Data information secure transmission method and system

Abstract

The invention relates to the technical field of digital information secure transmission, in particular to a data information secure transmission method and system. The method comprises the following steps of: performing differential quantization and coding on original measurement data of a power grid to obtain a parameter feature code vector; constructing a power grid disturbance response matrix according to the parameter feature code vector to obtain a disturbance response matrix; acquiring to-be-transmitted data; performing information channel modulation on the to-be-transmitted data according to the disturbance response matrix to obtain a modulation parameter sequence; performing time interleaving analysis according to the modulation parameter sequence to obtain a time interleaving data stream; obtaining power grid topology information; performing space-time fusion coding transmission according to the power grid topology information and the time-interleaved data stream to obtain a space-time interleaved code stream; and performing spatio-temporal de-interleaving processing on the spatio-temporal interleaving code stream to obtain a channel data block set. According to the invention, the security, reliability, concealment and high adaptability of data information transmission are improved through a digital information security transmission technology.

Description

Technical Field

[0001] The present invention relates to the technical field of digital information security transmission, and particularly to a data information security transmission method and system. Background Art

[0002] Traditional power data security transmission methods rely too much on the key mechanism, resulting in a series of security risks. In large power grid systems, a large number of keys need to be managed, leading to complex management and easy errors; the key distribution process faces the risk of man-in-the-middle attacks; the dependence on specific encryption algorithms forms a technology lock. Once the algorithm is cracked, the overall security collapses; the key synchronization problem frequently occurs in distributed systems, causing communication interruptions. This strong dependence on artificially set keys makes the security foundation of the system fragile and unsustainable.

[0003] Existing methods generally adopt a single security mechanism. Once this mechanism is cracked, the entire system defense line will collapse; the centralized key management server becomes the preferred target of attackers; at the same time, these methods lack an understanding of the physical characteristics of the power grid and cannot distinguish normal disturbances from malicious attacks, resulting in a large number of false alarms triggered by normal power grid phenomena such as electromagnetic interference and harmonic pollution; fixed security policies cannot adapt to the changes in the operation state of the power grid, and use the same security level for all data, wasting resources and unable to guarantee the security of key data; in extreme conditions, the security mechanism may instead block normal communication and exacerbate the impact of accidents.

[0004] In summary, there are problems of strong key dependence, single-point vulnerability, and disturbance sensitivity in the existing technologies that need to be solved urgently. Summary of the Invention

[0005] Based on this, it is necessary to provide a data information security transmission method and system to solve at least one of the above technical problems.

[0006] To achieve the above object, a data information security transmission method includes the following steps:

[0007] Step S1: Collect the original measurement data of the power grid; perform differential quantization and coding on the original measurement data of the power grid to obtain a parameter feature code vector;

[0008] Step S2: Construct a power grid disturbance response matrix according to the parameter feature code vector to obtain a disturbance response matrix; obtain the data to be transmitted; perform information channel modulation on the data to be transmitted according to the disturbance response matrix to obtain a modulation parameter sequence;

[0009] Step S3: Perform time interleaving analysis according to the modulation parameter sequence to obtain a time interleaved data stream; obtain the power grid topology information; perform space-time fusion coding transmission according to the power grid topology information and the time interleaved data stream to obtain a space-time interleaved code stream;

[0010] Step S4: Perform spatio-temporal deinterleaving processing on the spatio-temporal interleaved code stream to obtain a set of channel data blocks; perform multi-point and multi-dimensional electrical parameter verification on the set of channel data blocks to obtain a data integrity report; perform adaptive data reorganization and recovery on the set of channel data blocks according to the data integrity report to obtain a check result flag, so as to achieve the task of secure transmission of power grid data information.

[0011] The present invention ensures the accuracy and reliability of the original data by deploying high-precision synchronous PMUs and following strict standards to collect various power grid parameters at a high sampling rate and high time accuracy. The mutual information is used to screen the differential parameter set, reducing the data redundancy. Differentiated sampling and quantization are performed according to different parameter characteristics, taking into account both the information integrity and the data compression ratio. The multi-mode coding transformation (Gray code, Hamming code / CRC code) is applied to enhance the anti-interference and error detection and correction capabilities of the data. The finally generated parameter feature code vector is a compact, robust and highly secure representation of the power grid state, laying a solid foundation for the subsequent steps. By collecting the historical data of typical disturbance events and extracting the parameter feature code vectors, a three-dimensional disturbance response matrix framework is constructed. Multi-scale feature extraction and disturbance-parameter pair screening are adopted to quantify the sensitivity, response delay and duration of different parameters to different disturbances, generating the disturbance response matrix. The data to be transmitted is classified and segmented, and the parameter channel matching analysis is performed according to the disturbance response matrix, realizing the optimal configuration of data and channels. The micro-disturbance parameter modulation method is used to covertly embed the data information into the tiny changes of the power grid parameters, getting rid of the dependence on traditional keys and improving the security and concealment of the system. Through the channel correlation analysis and channel characteristic evaluation, a basis is provided for the subsequent optimization of the transmission strategy. The depth-variable time interleaving and pseudo-random interleaving algorithms are adopted to disrupt the order of the data in time. Combining the power grid topology information for spatial distribution strategy optimization, the data is dispersed in space. A spatio-temporal matrix is constructed and the key shards are embedded. Using the bidirectional expansion algorithm and synchronous markers, the deep fusion of the time and space dimensions is realized. According to the channel characteristic evaluation table, dynamic channel selection, redundant transmission, integrity check and encryption are performed to generate a spatio-temporal interleaved code stream with high security and anti-attack ability. The receiving end reliably receives the spatio-temporal interleaved code stream through multi-channel parallel reception and high-precision synchronization. The key shards are extracted and spatio-temporal de-interleaving is performed to reconstruct the modulation parameter sequence. The updated disturbance response matrix is obtained and channel demodulation is performed to extract the data information. The receiving end independently measures the local parameters to generate the local parameter feature code vector, providing a benchmark for multi-level comparison and verification. Through multi-level comparison such as syntax layer check, value range rationality check, time coherence check, power flow constraint check, state estimation verification and multi-point data cross-verification, the integrity and credibility of the data are comprehensively evaluated. An integrity analysis report is generated according to the comprehensive credibility evaluation, and adaptive data recombination and recovery are performed according to the report. Using the hierarchical error correction strategy, the final check result flag is obtained, ensuring the reliability and availability of the data. Therefore, the present invention provides a data information secure transmission method, which solves the two major drawbacks of strong key dependence and single-point vulnerability and disturbance sensitivity existing in the existing technologies by constructing a multi-factor security system independent of traditional keys and introducing a refined disturbance response mechanism.On the one hand, through multi-dimensional parameter coding, information channel modulation, and multi-channel fusion transmission, data is covertly dispersed in the minute changes and spatio-temporal dimensions of grid parameters, getting rid of the dependence on a single key and eliminating the risk of single-point attacks. On the other hand, by constructing a disturbance response matrix (DRM) to achieve adaptive channel selection and micro-disturbance modulation, and adopting multi-point multi-dimensional electrical parameter verification at the receiving end, normal disturbances and malicious attacks can be effectively distinguished, avoiding misjudgment and improving the system's robustness to grid disturbances. The security, reliability, concealment, and high self-adaptability of data information transmission are improved.

[0012] Preferably, the differential quantization and coding in step S1 include:

[0013] Select multiple grid parameters with significant differences for acquisition to obtain the original grid measurement data; perform parameter difference screening on the original grid measurement data to obtain a set of difference parameters;

[0014] Perform differential sampling processing on the set of difference parameters to obtain time-domain feature data;

[0015] Perform differential quantization processing based on the time-domain feature data to obtain a set of quantization features;

[0016] Perform multi-mode coding transformation on the set of quantization features to obtain a multi-mode coding table;

[0017] Integrate the feature vectors of the set of quantization features according to the multi-mode coding table to obtain a parameter feature code vector.

[0018] The present invention collects various power grid parameters (voltage, current, sequence components, frequency and its rate of change) with a high sampling rate (4.8 kHz) and high time synchronization accuracy (better than 1 microsecond) by deploying high-precision synchronized phasor measurement units (PMUs) and strictly following the IEC 61850-9-2 standard, ensuring the accuracy, integrity and timeliness of the original data. The direct optical fiber transmission method avoids the contamination and tampering of data during the transmission process, providing a reliable data basis for subsequent processing. By calculating the mutual information between parameters and setting a threshold (0.1) for screening, the redundancy between parameters is effectively reduced, highlighting the differences between parameters. Parameters that are more sensitive to disturbances (such as the rate of change of frequency) are selected and retained in the set of differential parameters, improving the efficiency and accuracy of subsequent disturbance analysis. The finally formed set of differential parameters (positive sequence voltage, negative sequence voltage, positive sequence current, negative sequence current, rate of change of frequency, active power) contains both the key information of the steady-state operation of the power grid and has a high sensitivity to power grid disturbances. For the dynamic characteristics of different parameters, different sampling windows and sampling frequencies are adopted (for example, shorter windows and higher sampling frequencies are used for negative sequence components), which not only fully capture the signal characteristics of fast changes but also avoid oversampling of slowly changing signals, reducing the data volume. This differential sampling strategy improves the data processing efficiency on the premise of ensuring information integrity. For the characteristics of different parameters, different quantization methods and quantization levels are adopted (for example, vector quantization based on K-means clustering is used for voltage and current, scalar quantization is used for the rate of change of frequency, and non-uniform quantization is used for active power), taking into account both quantization accuracy and data compression ratio. Vector quantization makes full use of the pattern characteristics of voltage and current waveforms, while scalar quantization and non-uniform quantization respectively adapt to the characteristics of the rate of change of frequency and small-range fluctuations as well as the large-range changes of active power. This differential quantization strategy effectively reduces the dimension of the data on the premise of ensuring that key information is not lost. Multiple coding methods (Gray code, Hamming code / CRC code) are applied to transform the quantization feature groups, enhancing the anti-interference ability and error detection and correction ability of the data. The Gray code reduces the error codes caused by quantization errors, the Hamming code can correct one error, and the CRC code can detect multiple errors. The generation of the multi-mode coding table provides multiple protections for subsequent data transmission, improving the reliability of data transmission. The coding values of each parameter in the multi-mode coding table are spliced in a predefined order to form an 118-bit parameter feature code vector. This vector is a compact, robust and highly secure representation of the current power grid state. It not only contains the information of multiple key power grid parameters but also is protected by multiple coding methods, and can effectively resist noise, interference and potential attacks during transmission. This feature code vector lays a solid foundation for subsequent disturbance analysis, information modulation and secure transmission.

[0019] Preferably, the generation of the disturbance response matrix in step S2 includes:

[0020] Collect the historical data of typical disturbance events of the power grid; perform disturbance characteristic analysis based on the parameter characteristic code vectors and the historical data of typical disturbance events to obtain a disturbance characteristic data set, where the disturbance characteristic data set includes load switching disturbance, power supply fluctuation disturbance, line fault disturbance, and harmonic injection disturbance;

[0021] Initialize the disturbance framework according to the disturbance characteristic data set to obtain an initial disturbance response matrix framework;

[0022] Extract multi-scale response characteristics from the disturbance characteristic data set to obtain a multi-scale response characteristic set;

[0023] Screen the disturbance-parameter pairs for the disturbance characteristic data set according to the multi-scale response characteristic set to obtain an effective disturbance-parameter pair set;

[0024] Calculate the sensitivity coefficient for the disturbance characteristic data set according to the effective disturbance-parameter pair set to obtain a sensitive disturbance response matrix;

[0025] Calculate the response delay for the disturbance characteristic data set according to the effective disturbance-parameter pair set to obtain a delay disturbance response matrix;

[0026] Calculate the duration for the disturbance characteristic data set according to the effective disturbance-parameter pair set to obtain a continuous disturbance response matrix;

[0027] Normalize the sensitive disturbance response matrix, the delay disturbance response matrix, and the continuous disturbance response matrix to obtain a disturbance response matrix.

