An intelligent substation merging unit simulation modeling method

Abstract

The application relates to the field of intelligent substation simulation modeling, and discloses a kind of intelligent substation merging unit simulation modeling method, comprising the following steps: parsing engineering configuration file to establish sampling value message protocol template, and based on the state edge to be identified to establish disturbance coverage matrix;According to the disturbance coverage matrix, generate homologous disturbance fingerprint, collect reference input data and sampling value message data and generate double-domain observation tuple;Based on double-domain observation tuple, cross-domain causal separation, channel binding verification and electrical sampling chain parameter inversion are carried out.The application establishes the linkage model between merging unit sampling value field, synchronization field, quality bit field, sampling count field and message timing by disturbance coverage matrix, double-domain observation tuple, cross-domain causal separation, phase residual table, state consistency constraint graph and frame-level joint constraint solving, improves the support capability of simulation model to boundary operating condition output behavior.

Description

Technical Field

[0001] This invention relates to the field of intelligent substation simulation modeling technology, specifically to a method for merging unit simulation modeling of intelligent substations. Background Technology

[0002] In smart substations, merging units are used to convert voltage and current sampling data from the primary equipment side into sampled value messages and send them to secondary equipment such as protection and measurement and control systems. To reproduce the output behavior of the merging units during equipment commissioning, protection testing, and abnormal operating condition verification, it is necessary to establish a merging unit simulation model that reflects the relationships between the sampled value field, synchronization field, quality bit field, sample count field, and message timing. Therefore, a simulation modeling method for merging units in smart substations has been developed.

[0003] In existing technologies, sampled value message templates are typically generated based on engineering configuration files, and then simulated sampled value messages are generated according to the set sampling rate, channel mapping, number of ASDUs, and communication parameters. Other solutions apply voltage and current signals through a tester or simulation platform and collect the output messages of the merging unit to verify whether the sampled values, smpCnt field, smpSynch field, quality field, and message interval meet the test requirements.

[0004] While existing technologies can generate sampled value messages and verify regular messages, some shortcomings remain: existing modeling processes largely rely on engineering configurations and preset message rules, but in actual merging units, under boundary conditions such as synchronization loss, synchronization recovery, quality bit changes, and abnormal sampling counts, the processing of sampled values, field changes, and message observation timing often exhibit a linkage relationship; if messages are only configured separately by field or generated according to fixed rules, it is difficult to identify the triggering source of field changes, and it is also difficult to reflect situations such as inconsistencies between channel configuration and actual message field responses, and changes in message timing residuals with event states, thereby affecting the simulation model's support for actual output behavior. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a simulation modeling method for merging units in intelligent substations. This method solves the problem that existing merging unit simulation modeling methods struggle to determine the linkage between the sampled value field, smpCnt field, smpSynch field, quality field, and message timing based on actual observation data, resulting in insufficient support for the output behavior of merging unit boundary conditions in the simulation model.

[0006] To achieve the above objectives, the present invention provides a simulation modeling method for merging units in intelligent substations, comprising the following steps: The project configuration file is parsed to establish a sampled value message protocol template, and a disturbance coverage matrix is ​​established based on the state edge to be identified; Generate homogeneous disturbance fingerprints based on the disturbance coverage matrix, collect reference input data and sampled value message data, and generate dual-domain observation tuples; Cross-domain causality separation, channel binding verification, and electrical sampling chain parameter inversion are performed based on dual-domain observation tuples; A phase residual table, a state consistency constraint diagram, and a boundary frame action template library are established based on the dual-domain observation tuples. The simulation sampled value message sequence is generated by solving the frame-level joint constraints, and the simulation sampled value message sequence is then checked in a closed loop.

[0007] Preferably, the project configuration file includes an SCD file, an ICD file, or a CID file; The sampled value message protocol template includes an Ethernet field template, a VLAN field template, an SV message header field template, an ASDU field template, a sampled value field template, a smpCnt field template, a smpSynch field template, and a quality field template; The state edges to be identified include synchronization state change edges, mass bit change edges, sampling count change edges, sampling value processing state change edges, and message observation timing change edges. The perturbation coverage matrix is ​​indexed by the state edge to be identified and the perturbation type, and is used to record whether the corresponding state edge to be identified is effectively observed for each perturbation type.

[0008] Preferably, the homogeneous disturbance fingerprint includes electrical disturbance sequences, synchronization disturbance sequences, and network observation link disturbance sequences; The electrical disturbance sequence is applied to the voltage input channel or the current input channel, the synchronization disturbance sequence is applied to the time synchronization input terminal, and the network observation link disturbance sequence is applied to the message acquisition link; The electrical disturbance sequence, synchronization disturbance sequence, and network observation link disturbance sequence are arranged according to a unified time base, so that the reference input data, synchronization state changes, and sampled value message data have corresponding time stamps.

[0009] Preferably, the dual-domain observation tuple consists of reference input data, a set of sampled value fields, a set of sampled value message control fields, a set of time fields, and a set of disturbance records under the same observation index; The set of control fields for the sampled value message includes the smpCnt field, the smpSynch field, the quality field, and the ASDU sequence number; The time field set includes the reference input time and the message observation time; The disturbance record set includes a disturbance plan identifier, a disturbance type, a disturbance target, and a disturbance tag.

[0010] Preferably, the cross-domain causal separation includes: dividing the disturbance observation window according to the disturbance record set, dividing the disturbance type into electrical disturbance domain, synchronization disturbance domain, network observation link disturbance domain and combined disturbance domain, and determining the triggering source of sample value change, smpCnt field change, smpSynch field change, quality field change and message observation timing change according to the disturbance mark, the order of field change and the triggering result of the state edge to be identified; Among them, samples whose triggering source belongs to the electrical disturbance domain or a combined disturbance domain containing the electrical disturbance domain are used for electrical sampling chain parameter inversion.

[0011] Preferably, the channel binding verification includes: forming a configuration-side channel set according to the engineering configuration file, forming an observation-side field set according to the sampled value message data, applying a single-channel perturbation to the physical input channel, calculating the channel response correlation degree between the physical input channel and the sampled value message field, and generating a channel binding table in combination with the phase and measurement type in the engineering configuration file; The channel binding table includes a physical input channel index, a sampled value message field index, channel polarity, proportional relationship, and delay category; When the channel semantics in the engineering configuration file are inconsistent with the observed response, the electrical sampling chain parameter inversion and simulation sample value field generation use the channel binding table as the basis for channel mapping.

[0012] Preferably, the electrical sampling chain parameter inversion includes: determining the correspondence between the sampled value message field and the physical input channel according to the channel binding table; selecting samples triggered by the electrical disturbance domain or samples triggered by the combined disturbance domain containing the electrical disturbance domain from the dual-domain observation tuple as parameter inversion samples; and determining the scaling factor, bias, equivalent sampling delay and filter residual parameters of each channel based on the parameter inversion samples. Among them, samples triggered by synchronous disturbance domain or network observation link disturbance domain are not used as the main samples for electrical sampling chain parameter inversion.

[0013] Preferably, the phase residual table is established based on the message observation time, the ideal message time, and the basic delay of the observation link; The phase residual table uses the sample count modulus, synchronization status, ASDU encapsulation mode, and event status as phase residual indexes, and records the message observation time series residual, residual representative value, and sample source in the table entry corresponding to each phase residual index; When generating the simulation sampled value message sequence, if the current phase residual index does not have a corresponding entry, the adjacent entries are searched in the order of event state, synchronization state, ASDU encapsulation mode and sample count modulus. If no entry is found, the message output time is generated using the basic message period.

[0014] Preferably, the state consistency constraint graph includes a joint set of state nodes, a set of observed state edges, a set of state edge triggering conditions, and a set of state edge output actions; The joint state in the set of joint state nodes consists of synchronization state, quality bit state, sampling count state, sample value processing state, and message observation timing state; The boundary frame action template library is generated based on the observed state edges, and records the sampling value action, smpCnt action, smpSynch action, quality action and observation timing action before and after the state switching index; When generating the simulation sampled value message sequence, if the current output frame is within the corresponding boundary frame window, the boundary frame action template library is called to generate the corresponding field changes.

