Method, device, electronic device and storage medium for processing environmental information of power distribution lines

Abstract

This disclosure relates to the field of power distribution Internet of Things (IoT) technology, specifically a method, device, electronic device, and storage medium for processing environmental information on power distribution lines. The method includes: acquiring multiple time-series data of environmental information in the local area where an IoT-enabled feeder terminal is located, collected by multiple sensors; selecting target data from each time-series data item; inputting the multiple target data items into an OS-ELM model to obtain data fusion features output by the OS-ELM model; and inputting the data fusion features into a pre-stored D-S evidence theory model to obtain wildfire risk prediction results output by the D-S evidence theory model. This technical solution can accurately predict wildfire risk and is primarily used for wildfire prevention in distribution networks.

Description

Technical Field

[0001] The present disclosure relates to the technical field of power distribution Internet of Things, and in particular to a method, device, electronic device, and storage medium for processing environmental information of a power distribution line. Background Art

[0002] In recent years, global warming has become increasingly pronounced, leading to a surge in wildfires. Power failures are a major contributing factor to wildfires, and their early stages are often hidden, making them difficult to detect and potentially leading to major fires. Medium-voltage distribution networks are large, widely distributed, and have lower insulation and maintenance standards. Their probability of failure is two to three orders of magnitude higher than that of high-voltage transmission networks. Furthermore, to ensure high power supply reliability, medium-voltage distribution networks widely utilize low-current grounding. This approach allows single-phase grounding faults to remain untreated, allowing arcing and sparks to ignite surrounding combustibles. According to incomplete statistics, over 80% of wildfires ignited by power failures both domestically and internationally occur in medium-voltage distribution networks. Therefore, the key to preventing wildfires from power failures lies in the effective management of distribution network faults, particularly single-phase grounding faults. To address the potential for wildfires caused by medium-voltage distribution line faults, fault management measures such as single-phase grounding fault tripping, reduced protection delays, and the elimination of reclosing circuit breakers have been widely adopted, effectively curbing the potential for power failures to cause wildfires.

[0003] However, the inventors discovered that due to the complex factors that cause wildfires, distribution network faults can sometimes trigger circuit breakers without causing a wildfire. For example, even a single-phase ground fault during rain, which generates arcing, may not cause a wildfire. This can lead to mismatched fault handling measures, causing numerous unnecessary circuit breakers and severely degrading distribution network reliability. Therefore, resolving the conflict between wildfire prevention and reliable power supply in distribution networks, making fault handling more precise and significantly reducing unnecessary circuit breakers, has become a pressing technical challenge. Summary of the Invention

[0004] In order to solve the problems in the related art, embodiments of the present disclosure provide a method, device, electronic device, and storage medium for processing environmental information of a power distribution line.

[0005] In a first aspect, an embodiment of the present disclosure provides a method for processing environmental information of a distribution line, which is applied to an IoT feeder terminal, comprising:

[0006] Acquire multiple pieces of environmental information time series data of the local area where the IoT feeder terminal is located, collected by multiple sensors;

[0007] Select target data from each piece of environmental information time series data;

[0008] Inputting a plurality of selected target data into an online sequential extreme learning machine (OS-ELM) model to obtain data fusion features output by the OS-ELM model;

[0009] The data fusion features are input into a pre-stored DS evidence theory model to obtain a wildfire risk prediction result output by the DS evidence theory model.

[0010] In a possible implementation, selecting target data from each piece of environmental information time series data includes:

[0011] Preprocessing the plurality of environmental information time series data to obtain a plurality of environmental feature time series data that meet prediction requirements;

[0012] The particle swarm optimization (PSO) algorithm is used to select the target data from each environmental feature time series data.

[0013] In a possible implementation, the selecting the target data from each piece of environmental feature time series data using a particle swarm optimization (PSO) algorithm includes:

[0014] For each piece of environmental characteristic time series data, the environmental characteristic time series data is simulated as a particle swarm, and the speed and position of each particle in the particle swarm are determined according to the following formula:

[0015]

[0016] in, represents the velocity of the i-th particle at the k+1th update, represents the velocity of the i-th particle at the k-th update, represents the position of the i-th particle at the k+1th update, represents the position of the i-th particle at the k-th update, represents the individual extreme value of the i-th particle at the k-th update, represents the group extreme value at the kth update, c1 is the individual learning factor in the PSO algorithm, c2 is the social learning factor in the PSO algorithm, and w k is the inertia weight factor of the PSO algorithm at the kth update, rand(0,1) represents a random value between 0 and 1;

[0017] When the termination condition of the PSO algorithm is reached, the updating is stopped and the group extreme value at the last update is obtained as the target data selected from the environmental feature time series data.

[0018] In one possible implementation, the method further includes:

[0019] The inertia weight factor w in the PSO algorithm is calculated according to the following formula k :

[0020]

[0021] Among them, w min and w max is the minimum and maximum value of the preset inertia weight factor, n is the sliding window of the inertia weight factor, w j is the inertia weight factor at the jth update, β is the preset descent control parameter, T max is the maximum evolutionary generation;

[0022] The individual learning factor c1 and the social learning factor c2 in the PSO algorithm are calculated according to the following formula:

[0023]

[0024] Among them, c min and c max The minimum and maximum values ​​of the preset learning factor.