[0028] The present invention ensures the comprehensiveness and representativeness of the disturbance feature dataset by collecting historical data up to 5 years from SCADA and WAMS systems and covering four types of typical disturbance events (load switching, power fluctuation, line fault, harmonic injection). The parameter feature code vectors of each event are extracted as disturbance features, and the power grid disturbances are associated with the feature code vectors generated in step S1, providing a unified data basis for subsequent disturbance analysis and information modulation. A three-dimensional disturbance response matrix framework [disturbance type, power grid parameter, time] is constructed, providing a structured storage and processing space for subsequent disturbance characteristic analysis. This framework clearly defines three dimensions: disturbance type, power grid parameter, and time, and refines the time dimension to 2000 time steps (5 ms / step), capable of capturing the fast dynamic process of the disturbance. The initial value is set to 0, preparing for filling in specific disturbance response characteristics subsequently. Beneficial effects: The Daubechies 4 wavelet basis is used for 5-layer discrete wavelet transform (DWT) to decompose the original signal of each parameter into multiple scales with different frequency ranges, and multiple features (maximum value, minimum value, average value, standard deviation, peak-to-peak value, kurtosis, skewness) of each scale are extracted. This multi-scale feature extraction method can comprehensively and meticulously characterize the features of the disturbance at different time scales, improving the recognition ability of the disturbance. By calculating the variance of the multi-scale response features of each parameter under each disturbance and setting a threshold for screening, the parameters sensitive to specific disturbances are effectively identified. This screening method reduces redundant information, reduces the complexity of subsequent calculations, and highlights the key disturbance-parameter associations, laying a foundation for constructing a concise and efficient disturbance response matrix. By calculating the sensitivity coefficient (parameter change amount / disturbance amount) of each effective disturbance-parameter pair, the sensitivity of different parameters to different disturbances is quantified. The sensitivity coefficient is filled into the corresponding position of the disturbance response matrix framework to form a sensitive disturbance response matrix, providing a key basis for subsequent information modulation. The higher the sensitivity coefficient, the more suitable the parameter is for carrying the information corresponding to the disturbance. By calculating the response delay (the moment when the parameter significantly changes - the moment when the disturbance occurs) of each effective disturbance-parameter pair, the response speed of different parameters to different disturbances is quantified. The response delay is filled into the corresponding position of the disturbance response matrix framework to form a delay disturbance response matrix, providing an important reference for subsequent information modulation and demodulation. The shorter the response delay, the more suitable the parameter is for carrying information with high requirements for delay. By calculating the duration (the moment when the parameter returns to normal - the moment of significant change) of each effective disturbance-parameter pair, the duration of different parameters affected by different disturbances is quantified. The duration is filled into the corresponding position of the disturbance response matrix framework to form a duration disturbance response matrix. This matrix provides a basis for evaluating the influence range and duration of the disturbance, helping to optimize the information modulation strategy and avoid transmitting key information during the duration of the disturbance.The three perturbation response matrices are respectively normalized (the sensitivity matrix adopts the maximum-minimum normalization, and the time delay and duration matrices adopt the method of dividing by the maximum value), eliminating the influence of different dimensions and making the perturbation response characteristics of different types comparable. The three normalized matrices are added to obtain the final perturbation response matrix. This matrix synthesizes various information such as the sensitivity, response time delay, and duration of the perturbation, providing a comprehensive and accurate basis for information channel modulation.

[0029] Preferably, the information channel modulation described in step S2 includes:

[0030] Obtain the data to be transmitted; perform data grading and block processing on the data to be transmitted according to the perturbation response matrix to obtain a graded data block group;

[0031] Perform parameter channel matching analysis according to the perturbation response matrix to obtain the channel matching degree; generate a channel preference matching strategy for the graded data block group according to the channel matching degree to obtain a channel allocation strategy table;

[0032] Perform micro-perturbation parameter modulation on the graded data block group according to the channel allocation strategy table to obtain a modulation parameter sequence.

[0033] After obtaining the binary data to be transmitted (such as a 1024-bit protection trip command), the present invention classifies (Level 1: direct trip command, Level 2: trip permission command, Level 3: status information) and blocks (Level 1 is not blocked, Level 2 and Level 3 are respectively divided into 4 blocks and 10 blocks) according to the importance of the data. This classification and blocking processing method enables important data to be protected at a higher level and provides a basis for subsequent differential modulation. Using the perturbation response matrix generated in step S2, the matching degree of each data block with different grid parameters is evaluated. According to indicators such as response delay, sensitivity, and duration, the most suitable parameter channel is selected for data blocks with different priorities (such as Level 1 selects the negative sequence current I2 channel with the fastest response, Level 2 selects the positive sequence voltage U1 channel). This matching analysis fully considers the importance, timeliness requirements of the data, and the characteristics of the parameter channels, realizing the optimal configuration of data and channels. The generated channel allocation strategy table details the parameter channels corresponding to each data block, providing clear guidance for subsequent modulation. According to the channel allocation strategy table, different modulation methods are used to perform micro-perturbation parameter modulation on each data block (such as Level 1 uses BFSK modulation for I2, Level 2 uses BPSK modulation for U1, Level 3 uses BFSK, BPSK, or PWM according to the mapping parameters). This micro-perturbation modulation method hides the data information in the small changes of the grid parameters, and these changes are controlled within the range allowed by the normal operation of the grid, without triggering protection actions or alarms, realizing the covert transmission of information. The generated modulation parameter sequence contains the information of all data blocks, providing input for subsequent space-time interleaving processing. This modulation method gets rid of the dependence on traditional keys and improves the security of the system.

[0034] Preferably, step S3 includes the following steps:

[0035] Step S31: Perform channel correlation analysis on the modulation parameter sequence to obtain a channel characteristic evaluation table;

[0036] Step S32: Perform time interleaving analysis according to the modulation parameter sequence and the channel characteristic evaluation table to obtain a time-interleaved data stream;

[0037] Step S33: Obtain the grid topology information; optimize the spatial distribution strategy according to the time-interleaved data stream and the grid topology information to obtain a spatial distribution scheme;

[0038] Step S34: Perform space-time fusion coding processing on the time-interleaved data stream according to the spatial distribution scheme to obtain a space-time fusion code;

[0039] Step S35: Perform dynamic channel transmission on the spatio-temporal fusion coding according to the channel characteristic evaluation table to obtain a spatio-temporal interleaved bitstream.

[0040] The present invention quantifies the correlation between parameter channels by calculating the Pearson correlation coefficient between different parameter groups in the modulation parameter sequence. At the same time, the transmission characteristics (delay, bandwidth, bit error rate, anti-interference, concealment) of each parameter channel are evaluated, and these indicators are combined to generate a channel characteristic evaluation table. The table provides an important reference basis for subsequent time interleaving, spatial distribution and dynamic channel selection, which is helpful to optimize the transmission strategy and improve transmission efficiency and security. According to the length of the modulation parameter sequence and the security requirements of each channel in the channel characteristic evaluation table, a time interleaving matrix with variable depth is designed. A pseudo-random interleaving algorithm is used to interleave different time segments with different depths (such as an I2 channel with high security requirements using an interleaving depth of 50). This time interleaving process disrupts the order of data in time, reduces the risk of data being intercepted and deciphered, and improves the security of data transmission. The generated time interleaving data stream and interleaving mapping table provide a basis for subsequent spatial distribution and deinterleaving at the receiving end. The power grid topology information is obtained from the power grid geographic information system (GIS) to construct a node connection diagram. The time interleaving data stream is divided into spatial slices, and the slice size is allocated according to the communication capability, storage capacity and reliability of the node. A multi-path routing algorithm is used to select multiple paths for important data slices, and data slices with high correlation are transmitted through long-distance nodes. This spatial distribution strategy optimizes the spatial distribution of data, reduces the risk of data being intercepted in a centralized manner, and improves the anti-destruction capability of data transmission. The generated spatial distribution scheme provides guidance for subsequent spatiotemporal fusion coding and dynamic channel transmission. A two-dimensional spatiotemporal matrix (rows: time slices, columns: spatial nodes) is constructed, and the data slices in the time-interleaved data stream are filled into the matrix, and key slices (one for each node) are embedded. A bidirectional expansion algorithm (cyclic shift in the time dimension, XOR operation in the space dimension) is used, and a synchronization mark (Barker code) is added to the header of the data block. The spatiotemporal mixing degree (the average value of row and column entropy) is calculated and ensured to meet the standard. This spatiotemporal fusion coding process fully mixes and disperses the data in both time and space dimensions, further improving the security of the data, making it difficult for attackers to recover the complete information even if they obtain part of the data. The transmission channel is dynamically selected based on the channel characteristic evaluation table and the current operating status of the power grid. Multi-channel redundant transmission is selected for important data slices, and channels with low latency and high reliability are given priority. Low-latency channels are selected for time-sensitive data, and high-bandwidth channels are selected for large-capacity data. Routing information and checksums are added to data packets for integrity verification and encryption (AES-256). This dynamic channel transmission strategy fully utilizes the communication resources of the power grid and improves transmission efficiency and reliability. The resulting time-space interleaved code stream is fully interleaved, encoded, and protected in both time and space, and has high security and anti-attack capabilities.

[0041] Preferably, the spatiotemporal deinterleaving process in step S4 includes:

[0042] The receiving end simultaneously monitors multiple physical communication channels, synchronously receives the spatio-temporal interleaved code stream, and obtains a set of synchronized data frames;

[0043] Perform spatio-temporal deinterleaving reconstruction on the set of synchronized data frames to obtain a reconstructed modulation sequence;

[0044] Obtain an updated perturbation response matrix; perform channel demodulation extraction on the reconstructed modulation sequence according to the updated perturbation response matrix to obtain a set of channel data blocks.

[0045] In the present invention, by deploying multiple receivers at the receiving end to simultaneously monitor multiple physical communication channels, it is possible to receive spatio-temporal interleaved code stream segments from different spatial nodes in parallel. By detecting synchronization markers (Barker codes) and using phase-locked loop (PLL) technology to achieve high-precision time synchronization (with precision controlled within 1 microsecond), it ensures the temporal alignment of the received data frames. This multi-channel parallel reception and high-precision synchronization mechanism improve the reliability and efficiency of data reception, laying a foundation for subsequent deinterleaving processing. Extract and splice key fragments to restore the complete interleaved key. Then, perform spatial dimension deinterleaving (XOR operation) and temporal dimension deinterleaving (reverse cyclic shift) in the reverse order of the sending end. Extract data blocks according to the spatio-temporal matrix filling rule and splice them in the order of the sending end. This spatio-temporal deinterleaving operation can accurately restore the modulated parameter sequence at the sending end. The success of deinterleaving is a prerequisite for subsequent channel demodulation and data recovery. The receiving end regularly obtains the latest power grid operation status information and recalculates the perturbation response matrix based on this to obtain an updated perturbation response matrix. This update mechanism ensures that the perturbation response matrix can reflect the latest state of the power grid and improves the accuracy of demodulation. According to the updated perturbation response matrix, group the reconstructed modulation sequence by parameter type and perform demodulation according to their respective modulation methods (such as BFSK demodulation for I2, BPSK demodulation for U1, PWM demodulation for P). This channel demodulation extraction can accurately restore the data information embedded at the sending end from the small changes in power grid parameters to obtain a set of channel data blocks. The success of demodulation is the ultimate goal of the entire data transmission process.

[0046] Preferably, the multi-point multi-dimensional electrical parameter verification in step S4 includes:

[0047] The receiving end measures the local power grid parameters to obtain a local parameter signature vector;

[0048] Perform syntax layer verification on the set of channel data blocks according to the local parameter signature vector to obtain a primary comparison result;

[0049] Check the rationality of the value range of the primary verification result and perform a time coherence check to obtain a secondary comparison result;

[0050] Perform power flow constraint check on the secondary comparison result and conduct state estimation verification to obtain the physical constraint status data;

[0051] Perform multi-point data cross-verification based on the physical constraint status data to obtain the tertiary comparison result;

[0052] Conduct comprehensive credibility evaluation based on the primary comparison result, secondary comparison result, and tertiary comparison result to obtain the data block credibility set;

[0053] Generate an integrity analysis report based on the data block credibility level to obtain the data integrity report.