[0015] Preferably, the frame-level joint constraint solution includes: reading the sampled value message protocol template, channel binding table, electrical sampling chain parameters, phase residual table, state consistency constraint diagram and boundary frame action template library, and using the same message frame or intra-frame ASDU as the processing object to determine the sampled value field, smpCnt field, smpSynch field, quality field, ASDU arrangement information and message output time; When different constraints produce different results for the same field, the output result is determined in the order of protocol template constraint, boundary frame action constraint, state consistency constraint, sample count constraint, sample chain constraint, and phase residual constraint. During closed-loop verification, the simulated sampled value message sequence is aligned with the measured sampled value message sequence, and the corresponding model data is updated based on field differences and message timing differences.

[0016] This invention provides a simulation modeling method for merging units in intelligent substations. It has the following beneficial effects: 1. This invention establishes a disturbance coverage matrix and generates homogeneous disturbance fingerprints based on the disturbance coverage matrix, enabling electrical disturbances, synchronization disturbances, and network observation link disturbances to act on the modeled and merged units according to a unified time reference. This allows the reference input data, disturbance records, and sampled value message data to form corresponding dual-domain observation tuples, facilitating the identification of state edge changes under different disturbance conditions.

[0017] 2. This invention uses cross-domain causal separation and channel binding verification to assign changes in sampled values, smpCnt field, smpSynch field, quality field, and message observation timing to their respective disturbance sources. It also generates a channel binding table based on the single-channel disturbance response, thereby reducing the impact of inconsistencies between the engineering configuration channel semantics and the actual message field responses on model building.

[0018] 3. This invention solves the joint constraints by using a phase residual table, a state consistency constraint diagram, a boundary frame action template library, and frame-level joint constraints. This ensures that the sampled value field, smpCnt field, smpSynch field, quality field, and message output time in the simulated sampled value message sequence satisfy the joint constraints in the same generation process. Furthermore, it writes back the field differences and message timing differences to the corresponding model data through closed-loop verification. Attached Figure Description

[0019] Figure 1 This is a flowchart illustrating the overall process of a simulation modeling method for merging units in an intelligent substation according to the present invention. Figure 2 This is a flowchart of the perturbation coverage matrix update process of the present invention. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0021] It should be noted that, unless otherwise defined, the technical or scientific terms used in this invention should have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0022] Example: See attached document Figure 1 - Appendix Figure 2 This invention provides a simulation modeling method for merging units in intelligent substations, comprising the following steps: S100. Establish a sampled value message protocol template according to the project configuration file, and establish a disturbance coverage matrix.

[0023] The computing device reads SCD, ICD, or CID files, extracts the sampled value control block, APPID, VLAN, target MAC address, dataset, channel configuration, sampling rate, number of ASDUs per frame, quality bit field configuration, and sampling counting rules, and generates a sampled value message protocol template. Using synchronization state, quality bit state, sampling counting state, sampled value processing state, and message observation timing state as a joint state, the computing device establishes a set of state edges to be identified, and generates a disturbance coverage matrix indexed by state edges and disturbance type.

[0024] S200. Generate homogeneous perturbation fingerprints based on the perturbation coverage matrix, and generate dual-domain observation tuples.

[0025] The computing device generates electrical disturbance sequences, synchronization disturbance sequences, and network observation link disturbance sequences based on the disturbance coverage matrix. The electrical disturbance sequence applies to the voltage and current input terminals of the modeled merging unit; the synchronization disturbance sequence applies to the time synchronization input terminal; and the network observation link disturbance sequence applies to the message acquisition link. The computing device acquires reference input data and sampled value messages output by the merging unit, aligns them according to a unified time base, and generates a dual-domain observation tuple. The dual-domain observation tuple includes reference input data, reference input time, sampled value field, sample count field, synchronization field, quality bit field, ASDU sequence number, frame sequence number, and message observation time.

[0026] S300: Perform cross-domain causal separation, channel binding verification, and electrical sampling chain parameter inversion based on dual-domain observation tuples.

[0027] The computing device processes observation results under electrical single-domain disturbances, synchronous single-domain disturbances, network observation link single-domain disturbances, and combined disturbances. Based on the order of output field changes, state edge triggering, and field linkage relationships, it determines the triggering sources for changes in sampled values, sample counts, synchronous fields, quality bits, and message observation timing. According to the channel semantics and sampled value message field responses to single-channel disturbances in the project configuration file, the computing device generates a channel binding table. Based on the channel binding table and dual-domain observation tuples, it inverts the scaling factor, bias, equivalent sampling delay, and filter residual parameters for each channel.

[0028] S400. Establish a phase residual table, state consistency constraint diagram and boundary frame action template library based on the dual-domain observation tuple, and update the disturbance coverage matrix according to the state edge coverage.

[0029] The computing device calculates timing residuals based on ideal message time, message observation time, and observation link baseline delay, and establishes a phase residual table using sample count modulus, synchronization status, ASDU encapsulation mode, and event status as indexes. The computing device establishes a state consistency constraint graph based on multi-field state changes, and extracts sample value actions, sample count actions, synchronization field actions, quality bit actions, and timing actions before and after state switching frames to form a boundary frame action template library. For uncovered state edges, the computing device updates the perturbation coverage matrix and generates a secondary perturbation sequence; for state edges not triggered under preset perturbation conditions, the computing device marks them as unreachable state edges and saves them.

[0030] S500 generates a simulation sample value message sequence based on the model data and performs closed-loop verification.

[0031] The frame-level joint constraint solving unit generates a joint output vector corresponding to each frame of sampled value messages based on the protocol template, channel binding table, electrical sampling chain parameters, phase residual table, state consistency constraint diagram, and boundary frame action template library. The joint output vector includes the sampled value, sample count, synchronization field, quality bit field, ASDU sequence number, and message output time. The message encoding output unit generates a simulated sampled value message sequence based on the joint output vector and protocol template. The closed-loop verification and update unit compares the simulated sampled value messages with the measured sampled value messages and updates the corresponding model data based on field differences and timing differences.

[0032] The technical implementation details of the above steps are explained in detail below with reference to specific embodiments: In step S100, the computing device reads the intelligent substation engineering configuration file, which includes an SCD file, an ICD file, or a CID file. The computing device extracts configuration content related to the merging unit sample value message from the engineering configuration file and converts the configuration content into structured configuration data. For the XML parsing, node traversal, and field reading of the engineering configuration file, those skilled in the art can use existing parsing methods, which are well-known technologies in the field and will not be elaborated upon here.

[0033] S101. The computing device reads the project configuration file and extracts the sampled value message configuration parameters.

[0034] The sampled value message configuration parameters include the sampled value control block identifier, APPID, VLAN identifier, target MAC address, dataset identifier, channel configuration, sampling rate, number of ASDUs per frame, sample count field rules, synchronization field rules, and quality bit field configuration.

[0035] Channel configuration includes channel name, phase, measurement type, data field location, scale configuration, and quality bit field location. Measurement types include voltage, current, and other sampling quantities defined in the engineering configuration file. Channel configuration is used to establish the configured channel set for subsequent channel binding verification.

[0036] S102. The computing device generates a sampled value message protocol template based on the configuration parameters.

[0037] The sampled value message protocol template is used to define the field structure of the simulation message, including Ethernet field template, VLAN field template, SV message header field template, ASDU field template, sampled value field template, sampled count field template, synchronization field template, and quality bit field template.

[0038] The computing device establishes a sample value field template based on the field order in the channel configuration, a sample count field template based on the sample counting rules, a synchronization field template based on the synchronization field rules, and a quality bit field template based on the quality bit field configuration. When each frame contains multiple ASDUs, the computing device establishes an intra-frame ASDU arrangement rule according to the number of ASDUs per frame, ensuring that each ASDU within the same Ethernet frame has a defined field position and counting relationship.

