Parallel error prevention and checking method and system for objectification data of power grid dispatching instructions

Abstract

The application discloses a kind of parallel error prevention check methods of power grid dispatching instruction objectification data, it is related to power system dispatching automation technical field, including the following steps: according to dispatching log data generation operation time sequence diagram, parse operation sequence, and extract instruction conflict feature of operation time sequence diagram;Based on beam forming technology, the vibration signal of the mechanical vibration signal of substation equipment obtained in advance is focused, and a voiceprint feature is generated;Operation sequence and instruction conflict feature are encoded to obtain operation time sequence data;Parallel check processing operation time sequence data and voiceprint feature, to generate a check result.The application is used to solve the technical problem that single dependence logical rule check cannot effectively identify device actual physical state anomaly, leading to misjudgment and security risk.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power system dispatch automation, more particularly, the present application relates to a parallel anti-misoperation checking method for power grid dispatch instruction object data. BACKGROUND

[0002] Power grid dispatching is an important link to ensure the safe and stable operation of the power system, and dispatching instructions need to be accurately conveyed and executed in complex operation sequences. However, in actual application scenarios, dispatching operations often involve the coordinated execution of multiple devices and multiple links, and if only relying on dispatching logs for post-checking, it is difficult to timely discover instruction conflicts, sequence errors or device abnormal execution and other problems. With the increasing intelligence of the power grid, the real-time, accuracy and multi-dimensional verification capabilities of dispatching links for instruction checking are put forward with higher demands, requiring the ability to identify potential risks during operation to avoid large-scale power grid accidents caused by instruction mismatch or device abnormalities.

[0003] To meet the above needs, existing solutions propose to build an object-based operation model based on log data, to structure the dispatching instructions, and to combine instruction conflict detection algorithms to realize logical checking of the operation sequence. This method can to some extent find the logical inconsistency of the instructions and avoid misoperation caused by instruction conflicts. At the same time, some research attempts to introduce device operation information collected by sensors into the checking process to make up for the shortcomings of single logical checking, so as to realize auxiliary verification of dispatching operations.

[0004] However, existing solutions generally rely on single-threaded data sources, focus on logical-level instruction checking, lack direct association with device physical execution process, and are difficult to achieve synchronous verification of operation authenticity. When there are mechanical abnormalities, voiceprint feature shifts or failure to respond to instructions during device execution, relying solely on logical checking cannot accurately identify. Especially in multi-device parallel operation and complex working conditions, existing methods cannot establish a correspondence between logical states and physical states, resulting in insufficient accuracy and real-time performance of anti-misoperation checking, and potential safety hazards.

[0005] For example, the invention patent announcement with announcement number: CN114744617A, a kind of power distribution network based on the anti-misoperation check method of primary instruction simulation operation, by constructing visual distribution network model, and the actual input parameter of distribution network is as the operation input parameter of distribution network model;Primary instruction information is obtained, and the primary instruction simulation operation is input into distribution network model;Distribution network model carries out simulation operation to primary instruction simulation operation;Distribution network model according to the feedback output instruction of simulation result to distribution network.The invention establishes distribution network model, uses simulation environment to carry out primary equipment instruction simulation operation, uses the running mode under simulation environment to carry out safety anti-misoperation check, simulation environment and real-time environment isolation, any operation under simulation environment does not affect real-time power grid operation environment, according to simulation result, operation is carried out to actual power grid, effectively isolation and anti-misoperation effect are played.

[0006] For example, the invention patent announcement with announcement number: CN118336927A, a kind of substation one-key sequence control anti-misoperation check method, including the following steps: using anti-misoperation check device to import the point table of communication between monitoring host computer and anti-misoperation host computer, establishing connection and signal verification, connecting anti-misoperation check device and intelligent anti-misoperation host computer, sending test signal to intelligent anti-misoperation host computer through anti-misoperation check device, observing the response signal and corresponding signal photon card movement prompted by intelligent anti-misoperation host computer alarm frame, verifying whether the sent signal and the signal responded by intelligent anti-misoperation host computer are consistent;The invention uses anti-misoperation check device to import the point table of communication between monitoring host computer and anti-misoperation host computer, can ensure the accuracy and integrity of point table data, lay a solid foundation for subsequent check work, connect anti-misoperation check device and intelligent anti-misoperation host computer, and verify the consistency of signal, effectively guarantee the stability and real-time performance of data transmission.

[0007] The above disclosed technical solutions at least have the following technical problems:

[0008] Lack of direct association with physical execution process of equipment, unable to realize multi-dimensional synchronous check of instruction authenticity;The check method of simulation environment or communication level cannot find problems such as device non-response, mechanical abnormality or voiceprint feature deviation in actual operation process in real time;In the process of multi-device parallel operation and complex working conditions, the corresponding relationship between logical state and physical state is lacked, which leads to insufficient accuracy and real-time performance of anti-misoperation check.