[0054] In the present invention, the receiving end is equipped with a PMU of the same model and configuration as the sending end, which independently measures the local power grid parameters and performs the same "differential quantization and coding" process as the sending end to generate a local parameter signature vector. This vector represents the independent measurement result of the local power grid state at the receiving end, providing a benchmark for subsequent multi-level comparison and verification, and is the key to realizing data integrity verification. Perform syntax-level verification on the channel data block set, including length check, coding format check (Hamming code check, CRC check), and comparison of parameter values with local measurement values. This verification can detect whether errors or tampering occur during data transmission, such as data bit loss, coding errors, or abnormal parameter values. The result of the first-level comparison provides a basis for the preliminary screening of data. Perform range rationality check (whether the parameter value is within the reasonable range of power grid operation) and time coherence check (whether the change of parameter value conforms to the law of power grid operation) on the data blocks passing the first-level verification. This check can detect whether the data conforms to the basic physical laws of power grid operation, such as whether the voltage is too high or too low, whether the frequency exceeds the allowable range, and whether the parameter change is too drastic. The result of the second-level comparison further improves the accuracy of data verification. Perform power flow constraint check (whether the calculation result meets the physical constraints of power grid operation) and state estimation verification (compare with the state estimation result measured by the local PMU at the receiving end) on the data blocks passing the second-level verification. This check can verify whether the data conforms to the overall operation state of the power grid, such as whether the node voltage is qualified, whether the line power is over-limit, and whether the generator output is reasonable. The physical constraint state data provides a higher-level basis for the further verification of data. Utilize the power grid topology information to select the receiving ends of multiple adjacent substations, compare their respective physical constraint state data, and calculate the correlation coefficient of the parameters. This multi-point data cross-verification utilizes the spatial correlation of the power grid and can detect abnormalities that are difficult to detect by a single receiving end, such as malicious tampering of data during transmission. The result of the third-level comparison further improves the reliability and accuracy of data verification. Adopt a weighted scoring method to comprehensively consider the results of the first-level, second-level, and third-level comparisons, evaluate the credibility of each data block, and obtain a credibility score between 0 and 1. This comprehensive evaluation method fully considers the verification results of data at different levels and provides a quantitative basis for the final determination of data. The data block credibility set provides guidance for subsequent data reorganization and recovery. Generate an integrity analysis report according to the data block credibility set, which details the credibility score of each data block, the comparison results at each level, the analysis of abnormal reasons (if any), and the overall credibility evaluation. For untrusted data blocks, the report will put forward processing suggestions (such as retransmission, using redundant data, etc.). This report provides comprehensive and detailed information about the quality of the received data for the receiving end, helps the receiving end make correct decisions, and ensures the safe and reliable operation of the power system.

[0055] Preferably, the present invention further provides a data information secure transmission system for executing the data information secure transmission method as described above. The data information secure transmission system includes:

[0056] A multi-dimensional parameter encoding module, configured to collect original power grid measurement data; perform differential quantization and encoding on the original power grid measurement data to obtain a parameter feature code vector;

[0057] An information channel modulation module, configured to construct a power grid disturbance response matrix according to the parameter feature code vector to obtain a disturbance response matrix; acquire data to be transmitted; perform information channel modulation on the data to be transmitted according to the disturbance response matrix to obtain a modulation parameter sequence;

[0058] A multi-channel fusion transmission module, configured to perform time interleaving analysis according to the modulation parameter sequence to obtain a time interleaved data stream; acquire power grid topology information; perform spatio-temporal fusion encoding transmission according to the power grid topology information and the time interleaved data stream to obtain a spatio-temporal interleaved code stream;

[0059] A receiving end demodulation and decoding module, configured to perform spatio-temporal de-interleaving processing on the spatio-temporal interleaved code stream to obtain a channel data block set; perform multi-point multi-dimensional electrical parameter verification on the channel data block set to obtain a data integrity report; perform adaptive data recombination and recovery on the channel data block set according to the data integrity report to obtain a verification result flag, so as to implement the power grid data information secure transmission task. Description of the Drawings

[0060] Figure 1 It is a schematic diagram of the step flow of a data information secure transmission method;

[0061] Figure 2 It is a schematic diagram of the detailed implementation step flow of step S3 in the present invention.

[0062] The implementation, functional features and advantages of the present invention will be further described in conjunction with the embodiments with reference to the drawings. Detailed Embodiments

[0063] The technical method of the present invention will be clearly and completely described below with reference to the drawings. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0064] In addition, the accompanying drawings are only schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and thus repeated descriptions thereof will be omitted. Some of the block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities may be implemented in software form, or in one or more hardware modules or integrated circuits, or in different networks and / or processor methods and / or microcontroller methods.

[0065] It should be understood that although terms such as "first" and "second" may be used herein to describe various units, these units should not be limited by these terms. These terms are only used to distinguish one unit from another. For example, without departing from the scope of the exemplary embodiments, the first unit may be referred to as the second unit, and similarly the second unit may be referred to as the first unit. The term "and / or" used herein includes any and all combinations of one or more of the listed associated items.

[0066] In an embodiment of the present invention, reference is made to Figure 1 As shown, it is a schematic flow chart of the steps of the data information secure transmission method of the present invention. In this example, the data information secure transmission method includes the following steps:

[0067] Step S1: Collect the original measurement data of the power grid; perform differential quantization and coding on the original measurement data of the power grid to obtain a parameter feature code vector;

[0068] In an embodiment of the present invention, a high-precision synchronous phasor measurement unit (PMU) is deployed on the 110 kV substation bus, configured strictly in accordance with the IEC 61850-9-2 standard, and parameters such as three-phase voltage, current, positive and negative zero-sequence voltage and current, frequency, and rate of change of frequency are collected at a sampling rate of 4.8 kHz, and with a GPS synchronous timestamp (accuracy better than 1 microsecond). The collected original data is directly transmitted to the data processing center through optical fibers. After preprocessing the original data (removing bad data and imputing missing values), calculate the mutual information between parameters, and screen out the set of differential parameters with mutual information lower than 0.1 (positive-sequence voltage, negative-sequence voltage, positive-sequence current, negative-sequence current, rate of change of frequency, active power). For different parameter characteristics, adopt different sampling windows and frequencies (for example, the negative-sequence component uses a 100 ms window and samples once every 2 ms). Adopt a differential quantization method for the time-domain feature data (for example, voltage and current use vector quantization based on K-means clustering, the rate of change of frequency uses scalar quantization, and active power uses non-uniform quantization). Perform multi-mode coding transformation (Gray code, Hamming code / CRC code) on the quantization feature group, and splice the encoded values of each parameter in a predefined order to form a 118-bit parameter feature code vector.

[0069] Step S2: Construct a power grid disturbance response matrix based on the parameter feature code vectors to obtain a disturbance response matrix; acquire the data to be transmitted; perform information channel modulation on the data to be transmitted according to the disturbance response matrix to obtain a modulation parameter sequence;

[0070] In the embodiment of the present invention, historical data for the past 5 years is collected from SCADA and WAMS, including four types of disturbance events: load switching, power fluctuation, line fault, and harmonic injection. The parameter feature code vectors of each event are extracted as disturbance features. Initialize a three-dimensional disturbance response matrix framework [disturbance type, power grid parameter, time (2000 time steps, 5 ms / step)]. Perform 5-layer Daubechies 4 wavelet decomposition on the disturbance features, extract 7 features (maximum value, minimum value, etc.) of each coefficient to form a multi-scale response feature set. Calculate the variance of the multi-scale response features of each parameter under each type of disturbance, and screen out the effective disturbance-parameter pairs. For the effective disturbance-parameter pairs, calculate the sensitivity coefficient (parameter change amount / disturbance amount), response delay (the moment when the parameter significantly changes - the moment when the disturbance occurs), and duration (the moment when the parameter returns to normal - the moment when the significant change occurs), and fill them into the disturbance response matrix framework respectively, and perform normalization processing to obtain the disturbance response matrix. Acquire a 1024-bit binary sequence to be transmitted (such as a protection trip instruction), and divide it into blocks according to importance levels (Level 1: 128 bits, Level 2: 4×64 bits, Level 3: 10×64 bits). According to the disturbance response matrix, evaluate the matching degree between the data block and the parameter channel (such as Level 1 matches negative sequence current I2), and generate a channel allocation strategy table. According to the strategy table, perform micro-disturbance parameter modulation on the data block (such as using BFSK to modulate I2 for Level 1, and using BPSK to modulate U1 for Level 2) to obtain a modulation parameter sequence.

[0071] Step S3: Perform time interleaving analysis according to the modulation parameter sequence to obtain a time-interleaved data stream; acquire the power grid topology information; perform spatio-temporal fusion coding transmission according to the power grid topology information and the time-interleaved data stream to obtain a spatio-temporal interleaved code stream;

[0072] In the embodiments of the present invention, the modulation parameter sequence is grouped according to parameter types, the Pearson correlation coefficient between parameter groups is calculated, and the transmission characteristics (delay, bandwidth, bit error rate, anti-interference ability, concealment) of each channel are evaluated to generate a channel characteristic evaluation table. The modulation parameter sequence is divided into time segments, the interleaving depth is dynamically adjusted according to the channel security requirements (for example, the depth of the I2 channel is 50, and the depth of the P channel is 10), and the pseudo-random interleaving algorithm is used to shuffle the order of elements within the segments to generate a time-interleaved data stream and an interleaving mapping table. The power grid topology information is obtained from GIS to construct a node connection graph. The time-interleaved data stream is segmented into spatial slices, and the slice size is determined according to the node communication capabilities and storage capacities. The multi-path routing algorithm is used to select multiple paths for transmitting important data slices, select long-distance nodes for data slices with high correlation, and preferentially select nodes with high reliability to generate a spatial distribution scheme. A two-dimensional spatio-temporal matrix (rows: time segments, columns: spatial nodes) is constructed, the data slices are filled into the matrix, and key slices (one for each node) are embedded. Bidirectional expansion (circular shift in the time dimension and exclusive OR operation in the spatial dimension) is performed, a synchronization marker (Barker code) is added, and the spatio-temporal mixing degree (average value of row and column entropy) is calculated to ensure that the mixing degree meets the standard. According to the channel characteristic evaluation table and the current state of the power grid, the transmission channel is dynamically selected, routing information and a check code are added to the data packet, integrity verification and encryption (AES-256) are performed, and the spatio-temporal interleaved code stream is sent.

[0073] Step S4: Perform spatio-temporal de-interleaving processing on the spatio-temporal interleaved code stream to obtain a set of channel data blocks; perform multi-point multi-dimensional electrical parameter verification on the set of channel data blocks to obtain a data integrity report; perform adaptive data recombination and recovery on the set of channel data blocks according to the data integrity report to obtain a check result flag;

[0074] In the embodiments of the present invention, multiple receivers at the receiving end simultaneously monitor multiple channels, detect synchronization markers (Barker codes), and synchronize data frames using PLL technology. Extract key fragments and restore the complete interleaved key. Perform deinterleaving in the spatial dimension (XOR operation) and deinterleaving in the time dimension (reverse cyclic shift), extract data blocks according to the space-time matrix filling rules, and reconstruct the modulation sequence. The receiving end regularly obtains the latest power grid state information and recalculates the disturbance response matrix. According to the updated disturbance response matrix, group the reconstructed modulation sequence by parameter type and perform demodulation (such as BFSK demodulation for I2, BPSK demodulation for U1, PWM demodulation for P) to obtain a set of channel data blocks. The receiving end independently measures local power grid parameters to generate a local parameter signature vector. Perform syntax layer verification (comparing length, encoding, parameter values with local measurement values), value range rationality check, and time coherence check on the set of channel data blocks. Perform power flow constraint check and state estimation verification to obtain physically constrained state data. Utilize power grid topology information to perform multi-point data cross-verification (construct a node adjacency matrix, calculate electrical distances, screen associated node pairs, align time series of similar parameters, calculate weighted correlation coefficients, set dynamic deviation thresholds, and mark abnormal data). Integrate the results of multi-level comparison, perform weighted scoring to obtain a set of data block credibility. Generate an integrity analysis report, mark untrusted data blocks, and recommend handling measures. According to the data integrity report, classify and process the data blocks (directly incorporate completely trusted ones, apply error correction algorithms to partially trusted ones, attempt to recover or retransmit untrusted ones), adopt a hierarchical error correction strategy (use Reed-Solomon codes for critical instructions, convolutional codes for general data, and utilize the power grid physical model to assist in error correction), reorganize the data blocks, perform a final verification, and generate a verification result flag.

