Intelligent equipment function reliability verification method and device, equipment and storage medium

Abstract

The invention relates to an intelligent equipment function reliability verification method and device, equipment and a storage medium, and the method comprises the steps: obtaining operation environment parameters and equipment state data of intelligent equipment, and carrying out the construction evaluation of a redundancy verification architecture, and obtaining an initial verification reference; acquiring real-time operation data of the intelligent equipment, and performing function association analysis on the real-time operation data and the initial verification reference to obtain function abnormality features; acquiring state information of associated equipment of the intelligent equipment, performing cross validation on the state information of the associated equipment according to the function abnormality characteristics, and performing fault analysis on the state information of the associated equipment and the function abnormality characteristics to obtain an evolution fault injection scheme; and carrying out abnormal behavior identification on the real-time operation data to obtain an equipment repair requirement, and carrying out feasibility evaluation on the equipment repair requirement based on the function abnormal characteristics to obtain a dynamic repair strategy. According to the invention, a comprehensive guarantee mechanism can be provided for reliable operation of the intelligent equipment.

Description

Technical Field

[0001] The present invention relates to the technical field of smart devices, and in particular to a method, apparatus, device and storage medium for verifying the functional reliability of a smart device. Background Art

[0002] With the rapid development of intelligent technology, intelligent devices have been widely used in industrial production, daily life and other fields. The reliability of intelligent devices is directly related to the safe operation and performance of the entire system, so it is particularly important to verify the functional reliability of intelligent devices. At present, the reliability verification methods of intelligent devices mainly focus on single-dimensional performance indicators or functional tests. For example, only the basic functions or certain specific parameters of the device are tested, while ignoring the comprehensive influence of multi-dimensional factors such as the device operating environment and the status of related devices. This single verification method is difficult to fully reflect the various problems that intelligent devices may encounter during actual operation, which can easily lead to a large deviation between the verification results and the actual operating conditions, and cannot effectively guarantee the reliability of intelligent devices. Summary of the Invention

[0003] The main purpose of the present invention is to provide a method, device, equipment and storage medium for verifying the functional reliability of smart devices, which can provide a comprehensive guarantee mechanism for the reliable operation of smart devices and improve the overall reliability level of the devices.

[0004] To achieve the above objectives, the present invention provides a method for verifying the functional reliability of an intelligent device, comprising: Obtain the operating environment parameters and device status data of the smart device, and conduct a redundant verification architecture construction assessment to obtain the initial verification benchmark; Collecting real-time operating data of the smart device, performing functional correlation analysis with the initial verification benchmark, and obtaining functional abnormality characteristics; Acquire associated device status information of the smart device, cross-validate the associated device status information based on the functional abnormality characteristics, and perform fault analysis with the functional abnormality characteristics to obtain an evolutionary fault injection solution; Identifying abnormal behavior of the real-time operation data to obtain equipment repair requirements, and conducting a feasibility assessment of the equipment repair requirements based on the functional abnormality characteristics to obtain a dynamic repair strategy; The evolutionary fault injection scheme and the dynamic repair strategy are integrated and analyzed to obtain a reliability verification report.

[0005] Furthermore, the acquisition of operating environment parameters and device status data of the smart device and the evaluation of redundant verification architecture construction to obtain an initial verification benchmark include: Collecting temperature data, humidity data, vibration data, voltage data, and current data of the smart device, and performing multi-dimensional parameter integration to obtain the operating environment parameters; Sampling the state of the smart device according to the operating environment parameters to obtain the device state data; Performing cluster identification on the device status data to obtain status node data; Performing redundant data node construction on the state node data to obtain a multi-dimensional verification node matrix; Constructing a verification architecture topology structure according to the multi-dimensional verification node matrix to obtain a redundant verification architecture; A reliability evaluation calculation is performed on the redundant verification architecture to obtain the initial verification benchmark.

[0006] Furthermore, the real-time operating data of the smart device is collected and functional correlation analysis is performed with the initial verification benchmark to obtain functional abnormality characteristics, including: Performing data normalization processing on the real-time operation data to obtain normalized operation data; performing dynamic benchmark matching processing on the normalized operating data according to the initial verification benchmark to generate a benchmark deviation feature; Performing multi-dimensional feature decoupling on the baseline deviation feature to obtain potential abnormal components; Performing pattern density clustering on the potential abnormal components to obtain an abnormal feature cluster set; Performing functional association on the abnormal feature cluster set to obtain core abnormal feature data; Dynamic weight allocation is performed based on the core abnormal feature data to obtain functional abnormality features.

[0007] Furthermore, the acquiring of the associated device status information of the smart device, cross-verifying the associated device status information according to the functional abnormality characteristics, and performing fault analysis with the functional abnormality characteristics to obtain an evolving fault injection scheme includes: Scan the device connection relationship address of the smart device to obtain the collection address; Performing associated data collection on the associated device status information according to the collection address to obtain an associated status data set; Performing cross-validation calculation on the functional abnormality features according to the associated state data set to obtain a device interaction influence matrix; Performing an abnormal mapping operation on the device interaction influence matrix to obtain an abnormal transmission link; Classify and identify the functional abnormality features according to the abnormal transmission link to obtain an abnormal pattern recognition result; Performing statistical analysis on the abnormal pattern recognition results to obtain abnormal occurrence frequency distribution; Analyzing the anomaly propagation law of the anomaly occurrence frequency distribution to obtain anomaly evolution characteristics; Performing conditional analysis on the abnormal evolution characteristics to obtain a fault injection condition set; A fault test sequence scheme is constructed according to the fault injection condition set to obtain the evolutionary fault injection scheme.

[0008] Furthermore, the conditional analysis of the abnormal evolution characteristics is performed to obtain a fault injection condition set, including: Segmenting the abnormal evolution characteristics to obtain a conditional path spectrum; Calculating the state transition entropy value based on the conditional path spectrum to obtain a critical trigger condition set; Identifying bifurcation points on the critical trigger condition set to obtain a critical trigger domain; Performing dynamic constraint solving of fault parameters according to the critical triggering domain to obtain a fault parameter constraint set; A multi-objective optimization combination configuration is performed on the fault parameter constraint set to obtain the fault injection condition set.

[0009] Furthermore, the abnormal behavior identification is performed on the real-time operation data to obtain equipment repair requirements, and the feasibility assessment of the equipment repair requirements is performed based on the functional abnormality characteristics to obtain a dynamic repair strategy, including: Performing abnormal behavior identification on the real-time operation data to obtain abnormal behavior characteristics; Constructing a behavior deviation matrix based on the abnormal behavior characteristics and performing threshold analysis to obtain abnormal behavior determination information; Calculating equipment repair requirements based on the abnormal behavior determination information to obtain the equipment repair requirements; Performing resource availability assessment on the equipment repair requirements based on the functional abnormality characteristics to obtain repair resource conditions; Calculating the repair cost of the repair resource conditions to obtain repair priority information; Constructing a plan based on the repair priority information to obtain repair execution measures; Verifying the repair execution measures to obtain repair effect evaluation parameters; The repair execution measures are dynamically repaired according to the repair effect evaluation parameters to obtain the dynamic repair strategy.