[0025] In one possible implementation, the method further includes:

[0026] Receive a disposal strategy table issued by a master station device, wherein the disposal strategy table records disposal strategies corresponding to wildfire risk prediction results;

[0027] Based on the handling strategy table, the handling strategy corresponding to the wildfire risk prediction result is executed.

[0028] In one possible implementation, the method further includes:

[0029] Sending the data fusion feature and its corresponding wildfire risk prediction result to the master station device so that the master station device can re-predict a new wildfire risk prediction result, and sending the new wildfire risk prediction result to the IoT feeder terminal;

[0030] Receive new wildfire risk prediction results from the master station equipment;

[0031] Based on the handling strategy table, the handling strategy corresponding to the new wildfire risk prediction result is executed.

[0032] In one possible implementation, the method further includes:

[0033] An updated DS evidence theory model is received from a master station device, where the updated DS evidence theory model is trained by the master station device based on data fusion features uploaded by the IoT feeder terminal and actual wildfire results.

[0034] In a second aspect, an embodiment of the present disclosure provides a method for processing environmental information of a distribution line, which is applied to a master station device, comprising:

[0035] Receive data fusion features uploaded by each IoT feeder terminal and its corresponding wildfire risk prediction result, wherein the data fusion features are obtained by the IoT feeder terminal performing data selection and feature extraction fusion of an online sequential extreme learning machine (OS-ELM) model on multiple environmental information time series data of the local area where the IoT feeder terminal is located collected by multiple sensors; the wildfire risk prediction result is obtained by the IoT feeder terminal inputting the data fusion features into a pre-stored DS evidence theory model and executing the DS evidence theory model;

[0036] Based on the data fusion characteristics uploaded by each IoT feeder terminal and its corresponding wildfire risk prediction result, determine whether the wildfire risk prediction result of the IoT feeder terminal is accurate;

[0037] When the wildfire risk prediction result of the Internet of Things feeder terminal is inaccurate, a new wildfire risk prediction result is obtained and the new wildfire risk prediction result is sent to the Internet of Things feeder terminal.

[0038] In one possible implementation, the determining whether the wildfire risk prediction result of the IoT feeder terminal is accurate based on the data fusion characteristics uploaded by each IoT feeder terminal and its corresponding wildfire risk prediction result includes:

[0039] For each IoT feeder terminal, obtaining wildfire risk prediction results of other IoT feeder terminals within a predetermined distance range from the IoT feeder terminal;

[0040] Determining whether the wildfire risk prediction result of the IoT feeder terminal is correct based on the wildfire risk prediction results of the other IoT feeder terminals;

[0041] When the wildfire risk prediction result of the IoT feeder terminal is inaccurate, obtaining a new wildfire risk prediction result includes:

[0042] When the wildfire risk prediction result of the Internet of Things feeder terminal is inaccurate, a new wildfire risk prediction result of the Internet of Things feeder terminal is determined based on the wildfire risk prediction results of the other Internet of Things feeder terminals.

[0043] In one possible implementation, the determining whether the wildfire risk prediction result of the IoT feeder terminal is accurate based on the data fusion characteristics uploaded by each IoT feeder terminal and its corresponding wildfire risk prediction result includes:

[0044] For each IoT feeder terminal, obtaining data fusion features of other IoT feeder terminals within a predetermined distance range from the IoT feeder terminal;

[0045] Inputting the data fusion features of the other IoT feeder terminals and the data fusion features of the IoT feeder terminal into the DS evidence theory model to obtain a new wildfire risk prediction result output by the DS evidence theory model;

[0046] Compare the wildfire risk prediction result of the IoT feeder terminal with the new wildfire risk prediction result to determine whether the wildfire risk prediction result of the IoT feeder terminal is accurate.

[0047] In one possible implementation, the method further includes:

[0048] Based on the data fusion features uploaded by each IoT feeder terminal and its corresponding real wildfire results, the DS evidence theory model is trained to obtain an updated DS evidence theory model;

[0049] The updated DS evidence theory model is distributed to each IoT feeder terminal.

[0050] In a third aspect, an embodiment of the present disclosure provides an environmental information processing device for a distribution line, which is applied to an IoT feeder terminal and includes:

[0051] An acquisition module is configured to acquire a plurality of environmental information time series data of a local area where the IoT feeder terminal is located, collected by a plurality of sensors;

[0052] A selection module is configured to select target data from each piece of environmental information time series data;

[0053] a fusion module configured to input a plurality of selected target data into an online sequential extreme learning machine (OS-ELM) model to obtain data fusion features output by the OS-ELM model;

[0054] The prediction module is configured to input the data fusion features into a pre-stored DS evidence theory model to obtain a wildfire risk prediction result output by the DS evidence theory model.

[0055] In a possible implementation, the selection module is configured to:

[0056] Preprocessing the plurality of environmental information time series data to obtain a plurality of environmental feature time series data that meet prediction requirements;

[0057] The particle swarm optimization (PSO) algorithm is used to select the target data from each environmental feature time series data.