[0009] In view of the above problems, the present application provides a solution. SUMMARY

[0010] In order to overcome the above-mentioned defects of the prior art, embodiments of the present application provide a parallel error prevention and checking method for power grid dispatching instruction objectification data, which objectifies the dispatching instruction, introduces multi-dimensional physical feedback data, constructs a parallel correspondence relationship between the logical state of the dispatching instruction and the physical execution state of the equipment, and realizes real-time checking of the instruction and the feedback during operation, so as to solve the technical problem that the abnormality of the actual physical state of the equipment cannot be effectively identified due to the single dependence on logical rule checking, thereby causing misjudgment and safety hazards.

[0011] In order to achieve the above-mentioned purpose, the present application provides the following technical solutions:

[0012] A parallel error prevention and checking method for power grid dispatching instruction objectification data, characterized in that the method comprises the following steps: generating an operation time sequence diagram according to dispatching log data, analyzing an operation sequence, and extracting instruction conflict characteristics of the operation time sequence diagram; focusing on a vibration signal of a pre-acquired mechanical vibration signal of a substation equipment based on a beamforming technology to generate a voiceprint feature; encoding the operation sequence and the instruction conflict characteristics to obtain operation time sequence data; and checking the operation time sequence data and the voiceprint feature in parallel to generate a checking result.

[0013] In a preferred embodiment, the voiceprint feature comprises a frequency spectrum distribution feature, an energy attenuation feature, and a time duration feature.

[0014] In a preferred embodiment, the generation of the voiceprint feature comprises the following specific steps: based on the spatial distribution relationship of an acoustic sensor array, performing sound source positioning on the collected vibration signal to obtain sound source position information, wherein the vibration signal comprises vibration waveform data; performing phase alignment processing on the vibration waveform data to obtain coherent waveform data; combining the sound source position information and the coherent waveform data to perform waveform synthesis to generate an enhanced sound wave signal; extracting the frequency domain feature and the time domain feature of the enhanced sound wave signal, and performing feature fusion to obtain a pattern voiceprint feature.

[0015] In a preferred embodiment, the parallel checking of the operation time sequence data and the voiceprint feature to generate a checking result comprises: performing logical compliance determination on the operation time sequence data to obtain a logical instruction state; performing feature determination on the voiceprint information to obtain a physical execution state; performing action type matching on the operation instruction type and the pattern voiceprint feature; performing double-state synchronous receiving and combination mapping processing on the logical instruction state and the physical execution state to generate a joint state identifier; and performing state decision logic processing on the joint state identifier to generate a checking result.

[0016] In a preferred implementation, the operation sequence and instruction conflict feature are encoded to obtain operation timing data, specifically: the operation sequence is reorganized based on the time correlation between operation instructions to obtain a reorganized operation sequence; the pattern feature of the instruction conflict feature is extracted and enhanced to obtain an enhanced conflict feature; the reorganized operation sequence and the enhanced conflict feature are fused and jointly encoded to generate fused encoding data.

[0017] In a preferred implementation, the waveform synthesis includes: establishing beamforming weights according to the sound source position information to obtain spatial filtering parameters; and dynamically weighting and synthesizing the coherent waveform data using the spatial filtering parameters.

[0018] In a preferred implementation, the operation instruction type and mode voiceprint feature are matched to obtain the action type, specifically: based on the action category of the operation instruction, a standard feature set in the voiceprint feature template library is called; the frequency spectrum distribution feature of the to-be-tested signal is matched with the reference frequency spectrum pattern in the standard feature set to calculate the frequency spectrum similarity; the energy attenuation feature of the to-be-tested signal is dynamically time-warped with the reference energy envelope curve in the standard feature set to quantify the shape consistency; the time duration feature of the to-be-tested signal is compared with the legal operation time window to generate a time sequence compliance factor; the frequency spectrum similarity, the shape consistency, and the time sequence compliance factor are fused, and the final action type matching degree is calculated through a weighted decision model, and a matching result is output.

[0019] In a preferred implementation, the logical instruction state and the physical execution state are synchronously received and combined and mapped to generate a joint state identifier, specifically: a time synchronization window between the logical instruction state sequence and the physical execution state sequence is established, and the two sequences are time-aligned; the aligned logical instruction state sequence is traversed, for each instruction state, a physically associated execution state in the synchronization window is searched to form an associated set; based on the operation instruction type and the action type matching result, the matching confidence of the logical and physical state combination in the associated set is calculated to generate a state matching degree matrix; and based on the state matching degree matrix, the logical instruction state sequence and the physical execution state sequence are combined and mapped to generate a unique joint state identifier for a successfully mapped state pair.

