A heterogeneous protocol concurrent test method based on physical topology and cascading failure modeling

Abstract

The application provides a heterogeneous protocol concurrent test method based on physical topology and chain failure modeling, relates to the technical field of water, wind, light and storage large-scale regulation and control systems, and aims to solve the problems of spatial and temporal evolution causality correlation when network pressure is missing and semantic conflicts are prone to occur. The application improves the deficiencies. The method acquires device topology data, constructs a state transition probability matrix to simulate the transition process, executes the Hox process to generate an initial concurrent network message when the transition occurs, performs dynamic semantic deconstruction and timestamp calibration preprocessing on the message to generate joint test data and send the joint test data to a verification platform for pressure, generates evaluation data through a probe and calculates a dynamic quantitative index set, generates a three-dimensional evaluation matrix and dynamically adjusts the excitation parameters to perform automatic optimization control. The application avoids invalid injection and effectively evaluates the scheduling capability of the platform.

Description

Technical Field

[0001] This invention relates to the field of large-scale control systems for hydro, wind, solar, and energy storage, and in particular to a heterogeneous protocol concurrent testing method based on physical topology and cascading fault modeling. Background Technology

[0002] For panoramic dynamic verification scenarios of large-scale control systems such as hydropower, wind power, solar power, and energy storage, existing hardware-in-the-loop (HIL) simulation systems typically rely on the conversion of basic hardware interfaces and the adaptation of data bus protocols to provide integrated access for link-layer devices. The limitations of this conventional approach in practical engineering applications lie in its focus on basic hardware compatibility and signal connectivity, lacking application-layer high-concurrency pressure generation control mechanisms and the ability to dynamically drive testing and evaluation based on the evolution of physical system faults. When faced with complex verification requirements involving multiple heterogeneous protocols, existing concurrent testing mechanisms often employ independent network traffic generators for static script-based pressure application. This unidirectional network-layer pressure injection method, lacking physical space constraints and causal relationships of temporal evolution, struggles to effectively test and verify the computational performance and scheduling mechanisms of large-scale control systems when simulating abnormal network conditions.

[0003] A general intelligent I / O data access method for a hardware-in-the-loop simulation system, disclosed in Chinese patent application CN118707869A, involves inserting a customized interface fiber optic protocol into fiber optic data frames. This allows the intelligent I / O interface to occupy a fixed dedicated address in the fiber optic reflection memory, and stores corresponding data at a fixed offset address in the corresponding node based on the signal type. Analysis of this cited literature and similar prior art reveals its shortcomings: this method is limited to static spatial data mapping between the interface adaptation and the reflection memory network, failing to employ a generation method that couples concurrent network traffic with the actual power grid physical topology fault evolution process. When simulating complex system fault conditions involving concurrent multi-source heterogeneous industrial protocols such as IEC 104 and IEC 61850, existing technologies treat data packets as independent sender-receiver objects, neglecting to consider the dynamic timing characteristics of the power grid physical equipment. This results in generated concurrent test data prone to data conflicts in timestamps and physical states due to the lack of physical spatial propagation order and causal correlation in temporal evolution. When heterogeneous data containing temporal misalignments and spatial inconsistencies enters the target panoramic dynamic verification platform, it is directly filtered or discarded by the target system according to data cleaning principles. This directly results in test traffic generated based on this method failing to effectively reach the verification platform's computing engine, making it impossible to conduct effective testing on heterogeneous interface compatibility and data coupling processing capabilities under high concurrency conditions. Summary of the Invention

[0004] To address the shortcomings of the existing technologies, the technical problem to be solved by this invention is to provide a heterogeneous protocol concurrent testing method based on physical topology and cascading fault modeling. In the panoramic dynamic verification scenario of a large-scale control system for hydro-wind-solar-storage systems with multiple heterogeneous protocols, when simulating large-scale cascading faults caused by local grid anomalies in hydro-wind-solar-storage nodes, the method can automatically generate and inject high-concurrency heterogeneous protocol messages with causal correlation and spatiotemporal synchronization based on the spatial propagation sequence and precise time nodes of the actual grid physical topology. This avoids high-voltage data being filtered by the target verification system due to spatiotemporal semantic conflicts, thereby effectively evaluating the functional completeness and latency determinism of the verification platform.

[0005] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows: The present invention provides a heterogeneous protocol concurrency testing method based on physical topology and cascading failure modeling, comprising the following steps: S1: Obtain device topology data of the target system According to device topology data Construct the state transition probability matrix According to the state transition probability matrix Simulate the state transition process of power grid topology fault evolution; when a topology state transition occurs, execute a Hawkes process to generate initial concurrent network messages. ; S2: For the initial concurrent network packets Perform dynamic semantic destructuring to generate a first-type state message. With the first type of sampling message and the second type of remote control message Calling a multi-dimensional time series matrix With sliding window For the first type of state report Type 1 sampling message and the second type of remote control message Perform timestamp calibration preprocessing to generate joint network test data. ; S3: Will jointly run network test data The data is sent to the panoramic dynamic verification platform for stress testing, and full-lifecycle tracking and sampling are performed by probes deployed at the data inlet and command outlet of the panoramic dynamic verification platform to generate probe evaluation data. ; S4: Based on probe evaluation data Perform quantitative calculations to generate a dynamic set of quantitative indicators. ; S5: Based on the dynamic quantitative indicator set Generate a three-dimensional evaluation matrix , the three-dimensional evaluation matrix The excitation parameters of the Hawkes process in step S1 are determined as feedback parameters and dynamically adjusted to lock the performance threshold node of the target system, thereby executing automated optimization control of the verification pressure.

[0006] In the preferred embodiment, in step S1, the device topology data of the target system is obtained. According to device topology data Construct the state transition probability matrix This includes the following sub-steps: S11: Extract the physical connections and communication routing relationships of the target large-scale hydropower, wind power, solar power, and energy storage control system to construct a topology node set. The topology node set includes power equipment nodes and communication network nodes. Perform graph theory modeling on the topology node set to construct a directed weighted graph, and determine the directed weighted graph as the equipment topology data. ; S12: Extract device topology data For any two adjacent nodes in the system, combining the historical fault sample set of the power system with the relay protection settings, extract the fault frequency statistics of adjacent nodes from the historical fault sample set; extract equipment topology data. Electrical distance and electrical coupling strength coefficient between adjacent nodes and topological relevance weighting coefficient The statistical values ​​of fault frequency of adjacent nodes, electrical distance and electrical coupling strength coefficient are combined. and topological relevance weighting coefficient Perform conditional probability calculations to generate state transition probabilities for adjacent nodes; arrange the state transition probabilities of all adjacent nodes into a matrix to generate the Markov chain space state transition matrix, and then define the Markov chain space state transition matrix as the state transition probability matrix. .

[0007] In the preferred scheme, in step S1, based on the state transition probability matrix... Simulate the state transition process of power grid topology fault evolution; when a topology state transition occurs, execute a Hawkes process to generate initial concurrent network messages. This includes the following sub-steps: S13: Inject initial disturbance sources onto the virtual simulation timeline, and use the Monte Carlo time-step simulation method combined with the state transition probability matrix. The probability of state transitions between nodes is calculated, and the state transition results of the neighboring nodes of the disturbed node are calculated. When the running state word of the neighboring node changes from steady state to transient or fault state, the state transition is recorded and identified as a topology state transition event. ; S14: Extract topology state transition events Timestamp sequence update history trigger time Extract the baseline value of the basic communication background traffic of the target system under steady state. The excitation parameter represents the weighting of the instantaneous surge intensity contribution. and the decay parameter representing the exponential decay rate. ; triggering historical moments Basic communication background traffic baseline value Excitation parameters and attenuation parameters Substituting the mathematical model of a stochastic point process with self-excitation characteristics into the model and performing mapping operations, the current time step is generated. Communication traffic condition strength function According to the communication traffic condition strength function The real-time distribution of data is achieved by calling the network packet sending module to dynamically map and package the corresponding test data payload to output the raw data frame, which is then used as the initial concurrent network packets. .

[0008] In the preferred embodiment, step S2 includes the following sub-steps: S21: Construct a general semantic mapping abstraction layer for multi-source heterogeneous protocols to identify initial concurrent network packets. The protocol types belong to physical devices; for communication nodes adopting substation communication network and system series standards, the abstract syntax notation and basic encoding rules are dynamically deconstructed to parse substation event messages oriented towards general objects, and these general object-oriented substation event messages are identified as the first type of status messages. Parse the sampled value message and identify it as a first-type sampled message. ; S22: For communication nodes using remote control equipment and system transmission protocols, identify the initial concurrent network packets. The content of the application service data unit in the message; based on the logic state machine of the target system, dynamically generate application service data units containing valid time tags and sequence numbers, and identify application service data units containing valid time tags and sequence numbers as the second type of remote control message. ; S23: In the set of virtualized sound-emitting nodes In this paper, a multidimensional time series matrix is ​​constructed to map protocol data elements across multiple time scales for heterogeneous data streams from communication nodes. Define the observation interval that slides along the virtual simulation time axis, and determine the observation interval as the sliding window. ; S24: Extract the inherent inter-module communication delay characteristics within the hardware-in-the-loop system and generate link delay correction coefficients. Extract the first type of status message Type 1 sampling message and the second type of remote control message Original generation time Extract the sliding window The starting reference time and time window length configuration value ; the original generation time Starting reference time and time window length configuration value Perform division and floor operations, combined with the link delay correction factor. Perform timestamp compensation calculations to generate synchronized timestamps after timestamp calibration preprocessing. The synchronization timestamp has been updated. Type 1 status message Type 1 sampling message and the second type of remote control message Data frames are merged and packaged to generate a heterogeneous test data stream that is self-consistent on the timeline. This heterogeneous test data stream is then used as the joint network test data. .

