Model and device for predicting influence of meteorological event on power distribution network, and medium

Abstract

The invention discloses a prediction model for the influence of meteorological events on a power distribution network, equipment and a medium. The prediction model is constructed by the following steps: S1, obtaining historical meteorological events and historical power distribution network fault data information, and carrying out data preprocessing; s2, according to the historical meteorological events, the historical power distribution network fault data information and the current meteorological information, establishing a prediction model of the influence of the meteorological events on the power distribution network, and training the prediction model of the influence of the meteorological events on the power distribution network; s3, inputting the historical meteorological event, the historical power distribution network fault data information and the current meteorological event into the trained prediction model of the influence of the meteorological event on the power distribution network, and evaluating the operation state of the power distribution network; the prediction result of the prediction model serves as the basis of power distribution network fault occurrence, related equipment is deployed at the corresponding position in advance, the influence of meteorological events on the power distribution network is avoided, and stable transmission of power is guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of distribution network failure, and in particular to a prediction model, equipment and medium for the impact of meteorological events on distribution networks. Background Art

[0002] When people burn fossil fuels, such as oil and coal, or cut down forests and burn them, they produce a lot of carbon dioxide, that is, greenhouse gases. These greenhouse gases are highly transparent to visible light from solar radiation, but highly absorptive to long-wave radiation emitted by the earth. They can strongly absorb infrared radiation from the ground, causing the earth's temperature to rise. With the exploitation and use of the earth's resources by humans, the greenhouse effect continues to accumulate, resulting in an imbalance in the energy absorbed and emitted by the earth-atmosphere system. Energy continues to accumulate in the earth-atmosphere system, resulting in rising temperatures and global warming. With the changes in the global climate, extreme weather events occur frequently. As the main component of power transmission, the distribution network plays an important role in people's production and life, and is widely used in daily life and work. Therefore, the importance of safe and stable operation of the distribution network is self-evident.

[0003] However, the extreme weather environment is harsh, and the safe operation of the distribution network is seriously threatened. Possible extreme weather conditions include strong winds, heavy rains, blizzards, etc. The frequent occurrence of extreme weather leads to mountain torrents, landslides, etc., posing a threat to the safety of people's lives in the corresponding areas. At the same time, extreme weather conditions pose a serious threat to the safe and stable operation of the distribution network.

[0004] The traditional distribution network operation and maintenance method is to receive a fault notification and then handle it when a distribution network failure occurs. This method has low emergency repair efficiency in areas where faults frequently occur. At the same time, there is a shortage of personnel and the problem of inability to carry out timely emergency repairs, which has a serious impact on the lives of people in the relevant areas. The traditional distribution network operation and maintenance method is difficult to effectively respond to sudden meteorological disasters, resulting in frequent power outages and affecting social production and life.

[0005] Therefore, developing a method that can predict the impact of meteorological disasters on distribution networks in advance and carry out reinforcement and prevention at corresponding locations in advance will be of great benefit to ensuring the stable operation of distribution networks. Summary of the invention

[0006] In order to overcome the above problems, the purpose of the present invention is to provide a prediction model, device and medium for the impact of meteorological events on distribution networks. The prediction model of meteorological events on distribution networks is based on historical meteorological events and immediate distribution network fault data to build a prediction model, and a multi-layer recurrent neural network is used to train the prediction model to improve the accuracy of the prediction results of the prediction model. The prediction results of the prediction model are used as the basis for the occurrence of distribution network faults, and related equipment is deployed at corresponding locations in advance to avoid the impact of meteorological events on the distribution network and ensure the stable transmission of electricity.

[0007] The technical solution adopted by the present invention is:

[0008] A prediction model for the impact of meteorological events on a distribution network, wherein the steps for constructing the prediction model are as follows:

[0009] S1: Obtain historical meteorological events and historical distribution network fault data information and perform data preprocessing;

[0010] S2: establishing a prediction model for the impact of meteorological events on the distribution network based on the historical meteorological events, historical distribution network fault data information and current meteorological information, training the prediction model for the impact of meteorological events on the distribution network, and obtaining a trained prediction model for the impact of meteorological events on the distribution network;

[0011] S3: Re-acquire historical meteorological events and historical distribution network fault data information, input the historical meteorological events, historical distribution network fault data information and current meteorological events into the trained prediction model of the impact of meteorological events on the distribution network, obtain the time series data of the current distribution network, and evaluate the operation status of the distribution network according to the time series data of the current distribution network;

[0012] In step S2, a prediction model for the impact of meteorological events on the distribution network is established. The prediction model uses a multi-layer recurrent neural network, including an input layer, a hidden layer, and an output layer. The specific steps are as follows:

[0013] S21: The input layer receives historical meteorological events and historical power distribution network failure data information after data preprocessing, wherein the preprocessed data is characteristic sequence data of meteorological events and power distribution network failures;

[0014] S22: There are two hidden layers, the maximum number of iterations is 200, the activation function is the tanh function, and the number of neurons in each hidden layer is the same.