[0020] In a preferred implementation, based on the spatial distribution relationship of the acoustic sensor array, the collected vibration signal is subjected to sound source positioning to obtain sound source position information, including: based on the geometric distribution relationship of the acoustic sensor array, the arrival time difference of the vibration waveform is calculated to obtain sound source direction data; and based on the sound wave propagation characteristics, the sound source direction data is subjected to three-dimensional space mapping to generate the sound source position information.

[0021] A parallel error prevention checking system for power grid dispatching instruction objectified data, comprising: a timing analysis module, configured to generate an operation timing diagram according to dispatching log data, analyze an operation sequence, and extract instruction conflict features of the operation timing diagram; a voiceprint generation module, configured to perform focusing processing on a vibration signal of a pre-acquired mechanical vibration signal of a substation device based on a beamforming technology, and generate voiceprint features; a timing encoding module, configured to encode the operation sequence and the instruction conflict features to obtain operation timing data; and a parallel checking module, configured to perform parallel checking processing on the operation timing data and the voiceprint features to generate a checking result.

[0022] The technical effects and advantages of the parallel error prevention checking method and system for power grid dispatching instruction objectified data are as follows:

[0023] The present application generates an operation timing diagram by acquiring dispatching log data, generates voiceprint information by processing a mechanical vibration signal based on a beamforming technology, encodes and converts the operation sequence and instruction conflict features to obtain operation timing data, and performs parallel checking processing on the operation timing data and voiceprint information, thereby achieving double protection of logical rule checking and device physical state verification. Then, the corresponding relationship between logical instruction states and physical execution states is established through time alignment and action type matching, the joint state identifier is generated through double-state synchronous receiving and combination mapping processing, and finally the checking result is generated through state decision logic processing. Through multi-dimensional feature fusion and parallel processing mechanism, the accuracy and reliability of power grid dispatching instruction checking can be significantly improved, and the technical problem that the actual physical state of the device cannot be effectively identified due to single dependence on logical rule checking, thereby causing misjudgment and safety hazards, is effectively solved. BRIEF DESCRIPTION OF DRAWINGS

[0024] Figure 1 A parallel error prevention checking method flowchart for power grid dispatching instruction objectified data is provided for the embodiments of the present application.

[0025] Figure 2 A parallel error prevention checking system module structure diagram for power grid dispatching instruction objectified data is provided for the embodiments of the present application. DETAILED DESCRIPTION

[0026] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0027] Embodiment 1, Figure 1 A parallel error prevention checking method for power grid dispatching instruction objectified data is provided, comprising the following steps:

[0028] S1, generating an operation timing diagram according to the dispatch log data, analyzing the operation sequence, and extracting instruction conflict features of the operation timing diagram;

[0029] S2, focusing the vibration signal of the substation equipment mechanical vibration signal pre-acquired based on the beamforming technology to generate a voiceprint feature;

[0030] S3, encoding the operation sequence and the instruction conflict features to obtain operation timing data;

[0031] S4, parallelly verifying the operation timing data and the voiceprint feature to generate a verification result.

[0032] The embodiment can structure and time sequence analyze the dispatch operation process by acquiring the dispatch log data and generating the operation timing diagram, realize intuitive modeling and traceability of the operation sequence, improve the accuracy of logical verification by extracting the instruction conflict features of the operation timing diagram and combining the encoding conversion processing, collect the mechanical vibration signal of the substation equipment, focus the vibration waveform data by using the beamforming technology, generate the voiceprint information of the equipment, enable the real execution state of the equipment to be accurately captured in the form of the voiceprint feature, realize double perception of the logical layer and the physical layer, and parallelly verify the operation timing data and the voiceprint information to significantly improve the authenticity, real-time performance and multi-dimensional verification capability of the error prevention verification.

[0033] S1, generating an operation timing diagram according to the dispatch log data, analyzing the operation sequence, and extracting instruction conflict features of the operation timing diagram;

[0034] Specifically:

[0035] The dispatch data interface is accessed by data, such as database direct connection or message bus subscription, to acquire the dispatch log in a target time period in real time or in batches;

[0036] The redundant items, missing values and abnormal time stamps in the log data are corrected or removed to ensure the integrity and consistency of the input data;

[0037] The core elements are extracted from the log, including operation instructions such as closing, opening, switching, device identification, execution time stamp and execution feedback state;

[0038] After the log data is cleaned and extracted, the operation timing diagram is constructed based on the extracted time stamp and instruction sequence, and the operation timing diagram is analyzed after being generated.

[0039] It should be noted that the dispatching log data refers to a set of operation information recorded by the power grid dispatching automation system in real time, and the content includes operation instructions issued by dispatchers, execution time stamps, target device identifiers, execution state feedback, etc. The dispatching log data is generally stored in the form of a structured or semi-structured database, such as a relational database table, a distributed log file, or a message queue, and can completely reflect the whole process of dispatching operations.