[0058] In a possible implementation, the selection module uses a particle swarm optimization (PSO) algorithm to select the target data from each piece of environmental feature time series data, and is configured as follows:

[0059] For each piece of environmental characteristic time series data, the environmental characteristic time series data is simulated as a particle swarm, and the speed and position of each particle in the particle swarm are determined according to the following formula:

[0060]

[0061] in, represents the velocity of the i-th particle at the k+1th update, represents the velocity of the i-th particle at the k-th update, represents the position of the i-th particle at the k+1th update, represents the position of the i-th particle at the k-th update, represents the individual extreme value of the i-th particle at the k-th update, represents the group extreme value at the kth update, c1 is the individual learning factor in the PSO algorithm, c2 is the social learning factor in the PSO algorithm, and w k is the inertia weight factor of the PSO algorithm at the kth update, and rand(0,1) represents a random value between 0 and 1;

[0062] When the termination condition of the PSO algorithm is reached, the updating is stopped and the group extreme value at the last update is obtained as the target data selected from the environmental feature time series data.

[0063] In a possible implementation, the device further includes:

[0064] The parameter determination module is configured to calculate the inertia weight factor w in the PSO algorithm according to the following formula k :

[0065]

[0066] Among them, w min and w max is the minimum and maximum value of the preset inertia weight factor, n is the sliding window of the inertia weight factor, w j is the inertia weight factor at the jth update, β is the preset descent control parameter, T max is the maximum evolutionary generation;

[0067] The individual learning factor c1 and the social learning factor c2 in the PSO algorithm are calculated according to the following formula:

[0068]

[0069] Among them, c min and c max The minimum and maximum values ​​of the preset learning factor.

[0070] In a possible implementation, the device further includes:

[0071] A receiving module is configured to receive a handling strategy table issued by a master station device, wherein the handling strategy table records handling strategies corresponding to wildfire risk prediction results;

[0072] The strategy execution module is configured to execute the handling strategy corresponding to the wildfire risk prediction result based on the handling strategy table.

[0073] In a possible implementation, the device further includes:

[0074] A data reporting module is configured to send the data fusion feature and its corresponding wildfire risk prediction result to the master station device so that the master station device can re-predict a new wildfire risk prediction result and send the new wildfire risk prediction result to the IoT feeder terminal;

[0075] A result receiving module is configured to receive new wildfire risk prediction results issued by the master station device;

[0076] The new strategy execution module is configured to execute the handling strategy corresponding to the new wildfire risk prediction result based on the handling strategy table.

[0077] In a possible implementation, the device further includes:

[0078] The model structure module is configured to receive an updated DS evidence theory model issued by a master station device, wherein the updated DS evidence theory model is trained by the master station device based on the data fusion features uploaded by the IoT feeder terminal and the actual wildfire results.

[0079] In a fourth aspect, an embodiment of the present disclosure provides an environmental information processing device for a power distribution line, which is applied to a master station device, including:

[0080] The data receiving module is configured to receive data fusion features uploaded by each IoT feeder terminal and its corresponding wildfire risk prediction result, wherein the data fusion features are obtained by the IoT feeder terminal performing data selection and feature extraction fusion of an online sequential extreme learning machine (OS-ELM) model on multiple environmental information time series data of the local area where the IoT feeder terminal is located collected by multiple sensors; the wildfire risk prediction result is obtained by the IoT feeder terminal inputting the data fusion features into a pre-stored DS evidence theory model and executing the DS evidence theory model;

[0081] A result judgment module is configured to judge whether the wildfire risk prediction result of the IoT feeder terminal is accurate based on the data fusion characteristics uploaded by each IoT feeder terminal and its corresponding wildfire risk prediction result;

[0082] The result sending module is configured to obtain a new wildfire risk prediction result and send the new wildfire risk prediction result to the Internet of Things feeder terminal when the wildfire risk prediction result of the Internet of Things feeder terminal is inaccurate.

[0083] In a possible implementation, the result determination module is configured to:

[0084] For each IoT feeder terminal, obtaining wildfire risk prediction results of other IoT feeder terminals within a predetermined distance range from the IoT feeder terminal;

[0085] Determining whether the wildfire risk prediction result of the IoT feeder terminal is correct based on the wildfire risk prediction results of the other IoT feeder terminals;

[0086] The part of the result sending module that obtains a new wildfire risk prediction result when the wildfire risk prediction result of the IoT feeder terminal is inaccurate is configured as follows:

[0087] When the wildfire risk prediction result of the Internet of Things feeder terminal is inaccurate, a new wildfire risk prediction result of the Internet of Things feeder terminal is determined based on the wildfire risk prediction results of the other Internet of Things feeder terminals.

[0088] In a possible implementation, the result determination module is configured to:

[0089] For each IoT feeder terminal, obtaining data fusion features of other IoT feeder terminals within a predetermined distance range from the IoT feeder terminal;

[0090] Inputting the data fusion features of the other IoT feeder terminals and the data fusion features of the IoT feeder terminal into the DS evidence theory model to obtain a new wildfire risk prediction result output by the DS evidence theory model;

[0091] Compare the wildfire risk prediction result of the IoT feeder terminal with the new wildfire risk prediction result to determine whether the wildfire risk prediction result of the IoT feeder terminal is accurate.

[0092] In a possible implementation, the device further includes:

[0093] A model training module is configured to train the DS evidence theory model based on the data fusion features uploaded by each IoT feeder terminal and its corresponding real wildfire results to obtain an updated DS evidence theory model;

[0094] The model delivery module is configured to deliver the updated DS evidence theory model to each IoT feeder terminal.

