Intelligent power-off control method and system for high-voltage transmission induced electricity monitoring data fusion

Abstract

The invention discloses an intelligent power-off control method and system for high-voltage transmission induced electricity monitoring data fusion, and belongs to the technical field of power system safety protection. The method comprises the following steps: constructing a heterogeneous data set through a multi-dimensional induced electricity monitoring array; extracting an induced electric intensity vector and an environment parameter vector based on the heterogeneous data set, and determining a first risk level based on the induced electric intensity vector and the environment parameter vector; constructing a dynamic Bayesian risk network according to the induced electric strength vector and the environmental parameter vector, and performing real-time risk assessment to obtain a second risk level; and performing weighted integration on the first risk level and the second risk level to obtain an integrated risk level, and determining a protection action of the circuit breaker based on the integrated risk level. According to the method, fusion of the multi-source data in the feature layer and the decision layer is realized, the depth and the real-time performance of data fusion are improved, and the problem of low fusion degree of the multi-source data in the prior art is solved.

Description

Technical Field

[0001] The present invention belongs to the technical field of power system safety protection, and in particular relates to an intelligent power-off control method and system for fusing high-voltage transmission induction power monitoring data. Background Art

[0002] The demand for electric energy is growing. As critical infrastructure for long-distance, high-capacity power transmission, high-voltage transmission lines are expanding in coverage and becoming increasingly widespread. During operation, high-voltage transmission lines generate induced electricity. This phenomenon poses significant safety risks to on-site workers in many practical scenarios, such as construction and routine maintenance around the lines. To minimize this risk and ensure personnel safety and the normal operation of transmission lines, effective monitoring and control of induced electricity in high-voltage transmission lines has become a critical issue receiving significant attention in the power industry.

[0003] Existing monitoring and control technologies for high-voltage transmission induced electricity often focus on the acquisition and utilization of a single type of data. Due to the relative independence of each monitoring method, the acquired data lacks effective integration and deep fusion. Most technologies analyze a specific type of data in isolation to determine induced electricity conditions, ignoring the correlations and interactions between different data sources. This results in a low degree of multi-source data integration. Summary of the Invention

[0004] In view of the above-mentioned problems, the purpose of the present invention is to provide an intelligent power outage control method and system for the fusion of high-voltage transmission inductive power monitoring data, which realizes the fusion of multi-source data at the feature layer and decision layer, improves the depth and real-time performance of data fusion, and solves the problem of low multi-source data fusion in the existing technology.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions: An intelligent power outage control method based on high-voltage transmission induction power monitoring data fusion includes the following steps: A multi-dimensional induction electrical monitoring array is used to synchronously collect the conductor's axial electric field strength, radial magnetic field component, surface potential gradient, and environmental parameters to construct a heterogeneous data set. Environmental parameters include temperature, humidity, air pressure, light intensity, and wind speed. extracting an induced electric intensity vector and an environmental parameter vector based on the heterogeneous data set, and determining a first risk level based on the induced electric intensity vector and the environmental parameter vector; A dynamic Bayesian risk network is constructed based on the induced electric intensity vector and the environmental parameter vector to conduct real-time risk assessment and derive the second risk level; The first risk level and the second risk level are weighted and integrated to obtain a comprehensive risk level, and a protective action of the circuit breaker is determined based on the comprehensive risk level.

[0006] Preferably, the multi-dimensional induction electrical monitoring array includes a distributed eddy electric field sensor group, a ring fluxgate array, a nanocomposite dielectric sensor and an integrated environmental sensor: The distributed eddy electric field sensor group measures the axial electric field strength of the conductor at each time point; The annular fluxgate array captures the radial magnetic field component at each time point; Nanocomposite dielectric sensors to obtain surface potential gradients; Integrated environmental sensors acquire environmental parameters at each time point; The conductor axial electric field intensity, radial magnetic field component, surface potential and environmental parameters are standardized and merged into a heterogeneous dataset.

[0007] Based on distributed eddy electric field sensor groups, annular fluxgate arrays and other equipment, accurate and synchronous acquisition of multiple physical quantities (electric field intensity, magnetic field components, etc.) is achieved; format standardization processing solves the problem of inconsistent multi-source data structures, laying the foundation for subsequent fusion.