[0009] In the preferred embodiment, the probe evaluation data in step S3 Includes parsing message status data Timestamp data and resource status usage data S3 specifically includes the following sub-steps: S31: Identify the panoramic dynamic verification platform to be tested and designate it as the verification platform system. The interface protocol converter of the hardware-in-the-loop testbed is invoked to jointly run network test data. Injected into the verification platform system via Ethernet interface and Synchronous Digital Architecture (SDA) interface. ; will receive test data from the joint operation network The network port or virtual gateway node is determined as the data entry node. It will be through the data entry node The continuously injected message sequence was identified as the injection stream sequence. ; S32: In the verification platform system Data entry node Deploy entry probes at the location In the verification platform system instruction export node Deploy exit probes at the location Extracting data entry nodes The physical network interface card (NIC) characteristics or port attribute sequence generates a constant salt value used to prevent hash collisions. Extract test data from the joint operation network. Data frame header characteristics of Chinese messages Data frame header features With constant salt value Perform an XOR operation with a secure hash function to generate an independent trace identifier. ; S33: Based on entry probe Monitoring and Verification Platform System The protocol stack receive buffer is used to verify the platform system based on the message. The system assigns a resolution status word to the actual situation where data is successfully received and stored in the corresponding memory area, or is discarded due to concurrency overload or semantic conflicts. ; Independent tracking identifier With the corresponding parsing status word Perform mapping and packaging to generate parsed message status data. ; S34: Based on entry probe Record message arrival verification platform system Precise time generation entry timestamp Based on the exit probe Capture and independent tracking identifiers Associated response commands and record the precise time of their generation to generate an export timestamp. ; Independent tracking identifier Entry timestamp Export timestamp Perform pairing and combination to generate timestamp marker data. ; S35: Based on entry probe With exit probe Read verification platform system The resource consumption state vector is constructed from the CPU utilization, memory heap resident capacity, and number of concurrent thread lock waits. All resource consumption state vectors recorded during the test period Resource status and usage data are generated by assembling data in chronological order. ; S36: Parse message status data Timestamp data and resource status and usage data The data is structured, aggregated, and persistently stored, then encapsulated into a comprehensive evaluation data object with a unified time-aligned index. This comprehensive evaluation data object is then designated as the probe evaluation data. .

[0010] In the preferred scheme, the set of dynamic quantitative indicators in step S4 Including protocol matching success rate Deterministic delay quantization value With state convergence index S4 specifically includes the following sub-steps: S41: Within the set quantitative evaluation time window Internally, read and parse message status data. Includes independent tracking identifiers Extract the parsing state word corresponding to the data tuple. ; Quantitative assessment time window All parsed status words The values ​​are summed to obtain the total number of successfully parsed packets; the total number of successfully parsed packets is then compared with the quantitative evaluation time window. Total number of evaluation window messages injected into the system Perform ratio calculations to generate the protocol matching pass rate. ; S42: Extract timestamp data Entry timestamp Export timestamp For entry timestamp Export timestamp The end-to-end processing delay is calculated. Calling the set kernel density function to process end-to-end latency The probability distribution is smoothly fitted to generate the probability density distribution function. In the probability density distribution function An integration operation is performed over the distribution interval using preset confidence quantiles to generate a deterministic delay quantization value that satisfies the tail delay value. ; S43: Read resource status and usage data Resource consumption state vector in continuous time series Selecting the initial observation time in phase space Extract the initial observation time Corresponding initial state trajectory deviation Tracking deviations in the evolutionary trajectory caused by initial perturbations. Calculate the initial trajectory deviation Deviation from evolutionary state trajectory The exponential separation rate during time evolution generates the state convergence exponent. ; S44: During the quantitative assessment time window At the end, the protocol matching pass rate will be calculated. Deterministic delay quantization value and the state convergence index The data structure is aligned and normalized, combined into a three-dimensional array with timestamp association, and then the three-dimensional array is encapsulated and defined as a set of dynamic quantitative indicators. .

[0011] In the preferred embodiment, step S5 includes the following sub-steps: S51: Extracting the set of dynamic quantitative indicators Protocol matching pass rate included Deterministic delay quantization value With state convergence index Call the range normalization function to calculate the protocol matching pass rate. Deterministic delay quantization value With state convergence index Perform data mapping operations to match protocol success rates. Mapping to match normalized values Quantize the deterministic delay value Mapped to normalized delay value The state convergence exponent Mapped to convergent normalized value Match normalized values Delay normalization value With convergence normalization value Assign orthogonal basis vectors to the evaluation space to construct a matrix feature with a three-dimensional diagonal structure, and then define the matrix feature with the three-dimensional diagonal structure as the three-dimensional evaluation matrix. ; S52: Define a zero matrix representing the ideal, fault-free operating state of the target system under test, and determine the zero matrix as the ideal reference matrix. Extract the three-dimensional evaluation matrix With the ideal reference matrix Construct a difference matrix, perform spatial algebraic distance quantization on the difference matrix to calculate the Frobenius norm, generate a spatial algebraic distance scalar, and determine the spatial algebraic distance scalar as the performance boundary deviation. .

[0012] In the preferred embodiment, the excitation parameters in step S5 include parameters and parameters S5 also includes the following sub-steps: S53: Monitor performance boundary deviation Optimization operation cycle Characteristics of the derivative slope under varying conditions; deviation from performance boundaries When the set upper limit is not reached, extract the excitation parameters that control the intensity of the instantaneous flow surge in the target system. And the excitation parameters that control the flow calming decay rate Extracting performance boundary deviation For excitation parameters The first partial derivative value Performance boundary deviation For excitation parameters The second partial derivative value and the update step size of the scaling multiplier coefficient that controls the gradient adjustment magnitude. Combined with the first partial derivative value The second partial derivative value With update step size The discrete difference method is used for approximate numerical estimation to calculate the excitation parameters for the next operation cycle. With excitation parameters ; S54: Set the activation parameters for the next computation cycle. With excitation parameters The concurrent stress generation engine configuration module of the hardware-in-the-loop co-simulation test platform is synchronized in real time, and the process parameters are hot-swapped to control the concurrent stress generation engine according to the activation parameters of the next calculation cycle. Adjust the peak value of heterogeneous protocol message bursts caused by a single topology state transition event, and adjust it according to the excitation parameters of the next operation cycle. Adjust the duration of the network communication traffic tail over time; continuously read performance boundary deviation. The trend is used to determine the extreme threshold, and the deviation from the performance boundary is considered. When the verification platform system has reached its performance inflection point, it extracts the proportion of various industrial control protocol messages in the current mixed data network stream, and determines the current protocol allocation combination by combining the proportions of various industrial control protocol messages. The concurrent pressure generation engine controls the current protocol ratio combination. Under the constraints of the operating conditions, the high-frequency flow excitation state is maintained according to the optimized excitation parameters, and the extreme capacity level load impact is continuously applied to the verification platform system to expose the shortcomings of the system's scheduling capabilities.

[0013] In a preferred embodiment, the present invention also provides a computer device, the computer device including at least one processor coupled to at least one memory, the memory storing at least one computer program or instruction, wherein the computer program or instruction is loaded and executed by the processor to implement the steps of the heterogeneous protocol concurrent testing method based on physical topology and cascading fault modeling as described in any of the preceding embodiments.

[0014] In a preferred embodiment, the present invention further provides a computer-readable storage medium, wherein a computer program or instructions are stored on the computer-readable storage medium, and when the computer program or instructions are executed by a processor, the steps of the heterogeneous protocol concurrent testing method based on physical topology and cascading fault modeling as described in any of the preceding embodiments are implemented.

[0015] In a preferred embodiment, the present invention further provides a computer program product, including a computer program or instructions, wherein when the computer program or instructions are executed by a processor, they implement the steps of the heterogeneous protocol concurrent testing method based on physical topology and cascading fault modeling as described in any of the preceding embodiments.

[0016] This invention provides a heterogeneous protocol concurrency testing method based on physical topology and cascading failure modeling. Through the coordination between the aforementioned structures, it offers the following advantages compared to existing methods: First, this solution simulates the power grid topology transition process by acquiring equipment topology data and constructing a state transition probability matrix. Combined with Hawkes processes, it generates initial concurrent network messages, ensuring that the generated concurrent test traffic exhibits a causal relationship between the spatial propagation order and temporal evolution of the actual power grid topology. This evolution-driven mechanism ensures that the test messages correspond to the characteristics of real power grid cascade faults, avoiding ineffective pressure injection due to missing constraints and increasing the effectiveness of test cases in panoramic dynamic verification scenarios for large-scale control systems such as hydropower, wind power, solar power, and energy storage.

[0017] Secondly, this solution performs dynamic semantic deconstruction processing on the initial concurrent network packets and invokes a multi-dimensional time-series matrix and sliding window to perform timestamp calibration preprocessing to generate jointly run network test data. This spatiotemporal dynamic synchronization mechanism ensures that multi-source heterogeneous protocol data remains self-consistent in terms of time axis and state, avoiding the situation where high-pressure test data is filtered or discarded by the target verification system due to spatiotemporal semantic conflicts. It guarantees the concurrent generation of high-load heterogeneous data in a non-blocking state, thereby enabling effective evaluation of the verification platform's interface compatibility and data processing capabilities under concurrent conditions.

[0018] Third, this solution employs probe deployment for full lifecycle tracking and sampling, and generates a dynamic set of quantitative indicators based on probe evaluation data. This, in turn, constructs a three-dimensional evaluation matrix to adjust the excitation parameters of the Hawkes process. This multi-dimensional quantitative evaluation mechanism, customized for power industry operations, integrates protocol matching pass rate, end-to-end latency jitter distribution, and system state convergence, performing automated optimization control of verification pressure. This feedback adjustment improves the early warning and positioning accuracy of computational latency jitter and resource state convergence on the panoramic dynamic verification platform, enabling high-fidelity stress testing and operational stability verification of the core computing and scheduling capabilities of large-scale control systems. Attached Figure Description

[0019] The present invention will be further described below with reference to the accompanying drawings and embodiments: Figure 1 This is a main view diagram of the process structure of this invention; Figure 2 A schematic diagram of the structure of the computer device of the present invention. Detailed Implementation

[0020] To better understand the purpose, system architecture, and functional implementation of this embodiment, the embodiments and features described herein can be combined with each other without conflict. The exemplary embodiments disclosed herein will be described below with reference to the accompanying drawings, including specific technical details disclosed to aid understanding; however, these details should be considered exemplary rather than restrictive. Therefore, those skilled in the art should understand that various improvements and adjustments can be made to the embodiments described herein without departing from the scope and core ideas of the invention. Similarly, for clarity, detailed descriptions of well-known technologies, functions, and structures are omitted in the following description.