[0095] In a fifth aspect, an embodiment of the present disclosure provides an electronic device comprising a memory and a processor, wherein the memory is used to store one or more computer instructions, and wherein the one or more computer instructions are executed by the processor to implement a method as described in any one of the first aspect or the second aspect.

[0096] In a sixth aspect, an embodiment of the present disclosure provides a computer-readable storage medium on which computer instructions are stored. When the computer instructions are executed by a processor, the method as described in any one of the first aspect or the second aspect is implemented.

[0097] According to the technical solution provided by the embodiment of the present disclosure, multiple environmental information time series data of the local area where the IoT feeder terminal is located, collected by multiple sensors, can be obtained through the IoT feeder terminal, and target data can be selected from each environmental information time series data; the multiple target data are input into the OS-ELM model to obtain the data fusion features output by the OS-ELM model; the data fusion features are input into the preset DS evidence theory model to obtain the wildfire risk prediction results output by the DS evidence theory model; in this way, by processing the environmental information of the local area where the IoT feeder terminal in the distribution line is located, accurate prediction of wildfires can be achieved, and then accurate disposal can be carried out based on the prediction, reducing unnecessary tripping caused by mismatch and mismatch of disposal measures.

[0098] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0099] Other features, objectives and advantages of the present disclosure will become more apparent through the following detailed description of non-limiting embodiments in conjunction with the accompanying drawings. In the accompanying drawings:

[0100] Figure 1 A flow chart of a method for processing environmental information of a power distribution line provided by an embodiment of the present disclosure is shown.

[0101] Figure 2 A flowchart of an environmental information processing method for a power distribution line applied to a master station device is shown in an embodiment of the present disclosure.

[0102] Figure 3A structural block diagram of an environmental information processing device for a power distribution line provided by an embodiment of the present disclosure is shown.

[0103] Figure 4 A structural block diagram of an environment information processing device for a power distribution line applied to a master station device provided by an embodiment of the present disclosure is shown.

[0104] Figure 5 A structural block diagram of an electronic device according to an embodiment of the present disclosure is shown.

[0105] Figure 6 A schematic diagram showing the structure of a computer system suitable for implementing the method of the embodiment of the present disclosure is shown. DETAILED DESCRIPTION

[0106] Hereinafter, exemplary embodiments of the present disclosure will be described in detail with reference to the accompanying drawings so that those skilled in the art can easily implement them. In addition, for the sake of clarity, parts not related to the description of the exemplary embodiments are omitted in the accompanying drawings.

[0107] In the present disclosure, it should be understood that terms such as "include" or "have" are intended to indicate the presence of features, numbers, steps, actions, components, parts, or combinations thereof disclosed in the present specification, and are not intended to exclude the possibility that one or more other features, numbers, steps, actions, components, parts, or combinations thereof exist or are added.

[0108] It should also be noted that, in the absence of conflict, the embodiments and features of the embodiments of the present disclosure may be combined with each other. The present disclosure will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0109] Figure 1 FIG. 1 is a flow chart showing a method for processing environmental information of a power distribution line provided by an embodiment of the present disclosure. Figure 1 As shown, the method includes the following steps S101-S104:

[0110] In step S101, multiple environmental information time series data of the local area where the IoT feeder terminal is located, collected by multiple sensors, are obtained;

[0111] In step S102, target data is selected from each piece of environmental information time series data;

[0112] In step S103, a plurality of selected target data are input into an online sequential extreme learning machine (OS-ELM) model to obtain data fusion features output by the OS-ELM model;

[0113] In step S104, the data fusion features are input into a pre-stored DS evidence theory model to obtain a wildfire risk prediction result output by the DS evidence theory model.

[0114] In one possible implementation, the environmental information processing method of the distribution line is applicable to an IoT feeder terminal that can perform environmental information processing of the distribution line. The IoT feeder terminal can be an intelligent terminal device installed in a distribution room or on a feeder line, and can communicate with a remote distribution master station device.

[0115] In one possible implementation, mountainous, forested, and grassland areas often have complex environments, changeable weather, and prominent micro-meteorological characteristics. Traditional meteorological monitoring stations are expensive, and communication and power supply are difficult. It is difficult to achieve extensive monitoring of the micro-environment of distribution lines, and it is also difficult to quickly interact with the distribution automation system. Therefore, this implementation provides an Internet of Things feeder terminal, which can obtain environmental information collected by these sensors from multiple sensors. The environmental information is various environmental information of the local area where the Internet of Things feeder terminal is located. The environmental information includes various information that has a certain impact on the occurrence of wildfires, and can include temperature information, humidity information, smoke information, rainfall information, wind speed information, and other environmental information. Each sensor can periodically collect corresponding environmental information and form an environmental information time series data in chronological order; in this way, multiple environmental information time series data can be obtained from multiple sensors.

[0116] In one possible implementation, the multiple sensors may include multiple sensors for collecting various environmental information, such as a temperature sensor for collecting temperature information, a humidity sensor for collecting humidity information, a smoke sensor for collecting smoke information, a rainfall sensor for collecting rainfall information, a wind speed sensor for collecting wind speed information, and the like. To collect environmental information of a local area, multiple sensors of each type of sensor may be provided and distributed in the local area.

[0117] In one possible implementation, target data can be selected from each piece of environmental information time series data and input into the OS-ELM (Online Sequential Extreme Learning Machine) model. The target data can be a piece of data randomly selected from each piece of environmental information time series data, or a piece of data selected using a preset algorithm.