[0008] Preferably, extracting the induced electric intensity vector and the environmental parameter vector based on the heterogeneous data set includes the following steps: For the axial electric field strength of the conductor at each time point, the maximum value is taken as the electric field characteristic; For the radial magnetic field component at each time point, the peak value is taken as the magnetic field feature; For the surface potential gradient, the maximum value is taken as the potential characteristic; The induced electric intensity vector is obtained by combining the electric field characteristics, magnetic field characteristics and potential characteristics; For each time point, the temperature, humidity, air pressure, light intensity and wind speed are arranged in order to form an environmental parameter vector.

[0009] Key features (such as electric field maximum value and magnetic field peak value) are extracted from the original data, and complex multi-dimensional data are compressed into induced electric intensity vectors and environmental parameter vectors, reducing data complexity and facilitating subsequent risk level calculations and network model processing.

[0010] Preferably, determining the first risk level based on the induced electric intensity vector and the environmental parameter vector comprises the following steps: Obtaining the electric intensity vector-electric intensity risk level mapping table stored in the database, and obtaining the electric intensity risk level based on the current induced electric intensity vector comparison mapping; Obtaining a reference environmental parameter vector, processing the environmental parameter vector with the reference environmental parameter vector to obtain an environmental parameter processing coefficient; Obtaining an environmental parameter processing coefficient-environmental risk level mapping table stored in a database, comparing and mapping the current environmental parameter processing coefficient with the environmental parameter processing coefficient-environmental risk level mapping table, and determining the environmental risk level; A weighted sum of the electrical intensity risk level and the environmental risk level is performed to obtain a first risk level.

[0011] Preferably, obtaining the environmental parameter processing coefficient comprises the following steps: Get the environmental parameter vector of the t time points before the current moment; Calculate the ratios of the parameters in the environmental parameter vector at t time points to the parameters in the baseline environmental parameter vector, and obtain t ratio calculation vectors; Perform weighted summation on the ratios in the t ratio calculation vectors respectively to obtain t ratio results; The t ratio results are averaged to obtain the environmental parameter treatment coefficient.

[0012] Through mapping tables and weighted summation, the induced electrical intensity and environmental parameters are converted into comparable risk levels, realizing the quantitative fusion of multi-source data at the feature layer. Through dynamic comparison of historical environmental parameters with baseline values (ratio calculation, weighted summation, and mean processing), the environmental change characteristics in the time dimension are introduced to enhance the dynamic nature of data fusion.

[0013] Preferably, a dynamic Bayesian risk network is constructed based on the induced electric intensity vector and the environmental parameter vector to perform real-time risk assessment and obtain the second risk level. The process is as follows: Set the hidden variables to the second risk level, the observable variables to the induced electric intensity vector and the environmental parameter vector, and construct the time series dependency relationship between the hidden variables and the observable variables; determining an initial probability distribution for the second risk level; The induced electric intensity vector and environmental parameter vector at the current moment are used as network inputs. Based on the statistical variation law of the second risk level, the transition probability from the previous moment to the current moment is determined. Based on the correlation statistics of observable variables and hidden variables, the observation probability of observing the induced electric intensity vector and the environmental parameter vector at the current moment under the set second risk level is determined; The posterior probability is determined based on recursive calculation, and the second risk level with the highest posterior probability is selected as the current result.

[0014] Preferably, determining the protection action of the circuit breaker based on the comprehensive risk level includes the following steps: Get the risk level decision set stored in the database ,in, is the first-level response threshold, is the secondary response threshold; Determine whether the current comprehensive risk level belongs to the risk level decision set If yes, the reactor impedance adjustment protection and dynamic reactive power compensation are triggered; If the current comprehensive risk level is greater than , then the circuit breaker sequence action protection is started; If the current comprehensive risk level is less than or equal to , only the optical alarm signal is triggered.