[0021] In this embodiment, for the panoramic dynamic verification scenario of large-scale control systems such as water, wind, solar and energy storage, existing hardware-in-the-loop simulation systems usually rely on the conversion of basic hardware interfaces and the protocol matching of data buses to provide full-process device integration access.

[0022] Furthermore, the limitations of this type of conventional solution in practical applications are that it is mainly based on basic hardware compatibility and signal connectivity, does not adopt a control mechanism for generating high-concurrency pressure at the application layer, and does not have the ability to dynamically drive testing and evaluation based on system fault evolution.

[0023] In practice, when faced with complex verification requirements involving multiple heterogeneous protocols, existing concurrent testing mechanisms typically employ independent network traffic generators to apply static script pressure. Consequently, this unidirectional network layer pressure injection method, lacking spatial constraints and causal relationships regarding temporal evolution, struggles to effectively test and verify the computational performance and scheduling mechanisms of large-scale control systems when simulating abnormal network conditions.

[0024] Specifically, existing technologies treat data packets as independent send-and-receive objects, without considering the dynamic timing characteristics of power grid equipment. This results in concurrent test data that lacks spatial spread sequence and causal relationships of temporal evolution, easily leading to data conflicts in timestamps and physical states.

[0025] Preferably, in hardware-in-the-loop (HIL) co-simulation, there is a complex interactive coupling between the continuous-time power system and the discrete-event communication network system.

[0026] Preferably, when simulating complex system fault conditions involving multiple heterogeneous industrial protocols such as IEC 104 and IEC 61850, if the test flow is not aligned with the cascaded fault topology of the actual power grid, heterogeneous data containing timing misalignment and spatial inconsistency will be directly filtered or discarded by the target system according to the data cleaning principle when it enters the target panoramic dynamic verification platform.

[0027] Consequently, the test traffic generated by this method is difficult to effectively reach the computing engine of the verification platform, and effective testing of heterogeneous interface compatibility and data coupling processing capabilities under high concurrency conditions cannot be carried out.

[0028] In this embodiment, to overcome the shortcomings of the existing technology, a joint verification model based on physical driving and spatiotemporal alignment is adopted. Specifically, this model does not use the traditional static script playback and simple network layer packet sending mechanism, but instead uses the Markov evolution model and Hawkes process in the real power system as pressure injection sources, so that concurrent network packets have a causal relationship between spatial spread order and temporal evolution. In addition, with the dynamic semantic deconstruction and sliding window mechanism of heterogeneous protocols, multi-source data is spatiotemporally synchronized to avoid high-voltage data being filtered by the target verification system due to spatiotemporal semantic conflicts. Therefore, finally, combined with a customized multi-dimensional comprehensive evaluation matrix, the verification platform is quantitatively evaluated and pressure is automatically optimized in a closed loop.

[0029] Example 1: like Figure 1 As shown, this embodiment provides a heterogeneous protocol concurrency testing method based on topology and cascading failure modeling. The technical solution of this embodiment will be elaborated in detail below with specific steps: S1: Obtain device topology data of the target system According to device topology data Construct the state transition probability matrix According to the state transition probability matrix Simulate the power grid topology transition process; when a topology transition occurs, execute the Hawkes procedure to generate initial concurrent network messages. .

[0030] In practice, step S1 above includes the following sub-steps: S11: Obtain device topology data of the target system .

[0031] Specifically, the input is the basic ledger and network architecture configuration information of the target large-scale water, wind, solar, and storage control system.

[0032] In this embodiment, a set of topological nodes containing physical connection relationships and communication routing relationships is extracted from the database or static model file of the target large-scale water, wind, solar and storage control system.

[0033] The topology node set includes power equipment nodes such as generator sets, transformers, photovoltaic inverters, wind turbine main control units, and energy storage converters, as well as communication network nodes such as Synchronous Digital Hierarchy (SDH) optical transceivers, routers, and switches.

[0034] In practice, the system performs graph theory modeling on the extracted set of topology nodes, constructs a directed weighted graph, and defines the directed weighted graph as device topology data. .

[0035] Furthermore, the improvement of step S11 over the prior art lies in addressing the limitation of conventional IT testing that only focuses on logical IP addresses and port numbers. This embodiment adopts a unified data description for the energy flow topology and the upper-layer information flow communication topology in the Cyber-Physical System (CPS).

[0036] S12: Based on device topology data Construct the state transition probability matrix .

[0037] In this embodiment, when acquiring device topology data Subsequently, based on the historical fault sample set of the power system and the relay protection settings, the equipment topology data was quantitatively evaluated. The probability of fault propagation at each node under different operating conditions.

[0038] Among them, the system extracts device topology data. Any two adjacent nodes and Calculation at node When an abnormal state occurs, the node Follow the conditional probability of a state change.

[0039] Preferably, the above-mentioned state change is a drop from the normal operating state to the low voltage ride-through state, or triggering the overcurrent protection trip state.

[0040] In practice, the system arranges all the calculated conditional probability values ​​into a matrix to generate a Markov chain space state transition matrix, and defines this matrix as the state transition probability matrix. .

[0041] In one feasible approach, during this process, the state transition probability matrix Internal elements Specifically, it is calculated using the following formula: (1); In formula (1), For nodes When a fault occurs, it causes the node to... The transfer probability of a cascading failure; For the historical fault sample set, there are nodes Fault-causing node Fault frequency statistics; For historical fault sample set nodes Total frequency of failures; For device topology data Middle node With nodes The electrical distance and electrical coupling strength coefficient between them; This is the topological relevance weighting coefficient, with a value range of [value range missing]. .

[0042] Furthermore, the improvement in step S12 lies in organically connecting isolated device nodes through a probabilistic model, so that the generated test concurrent behavior is no longer a random discrete event, but a spatially linked event that can follow the Kirchhoff laws of the power grid and the relay protection action logic.

[0043] S13: Based on the state transition probability matrix Simulate the power grid topology transfer process.

[0044] In this embodiment, the Monte Carlo time-stepping simulation method is used to inject an initial disturbance source onto the virtual simulation time axis.

[0045] Preferably, the initial disturbance source is a simulation of a three-phase short-circuit fault occurring on a high-voltage tie line due to extreme weather.

[0046] In practical implementation, based on the state transition probability matrix Transition probabilities between nodes Within each simulation time step, the state transition results of the adjacent nodes of the disturbed node are calculated. Specifically, when the calculation determines that the operating state word of a node changes from a steady state to a transient or fault state, the system records this state transition and defines it as a topology state transition event. .

[0047] Furthermore, the improvement in step S13 lies in its ability to fully reproduce the evolution trajectory of a local fault in the power system leading to a large-scale chain disconnection, filling the gap in the panoramic dynamic verification platform's lack of real extreme operating condition test sequences.

[0048] S14: When a topology state transition is determined, execute the Hawkes procedure to generate initial concurrent network packets. .

[0049] In this embodiment, when the system detects the topology state transition event output in step S13... At that time, the test stress generation engine is activated synchronously. In actual control systems, power grid faults not only trigger single network data packets but also cause a cascading surge in communication traffic.

[0050] Preferably, the aforementioned avalanche-like surge in communication traffic manifests as an abnormality in instantaneous current interruption, generating a massive number of tripping signals within microseconds, and a large number of remote signaling changes and alarm events emerging within milliseconds.

[0051] In one feasible approach, the system introduces a mathematical model of a Hawkes process, a stochastic point process with self-excitation properties, to map this bursty flow behavior.

[0052] In practice, the system relies on topology state transition events. Timestamp sequence update history trigger time And calculate the current time. Communication traffic condition strength function : (2); In formula (2), For the current moment The concurrent message generation rate; The baseline value of the basic communication background traffic of the target system under steady state; For each topology state transition event recorded in step S13 The historical trigger moment that occurred; The excitation parameter of the Hawkes process represents a single topological state transition event. The contribution weight to the intensity of instantaneous surges in communication traffic; is the decay parameter of the Hawkes process, representing the exponential decay rate of communication traffic over time after a fault is triggered.

[0053] Therefore, the test stress generation engine is based on the conditional strength function. The system performs real-time distribution, calls the network packet sending module, dynamically maps and packages the corresponding test data payload, outputs a large number of raw data frames, and defines these raw data frames as initial concurrent network packets. .

[0054] Furthermore, the improvement in step S14 lies in not using the conventional stress testing tool's blind stress-increasing method based on a linearly increasing number of threads; where the linearly increasing number of threads is a step-by-step stress-increasing method.

[0055] Therefore, the concurrent network traffic generated by the Hawkes process model has extremely low normal operating conditions and exponentially explosive failures, resulting in a long-tailed, nonlinear characteristic. This allows the pressure injected into the target verification platform to highly replicate the actual communication avalanche conditions of the hydro-wind-solar-storage system, significantly increasing the effectiveness of stress test cases.

[0056] In one feasible approach, the initial concurrent network packets generated in step S1 above... Perform the following steps S2 for processing: S2: For the initial concurrent network packets Perform dynamic semantic destructuring to generate a first-type state message. With the first type of sampling message and the second type of remote control message Calling a multi-dimensional time series matrix With sliding window For the first type of state report Type 1 sampling message and the second type of remote control message Perform timestamp calibration preprocessing to generate joint network test data. .

[0057] In practice, step S2 above includes the following sub-steps: S21: For the initial concurrent network packets Perform dynamic semantic destructuring to generate a first-type state message. With the first type of sampling message .

[0058] In this embodiment, a general semantic mapping abstraction layer for multi-source heterogeneous protocols is constructed to identify initial concurrent network packets. The protocol type is subordinate to the physical device. For communication nodes adopting the International Electrotechnical Commission (IEC) Substation Communication Networks and Systems Series Standards (IEC 61850), the Abstract Syntax Notation – Basic Encoding Rules (ASN.1 BER) are dynamically deconstructed to parse substation event messages oriented towards general objects. These general-object-oriented substation event messages are defined as Type I status messages. Similarly, high-sampling-rate sampled value messages are parsed out and defined as the first type of sampled message. .