[0039] The sampled value message protocol template is only used to limit the message format, field position and encapsulation relationship, and does not directly represent the actual output behavior of the modeled and merged unit when there are synchronization anomalies, quality bit changes, sample count changes and message timing changes.

[0040] S103. The computing device establishes a joint state to describe the observable state changes of the merging units.

[0041] The joint state consists of a synchronization state, a quality bit state, a sampling count state, a sample value processing state, and a message observation timing state. The synchronization state corresponds to the smpSynch field; the quality bit state corresponds to the quality field; the sampling count state corresponds to the continuous, repeating, jumping, or wraparound state of the smpCnt field; the sample value processing state corresponds to the normal output, hold, zeroing, or limiting state of the sample value field; and the message observation timing state corresponds to the timing state of the message observation time relative to the ideal message time.

[0042] The joint state is represented as: In the formula, Indicates the first The joint state corresponding to each observation index; Indicates synchronization status; Indicates the state of the mass; Indicates the sampling count status; Indicates the processing status of the sampled values; Indicates the timing status of message observation; Indicates the observation index.

[0043] Each sub-state in the above-mentioned joint state is calculated from the observable field of the sampled value message or the message observation time, without needing to read the internal program of the merging unit.

[0044] S104. The computing device establishes a set of edges of the states to be identified based on the joint state.

[0045] State edges are used to describe the observable relationship of a joint state changing from its pre-transition state to its post-transition state. Each state edge includes the pre-transition state, the post-transition state, the triggering condition, and the output action.

[0046] The set of state edges is represented as: In the formula, Represents the set of edges in the state to be identified; Indicates the number of state edges; Indicates the state edge index.

[0047] No. A state edge is represented as: In the formula, Indicates the first State edges; Indicates the number of state edges; Indicates the state edge index; Indicates the first The state before the transition of a state edge; Indicates the first The state after the transition of a state edge; Indicates the first Triggering conditions for state edges; Indicates the first Output action of the state edge.

[0048] The trigger condition records at least one of the following: the type of disturbance that caused the state change, the time of the disturbance, the synchronization state change, the change in the input sample value, and the change in the message observation timing. The output action records how the sample value field, smpCnt field, smpSynch field, quality field, and message observation time changed after the state change.

[0049] The identified state edges include synchronization state change edges, quality bit change edges, sample count change edges, sample value processing state change edges, and message observation timing change edges. Synchronization state change edges describe the transitions between synchronization active, synchronization held, synchronization lost, and synchronization recovered. Quality bit change edges describe the transitions between active, doubtful, invalid, test, and old data states. Sample count change edges describe continuous incrementing, repetition, jump, and wraparound behaviors. Sample value processing state change edges describe normal output, hold, zeroing, and limiting behaviors. Message observation timing change edges describe the changes in message observation time under different event states.

[0050] S105. The computing device establishes a perturbation coverage matrix based on the set of state edges to be identified and the set of perturbation types.

[0051] The disturbance type set includes electrical disturbances, synchronization disturbances, network observation link disturbances, and combined disturbances. Electrical disturbances affect voltage and current inputs; synchronization disturbances affect time synchronization inputs; and network observation link disturbances affect message acquisition links. Network observation link disturbances are used to distinguish the timing effects on message observation links and are not limited to changes in the transmission queue within the merging unit.

[0052] The perturbation coverage matrix is ​​represented as: In the formula, Represents the perturbation coverage matrix; Indicates the first Class of perturbations on the first The coverage state of the edge; Indicates the number of state edges; Indicates the number of disturbance types; Indicates the state edge index; Indicates the perturbation type index.

[0053] During initialization, the computing device sets each matrix element to an uncovered state: ;in It indicates that the first [stage / process] has not yet been passed. The first type of perturbation was effectively observed. A state edge.

[0054] After applying the perturbation and collecting the dual-domain observation tuples, if the first... The first type of perturbation appears in the observation data. If a state edge has a state before transition and a state after transition, and this state change satisfies the corresponding triggering condition, then the computing device will update the corresponding matrix element to the overridden state: ;in Indicates the first The state edge has been in the first Effective observation results were obtained under perturbation-like conditions.

[0055] Through the above processing, S100 generates a sampled value message protocol template and a perturbation coverage matrix. The sampled value message protocol template is used to constrain the field structure of subsequent simulation messages; the perturbation coverage matrix is ​​used to guide the generation of fingerprints of homogeneous perturbations and to record the coverage of state edges by different perturbation types. Subsequent steps generate perturbation sequences based on the perturbation coverage matrix and update the state edge coverage results after collecting observation data.

[0056] In step S200, the computing device determines the state edge to be covered based on the disturbance coverage matrix obtained in S100, and arranges electrical disturbances, synchronization disturbances, and network observation link disturbances according to a unified time base to form a common-source disturbance fingerprint. This common-source disturbance fingerprint is used to give the reference input data, synchronization state changes, and sampled message observation data corresponding time stamps. The network observation link disturbance is applied to the message acquisition link to record the impact of the acquisition path on the message observation timing, but is not used to limit changes in the transmission queue within the merging unit.

[0057] S201. The computing device determines the edge of the state to be covered based on the disturbance coverage matrix.

[0058] The computing device reads the coverage status of each state edge in the disturbance coverage matrix. For any state edge, if it has not formed a coverage record under any type of disturbance, the state edge is determined as a state edge to be covered.

[0059] The marker to be covered is represented as: In the formula, Indicates the first The markers to be covered for the state edges; Indicates the th in the perturbation coverage matrix Class of perturbations on the first The coverage state of the edge; Indicates the state edge index; Indicates the perturbation type index; Indicates the number of disturbance types. If... , indicating the first The edge of the state has not yet been covered; if , indicating the first The state edge has formed a covered record.

[0060] The computing device selects the disturbance type based on the state edge type. Electrical disturbances correspond to the sampled value processing state edge and the quality bit state edge related to sampled value anomalies; synchronization disturbances correspond to the synchronization state edge, quality bit state edge, and sample counting state edge; network observation link disturbances correspond to the message observation timing state edge. The above correspondences are stored in the form of a rule table, which records at least the state edge type, the optional disturbance type, and the object affected by the disturbance.

[0061] S202, The computing device generates a disturbance plan based on the state to be covered.

[0062] The computing device generates a disturbance plan based on the type of state edge to be covered. The disturbance plan includes at least a disturbance plan identifier, disturbance type, disturbance target, disturbance start time, disturbance end time, and disturbance flag. The disturbance flag is written into subsequent observation records to identify the correspondence between disturbances and state edge changes.

[0063] The disturbance plan is represented as follows: In the formula, Indicates the first A disturbance plan; Indicates the disturbance plan identifier; Indicates the type of disturbance; Indicates the object of the disturbance; Indicates the start time of the disturbance; Indicates the end time of the disturbance; Indicates disturbance marker; This indicates the perturbation plan index.

[0064] The objects affected by the disturbance include the voltage input channel, the current input channel, the time synchronization input terminal, and the message acquisition link. For the signal output of electrical disturbances, those skilled in the art can achieve this using a relay protection tester or a real-time simulator; the method of outputting the test signal is well-known in the field and will not be elaborated upon here.

[0065] S203, The computing device will arrange the perturbation plan into a homogeneous perturbation fingerprint.

[0066] The computing device arranges the disturbance plans according to a unified time base, forming a fingerprint of the same source disturbance. The unified time base is provided by a time synchronization device, a relay protection tester, or a unified clock of the computing device.

[0067] Homologous perturbation fingerprints are represented as follows: In the formula, Indicates homologous perturbation fingerprints; Indicates an electrical disturbance sequence; Indicates a sequence of synchronous perturbations; Indicates the network observation link disturbance sequence; It indicates a unified time base.

[0068] Electrical disturbance sequences include at least one of amplitude variation, phase variation, frequency variation, harmonic injection, single-channel input disturbance, and short-term input interruption. Synchronization disturbance sequences include at least one of synchronization loss, synchronization maintenance, and synchronization recovery. Network observation link disturbance sequences include at least one of message acquisition link load change, acquisition port status record, and acquisition link delay marker.