[0015] S23: Use the back propagation algorithm over time to update the training set preprocessed by S21 to minimize the loss function, so that the prediction result of the model is infinitely close to the actual output;

[0016] S24: using the model trained in S23 to predict the distribution network fault according to the current actual meteorological events, and obtaining the prediction result of the model at the prediction time;

[0017] S25: Perform autocorrelation analysis on historical data of distribution network fault parameters to obtain an appropriate number of evidences.

[0018] As a further description of the present invention, the historical meteorological events in step S1 are natural disaster events, including earthquakes, rainstorms, floods, and thunderstorms;

[0019] The historical distribution network faults include cable abnormal point faults and substation faults. The data information of the cable abnormal point faults includes cable construction time, abnormal point cross-sectional size, and abnormal point environmental parameters. The substation faults include voltage transformer faults, current transformer faults, DC system grounding faults, busbar faults, and circuit breaker faults.

[0020] The specific steps of data preprocessing in step S1 are:

[0021] S11: Data cleaning,

[0022] S12: Noise data removal;

[0023] S13: Data standardization: the data after steps S11 and S12 are used to generate feature sets according to meteorological events and distribution network faults, and further generate feature sequences.

[0024] As a further description of the present invention, the specific process of the back propagation algorithm in step S23 is:

[0025] S231: forward calculation of the output value of each neuron;

[0026] S232: Reversely calculate the error term value of each neuron, which is the partial derivative of the loss function with respect to the weighted input of the neuron;

[0027] S233: Calculate the gradient of each weight, and then use the batch gradient descent algorithm to update the weight, and determine whether the loss function converges or whether the maximum number of iterations has been reached. If the loss function has not converged and the maximum number of iterations has not been reached, return to step S231, otherwise end the training.

[0028] As a further description of the present invention, the specific process of evaluating the operating status of the distribution network in step S3 is:

[0029] S31: Setting meteorological thresholds that affect the stable operation of the distribution network and setting early warning rules within different ranges of the thresholds;

[0030] S32: Adopt corresponding early warning response plans according to the prediction results of the prediction model.

[0031] As a further description of the present invention, the thresholds include maximum temperature, minimum temperature, maximum precipitation, maximum snowfall, and maximum water level.

[0032] A computer device comprises a memory, a processor and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of establishing a distribution network meteorological disaster prediction model are implemented.

[0033] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of establishing a distribution network meteorological disaster prediction model.

[0034] Beneficial effects of the present invention:

[0035] The present invention provides a prediction model for the impact of meteorological events on distribution networks. The model obtains historical meteorological events and historical distribution network fault data information; establishes a prediction model for meteorological events on distribution networks according to the historical meteorological events and historical distribution network fault data information, trains the prediction model for meteorological events on distribution networks, and obtains the trained prediction model for meteorological events on distribution networks; re-acquires historical meteorological events and historical distribution network fault data information, inputs the historical meteorological events and historical distribution network fault data information into the trained prediction model for meteorological events on distribution networks, obtains time series data of the current distribution network, and evaluates the operation status of the distribution network according to the time series data of the current distribution network, thereby realizing the evaluation of distribution network operation faults, being able to predict the faults of the distribution network in advance, and deploying corresponding measures in advance to deal with them according to the prediction results, thereby avoiding the impact of meteorological events on the distribution network and ensuring the safe and stable operation of the distribution network. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 A schematic diagram of the process flow of constructing a prediction model for the impact of meteorological events on a distribution network proposed by the present invention;

[0037] Figure 2 A specific process flow chart of step S2 of constructing a prediction model of the impact of meteorological events on the distribution network proposed by the present invention;

[0038] Figure 3 This is a specific process flow chart of the back propagation algorithm in step S23 of the prediction model for the impact of meteorological events on the distribution network proposed by the present invention. DETAILED DESCRIPTION

[0039] The specific implementation of the present invention is described below in conjunction with the accompanying drawings and embodiments:

[0040] It should be noted that the structures, proportions, sizes, etc. illustrated in the drawings of this specification are only used to match the contents disclosed in the specification for people familiar with this technology to understand and read, and are not used to limit the conditions under which the present invention can be implemented. Any structural modification, change in proportional relationship or adjustment of size should still fall within the scope of the technical contents disclosed in the present invention without affecting the effects and purposes that can be achieved by the present invention.