[0075] Preferably, the differential quantization and coding in step S1 include:

[0076] Select multiple significantly different power grid parameters for acquisition to obtain original power grid measurement data; perform parameter difference screening on the original power grid measurement data to obtain a set of difference parameters;

[0077] Perform difference sampling processing on the set of difference parameters to obtain time-domain feature data;

[0078] Perform differential quantization processing according to the time-domain feature data to obtain a set of quantization features;

[0079] Perform multi-mode coding transformation on the set of quantization features to obtain a multi-mode coding table;

[0080] Integrate the quantization features according to the multi-mode coding table to obtain a parameter signature vector.

[0081] In the embodiments of the present invention, high-precision synchronized phasor measurement units (PMUs) are deployed on the busbars of a 110 kV substation. These PMUs are configured strictly in accordance with the IEC 61850-9-2 standard to ensure the time synchronization of data. The following grid parameters are specifically collected: three-phase voltages (Ua, Ub, Uc), three-phase currents (Ia, Ib, Ic), positive-sequence voltage (U1), negative-sequence voltage (U2), zero-sequence voltage (U0), positive-sequence current (I1), negative-sequence current (I2), zero-sequence current (I0), frequency (f), and frequency change rate (df / dt). The sampling rate of the PMU is set to 4800 sampling points per second (i.e., 4.8 kHz) to capture fast transient processes. The data of each sampling point contains a 64-bit timestamp, which is synchronized with the GPS signal, and the synchronization accuracy is better than 1 microsecond. The collected raw data is directly transmitted to the data processing center of the substation through an optical fiber channel without passing through any intermediate links to avoid data contamination. The data is stored in binary format, and each parameter value is represented by a 32-bit floating-point number, ensuring sufficient numerical accuracy. After receiving the raw measurement data uploaded by the PMU, data preprocessing is first performed, including bad data rejection and missing value imputation. Bad data rejection uses a statistics-based method, and data points exceeding 3 times the standard deviation are regarded as outliers and rejected. Missing value imputation uses the cubic spline interpolation method, and interpolation is performed using 5 data points before and after the missing point. Then, the mutual information between every two parameters is calculated. The mutual information calculation formula is: I(X; Y) = ΣΣp(x, y)log(p(x, y) / (p(x)p(y))), where p(x, y) is the joint probability distribution of X and Y, and p(x) and p(y) are the marginal probability distributions of X and Y respectively. The lower the mutual information value, the higher the redundancy and the greater the difference between the two parameters. The mutual information threshold is set to 0.1. Among the parameter pairs with mutual information lower than 0.1, the parameter that is more sensitive to disturbances is retained (for example, the frequency change rate is usually more sensitive than the frequency itself). Finally, the selected set of differential parameters includes: positive-sequence voltage (U1), negative-sequence voltage (U2), positive-sequence current (I1), negative-sequence current (I2), frequency change rate (df / dt), and active power (P). For each parameter in the set of differential parameters, different sampling windows and sampling frequencies are adopted. For the positive-sequence voltage (U1) and positive-sequence current (I1), a sliding window with a length of 200 ms is adopted, and sampling is performed every 5 ms, obtaining 40 sampling points. For the negative-sequence voltage (U2) and negative-sequence current (I2), a sliding window with a length of 100 ms is adopted, and sampling is performed every 2 ms, obtaining 50 sampling points. This is because the negative-sequence component usually changes faster during a fault. For the frequency change rate (df / dt), a sliding window with a length of 50 ms is adopted, and sampling is performed every 1 ms, obtaining 50 sampling points.For the active power (P), a sliding window with a length of 500 ms is adopted, and sampling is carried out every 10 ms to obtain 50 sampling points. This differential sampling strategy fully considers the dynamic characteristics of different parameters, ensuring both the capture of rapidly changing signals and the avoidance of oversampling of slowly changing signals. The sampled data constitutes time-domain feature data. For the time-domain feature data of different parameters, different quantization methods and quantization levels are adopted. For voltage and current parameters, a vector quantization method based on K-means clustering is used. First, a large amount of historical voltage and current data are subjected to K-means clustering to obtain 64 cluster centers, and each cluster center represents a typical voltage or current waveform pattern. Then, the current time-domain feature data is compared with these 64 cluster centers, and the cluster center with the closest distance is selected as the quantization result, which is represented by a 6-bit binary number. For the frequency change rate (df / dt), a scalar quantization method is adopted, and its change range is divided into 32 levels, each level being represented by a 5-bit binary number. For the active power (P), due to its large change range, a non-uniform quantization method is used. The power change range is divided into 128 non-uniform intervals, with the low-power region divided more densely and the high-power region divided more sparsely, and each interval is represented by a 7-bit binary number. This differential quantization strategy takes into account both quantization accuracy and data compression ratio. For the quantization feature group obtained in the previous step, various coding methods are applied for transformation to enhance the anti-interference ability of the data. First, all quantization values are encoded with Gray code. The characteristic of Gray code is that only one bit is different between adjacent codewords, which can reduce bit errors caused by quantization errors. Then, for the quantization values of voltage and current, Hamming code encoding is additionally applied. Hamming code is an error-correcting code that can detect and correct one-bit errors. For the quantization values of frequency change rate and active power, cyclic redundancy check (CRC) encoding is additionally applied. CRC code can detect multiple-bit errors but does not have the ability to correct errors. The original quantization values, Gray code encoded values, and Hamming code / CRC encoded values of each parameter are combined together to form a multi-mode coding table. For example, the coding table of positive-sequence voltage contains: the original 6-bit quantization value, 6-bit Gray code encoded value, and 10-bit Hamming code encoded value. The encoded values of each parameter in the multi-mode coding table are concatenated in a predefined order to form a long binary vector. The concatenation order is: positive-sequence voltage (6 + 6 + 10 bits), negative-sequence voltage (6 + 6 + 10 bits), positive-sequence current (6 + 6 + 10 bits), negative-sequence current (6 + 6 + 10 bits), frequency change rate (5 + 5 + 8 bits), active power (7 + 7 + 16 bits). The finally formed parameter feature code vector has a length of 118 bits. This 118-bit binary vector is a compact and robust representation of the power grid state at the current moment. It contains the key information of multiple parameters and is protected by various coding methods, capable of resisting noise and interference during transmission. This vector will be used as the input for subsequent steps.

[0082] Preferably, the generation of the disturbance response matrix in step S2 includes:

[0083] Collect historical data of typical disturbance events of the power grid; conduct disturbance characteristic analysis based on the parameter characteristic code vector and the historical data of typical disturbance events to obtain a disturbance characteristic data set, where the disturbance characteristic data set includes load switching disturbance, power source fluctuation disturbance, line fault disturbance, and harmonic injection disturbance;

[0084] Initialize the disturbance framework according to the disturbance characteristic data set to obtain an initial disturbance response matrix framework;

[0085] Extract multi-scale response characteristics from the disturbance characteristic data set to obtain a multi-scale response characteristic set;

[0086] Screen disturbance-parameter pairs according to the multi-scale response characteristic set from the disturbance characteristic data set to obtain an effective disturbance-parameter pair set;

[0087] Calculate the sensitivity coefficient according to the effective disturbance-parameter pair set from the disturbance characteristic data set to obtain a sensitive disturbance response matrix;

[0088] Calculate the response delay according to the effective disturbance-parameter pair set from the disturbance characteristic data set to obtain a delay disturbance response matrix;

[0089] Calculate the duration according to the effective disturbance-parameter pair set from the disturbance characteristic data set to obtain a continuous disturbance response matrix;

[0090] Normalize the sensitive disturbance response matrix, the delay disturbance response matrix, and the continuous disturbance response matrix to obtain a disturbance response matrix.

[0091] In the embodiments of the present invention, historical data for the past 5 years is collected from the SCADA system and the Wide Area Measurement System (WAMS) of the power grid dispatching center. The data includes: 1) Load switching disturbance: Record the switching operations of large industrial users (such as steel mills, electrolytic aluminum plants), and the changes in parameters such as voltage, current, and frequency on the substation bus connected to the user within 10 minutes before and after the switching. Each switching operation is recorded as an event. 2) Power fluctuation disturbance: Record the changes in active and reactive power outputs of large generating units (such as thermal power units, hydropower units), and the changes in parameters such as voltage, current, and frequency on the substation bus connected to the unit within 10 minutes before and after the change. Each output change exceeding 5% of the rated capacity is recorded as an event. 3) Line fault disturbance: Record the short-circuit fault events (single-phase grounding, two-phase short-circuit, three-phase short-circuit) of transmission lines, and the changes in parameters such as voltage, current, and frequency on the substation buses at both ends of the fault line within 5 seconds before and after the fault. Each fault is recorded as an event. 4) Harmonic injection disturbance: Record the harmonic injection events caused by non-linear loads (such as frequency converters, electric arc furnaces), and the changes in the total harmonic distortion rate (THD) of voltage and current on the substation bus connected to the load within 10 minutes before and after the injection. Each THD change exceeding 5% is recorded as an event. For each event, the parameter feature code vector generated in step S1 is extracted as the disturbance feature. The disturbance features of all events are combined to form a disturbance feature data set. The initial disturbance response matrix framework is a three-dimensional matrix, with dimensions: disturbance type, grid parameters, and time. The disturbance types include: load switching, power fluctuation, line fault (subdivided into single-phase grounding, two-phase short-circuit, three-phase short-circuit), and harmonic injection. The grid parameters are consistent with the differential parameter set selected in step S1, namely: positive sequence voltage (U1), negative sequence voltage (U2), positive sequence current (I1), negative sequence current (I2), rate of change of frequency (df / dt), active power (P). The time dimension is determined according to the longest duration of the disturbance event, set to 10 seconds, with a time step of 5 ms, for a total of 2000 time steps. Therefore, the dimension of the initial disturbance response matrix framework is [4, 6, 2000] (the line fault is subdivided into 3 categories, for a total of 4 types of disturbances). The initial value of each element in the matrix is set to 0.Perform multi-scale response feature extraction on the disturbance feature dataset, including electromagnetic transient characteristic analysis, electrical transient characteristic analysis, and steady-state characteristic analysis. After fusing the characteristics of each time scale, obtain the electrical characteristic correlation matrix and the multi-scale response feature set. Specifically, it includes: performing electromagnetic transient analysis with high-frequency sampling (10 kHz - 100 kHz) on grid parameters within a millisecond-level (0.1 ms - 10 ms) time window, and extracting the front steepness, peak amplitude, decay time constant, and oscillation frequency of the transient waveform; performing electrical transient analysis with medium-frequency sampling (1 kHz - 10 kHz) on parameters within a second-level (10 ms - 5 s) time window, and extracting the rise time, overshoot, settling time, and oscillation decay rate; performing steady-state analysis with low-frequency sampling (10 Hz - 100 Hz) on parameters within a minute-level (5 s - 300 s) time window, and calculating the steady-state deviation, recovery time, and steady-state fluctuation range; using wavelet transform to map the features of the three time scales in the time-frequency domain, calculating the time-scale feature transfer function relationship under different disturbance types, and constructing the electrical characteristic correlation matrix to characterize the mapping relationship among disturbance - parameter - time scale. Screen the disturbance - parameter pairs of the disturbance feature dataset according to the multi-scale response feature set and the electrical characteristic correlation matrix to obtain the effective disturbance - parameter pair set. Specifically, it includes: using the electrical characteristic correlation matrix to analyze the physical effectiveness of each disturbance - parameter pair, evaluating the physical rationality of the response based on the power system stability theory, and establishing a set of physical constraint conditions including power balance constraint, node volt-ampere constraint, and system stability constraint; introducing the time-scale dimension, calculating the mutual information of each disturbance - parameter pair at the electromagnetic transient, electrical transient, and steady-state scales, and constructing a three-dimensional mutual information matrix; using the K-means clustering method to automatically divide the high, medium, and low mutual information intervals, and dynamically adjusting the mutual information threshold according to the current grid operation state; calculating the physical sensitivity of the parameter pair to the disturbance in combination with the electrical characteristic correlation matrix, and identifying the parameter combinations with time-scale complementarity; comprehensively considering the mutual information, physical sensitivity, and time-scale characteristics, and applying the greedy algorithm to select the optimal disturbance - parameter pair on the premise of ensuring information complementarity. For example, [(load switching, U1), (load switching, I1), (power supply fluctuation, df / dt), (line fault, U2), (line fault, I2), (harmonic injection, U1), (harmonic injection, I1)]. For each element in the effective disturbance - parameter pair set, such as (load switching, U1), calculate the sensitivity coefficient of U1 under the load switching disturbance. The specific method is: for each load switching event, calculate the change in U1 before and after the load switching, ΔU1, and the power change in the load switching, ΔP. The sensitivity coefficient is defined as S = ΔU1 / ΔP. Calculate the average of the S values obtained for all load switching events to obtain the final sensitivity coefficient. Fill the sensitivity coefficient into the corresponding position of the initial disturbance response matrix framework.For example, the sensitivity coefficient corresponding to (load switching, U1) is filled in the position [0, 0, :] of the matrix (assuming that load switching is the 0th type of disturbance and U1 is the 0th parameter). Repeat this process for all valid disturbance-parameter pairs to obtain the sensitive disturbance response matrix. For each element in the set of valid disturbance-parameter pairs, such as (line fault, U2), calculate the response delay of U2 under the line fault disturbance. The specific method is as follows: for each line fault event, find the occurrence time T0 of the fault and the time T1 when U2 starts to change significantly (for example, the change amount of U2 exceeds 3 times its standard deviation). The response delay is defined as τ = T1 - T0. Calculate the average of the τ values obtained for all line fault events to get the final response delay. Convert the response delay to the number of time steps (τ / 5ms) and fill it in the corresponding position of the initial disturbance response matrix framework. For example, the response delay corresponding to (line fault, U2) is filled in the position [2, 1, :] of the matrix (assuming that the line fault is the 2nd type of disturbance and U2 is the 1st parameter). Repeat this process for all valid disturbance-parameter pairs to obtain the delay disturbance response matrix. Note that the values in the delay disturbance response matrix are delay points, not time series. For each element in the set of valid disturbance-parameter pairs, such as (power fluctuation, df / dt), calculate the duration of df / dt under the power fluctuation disturbance. The specific method is as follows: for each power fluctuation event, find the time T1 when df / dt starts to change significantly and the time T2 when df / dt returns to the normal range (for example, the change amount of df / dt is less than 3 times its standard deviation). The duration is defined as D = T2 - T1. Calculate the average of the D values obtained for all power fluctuation events to get the final duration. Convert the duration to the number of time steps (D / 5ms) and fill it in the corresponding position of the initial disturbance response matrix framework. For example, the duration corresponding to (power fluctuation, df / dt) is filled in the position [1, 4, :] of the matrix (assuming that the power fluctuation is the 1st type of disturbance and df / dt is the 4th parameter). Repeat this process for all valid disturbance-parameter pairs to obtain the duration disturbance response matrix. Note that the values in the duration disturbance response matrix are disturbance duration points, not time series. Normalize the sensitive disturbance response matrix, the delay disturbance response matrix, and the duration disturbance response matrix respectively. For the sensitive disturbance response matrix, use the maximum-minimum normalization method. Find the maximum value Smax and the minimum value Smin in the matrix, and for each element Sij, calculate the normalized value S'ij = (Sij - Smin) / (Smax - Smin). For the delay disturbance response matrix and the duration disturbance response matrix, since their values represent the number of time steps, use the method of dividing by the maximum value for normalization. Find the maximum value Tmax in the matrix, and for each element Tij, calculate the normalized value T'ij = Tij / Tmax. Add the three normalized matrices together to obtain the disturbance response matrix.