[0010] Furthermore, the integrated analysis of the evolutionary fault injection scheme and the dynamic repair strategy to obtain a reliability verification report includes: Classifying the fault types of the evolutionary fault injection scheme to obtain a fault type matrix; Performing a quantitative evaluation of the repair effect of the dynamic repair strategy according to the fault type matrix to obtain a repair effect score; Optimizing and analyzing the repair effect scores to obtain repair strategy optimization parameters; Calculating the repair coverage of the evolutionary fault injection scheme according to the repair strategy optimization parameters to obtain a repair coverage index; Performing a time series correlation analysis on the repair coverage index to obtain a fault repair time series diagram; Performing collaborative optimization calculation on the evolutionary fault injection scheme and the dynamic repair strategy according to the fault repair timing diagram to obtain a strategy collaboration index; Optimizing and integrating the evolutionary fault injection scheme and the dynamic repair strategy according to the strategy synergy index to obtain a fault repair integration scheme; A reliability verification and evaluation is performed on the fault repair integration solution to obtain a reliability verification report.

[0011] The present invention further provides a device for verifying the functional reliability of an intelligent device, which is applied to any of the above-mentioned methods for verifying the functional reliability of an intelligent device, comprising: An acquisition module is used to obtain operating environment parameters and device status data of smart devices, and to perform redundant verification architecture construction and evaluation to obtain an initial verification benchmark; An analysis module, configured to collect real-time operating data of the smart device, perform functional correlation analysis with the initial verification benchmark, and obtain functional abnormality characteristics; an association module, the association module being configured to obtain associated device status information of the smart device, cross-validate the associated device status information based on the functional abnormality characteristics, and perform fault analysis with the functional abnormality characteristics to obtain an evolutionary fault injection solution; a processing module, the processing module being configured to obtain associated device status information of the smart device, cross-validate the associated device status information based on the functional abnormality characteristics, and perform fault analysis with the functional abnormality characteristics to obtain an evolutionary fault injection solution; A control module is used to integrate and analyze the evolutionary fault injection scheme and the dynamic repair strategy to obtain a reliability verification report.

[0012] The present invention also provides a device for verifying the functional reliability of an intelligent device, comprising: Memory, used to store programs; The processor is used to execute the program to implement each step of any one of the above-mentioned methods for verifying the functional reliability of a smart device.

[0013] The present invention also provides a storage medium storing computer instructions, wherein the computer instructions are used to enable a computer to execute any of the above methods.

[0014] The present invention provides a method, apparatus, device, and storage medium for verifying the reliability of intelligent device functions, which have the following beneficial effects: By acquiring operating environment parameters and device status data for smart devices and conducting redundant verification architecture assessment, comprehensive monitoring and assessment of the device operating environment is achieved, providing more complete and accurate baseline data for reliability verification and effectively addressing the inadequate consideration of the operating environment in existing verification methods. By collecting real-time operating data and performing functional correlation analysis with the initial verification benchmark, abnormalities in device operation can be promptly identified, enabling dynamic monitoring of device functional status and improving the real-time and accuracy of the verification process. By acquiring correlated device status information and conducting cross-validation and fault analysis, a correlation assessment mechanism between devices is established, effectively addressing the existing method's neglect of inter-device interactions and enhancing the reliability and integrity of the verification results. By identifying abnormal behavior in real-time operating data and formulating dynamic remediation strategies, rapid response and handling of device faults is achieved, enhancing the system's fault handling capabilities and improving the operational stability of smart devices. By integrating and analyzing evolving fault injection schemes with dynamic remediation strategies, a comprehensive reliability verification system is formed, providing a comprehensive guarantee mechanism for the reliable operation of smart devices and effectively improving the overall reliability of the devices. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 This is a flow chart of a method for verifying the functional reliability of an intelligent device provided by the present invention; Figure 2 This is a structural diagram of a device for verifying the functional reliability of an intelligent device provided by the present invention; Figure 3 This is a structural diagram of a smart device functional reliability verification device provided by the present invention.

[0016] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0017] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0018] The present invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0019] Reference Figure 1 The present invention provides a method for verifying the functional reliability of an intelligent device, comprising: Step S1: Obtain operating environment parameters and device status data of the smart device, and perform redundant verification architecture construction and evaluation to obtain an initial verification benchmark; Step S2: collecting real-time operating data of the smart device, performing functional correlation analysis with the initial verification benchmark, and obtaining functional abnormality characteristics; Step S3: Obtain the associated device status information of the smart device, cross-validate the associated device status information based on the functional abnormality characteristics, and perform fault analysis with the functional abnormality characteristics to obtain an evolutionary fault injection plan; Step S4: Identify abnormal behavior of real-time operation data to obtain equipment repair requirements, conduct feasibility assessment of equipment repair requirements based on functional abnormality characteristics, and obtain a dynamic repair strategy; Step S5: Integrate and analyze the evolutionary fault injection scheme and the dynamic repair strategy to obtain a reliability verification report.

[0020] Based on the above steps, the detailed process is as follows: Step S1: Comprehensive monitoring of the operating environment of smart devices is achieved through multi-dimensional data collection. Operating environment parameters include physical environmental indicators such as temperature, humidity, vibration, and electromagnetic interference. Device status data covers core operating indicators such as CPU usage, memory usage, network traffic, and power status. The collected data undergoes data preprocessing to eliminate outliers and redundant data. The redundant verification architecture assessment adopts a multiple verification mechanism, and the collected data is input into different verification modules for cross-validation. The verification module includes data consistency check, timing correlation analysis, parameter boundary verification, etc. By setting a reasonable threshold range, a normal data fluctuation range is established. Based on the statistical analysis of historical data and combined with the standard parameters in the device specification manual, a reference benchmark for the normal operation of the device is constructed. This benchmark covers multiple dimensions such as the normal value range, change trend, and correlation characteristics of each parameter, providing an important basis for subsequent functional abnormality detection.

[0021] Step S2: The real-time data acquisition system continuously monitors the operating status of the equipment and records the dynamic changes of various functional indicators of the equipment. The collected data includes the operating parameters of the equipment functional modules, operation response time, resource usage, etc. By comparing and analyzing with the initial verification benchmark, abnormal data points that deviate from the normal range are identified. Functional correlation analysis uses data mining technology to establish a correlation model between functional parameters and discover the law of parameter changes and mutual influence relationships. Abnormal feature extraction uses machine learning algorithms, including cluster analysis, anomaly detection and other methods to identify abnormal patterns from multidimensional data. Functional abnormality characteristics describe the specific manifestations of equipment function degradation or failure, including information such as the time characteristics of the abnormality, parameter change characteristics, and the scope of influence. These characteristics provide important input for subsequent fault analysis and repair strategy formulation.

[0022] Step S3: The collection of associated device status information covers peripheral devices that have data interaction and functional dependence with the target device. Cross-validation analysis is performed to analyze the relationship between the operating status of the associated devices and the functional anomaly of the target device to verify whether the anomaly originates from the device itself or is affected by external devices. Fault analysis uses the fault tree analysis method, combined with the characteristics of functional anomalies, to construct the fault propagation path. By analyzing the diffusion pattern and impact mode of the fault, a targeted fault injection scheme is designed. The evolutionary fault injection scheme includes specific implementation details such as fault type, injection timing, and injection method. This scheme simulates various possible fault scenarios, verifies the response performance of the device under different fault conditions, and evaluates the fault tolerance and reliability level of the device. The design of the fault injection scheme takes into account the evolutionary characteristics of the fault and simulates the propagation process of the fault in time and space.