[0041] At the same time, the terms such as "upper", "lower", "left", "right", "middle" and "one" cited in this specification are only for the convenience of description and are not used to limit the scope of implementation of the present invention. Changes or adjustments to their relative relationships should be regarded as the scope of implementation of the present invention without substantially changing the technical content.

[0042] like Figures 1 to 3 As shown, it shows a specific embodiment of the present invention:

[0043] Embodiment 1

[0044] A prediction model for the impact of meteorological events on a distribution network, wherein the steps for constructing the prediction model are as follows:

[0045] S1: Obtain historical meteorological events and historical distribution network fault data information and perform data preprocessing;

[0046] S2: establishing a prediction model for the impact of meteorological events on the distribution network based on the historical meteorological events, historical distribution network fault data information and current meteorological information, training the prediction model for the impact of meteorological events on the distribution network, and obtaining a trained prediction model for the impact of meteorological events on the distribution network;

[0047] S3: Re-acquire historical meteorological events and historical distribution network fault data information, input the historical meteorological events, historical distribution network fault data information and current meteorological events into the trained prediction model of the impact of meteorological events on the distribution network, obtain the time series data of the current distribution network, and evaluate the operation status of the distribution network according to the time series data of the current distribution network;

[0048] In this embodiment, Figure 1 As shown, the prediction model obtains historical meteorological events and historical distribution network fault data information; establishes a prediction model of meteorological events on the distribution network based on the historical meteorological events and historical distribution network fault data information, trains the prediction model of meteorological events on the distribution network, and obtains the trained prediction model of meteorological events on the distribution network; re-acquires historical meteorological events and historical distribution network fault data information, inputs the historical meteorological events and historical distribution network fault data information into the trained prediction model of meteorological events on the distribution network, obtains the time series data of the current distribution network, and evaluates the operation status of the distribution network based on the time series data of the current distribution network, thereby realizing the evaluation of the operation fault of the distribution network, being able to predict the fault of the distribution network in advance, and deploying corresponding measures in advance according to the prediction results to avoid the impact of the occurrence of meteorological events on the distribution network and ensure the safe and stable operation of the distribution network.

[0049] In step S2, a prediction model for the impact of meteorological events on the distribution network is established. The prediction model uses a multi-layer recurrent neural network, including an input layer, a hidden layer, and an output layer. The specific steps are as follows:

[0050] S21: The input layer receives historical meteorological events and historical power distribution network failure data information after data preprocessing, wherein the preprocessed data is characteristic sequence data of meteorological events and power distribution network failures;

[0051] S22: There are two hidden layers, the maximum number of iterations is 200, the activation function is the tanh function, and the number of neurons in each hidden layer is the same.

[0052] S23: Use the back propagation algorithm over time to update the training set preprocessed by S21 to minimize the loss function, so that the prediction result of the model is infinitely close to the actual output;

[0053] S24: using the model trained in S23 to predict the distribution network fault according to the current actual meteorological events, and obtaining the prediction result of the model at the prediction time;

[0054] S25: Perform autocorrelation analysis on historical data of distribution network fault parameters to obtain an appropriate number of evidences.

[0055] In this embodiment, Figure 2 As shown in FIG, a multi-layer recurrent neural network is used to train the prediction model. The multi-layer recurrent neural network can better capture the complex features and patterns in the sequence data by increasing the number of hidden layers, thereby improving the generalization ability of the model.

[0056] In this embodiment, the historical meteorological events contain multiple characteristics and their parameters are complex. The historical distribution network fault data also contains multiple characteristic information. The recurrent neural network is used for training the responsible characteristic information in the historical meteorological events and historical distribution network fault data. The complex characteristics in the data can be captured, making the training results of the model more reliable and effective, so that reliable prediction results can be obtained when the model is used for prediction. Subsequently, relevant deployments are performed based on the prediction results to avoid the impact of meteorological events on the stable operation of the distribution network, avoid causing corresponding faults, and ensure the safe, stable and effective operation of the distribution network.