[0092] Preferably, the information channel modulation in step S2 includes:

[0093] Obtain the data to be transmitted; perform data classification and block processing on the data to be transmitted according to the perturbation response matrix to obtain a hierarchical data block group;

[0094] Perform parameter channel matching analysis according to the perturbation response matrix to obtain the channel matching degree; generate a channel optimal matching strategy for the hierarchical data block group according to the channel matching degree to obtain a channel allocation strategy table;

[0095] Perform micro-perturbation parameter modulation on the hierarchical data block group according to the channel allocation strategy table to obtain a modulation parameter sequence.

[0096] In an embodiment of the present invention, it is assumed that the data to be transmitted is a binary sequence with a length of 1024 bits, representing a protection tripping instruction between substations. First, the data is classified according to its importance. The first 128 bits are defined as the highest priority (Level1), representing a direct tripping instruction; the next 256 bits are defined as a higher priority (Level2), representing a tripping permission instruction; the remaining 640 bits are defined as a normal priority (Level3), representing status information. Then, the data with different priorities is block-divided. The Level1 data is not block-divided and maintains the integrity of 128 bits. The Level2 data is divided into 4 blocks, each with 64 bits. The Level3 data is divided into 10 blocks, each with 64 bits. In this way, a hierarchical data block group is obtained, including 1 Level1 block, 4 Level2 blocks, and 10 Level3 blocks, a total of 15 data blocks. Using the perturbation response matrix generated in step S2, the matching degree of each data block with different power grid parameters is evaluated. For the Level1 data block, the parameter channel with the fastest response and the highest sensitivity is selected. By referring to the time-delay perturbation response matrix, it is found that the negative sequence current (I2) has the shortest response time-delay to line faults. By referring to the sensitivity perturbation response matrix, it is also found that the sensitivity coefficient of the negative sequence current (I2) to line faults is relatively high. Therefore, the Level1 data block is matched with the negative sequence current (I2) channel, and the matching degree is rated as "high". For the Level2 data block, a parameter channel with a relatively fast response, relatively high sensitivity, and low correlation with the channel selected for the Level1 data block is selected. Considering the time-delay, sensitivity, and continuous perturbation response matrix comprehensively, the positive sequence voltage (U1) is selected as the matching channel for the Level2 data block, and the matching degree is rated as "medium". For the Level3 data block, the remaining parameter channels are selected, including the positive sequence current (I1), the rate of change of frequency (df / dt), and the active power (P), and are allocated according to their response characteristics to different perturbation types, and the matching degree is rated as "low". According to the matching degree, a channel allocation strategy table is generated. This table details the parameter channels corresponding to each data block, for example: [Level 1: I2, Level 2: U1, U1, U1, U1, Level 3: I1, I1,..., df / dt, P, P]. According to the channel allocation strategy table, specific parameter modulation is performed on each data block. For the Level1 data block (mapped to I2), binary frequency shift keying (BFSK) modulation is used. The "0" in the data block is mapped to a slight decrease ΔI2 in the negative sequence current amplitude (for example, 0.1% of the rated value), with a duration of T1 (for example, 5 ms); the "1" is mapped to a slight increase ΔI2 in the negative sequence current amplitude, with a duration of T1. For the Level2 data block (mapped to U1), binary phase shift keying (BPSK) modulation is used.Map the "0" in the data block to a slightly lagging positive sequence voltage phase Δφ (e.g., 1 degree) for a duration of T2 (e.g., 10 ms); map the "1" to a slightly leading positive sequence voltage phase Δφ for a duration of T2. For Level 3 data blocks, different modulation methods are adopted according to the mapped parameter channels. For example, the data block mapped to I1 adopts BFSK modulation similar to I2, the data block mapped to df / dt adopts BPSK modulation similar to U1, and the data block mapped to P adopts pulse width modulation (PWM). Combine the modulated parameter changes in chronological order to form a modulation parameter sequence. This sequence contains the information of all data blocks, but this information is hidden in the slight changes of the grid parameters. These slight changes are controlled within the range allowed by the normal operation of the grid and will not trigger any protection actions or alarms.

[0097] As an example of the present invention, refer to Figure 2 shown. In this example, step S3 includes:

[0098] Step S31: Perform channel correlation analysis on the modulation parameter sequence to obtain a channel characteristic evaluation table;

[0099] In the embodiment of the present invention, receive the modulation parameter sequence generated in step S2, and this sequence contains the modulation information of different grid parameters (U1, U2, I1, I2, df / dt, P). First, group the sequence according to the parameter type. Then, calculate the correlation coefficient between every two parameter groups. The Pearson Correlation Coefficient is used to measure the linear correlation. The calculation formula is: ρ(X, Y) = cov(X, Y) / (σX * σY), where cov(X, Y) is the covariance of X and Y, and σX and σY are the standard deviations of X and Y respectively. The closer the absolute value of the correlation coefficient is to 1, the stronger the correlation; the closer it is to 0, the weaker the correlation. At the same time, evaluate the transmission characteristics of each parameter channel. The transmission delay is determined by measuring the propagation time of the signal from the sending end to the receiving end. The bandwidth limitation is determined by analyzing the frequency response characteristics of the channel. The bit error rate is determined by transmitting a known test sequence in the channel and comparing the received sequence with the original sequence. The anti-interference ability is evaluated by injecting different types of noise (such as Gaussian white noise, pulse noise) into the channel and measuring the bit error rate. The information concealment is evaluated by analyzing the power spectral density (PSD) of the modulated signal. The flatter the PSD, the better the concealment. Finally, comprehensively consider the correlation, transmission delay, bandwidth limitation, bit error rate, anti-interference ability, and information concealment to generate a channel characteristic evaluation table. Each row in the table represents a parameter channel, each column represents a characteristic index, and each index is scored (e.g., 1 - 5 points, 5 points indicating the best).

[0100] Step S32: Perform time interleaving analysis based on the modulation parameter sequence and the channel characteristic evaluation table to obtain the time-interleaved data stream;

[0101] In the embodiment of the present invention, according to the length of the modulation parameter sequence and the channel characteristic evaluation table, a time interleaving matrix with variable depth is designed. Assume that the length of the modulation parameter sequence is 1000 units (each unit represents a modulation symbol). First, divide the sequence into 10 time segments, each segment being 100 units. For each time segment, dynamically adjust the interleaving depth according to the security requirements of the corresponding parameter channel (determined by the channel characteristic evaluation table). For example, for the segment mapped to the I2 channel (high security requirement), an interleaving with a depth of 50 is adopted; for the segment mapped to the P channel (low security requirement), an interleaving with a depth of 10 is adopted. The pseudo-random interleaving algorithm is used. Specifically, generate a pseudo-random number sequence with the same length as the segment, and sort the elements within the segment according to the size of the pseudo-random numbers. The sorted sequence is the interleaved sequence. Record the original position and the interleaved position of each element to generate an interleaving mapping table. Concatenate the interleaved sequences of all time segments in order to obtain the time-interleaved data stream. At the same time, generate a total interleaving mapping table, which records the interleaving relationship of the entire time-interleaved data stream.

[0102] Step S33: Obtain the power grid topology information; optimize the spatial distribution strategy according to the time-interleaved data stream and the power grid topology information to obtain the spatial distribution scheme;

[0103] In the embodiment of the present invention, obtain the power grid topology information from the power grid geographic information system (GIS). Specifically, it includes: the location coordinates of substations, the connection relationships of transmission lines, the types and capacities of transformers, etc. According to the topology information, construct a node connection graph of the power grid. The nodes in the graph represent substations, and the edges represent transmission lines. Divide the time-interleaved data stream into multiple spatial slices. The size of the slices is determined according to the communication capabilities and storage capacities of the nodes. For example, for nodes with large communication bandwidth and high storage capacity, allocate larger slices; for nodes with small communication bandwidth and low storage capacity, allocate smaller slices. The multi-path routing algorithm is used. For important data slices (for example, slices containing Level 1 data), select multiple physical paths for transmission to increase redundancy. For data slices with high correlation (for example, slices mapped to U1 and I1), select nodes with a relatively long physical distance for transmission to reduce the risk of being intercepted simultaneously. Preferentially select nodes with high reliability (for example, nodes with a low historical failure rate) for transmission. Assign a unique spatial location identifier to each data slice and generate routing information (including the transmission path and the next-hop node). Combine the allocation information and routing information of all slices to generate the spatial distribution scheme.