[0040] The operation timing diagram refers to a directed graph constructed by taking time as the horizontal axis, taking devices or operation objects as the vertical axis, labeling dispatching instructions as event nodes, and constructing time-dependent relationships and logical constraint relationships between events. The operation sequence refers to a set of operation steps extracted in time sequence or logical constraints from the operation timing diagram.

[0041] In this embodiment, based on the generated operation timing diagram, a set of formal conflict rules is predefined based on power grid safety operation procedures such as the five-prevention rule and the power grid topology structure.

[0042] Using a graph traversal algorithm, all nodes (device operation instructions) and edges (timing and logical relationships between instructions) in the operation timing diagram are systematically accessed, the execution order of the instructions is simulated, and the context environment of each instruction is checked to see if it meets the risk criteria.

[0043] When traversing each node of the graph, the current operation instruction and its context, such as predecessor instructions, successor instructions, and the current state of associated devices, are matched in real time with the conflict rule library.

[0044] When an operation instruction is detected to conflict with a rule, the system generates a structured conflict feature record.

[0045] All detected and marked conflict features are summarized to form a complete instruction conflict feature set.

[0046] In this embodiment, a set of formal conflict rules is predefined, which are used to determine whether the instruction sequence meets the risk criteria, specifically:

[0047] Electrical interlocking rule: specifies the operation interlocking relationship between devices with different electrical states, for example: only when the circuit breaker is in the open state, can the downstream disconnecting switch or live grounding knife switch be operated;

[0048] Logical interlocking rule: specifies the logical order between operation instructions, for example: the load switch must be opened first, and then the fuse is operated;

[0049] Parallel power supply conflict rule: checks whether power points from different sources can be connected together irregularly;

[0050] Load loss rule: check if a series of operations will accidentally cause an important load to lose power.

[0051] In this embodiment, the conflict feature record is specifically:

[0052] Conflict position: occurring on the corresponding instruction node in the operation timing diagram;

[0053] Conflict type: specific rule violated;

[0054] Conflict description: detailed text description of the conflict reason;

[0055] Risk level: severity of the consequences that the conflict may cause;

[0056] It should be noted that the instruction conflict feature is identified from the operation timing diagram to identify the instruction combination or timing exception that does not conform to the power grid operation regulation.

[0057] S2, based on the beam forming technology, focusing processing is performed on the vibration signal of the substation equipment mechanical vibration signal obtained in advance, and a voiceprint feature is generated.

[0058] Specifically,

[0059] Based on the spatial distribution relationship of the acoustic sensor array, the vibration signal collected is subjected to sound source positioning to obtain sound source position information, and the vibration signal includes vibration waveform data;

[0060] The vibration waveform data is subjected to phase alignment processing to obtain coherent waveform data;

[0061] The sound source position information and the coherent waveform data are combined for waveform synthesis to generate an enhanced sound wave signal;

[0062] The frequency domain features and time domain features of the enhanced sound wave signal are extracted, and feature fusion is performed to obtain a mode voiceprint feature.

[0063] In this embodiment, based on the spatial distribution relationship of the acoustic sensor array, the vibration signal collected is subjected to sound source positioning to obtain sound source position information, including:

[0064] Based on the geometric distribution relationship of the acoustic sensor array, the arrival time difference of the vibration waveform is calculated to obtain sound source azimuth data;

[0065] After obtaining the sound source azimuth data, three-dimensional mapping needs to be performed in combination with the sound wave propagation characteristics, and in the two-dimensional plane, the azimuth angle can be inversely deduced through the arrival time difference;

[0066] Under the condition of a three-dimensional array, the sound source spatial coordinates can be solved by using a plurality of TDOA equations;

[0067] By combining the characteristics of sound wave propagation, the sound source location data is mapped in three-dimensional space to generate sound source location information.

[0068] In this embodiment, the specific formula for calculating the arrival time difference of the vibration waveform is as follows:

[0069]

[0070] Distance from the sound source to the sensor:

[0071]

[0072] In the formula, , , Let be the coordinates of the sound source location to be determined. Let be the propagation distance from the sound source to the i-th sensor;

[0073] The specific formula for the propagation characteristics of sound waves is:

[0074]

[0075] In the formula, c is the speed at which sound waves propagate in air or solids.

[0076] In this embodiment, the azimuth formula is:

[0077]

[0078] In the formula, , The coordinates are for reference sensor.

[0079] In this embodiment, the TDOA equation formula is:

[0080]

[0081] In the formula, This represents the solution function based on geometric optimization.