[0057] As a further description of the present invention, the historical meteorological events in step S1 are natural disaster events, including earthquakes, rainstorms, floods, and thunderstorms;

[0058] The historical distribution network faults include cable abnormal point faults and substation faults. The data information of the cable abnormal point faults includes cable construction time, abnormal point cross-sectional size, and abnormal point environmental parameters. The substation faults include voltage transformer faults, current transformer faults, DC system grounding faults, busbar faults, and circuit breaker faults.

[0059] In this embodiment, the natural disaster events may cause distribution network failures, wherein earthquakes, rainstorms, floods, thunderstorms, etc. may all cause distribution network failures, and there are certain differences between the parameter standards of different meteorological events.

[0060] In this embodiment, the distribution network is an important medium connecting the substation and the users. Failures in the equipment of the distribution network will lead to failures in the distribution network, thereby affecting power transmission and affecting the lives of people in the disaster-stricken areas.

[0061] In this embodiment, there are multiple types of meteorological events and multiple types of distribution network faults, and the parameters of each fault condition are different, which makes the data of the entire meteorological event and distribution network fault complex.

[0062] The specific steps of data preprocessing in step S1 are:

[0063] S11: Data cleaning,

[0064] S12: Noise data removal;

[0065] S13: Data standardization: the data after steps S11 and S12 are used to generate feature sets according to meteorological events and distribution network faults, and further generate feature sequences.

[0066] In this embodiment, the data cleaning is used to identify and correct errors, inconsistencies and handle missing values ​​in the data set, so that the data meets the requirements and improves accuracy and reliability.

[0067] Specifically include:

[0068] Deduplication: Identify and remove duplicate records in a dataset to avoid introducing bias in the analysis process.

[0069] Handling missing values: Fill in missing data or delete records containing missing values. There are many ways to fill in missing values, which can be done by using statistics such as mean, median, mode, etc., or by inferring based on logic.

[0070] In this embodiment, the noise data removal is part of data cleaning, which requires more calculations than data cleaning and is used to identify and remove outliers or erroneous data in the data to ensure the accuracy and reliability of the data. It is mainly to correct data errors and correct errors and inconsistencies in the data, which is achieved through data verification, rule definition, etc.

[0071] In this embodiment, the data standardization is to generate a feature set from historical meteorological events and historical distribution network fault data, and further generate a feature sequence. The standardized data facilitates the establishment and training of the prediction model.

[0072] As a further description of the present invention, the specific process of the back propagation algorithm in step S23 is:

[0073] S231: forward calculation of the output value of each neuron;

[0074] S232: Reversely calculate the error term value of each neuron, which is the partial derivative of the loss function with respect to the weighted input of the neuron;

[0075] S233: Calculate the gradient of each weight, and then use the batch gradient descent algorithm to update the weight, and determine whether the loss function converges or whether the maximum number of iterations has been reached. If the loss function has not converged and the maximum number of iterations has not been reached, return to step S231, otherwise end the training.

[0076] In this embodiment, Figure 3 As shown, the back propagation algorithm is used to clarify the training process of the prediction model, so that the training results of the prediction model are closest to the actual values, ensuring the accuracy of the final prediction results of the prediction model.

[0077] As a further description of the present invention, the specific process of evaluating the operating status of the distribution network in step S3 is:

[0078] S31: Setting meteorological thresholds that affect the stable operation of the distribution network and setting early warning rules within different ranges of the thresholds;

[0079] S32: Adopt corresponding early warning response plans according to the prediction results of the prediction model.

[0080] As a further description of the present invention, the thresholds include maximum temperature, minimum temperature, maximum precipitation, maximum snowfall, and maximum water level.

[0081] In this embodiment, the distribution network has a certain ability to resist meteorological events during operation. When the threshold of the corresponding parameter is exceeded, certain measures need to be taken. In practice, deployment is carried out in advance based on the prediction results of the prediction model to ensure the safe and stable operation of the distribution network.

[0082] Example 2

[0083] A computer device comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of establishing a distribution network meteorological disaster prediction model as described in Example 1 when executing the computer program.

[0084] Example 3

[0085] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of establishing a distribution network meteorological disaster prediction model as described in Example 1.