[0023] Step S4: Abnormal Behavior Identification uses a deep learning model to analyze real-time operating data and identify abnormal behavior patterns during device operation. The model builds a behavioral baseline by learning from normal behavior characteristics in historical data, enabling real-time detection of abnormal behavior. Abnormal behavior includes various types, such as abnormal functional response, performance degradation, and resource exhaustion. Based on the detected abnormal behavior and combined with the device's functional specifications and operating requirements, a list of device repair requirements is generated. These repair requirements cover functional restoration, performance optimization, and resource allocation. The feasibility assessment comprehensively analyzes the repair requirements based on technical implementation difficulty, resource consumption, and repair risk. The assessment integrates the functional abnormality characteristics obtained in step S2 to analyze the root cause and impact of the abnormal behavior. Based on the feasibility assessment results, a dynamic repair strategy develops a targeted repair plan. The repair strategy includes various methods, such as fault isolation, functional degradation, resource reallocation, and software updates, and dynamically adjusts repair measures based on the device's operating status.

[0024] Step S5: Integrated Analysis: A systematic comparative study of the evolutionary fault injection scheme and the dynamic repair strategy is conducted. The analysis includes the repair strategy's coverage of various fault types, a quantitative assessment of the repair effect, and the time overhead of the repair process. Simulation experiments verify the effectiveness of the repair strategy in different fault scenarios and evaluate its adaptability and robustness. The reliability verification report summarizes the equipment's response capabilities and recovery effects in the face of various faults. The report includes fault type statistics, fault impact assessment, repair strategy effectiveness analysis, and reliability index calculations. The verification report also provides early warnings for potential risks to equipment reliability and proposes recommendations for improving reliability. The report uses visualization to present verification results, intuitively demonstrating the equipment's reliability level. The verification report provides a decision-making basis for continuous improvement of equipment reliability and guides subsequent optimization work.

[0025] The present invention provides a method for verifying the functional reliability of intelligent devices. By acquiring the operating environment parameters and device status data of intelligent devices and conducting a redundant verification architecture assessment, this method achieves comprehensive monitoring and assessment of the device operating environment, providing more complete and accurate basic data for reliability verification and effectively addressing the problem of existing verification methods' insufficient consideration of the operating environment. By collecting real-time operating data and performing functional correlation analysis with the initial verification benchmark, abnormal characteristics during device operation can be promptly detected, enabling dynamic monitoring of the device's functional status and improving the real-time and accuracy of the verification process. By acquiring associated device status information and conducting cross-validation and fault analysis, a correlation assessment mechanism between devices is established, effectively addressing the problem of existing methods ignoring the mutual influence between devices and improving the reliability and integrity of the verification results. By identifying abnormal behavior in real-time operating data and formulating dynamic repair strategies, rapid response and handling of device faults are achieved, enhancing the system's fault handling capabilities and improving the operational stability of intelligent devices. By integrating and analyzing the evolving fault injection scheme and dynamic repair strategies, a complete reliability verification system is formed, providing a comprehensive guarantee mechanism for the reliable operation of intelligent devices and effectively improving the overall reliability of the devices.

[0026] In one embodiment, the operating environment parameters and device status data of the smart device are obtained, and a redundant verification architecture is constructed and evaluated to obtain an initial verification benchmark, including: Temperature data is collected using a PT100 temperature sensor with a measurement range of -50°C to 150°C, a sampling frequency of 1 time / second, and a measurement accuracy of ±0.1°C. Humidity data is collected using a capacitive humidity sensor with a measurement range of 0-100%RH, a sampling frequency of 1 time / second, and a measurement accuracy of ±2%RH. Vibration data is collected using a triaxial accelerometer with a measurement range of ±16g, a sampling frequency of 100Hz, and a measurement accuracy of 0.01g. Voltage data is collected using a high-precision voltage sensor with a measurement range of 0-380V, a sampling frequency of 10 times / second, and a measurement accuracy of ±0.1V. Current data is collected using a Hall effect current sensor with a measurement range of 0-100A, a sampling frequency of 10 times / second, and a measurement accuracy of ±0.1A. Data fusion uses a weighted average algorithm to assign different weight coefficients to each parameter, with the temperature weight being 0.3, the humidity weight being 0.2, the vibration weight being 0.2, the voltage weight being 0.15, and the current weight being 0.15, to obtain the operating environment parameters.

[0027] Status sampling is based on the operating environment parameters, with the sampling interval shortened when environmental parameters change significantly and extended when the environment is stable. The sampling interval ranges from 100ms to 1s. Sampled data includes the device's operating status (normal / abnormal / faulty), performance metrics (response time, processing power, resource utilization), and functional parameter data (operating mode, output power, control accuracy). Sampled data is stored in the data buffer using timestamps as the index, forming a time-series status data stream.

[0028] In the K-means clustering algorithm, the K value is determined using the silhouette coefficient method to determine the optimal number of clusters, with a range of 3-10. Cluster feature dimensions include operational status, performance indicators, and functional parameters. Cluster distances are calculated using Euclidean distance for similarity, and cluster centers are determined through iterative optimization. The clustering results form state node data, with each node containing information such as the cluster center value, the number of samples contained in the node, and the node feature vector.

[0029] A triple-redundant structure is constructed for each state node data, including master node data, mirror node data, and check node data. The master node stores the original state data, the mirror node stores a copy of the data, and the check node stores the checksum. A CRC32 checksum algorithm is used between nodes to ensure data consistency. The multi-dimensional verification node matrix adopts an N×M structure, where N represents the number of state nodes and M represents the number of redundant dimensions. Matrix elements contain node identifiers, data content, and checksum information.

[0030] The verification architecture adopts a layered structure, consisting of a data acquisition layer, a node verification layer, and a result output layer. The data acquisition layer receives status data, the node verification layer performs data verification, and the result output layer generates verification results. The layers are connected via a data bus, using an asynchronous communication mechanism. Bidirectional verification channels are established between nodes, supporting cross-validation. The verification process utilizes a parallel processing mechanism to improve verification efficiency.

[0031] The reliability assessment uses a comprehensive scoring model with scoring metrics including data integrity (weight 0.3), verification accuracy (weight 0.3), system response time (weight 0.2), and resource utilization (weight 0.2). Data integrity is calculated using the data loss rate, verification accuracy is calculated using the error detection rate, system response time is calculated using the average processing delay, and resource utilization is calculated using the system load factor. Each metric is quantified using a percentage, and a weighted average is taken to arrive at the final assessment score, which serves as the initial verification benchmark. The assessment benchmark score ranges from 0 to 100, with 90-100 being excellent, 80-89 being good, 70-79 being acceptable, and below 70 being unacceptable.