[0015] Preferably, the protection and dynamic reactive power compensation are triggered when the reactor impedance is adjusted, and the adjusted impedance value and the dynamic reactive power compensation adjustment amount are obtained: ; in, To adjust the impedance, is the reference impedance, To adjust the gain coefficient, is the current comprehensive risk level, is the phase compensation factor, is the temperature drift, is the system angular frequency; ; in, is the dynamic reactive compensation adjustment amount, is the maximum capacity of the reactive compensation device, is the Gaussian error function, is the sensitivity coefficient.

[0016] Preferably, when the circuit breaker sequence action protection is started, the circuit breaker action delay is determined: ; in, is the circuit breaker action delay, is the reference delay constant, e is the natural constant, is the attenuation factor.

[0017] An intelligent power outage control system integrating high-voltage transmission induction power monitoring data, used to implement the above method, includes: The data acquisition module is used to construct a heterogeneous data set by synchronously collecting the conductor's axial electric field strength, radial magnetic field component, surface potential gradient, and environmental parameters through a multi-dimensional induction electrical monitoring array. The environmental parameters include temperature, humidity, air pressure, light intensity, and wind speed. a first risk level determination module, configured to extract an induced electric intensity vector and an environmental parameter vector based on a heterogeneous data set, and determine a first risk level based on the induced electric intensity vector and the environmental parameter vector; A second risk level determination module is used to construct a dynamic Bayesian risk network based on the induced electric intensity vector and the environmental parameter vector, perform real-time risk assessment, and obtain a second risk level; The multi-level protection module performs weighted integration of the first risk level and the second risk level to obtain a comprehensive risk level, and determines the protection action of the circuit breaker based on the comprehensive risk level.

[0018] The beneficial effects of the present invention are: The present invention collects multi-source data through a multi-dimensional induction electrical monitoring array to construct a heterogeneous data set, combines feature extraction to determine the first risk level, and uses a dynamic Bayesian risk network to conduct real-time evaluation to derive the second risk level. Weighted synthesis is used to obtain a comprehensive risk level to achieve intelligent control of circuit breaker protection actions. This realizes the fusion of multi-source data at the feature layer and the decision layer, improves the depth and real-time performance of data fusion, and solves the problem of low multi-source data fusion in the existing technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 Schematic diagram of the method in Example 1; Figure 2 This is a schematic diagram of the structure of the system in Example 2. DETAILED DESCRIPTION

[0020] The above technical solution is described in detail below with reference to the accompanying drawings and specific implementation methods.

[0021] Example 1: Figure 1 As shown, the intelligent power outage control method based on the fusion of high-voltage transmission induction power monitoring data includes the following steps: A multi-dimensional induction electrical monitoring array is used to synchronously collect the conductor's axial electric field strength, radial magnetic field component, surface potential gradient, and environmental parameters to construct a heterogeneous data set. Environmental parameters include temperature, humidity, air pressure, light intensity, and wind speed. extracting an induced electric intensity vector and an environmental parameter vector based on the heterogeneous data set, and determining a first risk level based on the induced electric intensity vector and the environmental parameter vector; A dynamic Bayesian risk network is constructed based on the induced electric intensity vector and the environmental parameter vector to conduct real-time risk assessment and derive the second risk level; The first risk level and the second risk level are weighted and integrated to obtain a comprehensive risk level, and a protective action of the circuit breaker is determined based on the comprehensive risk level.

[0022] The multi-dimensional inductive electrical monitoring array includes a distributed eddy electric field sensor group, a ring fluxgate array, a nanocomposite dielectric sensor, and an integrated environmental sensor: The distributed eddy electric field sensor group measures the axial electric field strength of the conductor at each time point; The annular fluxgate array captures the radial magnetic field component at each time point; Nanocomposite dielectric sensors to obtain surface potential gradients; Integrated environmental sensors acquire environmental parameters at each time point; The conductor axial electric field intensity, radial magnetic field component, surface potential and environmental parameters are standardized and merged into a heterogeneous dataset.