[0069] S204. The computing device controls the test equipment to apply the same source perturbation fingerprint.

[0070] The computing device sends control commands to the relay protection tester, time synchronization device, and message acquisition device based on the fingerprint of the same source disturbance. The relay protection tester outputs voltage and current test signals based on the electrical disturbance sequence; the time synchronization device controls the time synchronization input status based on the synchronization disturbance sequence; and the message acquisition device records and acquires the link status based on the network observation link disturbance sequence.

[0071] S205. The computing device collects reference input data and sampled value message data.

[0072] The reference input data includes the voltage input value, current input value, disturbance flag, and reference input time. The sampled value message data includes the sampled value field, smpCnt field, smpSynch field, quality field, ASDU sequence number, frame sequence number, and message observation time. For sampled value message parsing, those skilled in the art can use the IEC 61850 sampled value message parsing method, which will not be elaborated upon here.

[0073] The message observation time is provided by the message acquisition device. When using hardware timestamps, the calculation device records the hardware timestamp; when using a calibrated acquisition link, the calculation device records the calibrated observation time. The time recorded in a normal packet capture environment is used as the message observation time, not as the internal transmission time of the merging unit.

[0074] S206. The computing device performs time alignment between the reference input data and the sampled value message data.

[0075] The computing device establishes a correspondence between reference input data and sampled value message data based on a unified time base. For each set of sampled value messages (ASDUs), the computing device determines the corresponding reference input sampling point and disturbance marker based on the frame sequence number, ASDU sequence number, sampling count field, and message observation time. The conversion between the observation index and the message frame and ASDU sequence number can be performed according to the number of ASDUs per frame. This index conversion is a standard data processing method, and the formula is not listed here.

[0076] S207, The computing device generates dual-domain observation tuples.

[0077] The computing device combines the reference input data, disturbance records, sampled value message fields, and message observation times under the same observation index into a two-domain observation tuple.

[0078] The dual-domain observation tuple is represented as: In the formula, Indicates the first One dual-domain observation tuple; Indicates reference input data; Represents a set of sampled value fields; This represents the set of control fields for the sampled value message; Represents a collection of time fields; Represents the set of disturbance records; Indicates the observation index.

[0079] The reference input data includes the voltage and current input values ​​at the corresponding time; the sampled value field set includes the sampled values ​​of each channel in the sampled value message; the sampled value message control field set includes the smpCnt field, smpSynch field, quality field, and ASDU serial number; the time field set includes the reference input time and the message observation time; and the disturbance record set includes the disturbance plan identifier, disturbance type, disturbance target, and disturbance flag.

[0080] Through the above processing, S200 converts the perturbation coverage matrix into a homogeneous perturbation fingerprint and organizes the reference input data and sampled message data into a dual-domain observation tuple. This dual-domain observation tuple contains input-side data, message-side data, time data, and perturbation records, which are used for subsequent identification of state edge trigger sources, establishment of channel binding relationships, and generation of behavioral models.

[0081] In step S300, the computing device uses the dual-domain observation tuples generated in S200 as the data basis to determine the triggering source of output field changes, the correspondence between physical input channels and sampled value message fields, and the electrical sampling chain parameters of each channel. Cross-domain causal separation is used to distinguish the impact of electrical disturbances, synchronization disturbances, and network observation link disturbances on message output fields; channel binding verification is used to verify the correspondence between the engineering configuration channel semantics and the actual message field response; electrical sampling chain parameter inversion is used to obtain the scaling, bias, delay, and filter residual parameters required for simulating sampled value fields.

[0082] S301, The computing device divides the disturbance observation window according to the dual-domain observation tuple.

[0083] The computing device reads the set of disturbance records from the dual-domain observation tuple and establishes a disturbance observation window according to the disturbance plan identifier, disturbance type, disturbance target, disturbance start time, and disturbance end time. Each disturbance observation window corresponds to a disturbance plan and is used to extract reference input data and sampled value message data during the period of the disturbance plan and its adjacent time range.

[0084] The disturbance observation window is represented as ;in Indicates the first One disturbance observation window; Indicates the observation index; Indicates the first The time corresponding to each observation index; Indicates the first The start time of the disturbance plan; Indicates the first The end time of the disturbance plan; This indicates the perturbation plan index.

[0085] When it is necessary to identify field changes before and after a state switch, the computing device retains adjacent observation data before and after the disturbance observation window to determine the order of field changes.

[0086] S302, Changes in the output fields of the computing device record.

[0087] The computing device extracts the changes in the sampled value field, smpCnt field, smpSynch field, quality field, and message observation time within the disturbance observation window. Field changes are obtained through comparison of adjacent observation data, determination of field state changes, or state edge triggering results. For sampled value differentiation, field state comparison, and time difference calculation, those skilled in the art can use conventional data processing methods, which will not be elaborated upon here.

[0088] The set of output field changes is represented as ;in Indicates the first The set of output field changes corresponding to each observation index; This indicates a change in the sampled value field; This indicates a change in the smpCnt field; This indicates a change in the smpSynch field; This indicates a change in the quality field; Indicates changes in message observation time; Indicates the observation index.

[0089] S303, Computing devices perform cross-domain causality separation.

[0090] The computing device classifies disturbance types into electrical disturbance domain, synchronization disturbance domain, network observation link disturbance domain, and combined disturbance domain. The electrical disturbance domain corresponds to voltage and current input disturbances; the synchronization disturbance domain corresponds to timing input disturbances; the network observation link disturbance domain corresponds to message acquisition link disturbances; and the combined disturbance domain corresponds to situations where at least two disturbance domains act together within the same observation window.

[0091] The computing device assigns the trigger source to the output field changes based on the disturbance flag, the order in which field changes occur, and the state edge triggering results. The trigger source attribution is represented as follows: ;in Indicates changes in output fields The trigger source; This indicates a change in the output field to be assigned to; Indicates the perturbation domain; Represents the set of perturbation domains; Indicates changes in output fields With disturbance domain Match score.

[0092] Match score is expressed as: In the formula, Indicates changes in output fields Does it fall into the disturbance domain? The corresponding disturbance mark range; Indicates changes in output fields With disturbance domain Time sequence matching results; Indicates changes in output fields Does it correspond to the perturbation domain? The associated state edges; , and This indicates the weight of each matching item.

[0093] When the output field change occurs within the time range corresponding to the disturbance marker, the corresponding disturbance marker matching item takes the matching state; when the output field change occurs after the disturbance and within the preset observation window, the corresponding time sequence matching item takes the matching state; when the state edge corresponding to the output field change belongs to the state edge type associated with the disturbance domain, the corresponding state edge matching item takes the matching state. Each matching item can use binary values ​​or hierarchical values, and the value method is preset in the modeling parameters.

[0094] If multiple disturbance domains have the same matching score, or if the change in the output field cannot be explained by a single disturbance domain, the computing device marks the change in the output field as a combined disturbance source or an unknown source. Unknown sources are not used as the sample basis for electrical sampling chain parameter inversion, but are retained in the state consistency constraint diagram and closed-loop verification processing.

[0095] S304. The computing device filters parameter inversion samples based on the cross-domain causal separation results.

[0096] The computing device selects samples from the dual-domain observation tuples whose triggering sources belong to the electrical disturbance domain, or to a combined disturbance domain that includes the electrical disturbance domain, as samples for electrical sampling chain parameter inversion. Samples whose triggering sources belong to the synchronization disturbance domain or the network observation link disturbance domain are not used as the main samples for electrical sampling chain parameter inversion. This process is used to avoid including synchronization state changes and message observation link changes in the electrical sampling chain parameters.

[0097] S305. The computing device establishes a channel binding relationship between configuration semantics and observation response.

[0098] The computing device generates a configuration-side channel set based on the engineering configuration file and an observation-side field set based on the sampled value messages. Each channel in the configuration-side channel set has a channel name, phase, measurement type, and configuration field position. Each field in the observation-side field set has a message field position, sampled value sequence number, and quality bit position.

