credibility index sets of all operating state main variables are collected and mapped into a state credibility multi-dimensional matrix map. According to the index distribution of each operating state main variable in the map, a weighted score is performed to generate a corresponding credibility level hierarchical label.

6. The industrial internet device operating state evaluation method according to claim 5, characterized in that: For the operating state main variable whose three-dimensional credibility index set weighted score result is lower than the preset credibility threshold, combined with its structure annotation state in the semantic perception dead zone node set, and based on its corresponding path level coverage range in the wrong layer mapping diffusion chain, control signal freezing, operating state evaluation process interruption and path replacement switching operation are implemented to block the operating state main variable from entering the subsequent stage of the operating state evaluation process under the semantic indeterminacy condition; The credibility level hierarchical label generated by the three-dimensional credibility index set weighted score is input into the operating state scheduling logic control process as structured feedback information, which is used to trigger the revision action of the semantic structure modeling rule, the adjustment execution of the reference structure configuration parameter, and the dynamic update of the operating state risk evaluation strategy; The structured feedback information input guides the control process output to generate the credibility calibration result and risk response instruction of the operating state main variable, forming an operating state risk evaluation control closed loop based on semantic reference consistency and behavior closed loop stability.

7. An industrial internet device operating state evaluation system, comprising a main variable identification module, a reference dependence registration module, a mapping binding construction module, a binding chain map construction module, and a reference consistency checking module, characterized in that: The main variable identification module is used to identify the operating state parameters collected from the industrial equipment as operating state main variables, and record the source equipment, measurement interface, sampling method and update period of each operating state main variable; The reference dependence registration module is used to extract the external input environment variables that each operating state main variable depends on, structure register the external input environment variables as state perception dependencies, and uniformly identify each state perception dependency as an environmental reference variable. Record the sensor path, measurement physical unit, sampling response delay and source configuration strategy content of each environmental reference variable; The mapping binding construction module is used to establish the mapping relationship between the operating state main variable and the environmental reference variable, and encapsulate the dependence mode, unit logic, model version and trigger condition in the mapping relationship as a binding parameter set; The binding chain map construction module constructs a state structure binding chain map based on the operating state main variable, the environmental reference variable and the binding parameter set. The state structure binding chain map is used as the initial structure input for subsequent semantic analysis and reference checking; The reference consistency checking module establishes a reference structure consistency checking mechanism based on the state structure binding chain atlas, and the reference structure consistency checking mechanism is used to identify whether the environment reference structure called by the running state in the evaluation process remains version consistent and path stable; The reference structure consistency checking mechanism comprises: All environment reference variables in the state structure binding chain atlas are converted into reference structures, and the reference structures need to include information fields of unit analysis method, boundary strategy, mapping structure and configuration source; taking the current evaluation time as a benchmark, all reference structures called by the main variables of the running state are read, and a reference structure call snapshot matrix is generated, which is used for subsequent structure consistency comparison; Structure comparison is performed on each pair of reference structures in the reference structure call snapshot matrix, and in the structure comparison, if there is a version difference or path drift or structure break, the call path corresponding to the compared pair of reference structures in the reference structure call snapshot matrix is marked as a reference binding inconsistent path; The main variables of the running state with the reference binding inconsistent path are marked as semantic reference instability nodes, and are marked in the state structure binding chain atlas; Control behavior data and physical feedback data starting from the main variables of the running state are extracted from the device historical running log to construct a running behavior data sequence group; For each running behavior data sequence in the running behavior data sequence group, according to the state main variable change event recorded in the sequence, the control signal response record, the feedback variable change process and the triggered response behavior result accompanied by the change event are extracted as path nodes, and the nodes of the state main variable change event, the nodes of the control signal response record, the nodes of the feedback variable change process and the nodes of the response behavior result are connected in the actual occurrence time sequence of the events in the sequence to construct a state semantic behavior evolution path corresponding to the sequence.

Images