[0023] Extracting the induced electric intensity vector and environmental parameter vector based on heterogeneous data sets includes the following steps: For the axial electric field strength of the conductor at each time point, the maximum value is taken as the electric field characteristic; For the radial magnetic field component at each time point, the peak value is taken as the magnetic field feature; For the surface potential gradient, the maximum value is taken as the potential characteristic; The induced electric intensity vector is obtained by combining the electric field characteristics, magnetic field characteristics and potential characteristics; For each time point, the temperature, humidity, air pressure, light intensity and wind speed are arranged in order to form an environmental parameter vector.

[0024] Determining a first risk level based on the induced electric intensity vector and the environmental parameter vector includes the following steps: Obtaining the electric intensity vector-electric intensity risk level mapping table stored in the database, and obtaining the electric intensity risk level based on the current induced electric intensity vector comparison mapping; Obtaining a reference environmental parameter vector, processing the environmental parameter vector with the reference environmental parameter vector to obtain an environmental parameter processing coefficient; Obtaining an environmental parameter processing coefficient-environmental risk level mapping table stored in a database, comparing and mapping the current environmental parameter processing coefficient with the environmental parameter processing coefficient-environmental risk level mapping table, and determining the environmental risk level; A weighted sum of the electrical intensity risk level and the environmental risk level is performed to obtain a first risk level.

[0025] Obtaining the environmental parameter processing coefficient includes the following steps: Get the environmental parameter vector of the t time points before the current moment; Calculate the ratios of the parameters in the environmental parameter vector at t time points to the parameters in the baseline environmental parameter vector, and obtain t ratio calculation vectors; Perform weighted summation on the ratios in the t ratio calculation vectors respectively to obtain t ratio results; The t ratio results are averaged to obtain the environmental parameter treatment coefficient.

[0026] A dynamic Bayesian risk network is constructed based on the induced electric intensity vector and the environmental parameter vector to perform real-time risk assessment and derive the second risk level. The process is as follows: Set the hidden variables to the second risk level, the observable variables to the induced electric intensity vector and the environmental parameter vector, and construct the time series dependency relationship between the hidden variables and the observable variables; determining an initial probability distribution for the second risk level; The induced electric intensity vector and environmental parameter vector at the current moment are used as network inputs. Based on the statistical variation law of the second risk level, the transition probability from the previous moment to the current moment is determined. Based on the correlation statistics of observable variables and hidden variables, the observation probability of observing the induced electric intensity vector and the environmental parameter vector at the current moment under the set second risk level is determined; The posterior probability is determined based on recursive calculation, and the second risk level with the highest posterior probability is selected as the current result.

[0027] By leveraging the probabilistic reasoning capabilities of Bayesian networks, we model the time series dependency between hidden variables (risk level) and observable variables (induced electrical intensity, environmental parameters), capture the temporal correlation and causal relationship between data, and improve the real-time and accuracy of fusion.

[0028] Determining the protective action of the circuit breaker based on the comprehensive risk level includes the following steps: Get the risk level decision set stored in the database ,in, is the first-level response threshold, is the secondary response threshold; Determine whether the current comprehensive risk level belongs to the risk level decision set If yes, the reactor impedance adjustment protection and dynamic reactive power compensation are triggered; If the current comprehensive risk level is greater than , then the circuit breaker sequence action protection is started; If the current comprehensive risk level is less than or equal to , only the optical alarm signal is triggered.

[0029] Through the risk level decision set (threshold division), the fused comprehensive risk is converted into the graded actions of the circuit breaker (alarm, impedance adjustment, power off), realizing the practical application of the data fusion results.

[0030] When the reactor impedance adjustment is triggered, protection and dynamic reactive compensation are performed to obtain the adjusted impedance value and dynamic reactive compensation adjustment amount: ; in, To adjust the impedance, is the reference impedance, To adjust the gain coefficient, is the current comprehensive risk level, is the phase compensation factor, is the temperature drift, is the system angular frequency; ; in, is the dynamic reactive compensation adjustment amount, is the maximum capacity of the reactive compensation device, is the Gaussian error function, is the sensitivity coefficient.

[0031] When starting the circuit breaker sequence action protection, determine the circuit breaker action delay: ; in, is the circuit breaker action delay, is the reference delay constant, e is the natural constant, is the attenuation factor.