[0032] This embodiment achieves comprehensive monitoring of the operating environment of smart devices by adopting a data fusion method that integrates multi-dimensional parameters, ensuring the accuracy and completeness of environmental parameter collection. The state sampling mechanism based on adaptive sampling intervals effectively reduces data redundancy, improves sampling efficiency, and enables state data to more accurately reflect the real-time operating status of the device. Through the state node identification of the K-means clustering algorithm, accurate classification of device status is achieved, providing a reliable data foundation for subsequent reliability verification. The data node construction method using a triple redundant structure significantly improves the reliability and fault tolerance of the data, and effectively prevents data loss and errors. The design of a layered verification architecture, combined with asynchronous communication and parallel processing mechanisms, greatly improves verification efficiency and reduces system response delays. The application of a comprehensive scoring model realizes the quantitative evaluation of equipment reliability and provides a scientific decision-making basis for equipment reliability management.

[0033] In one embodiment, real-time operating data of smart devices is collected and functional correlation analysis is performed with the initial verification benchmark to obtain functional abnormality characteristics, including: During the functional reliability verification process for smart devices, real-time operational data is collected through sensor networks, including physical parameters such as temperature, humidity, vibration, voltage, and current. This collected data is normalized in three steps: data cleaning to remove outliers, missing values, and duplicate values; normalization to map the data to the interval [0, 1]; and standardization to convert the data to a standard normal distribution with a mean of 0 and a variance of 1. The resulting normalized operational data has a uniform numerical scale and distribution characteristics.

[0034] Dynamic benchmark matching is performed on the normalized operating data based on the initial verification benchmark. The initial verification benchmark contains information such as standard parameter thresholds, parameter variation ranges, and parameter correlations during normal equipment operation. Dynamic benchmark matching uses Euclidean distance to calculate the distance matrix between the normalized operating data and the initial verification benchmark. The Mahalanobis distance is used to calculate the correlation deviation between the data. The covariance matrix is used to calculate the degree of deviation in parameter variation trends, generating a benchmark deviation signature. The benchmark deviation signature represents the deviation between the current operating state and the normal state in vector form.

[0035] Multidimensional feature decoupling analysis was performed on the baseline deviation features, and principal component analysis (PCA) was used for feature decomposition. By calculating eigenvalues and eigenvectors, principal components with a cumulative contribution rate of 85% were selected, and the high-dimensional feature space was mapped to a low-dimensional feature space. In this low-dimensional feature space, independent feature components were extracted using the maximum variance method to obtain potential abnormal components. These potential abnormal components reflect the primary factors contributing to functional abnormalities.

[0036] The DBSCAN density clustering algorithm was used to cluster potential anomalies. Based on the principle of density reachability, the neighborhood radius ε was set to 0.5 and the minimum number of samples, MinPts, was set to 4. Density-connected data points were grouped into the same cluster. By calculating local density and relative density differences, high-density core points and low-density boundary points were identified to form a set of anomaly feature clusters. Each cluster in the set represents a typical anomaly pattern.

[0037] Functional association analysis was performed on the abnormal feature clusters, and the Apriori association rule mining algorithm was used to establish the association between abnormal features and functional failures. The minimum support threshold was set at 0.3, and the minimum confidence threshold was set at 0.8. Frequent item sets that met the threshold conditions were mined. Association rules were generated based on the frequent item sets, and the rule lift was calculated. Strong association rules with a lift greater than 1.5 were selected to obtain core abnormal feature data. This core abnormal feature data includes information such as abnormal features, failure type, and association strength.

[0038] Dynamic weighting is performed based on core abnormality feature data, using the entropy weight method to determine feature weights. The information entropy of each feature is calculated; lower entropy indicates greater information content. A weight coefficient is calculated based on the information entropy, with the weight coefficient being inversely proportional to the information entropy. The weight coefficient is multiplied by the feature value to produce a weighted functional abnormality feature. Functional abnormality features include quantitative indicators such as abnormality severity, abnormality type probability, and reliability score.

[0039] This embodiment achieves unified standardization of different physical quantity parameters by adopting a multi-dimensional data normalization processing mechanism, eliminates dimensionality effects, and improves data comparability and analysis accuracy. Through dynamic benchmark matching processing, a multi-level deviation calculation model is established, which achieves accurate comparison between the operating state and the normal state, and improves the accuracy of anomaly detection. Through multi-dimensional feature decoupling analysis, the data dimension is reduced, the main influencing factors are extracted, the feature space is simplified, and the computational efficiency is improved. Through the application of density clustering algorithm, the automatic discovery of abnormal patterns is achieved, the subjectivity of manually setting thresholds is avoided, and the objectivity of anomaly identification is enhanced. Through association rule mining, a mapping relationship between abnormal features and functional faults is established, which achieves accurate identification of fault types and improves the accuracy of diagnosis. Through the dynamic allocation of feature weights by the entropy weight method, a quantitative assessment of the degree of abnormality is achieved, providing a scientific basis for equipment reliability verification.

[0040] In one embodiment, the associated device status information of the smart device is obtained, the associated device status information is cross-validated based on the functional abnormality characteristics, and the fault analysis is performed with the functional abnormality characteristics to obtain an evolutionary fault injection scheme, including: During the device connection address scanning phase, the system uses network scanning technology to scan smart devices. The specific process involves: scanning active devices within the local network segment using the ARP protocol to obtain their MAC addresses; using the ICMP protocol to detect the device's network connectivity and confirm its IP address; applying port scanning technology to identify the device's open service ports; and obtaining device system information using the SNMP protocol. The scan results form a collection address list, which records device identification information.

[0041] During the associated data collection process, data acquisition is performed based on the collection address list. The processing flow includes: establishing a communication connection with the target device; sending data requests according to the communication protocols supported by the device (such as Modbus and OPC UA); periodically collecting device operating parameters, including CPU usage, memory usage, network traffic, and process status; and storing the collected data in time series to construct an associated status dataset.

[0042] During the cross-validation calculation phase, the system analyzes the associated state dataset. This includes calculating the correlation coefficient matrix of device state parameters; using principal component analysis to reduce dimensionality and extract key features; applying association rule mining algorithms to analyze inter-device relationships; and constructing a device association network using graph theory. The analysis results form a device interaction influence matrix.

[0043] The anomaly mapping process uses a network propagation model for analysis. The steps are: converting the device interaction matrix into a directed weighted graph; calculating the anomaly propagation path using a shortest path algorithm; simulating the anomaly diffusion process using a propagation dynamics model; and evaluating the spatiotemporal distribution of the anomaly's impact. The analysis results in a graph of the anomaly transmission chain.

[0044] The abnormal pattern recognition process applies pattern recognition technology to classify faults. This includes constructing a fault feature vector; performing pattern matching using classification algorithms such as support vector machines; calculating the similarity between the fault signature and predefined patterns; and determining the fault type and severity. The recognition results form a description of the abnormal pattern.

[0045] The statistical analysis phase conducts frequency analysis of abnormal patterns. This process includes: counting the number of occurrences of various fault types; calculating the temporal distribution of faults; analyzing the spatial distribution characteristics of faults; and assessing the probability distribution of faults. The analysis results are reported as an abnormality frequency distribution report.