[0059] Preferably, the aforementioned communication nodes are wind power main control nodes and substation nodes.

[0060] The variables used in this step include the initial concurrent network packets. Type 1 status message Type 1 sampling message .

[0061] In addition, step S21 addresses the problem that traditional stress testing tools treat data packets as blindly sent units by using dynamic semantic reconstruction of the industrial protocol stack to prevent messages from being identified as illegal formats by the target system, thereby improving the availability of the test stress.

[0062] S22: For the initial concurrent network packets Perform dynamic semantic destructuring to generate the second type of remote control message. .

[0063] In this embodiment, for communication nodes using the Telemetry and Control System Transmission Protocol (IEC 104), the initial concurrent network packets are identified. The content of the application service data unit. Based on the logic state machine of the target control system, application service data units containing valid time tags and sequence numbers are dynamically generated, and these application service data units containing valid time tags and sequence numbers are defined as the second type of remote control message. .

[0064] Preferably, the aforementioned communication node is a scheduling master station node.

[0065] Preferably, the above-mentioned valid time stamp adopts the CP56Time2a format.

[0066] In addition, step S22 ensures that the generated concurrent traffic meets the parsing specifications of large-scale control systems in terms of protocol semantics, and effectively improves the problem that network replay test data is easily discarded due to sequence number errors.

[0067] S23: Constructing a multidimensional time series matrix With sliding window .

[0068] Specifically, the input is a Type I status message. Type 1 sampling message Type II remote control messages .

[0069] In this embodiment, in the virtualized sound-emitting node set In the middle, targeting nodes Heterogeneous data streams, constructing a multidimensional time series matrix This multidimensional time series matrix This is used to map protocol data elements across multiple time scales. The system defines an observation interval that slides along the virtual simulation time axis and defines this observation interval as the sliding window. .

[0070] Preferably, the aforementioned multidimensional time scale covers both microsecond-level and millisecond-level data scales.

[0071] In addition, step S23 establishes a unified spatiotemporal calibration benchmark to address the timeline disorder problem that is prone to occur when heterogeneous protocols are transmitted concurrently.

[0072] S24: Call the multidimensional timing matrix With sliding window The message undergoes timestamp calibration preprocessing to generate joint operation network test data. .

[0073] In this embodiment, a virtualization mapping mechanism based on the Precision Time Protocol (IEEE 1588) is employed in the multi-dimensional timing matrix. Combined with sliding window For the first type of status message Type 1 sampling message and the second type of remote control message Perform timestamp calibration preprocessing.

[0074] The specific steps are as follows: extract the original generation time of the above three types of heterogeneous messages, and calculate the aligned timestamp under a unified benchmark by combining the hardware-in-the-loop link communication transmission delay compensation term.

[0075] In one feasible approach, the specific numerical calculations for the calibration process are obtained using the following formula: (3); In formula (3), For the first Each message is synchronized with a timestamp after timestamp calibration preprocessing. For the first The original generation time of each message; For sliding windows The starting reference time; For sliding windows The time window length configuration value; This is the link delay correction factor.

[0076] Specifically, a link delay correction coefficient is adopted. The reason is that there is an inherent inter-module communication delay within the hardware-in-the-loop system. Introducing compensation can reduce system errors and improve synchronization timestamps. The system will update the synchronization timestamps to ensure accuracy. Type 1 status message Type 1 sampling message and the second type of remote control message Data frames are merged and packaged to generate a heterogeneous test data stream that is self-consistent on the timeline, and this heterogeneous test data stream is defined as the jointly run network test data. .

[0077] Therefore, the output of the joint operation network test data It will serve as a test stimulus source for sending to the panoramic dynamic verification platform in step S3.

[0078] Furthermore, step S24 improves upon the technique used in multi-protocol data interface compatibility verification where protocol parsing is easy to complete but timing coupling is difficult. After timestamp calibration preprocessing, it ensures that the concurrent data received by the test platform is absolutely synchronized in time and logic, preventing dirty data filtering caused by time misalignment and enabling test traffic to effectively reach the verification platform's computing engine.

[0079] In this embodiment, the test data of the jointly operated network output in step S2 is used as an example. Then, perform the following steps S3 for processing.

[0080] To address the shortcomings of existing technologies in effectively capturing concurrent latency and parsing status of heterogeneous protocols, this embodiment does not adopt the traditional unidirectional network layer pressure injection test mode. Instead, it introduces a hardware-in-the-loop co-simulation architecture and combines it with non-intrusive probe technology to perform end-to-end monitoring.

[0081] S3: Will jointly run network test data The data is sent to the panoramic dynamic verification platform for stress testing, and full-lifecycle tracking and sampling are performed by probes deployed at the data inlet and command outlet of the panoramic dynamic verification platform to generate probe evaluation data. Probe evaluation data Includes parsing message status data Timestamp data and resource status usage data .

[0082] In practice, step S3 above includes the following sub-steps: S31: Will jointly run network test data Send to the verification platform system .

[0083] Specifically, the input is the joint operation network test data generated in step S2. .

[0084] In this embodiment, the panoramic dynamic verification platform under test is determined and defined as the verification platform system. Verification platform system It covers both the main station control center node and the substation execution nodes. The system calls the hardware-in-the-loop testbed's interface protocol converter, based on the System-in-the-loop (SITL) virtualization mapping mechanism, to convert jointly running network test data containing multiple heterogeneous protocols. Injected into the verification platform system via Ethernet interface and Synchronous Digital Hierarchy (SDH) optical transceiver. .

[0085] In practice, the system will receive test data from the joint operation network. The network port or virtual gateway node is defined as the data entry node. During this injection process, network test data is run concurrently. Based on the extreme nonlinear long-tail characteristics and topological evolution order generated by the Hawkes process in the preceding step S1, the verification platform system is... This creates high concurrency pressure.

[0086] The system will access data through the data entry node. A continuously injected message sequence is defined as an injection stream sequence. .

[0087] The variables used in this step include joint network test data. Verification platform system Data entry node Injection stream sequence .

[0088] Furthermore, step S32 addresses the deficiency in conventional testing that does not employ an application-layer high-concurrency pressure generation mechanism. By introducing a hardware-in-the-loop interface and system-in-the-loop mapping mechanism, concurrent test messages generated at the application layer can be directly used in the communication links of real control devices, solving the problem that heterogeneous protocol messages cannot be transmitted at high concurrency in real-world channels.

[0089] S32: In the verification platform system Deploy entry probes in the middle With exit probe And establish a tracking and identification mechanism. .

[0090] Specifically, the input is the verification platform system. Data entry node Injection stream sequence .

[0091] In this embodiment, in the verification platform system Data entry node At this point, a data packet capture and status parsing proxy module is deployed in bypass mode, and this proxy module is defined as the entry probe. Meanwhile, in verifying the platform system Another agent module is deployed at the output node that sends out power grid control commands or protection action signals, and the system defines this output node as the command output node. The agent module deployed there is defined as an exit probe. .

[0092] Preferably, the aforementioned power grid control commands include generator trip commands, energy storage node disconnection commands, or load unloading commands.

[0093] In one feasible approach, for the injection stream sequence The massive amount of heterogeneous concurrent messages can easily lead to data corruption and mismatch issues. Therefore, the system establishes a tracking and identification mechanism based on hash mapping. This tracking and identification mechanism For each person entering the verification platform system The message generates a globally unique identity sequence number.

[0094] Specifically, tracking and identification mechanism The specific identifier calculation is achieved through the following formula: (4); In formula (4), For tracking and identification mechanisms For the injection stream sequence The Middle Each message generates an independent tracking identifier; For secure hash functions; For joint operation of network test data The Middle The data frame header characteristics of a message; It is the XOR operator; A constant salt value to prevent hash collisions.

[0095] S33: Based on entry probe Perform semantic state sampling to generate parsed message state data. .

[0096] In this embodiment, when the injected stream sequence After data entry node At that time, the inlet probe Real-time reading of message data and monitoring and verification of the platform system. The protocol stack receive buffer determines whether the message has been successfully parsed and stored.

[0097] In practice, if the message passes the syntax check and is verified by the platform system... Normal reception and storage in the corresponding memory area, entry probe Assignment to resolve state word If a message is discarded by filtering rules due to concurrency overload or semantic conflicts, the ingress probe... Assignment to resolve state word The system will track identifiers independently. With the corresponding parsing status word The data is mapped and packaged to form a dataset containing the acceptance status of multi-source heterogeneous protocols, and this dataset is defined as parsed message state data. .

[0098] Therefore, the output parsed message status data Will be incorporated into probe evaluation data In step S4, the heterogeneous protocol adaptation pass rate is calculated.

[0099] S34: Based on entry probe With exit probe Perform end-to-end time consumption sampling and generate timestamp data. .

[0100] In this embodiment, the inlet probe At the data entry node The system records the arrival of messages to the verification platform. The precise time, and define that time as the entry timestamp. At the same time, verify the platform system. Execute internal computation logic based on the successfully parsed message, and at the instruction exit node... Generate the corresponding business response instructions.

[0101] In practical implementation, the exit probe Listening command exit node When captured with an independent tracking identifier When a related response command is received, the exact time when the command was generated is recorded and defined as the exit timestamp. The system will track identifiers independently. Entry timestamp Export timestamp The time series data are paired and combined to form a time series set that reflects end-to-end delay characteristics, and this time series set is defined as timestamped data. .

[0102] The variables used in this step include the injection stream sequence. , entrance probe Command export node Exit probe Independent tracking identifier Entry timestamp Export timestamp Timestamp data .

[0103] Specifically, the output is timestamp-marked data. .

[0104] Therefore, the output timestamp data Will be incorporated into probe evaluation data This is used in step S4 to calculate the deterministic delay quantization value.

[0105] Furthermore, step S34 addresses the shortcomings of traditional tests that cannot observe the internal processing time of the system. It achieves accurate digital sampling of computation time in a non-blocking state, truly reflecting the verification platform system. Time determinism in a hardware-in-the-loop simulation environment.