[0104] Step S34: Perform spatio-temporal fusion encoding on the time-interleaved data stream according to the spatial distribution scheme to obtain the spatio-temporal fusion encoding;

[0105] In the embodiment of the present invention, according to the spatial distribution scheme, each slice in the time-interleaved data stream is assigned to different spatial nodes. A two-dimensional spatio-temporal matrix is constructed. The rows of the matrix represent time (corresponding to the time segments in the time-interleaved data stream), and the columns represent space (corresponding to the nodes in the spatial distribution scheme). Each data slice is filled into the corresponding position in the matrix according to its time and space positions. Interleaved key slices are embedded in the matrix. The number of key slices is equal to the number of spatial nodes, and each node holds one slice. A two-way extension algorithm is adopted. First, in the time dimension, a cyclic shift operation is performed on each column (i.e., the data of each node), and the shift amount is determined by the key slice of that node. Then, in the spatial dimension, an exclusive OR operation is performed on each row (i.e., the data of each time segment), and the exclusive OR key is determined by the key slice corresponding to that time segment. Synchronization markers are added. A unique synchronization codeword is added to the head of each data slice for synchronization and data recovery at the receiving end. The spatio-temporal mixing degree is calculated. The mixing degree is evaluated by calculating the entropy of each row and each column in the matrix. The higher the entropy, the better the mixing degree. If the mixing degree is lower than the preset threshold, the parameters of the two-way extension algorithm (such as the number of bits of cyclic shift, the exclusive OR key) are adjusted, and the encoding is performed again until the mixing degree requirement is met. The finally generated matrix is the spatio-temporal fusion encoding.

[0106] Step S35: Perform dynamic channel transmission on the spatio-temporal fusion encoding according to the channel characteristic evaluation table to obtain the spatio-temporal interleaved code stream;

[0107] In the embodiments of the present invention, an electrical parameter wave impedance model is established according to the power grid topology information, the main electrical transmission paths are extracted through eigenvalue decomposition to obtain a set of main paths; a physical transmission feasibility evaluation is performed according to the set of main paths and the channel characteristic evaluation table to obtain a physically enhanced channel evaluation table; a dynamic channel transmission configuration is performed on the spatio-temporal fusion coding according to the physically enhanced channel evaluation table to obtain a spatio-temporal interleaved code stream with electrical and physical robustness; specifically including: establishing a π-type equivalent circuit model according to the power grid topology structure, calculating the characteristic impedance, propagation constant and wave impedance between each node to form a network wave impedance matrix; performing eigenvalue decomposition on the wave impedance matrix to extract the main physical transmission paths corresponding to the eigenvectors, and the main paths are composed of node sequences with the smallest transmission impedance; performing matching analysis on the set of main paths and the channel characteristic evaluation table, calculating the electrical robustness indexes (including anti-interference ability, signal integrity and physical delay) of each main path, and generating a physically enhanced channel evaluation table; performing dynamic transmission configuration according to the physically enhanced channel evaluation table and the data importance level, selecting multi-channel redundancy transmission with high electrical robustness for critical data (Level 1), selecting the channel with the lowest physical delay for time-sensitive data, and selecting the channel with the most matching physical wave impedance for large-capacity data; adding electrical characteristic adaptation identifiers, routing information and CRC check codes to each transmission data packet; encrypting the data packet using the AES-256 algorithm and performing integrity verification before transmission; sending the data packets in the order of the electrical main path priorities to achieve the transmission of a spatio-temporal interleaved code stream with electrical and physical robustness.

[0108] Preferably, step S34 includes:

[0109] Extract the number of time interleaving units in the time interleaved data stream, extract the number of spatial nodes participating in the transmission in the spatial distribution scheme, and construct a spatio-temporal matrix framework according to the number of time interleaving units and the number of spatial nodes;

[0110] Perform data block mapping and filling according to the spatio-temporal matrix framework to obtain an initial spatio-temporal matrix;

[0111] Divide the pre-generated interleaving key into multiple shards, and the number of shards is equal to the number of spatial nodes to obtain key shards;

[0112] Embed the key shards into the initial spatio-temporal matrix to obtain a spatio-temporal matrix with keys;

[0113] Perform time dimension expansion on the spatio-temporal matrix with keys, and perform space dimension expansion to obtain an expanded spatio-temporal matrix;

[0114] Add synchronization marks to the expanded spatio-temporal matrix to obtain a spatio-temporal matrix with marks;

[0115] Perform spatio-temporal mixing degree evaluation on the spatio-temporal matrix with marks to obtain spatio-temporal fusion coding.

[0116] In an embodiment of the present invention, the number of time-interleaved units is extracted from the time-interleaved data stream generated in step S32. Suppose the time-interleaved data stream is divided into 10 time segments, and each time segment is used as an interleaved unit, then the number of time-interleaved units is 10. The number of spatial nodes participating in transmission is extracted from the spatial distribution scheme generated in step S33. Suppose the spatial distribution scheme selects 5 substations as transmission nodes, then the number of spatial nodes is 5. A two-dimensional matrix framework is constructed, with the number of rows being the number of time-interleaved units (10) and the number of columns being the number of spatial nodes (5). This matrix framework will serve as the basis for spatio-temporal fusion coding. Each element of the matrix will be used to store a data block or a part thereof. According to the data block allocation strategy specified in the spatial distribution scheme, the data blocks in the time-interleaved data stream are filled into the spatio-temporal matrix framework one by one. The filling rule is as follows: In chronological order, the data blocks of each time segment are sequentially allocated to different spatial nodes. For example, the data block of the first time segment is allocated to the first node, the data block of the second time segment is allocated to the second node, and so on. If the number of nodes is less than the number of time segments, the allocation is cyclic. For example, the data block of the sixth time segment is allocated to the first node again. If the size of a data block exceeds the space allocated to the node, the data block is split into multiple sub-blocks and filled into multiple elements of the matrix respectively. After filling, the initial spatio-temporal matrix is obtained. Each element of this matrix contains a data block or a sub-block thereof. A 256-bit key is pre-generated as the total key for spatio-temporal interleaving. The generation of the key uses a hardware random number generator to ensure its true randomness. This 256-bit key is split into 5 shards, each shard being 51 or 52 bits (since 256 is not divisible by 5, some are 51 bits and some are 52 bits, keeping the total length of 256 bits unchanged). The number of shards is equal to the number of spatial nodes (5). The splitting method is sequential splitting, that is: the first 51 bits of the key are used as the first shard, the next 52 bits are used as the second shard, and so on. Each shard will be allocated to a spatial node for subsequent spatio-temporal expansion operations. The 5 key shards generated in the previous step are embedded into the initial spatio-temporal matrix. The embedding method is as follows: For each spatial node (corresponding to a column of the matrix), a fixed position (for example, the first element of this column) is selected to store the key shard of this node. The value of the key shard directly overwrites the original data block content at this position. In this way, each node holds a key shard and the position of the key shard is fixed. After embedding the key shards, the spatio-temporal matrix with keys is obtained. The spatio-temporal matrix with keys is expanded bidirectionally. First, the time dimension is expanded. For each column (corresponding to a spatial node), according to the key shard of this node, a cyclic shift operation is performed on the data of this column. The number of shifted bits is determined by the value of the key shard. For example, if the value of the key shard is 10, the data of this column is cyclically shifted upward by 10 bits. Then, the space dimension is expanded.For each row (corresponding to a time segment), according to the key shard corresponding to this time segment (modulo operation, taking the modulo of the row number by the number of shards 5, getting 0 - 4, corresponding to 5 shards), perform an exclusive OR operation on the data of this row. For example, if the modulo of the row number by 5 is 1, then use the second key shard and perform a bitwise exclusive OR operation between this key shard and each element of this row. After two-way expansion, an extended spatio-temporal matrix is obtained. Add a synchronization marker to the head of each data block (or sub-block) in the extended spatio-temporal matrix. The synchronization marker uses the Barker Code. The Barker Code is a binary sequence with good autocorrelation characteristics. Select a Barker Code with a length of 13: +1+1+1+1+1-1-1+1+1-1+1-1+1. Add this 13-bit Barker Code to the head of each data block. Since a data block may be split into multiple sub-blocks, each sub-block needs to add a synchronization marker. After adding the synchronization marker, a spatio-temporal matrix with markers is obtained. Perform a spatio-temporal mixing degree evaluation on the spatio-temporal matrix with markers. Calculate the entropy of each row and each column of the matrix respectively. The formula for entropy is: H = -Σp(x)log2(p(x)), where p(x) is the probability of the occurrence of data block x. For each row, regard the data blocks of this row as a set and calculate the entropy of the set. For each column, similarly regard the data blocks of this column as a set and calculate the entropy of the set. Calculate the average value of all row entropies and the average value of all column entropies. Add these two average values to obtain the spatio-temporal mixing degree index. Set a threshold (for example, 4.5, assuming the value range of the data block is 0 - 255 and the maximum entropy is 8). If the spatio-temporal mixing degree index is lower than this threshold, it is considered that the mixing degree is insufficient, and the parameters of the two-way expansion algorithm (such as the number of bits of cyclic shift, the exclusive OR key) need to be adjusted, and steps 5, 6, and 7 are performed again until the mixing degree index meets the threshold requirement. The finally spatio-temporal matrix with markers that meets the mixing degree requirement is the spatio-temporal fusion coding.

[0117] Preferably, the spatio-temporal deinterleaving process described in step S4 includes:

[0118] The receiving end simultaneously monitors multiple physical communication channels, synchronously receives the spatio-temporal interleaved code stream, and obtains a set of synchronous data frames;

[0119] Perform spatio-temporal deinterleaving reconstruction on the set of synchronous data frames to obtain a reconstructed modulation sequence;

[0120] Obtain an updated perturbation response matrix; perform channel demodulation extraction on the reconstructed modulation sequence according to the updated perturbation response matrix to obtain a set of channel data blocks.

[0121] In the embodiments of the present invention, multiple receivers are deployed at the receiving end, and each receiver corresponds to a physical communication channel (for example, different carrier frequencies or different optical fibers). These receivers simultaneously monitor all channels and receive spatio-temporal interleaved code stream segments from different spatial nodes. Each receiver continuously detects the synchronization marker (Barker code) in the received signal. Once the Barker code is detected, subsequent data is started to be recorded as a data frame. Due to delays and jitters during transmission, the data frames received by different receivers may be out of sync. Phase-locked loop (PLL) technology is used for synchronization. One receiver is selected as the master clock source, and the other receivers are used as slave clock sources. The PLL of each slave clock source aligns its local clock with the synchronization marker of the master clock source, thereby achieving time synchronization of all receivers. The synchronization accuracy is controlled within 1 microsecond. The data frames received by all receivers and synchronized in time are aggregated to form a synchronized data frame set. Spatio-temporal deinterleaving is performed on the synchronized data frame set. First, the key fragments in each data frame are extracted. According to the positions of the key fragments (the first elements of the columns corresponding to each node), 5 key fragments are extracted. These 5 key fragments are concatenated in the order of the transmitting end to restore the complete 256-bit interleaved key. Then, spatial dimension deinterleaving is performed. For each row (corresponding to a time segment), according to the key fragment corresponding to this time segment (obtained by modulo operation on the row number), the data in this row is XORed to restore the data before spatial dimension interleaving. Next, time dimension deinterleaving is performed. For each column (corresponding to a spatial node), according to the key fragment of this node, the data in this column is circularly shifted backward to restore the data before time dimension interleaving. According to the spatio-temporal matrix filling rule recorded in step S34, the data blocks (or sub-blocks) in the data frame are extracted from the matrix and concatenated in the order of the transmitting end. Finally, the reconstructed modulation sequence is obtained. This sequence should be consistent with the modulation parameter sequence generated in step S32 in an ideal situation. The receiving end regularly (for example, every 5 minutes) obtains the latest power grid operation status information from the power grid dispatching center, including generator output, load level, line parameters, etc. Using this latest information, the disturbance response matrix is recalculated. The recalculation process is the same as the disturbance response matrix generation process in step S2, but the latest power grid data is used. This new disturbance response matrix is called the updated disturbance response matrix. According to the updated disturbance response matrix, channel demodulation is performed on the reconstructed modulation sequence. First, the reconstructed modulation sequence is grouped according to parameter types (U1, U2, I1, I2, df / dt, P). Then, for each parameter group, demodulation is performed according to its modulation method. For example, for the I2 parameter group using BFSK modulation, the change in the negative sequence current amplitude is detected, and a slight decrease of ΔI2 is demodulated to "0", and a slight increase of ΔI2 is demodulated to "1".For the U1 parameter group using BPSK modulation, detect the change in the phase of the positive-sequence voltage, demodulate the slight lag Δφ as "0", and demodulate the slight lead Δφ as "1". For the P parameter group using PWM modulation, demodulate the corresponding data according to the pulse width. Combine the demodulation results of all parameter channels in the order of data block division in step S2 to obtain a set of channel data blocks. This set of data blocks should be consistent with the hierarchical data block group generated in step S2 ideally.