[0099] The computing device applies a single-channel perturbation to each physical input channel and records the response of each sampled value message field within the corresponding perturbation observation window. If a certain observation field changes accordingly with the physical input channel perturbation, then that observation field is used as a candidate binding field for that physical input channel.

[0100] The correlation of channel response is expressed as: In the formula, Indicates the first The physical input channel and the first The correlation of responses to each observed field; Indicates the first The physical input channel in the first Input values ​​for each observation index; Indicates the first The observation field in the _th ... Sample values ​​of each observation index; Indicates the first The mean value of each physical input channel within its corresponding observation window; Indicates the first The mean of each observation field within its corresponding observation window; Indicates the first A set of single-channel disturbance observation indexes for each physical input channel; Indicates the physical input channel index; This indicates the index of the observed field.

[0101] The computing device determines candidate binding fields based on response correlation from high to low, and performs consistency verification in conjunction with the phase and measurement type in the engineering configuration. For cases where one physical input channel corresponds to multiple candidate observation fields, the computing device selects the observation field whose response correlation meets preset conditions and is consistent with the channel semantics; for cases where one observation field corresponds to multiple physical input channels, the computing device determines a one-to-one binding relationship based on the response correlation and the time correspondence of the disturbance marker.

[0102] S306, Computing device generates channel binding table.

[0103] The channel binding table includes physical input channel indexes, observation field indexes, channel polarity, scaling factors, and delay categories. Channel polarity is determined based on the direction of input and observation field changes within a single channel disturbance window; scaling factors are determined based on the magnitude of input and observation field changes; and delay categories are determined based on the chronological relationship between the observation field response and the occurrence of the input disturbance. When the channel semantics in the project configuration file are inconsistent with the observed response, subsequent electrical sampling chain parameter inversion and simulation sample value field generation use the channel binding table as the basis for channel mapping.

[0104] S307, Calculate the electrical sampling chain parameters of the inverted equipment.

[0105] The computing device establishes an electrical sampling chain model for each channel based on the channel binding table and parameter inversion samples. The electrical sampling chain parameters include scaling factor, bias, equivalent sampling delay, and filter residual parameters.

[0106] No. The electrical sampling chain model for each channel is represented as follows: In the formula, Indicates the first The first channel in the The model output value for each observation index; Indicates the first The scaling factor for each channel; Indicates the first Each physical input channel has a delay The corresponding input value; Indicates the first The equivalent sampling delay of each channel; Indicates the first The bias of each channel; Indicates the first The first channel Filter residual parameters; Indicates the first The order of the filter residuals for each channel; Indicates the first The physical input channel in the first Input values ​​for each observation index; Indicates the physical input channel index; Indicates the observation index; This represents the index of the filter residual order.

[0107] The computing device uses the observation fields determined by the channel binding table as the target output and the reference input data in the parameter inversion samples as the input to calculate the scaling factor, bias, equivalent sampling delay, and filter residual parameters. For the minimum residual, piecewise loss, or outlier removal in the parameter solution, those skilled in the art can use mature numerical calculation methods, and the specific solution process will not be described in detail here.

[0108] S308, The computing device outputs cross-domain causality separation results, channel binding tables, and electrical sampling chain parameters.

[0109] The cross-domain causality separation results include the trigger sources corresponding to changes in each output field. The channel binding table includes physical input channel indices, observation field indices, polarity relationships, scaling relationships, and delay categories. The electrical sampling chain parameters include the scaling factor, bias, equivalent sampling delay, and filter residual parameters for each channel. These results are used for subsequent phase residual table generation, state consistency constraint graph generation, boundary frame action template library generation, and frame-level joint constraint solving.

[0110] In step S400, the computing device reads the dual-domain observation tuples generated in S200 and, combined with the cross-domain causal separation results, channel binding table, and perturbation coverage matrix obtained in S300, extracts the observation timing features, multi-field state transition relationships, and field actions before and after state switching from the merge unit's output message. The phase residual table records the residual between the message observation time and the ideal message time; the state consistency constraint graph records the linkage between the sampled value field, smpCnt field, smpSynch field, quality field, and message observation timing; the boundary frame action template library records the multi-field action sequences near the state switching frame. This data is used for frame-level joint constraint solving in S500.

[0111] S401. The computing device establishes a phase residual table.

[0112] The computing device reads the message frame sequence number, ASDU sequence number, smpCnt field, smpSynch field, quality field, message observation time, and disturbance record from the dual-domain observation tuple. The computing device determines the ideal message time based on the sampling rate, the number of ASDUs per frame, and the message frame sequence number. The calculation of the ideal message time can be determined by those skilled in the art based on the sampling rate and encapsulation rules of the sampled value message; this is a conventional method for timing processing of sampled value messages and will not be elaborated upon here.

[0113] The computing device calculates the message observation timing residual based on the message observation time, ideal message time, and observation link baseline delay: In the formula, Indicates the first The observation timing residuals of the frame message; Indicates the first The observation time of the frame message; Indicates the first Ideal time for frame messages; Indicates the basic delay of the observation link; Indicates the message frame index.

[0114] The message observation time is provided by the message acquisition device. When using hardware timestamps or a calibrated acquisition link, the computing device uses the corresponding observation time; in a normal packet capture environment, the packet capture time is used as the message observation time, not as the internal transmission time of the merging unit. The basic delay of the observation link can be obtained by calibrating the acquisition link, or it can be pre-configured by the test system.

[0115] The computing device establishes a phase residual index based on the sample count modulus, synchronization status, ASDU encapsulation mode, and event status. The ASDU encapsulation mode includes the number of ASDUs per frame, the current ASDU sequence number, and the intra-frame arrangement. The event status is determined based on the disturbance record and includes the pre-disturbance state, disturbance effect state, disturbance recovery state, and stable observation state.

[0116] The phase residual index is represented as: In the formula, Indicates the first Phase residual index of the frame message; Indicates the first The sampling count field corresponding to the frame message; Indicates the sampling counting cycle length; Indicates the first The synchronization status corresponding to the frame message; Indicates the first The ASDU encapsulation mode corresponding to the frame message; Indicates the first The event status corresponding to the frame message; Indicates the message frame index.

[0117] The computing device writes the observation time-series residuals with the same phase residual index into the same table entry. Each entry includes at least the index value, residual sample, representative residual value, and sample source. The representative residual value can be the mean, median, or statistical value after outlier removal. For the mean, median, and outlier removal, those skilled in the art can use conventional statistical methods, which will not be elaborated here.

[0118] When the phase residual table needs to be called during the simulation phase but no table entry has been formed in the current index, the computing device searches for adjacent table entries in the order of event state, synchronization state, ASDU encapsulation mode, and sample count modulus. If no table entry is found, the basic timing corresponding to the basic delay of the observation link is used as the timing constraint for the frame message.

[0119] S402. Establish a state consistency constraint diagram for the computing device.

[0120] The computing device determines the joint state corresponding to each observation index based on the dual-domain observation tuple. The joint state includes synchronization state, quality bit state, sample count state, sample value processing state, and message observation timing state. All of the above states are derived from the observable fields of the sample value message or the calculation results of the message observation time.

[0121] When a joint state change occurs in an adjacent observation index or adjacent observation window, the computing device takes the joint state before the change as the state before the transition and the joint state after the change as the state after the transition, and generates state edges by combining the disturbance record and the cross-domain causal separation result.

[0122] The state consistency constraint diagram is represented as follows: In the formula, Represents a state consistency constraint diagram; Represents the set of joint state nodes; Represents the set of observed state edges; This represents the set of conditions for triggering state edges; This represents the set of output actions for the state edge.

[0123] Each state edge in the observed state edge set corresponds to a set of trigger conditions in the trigger condition set and a set of output actions in the output action set. Each state edge records the joint state before the transition, the joint state after the transition, the trigger conditions, and the output actions. Trigger conditions include at least one of the following: disturbance type, disturbance flag, trigger source, input sample value change, synchronization state change, and message observation timing change. Output actions include sample value actions, smpCnt actions, smpSynch actions, quality actions, and observation timing actions.