[0082] Based on the geometric distribution of multiple sensors, the sound wave propagation model, and time difference calculation, the three-dimensional spatial coordinates of the sound source are obtained, forming the sound source location information.

[0083] It should be noted that the geometric distribution relationship of the acoustic sensor array refers to the specific arrangement and relative position coordinates of each acoustic sensor unit in the array in three-dimensional space. This relationship is a known prerequisite for spatial localization of the sound source. Usually, this geometric distribution relationship is determined by precise measurement after the array is deployed and is pre-stored in the system.

[0084] In this embodiment, the waveform synthesis includes:

[0085] The beamforming weight is established according to the sound source position information, and spatial filtering parameters are obtained;

[0086] The coherent waveform data is dynamically weighted and synthesized by using the spatial filtering parameters;

[0087] A beamforming weight vector is constructed according to the sound source position information and the geometric distribution relationship of the acoustic sensor array.

[0088] After the beamforming weight is obtained, the beamforming weight is weighted and synthesized with the coherent waveform data received by each sensor to obtain an enhanced acoustic signal.

[0089] In this embodiment, the weight vector is specifically as follows:

[0090]

[0091] In the formula, is a beamforming weight vector, is an array manifold vector, which describes the phase response of each sensor in the array in the target direction.

[0092] In this embodiment, the array flow vector expression is as follows:

[0093]

[0094] In the formula, is the frequency of the acoustic signal, is the acoustic velocity, is the spatial coordinate vector of the mth acoustic sensor, is the unit direction vector of the sound source.

[0095] In this embodiment, the acoustic signal is specifically as follows:

[0096]

[0097] In the formula, is the synthesized enhanced acoustic signal, is the conjugate transpose of the weight vector, is the weight factor corresponding to the mth sensor, is the coherent waveform signal received by the mth sensor.

[0098] It should be noted that the spatial filtering parameters refer to a set of weighting factors and delay compensation amounts set for different sensor received signals in the beamforming technology based on the acoustic sensor array, in order to suppress interference and highlight the target sound source in space.

[0099] In this embodiment, the physical characteristics of the operating acoustic fingerprint are specifically spectral distribution characteristics, energy attenuation characteristics and time duration characteristics.

[0100] The spectral distribution feature is obtained by pre-emphasizing, framing and windowing the enhanced sound wave signal after beamforming to obtain a short-time stationary signal segment;

[0101] The power spectral density of each frame signal is calculated by fast Fourier transform;

[0102] In this embodiment, the power spectral density calculation formula is:

[0103]

[0104] In the formula, Pm(k) is the power value of the mth frame signal at the kth frequency point, is the nth sampling point in the mth frame, is the window function, is the Fourier transform function;

[0105] In this embodiment, the energy decay feature extraction formula is:

[0106]

[0107] In the formula, is the mth frame energy, is the frame length;

[0108] In this embodiment, the time duration feature extraction formula is:

[0109]

[0110] In the formula, The monitoring start point and end point are included, and the duration is the difference between the two;

[0111] It should be noted that the spectral distribution feature refers to the energy distribution of different frequency components in the sound signal, which can uniquely identify which device and whether closing or opening operation is performed; the energy decay feature refers to the process mode of the gradual weakening of the sound signal intensity from the peak to the disappearance, which reflects the health status of the mechanical structure, and can effectively judge whether the operation process is smooth, complete, has jamming or abnormal impact; the time duration feature refers to the absolute time length consumed by one mechanical operation action from start to end, a simple but key compliance criterion, which is used to judge whether the operation speed is within the standard allowed safe time range, and too fast or too slow indicates potential failure.

[0112] S3, encode the operation sequence and instruction conflict feature to obtain operation timing data.

[0113] Specifically,

[0114] Reorganize the operation sequence based on the time correlation between operation instructions to obtain a reorganized operation sequence;

[0115] Extract the mode features of the instruction conflict features and perform feature enhancement to obtain enhanced conflict features;

[0116] Fuse the reorganized operation sequence and the enhanced conflict features and jointly encode to generate fusion encoding data.

[0117] In this embodiment, the original operation sequence is mapped to obtain a reorganized operation sequence, the conflict features are mapped to a feature space through a clustering recognition method, and a weighted amplification method is used to improve the conflict sensitivity to obtain enhanced conflict features;

[0118] According to the operation sequence and the instruction conflict features, the encoding conversion further includes mapping the enhanced conflict features into a conflict constraint matrix and mapping the reorganized operation sequence into a time sequence vector matrix, and jointly encoding to generate operation time sequence data.

[0119] In this embodiment, the conflict features are specifically:

[0120]

[0121] In the formula, is a feature weight matrix, is a conflict feature vector.