[0032] By quantifying protection action parameters through mathematical models (such as impedance adjustment formulas and time delay calculations), the fused risk level is accurately matched with the control strategy, improving the scientific nature of the system response.

[0033] The formulas involved in this embodiment can be dimensioned during calculation to simplify the calculation.

[0034] Example 2: This example is further improved on the basis of Example 1, specifically including: Building smart sensor networks based on wireless communication: A sensor network is constructed using multi-hop wireless communication technology, connecting multiple different types of sensor nodes (including distributed eddy electric field sensors, annular fluxgate sensors, nanocomposite dielectric sensors, and integrated environmental sensors) via wireless links. Each sensor node not only collects its own monitoring data but also acts as a data relay node, forwarding data from other nodes to the aggregation node, forming a multi-hop wireless sensor network.

[0035] This architectural design can expand the coverage of the sensor network, reduce the transmission distance of individual sensor nodes, lower energy consumption, and improve data transmission reliability. By rationally planning the network topology, the sensor network can ensure extensive coverage along high-voltage transmission lines, enabling comprehensive monitoring of induced electricity and related environmental parameters.

[0036] At the same time, each sensor node integrates a microprocessor, memory, and wireless communication module, giving it a certain level of data processing and storage capabilities. The sensor node can perform preliminary processing on the collected data, such as data filtering and feature extraction, to reduce the amount of data transmitted, and store the processed data in local memory for query and analysis when needed.

[0037] Dynamic collaboration and data fusion of sensor networks: During the monitoring process, sensor nodes dynamically collaborate based on pre-set collaboration rules. For example, when a sensor node detects an abnormal change in induced electrical intensity or environmental parameters, it sends a collaboration request to its neighboring nodes. Upon receiving the request, the neighboring nodes increase their sampling frequency and focus on changes in the relevant parameters, jointly conducting intensive monitoring of the area.

[0038] This collaborative monitoring mechanism can quickly capture early signs of induced electrocution risks, improving monitoring sensitivity and accuracy. Furthermore, data from multiple sensor nodes complement and verify each other, effectively reducing the uncertainty of individual node data and providing a more reliable foundation for subsequent data fusion.

[0039] Furthermore, a distributed data fusion algorithm is employed within the sensor network. Each sensor node performs local data fusion processing based on its own collected data and data received from neighboring nodes. This fusion algorithm comprehensively considers factors such as data relevance, reliability, and weighting, fusing multi-source data into more representative and accurate information. This distributed data fusion algorithm not only fully leverages the distributed nature of the sensor network, improving data processing efficiency and real-time performance, but also reduces information loss and error accumulation during data transmission. The fused data more accurately reflects the actual conditions of induced static electricity around high-voltage transmission lines, providing a more precise basis for risk assessment.

[0040] Data-driven sensor network optimization: A data quality assessment model is established to perform real-time quality assessments on data collected by the sensor network. Evaluation metrics include data accuracy, completeness, timeliness, and consistency. Based on the assessment results, feedback and adjustments are provided to sensor nodes with poor data quality, such as sensor calibration and communication link repair, to ensure that the data output from the sensor network meets quality requirements.

[0041] Through data quality assessment and feedback mechanisms, data quality issues in sensor networks can be promptly identified and resolved, improving the overall performance and reliability of the system. Furthermore, continuous monitoring and optimization of data quality can further enhance the effectiveness of data fusion and the accuracy of risk assessment.

[0042] Furthermore, machine learning algorithms analyze and mine sensor network operational data, automatically learning the performance characteristics and data distribution patterns of the sensor network. Based on these learning results, sensor network parameters (such as sampling frequency, transmission power, and routing strategies) are dynamically optimized to improve the network's energy efficiency, data transmission efficiency, and monitoring performance. For example, by analyzing historical data and network operating status, machine learning algorithms can predict the optimal sampling frequency and transmission power settings for the sensor network under different time periods and environmental conditions. This maximizes the lifespan of the sensor network while ensuring effective monitoring, reducing system maintenance costs.