[0124] Sampling actions include normal output, hold, zeroing, and limiting. smpCnt actions include continuous increment, repeat, jump, and wrapback. smpSynch actions include hold original value, set to synchronization, set to hold, set to out of synchronization, and set to recovery. quality actions include hold original value, set to valid, set to suspicious, set to invalid, set to test, and set to old data. Observation timing actions include using phase residual entries, using basic timing, repeating observation timing actions, and no-report observation actions.

[0125] The computing device writes the cross-domain causality separation results into the state edge triggering conditions. State edges triggered by the electrical disturbance domain are recorded as electrical triggers; state edges triggered by the synchronization disturbance domain are recorded as synchronization triggers; state edges triggered by the network observation link disturbance domain are recorded as observation link triggers; and state edges jointly triggered by multiple disturbance domains are recorded as combined triggers. State edges that cannot be uniquely attributed retain disturbance records and field change results, but are not used as the basis for updating electrical sampling chain parameters.

[0126] S403, Computing device generates boundary frame action template library.

[0127] The computing device determines the state transition index based on the observed state edges in the state consistency constraint graph. The state transition index represents the message frame position or ASDU position where the joint state changes from the pre-transition state to the post-transition state. For the same state edge, the computing device selects a boundary frame window before and after the state transition index and records multi-field actions within that window.

[0128] The boundary frame window is represented as: In the formula, Indicates the first The boundary frame window corresponding to each state edge; Indicates the observation index; Indicates the first The state transition index corresponding to each state edge; Indicates the observation span retained before the state switch; This indicates the observation span retained after a state switch; Indicates the state edge index.

[0129] The computing device extracts sampling value actions, smpCnt actions, smpSynch actions, quality actions, and observation timing actions within the boundary frame window, and writes them into the boundary frame action template library according to the state edge index and relative position. The relative position is determined by the difference between the current observation index and the state transition index.

[0130] The boundary frame action template is represented as follows: In the formula, Indicates the first State edges in relative positions Boundary frame action template; Indicates the sampling value action; This indicates the smpCnt action; This indicates the smpSynch action; Indicates the quality action; Indicates the timing of observations; Indicates the state edge index; Indicates the position relative to the state transition index.

[0131] The boundary frame action template library records the relative change order of multiple fields within the same state transition window. During simulation output, if the current output frame is within the boundary window of the corresponding state edge, the frame-level joint constraint solver calls the boundary frame action template corresponding to that state edge.

[0132] S404, Calculate the device to update the disturbance coverage matrix.

[0133] The computing device updates the perturbation coverage matrix based on the observed state edges and perturbation records in the state consistency constraint graph. If a state edge appears within the observation window corresponding to a certain perturbation type, and the state before the transition, the state after the transition, and the triggering condition of the state edge are all recorded, then the state edge is considered a coverage under the corresponding perturbation type.

[0134] The perturbation coverage matrix update is represented as: In the formula, Indicates the updated overwrite status; Indicates the overwrite status before the update; Indicates the first Whether the perturbation effectively observed the first State edges; Indicates the state edge index; Indicates the perturbation type index. If the first... The first type of perturbation was effectively observed. If there are 1 state edge, then If not effectively observed, then .

[0135] The criteria for determining a valid observation include: the existence of a joint state before the transition within the observation window; the existence of a joint state after the transition within the observation window; the change between the two satisfying the corresponding triggering conditions; and the output action containing a field change related to the state edge. When all of the above conditions are met, the computing device determines the state edge as a valid observation.

[0136] S405. The computing device generates a secondary perturbation sequence based on the uncovered state edge.

[0137] The computing device determines the uncovered state edges based on the updated perturbation coverage matrix. If a state edge does not form a coverage record under all perturbation types, the computing device writes the state edge into the set of uncovered state edges.

[0138] The set of uncovered edge states is represented as: In the formula, Represents the set of uncovered edge states; Indicates the first State edges; Indicates the th in the perturbation coverage matrix Class of perturbations on the first The coverage state of the edge; Indicates the state edge index; Indicates the perturbation type index; Indicates the number of disturbance types.

[0139] The computing device generates a secondary disturbance sequence based on the type of uncovered state edge, the type of disturbance already executed, the observed field actions, and the cross-domain causality separation results. For uncovered synchronization state edges, the computing device adjusts the occurrence time, hold interval, or recovery interval of the synchronization disturbance. For uncovered quality bit state edges, the computing device adjusts the combination of synchronization disturbance and electrical disturbance. For uncovered sample counting state edges, the computing device adjusts the position of the synchronization disturbance and the observation window. For uncovered sample value processing state edges, the computing device adjusts the application channel, disturbance duration interval, or input interruption method of the electrical disturbance. For uncovered message observation timing state edges, the computing device adjusts the recording interval or acquisition link status of network observation link disturbances.

[0140] For state edges that are not triggered under preset perturbation conditions, the computing device marks the state edge as an unreachable state edge and saves the state edge identifier, the perturbation record executed, and the reason for non-triggering. Unreachable state edges are not written into the mandatory state transition constraints of the final simulation model, but are retained in the modeling record.

[0141] Through the above processing, the computing device generates a phase residual table, a state consistency constraint diagram, and a boundary frame action template library, and updates the perturbation coverage matrix according to the state edge coverage. This data is used for frame-level joint constraint solving and simulation sampled value message generation in the S500.

[0142] The simulation sample value message sequence is generated based on the model data and closed-loop verification is performed.

[0143] S501: The computing device reads the sampled value message protocol template, channel binding table, electrical sampling chain parameters, phase residual table, state consistency constraint diagram, boundary frame action template library, and disturbance coverage matrix to generate a simulated sampled value message sequence. During message generation, the same message frame or intra-frame ASDU is used as the processing object, ensuring that the sampled value field, smpCnt field, smpSynch field, quality field, ASDU arrangement information, and message output time are determined in the same processing step, rather than being generated independently and then simply concatenated.

[0144] The frame-level joint output vector is represented as: In the formula, Indicates the first The first frame The combined output of each ASDU; Indicates the message frame index; Indicates the intra-frame ASDU index; Represents a set of sampled value fields; This represents the smpCnt field; This refers to the smpSynch field; This represents the quality field; Indicates the ASDU sequence number and intra-frame arrangement information; Indicates the first The output time of the frame message.

[0145] S502, the computing device determines the correspondence between the sampled value field and the physical input channel according to the channel binding table, and generates the sampled value field based on the electrical sampling chain parameters. For the first... Each sample value field has a simulated value represented as follows: In the formula, Indicates the first Each sample value field in the sample index Simulated sampled values ​​at the location; Indicates message encoding processing; Indicates the proportionality coefficient; Indicates the relationship with the first The input channel data is bound to each sample value field; This indicates the binding relationship between the sampled value field and the input channel; Indicates the sampling index; Indicates the equivalent sampling delay; Indicates bias; Represents the filter residual coefficients; Indicates the order of the filter residual; This represents the index of the filter residual term. The unit conversion, integerization, and message encoding of the sampled value field can be implemented according to the sampled value message encoding rules, which are well-known technologies in this field and will not be elaborated upon here.

[0146] S503. The computing device determines the synchronization state, quality bit state, sample count state, sample value processing state, and message timing state corresponding to the current message frame or the current ASDU based on the simulation input event, the joint state of the previous generation position, and the state consistency constraint diagram. When the simulation input event matches a recorded state edge, the computing device generates the smpCnt field, smpSynch field, quality field, and sample value processing action according to the triggering condition and output action of that state edge; when no state edge is matched, the computing device maintains the previous joint state and records the output in the unmatched state edge record.