[0118] In a possible implementation, the OS-ELM model is used to extract features from multiple input target data, mainly by grouping the input target data with similar attributes into one category according to the similarity of the attributes, and extracting features from the fused data to obtain data fusion features.

[0119] The OS-ELM model is divided into two parts. The first part is to calculate and initialize the output weight β using the ELM algorithm through a small amount of training samples. (0), the second part starts online learning. Every time a new target data arrives, a new output weight β is obtained through a recursive formula (k+1) In this way, when the selected multiple target data are input into the OS-ELM model, a new weight β can be calculated according to the recursive formula, and the data fusion features of the multiple target data can be obtained based on the weight.

[0120] In one possible implementation, OS-ELM belongs to a supervised learning method, that is, a selected set of data is used to train the network. Based on the principle of neural network, some excitation functions are often selected as the transfer function g(x) of the neuron. Such functions are selected because they are infinitely integrable functions, and the outputs of these functions are between positive and negative 1. Therefore, it can be considered that the data fusion characteristics of the output multiple target data are some probability t i , and this probability just meets the input of the DS evidence theory model. The data fusion characteristics of the multiple target data have certain fuzziness. The DS evidence theory model can further fuse these data to remove the fuzziness of the data to achieve the final decision output.

[0121] In one possible implementation, the Dempster-Shafer (DS) evidence theory model synthesizes evidence within a hypothesis space to achieve information fusion and infer useful facts. In this embodiment, the DS evidence theory model is primarily used to synthesize evidence, namely, the data fusion features of multiple target data, to determine whether wildfires will occur. This data fusion feature can be input into a preset DS evidence theory model to obtain the wildfire risk prediction result output by the DS evidence theory model.

[0122] For example, in this embodiment, the recognition framework of the DS evidence theory model can be a plurality of events composed of a free combination of open flame, no open flame, interference fire, no interference fire, smoldering fire, no ignition fire, etc. Each data fusion feature can be used as a piece of evidence to calculate the probability of occurrence of each event in the recognition framework under different evidence, and then according to the evidence synthesis rules, the different evidence (each data fusion feature) is fused, that is, evidence synthesis, to obtain the evidence synthesis result, that is, the probability of occurrence of each event in the recognition framework after evidence fusion.

[0123] This embodiment can obtain multiple environmental information time series data of the local area where the IoT feeder terminal is located, which are collected by multiple sensors through the IoT feeder terminal, and select target data from each environmental information time series data to be input into the OS-ELM model; input multiple target data into the OS-ELM model to obtain data fusion features output by the OS-ELM model; input the data fusion features into a preset DS evidence theory model to obtain a wildfire risk prediction result output by the DS evidence theory model; in this way, by processing the environmental information of the local area where the IoT feeder terminal is located in the distribution line, accurate prediction of wildfires can be achieved, and then accurate disposal can be carried out based on the prediction, reducing unnecessary tripping caused by mismatch and mismatch of disposal measures.

[0124] In a possible implementation, selecting target data from each piece of environmental information time series data includes:

[0125] Preprocessing the plurality of environmental information time series data to obtain a plurality of environmental feature time series data that meet prediction requirements;

[0126] The particle swarm optimization (PSO) algorithm is used to select the target data from each environmental feature time series data.

[0127] In this embodiment, multiple environmental information time series data can be preprocessed to obtain multiple environmental feature time series data that meet the prediction requirements. The preprocessing includes at least one of denoising, smoothing and normalization. Denoising refers to removing data that is obviously not within a reasonable range. For example, for temperature data, data with very high temperature values ​​such as above 50 degrees or very low temperature values ​​such as below minus 20 degrees can be removed. Under normal circumstances, the data values ​​in each environmental information time series data will gradually change over time. Therefore, the mutation data in the data can be removed by smoothing. Normalization refers to changing the data to a fixed interval (range). For example, the environmental information time series data can be mapped to between 0 and 1. A variety of algorithms such as maximum and minimum normalization, mean variance standardization, decimal calibration, and quantitative feature binarization can be used for normalization. Taking the maximum and minimum normalization algorithm as an example, the following formula can be used for normalization:

[0128]

[0129] Among them, x is the data value before normalization, x′ is the data value after normalization, and x min is the minimum value of the data, x max The maximum value among the data values.

[0130] In this implementation, the PSO algorithm simulates birds in a flock by designing massless particles with only two properties: speed and position. Speed ​​represents the speed of movement, and position represents the direction of movement. Each particle independently searches for the optimal solution in the search space and records it as its current individual extreme value. This individual extreme value is then shared with other particles in the swarm. The optimal individual extreme value found becomes the current global optimal solution for the entire swarm. All particles in the swarm then adjust their speed and position based on their individual extreme value and the current global optimal solution shared by the entire swarm.

[0131] In this embodiment, the PSO algorithm can be used to select the optimal data, i.e., the target data, from a piece of environmental feature time series data corresponding to each sensor to be input into the OS-ELM model. A particle swarm can be used to simulate each piece of environmental feature time series data. The global optimal solution of the particle swarm searched using the PSO algorithm is the target data selected from each piece of environmental feature time series data.