[0046] Anomaly propagation patterns are analyzed using a time series analysis method. The processing steps are: extracting the time series characteristics of fault evolution; analyzing the trend characteristics of fault development; identifying the spatial characteristics of fault propagation; and establishing a fault evolution prediction model. The analysis results generate a description of the anomaly evolution characteristics.

[0047] The conditional analysis process is conducted using a causal analysis approach. The process includes: identifying the triggering conditions for a fault, including hardware environmental conditions (temperature, humidity, voltage, etc.), software environmental conditions (system load, resource usage, etc.), and network environmental conditions (bandwidth, latency, etc.); analyzing the environmental factors that cause the fault; summarizing the necessary conditions for reproducing the fault; and establishing a mapping relationship between fault conditions and results. The analysis results form a set of fault injection conditions, which includes the fault triggering conditions, fault type definitions (functional fault, performance fault, resource fault), fault parameter settings (duration, frequency, scope of impact, severity), and fault dependencies.

[0048] The fault test sequence plan construction process utilizes a systematic test case design approach. This process analyzes the fault injection condition set, determines test coverage, identifies key fault scenarios, and designs fault combinations. The test scenario and execution environment design involves planning the test topology, configuring test environment parameters, preparing test datasets, and deploying the test toolchain. The fault injection operational process includes designing injection steps, configuring fault parameter values, planning execution timing, and developing recovery plans. The establishment of a test verification system encompasses the definition of verification metrics, designing data collection plans, developing judgment criteria, and establishing evaluation methods. Ultimately, a complete, evolved fault injection plan is formed, including a test scenario description (environmental configuration requirements, device networking topology, data preparation plan, tool configuration instructions), a fault injection implementation plan (injection timing, parameter configuration plan, combination design, and execution process), a test verification plan (metric definition, data collection plan, verification method, and judgment criteria), and a test result evaluation plan (analysis methods, evaluation metrics, report templates, and issue tracking mechanisms).

[0049] This embodiment achieves accurate identification and prediction of device failures by conducting comprehensive functional reliability verification of smart devices. Network scanning and multi-protocol data acquisition technology are used to ensure the integrity and accuracy of device status information collection. Through cross-validation calculations and anomaly mapping operations, an associated network and anomaly transmission link between devices are constructed, effectively revealing the propagation law and impact range of the fault. Based on pattern recognition and statistical analysis methods, a systematic fault classification system is established to improve the accuracy of fault diagnosis. Through timing analysis and causal analysis, a complete fault evolution feature description and injection condition set are formed, providing a reliable basis for fault prevention. The evolutionary fault injection scheme finally constructed realizes the systematization and standardization of fault testing, improves the reliability verification efficiency of smart devices, and provides a strong guarantee for the stable operation of the equipment.

[0050] In one embodiment, conditional analysis is performed on the abnormal evolution characteristics to obtain a set of fault injection conditions, including: In the process of conditional analysis of abnormal evolution characteristics to obtain a fault injection condition set, the abnormal evolution characteristics refer to the data characteristics of abnormal state changes that occur during the operation of intelligent devices.

[0051] During the segmented processing phase, the anomaly evolution characteristics are divided into multiple time windows according to the time series. Each window contains a set of state parameters, forming an evolution process sequence. The evolution process sequence reflects the dynamic process of device state changes over time. Segmented processing uses a sliding time window method. By setting the window size and sliding step size, feature extraction is performed on the time series data to construct an evolution process sequence matrix, which contains the temporal characteristics of the anomaly evolution.

[0052] State space projection maps the multidimensional state parameters of an evolutionary process sequence into a low-dimensional space, generating a state mapping point cloud. State mapping point clouds transform high-dimensional data into a visual 3D or 2D point set using dimensionality reduction methods such as principal component analysis.

[0053] Density field calculation is to perform kernel density estimation on the points in the state mapping point cloud, calculate the point density distribution in each area, and form an abnormal distribution manifold. Density field calculation uses the kernel density estimation method to calculate the local density of each point P(x, y, z) in the state mapping point cloud.

[0054] The specific calculation formula is: ; in, is the local density value at point P, is the Euclidean distance from point i to point j, is the cutoff distance parameter, n is the total number of points in the state mapping point cloud, : Natural exponential function.

[0055] After calculating the density value for each data point, a continuous density field distribution is constructed using trilinear interpolation. The anomaly distribution manifold describes the distribution characteristics of anomaly states in low-dimensional space.

[0056] Trigger condition topological association analyzes the connectivity between regions within the anomaly distribution manifold and constructs a conditional coupling network. This network is established based on a distance threshold method. By calculating the geodesic distance between pairs of points in the anomaly distribution manifold, a connecting edge is established when the distance is less than a set threshold, and the corresponding edge weight is assigned. All connecting edges form the topological structure of the conditional coupling network, which reflects the transition paths between different anomaly states.

[0057] The spectral decomposition operation performs eigenvalue decomposition on the conditional coupling network, extracts the main eigenvectors of the network, and obtains the conditional path spectrum. The conditional path spectrum characterizes the main patterns of abnormal state transitions.

[0058] The state transition entropy calculation is based on the conditional path spectrum, calculating the information entropy during the state transition process and determining the critical trigger condition set. The state transition entropy is calculated based on the conditional path spectrum. By constructing the state transition probability matrix and calculating the transition entropy of each state, the key state transition points of the system are determined. These points constitute the critical trigger condition set.

[0059] Bifurcation point identification involves finding critical points within the critical trigger condition set where the state undergoes a sudden change, thereby determining the critical trigger domain. Bifurcation point identification employs differential topology methods, calculating the derivatives of the state variables with respect to the control parameters and identifying locations where these derivatives exhibit singularities. These locations are known as bifurcation points, forming the boundaries of the critical trigger domain.

[0060] Dynamic fault parameter constraint solving involves establishing constraint equations between fault parameters and system states within the critical triggering domain and solving them to obtain a fault parameter constraint set. The fault parameter constraint set describes the range of values for each parameter when a fault occurs.

[0061] Multi-objective optimization combination configuration involves multi-objective optimization of the fault parameter constraint set, selecting the optimal parameter combination, and forming a fault injection condition set. The fault injection condition set provides specific test conditions for functional reliability verification of smart devices.

[0062] This embodiment achieves accurate conversion from raw data to fault injection conditions by systematically analyzing the conditions of abnormal evolution. The sliding time window method is used for segmented processing, combined with the state space projection technology, which effectively reduces the data dimension and improves the efficiency of feature extraction. The density field calculation method based on kernel density estimation accurately depicts the distribution characteristics of abnormal states, laying the foundation for subsequent analysis. Through trigger condition topological association and spectral decomposition operations, a conditional coupling network reflecting the abnormal state conversion relationship is constructed, making the abnormal evolution path clearer and more visible. Combined with the state transition entropy value calculation and bifurcation point identification technology, the critical state point of the system is accurately located, and the accuracy of fault prediction is improved. By establishing dynamic constraints on fault parameters and multi-objective optimization configuration, a scientific and reasonable set of fault injection conditions is formed, which provides a reliable test basis for the functional reliability verification of intelligent devices and significantly improves the verification efficiency and accuracy.