[0043] Example 3: Figure 2 As shown, the intelligent power outage control system for high-voltage transmission induction power monitoring data fusion is used to implement the method in Example 1, including: The data acquisition module is used to construct a heterogeneous data set by synchronously collecting the conductor's axial electric field strength, radial magnetic field component, surface potential gradient, and environmental parameters through a multi-dimensional induction electrical monitoring array. The environmental parameters include temperature, humidity, air pressure, light intensity, and wind speed. a first risk level determination module, configured to extract an induced electric intensity vector and an environmental parameter vector based on a heterogeneous data set, and determine a first risk level based on the induced electric intensity vector and the environmental parameter vector; A second risk level determination module is used to construct a dynamic Bayesian risk network based on the induced electric intensity vector and the environmental parameter vector, perform real-time risk assessment, and obtain a second risk level; The multi-level protection module performs weighted integration of the first risk level and the second risk level to obtain a comprehensive risk level, and determines the protection action of the circuit breaker based on the comprehensive risk level.

Claims

1. An intelligent power outage control method based on the fusion of high-voltage transmission induction power monitoring data, characterized in that: The following steps are involved: A multi-dimensional induction electrical monitoring array is used to synchronously collect the conductor's axial electric field strength, radial magnetic field component, surface potential gradient, and environmental parameters to construct a heterogeneous data set. Environmental parameters include temperature, humidity, air pressure, light intensity, and wind speed. extracting an induced electric intensity vector and an environmental parameter vector based on the heterogeneous data set, and determining a first risk level based on the induced electric intensity vector and the environmental parameter vector; A dynamic Bayesian risk network is constructed based on the induced electric intensity vector and the environmental parameter vector to conduct real-time risk assessment and derive the second risk level; The first risk level and the second risk level are weighted and integrated to obtain a comprehensive risk level, and a protective action of the circuit breaker is determined based on the comprehensive risk level.

2. The intelligent power outage control method based on high-voltage transmission induction power monitoring data fusion as claimed in claim 1 is characterized in that: The multi-dimensional inductive electrical monitoring array includes a distributed eddy electric field sensor group, a ring fluxgate array, a nanocomposite dielectric sensor, and an integrated environmental sensor: The distributed eddy electric field sensor group measures the axial electric field strength of the conductor at each time point; The annular fluxgate array captures the radial magnetic field component at each time point; Nanocomposite dielectric sensors to obtain surface potential gradients; Integrated environmental sensors acquire environmental parameters at each time point; The conductor axial electric field intensity, radial magnetic field component, surface potential and environmental parameters are standardized and merged into a heterogeneous dataset.

3. The intelligent power outage control method based on high-voltage transmission induction power monitoring data fusion as claimed in claim 2 is characterized in that: Extracting the induced electric intensity vector and environmental parameter vector based on heterogeneous data sets includes the following steps: For the axial electric field strength of the conductor at each time point, the maximum value is taken as the electric field characteristic; For the radial magnetic field component at each time point, the peak value is taken as the magnetic field feature; For the surface potential gradient, the maximum value is taken as the potential characteristic; The induced electric intensity vector is obtained by combining the electric field characteristics, magnetic field characteristics and potential characteristics; For each time point, the temperature, humidity, air pressure, light intensity and wind speed are arranged in order to form an environmental parameter vector.

4. The intelligent power outage control method based on high-voltage transmission induction power monitoring data fusion as claimed in claim 3 is characterized in that: Determining a first risk level based on the induced electric intensity vector and the environmental parameter vector includes the following steps: Obtaining the electric intensity vector-electric intensity risk level mapping table stored in the database, and obtaining the electric intensity risk level based on the current induced electric intensity vector comparison mapping; Obtaining a reference environmental parameter vector, processing the environmental parameter vector with the reference environmental parameter vector to obtain an environmental parameter processing coefficient; Obtaining an environmental parameter processing coefficient-environmental risk level mapping table stored in a database, comparing and mapping the current environmental parameter processing coefficient with the environmental parameter processing coefficient-environmental risk level mapping table, and determining the environmental risk level; A weighted sum of the electrical intensity risk level and the environmental risk level is performed to obtain a first risk level.