[0132] In a possible implementation, the selecting the target data from each piece of environmental feature time series data using the PSO algorithm includes:

[0133] For each piece of environmental characteristic time series data, the environmental characteristic time series data is simulated as a particle swarm, and the speed and position of each particle in the particle swarm are determined according to the following formula:

[0134]

[0135] in, represents the velocity of the i-th particle at the k+1th update, represents the velocity of the i-th particle at the k-th update, represents the position of the i-th particle at the k+1th update, represents the position of the i-th particle at the k-th update, represents the individual extreme value of the i-th particle at the k-th update, represents the group extreme value at the kth update, c1 is the individual learning factor in the PSO algorithm, c2 is the social learning factor in the PSO algorithm, and w k is the inertia weight factor of the PSO algorithm at the kth update, and rand(0,1) represents a random value between 0 and 1;

[0136] When the termination condition of the PSO algorithm is reached, the updating is stopped and the group extreme value at the last update is obtained as the target data selected from the environmental feature time series data.

[0137] In this embodiment, a particle swarm can be used to simulate an environmental feature time series data corresponding to a sensor, and the swarm extreme value (i.e., the global optimal solution) searched by the PSO algorithm is the target data searched from the environmental feature time series data.

[0138] In this embodiment, the above and The calculation formula updates the position and velocity of the particle, taking into account the individual extreme value and group extreme value of the particle; the objective function can be preset to evaluate the fitness of each particle, that is, the prediction accuracy or error of the OS-ELM model; according to the fitness of the particle, the individual extreme value and group extreme value are updated to determine whether the termination condition such as the maximum number of iterations or a certain preset accuracy requirement is met. When the termination condition of the PSO algorithm is met, the update is stopped, and the group extreme value at the last update is the optimal parameter input to the OS-ELM in the particle swarm, that is, the target data selected by the environmental feature time series data.

[0139] In one possible implementation, the method further includes:

[0140] The inertia weight factor w in the PSO algorithm is calculated according to the following formula k :

[0141]

[0142] Among them, w min and w max is the minimum and maximum value of the preset inertia weight factor, n is the sliding window of the inertia weight factor, w j is the inertia weight factor at the jth update, β is the preset descent control parameter, T max is the maximum evolutionary generation;

[0143] The individual learning factor c1 and the social learning factor c2 in the PSO algorithm are calculated according to the following formula:

[0144]

[0145] Among them, c min and c max The minimum and maximum values ​​of the preset learning factor.

[0146] In this embodiment, when the inertia weight factor in the PSO algorithm approaches 1, the global search capability is strong, which is suitable for the global search in the early stage of the PSO algorithm; and when the inertia weight factor approaches 0, the local search capability of the PSO algorithm is improved, the global search capability is reduced, and it is suitable for the local search in the later stage of the PSO algorithm. At the same time, the individual learning factor c1 and the social learning factor c2 in the PSO algorithm will also affect the search capability of the particle swarm. In order to maintain a high inertia weight factor in the early stage of the algorithm and take the smallest inertia weight factor in the later stage, an inertia weight factor reduction strategy based on a sliding average can be adopted to effectively eliminate the impact of data jitter and make the inertia weight factor adaptively decrease. The above-mentioned w aver The calculation formula is used to calculate the sliding mean w of the inertia weight factor aver , and calculate the inertia weight factor w at the kth update based on this k .

[0147] In this embodiment, the values ​​of the individual learning factor c1 and the social learning factor c2 in the PSO algorithm can be synchronously transformed with the inertia weight, which can further optimize the PSO algorithm and better search for the global optimal solution.

[0148] In one possible implementation, the method further includes:

[0149] Receive a disposal strategy table issued by a master station device, wherein the disposal strategy table records disposal strategies corresponding to wildfire risk prediction results;

[0150] Based on the handling strategy table, the handling strategy corresponding to the wildfire risk prediction result is executed.

[0151] In this embodiment, the IoT feeder terminal can communicate with the remote power distribution master station device and receive the handling strategy table issued by the master station device. The handling strategy table records the handling strategy corresponding to the wildfire risk prediction result. In this way, after determining the wildfire risk prediction result, the IoT feeder terminal can execute the handling strategy corresponding to the wildfire risk prediction result according to the record in the handling strategy table. For example, the handling strategy may include tripping when the determined probability of an open fire exceeds a first predetermined probability value, not tripping when the determined probability of an interference fire exceeds a second predetermined probability value and the probability of an open fire is less than a third predetermined probability value; etc.

[0152] In one possible implementation, the method further includes:

[0153] Sending the data fusion feature and its corresponding wildfire risk prediction result to the master station device so that the master station device can re-predict a new wildfire risk prediction result, and distribute the new wildfire risk prediction result to the IoT feeder terminal;

[0154] Receive new wildfire risk prediction results from the master station equipment;

[0155] Based on the handling strategy table, the handling strategy corresponding to the new wildfire risk prediction result is executed.