[0106] S35: Based on the probe, perform high-frequency sampling of the hardware device's operating status to generate resource status occupancy data. .

[0107] Specifically, the input is the verification platform system. Injection stream sequence .

[0108] In this embodiment, the platform system is verified during a long-term high-concurrency stress test. The computing units need to continuously allocate and reclaim system resources. (Ingress probe) With exit probe An embedded resource monitoring script verifies the platform system at fixed sampling intervals. The hardware operating status is read.

[0109] In practice, at each sampling time Probe reading verification platform system The CPU utilization, memory heap resident capacity, and number of concurrent thread lock waits are measured, and a state space feature vector is constructed based on these parameters. This feature vector is then defined as the resource consumption state vector. The system will record all resource consumption state vectors during the test period. The data is assembled sequentially to form a matrix, which is then defined as resource status occupancy data. .

[0110] Therefore, the output resource status occupancy data Will be incorporated into probe evaluation data In step S4, the system state convergence index is calculated.

[0111] In addition, step S35 addresses the shortcomings of existing IT stress testing that only focuses on network layer throughput while ignoring the evolution of physical machine states, providing data support for evaluating and verifying the long-term stability of the platform under cascading failure concurrency conditions.

[0112] S36: Aggregate sampling data from various dimensions to generate probe evaluation data. .

[0113] In this embodiment, the system outputs the parsed message status data in step S33. The timestamp data output in step S34 and the resource status and usage data output in step S35. The system performs structured aggregation and persistent storage processing. Through a data bus, it encapsulates these three types of feature data into a comprehensive evaluation data object with a unified time-aligned index, and defines this comprehensive evaluation data object as probe evaluation data. .

[0114] Among them, probe evaluation data The internal structure can be represented as: (5); In formula (5), Evaluation data for the generated probes; To parse message status data; Use timestamps to mark data; This is data on resource status and usage.

[0115] Therefore, the output probe evaluation data The data source will be identified and input into step S4 to perform dynamic quantitative indicator set calculations customized for power business, supporting the quantitative evaluation of the calculation and scheduling capabilities of the panoramic dynamic verification platform for water, wind, solar and energy storage.

[0116] In this embodiment, the probe evaluation data output in step S3 is... The following steps, S4, are performed for processing. In the panoramic dynamic verification scenario of a large-scale control system for water, wind, solar and energy storage with multiple heterogeneous protocols, traditional IT software stress testing usually only focuses on network throughput (TPS) or average response time, without adopting a spatiotemporal consistency and functional completeness evaluation method oriented towards the entire life cycle of control commands.

[0117] To address the shortcomings of traditional testing methods in effectively reflecting the true capabilities of devices in handling extreme avalanche traffic under a fusion information system (CPS) environment, this embodiment constructs a multi-dimensional comprehensive evaluation matrix.

[0118] S4: Based on probe evaluation data Perform quantitative calculations to generate a dynamic set of quantitative indicators. Dynamic quantitative indicator set Including protocol matching success rate Deterministic delay quantization value With state convergence index .

[0119] In practice, step S4 above includes the following sub-steps: S41: Based on parsing message status data Perform acceptance rate statistics and calculate the protocol matching pass rate. .

[0120] Specifically, the input is the probe evaluation data generated in step S3. Parsing message status data .

[0121] In this embodiment, within the set quantitative evaluation time window Inside, the system reads and parses message status data. All containing independent tracking identifiers The system extracts the parsing status word corresponding to each message from the data tuple. The system uses a quantitative evaluation time window. All parsed status words The values ​​are summed to obtain the verification platform system data. The total number of heterogeneous packets successfully parsed and persistently stored. Subsequently, the system compares this total number of successfully parsed packets with a quantitative evaluation time window. Total number of evaluation window messages injected into the system Perform a ratio calculation and define the calculated ratio result as the protocol matching pass rate. .

[0122] In one feasible approach, the protocol matching pass rate The specific calculation is achieved through the following formula: (6); In formula (6), For protocol matching success rate; For quantitative assessment time window Internally composed of an inlet probe Total number of captured evaluation window messages; For the first The parsing status word of each message, when the message is verified by the platform system. The value is 1 when the message is successfully parsed and accepted, and 0 when the message is filtered due to semantic conflict or concurrency blocking.

[0123] Therefore, the output protocol matching pass rate It will be packaged and entered into the dynamic quantitative indicator set. In step S5, the injection ratio of the heterogeneous protocol is adjusted.

[0124] In addition, step S41 addresses the issue that data from multi-source heterogeneous protocols (IEC 104 and IEC 61850) is easily discarded by the target verification system according to the cleaning principle during high-load concurrency, by providing a quantitative benchmark for the data interface compatibility and throughput acceptance capability of the verification platform, so as to avoid false high-concurrency injection that could lead to test failure.

[0125] S42: Mark data according to timestamp Perform kernel density fitting to calculate and generate deterministic time delay quantization values. .

[0126] Specifically, the input is probe evaluation data. Timestamp data in .

[0127] In this embodiment, during the simulation of a cascading fault in the power system, network communication traffic experiences an avalanche-like surge. For services such as power grid relay protection and automatic voltage control (AVC), the jitter distribution of latency is more decisive than the average absolute latency. The system extracts timestamp data. All entry timestamps Export timestamp Calculate the end-to-end processing delay for each pair of corresponding data. In order to obtain the verification platform system under high concurrency conditions. To address the time deterministic boundary, the system employs the kernel density estimation (KDE) model, a nonparametric estimation method, to determine the end-to-end processing delay. A smooth fit is performed on the true probability distribution to generate the probability density distribution function. .

[0128] In practical implementation, the probability density distribution function The calculation formula is as follows: (7); In formula (7), Let be the probability density distribution function of the end-to-end processing delay; The time delay is an independent variable; The total number of evaluation window messages that successfully parsed and generated response instructions; For the first End-to-end processing latency of a message; The kernel density function; The window smoothing bandwidth is used to control the smoothness of the fitted curve.

[0129] Preferably, the kernel density function A Gaussian kernel function is used.

[0130] Subsequently, the system in the probability density distribution function Perform integration over the distribution interval to calculate the quantiles that satisfy a specific confidence level. The tail delay value is determined and defined as the deterministic delay quantization value. .

[0131] Specifically, deterministic delay quantization value The following integral equation is satisfied: (8); In formula (8), To obtain the deterministic delay quantization value; It is the probability density distribution function; These are the preset confidence quantiles.

[0132] Preferably, confidence quantiles The value is set to 0.999 (i.e., 99.9%).

[0133] Therefore, the output deterministic delay quantization value It will be packaged and entered into the dynamic quantitative indicator set. In step S5, the timing pressure adjustment feedback is provided.

[0134] In addition, step S42 addresses the technical characteristics of large-scale hydro-wind-solar-storage control systems being extremely sensitive to delays by using a statistical model to solve the deterministic delay boundary under extreme operating conditions, thus resolving the problem that conventional testing methods cannot assess the worsening trend of control command jitter.

[0135] S43: Based on resource status and usage data Perform nonlinear dynamic evolution calculations to generate state convergence indices. .

[0136] Specifically, the input is probe evaluation data. Resource status and usage data .

[0137] In this embodiment, during the long-cycle topology cascading failure verification, the verification platform system... The allocation and release of internal resources are in a highly dynamic, non-linear state. The system reads resource status and occupancy data. Resource consumption state vector in continuous time series Instead of using instantaneous peak CPU or memory percentages as the criterion, the system introduces the Largest Lyapunov Exponent from dynamical system chaos theory to determine whether the system has the ability to converge and operate stably over a long period of time under high-pressure testing.

[0138] In practice, the system selects the initial observation time in phase space. Extract the initial state trajectory deviation at that moment. As the virtual simulation progresses, the system continuously tracks the deviation in the evolutionary state trajectory caused by the initial disturbance. The system calculates the exponential separation rate of the orbit during its time evolution to generate a value that determines the direction of resource state evolution, and defines this value as the state convergence exponent. .

[0139] In one feasible approach, the state convergence exponent The specific calculation formula is as follows: (9); In formula (9), The state convergence index; This represents the maximum computational length for evolutionary computation. For at any time The Euclidean norm of the deviation of the evolutionary state trajectory of the tracked resource consumption state vector; Initial observation time The Euclidean norm of the initial state trajectory deviation.

[0140] The system is based on the obtained state convergence index For the verification platform system Quantitatively assess health status: when When this occurs, it indicates that the system resource overhead gradually converges, demonstrating stability under long-term high load operation; when This indicates that the system resource consumption trajectory is diverging, suggesting a potential memory leak or thread deadlock.

[0141] Therefore, the output state convergence index It will be packaged and entered into the dynamic quantitative indicator set. In step S5, this is a key control condition for stopping pressurization or triggering an abnormal warning.

[0142] Furthermore, this sub-step addresses the shortcomings of conventional resource monitoring methods, which tend to overlook latent defects during long-term operation. By introducing a nonlinear dynamic model, it is possible to accurately detect resource depletion trends under concurrent pressure injection.

[0143] S44: Integrate quantitative evaluation results from various dimensions to generate a dynamic set of quantitative indicators. .

[0144] Specifically, the input is the protocol matching pass rate generated in step S41. The deterministic delay quantization value generated in step S42 and the state convergence index generated in step S43 .

[0145] In this embodiment, within the same quantitative evaluation time window At the end, the system performs data structure alignment and normalization on the feature parameters obtained independently by the multi-dimensional algorithm. The system then uses the protocol matching pass rate, which represents compatibility, as an indicator. Deterministic delay quantization value representing time determinism And the state convergence index, which represents long-term stability. Combined into a three-dimensional array with timestamp association, and this three-dimensional array is packaged and defined as a dynamic quantitative indicator set. .

[0146] In one feasible approach, a dynamic quantitative set of indicators The algebraic expression of it is as follows: (10); In formula (10), In order to conduct quantitative assessment within the time window An internally generated set of dynamic quantitative indicators; For protocol matching success rate; This is a deterministic delay quantization value; The convergence index is denoted by .