[0122] Preferably, the multi-point multi-dimensional electrical parameter verification in step S4 includes:

[0123] The receiving end measures the local power grid parameters to obtain a local parameter signature vector;

[0124] Perform a syntax layer check on the set of channel data blocks according to the local parameter signature vector to obtain a first-level comparison result;

[0125] Check the rationality of the value range of the first-level verification result and perform a time coherence check to obtain a second-level comparison result;

[0126] Perform a power flow constraint check on the second-level comparison result and perform a state estimation verification to obtain physically constrained state data;

[0127] Perform multi-point data cross-verification according to the physically constrained state data to obtain a third-level comparison result;

[0128] Perform a comprehensive credibility evaluation according to the first-level comparison result, the second-level comparison result, and the third-level comparison result to obtain a set of data block credibility;

[0129] Generate an integrity analysis report according to the data block credibility level to obtain a data integrity report.

[0130] In the embodiments of the present invention, the receiving end is equipped with PMUs of the same model and configuration as those at the sending end. These PMUs independently measure the local power grid parameters of the substation where the receiving end is located: three-phase voltages (Ua, Ub, Uc), three-phase currents (Ia, Ib, Ic), positive-sequence voltage (U1), negative-sequence voltage (U2), zero-sequence voltage (U0), positive-sequence current (I1), negative-sequence current (I2), zero-sequence current (I0), frequency (f), and rate of change of frequency (df / dt). The sampling rate of the PMUs is also set to 4800 samples per second and is synchronized with the GPS signal. The same "differential quantization and coding" process as in step S1 is performed on these local measurement data, including parameter difference screening, differential sampling, differential quantization, multi-mode coding transformation, and eigenvector integration. Finally, a local parameter feature code vector with exactly the same format as the parameter feature code vector at the sending end is obtained. This vector represents the independent measurement result of the local power grid state at the receiving end. Perform a syntax layer check on the channel data block set obtained by "channel demodulation extraction" in step S4. First, check whether the length of each data block conforms to the predefined specification (for example, the Level 1 data block should be 128 bits, the Level 2 data block should be 64 bits, and the Level 3 data block should be 64 bits). Then, check whether the coding format of each data block is correct. For example, for a data block encoded with Hamming code, check whether its parity bits are correct; for a data block encoded with CRC code, check whether its CRC checksum is zero. Finally, compare the parameter value after decoding each data block with the current value of the corresponding parameter in the local parameter feature code vector. Calculate the difference between the two. If the difference exceeds the preset threshold (for example, the voltage difference exceeds 5% of the rated value, and the frequency difference exceeds 0.1 Hz), it is marked as a syntax layer anomaly. Record the verification results (whether the length is correct, whether the coding is correct, and whether the parameter value is close to the local measurement value) of all data blocks to form a first-level comparison result. Perform a range rationality check on the data blocks marked as "normal" in the first-level comparison result. For each parameter, set its reasonable value range according to the physical laws of power grid operation. For example, the voltage is usually between 90% and 110% of the rated value, and the frequency is usually between 49.5 Hz and 50.5 Hz. If the parameter value after decoding the data block exceeds these ranges, it is marked as a range anomaly. Then, perform a time coherence check. For each parameter, analyze its change trend over a period of time. If the parameter value after decoding a certain data block shows a sudden change (for example, the voltage suddenly increases or decreases by 10%) compared with the parameter values of its adjacent data blocks before and after, and this sudden change cannot be explained by the normal operation state change of the power grid, it is marked as a time coherence anomaly. Combine the results of the range rationality check and the time coherence check with the first-level comparison result to form a second-level comparison result. Perform a power flow constraint check on the data blocks marked as "normal" in the second-level comparison result.Based on the topological structure and line parameters of the power grid, a power flow calculation model is established. The parameter values such as voltage, current, and power after decoding the data block are used as inputs for power flow calculation. Check whether the calculation results meet the physical constraints of power grid operation, such as: whether the voltage at each node is within the allowable range, whether the power of each line exceeds its transmission capacity, and whether the generator output is within its rated range. If the power flow calculation results violate these constraints, they are marked as power flow constraint anomalies. Then, state estimation verification is carried out. Using the voltage and current phasors measured by the local PMU at the receiving end, state estimation is performed. The parameter values after decoding the data block are compared with the results of state estimation. If the deviation between the two exceeds the preset threshold (for example, the voltage deviation exceeds 1%, and the phase angle deviation exceeds 1 degree), it is marked as a state estimation anomaly. The results of power flow constraint check and state estimation verification are combined to obtain the physical constraint status data. Using the topological information of the power grid, multi-point data cross-verification is carried out. Select the receiving ends of multiple adjacent substations and compare their respective physical constraint status data. The parameters for comparison include: voltage amplitude, voltage phase angle, frequency, active power, reactive power, etc. Calculate the correlation coefficients of these parameters between different substations. If the distance between two substations is relatively close and the electrical connection is tight, their parameters should have a high correlation. If the correlation coefficient is lower than the preset threshold (for example, 0.8), it is considered abnormal. Further analyze the cause of the anomaly to determine whether it is caused by data errors, measurement errors, or power grid faults. The results of multi-point data cross-verification are combined with the physical constraint status data to form the three-level comparison results. Considering the first-level comparison results, second-level comparison results, and third-level comparison results comprehensively, the credibility of each data block is evaluated. The weighted scoring method is adopted. Different weights are assigned to different levels of comparison results. For example, the weight of the first-level comparison result is 0.2, the weight of the second-level comparison result is 0.3, and the weight of the third-level comparison result is 0.5. For each data block, if it is marked as "abnormal" in a certain level of comparison, it gets 0 points at that level; if it is marked as "normal", it gets 1 point at that level. Multiply the score of each level by its weight and then sum them up to get the comprehensive credibility score of the data block. The scoring range is 0 - 1, where 1 represents completely credible and 0 represents completely non-credible. The credibility scores of all data blocks are aggregated to form the data block credibility set. Based on the data block credibility set, an integrity analysis report is generated. The report content includes: the credibility score of each data block, the comparison results (normal / abnormal) of each level, the analysis of the cause of the anomaly (if there is an anomaly), and the overall credibility assessment (for example, 95% of the data blocks have a credibility higher than 0.9). For data blocks with a credibility lower than the threshold (for example, 0.5), they are marked as "non-credible", and it is recommended that the receiving end take corresponding measures (such as requesting retransmission, using redundant data, and performing fault diagnosis).The integrity analysis report provides the receiving end with detailed information about the quality of the received data, which helps the receiving end make correct decisions.

[0131] Preferably, the multi-point data cross-verification includes:

[0132] Construct a node adjacency matrix according to the power grid topology information, compare the electrical distance thresholds based on the node adjacency matrix and the physical constraint status data, and perform a correlation score for the node pairs to obtain a list of associated node pairs and the correlation scores of the node pairs;

[0133] Align the time series of the same type of parameters for the physical constraint status data according to the list of associated node pairs to obtain a sequence of the same type of parameter pairs;

[0134] Calculate the correlation coefficient according to the sequence of the same type of parameter pairs and the correlation scores of the node pairs to obtain a weighted correlation coefficient matrix;

[0135] Set the deviation threshold according to the weighted correlation coefficient matrix to obtain a dynamic deviation threshold table;

[0136] Mark the abnormal data according to the dynamic deviation threshold table and the weighted correlation coefficient matrix to obtain a three-level comparison result.

[0137] In the embodiments of the present invention, power grid topology information is obtained from the power grid Geographic Information System (GIS), including the locations of substations, the connection relationships of transmission lines, and the types and parameters of transformers. According to this information, a node adjacency matrix A is constructed. A is an N×N square matrix, where N is the number of substations (nodes) in the power grid. If there is a direct transmission line connection between node i and node j, then Aij = 1; otherwise, Aij = 0. According to the physical constraint status data, the status information (voltage, current, power, etc.) of each node is obtained. The electrical distance Dij between any two nodes i and j is calculated. The electrical distance can be defined as the sum of the impedances of the shortest path between the two nodes. The Dijkstra algorithm or the Floyd-Warshall algorithm is used to calculate the shortest path. An electrical distance threshold Dthreshold is set (for example, 0.5 pu, per-unit value). If Dij < Dthreshold, then nodes i and j are considered strongly correlated. According to the electrical distance and the node status information, a correlation score is given to the node pair. The scoring rules are as follows: the smaller the electrical distance, the higher the correlation score; the more similar the node status (for example, the closer the voltage amplitude and phase angle), the higher the correlation score. For example, the following formula can be used for scoring: Scoreij = k1*(1 - Dij / Dmax) + k2*exp(-||Vi - Vj||2), where Dmax is the maximum electrical distance, Vi and Vj are the voltage vectors of nodes i and j respectively, and k1 and k2 are weight coefficients. All node pairs with a correlation score higher than the threshold (for example, 0.8) are added to the associated node pair list, and their correlation scores are recorded. For each node pair (i, j) in the associated node pair list, the same type of parameter data of nodes i and j is extracted from the physical constraint status data. For example, if nodes i and j are both substations, then their voltage amplitudes, voltage phase angles, active power, reactive power, etc. are extracted. Since there may be a slight time deviation in the PMU sampling of different nodes, time series alignment is required. The cross-correlation method is used for alignment. The cross-correlation function of the time series of the same parameter of nodes i and j is calculated. The time delay corresponding to the peak of the cross-correlation function is found. According to this time delay, the parameter time series of one of the nodes (for example, node j) is shifted to align it with the parameter time series of the other node (node i). The aligned same type of parameter data of nodes i and j are grouped into a pair to form a same type of parameter pair sequence. For example, (Ui(t), Uj(t)), (Pi(t), Pj(t)), etc., where t represents time. For each same type of parameter pair sequence, its correlation coefficient is calculated. The Pearson Correlation Coefficient is used. For example, for the voltage amplitude pair sequence (Ui(t), Uj(t)), its correlation coefficient ρ(Ui, Uj) is calculated. The calculated correlation coefficient is weighted.The weight is determined by the relevance score of the node pair. For example, if the relevance score of the node pair (i, j) is Scoreij, then the weight of the voltage amplitude correlation coefficient is Scoreij. The weighted correlation coefficient is Scoreij * ρ(Ui, Uj). The weighted correlation coefficients of all parameters of all node pairs are combined into a matrix to obtain the weighted correlation coefficient matrix R. The dimension of R is M×M, where M is the number of nodes participating in the multi-point data cross-validation. Rij represents the average value of the weighted correlation coefficients of all parameters between node i and node j. According to the weighted correlation coefficient matrix R, a dynamic deviation threshold is set for each parameter of each node pair. The principle for setting the threshold is as follows: the higher the relevance, the smaller the threshold; the lower the relevance, the larger the threshold. For example, the following formula can be used to set the deviation threshold of the voltage amplitude: ΔUthreshold,ij = k * (1 - Rij) * Unominal, where Unominal is the rated value of the voltage and k is a coefficient (for example, 0.05, indicating a 5% deviation is allowed). For other parameters (voltage phase angle, active power, reactive power, etc.), a similar formula is used to set the deviation threshold, but the value of the coefficient k may be different. The deviation thresholds of all parameters of all node pairs are combined into a table to obtain the dynamic deviation threshold table. According to the dynamic deviation threshold table, abnormal data marking is performed on each parameter of each node pair. For the voltage amplitudes Ui(t) and Uj(t) of the node pair (i, j), if |Ui(t) - Uj(t)| > ΔUthreshold,ij, then the voltage amplitude data at this moment is marked as abnormal. Similar comparisons and markings are also performed for other parameters. If a certain node is marked as abnormal in multiple parameters or in comparisons with multiple nodes, then the data of this node at this moment is marked as suspicious. If a certain node is marked as abnormal in most parameters and in comparisons with most nodes, then the data of this node at this moment is marked as abnormal. The abnormal marking results of all nodes are combined to obtain the three-level comparison result. The three-level comparison result details the data status (normal, suspicious, abnormal) of each node at each moment.