[0156] In this embodiment, the IoT feeder terminal can also send the data fusion characteristics and the corresponding wildfire risk prediction results to the master station device. The master station device can receive the data fusion characteristics and the corresponding wildfire risk prediction results reported by multiple IoT feeder terminals, the positions of each IoT feeder terminal pre-stored in the master station device, and for each IoT feeder terminal, the master station device can obtain the wildfire risk prediction results of other IoT feeder terminals within a predetermined distance range from the IoT feeder terminal, and determine whether the wildfire risk prediction result of the IoT feeder terminal is correct based on the wildfire risk prediction results of other IoT feeder terminals. If the similarity between the wildfire risk prediction result of the Internet of Things feeder terminal and the wildfire risk prediction results of other Internet of Things feeder terminals exceeds the predetermined threshold, then the wildfire risk prediction result of the Internet of Things feeder terminal is judged to be correct. If the similarity between the wildfire risk prediction result of the Internet of Things feeder terminal and the wildfire risk prediction results of other Internet of Things feeder terminals does not exceed the predetermined threshold, then the wildfire risk prediction result of the Internet of Things feeder terminal is judged to be incorrect. At this time, the master station device can calculate the average value of the wildfire risk prediction results of other Internet of Things feeder terminals as the new wildfire risk prediction result of the Internet of Things feeder terminal. Alternatively, the master station device can input the data fusion characteristics of the IoT feeder terminal and other IoT feeder terminals within a predetermined distance range from the IoT feeder terminal into the DS evidence theory model, execute the DS evidence theory model, obtain a new wildfire risk prediction result output by the DS evidence theory model, and directly calculate the similarity between the new wildfire risk prediction result and the wildfire risk prediction result of the IoT feeder terminal. If the similarity is high and exceeds a predetermined threshold, it is judged that the wildfire risk prediction result of the IoT feeder terminal is accurate; if the similarity is low, it is judged that the wildfire risk prediction result of the IoT feeder terminal is inaccurate.

[0157] In this embodiment, the master station device can comprehensively consider the data of multiple IoT feeder terminals in an area to judge the wildfire risk prediction results of the corresponding IoT feeder terminals. When a measurement error occurs in the sensor corresponding to a certain IoT feeder terminal, resulting in an erroneous wildfire risk prediction result, the master station device can promptly detect the error and send a new wildfire risk prediction result to the IoT feeder terminal. The IoT feeder terminal can execute the handling strategy corresponding to the new wildfire risk prediction result based on the handling strategy table; for example, when an erroneous prediction occurs in the IoT feeder terminal and tripping is performed, the master station device can promptly detect the error and send a new wildfire risk prediction result to make the IoT feeder terminal reclose; or, when an erroneous prediction occurs in the IoT feeder terminal and it is judged that the error did not trip when there is a wildfire risk, the master station device can promptly detect the error and send a new wildfire risk prediction result to make the IoT feeder terminal perform a tripping operation.

[0158] In one possible implementation, the method further includes:

[0159] An updated DS evidence theory model is received from a master station device, where the updated DS evidence theory model is trained by the master station device based on data fusion features uploaded by the IoT feeder terminal and actual wildfire results.

[0160] In this embodiment, as time goes by, the master station device can continuously obtain the data fusion features uploaded by each IoT feeder terminal and obtain the real wildfire results. Therefore, the master station device can train an updated DS evidence theory model based on the data fusion features uploaded by the IoT feeder terminal and the real wildfire results. In this way, as time goes by, the more training data there is, the more accurate the prediction of the DS evidence theory model will be.

[0161] Figure 2 FIG. 1 is a flow chart showing a method for processing environmental information of a power distribution line applied to a master station device according to an embodiment of the present disclosure. Figure 2 As shown, the method includes the following steps S201-S203:

[0162] In step S201, the data fusion features uploaded by each IoT feeder terminal and their corresponding wildfire risk prediction results are received;

[0163] In step S202, based on the data fusion characteristics uploaded by each IoT feeder terminal and its corresponding wildfire risk prediction result, it is determined whether the wildfire risk prediction result of the IoT feeder terminal is accurate;

[0164] In step S203, when the wildfire risk prediction result of the IoT feeder terminal is inaccurate, a new wildfire risk prediction result is obtained and sent to the IoT feeder terminal.

[0165] In a possible implementation, the method for processing environmental information of a distribution line is applicable to a master station device capable of performing environmental information processing of a distribution line, and the master station device can communicate with an Internet of Things feeder terminal.

[0166] In one possible implementation, mountainous, forested, and grassland areas often have complex environments, changeable weather, and prominent micro-meteorological characteristics. Traditional meteorological monitoring stations are expensive, and communication and power supply are difficult. It is difficult to achieve extensive monitoring of the micro-environment of distribution lines, and it is also difficult to quickly interact with the distribution automation system. Therefore, this implementation provides an Internet of Things feeder terminal, which can obtain environmental information collected by these sensors from multiple sensors. The environmental information is various environmental information of the local area where the Internet of Things feeder terminal is located. The environmental information includes various information that has a certain impact on the occurrence of wildfires, and can include temperature information, humidity information, smoke information, rainfall information, wind speed information, and other environmental information. Each sensor can periodically collect corresponding environmental information and form an environmental information time series data in chronological order; in this way, multiple environmental information time series data can be obtained from multiple sensors.

[0167] In a possible implementation, the IoT feeder terminal can select target data from each piece of environmental information time series data and input it into the online sequential extreme learning machine OS-ELM model; input the multiple target data into the OS-ELM model to obtain data fusion features output by the OS-ELM model; input the data fusion features into a pre-stored DS evidence theory model to obtain a wildfire risk prediction result output by the DS evidence theory model; and then send the data fusion features and their corresponding wildfire risk prediction results to the master station device.