[0122] In this embodiment, the specific formula is:

[0123]

[0124]

[0125] In the formula, represents an embedding vector, is a conflict constraint matrix, is a time sequence matrix of the operation sequence;

[0126] In this embodiment, the fusion encoding data is specifically:

[0127]

[0128] In the formula, b is a bias vector, and a baseline offset is introduced for each output dimension;

[0129] It should be noted that the operation time sequence data refers to the operation information sequence organized and encoded in time sequence during the operation of the execution device, which not only contains the sequence of each operation, but also fuses the conflict features related to the operation.

[0130] S4, parallelly check the operation timing data and the voiceprint feature to generate a check result.

[0131] Specifically,

[0132] perform logical compliance determination on the operation timing data to obtain a logical instruction state;

[0133] perform feature determination on the voiceprint information to obtain a physical execution state;

[0134] match the action type according to the operation instruction type and the mode voiceprint feature;

[0135] perform double-state synchronization receiving and combined mapping processing on the logical instruction state and the physical execution state to generate a joint state identifier;

[0136] perform state decision logic processing on the joint state identifier to generate a check result.

[0137] In the embodiment, the matching of the action type according to the operation instruction type and the mode voiceprint feature is specifically:

[0138] based on the action category of the operation instruction, a standard feature set in a voiceprint feature template library is called;

[0139] the frequency spectrum distribution feature of the to-be-tested signal is matched with a reference frequency spectrum mode in the standard feature set to calculate a frequency spectrum similarity;

[0140] the energy attenuation feature of the to-be-tested signal is dynamically time-warped with a reference energy envelope curve in the standard feature set to quantify the shape consistency;

[0141] the time duration feature of the to-be-tested signal is compared with a legal operation time window to generate a timing compliance factor;

[0142] the frequency spectrum similarity, the shape consistency and the timing compliance factor are fused, a final action type matching degree is calculated through a weighted decision model, and a matching result is output.

[0143] the action category indicated by the operation instruction is parsed based on instruction text or objectified data structure, an action type corresponding to the instruction is extracted from a predefined action set, and the action type is specifically:

[0144] operation records are extracted from the scheduling log, and key fields, i.e., action type, device ID, action initiation timestamp, initiator and permission, pre-state and expected post-state, are parsed;

[0145] the timestamps of the records are normalized to a unified clock, and the monotonicity and overlap conflict of the time sequence are detected;

[0146] Verify whether the initiator and the ticket have the permission to perform the action on the target device; check whether the action type is in the allowed table;

[0147] Verify whether the state transition before and after the action is consistent with the device model;

[0148] Match the joint state with the action matching degree as a decision input, and use a table lookup rule or a risk score function to determine the final verification result;

[0149] Compare other instructions in the same control domain within a similar time to identify mutually exclusive or conflicting instructions.

[0150] In this embodiment, the spectral similarity is specifically as follows:

[0151]

[0152] In the formula, is the spectral similarity, is the amplitude of the to-be-tested signal at the ith frequency point, is the reference amplitude of the template at the ith frequency point; is the number of frequency points.

[0153] In this embodiment, the shape consistency is specifically as follows:

[0154]

[0155] In the formula, is the energy shape consistency score, is the to-be-tested energy envelope sequence, is the template energy envelope sequence, is the minimum cumulative cost of the two sequences, is the sequence length.

[0156] In this embodiment, the timing compliance factor is specifically as follows:

[0157]

[0158] In the formula, is the timing compliance factor, is the template standard duration.

[0159] According to the business importance, weights are assigned to the three types of indexes, and the three types of indexes are fused into a single action matching degree, which is specifically as follows:

[0160]

[0161] In the formula, is the final action matching degree, , , is the weight of the three types of features.

[0162] In this embodiment, the risk score formula is:

[0163]

[0164] In the formula, is a comprehensive risk score, , , is a weight coefficient, reflecting the contribution of logic, physics and matching degree to risk.

[0165] It should be noted that the risk score is constructed, and the threshold interval is mapped to a specific disposal level to generate a verification result R; the final verification result is compared through a preset risk threshold interval, if the risk score Q is less than the preset threshold, the verification passes; the risk score is between the preset threshold, listed as a suspected result, and the verification is audited; the risk score is higher than the threshold, and does not pass and reports to the system.

[0166] In this embodiment, the double-state synchronization receiving and combined mapping processing of the logical instruction state and the physical execution state is specifically:

[0167] A time synchronization window is established between the logical instruction state sequence and the physical execution state sequence, and the two sequences are time-aligned;

[0168] The aligned logical instruction state sequence is traversed, and for each instruction state, all time-related physical execution states in the synchronization window are searched to form an associated set;

[0169] Based on the matching result of the operation instruction type and the action type, the matching confidence of the logical and physical state combination in the associated set is calculated, and a state matching degree matrix is generated;

[0170] According to the state matching degree matrix, the logical instruction state sequence and the physical execution state sequence are optimally combined, and a unique joint state identifier is generated for each successfully mapped state pair.