5. The intelligent power outage control method based on high-voltage transmission induction power monitoring data fusion as claimed in claim 1 is characterized in that: Obtaining the environmental parameter processing coefficient includes the following steps: Get the environmental parameter vector of the t time points before the current moment; Calculate the ratios of the parameters in the environmental parameter vector at t time points to the parameters in the baseline environmental parameter vector, and obtain t ratio calculation vectors; Perform weighted summation on the ratios in the t ratio calculation vectors respectively to obtain t ratio results; The t ratio results are averaged to obtain the environmental parameter treatment coefficient.

6. The intelligent power outage control method based on high-voltage transmission induction power monitoring data fusion as claimed in claim 1, characterized in that: A dynamic Bayesian risk network is constructed based on the induced electric intensity vector and the environmental parameter vector to perform real-time risk assessment and derive the second risk level. The process is as follows: Set the hidden variables to the second risk level, the observable variables to the induced electric intensity vector and the environmental parameter vector, and construct the time series dependency relationship between the hidden variables and the observable variables; determining an initial probability distribution for the second risk level; The induced electric intensity vector and environmental parameter vector at the current moment are used as network inputs. Based on the statistical variation law of the second risk level, the transition probability from the previous moment to the current moment is determined. Based on the correlation statistics of observable variables and hidden variables, the observation probability of observing the induced electric intensity vector and the environmental parameter vector at the current moment under the set second risk level is determined; The posterior probability is determined based on recursive calculation, and the second risk level with the highest posterior probability is selected as the current result.

7. The intelligent power outage control method based on high-voltage transmission induction power monitoring data fusion as claimed in claim 1, characterized in that: Determining the protective action of the circuit breaker based on the comprehensive risk level includes the following steps: Get the risk level decision set stored in the database ,in, is the first-level response threshold, is the secondary response threshold; Determine whether the current comprehensive risk level belongs to the risk level decision set If yes, the reactor impedance adjustment protection and dynamic reactive power compensation are triggered; If the current comprehensive risk level is greater than , then the circuit breaker sequence action protection is started; If the current comprehensive risk level is less than or equal to , only the optical alarm signal is triggered.

8. The intelligent power outage control method based on high-voltage transmission induction power monitoring data fusion as claimed in claim 7, characterized in that: When the reactor impedance adjustment is triggered, protection and dynamic reactive compensation are performed to obtain the adjusted impedance value and dynamic reactive compensation adjustment amount: ； in, To adjust the impedance, is the reference impedance, To adjust the gain coefficient, is the current comprehensive risk level, is the phase compensation factor, is the temperature drift, is the system angular frequency; ； in, is the dynamic reactive compensation adjustment amount, is the maximum capacity of the reactive compensation device, is the Gaussian error function, is the sensitivity coefficient.

9. The intelligent power outage control method based on high-voltage transmission induction power monitoring data fusion as claimed in claim 7, characterized in that: When starting the circuit breaker sequence action protection, determine the circuit breaker action delay: ； in, is the circuit breaker action delay, is the reference delay constant, e is the natural constant, is the attenuation factor.

10. An intelligent power outage control system based on the fusion of high-voltage transmission induction power monitoring data, used to implement the method according to any one of claims 1 to 9, characterized in that: include: The data acquisition module is used to construct a heterogeneous data set by synchronously collecting the conductor's axial electric field strength, radial magnetic field component, surface potential gradient, and environmental parameters through a multi-dimensional induction electrical monitoring array. The environmental parameters include temperature, humidity, air pressure, light intensity, and wind speed. a first risk level determination module, configured to extract an induced electric intensity vector and an environmental parameter vector based on a heterogeneous data set, and determine a first risk level based on the induced electric intensity vector and the environmental parameter vector; A second risk level determination module is used to construct a dynamic Bayesian risk network based on the induced electric intensity vector and the environmental parameter vector, perform real-time risk assessment, and obtain a second risk level; The multi-level protection module performs weighted integration of the first risk level and the second risk level to obtain a comprehensive risk level, and determines the protection action of the circuit breaker based on the comprehensive risk level.

Images