[0147] Therefore, the output dynamic quantitative index set The input data identified as the basis for full-process feedback control is input into step S5. The system will then use the dynamic quantitative indicator set... The process involves constructing a three-dimensional evaluation matrix and dynamically adjusting the excitation parameters of the Hawkes process in the preceding step S1, forming a complete cyclical processing flow that includes four stages: dynamic generation, injection testing, quantitative evaluation, and parameter optimization.

[0148] In this embodiment, the dynamic quantitative index set output in step S4 is... Then, proceed with step S5 for processing. Before going into the specific technical steps, first clarify the technical problem and solution approach addressed in this step.

[0149] To address the shortcomings of existing technologies that use independent network traffic generators for static script stress application, which makes it difficult to effectively verify the computational performance and scheduling mechanisms of large-scale control systems, this embodiment, based on autonomous computing theory, constructs a feedback control mechanism for verifying stress in a panoramic stress test of an information fusion environment. Instead of employing traditional testing methods that involve manually and blindly adjusting packet transmission frequencies, this embodiment uses the multi-dimensional quantitative evaluation results generated in step S4 as feedback input to establish an automated computational model capable of adaptively tracing and verifying the performance inflection points of the platform.

[0150] S5: Based on the dynamic quantitative indicator set Generate a three-dimensional evaluation matrix According to the three-dimensional evaluation matrix Adjust the excitation parameters of the Hawkes process in step S1, and perform automated optimization control of the verification pressure. The excitation parameters include the excitation parameters... and excitation parameters .

[0151] In practice, step S5 above includes the following sub-steps: S51: Based on the dynamic quantitative indicator set Perform normalization mapping to generate a three-dimensional evaluation matrix. .

[0152] Specifically, the input is the set of dynamic quantitative indicators generated in step S4. .

[0153] In this embodiment, the system extracts a set of dynamic quantitative indicators. Protocol matching pass rate included Deterministic delay quantization value With state convergence index To address the issue of the three independent quantification indicators having different dimensions and convergence directions, the system introduces a range standardization function to perform a data mapping operation. This mapping operation is used to avoid the interference of dimensional differences on the matrix space distance calculation results. The system will then determine the protocol matching pass rate. Mapping to match normalized values Quantize the deterministic delay value Mapped to normalized delay value and the state convergence exponent Mapped to convergent normalized value .

[0154] Preferably, the range standardization function is a linear minimax mapping transformation function matrix, or a nonlinear Sigmoid function family.

[0155] In practice, the system will use the normalized matching value obtained from the mapping calculation. Delay normalization value With convergence normalization value By assigning orthogonal basis vectors to the evaluation space, a matrix feature with a three-dimensional diagonal structure is constructed, and this matrix feature is defined as the three-dimensional evaluation matrix. .

[0156] In one feasible approach, a three-dimensional evaluation matrix The specific construction formula is as follows: (11); In formula (11), the three-dimensional evaluation matrix To construct a state algebraic representation of the real-time coordinates of the verification platform system in three-dimensional space; to match normalized values. To improve the protocol matching success rate Numerical features after mapping processing; time delay normalized value Quantization value of deterministic delay Numerical features after mapping processing; convergent normalized values The state convergence exponent Numerical features after mapping processing.

[0157] Therefore, the output three-dimensional evaluation matrix The matrix norm and deviation will be used in step S52 for calculation.

[0158] In addition, step S51 improves upon the shortcomings of conventional testing methods that rely solely on a single discrete performance index for rough evaluation by establishing a multi-dimensional comprehensive digital characterization model, enabling discrete platform operating parameters to have a unified spatial algebraic representation.

[0159] S52: Based on a three-dimensional evaluation matrix Perform matrix norm calculation to generate performance boundary deviation. .

[0160] In this embodiment, under the high-concurrency verification scenario of multi-source heterogeneous protocols, the computational resource consumption state and network response efficiency of the verification platform system exhibit a non-linear coupling relationship. When simulating a large-scale cascading failure of hydro-wind-solar-storage nodes caused by a local power grid anomaly, the massive influx of data will lead to queuing and congestion of the system's internal processing threads. The system sets a zero matrix representing the ideal fault-free operating state of the test target system and defines this zero matrix as the ideal baseline matrix. The system evaluates the three-dimensional evaluation matrix. With the ideal reference matrix The difference matrix between them is used to perform spatial algebraic distance quantization.

[0161] Preferably, the spatial algebraic distance quantization operation employs the Frobenius norm algorithm or the spectral norm algorithm.

[0162] In practice, the system defines the calculated spatial algebraic distance scalar as the performance boundary deviation. Performance boundary deviation The value indicates the degree of overall health degradation of the current verification platform under extreme concurrent stress load.

[0163] In one feasible approach, when employing the Frobenius norm algorithm, the performance boundary deviation is... The specific calculation formula is as follows: (12); In formula (12), the deviation from the performance boundary is... A distance scalar representing the degree of degradation at the current system operating boundary; a three-dimensional evaluation matrix. The currently acquired spatial feature matrix; the ideal reference matrix. This is an ideal reference matrix in which all diagonal elements are zero. The Frobenius norm operator for matrices; matches normalized values. Three-dimensional evaluation matrix The first diagonal element; the time delay normalization value Three-dimensional evaluation matrix The second diagonal element; convergent normalized value Three-dimensional evaluation matrix The third diagonal element.

[0164] Therefore, the deviation of the output performance boundary The identified feedback control signal is passed to step S53 to guide the update of process parameters in the basic communication packet sending mathematical model.

[0165] In addition, step S52 transforms the complex spatiotemporal digital characteristics of the verification platform into a continuous monotonic scalar signal, solving the problem that the performance inflection point of the target system is difficult to be accurately perceived and mathematically described under high dynamic pressure.

[0166] S53: Based on performance boundary deviation Perform parameter optimization calculations to adjust the excitation parameters of the Hawkes process. With excitation parameters .

[0167] In this embodiment, based on the performance boundary deviation... Based on the feedback of numerical changes, the system executes a gradient ascent operation with automated optimization features. This gradient ascent operation is used to dynamically adjust the Hawkes process mathematical model used in the preceding step S1 to simulate the burst characteristics of avalanche communication traffic. According to the principle of autonomous computing, the system continuously monitors the deviation from the performance boundary. During the optimization operation cycle Characteristics of the slope of the derivative under varying conditions.

[0168] Preferably, the principle of autonomous computing involves a mathematical feedback architecture that enables industrial network systems to possess self-awareness and autonomous state optimization.

[0169] In practical implementation, when the performance boundary deviation is... When the set upper limit has not been reached and there is still room for increase, the system increases the value of the parameter controlling the intensity of the instantaneous flow surge according to the set update step size, while decreasing the value of the parameter controlling the flow calming and decay rate. The system defines the process parameter controlling the intensity of the instantaneous flow surge as the excitation parameter. The process parameter controlling the rate of flow calming and decay is defined as the excitation parameter. The system extracts the deviation from the performance boundary. The corresponding partial derivative values ​​generate the parameter settings for the next calculation cycle.

[0170] In one feasible approach, the gradient-based update formula is specifically as follows: (13); (14); In formulas (13) and (14), the excitation parameters are... The excitation parameters for the next computation cycle generated after the update calculation. Excitation parameters For the current number Excitation parameters implemented during the second optimization calculation cycle Excitation parameters The excitation parameters for the next computation cycle generated after the update calculation. Excitation parameters For the current number Excitation parameters implemented during the second optimization calculation cycle Update step size The scaling multiplier coefficient is used to control the gradient adjustment magnitude; the first partial derivative value... Deviation from performance boundary For excitation parameters The gradient partial derivative values; the second partial derivative values Deviation from performance boundary For excitation parameters The gradient partial derivative values.

[0171] Preferably, the value of the first partial derivative is... With the second partial derivative value The numerical calculation uses the discrete difference method within two adjacent optimization calculation cycles to perform approximate numerical estimation.

[0172] Therefore, the updated excitation parameters are output. With excitation parameters This will be input into step S54, which is used to perform the critical parameter overwrite operation of the concurrent stress generation engine.

[0173] Furthermore, step S53 addresses the shortcomings of conventional software stress testing, which uses a linearly increasing number of threads to apply fixed pressure. By introducing an automatic adjustment mechanism for Hawkes process parameters with a real power grid topology mapping, the generated network communication pressure can autonomously focus on the vulnerable computational components of the system under test.

[0174] S54: Update the excitation parameters With excitation parameters Inject a concurrent pressure generation engine to perform automated optimization control of the verification pressure.

[0175] In this embodiment, the system will calculate the obtained next-generation excitation parameters. With excitation parameters The concurrent stress generation engine configuration module of the underlying hardware-in-the-loop co-simulation test platform is synchronized in real time to perform hot-swap of process parameters. After receiving and loading new parameters, the stress test engine executes the test according to the latest excitation parameters. Control the peak generation of heterogeneous protocol messages caused by a single topology fault state transition event, and according to the latest excitation parameters. The duration of the tail of the avalanche flow distribution over time is determined.

[0176] In practice, the system continuously reads the performance boundary deviation. The trend. When determining the deviation from the performance boundary. When the verification platform system has reached its performance inflection point, indicating the limits of its processing capacity, the test engine extracts the proportion of various industrial control protocol messages in the current mixed data network stream and defines the combination of these proportions as the current protocol configuration. .

[0177] Preferably, the current protocol ratio combination Its internal components include 70% IEC 104 remote control protocol messages and 30% GOOSE substation event protocol messages for general objects.

[0178] At this point, the concurrent pressure generation engine is locking the current protocol ratio combination. Under the constraints of the operating conditions, a high-frequency flow excitation state is maintained according to the optimized excitation parameters. This operation can sustainably and precisely apply extreme capacity-level load impacts to the verification platform, revealing the system's shortcomings under the interplay of specific protocol combinations and avalanche pressure, thereby achieving automated optimization control of the verification pressure.

[0179] Therefore, by establishing a feedback control system that extends from protocol parsing to dynamic traffic generation and adaptive adjustment of feedback parameters, this embodiment forms an intelligent and automated verification process targeting the performance boundaries of the verification platform. This execution mechanism ensures that the panoramic dynamic verification platform can demonstrate functional deficiencies and latency jitter risks under simulated complex multi-source concurrent cascading failure conditions, thus completing the stress test task on the scheduling capability of the target test object.