[0171] After time alignment, the logical instruction state sequence is traversed, and for each logical instruction state, all time-related physical execution states in the synchronization window are searched to form an associated set of logical states and multiple physical states;

[0172] Based on the matching rules of the operation instruction type and the action type, each logical-physical associated set is combined and operated, the matching confidence of each combination is calculated, and a state matching degree matrix describing the global association relationship is obtained.

[0173] In this embodiment, the time alignment formula is:

[0174]

[0175] In the formula, is a logical instruction state sequence, is a physical execution state sequence, is a time synchronization window.

[0176] In the embodiment, the associated set formula is:

[0177]

[0178]

[0179] In the formula, is a logical instruction state physical associated set; is a logical instruction state physical associated set;

[0180] In the embodiment, the state matching degree matrix formula is:

[0181]

[0182] In the embodiment, the joint state identification formula is:

[0183]

[0184]

[0185] In the formula, is a logical state i and j matching degree, is a mapping threshold value, used to determine whether the logical and physical states can establish effective mapping.

[0186] It should be noted that the operation timing data refers to the action sequence and metadata issued and recorded by the scheduling system, the logical instruction state refers to the instruction compliance determined based on the operation timing and rules, the physical execution state refers to whether the action determined based on the acoustic fingerprint comparison has actually occurred in the physical layer; the bidirectional associated set is a multiple correspondence relationship that can be formed between the logical state and the physical state within the time synchronization window, one logical state can correspond to multiple physical states, and one physical state can also correspond to multiple logical states; the state matching degree matrix is a two-dimensional matrix used to describe the matching degree of the logical state and the physical state in the action category and the time characteristic, and embodies the many-to-many relationship.

[0187] Embodiment 2, Figure 2 A system of a parallel error prevention verification method of power grid scheduling instruction objectification data is given, which comprises a timing analysis module, a conflict extraction module, a voiceprint generation module, a timing coding module and a parallel verification module, and there is a connection between the modules:

[0188] a timing analysis module, configured to generate an operation timing diagram according to the scheduling log data, analyze the operation sequence, and extract instruction conflict features of the operation timing diagram;

[0189] a voiceprint generation module, configured to perform focusing processing on the vibration signals of the pre-acquired mechanical vibration signals of the substation equipment based on a beamforming technology, and generate voiceprint features;

[0190] a timing encoding module, configured to encode the operation sequence and the instruction conflict features, and obtain operation timing data;

[0191] a parallel verification module, configured to perform parallel verification processing on the operation timing data and the voiceprint features, and generate a verification result.

[0192] The above formulas are all dimensionless numerical calculations, the formulas are obtained by software simulation of a large amount of data to obtain a formula of the most recent real situation, and the preset parameters in the formula are set by a person skilled in the art according to the actual situation.

[0193] The above embodiments can be realized wholly or partially by software, hardware, firmware or any combination thereof. When realized by software, the above embodiments can be realized wholly or partially in the form of a computer program product.

[0194] Those skilled in the art can realize that the modules and algorithm steps of the examples described in combination with the embodiments disclosed herein can be realized by electronic hardware or a combination of computer software and electronic hardware. Whether the functions are realized by hardware or software depends on the specific application and design constraints of the technical solutions. A person skilled in the art can use different methods to realize the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0195] In addition, each functional module in each embodiment of the present application can be integrated in one processing module, or each module can exist physically alone, or two or more modules can be integrated in one module.

[0196] The above is merely specific embodiments of the present application, but the protection scope of the present application is not limited thereto, any person skilled in the art can easily think of changes or replacements within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

[0197] Finally, the above is only the preferred embodiments of the present application, and is not used to limit the present application, any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application should be included in the protection scope of the present application.

Claims

1. A parallel error prevention checking method for objectifying data of power grid dispatching instructions, characterized in that, The method comprises the following steps: According to the scheduling log data, the operation time sequence diagram is generated, the operation sequence is analyzed, and the instruction conflict characteristics of the operation time sequence diagram are extracted; Based on the beam forming technology, the vibration signal of the pre-acquired mechanical vibration signal of the transformer substation equipment is focused, and the voiceprint characteristics are generated; The operation sequence and the instruction conflict characteristics are coded to obtain the operation time sequence data; The operation time sequence data and the voiceprint characteristics are processed in parallel to generate a verification result.

2. The parallel mischeck prevention method for grid dispatch instruction objectification data according to claim 1, characterized in that, The voiceprint characteristics include spectral distribution characteristics, energy attenuation characteristics, and time duration characteristics.