[0138] Preferably, the present invention further provides a data information secure transmission system for performing the data information secure transmission method as described above. The data information secure transmission system includes:

[0139] A multi-dimensional parameter encoding module for collecting the original power grid measurement data; performing differential quantization and encoding on the original power grid measurement data to obtain a parameter feature code vector;

[0140] An information channel modulation module for constructing a power grid disturbance response matrix according to the parameter feature code vector to obtain a disturbance response matrix; acquiring the data to be transmitted; performing information channel modulation on the data to be transmitted according to the disturbance response matrix to obtain a modulation parameter sequence;

[0141] A multi-channel fusion transmission module is used to perform time-interleaving analysis according to a modulation parameter sequence to obtain a time-interleaved data stream; acquire power grid topology information; perform spatio-temporal fusion coding transmission based on the power grid topology information and the time-interleaved data stream to obtain a spatio-temporal interleaved code stream;

[0142] A receiving-end demodulation and decoding module is used to perform spatio-temporal de-interleaving processing on the spatio-temporal interleaved code stream to obtain a set of channel data blocks; perform multi-point multi-dimensional electrical parameter verification on the set of channel data blocks to obtain a data integrity report; perform adaptive data reorganization and recovery on the set of channel data blocks according to the data integrity report to obtain a check result flag, so as to implement the secure transmission task of power grid data information.

[0143] Therefore, from any perspective, the embodiments should be regarded as exemplary and non-restrictive. The scope of the present invention is defined by the appended claims rather than the above description. Therefore, all changes falling within the meaning and scope of the equivalent elements of the application documents are intended to be encompassed within the present invention.

[0144] The above description is only a specific implementation manner of the present invention, enabling those skilled in the art to understand or implement the present invention. Various modifications to these embodiments will be obvious to those skilled in the art. The general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to these embodiments shown herein, but rather will conform to the broadest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for secure transmission of data information, characterized in that, It includes the following steps: Step S1: Collect the original grid measurement data; perform differential quantization and coding on the original grid measurement data to obtain a parameter feature code vector; Step S2: Construct a grid disturbance response matrix based on the parameter feature code vector to obtain a disturbance response matrix; acquire the data to be transmitted; perform information channel modulation on the data to be transmitted according to the disturbance response matrix to obtain a modulation parameter sequence; Step S3: Perform time interleaving analysis according to the modulation parameter sequence to obtain a time interleaved data stream; Acquire the grid topology information; perform spatio-temporal fusion coding transmission according to the grid topology information and the time interleaved data stream to obtain a spatio-temporal interleaved code stream; Step S4: Perform spatio-temporal deinterleaving processing on the spatio-temporal interleaved code stream to obtain a channel data block set; perform multi-point multi-dimensional electrical parameter verification on the channel data block set to obtain a data integrity report; perform adaptive data reorganization and recovery on the channel data block set according to the data integrity report to obtain a check result flag, so as to achieve the task of secure transmission of grid data information.

2. The data information secure transmission method according to claim 1, characterized in that The differential quantization and coding described in Step S1 includes: Select multiple significantly different grid parameters for acquisition to obtain the original grid measurement data, where the original grid measurement data includes frequency, voltage amplitude, voltage phase angle, amplitudes and phases of each harmonic; perform parameter difference screening on the original grid measurement data to obtain a set of difference parameters; Perform difference sampling processing on the set of difference parameters to obtain time domain feature data; Perform differential quantization processing according to the time domain feature data to obtain a quantization feature group, and the differential quantization processing includes: performing high-precision quantization on the frequency, medium-precision quantization on the voltage amplitude, low-precision quantization on the voltage phase angle, and quantizing the amplitudes and phases of each harmonic respectively; Perform multi-mode coding transformation on the quantization feature group to obtain a multi-mode coding table; Integrate the quantization feature group according to the multi-mode coding table to obtain a parameter feature code vector.

3. The data information secure transmission method according to claim 1, wherein The generation of the disturbance response matrix described in Step S2 includes: Collect historical data of typical disturbance events of the grid; perform disturbance characteristic analysis according to the parameter feature code vector and the historical data of typical disturbance events to obtain a disturbance characteristic data set, where the disturbance characteristic data set includes load switching disturbance, power supply fluctuation disturbance, line fault disturbance, and harmonic injection disturbance; Perform disturbance framework initialization according to the disturbance characteristic data set to obtain an initial disturbance response matrix framework; Extract multi-scale response characteristics from the disturbance characteristic data set, including electromagnetic transient characteristic analysis, electrical transient characteristic analysis, and steady-state characteristic analysis, and fuse the characteristics of each time scale to obtain an electrical characteristic correlation matrix and a multi-scale response characteristic set; Perform disturbance-parameter pair screening on the disturbance characteristic data set according to the multi-scale response characteristic set and the electrical characteristic correlation matrix to obtain an effective disturbance-parameter pair set; Calculate the sensitivity coefficient of the disturbance characteristic data set according to the effective disturbance-parameter pair set to obtain a sensitive disturbance response matrix; Calculate the response delay of the disturbance characteristic data set according to the effective disturbance-parameter pair set to obtain a delay disturbance response matrix; Calculate the duration of the disturbance characteristic data set according to the effective disturbance-parameter pair set to obtain a continuous disturbance response matrix; The sensitivity disturbance response matrix, time-delay disturbance response matrix, and continuous disturbance response matrix are normalized to obtain the disturbance response matrix.

4. The data information secure transmission method according to claim 1, wherein, The information channel modulation described in step S2 includes: Obtain the data to be transmitted; perform data classification and block processing on the data to be transmitted according to the disturbance response matrix to obtain a classified data block group; Perform parameter channel matching analysis according to the disturbance response matrix to obtain the channel matching degree; generate a channel preference matching strategy for the classified data block group according to the channel matching degree to obtain a channel allocation strategy table; Perform micro-disturbance parameter modulation on the classified data block group according to the channel allocation strategy table to obtain a modulation parameter sequence.

5. The data information secure transmission method according to claim 1, wherein Step S3 includes the following steps: Step S31: Perform channel correlation analysis on the modulation parameter sequence to obtain a channel characteristic evaluation table; Step S32: Perform time interleaving analysis according to the modulation parameter sequence and the channel characteristic evaluation table to obtain a time-interleaved data stream; Step S33: Obtain the power grid topology information; optimize the spatial distribution strategy according to the time-interleaved data stream and the power grid topology information to obtain a spatial distribution scheme; Step S34: Perform spatio-temporal fusion coding processing on the time-interleaved data stream according to the spatial distribution scheme to obtain spatio-temporal fusion coding; Step S35: Establish an electrical parameter wave impedance model according to the power grid topology information, extract the main electrical transmission path through eigenvalue decomposition to obtain a main path set; perform physical transmission feasibility evaluation according to the main path set and the channel characteristic evaluation table to obtain a physically enhanced channel evaluation table; perform dynamic channel transmission configuration on the spatio-temporal fusion coding according to the physically enhanced channel evaluation table to obtain a spatio-temporal interleaved code stream with electrical and physical robustness.

6. The data information secure transmission method according to claim 5, wherein Step S34 includes: Extract the number of time-interleaved units in the time-interleaved data stream, extract the number of spatial nodes participating in transmission in the spatial distribution scheme, and construct a spatio-temporal matrix framework according to the number of time-interleaved units and the number of spatial nodes; Perform data block mapping and filling according to the spatio-temporal matrix framework to obtain an initial spatio-temporal matrix; Divide the pre-generated interleaving key into multiple shards, and the number of shards is equal to the number of spatial nodes to obtain key shards; Embed the key shards into the initial spatio-temporal matrix to obtain a spatio-temporal matrix containing the key; Perform time dimension expansion on the spatio-temporal matrix containing the key, and perform spatial dimension expansion to obtain an expanded spatio-temporal matrix; Add synchronization marks to the expanded spatio-temporal matrix to obtain a marked spatio-temporal matrix; Perform spatio-temporal mixing degree evaluation on the marked spatio-temporal matrix to obtain spatio-temporal fusion coding.

7. The data information secure transmission method according to claim 1, wherein The spatio-temporal de-interleaving process described in step S4 includes: The receiving end simultaneously monitors multiple physical communication channels, synchronously receives the spatio-temporal interleaved code stream to obtain a set of synchronous data frames; Perform spatio-temporal de-interleaving reconstruction on the set of synchronous data frames to obtain a reconstructed modulation sequence; Obtain an updated disturbance response matrix; perform channel demodulation extraction on the reconstructed modulation sequence according to the updated disturbance response matrix to obtain a set of channel data blocks.

8. The data information secure transmission method according to claim 1, wherein The multi-point multi-dimensional electrical parameter verification described in step S4 includes: The receiving end measures the local power grid parameters to obtain a local parameter characteristic code vector; Perform syntax layer verification on the set of channel data blocks according to the local parameter characteristic code vector to obtain a primary comparison result; Perform range rationality check on the primary verification result and time coherence check to obtain the secondary comparison result; Perform power flow constraint check on the secondary comparison result and state estimation verification to obtain the physically constrained state data; Perform multi-point data cross-verification based on the physically constrained state data to obtain the tertiary comparison result; Perform comprehensive credibility evaluation based on the primary comparison result, secondary comparison result and tertiary comparison result to obtain the data block credibility set; Generate an integrity analysis report according to the data block credibility level to obtain the data integrity report.

9. The data information secure transmission method according to claim 8, wherein, The multi-point data cross-verification includes: Construct a node adjacency matrix according to the power grid topology information, perform electrical distance threshold comparison based on the node adjacency matrix and the physically constrained state data, and perform node pair correlation scoring to obtain the associated node pair list and node pair correlation score; Perform time series alignment of the same type of parameters on the physically constrained state data according to the associated node pair list to obtain the same type of parameter pair sequence; Calculate the correlation coefficient according to the same type of parameter pair sequence and the node pair correlation score to obtain the weighted correlation coefficient matrix; Set the deviation threshold according to the weighted correlation coefficient matrix to obtain the dynamic deviation threshold table; Mark abnormal data according to the dynamic deviation threshold table and the weighted correlation coefficient matrix to obtain the tertiary comparison result.

10. A data information secure transmission system, characterized in that For executing the data information security transmission method as claimed in claim 1, the data information security transmission system includes: A multi-dimensional parameter encoding module, configured to collect the original power grid measurement data; perform differential quantization and encoding on the original power grid measurement data to obtain a parameter feature code vector; An information channel modulation module, configured to construct a power grid disturbance response matrix according to the parameter feature code vector to obtain a disturbance response matrix; obtain the data to be transmitted; perform information channel modulation on the data to be transmitted according to the disturbance response matrix to obtain a modulation parameter sequence; A multi-channel fusion transmission module, configured to perform time interleaving analysis according to the modulation parameter sequence to obtain a time interleaved data stream; obtain the power grid topology information; perform spatio-temporal fusion coding transmission according to the power grid topology information and the time interleaved data stream to obtain a spatio-temporal interleaved code stream; A receiving end demodulation and decoding module, configured to perform spatio-temporal de-interleaving processing on the spatio-temporal interleaved code stream to obtain a channel data block set; perform multi-point multi-dimensional electrical parameter verification on the channel data block set to obtain a data integrity report; perform adaptive data reorganization and recovery on the channel data block set according to the data integrity report to obtain a verification result flag, so as to implement the power grid data information security transmission task.

Images