[0147] S504. When the current generation location is near a state transition frame, the computing device calls the boundary frame action template library. The boundary frame action template includes sampling value actions, smpCnt actions, smpSynch actions, quality actions, and message timing actions. Sampling value actions include maintaining the previous valid value, clearing, limiting, and normal output; smpCnt actions include continuous increment, repetition, transition, and wraparound; smpSynch and quality actions are written according to the field change order recorded in the template. For fields already defined in the boundary frame action template, general state rules are no longer used as substitutes.

[0148] S505. The computing device determines the message output time based on the phase residual table: In the formula, Indicates the first Frame message output time; Indicates the simulation start reference time; Indicates the ideal message frame period; Indicates the equivalent path delay; This represents the residual value obtained according to the phase residual index; This represents an index composed of the smpCnt modulus, synchronization status, ASDU encapsulation mode, and event status. When a corresponding entry does not exist in the phase residual table, the computing device searches for the residual by the modulus of adjacent sample counts under the same synchronization status; if it still does not exist, the output time is generated using the basic message period, and this index is recorded in the residual gap record.

[0149] S506. The computing device writes the sampled value field, smpCnt field, smpSynch field, quality field, and ASDU arrangement information into the sampled value message protocol template to form a simulated sampled value message. The generation of Ethernet frame encapsulation, VLAN field, APPID field, length field, and checksum field can adopt the conventional encapsulation method of sampled value messages, which is a well-known technology in this field and will not be described in detail here.

[0150] S507. When different constraints yield different results for the same field, the computing device determines the output in the following order: protocol template constraint, boundary frame action constraint, state consistency constraint, sample count constraint, sampling chain constraint, and phase residual constraint. The protocol template constraint ensures the field structure and number of fields; the boundary frame action constraint ensures the order of field changes within the state switching window; and the state consistency constraint ensures the correspondence between the synchronization state, quality bit state, sample count state, and sample value processing state.

[0151] S508, the computing device aligns the simulated sampled value message sequence with the measured sampled value message sequence, and compares the sampled value field, smpCnt field, smpSynch field, quality field, and message timing. Sampled value differences are written back to the channel binding table or electrical sampling chain parameters; smpCnt differences are written back to the sample count state edge; smpSynch and quality field differences are written back to the state consistency constraint graph or boundary frame action template library; message timing differences are written back to the phase residual table; unmatched state edges and residual gaps are written to the disturbance coverage matrix for generating subsequent secondary disturbance sequences.

[0152] In one specific embodiment, the computing device is an industrial control computer with a gigabit Ethernet interface. A relay protection tester inputs three-phase voltage and three-phase current to the modeled merging unit, a time synchronization device provides a synchronization signal to the modeled merging unit, and a message acquisition device is connected to the sampled value message output port of the merging unit. In the project configuration file, the sampled value control block is identified as MU_SV01, the APPID is 4001, the VLAN identifier is 100, each frame contains one ASDU, the sampling rate is 4000Hz, and the channel configuration includes A-phase voltage, B-phase voltage, C-phase voltage, A-phase current, B-phase current, and C-phase current. The above parameters are used to describe the data organization method in this embodiment and do not limit this method to only be applicable to this sampling rate, number of channels, or ASDU encapsulation method.

[0153] The computing device generates a first-round disturbance plan based on the disturbance coverage matrix. The disturbance plan includes: applying amplitude disturbance to the A-phase voltage input channel from 2.000s to 2.200s; controlling the time synchronization input to enter a synchronization loss state from 3.000s to 3.120s; controlling the time synchronization input to restore synchronization from 3.120s to 3.300s; and recording the packet acquisition link load flag from 4.000s to 4.100s. Each disturbance plan records the disturbance plan identifier, disturbance type, disturbance target, disturbance start time, disturbance end time, and disturbance flag.

[0154] In this embodiment, the collected sampled message observation data are as follows: For frame 8000, the smpCnt field is 0, the smpSynch field is synchronized, the quality field is valid, the A-phase voltage sample value is 5735, and the message observation time is 2.000000s; For frame 8001, the smpCnt field is 1, the smpSynch field is synchronized, the quality field is valid, the A-phase voltage sample value is 5792, and the message observation time is... 2.000250s; In frame 12000, the smpCnt field is 0, the smpSynch field is out of step, the quality field is suspicious, and the A-phase voltage sample value field retains the previous valid value. The message observation time is 3.000000s. In frame 12001, the smpCnt field is 1, the smpSynch field is out of step, the quality field is suspicious, and the A-phase voltage sample value field retains the previous valid value. The message observation time is 3.000251s. The computing device combines the above message observation data with the reference input time, disturbance flag, and reference input sample value into a dual-domain observation tuple.

[0155] Based on the response results within the single-channel disturbance window, the computing device obtains a record in the channel binding table: the physical input channel is the A-phase voltage input channel, corresponding to the first sampled value field in the sampled value message, the channel polarity is positive, the proportional relationship is determined by the input change and the sampled value field change, and the delay category is the response within the same sampling period. If the position of the A-phase voltage field recorded in the engineering configuration file is inconsistent with the position determined by the observed response, the subsequent electrical sampling chain parameter inversion and simulation message generation will be based on this channel binding table.

[0156] Based on the samples corresponding to the electrical disturbance domain, the computing device obtains the electrical sampling chain parameter record for the A-phase voltage channel, including the proportional coefficient, bias, equivalent sampling delay, and filter residual parameters. This parameter record is used to generate the A-phase voltage sample value field in S502. For samples corresponding to the synchronization disturbance domain or message acquisition link disturbance domain, the computing device does not use them as the main samples for the inversion of the A-phase voltage channel electrical sampling chain parameters to avoid changes in synchronization state or acquisition link entering the electrical sampling chain parameters.

[0157] Based on the field changes within the synchronization loss disturbance window, the computing device generates a state edge record. Before the transition, the state edge's state is: synchronization valid, quality valid, smpCnt continuously increasing, and normal sampled value output. After the transition, the state is: synchronization failed, quality questionable, smpCnt continuously increasing, and sampled value remaining at the previous valid value. The trigger condition for this state edge is that the synchronization disturbance flag enters the synchronization loss state. The output actions are: the smpSynch field is set to out of synchronization, the quality field is set to questionable, the sampled value field remains at the previous valid value, and the smpCnt field continues to increase according to the sampling counting rules.

[0158] Based on the message data before and after the state edge, the computing device generates a boundary frame action template. This template records the order of field changes in the state transition frame and adjacent frames, including the smpSynch field entering the out-of-synchronization state first, the quality field entering the suspicious state, the sampled value field retaining the previous valid value, and the smpCnt field maintaining a continuous increment. During simulation output, if the current simulation event matches the state edge, and the current output frame is within the boundary frame window corresponding to the state edge, the frame-level joint constraint solver calls the boundary frame action template to generate the corresponding fields.

[0159] The computing device establishes a phase residual table based on the message observation time and the ideal message time. Each entry in the phase residual table includes the sample count modulus, synchronization status, ASDU encapsulation mode, event status, and a representative residual value. For example, when the synchronization status is out of step, the event status is during synchronization loss, and the ASDU encapsulation mode is single ASDU encapsulation, the computing device writes the observation timing residuals under the same index into the same entry and uses this entry to constrain the output time of the corresponding message frame during the simulation phase. When the corresponding index does not exist during the simulation phase, the computing device searches for adjacent entries in the order of event status, synchronization status, ASDU encapsulation mode, and sample count modulus. If no entry is found, the basic message period is used to generate the output time, and this index is written into the residual gap record.

[0160] During closed-loop verification, the computing device aligns the simulated sampled value message sequence with the measured sampled value message sequence according to the smpCnt field, ASDU sequence number, disturbance flag, and message observation time. If the differences in the A-phase voltage sampled value field are concentrated within the electrical disturbance window, the computing device updates the channel binding table or electrical sampling chain parameters; if the differences in the smpCnt field are within the synchronization loss or synchronization recovery window, the computing device updates the sample count state edge; if the differences in the smpSynch field or quality field are near the state switching frame, the computing device updates the state consistency constraint graph or boundary frame action template library; if the message observation time difference is related to a specific synchronization state or event state, the computing device updates the phase residual table; if a certain state edge is not effectively triggered in the current disturbance plan, the computing device writes it into the uncovered state edge record in the disturbance coverage matrix and generates a secondary disturbance sequence accordingly.