3. The parallel mischeck prevention method for grid dispatch instruction objectification data according to claim 2, characterized in that, The voiceprint characteristics are generated in the following steps: Based on the spatial distribution relationship of the acoustic sensor array, the sound source position information is obtained by positioning the collected vibration signal, and the vibration waveform data is included; The phase alignment processing is performed on the vibration waveform data to obtain coherent waveform data; The waveform synthesis is performed in combination with the sound source position information and the coherent waveform data to generate an enhanced acoustic wave signal; The frequency domain characteristics and the time domain characteristics of the enhanced acoustic wave signal are extracted, and the feature fusion is performed to obtain the pattern voiceprint characteristics.

4. The parallel mischeck prevention method for grid dispatch instruction objectification data according to claim 3, characterized in that, The operation time sequence data and the voiceprint characteristics are processed in parallel to generate a verification result, which includes: Logical compliance determination is performed on the operation time sequence data to obtain a logical instruction state; Feature determination is performed on the voiceprint information to obtain a physical execution state; Action type matching is performed according to the operation instruction type and the pattern voiceprint characteristics; The logical instruction state and the physical execution state are received and combined to generate a joint state identifier; State decision logic processing is performed on the joint state identifier to generate a verification result.

5. The parallel mischeck prevention method for grid dispatch instruction objectification data according to claim 4, characterized in that, The operation sequence and the instruction conflict characteristics are coded to obtain the operation time sequence data, which includes: Based on the time correlation between the operation instructions, the operation sequence is reorganized to obtain a reorganized operation sequence; The pattern characteristics of the instruction conflict characteristics are extracted and the characteristics are strengthened to obtain strengthened conflict characteristics; The reorganized operation sequence and the strengthened conflict characteristics are fused and jointly coded to generate fused coding data.

6. The parallel mischeck prevention method for grid dispatch instruction objectification data according to claim 5, characterized in that, The waveform synthesis includes: The beam forming weight is established according to the sound source position information to obtain spatial filtering parameters; The spatial filtering parameters are used to dynamically weight and synthesize the coherent waveform data.

7. The parallel mischeck prevention method for grid dispatch instruction objectification data according to claim 6, characterized in that, The action type matching is performed according to the operation instruction type and the pattern voiceprint characteristics, which includes: Based on the action category of the operation instruction, the standard feature set in the voiceprint characteristic template library is called; The spectral distribution characteristics of the to-be-tested signal are matched with the reference spectral mode in the standard feature set to calculate the spectral similarity; The energy attenuation characteristics of the to-be-tested signal are dynamically time-warped with the reference energy envelope curve in the standard feature set to quantify the shape consistency; The time duration characteristics of the to-be-tested signal are compared with the legal operation time window to generate a time sequence compliance factor; The spectral similarity, the shape consistency, and the time sequence compliance factor are fused, and the final action type matching degree is calculated through a weighted decision model, and the matching result is output.

8. The parallel mischeck prevention method of grid dispatch instruction objectified data according to claim 7, characterized in that, The logical instruction state and the physical execution state are received and combined to generate a joint state identifier, which includes: A time synchronization window between the logical instruction state sequence and the physical execution state sequence is established, and the two sequences are time-aligned; The aligned logical instruction state sequence is traversed, and for each instruction state, a time-associated physical execution state is searched in the synchronization window to form an associated set; Based on the matching result of the operation instruction type and the action type, the matching confidence of the logical and physical state combination in the associated set is calculated, and a state matching degree matrix is generated; According to the state matching degree matrix, the logical instruction state sequence and the physical execution state sequence are combined and mapped, and a unique joint state identifier is generated for the successfully mapped state pairs.

9. The parallel mischeck prevention method for grid dispatch instruction objectification data according to claim 8, wherein, Based on the spatial distribution relationship of the acoustic sensor array, the sound source position information is obtained by sound source positioning of the collected vibration signal, including: Based on the geometric distribution relationship of the acoustic sensor array, the arrival time difference of the vibration waveform is calculated to obtain the sound source direction data; Combined with the sound wave propagation characteristics, the sound source direction data is mapped in three-dimensional space to generate the sound source position information.

10. A system for parallel misoperation prevention check of grid dispatch instruction objectification data using the method according to any one of claims 1 to 9, characterized in that, It includes: The time sequence analysis module is used for generating operation time sequence diagram according to the scheduling log data, analyzing operation sequence, and extracting instruction conflict features of operation time sequence diagram; The voiceprint generation module is used for focusing processing of the vibration signal of the pre-acquired mechanical vibration signal of the substation equipment based on the beamforming technology to generate voiceprint features; The time sequence coding module is used for coding operation sequence and instruction conflict features to obtain operation time sequence data; The parallel verification module is used for parallel verification processing of operation time sequence data and voiceprint features to generate a verification result.

Images