[0063] In one embodiment, abnormal behavior is identified in real-time operation data to obtain equipment repair requirements. The feasibility of the equipment repair requirements is evaluated based on the functional abnormality characteristics to obtain a dynamic repair strategy, including: Sensors collect real-time operational data from devices, including information such as device operating status, performance parameters, and environmental parameters. This data is used to identify abnormal behaviors, employing a multi-layered anomaly detection mechanism. At the numerical level, statistical analysis methods, including mean deviation analysis, variance analysis, and trend analysis, are used to identify data anomalies. At the pattern level, pattern matching algorithms, including temporal pattern analysis and association rule analysis, are used to identify behavioral anomalies. At the semantic level, knowledge graphs are used to understand the semantic meaning of anomalies. Abnormal behavior characteristics are represented by feature vectors, which contain information such as the time of occurrence, abnormal value, abnormal pattern type, and abnormality severity. The system compares the abnormal characteristics with pre-set normal behavior templates to generate a description of the abnormal behavior characteristics.

[0064] Based on the extracted abnormal behavior features, a behavior deviation matrix is constructed. This is a multidimensional data structure that characterizes the degree of discrepancy between a device's actual operating behavior and its expected behavior. Threshold analysis is performed by setting deviation thresholds. When the deviation exceeds the preset threshold, abnormal behavior determination information is generated. This abnormal behavior determination information includes key indicators such as the abnormality type, severity, and duration.

[0065] Based on abnormal behavior information and the device's functional model, the device repair requirement is calculated. This refers to the specific measures required to restore normal device functionality, including hardware replacement, software upgrades, parameter adjustments, and other repair options. The repair requirement is calculated based on factors such as the severity of the abnormal behavior, the scope of impact, and the urgency of the repair.

[0066] Resource availability assessments for equipment repair needs are conducted using a multi-dimensional evaluation system. Hardware resource assessments include spare parts inventory, spare parts quality, and equipment status assessments; software resource assessments include system capacity, performance, and compatibility assessments; and human resource assessments include technical capabilities, time commitment, and professional compatibility assessments. The evaluation process uses the Analytic Hierarchy Process (AHP) to establish an evaluation index system, and fuzzy comprehensive evaluation methods are used to calculate the availability score of each resource. The evaluation results are presented in the form of a resource availability matrix, with the matrix elements representing the availability of each resource, forming the repair resource conditions.

[0067] Calculate the repair cost based on the repair resource requirements. This cost includes material costs, labor costs, and time costs. Through cost-benefit analysis and considering the severity of the abnormal behavior, determine the repair priority. This repair priority information is used to guide the sequencing of repair tasks and resource allocation.

[0068] Repair plans are constructed based on repair priority information, using a combination of case-based and rule-based reasoning. The case library stores historical repair plans, including fault characteristics, repair measures, and repair results. The rule library contains rule-based knowledge, such as equipment maintenance procedures, technical standards, and safety regulations. The plan construction process searches the case library for similar cases and extracts effective repair measures. This rule-based knowledge is then applied to constrain and optimize the repair measures. Specific repair execution measures are then formulated based on the current equipment status and resource conditions. These repair execution measures include detailed operational steps, resource allocation plans, and schedules.

[0069] The proposed repair measures are verified, and the feasibility and effectiveness of the repair plan are evaluated through simulation testing or small-scale trials to obtain repair effect evaluation parameters. The repair effect evaluation parameters include quantitative indicators such as performance indicators after repair, degree of functional recovery, and stability.

[0070] Based on repair effect evaluation parameters, a closed-loop control mechanism is used for dynamic repair. During the repair process, equipment status data and repair effect data are continuously collected, and the repair progress and effectiveness are calculated through a real-time evaluation mechanism. Evaluation indicators include functional recovery, performance improvement, stability, and so on. When the evaluation results show that the repair effect is unsatisfactory, the dynamic adjustment mechanism of the repair strategy is triggered. The adjustment mechanism is based on a reinforcement learning algorithm, which optimizes repair parameters and measures through feedback from historical repair experience and current repair effects. The dynamic repair strategy is expressed through a decision tree model, and the decision nodes include elements such as repair conditions, repair measures, and effect evaluation. Based on the reasoning results of the decision tree, the system adjusts the repair strategy in real time until the expected repair effect is achieved. The entire process uses a distributed computing framework to ensure real-time performance. The decision-making process uses a multi-agent collaborative mechanism to improve decision accuracy. The execution process uses automated control technology to improve execution efficiency, forming a complete closed-loop control system to achieve continuous optimization of equipment functional reliability.

[0071] This embodiment achieves accurate identification and feature extraction of abnormal device behavior by adopting a multi-level anomaly detection mechanism and a feature vector representation method, thereby improving the accuracy and real-time performance of anomaly detection. The method based on the behavior deviation matrix and threshold analysis makes the determination of abnormal behavior more objective and quantitative, providing a reliable basis for subsequent repair decisions. The multi-dimensional resource evaluation system and fuzzy comprehensive evaluation method are adopted to achieve a comprehensive evaluation of repair resources and ensure the feasibility of the repair plan. The solution construction method that combines case reasoning and rule reasoning improves the scientificity and reliability of the repair plan. The dynamic repair strategy based on the closed-loop control mechanism and the reinforcement learning algorithm enables the repair process to be adaptively adjusted according to real-time feedback, significantly improving the repair effect. The entire method adopts a distributed computing framework and a multi-agent collaborative mechanism, which greatly improves the real-time performance and decision-making accuracy of the system, and provides an effective guarantee for the continuous optimization of the functional reliability of smart devices.

[0072] In one embodiment, the evolutionary fault injection scheme and the dynamic repair strategy are integrated and analyzed to obtain a reliability verification report, including: Specific implementation plans for various fault types are introduced through manual settings or automatic triggering. Dynamic repair strategies refer to real-time repair solutions formulated for various fault types.

[0073] When integrating and analyzing the evolving fault injection scheme with the dynamic repair strategy, fault types are categorized and organized, and a fault type matrix is constructed based on dimensions such as hardware faults, software faults, and network faults. The fault type matrix is a multidimensional data structure that contains attribute information such as fault type, fault level, and fault impact range. Each fault type is subdivided into multiple specific fault items.

[0074] Based on the constructed fault type matrix, the dynamic repair strategy's repair effectiveness is quantitatively evaluated. This evaluation utilizes a multi-dimensional scoring system, including three core metrics: repair time (RT), repair success rate (RS), and resource consumption (RC). The RT metric is normalized and converted to a percentage score. The RS metric uses the raw percentage value. The RC metric is reverse-calculated based on resource utilization, and a weighted calculation is performed to determine the repair effectiveness score. The repair effectiveness score is a value from 0 to 100, with higher scores indicating better repair effectiveness.

[0075] We optimize and analyze the restoration effect scores, extract key parameters that influence restoration effectiveness through mathematical modeling and data mining, and form restoration strategy optimization parameters. These parameters include key decision-making factors such as restoration timing, restoration method selection, and resource allocation.