[0161] The preset parameters in this embodiment include the disturbance observation window length, state edge judgment threshold, boundary frame window length, phase residual table backoff order, abnormal sample removal conditions, and channel response association conditions. The disturbance observation window length is used to determine the range of data participating in state edge identification before and after a disturbance occurs; the state edge judgment threshold is used to determine whether a field change constitutes a joint state change; the boundary frame window length is used to determine the range of messages participating in template extraction before and after a state switching frame; the phase residual table backoff order is used to select alternative entries when no entry is formed in the current index; the abnormal sample removal conditions are used to exclude observation data corresponding to missing messages, repeated acquisitions, or abnormal acquisition times; and the channel response association conditions are used to determine the binding relationship between the physical input channel and the sampled value message field.

[0162] The above formula describes the processing relationship of the computing device in calculating parameters, matching state constraints, and generating message fields based on observation data. The perturbation window length, state edge judgment threshold, boundary frame window length, phase residual table backoff order, and abnormal sample removal conditions in the formula can be determined based on the engineering configuration, test conditions, and acquired data of the modeled and merged unit. For sampled value field encoding, message encapsulation, time alignment, and routine statistical processing, those skilled in the art can use conventional implementation methods in sampled value message testing, which will not be elaborated upon here.

[0163] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented in software, the above embodiments can be implemented, in whole or in part, as a computer program product. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution.

[0164] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0165] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A simulation modeling method for merging units in an intelligent substation, characterized in that, Includes the following steps: The project configuration file is parsed to establish a sampled value message protocol template, and a disturbance coverage matrix is ​​established based on the state edge to be identified; Generate homogeneous disturbance fingerprints based on the disturbance coverage matrix, collect reference input data and sampled value message data, and generate dual-domain observation tuples; Cross-domain causality separation, channel binding verification, and electrical sampling chain parameter inversion are performed based on dual-domain observation tuples; The cross-domain causal separation includes: dividing the disturbance observation window according to the disturbance record set, dividing the disturbance type into electrical disturbance domain, synchronization disturbance domain, network observation link disturbance domain and combined disturbance domain, and determining the triggering source of sample value change, smpCnt field change, smpSynch field change, quality field change and message observation timing change according to the disturbance mark, the order of field change and the triggering result of the state edge to be identified; The channel binding verification includes: forming a configuration-side channel set based on the engineering configuration file, forming an observation-side field set based on the sampled value message data, applying a single-channel perturbation to the physical input channel, calculating the channel response correlation degree between the physical input channel and the sampled value message field, and generating a channel binding table in combination with the phase and measurement type in the engineering configuration file; A phase residual table, a state consistency constraint diagram, and a boundary frame action template library are established based on the dual-domain observation tuples. The state consistency constraint graph includes a joint set of state nodes, a set of observed state edges, a set of state edge triggering conditions, and a set of state edge output actions. The joint state in the set of joint state nodes consists of synchronization state, quality bit state, sampling count state, sample value processing state, and message observation timing state; The simulation sampled value message sequence is generated by solving the frame-level joint constraints, and the simulation sampled value message sequence is then checked in a closed loop. The frame-level joint constraint solution includes: reading the sampled value message protocol template, channel binding table, electrical sampling chain parameters, phase residual table, state consistency constraint diagram and boundary frame action template library, taking the same message frame or intra-frame ASDU as the processing object, and determining the sampled value field, smpCnt field, smpSynch field, quality field, ASDU arrangement information and message output time; When different constraints produce different results for the same field, the output result is determined in the order of protocol template constraint, boundary frame action constraint, state consistency constraint, sample count constraint, sample chain constraint, and phase residual constraint. During closed-loop verification, the simulated sampled value message sequence is aligned with the measured sampled value message sequence, and the corresponding model data is updated based on field differences and message timing differences.

2. The simulation modeling method for merging units in an intelligent substation according to claim 1, characterized in that, The project configuration file includes an SCD file, an ICD file, or a CID file; The sampled value message protocol template includes an Ethernet field template, a VLAN field template, an SV message header field template, an ASDU field template, a sampled value field template, a smpCnt field template, a smpSynch field template, and a quality field template; The state edges to be identified include synchronization state change edges, mass bit change edges, sampling count change edges, sampling value processing state change edges, and message observation timing change edges. The perturbation coverage matrix is ​​indexed by the state edge to be identified and the perturbation type, and is used to record whether the corresponding state edge to be identified is effectively observed for each perturbation type.

3. The simulation modeling method for merging units in an intelligent substation according to claim 1, characterized in that, The homogeneous disturbance fingerprint includes electrical disturbance sequences, synchronization disturbance sequences, and network observation link disturbance sequences; The electrical disturbance sequence is applied to the voltage input channel or the current input channel, the synchronization disturbance sequence is applied to the time synchronization input terminal, and the network observation link disturbance sequence is applied to the message acquisition link; The electrical disturbance sequence, synchronization disturbance sequence, and network observation link disturbance sequence are arranged according to a unified time base, so that the reference input data, synchronization state changes, and sampled value message data have corresponding time stamps.

4. The simulation modeling method for merging units in an intelligent substation according to claim 1, characterized in that, The dual-domain observation tuple consists of reference input data, a set of sampled value fields, a set of sampled value message control fields, a set of time fields, and a set of disturbance records under the same observation index. The set of control fields for the sampled value message includes the smpCnt field, the smpSynch field, the quality field, and the ASDU sequence number; The time field set includes the reference input time and the message observation time; The disturbance record set includes a disturbance plan identifier, a disturbance type, a disturbance target, and a disturbance tag.

5. The simulation modeling method for merging units in an intelligent substation according to claim 1, characterized in that, The trigger source is a sample belonging to the electrical disturbance domain or a combined disturbance domain containing the electrical disturbance domain, used for electrical sampling chain parameter inversion.

6. The simulation modeling method for merging units in an intelligent substation according to claim 1, characterized in that, The channel binding table includes a physical input channel index, a sampled value message field index, channel polarity, proportional relationship, and delay category; When the channel semantics in the engineering configuration file are inconsistent with the observed response, the electrical sampling chain parameter inversion and simulation sample value field generation use the channel binding table as the basis for channel mapping.

7. The simulation modeling method for merging units in an intelligent substation according to claim 1, characterized in that, The electrical sampling chain parameter inversion includes: determining the correspondence between the sampled value message field and the physical input channel according to the channel binding table; selecting samples triggered by the electrical disturbance domain or samples triggered by the combined disturbance domain containing the electrical disturbance domain from the dual-domain observation tuple as parameter inversion samples; and determining the scaling factor, bias, equivalent sampling delay and filter residual parameters of each channel based on the parameter inversion samples. Among them, samples triggered by synchronous disturbance domain or network observation link disturbance domain are not used as the main samples for electrical sampling chain parameter inversion.

8. The simulation modeling method for merging units in an intelligent substation according to claim 1, characterized in that, The phase residual table is established based on the message observation time, the ideal message time, and the basic delay of the observation link. The phase residual table uses the sample count modulus, synchronization status, ASDU encapsulation mode, and event status as phase residual indexes, and records the message observation time series residual, residual representative value, and sample source in the table entry corresponding to each phase residual index; When generating the simulation sampled value message sequence, if the current phase residual index does not have a corresponding entry, the adjacent entries are searched in the order of event state, synchronization state, ASDU encapsulation mode and sample count modulus. If no entry is found, the message output time is generated using the basic message period.

9. The simulation modeling method for merging units in an intelligent substation according to claim 1, characterized in that, The boundary frame action template library is generated based on the observed state edges, and records the sampling value action, smpCnt action, smpSynch action, quality action and observation timing action before and after the state switching index; When generating the simulation sampled value message sequence, if the current output frame is within the corresponding boundary frame window, the boundary frame action template library is called to generate the corresponding field changes.

Images