[0180] Example 2: Further explanation in conjunction with Example 1, such as Figure 2 The structure shown. Figure 2 A schematic diagram of the structure of a computer device provided in an embodiment of this application. The computer device includes: Processor, memory, communication bus, and computer programs stored in memory that can run on the processor.

[0181] The processor can call a computer program in memory, and when executing the program, implement the heterogeneous protocol concurrency testing method based on physical topology and cascading fault modeling provided in the above embodiments. The method includes: S1: acquiring device topology data of the target system. According to device topology data Construct the state transition probability matrix According to the state transition probability matrix Simulate the state transition process of power grid topology fault evolution; when a topology state transition occurs, execute a Hawkes process to generate initial concurrent network messages. ; S2: For the initial concurrent network packets Perform dynamic semantic destructuring to generate a first-type state message. With the first type of sampling message and the second type of remote control message Calling a multi-dimensional time series matrix With sliding window For the first type of state report Type 1 sampling message and the second type of remote control message Perform timestamp calibration preprocessing to generate joint network test data. ; S3: Will jointly run network test data The data is sent to the panoramic dynamic verification platform for stress testing, and full-lifecycle tracking and sampling are performed by probes deployed at the data inlet and command outlet of the panoramic dynamic verification platform to generate probe evaluation data. ; S4: Based on probe evaluation data Perform quantitative calculations to generate a dynamic set of quantitative indicators. ; S5: Based on the dynamic quantitative indicator set Generate a three-dimensional evaluation matrix , the three-dimensional evaluation matrix The excitation parameters of the Hawkes process in step S1 are determined as feedback parameters and dynamically adjusted to lock the performance threshold node of the target system, thereby executing automated optimization control of the verification pressure.

[0182] Furthermore, computer equipment also includes: The Communications Interface (CI) is used for communication between the memory and the processor.

[0183] The memory may include high-speed RAM, and may also include non-volatile memory, such as at least one disk drive.

[0184] If the memory, processor, and communication interface are implemented independently, they can be interconnected via a bus to communicate with each other. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of representation, Figure 2 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0185] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0186] Display devices are used to display images, videos, etc. Display devices may include display panels, which may employ liquid crystal displays (LCDs), organic light-emitting diodes (OLEDs), active-matrix organic light-emitting diodes (AMOLEDs), flexible light-emitting diodes (FLEDs), MiniLEDs, MicroLEDs, Micro-OLEDs, quantum dot light-emitting diodes (QLEDs), etc.

[0187] Alternatively, in a specific implementation, if the memory, processor, and communication interface are integrated on a single chip, then the memory, processor, and communication interface can communicate with each other through an internal interface.

[0188] On the other hand, embodiments of this application also provide a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the above-described heterogeneous protocol concurrent testing method based on physical topology and cascading fault modeling. The method includes: S1: acquiring device topology data of the target system. According to device topology data Construct the state transition probability matrix According to the state transition probability matrix Simulate the state transition process of power grid topology fault evolution; when a topology state transition occurs, execute a Hawkes process to generate initial concurrent network messages. ; S2: For the initial concurrent network packets Perform dynamic semantic destructuring to generate a first-type state message. With the first type of sampling message and the second type of remote control message Calling a multi-dimensional time series matrix With sliding window For the first type of state report Type 1 sampling message and the second type of remote control message Perform timestamp calibration preprocessing to generate joint network test data. ; S3: Will jointly run network test data The data is sent to the panoramic dynamic verification platform for stress testing, and full-lifecycle tracking and sampling are performed by probes deployed at the data inlet and command outlet of the panoramic dynamic verification platform to generate probe evaluation data. ; S4: Based on probe evaluation data Perform quantitative calculations to generate a dynamic set of quantitative indicators. ; S5: Based on the dynamic quantitative indicator set Generate a three-dimensional evaluation matrix , the three-dimensional evaluation matrix The excitation parameters of the Hawkes process in step S1 are determined as feedback parameters and dynamically adjusted to lock the performance threshold node of the target system, thereby executing automated optimization control of the verification pressure.

[0189] In another aspect, embodiments of this application also provide a computer program product, which includes a computer program that can be stored on a computer-readable storage medium. The computer program can execute computer instructions. When the computer program is executed by a processor, the computer can execute the heterogeneous protocol concurrency testing method based on physical topology and cascading fault modeling provided by the above methods. This method includes: S1: acquiring device topology data of the target system. According to device topology data Construct the state transition probability matrix According to the state transition probability matrix Simulate the state transition process of power grid topology fault evolution; when a topology state transition occurs, execute a Hawkes process to generate initial concurrent network messages. ; S2: For the initial concurrent network packets Perform dynamic semantic destructuring to generate a first-type state message. With the first type of sampling message and the second type of remote control message Calling a multi-dimensional time series matrix With sliding window For the first type of state report Type 1 sampling message and the second type of remote control message Perform timestamp calibration preprocessing to generate joint network test data. ; S3: Will jointly run network test data The data is sent to the panoramic dynamic verification platform for stress testing, and full-lifecycle tracking and sampling are performed by probes deployed at the data inlet and command outlet of the panoramic dynamic verification platform to generate probe evaluation data. ; S4: Based on probe evaluation data Perform quantitative calculations to generate a dynamic set of quantitative indicators. ; S5: Based on the dynamic quantitative indicator set Generate a three-dimensional evaluation matrix , the three-dimensional evaluation matrix The excitation parameters of the Hawkes process in step S1 are determined as feedback parameters and dynamically adjusted to lock the performance threshold node of the target system, thereby executing automated optimization control of the verification pressure.

[0190] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus or device (such as a computer-based system, a processor-included system or other system that can fetch and execute instructions from, an instruction execution system, apparatus or device).

[0191] For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit a program for use in or in conjunction with an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing. Furthermore, a computer-readable medium can even be paper or other suitable media on which the program can be printed, since the program can be obtained electronically by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in a computer memory.

[0192] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0193] The program code used to implement the methods of this disclosure may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0194] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as a data server), or computing systems that include middleware components (e.g., an application server), or computing systems that include frontend components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., a communication network). Examples of communication networks include local area networks (LANs), wide area networks (WANs), and the Internet.

[0195] Computer systems can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. Client-server relationships are created by computer programs running on the respective computers and having a client-server relationship with each other.

[0196] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in this disclosure can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution disclosed in this disclosure can be achieved, and this is not limited herein.

[0197] The specific embodiments described above do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of this application.

Claims

1. A heterogeneous protocol concurrency testing method based on physical topology and cascading failure modeling, characterized in that, Includes the following steps: S1: Obtain device topology data of the target system According to device topology data Construct the state transition probability matrix According to the state transition probability matrix Simulate the state transition process of power grid topology fault evolution; When a topology state transition is determined, a Hawkes procedure is executed to generate initial concurrent network packets. ; S2: For the initial concurrent network packets Perform dynamic semantic destructuring to generate a first-class state message. With the first type of sampling message and the second type of remote control message Calling a multi-dimensional time series matrix With sliding window For the first type of state report Type 1 sampling message and the second type of remote control message Perform timestamp calibration preprocessing to generate joint network test data. ; S3: Will jointly run network test data The data is sent to the panoramic dynamic verification platform for stress testing, and full lifecycle tracking and sampling are performed using probes deployed at the data inlet and command outlet of the panoramic dynamic verification platform to generate probe evaluation data. ; S4: Based on probe evaluation data Perform quantitative calculations to generate a dynamic set of quantitative indicators. ; S5: Based on the dynamic quantitative indicator set Generate a three-dimensional evaluation matrix , the three-dimensional evaluation matrix The excitation parameters of the Hawkes process in step S1 are determined as feedback parameters and dynamically adjusted to lock the performance threshold node of the target system, thereby executing automated optimization control of the verification pressure.

2. The heterogeneous protocol concurrency testing method based on physical topology and cascading failure modeling as described in claim 1, characterized in that, In step S1, the device topology data of the target system is obtained. According to device topology data Construct the state transition probability matrix This includes the following sub-steps: S11: Extract the physical connections and communication routing relationships of the target large-scale hydropower, wind power, solar power, and energy storage control system to construct a topology node set. The topology node set includes power equipment nodes and communication network nodes. Perform graph theory modeling on the topology node set to construct a directed weighted graph, and determine the directed weighted graph as the equipment topology data. ; S12: Extract device topology data For any two adjacent nodes in the system, combine the historical fault sample set of the power system with the relay protection settings to extract the fault frequency statistics of adjacent nodes in the historical fault sample set; Extract device topology data Electrical distance and electrical coupling strength coefficient between adjacent nodes and topological relevance weighting coefficient The statistical values ​​of fault frequency of adjacent nodes, electrical distance and electrical coupling strength coefficient are combined. and topological relevance weighting coefficient Perform conditional probability calculations to generate the state transition probabilities of adjacent nodes; The state transition probabilities of all adjacent nodes are arranged in a matrix to generate the Markov chain space state transition matrix. This Markov chain space state transition matrix is ​​then used as the state transition probability matrix. .

3. The heterogeneous protocol concurrency testing method based on physical topology and cascading failure modeling according to claim 2, characterized in that, In step S1, based on the state transition probability matrix Simulate the state transition process of power grid topology fault evolution; When a topology state transition is determined, a Hawkes procedure is executed to generate initial concurrent network packets. This includes the following sub-steps: S13: Inject initial disturbance sources onto the virtual simulation timeline, and use the Monte Carlo time-step simulation method combined with the state transition probability matrix. The probability of state abrupt transitions between nodes is calculated, and the state transition results of the neighboring nodes of the disturbed node are calculated. When the running state word of an adjacent node changes from a steady state to a transient or fault state, the state transition is recorded and identified as a topology state transition event. ; S14: Extract topology state transition events Timestamp sequence update history trigger time Extract the baseline value of the basic communication background traffic of the target system under steady state. The excitation parameter represents the weighting of the instantaneous surge intensity contribution. and the decay parameter representing the exponential decay rate. ; triggering historical moments Basic communication background traffic baseline value Excitation parameters and attenuation parameters Substituting the mathematical model of a stochastic point process with self-excitation characteristics into the model and performing mapping operations, the current time step is generated. Communication traffic condition strength function According to the communication traffic condition strength function The real-time distribution of data is achieved by calling the network packet sending module to dynamically map and package the corresponding test data payload to output the raw data frame, which is then used as the initial concurrent network packets. .