[0076] Based on the optimized parameters of the repair strategy, the repair coverage of the evolved fault injection scheme is calculated. This repair coverage calculation is based on a fault type matrix. By calculating the repair strategy coverage ratio for each type of fault and combining it with the fault type weight coefficient, the overall repair coverage ratio is derived. The repair coverage ratio is expressed as a percentage, reflecting the degree to which the existing repair strategy covers each type of fault.

[0077] Perform a time-series correlation analysis on the repair coverage metrics and create a fault repair timeline diagram. This diagram shows the time series relationship between fault occurrence, detection, and repair, including key time nodes and detailed information about the repair process.

[0078] Based on the fault repair sequence diagram, a collaborative optimization calculation is performed on the evolving fault injection scheme and the dynamic repair strategy. The collaborative optimization calculation adopts the matrix matching analysis method and is achieved through the correspondence analysis of the fault injection scheme matrix F and the repair strategy matrix R.

[0079] The formula for calculating the strategic synergy index is: ; : Strategy synergy index, which indicates the matching degree between the fault injection scheme and the repair strategy.

[0080] : Fault injection scheme matrix, which contains the injection parameters for various fault types.

[0081] : Repair strategy matrix, containing the corresponding repair strategy parameters.

[0082] : The i-th element in the fault injection scheme matrix.

[0083] : The i-th element in the repair strategy matrix.

[0084] : Matrix dimension, indicating the number of fault types.

[0085] : The modulus length of the fault injection scheme matrix.

[0086] : Modulus of the repair strategy matrix.

[0087] : The dot product of two matrices. : The angle between the two matrices. The value range is 0-1, and the closer the value is to 1, the better the synergy effect.

[0088] Based on the strategy synergy index, an iterative optimization method is used to optimize and integrate the evolving fault injection scheme and dynamic repair strategy. The optimization process focuses on three core indicators: the strategy synergy index, the repair effectiveness score, and the repair coverage rate. Through parameter adjustment and strategy optimization, overall performance is improved, resulting in an integrated fault repair solution. The integrated fault repair solution includes the optimized fault injection rules and the corresponding repair strategy combination.

[0089] Conduct reliability verification and assessment of the integrated fault repair solution. This assessment utilizes a multi-tiered system, comprehensively evaluating the solution's integrity, effectiveness, stability, and affordability. The evaluation process combines quantitative and qualitative analysis to verify the solution's performance indicators. Verification results are presented as a quantitative score, and a reliability verification report containing detailed analysis data is generated, providing a comprehensive assessment basis for the functional reliability of smart devices. Evaluation indicators and thresholds are based on industry standards, and the evaluation methodology and calculation system ensure the scientific and reliable verification results.

[0090] This embodiment realizes the coordinated optimization of fault injection schemes and repair strategies by constructing a functional reliability verification method for intelligent devices. This method adopts a multi-dimensional scoring system to quantitatively evaluate the repair effect, and effectively improves the accuracy and efficiency of fault repair through the comprehensive calculation of three core indicators: repair time, repair success rate, and resource consumption. The repair coverage calculation method based on the fault type matrix enables the system to comprehensively evaluate the coverage of the repair strategy for various types of faults, significantly improving the integrity of fault handling. The optimal matching of the fault injection scheme and the repair strategy is achieved through the calculation formula of the strategy synergy index, which greatly improves the fault handling capability of the system. The application of a multi-level evaluation system ensures the scientificity and reliability of the verification results, provides a comprehensive evaluation basis for the functional reliability of intelligent devices, effectively reduces the risk of equipment operation, and improves the overall stability of the system.

[0091] Reference Figure 2As shown, the present invention also provides a smart device function reliability verification device, which is applied to any of the above-mentioned smart device function reliability verification methods, including: The acquisition module is used to obtain the operating environment parameters and device status data of the smart device, and to conduct redundant verification architecture construction and evaluation to obtain the initial verification benchmark; The analysis module is used to collect real-time operation data of smart devices, perform functional correlation analysis with the initial verification benchmark, and obtain functional abnormality characteristics; The association module is used to obtain the associated device status information of the smart device, cross-validate the associated device status information based on the functional abnormality characteristics, and perform fault analysis with the functional abnormality characteristics to obtain an evolutionary fault injection plan; The processing module is used to obtain the associated device status information of the smart device, cross-validate the associated device status information based on the functional abnormality characteristics, and perform fault analysis with the functional abnormality characteristics to obtain an evolutionary fault injection plan; Control module,The control module is used to integrate and analyze the evolutionary fault injection scheme and the dynamic repair strategy to obtain a reliability verification report.

[0092] The present invention provides a functional reliability verification device for intelligent devices. By acquiring operating environment parameters and device status data of intelligent devices and performing redundant verification architecture construction and evaluation, this device comprehensively monitors and evaluates the operating environment. This provides more complete and accurate basic data for reliability verification, effectively addressing the problem of existing verification methods' insufficient consideration of the operating environment. By collecting real-time operating data and performing functional correlation analysis with the initial verification benchmark, abnormal characteristics during device operation can be promptly detected, enabling dynamic monitoring of the device's functional status and improving the real-time and accuracy of the verification process. By acquiring associated device status information and performing cross-validation and fault analysis, a correlation assessment mechanism between devices is established, effectively addressing the problem of existing methods ignoring the mutual influence between devices and improving the reliability and integrity of the verification results. By identifying abnormal behavior in real-time operating data and formulating dynamic repair strategies, rapid response and handling of device faults are achieved, enhancing the system's fault handling capabilities and improving the operational stability of intelligent devices. By integrating and analyzing the evolving fault injection scheme and dynamic repair strategies, a complete reliability verification system is formed, providing a comprehensive guarantee mechanism for the reliable operation of intelligent devices and effectively improving the overall reliability of the devices.

[0093] Reference Figure 3 As shown, the present invention also provides a smart device functional reliability verification device, comprising: Memory, used to store programs; The processor is used to execute the program to implement each step of any one of the above-mentioned methods for verifying the functional reliability of an intelligent device.

[0094] In this embodiment, the processor and memory may be connected via a bus or other means. The memory may include volatile memory, such as random access memory; the memory may also include non-volatile memory, such as read-only memory, flash memory, hard disk, or solid-state drive. The processor may be a general-purpose processor, such as a central processing unit, a digital signal processor, an application-specific integrated circuit, or one or more integrated circuits configured to implement the embodiments of the present invention.

[0095] The present invention also provides a storage medium storing computer instructions, wherein the computer instructions are used to enable a computer to execute any of the above methods.

[0096] It should be noted that, those skilled in the art will clearly understand that, for the sake of convenience and brevity of description, the specific working processes of the above-described system and each module can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0097] The above description is only a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made by using the contents of the present invention description and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A method for verifying the functional reliability of an intelligent device, characterized in that: include: Obtain the operating environment parameters and device status data of the smart device, and conduct a redundant verification architecture construction assessment to obtain the initial verification benchmark; Collecting real-time operating data of the smart device, performing functional correlation analysis with the initial verification benchmark, and obtaining functional abnormality characteristics; Acquire associated device status information of the smart device, cross-validate the associated device status information based on the functional abnormality characteristics, and perform fault analysis with the functional abnormality characteristics to obtain an evolutionary fault injection solution; Identifying abnormal behavior of the real-time operation data to obtain equipment repair requirements, and conducting a feasibility assessment of the equipment repair requirements based on the functional abnormality characteristics to obtain a dynamic repair strategy; The evolutionary fault injection scheme and the dynamic repair strategy are integrated and analyzed to obtain a reliability verification report.