[0086] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0087] The present invention is described with reference to the flowcharts and / or block diagrams of the methods and computer program products according to the embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of the processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0088] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0089] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0090] The preferred embodiments of the present invention are described in detail above in conjunction with the accompanying drawings, but the present invention is not limited to the above embodiments, and various changes can be made within the knowledge scope of ordinary technicians in this field without departing from the purpose of the present invention.

[0091] Many other changes and modifications may be made without departing from the concept and scope of the present invention.It should be understood that the present invention is not limited to the specific embodiments, and the scope of the present invention is defined by the appended claims.

Claims

1. A prediction model for the impact of meteorological events on distribution networks, characterized in that: The steps for constructing the prediction model are as follows: S1: Obtain historical meteorological events and historical distribution network fault data information and perform data preprocessing; S2: establishing a prediction model for the impact of meteorological events on the distribution network based on the historical meteorological events, historical distribution network fault data information and current meteorological information, training the prediction model for the impact of meteorological events on the distribution network, and obtaining a trained prediction model for the impact of meteorological events on the distribution network; S3: Re-acquire historical meteorological events and historical distribution network fault data information, input the historical meteorological events, historical distribution network fault data information and current meteorological events into the trained prediction model of the impact of meteorological events on the distribution network, obtain the time series data of the current distribution network, and evaluate the operation status of the distribution network according to the time series data of the current distribution network; In step S2, a prediction model for the impact of meteorological events on the distribution network is established. The prediction model uses a multi-layer recurrent neural network, including an input layer, a hidden layer, and an output layer. The specific steps are as follows: S21: The input layer receives historical meteorological events and historical power distribution network failure data information after data preprocessing, wherein the preprocessed data is characteristic sequence data of meteorological events and power distribution network failures; S22: There are two hidden layers, the maximum number of iterations is 200, the activation function is the tanh function, and the number of neurons in each hidden layer is the same. S23: Use the back propagation algorithm over time to update the training set preprocessed by S21 to minimize the loss function, so that the prediction result of the model is infinitely close to the actual output; S24: using the model trained in S23 to predict the distribution network fault according to the current actual meteorological events, and obtaining the prediction result of the model at the prediction time; S25: Perform autocorrelation analysis on historical data of distribution network fault parameters to obtain an appropriate number of evidences.

2. A meteorological disaster prediction model for distribution network assessment and early warning according to claim 1, characterized in that: The historical meteorological events in step S1 are natural disaster events, including earthquakes, rainstorms, floods, and thunderstorms; The historical distribution network faults include cable abnormal point faults and substation faults. The data information of the cable abnormal point faults includes cable construction time, abnormal point cross-sectional size, and abnormal point environmental parameters. The substation faults include voltage transformer faults, current transformer faults, DC system grounding faults, busbar faults, and circuit breaker faults. The specific steps of data preprocessing in step S1 are: S11: Data cleaning, S12: Noise data removal; S13: Data standardization: the data after steps S11 and S12 are used to generate feature sets according to meteorological events and distribution network faults, and further generate feature sequences.

3. A meteorological disaster prediction model for distribution network assessment and early warning according to claim 1, characterized in that: The specific process of the back propagation algorithm in step S23 is: S231: forward calculation of the output value of each neuron; S232: Reversely calculate the error term value of each neuron, which is the partial derivative of the loss function with respect to the weighted input of the neuron; S233: Calculate the gradient of each weight, and then use the batch gradient descent algorithm to update the weight, and determine whether the loss function converges or whether the maximum number of iterations has been reached. If the loss function has not converged and the maximum number of iterations has not been reached, return to step S231, otherwise end the training.

4. A meteorological disaster prediction model for distribution network assessment and early warning according to claim 1, characterized in that: The specific process of evaluating the operating status of the distribution network in step S3 is as follows: S31: Setting meteorological thresholds that affect the stable operation of the distribution network and setting early warning rules within different ranges of the thresholds; S32: Adopt corresponding early warning response plans according to the prediction results of the prediction model.

5. A meteorological disaster prediction model for distribution network assessment and early warning according to claim 4, characterized in that: The thresholds include maximum temperature, minimum temperature, maximum precipitation, maximum snowfall, and maximum water level.

6. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of establishing a distribution network meteorological disaster prediction model as described in any one of claims 1 to 5 are implemented.

7. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of establishing a distribution network meteorological disaster prediction model as described in any one of claims 1 to 5 are implemented.

Images