4. The heterogeneous protocol concurrency testing method based on physical topology and cascading failure modeling as described in claim 1 or 3, characterized in that, Step S2 includes the following sub-steps: S21: Construct a general semantic mapping abstraction layer for multi-source heterogeneous protocols to identify initial concurrent network packets. The protocol types belong to physical devices; for communication nodes adopting substation communication network and system series standards, the abstract syntax notation and basic encoding rules are dynamically deconstructed to parse substation event messages oriented towards general objects, and these general object-oriented substation event messages are identified as the first type of status messages. Parse the sampled value message and identify it as a first-type sampled message. ; S22: For communication nodes using remote control equipment and system transmission protocols, identify the initial concurrent network packets. The content of the application service data unit in the message; based on the logic state machine of the target system, dynamically generate application service data units containing valid time tags and sequence numbers, and identify application service data units containing valid time tags and sequence numbers as the second type of remote control message. ; S23: In the set of virtualized voice nodes In this paper, a multidimensional time series matrix is ​​constructed to map protocol data elements across multiple time scales for heterogeneous data streams from communication nodes. Define the observation interval that slides along the virtual simulation time axis, and determine the observation interval as the sliding window. ; S24: Extract the inherent inter-module communication delay characteristics within the hardware-in-the-loop system and generate link delay correction coefficients. ; Extracting Type 1 Status Messages Type 1 sampling message and the second type of remote control message Original generation time Extract the sliding window The starting reference time and time window length configuration value ; the original generation time Starting reference time and time window length configuration value Perform division and floor operations, combined with the link delay correction factor. Perform timestamp compensation calculations to generate synchronized timestamps after timestamp calibration preprocessing. ; The synchronization timestamp has been updated. Type 1 status message Type 1 sampling message and the second type of remote control message Data frames are merged and packaged to generate a heterogeneous test data stream that is self-consistent on the timeline. This heterogeneous test data stream is then used as the joint network test data. .

5. The heterogeneous protocol concurrency testing method based on physical topology and cascading failure modeling according to any one of claims 1 to 4, characterized in that, Probe evaluation data in step S3 Includes parsing message status data Timestamp data and resource status usage data S3 specifically includes the following sub-steps: S31: Identify the panoramic dynamic verification platform to be tested and designate it as the verification platform system. The interface protocol converter of the hardware-in-the-loop testbed is invoked to jointly run network test data. Injected into the verification platform system via Ethernet interface and Synchronous Digital Architecture (SDA) interface. ; will receive test data from the joint operation network The network port or virtual gateway node is determined as the data entry node. It will be through the data entry node The continuously injected message sequence was identified as the injection stream sequence. ; S32: In the verification platform system Data entry node Deploy entry probes at the location In the verification platform system instruction export node Deploy exit probes at the location Extracting data entry nodes The physical network interface card (NIC) characteristics or port attribute sequence generates a constant salt value used to prevent hash collisions. Extract test data from the joint operation network. Data frame header characteristics of Chinese messages Data frame header features With constant salt value Perform an XOR operation with a secure hash function to generate an independent trace identifier. ; S33: Based on entry probe Monitoring and Verification Platform System The protocol stack receive buffer is used to verify the platform system based on the message. The system assigns a resolution status word to the actual situation where data is successfully received and stored in the corresponding memory area, or is discarded due to concurrency overload or semantic conflicts. ; independent tracking identifier With the corresponding parsing status word Perform mapping and packaging to generate parsed message status data. ; S34: Based on entry probe Record message arrival verification platform system Precise time generation of entry timestamp Based on the exit probe Capture and independent tracking identifiers Associated response commands and record the precise time of their generation to generate an export timestamp. ; independent tracking identifier Entry timestamp Export timestamp Perform pairing and combination to generate timestamp marker data. ; S35: Based on entry probe With exit probe Read verification platform system The resource consumption state vector is constructed from the CPU utilization, memory heap resident capacity, and number of concurrent thread lock waits. All resource consumption state vectors recorded during the test period Resource status and usage data are generated by assembling data in chronological order. ; S36: Parse message status data Timestamp data and resource status and usage data The data is structured, aggregated, and persistently stored, then encapsulated into a comprehensive evaluation data object with a unified time-aligned index. This comprehensive evaluation data object is then designated as the probe evaluation data. .

6. The heterogeneous protocol concurrency testing method based on physical topology and cascading failure modeling according to claim 5, characterized in that, The set of dynamic quantitative indicators in step S4 Including protocol matching success rate Deterministic delay quantization value With state convergence index S4 specifically includes the following sub-steps: S41: Within the set quantitative evaluation time window Internally, read and parse message status data. Includes independent tracking identifiers Extract the parsing state word corresponding to the data tuple. ; Quantitative assessment time window All parsed status words The values ​​are summed to obtain the total number of successfully parsed packets; the total number of successfully parsed packets is then compared with the quantitative evaluation time window. Total number of evaluation window messages injected into the system Perform ratio calculations to generate the protocol matching pass rate. ; S42: Extract timestamp data Entry timestamp Export timestamp For entry timestamp Export timestamp The end-to-end processing delay is calculated. Calling the set kernel density function to process end-to-end latency The probability distribution is smoothly fitted to generate the probability density distribution function. ; In probability density distribution function An integration operation is performed over the distribution interval using preset confidence quantiles to generate a deterministic delay quantization value that satisfies the tail delay value. ; S43: Read resource status and usage data Resource consumption state vector in continuous time series Selecting the initial observation time in phase space Extract the initial observation time Corresponding initial state trajectory deviation Tracking deviations in the evolutionary trajectory caused by initial perturbations. Calculate the initial trajectory deviation Deviation from evolutionary state trajectory The exponential separation rate during time evolution generates the state convergence exponent. ; S44: During the quantitative assessment time window At the end, the protocol matching pass rate will be calculated. Deterministic delay quantization value and the state convergence index The data structure is aligned and normalized, combined into a three-dimensional array with timestamp association, and then the three-dimensional array is encapsulated and defined as a set of dynamic quantitative indicators. .

7. The heterogeneous protocol concurrency testing method based on physical topology and cascading failure modeling according to claim 6, characterized in that, Step S5 includes the following sub-steps: S51: Extracting the set of dynamic quantitative indicators Protocol matching pass rate included Deterministic delay quantization value With state convergence index Call the range normalization function to calculate the protocol matching pass rate. Deterministic delay quantization value With state convergence index Perform data mapping operations to match protocol success rates. Mapping to match normalized values Quantize the deterministic delay value Mapped to normalized delay value The state convergence exponent Mapped to convergent normalized value Match normalized values Delay normalization value With convergence normalization value Assign orthogonal basis vectors to the evaluation space to construct a matrix feature with a three-dimensional diagonal structure, and then define the matrix feature with the three-dimensional diagonal structure as the three-dimensional evaluation matrix. ; S52: Define a zero matrix representing the ideal, fault-free operating state of the target system under test, and determine the zero matrix as the ideal reference matrix. ; Extracting the three-dimensional evaluation matrix With the ideal reference matrix Construct a difference matrix, perform spatial algebraic distance quantization on the difference matrix to calculate the Frobenius norm, generate a spatial algebraic distance scalar, and determine the spatial algebraic distance scalar as the performance boundary deviation. .

8. The heterogeneous protocol concurrency testing method based on physical topology and cascading failure modeling according to claim 7, characterized in that, The excitation parameters in step S5 include parameters and parameters S5 also includes the following sub-steps: S53: Monitor performance boundary deviation Optimization operation cycle Characteristics of the derivative slope under varying conditions; deviation from performance boundaries When the set upper limit is not reached, extract the excitation parameters that control the intensity of the instantaneous flow surge in the target system. And the excitation parameters that control the flow calming decay rate Extracting performance boundary deviation For excitation parameters The first partial derivative value Performance boundary deviation For excitation parameters The second partial derivative value and the update step size of the scaling multiplier coefficient that controls the gradient adjustment magnitude. Combined with the first partial derivative value The second partial derivative value With update step size The discrete difference method is used for approximate numerical estimation to calculate the excitation parameters for the next operation cycle. With excitation parameters ; S54: Set the activation parameters for the next computation cycle. With excitation parameters The concurrent stress generation engine configuration module of the hardware-in-the-loop co-simulation test platform is synchronized in real time, and the process parameters are hot-swapped to control the concurrent stress generation engine according to the activation parameters of the next calculation cycle. Adjust the peak value of heterogeneous protocol message bursts caused by a single topology state transition event, and adjust it according to the excitation parameters of the next operation cycle. Adjust the duration of the tail of network communication traffic distribution over time; Continuous read performance boundary deviation The trend is used to determine the extreme threshold, and the deviation from the performance boundary is considered. When the verification platform system has reached its performance inflection point, it extracts the proportion of various industrial control protocol messages in the current mixed data network stream, and determines the current protocol allocation combination by combining the proportions of various industrial control protocol messages. The concurrent pressure generation engine controls the current protocol ratio combination. Under the constraints of the operating conditions, the high-frequency flow excitation state is maintained according to the optimized excitation parameters, and the extreme capacity level load impact is continuously applied to the verification platform system to expose the shortcomings of the system's scheduling capabilities.

9. A computer device comprising at least one processor coupled to at least one memory storing at least one computer program or instruction, characterized in that, The computer program or instructions are loaded and executed by the processor to implement the steps of the heterogeneous protocol concurrent testing method based on physical topology and cascading fault modeling as described in any one of claims 1 to 8.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program or instructions, which, when executed by a processor, implement the steps of the heterogeneous protocol concurrent testing method based on physical topology and cascading fault modeling as described in any one of claims 1 to 8.

Images