2. The method for verifying the functional reliability of an intelligent device according to claim 1, wherein: The process of obtaining the operating environment parameters and device status data of the smart device and conducting a redundant verification architecture construction evaluation to obtain an initial verification benchmark includes: Collecting temperature data, humidity data, vibration data, voltage data, and current data of the smart device, and performing multi-dimensional parameter integration to obtain the operating environment parameters; Sampling the state of the smart device according to the operating environment parameters to obtain the device state data; Performing cluster identification on the device status data to obtain status node data; Performing redundant data node construction on the state node data to obtain a multi-dimensional verification node matrix; Constructing a verification architecture topology structure according to the multi-dimensional verification node matrix to obtain a redundant verification architecture; A reliability evaluation calculation is performed on the redundant verification architecture to obtain the initial verification benchmark.

3. The method for verifying the functional reliability of an intelligent device according to claim 1, wherein: The collecting of the real-time operation data of the smart device and performing functional correlation analysis with the initial verification benchmark to obtain functional abnormality characteristics include: Performing data normalization processing on the real-time operation data to obtain normalized operation data; performing dynamic benchmark matching processing on the normalized operating data according to the initial verification benchmark to generate a benchmark deviation feature; Performing multi-dimensional feature decoupling on the baseline deviation feature to obtain potential abnormal components; Performing pattern density clustering on the potential abnormal components to obtain an abnormal feature cluster set; Performing functional association on the abnormal feature cluster set to obtain core abnormal feature data; Dynamic weight allocation is performed based on the core abnormal feature data to obtain functional abnormality features.

4. The method for verifying the functional reliability of an intelligent device according to claim 1, wherein: The acquiring of the associated device status information of the smart device, cross-verifying the associated device status information according to the functional abnormality characteristics, and performing fault analysis with the functional abnormality characteristics to obtain an evolving fault injection scheme includes: Scan the device connection relationship address of the smart device to obtain the collection address; Performing associated data collection on the associated device status information according to the collection address to obtain an associated status data set; Performing cross-validation calculation on the functional abnormality features according to the associated state data set to obtain a device interaction influence matrix; Performing an abnormal mapping operation on the device interaction influence matrix to obtain an abnormal transmission link; Classify and identify the functional abnormality features according to the abnormal transmission link to obtain an abnormal pattern recognition result; Performing statistical analysis on the abnormal pattern recognition results to obtain abnormal occurrence frequency distribution; Analyzing the anomaly propagation law of the anomaly occurrence frequency distribution to obtain anomaly evolution characteristics; Performing conditional analysis on the abnormal evolution characteristics to obtain a fault injection condition set; A fault test sequence scheme is constructed according to the fault injection condition set to obtain the evolutionary fault injection scheme.

5. The method for verifying the functional reliability of an intelligent device according to claim 4, wherein: The conditional analysis of the abnormal evolution characteristics is performed to obtain a fault injection condition set, including: Segmenting the abnormal evolution characteristics to obtain a conditional path spectrum; Calculating the state transition entropy value based on the conditional path spectrum to obtain a critical trigger condition set; Identifying bifurcation points on the critical trigger condition set to obtain a critical trigger domain; Performing dynamic constraint solving of fault parameters according to the critical triggering domain to obtain a fault parameter constraint set; A multi-objective optimization combination configuration is performed on the fault parameter constraint set to obtain the fault injection condition set.

6. The method for verifying the functional reliability of an intelligent device according to claim 1, wherein: The identifying abnormal behavior of the real-time operation data to obtain equipment repair requirements, and performing a feasibility assessment on the equipment repair requirements based on the functional abnormality characteristics to obtain a dynamic repair strategy include: Performing abnormal behavior identification on the real-time operation data to obtain abnormal behavior characteristics; Constructing a behavior deviation matrix based on the abnormal behavior characteristics and performing threshold analysis to obtain abnormal behavior determination information; Calculating equipment repair requirements based on the abnormal behavior determination information to obtain the equipment repair requirements; Performing resource availability assessment on the equipment repair requirements based on the functional abnormality characteristics to obtain repair resource conditions; Calculating the repair cost of the repair resource conditions to obtain repair priority information; Constructing a plan based on the repair priority information to obtain repair execution measures; Verifying the repair execution measures to obtain repair effect evaluation parameters; The repair execution measures are dynamically repaired according to the repair effect evaluation parameters to obtain the dynamic repair strategy.

7. The method for verifying the functional reliability of an intelligent device according to claim 1, wherein: The integrated analysis of the evolutionary fault injection scheme and the dynamic repair strategy to obtain a reliability verification report includes: Classifying the fault types of the evolutionary fault injection scheme to obtain a fault type matrix; Performing a quantitative evaluation of the repair effect of the dynamic repair strategy according to the fault type matrix to obtain a repair effect score; Optimizing and analyzing the repair effect scores to obtain repair strategy optimization parameters; Calculating the repair coverage of the evolutionary fault injection scheme according to the repair strategy optimization parameters to obtain a repair coverage index; Performing a time series correlation analysis on the repair coverage index to obtain a fault repair time series diagram; Performing collaborative optimization calculation on the evolutionary fault injection scheme and the dynamic repair strategy according to the fault repair timing diagram to obtain a strategy collaboration index; Optimizing and integrating the evolutionary fault injection scheme and the dynamic repair strategy according to the strategy synergy index to obtain a fault repair integration scheme; Perform reliability verification and evaluation on the fault repair integration solution to obtain a reliability verification report.

8. A device for verifying the functional reliability of an intelligent device, characterized in that: The method for verifying the functional reliability of a smart device as described in any one of claims 1 to 7 above comprises: An acquisition module is used to obtain operating environment parameters and device status data of smart devices, and to perform redundant verification architecture construction and evaluation to obtain an initial verification benchmark; An analysis module, configured to collect real-time operating data of the smart device, perform functional correlation analysis with the initial verification benchmark, and obtain functional abnormality characteristics; an association module, the association module being configured to obtain associated device status information of the smart device, cross-validate the associated device status information based on the functional abnormality characteristics, and perform fault analysis with the functional abnormality characteristics to obtain an evolutionary fault injection solution; a processing module, the processing module being configured to obtain associated device status information of the smart device, cross-validate the associated device status information based on the functional abnormality characteristics, and perform fault analysis with the functional abnormality characteristics to obtain an evolutionary fault injection solution; A control module is used to integrate and analyze the evolutionary fault injection scheme and the dynamic repair strategy to obtain a reliability verification report.

9. A device for verifying the functional reliability of an intelligent device, characterized in that: include: Memory, used to store programs; The processor is configured to execute the program to implement the steps of the method for verifying the functional reliability of an intelligent device as described in any one of claims 1 to 7.

10. A storage medium, characterized in that: Computer instructions are stored, and the computer instructions are used to make a computer execute the method according to any one of claims 1 to 7.

Images