[0168] In one possible implementation, the master station device can receive data fusion features and their corresponding wildfire risk prediction results reported by multiple IoT feeder terminals. After receiving this data, the master station device can comprehensively consider the data of multiple IoT feeder terminals in an area to determine the wildfire risk prediction result of the corresponding IoT feeder terminal. When a measurement error occurs in the sensor corresponding to a certain IoT feeder terminal, resulting in an erroneous wildfire risk prediction result, the master station device can promptly detect the error and send a new wildfire risk prediction result to the IoT feeder terminal. The IoT feeder terminal can execute the handling strategy corresponding to the new wildfire risk prediction result based on the handling strategy table. For example, when the IoT feeder terminal makes an erroneous prediction and performs a trip, the master station device can promptly detect the error and send a new wildfire risk prediction result to cause the IoT feeder terminal to reclose. Alternatively, when the IoT feeder terminal makes an erroneous prediction and determines that the error did not trip when there is a wildfire risk, the master station device can promptly detect the error and send a new wildfire risk prediction result to cause the IoT feeder terminal to perform a trip operation.

[0169] In one possible implementation, the determining whether the wildfire risk prediction result of the IoT feeder terminal is accurate based on the data fusion characteristics uploaded by each IoT feeder terminal and its corresponding wildfire risk prediction result includes:

[0170] For each IoT feeder terminal, obtaining wildfire risk prediction results of other IoT feeder terminals within a predetermined distance range from the IoT feeder terminal;

[0171] Determining whether the wildfire risk prediction result of the IoT feeder terminal is correct based on the wildfire risk prediction results of the other IoT feeder terminals;

[0172] When the wildfire risk prediction result of the IoT feeder terminal is inaccurate, obtaining a new wildfire risk prediction result includes:

[0173] When the wildfire risk prediction result of the Internet of Things feeder terminal is inaccurate, a new wildfire risk prediction result of the Internet of Things feeder terminal is determined based on the wildfire risk prediction results of the other Internet of Things feeder terminals.

[0174] In this embodiment, the positions of each IoT feeder terminal are pre-stored in the master station device. For each IoT feeder terminal, the master station device can obtain the wildfire risk prediction results of other IoT feeder terminals whose distance from the IoT feeder terminal is within a predetermined distance range, and determine whether the wildfire risk prediction result of the IoT feeder terminal is correct based on the wildfire risk prediction results of other IoT feeder terminals. If the similarity between the wildfire risk prediction result of the IoT feeder terminal and the wildfire risk prediction results of other IoT feeder terminals exceeds the predetermined threshold, then the wildfire risk prediction result of the IoT feeder terminal is determined to be correct. If the similarity between the wildfire risk prediction result of the IoT feeder terminal and the wildfire risk prediction results of other IoT feeder terminals does not exceed the predetermined threshold, then the wildfire risk prediction result of the IoT feeder terminal is determined to be incorrect. At this time, the master station device can calculate the average value of the wildfire risk prediction results of other IoT feeder terminals as the new wildfire risk prediction result of the IoT feeder terminal.

[0175] In one possible implementation, the determining whether the wildfire risk prediction result of the IoT feeder terminal is accurate based on the data fusion characteristics uploaded by each IoT feeder terminal and its corresponding wildfire risk prediction result includes:

[0176] For each IoT feeder terminal, obtaining data fusion features of other IoT feeder terminals within a predetermined distance range from the IoT feeder terminal;

[0177] Inputting the data fusion features of the other IoT feeder terminals and the data fusion features of the IoT feeder terminal into the DS evidence theory model to obtain a new wildfire risk prediction result output by the DS evidence theory model;

[0178] Compare the wildfire risk prediction result of the IoT feeder terminal with the new wildfire risk prediction result to determine whether the wildfire risk prediction result of the IoT feeder terminal is accurate.

[0179] In this embodiment, the master station device can input the data fusion characteristics of the IoT feeder terminal and other IoT feeder terminals within a predetermined distance range from the IoT feeder terminal into the DS evidence theory model, execute the DS evidence theory model, obtain a new wildfire risk prediction result output by the DS evidence theory model, and directly calculate the similarity between the new wildfire risk prediction result and the wildfire risk prediction result of the IoT feeder terminal. If the similarity is higher than a predetermined threshold, the wildfire risk prediction result of the IoT feeder terminal is judged to be accurate. If the similarity is lower, the wildfire risk prediction result of the IoT feeder terminal is judged to be inaccurate.

[0180] In one possible implementation, the method further includes:

[0181] Based on the data fusion features uploaded by each IoT feeder terminal and its corresponding real wildfire results, the DS evidence theory model is trained to obtain an updated DS evidence theory model;

[0182] The updated DS evidence theory model is distributed to each IoT feeder terminal.

[0183] In this embodiment, as time goes by, the master station device can continuously obtain the data fusion features uploaded by each IoT feeder terminal and obtain the real wildfire results. Therefore, the master station device can train an updated DS evidence theory model based on the data fusion features uploaded by the IoT feeder terminal and the real wildfire results. In this way, as time goes by, the more training data there is, the more accurate the prediction of the DS evidence theory model will be.

[0184] The present disclosure also provides an environmental information processing device for a power distribution line, Figure 3 The following is a block diagram showing the structure of the environmental information processing device for the power distribution line provided by the embodiment of the present disclosure. The device can be implemented as part or all of the electronic equipment through software, hardware or a combination of both. Figure 3 As shown, the environmental information processing device of the power distribution line includes:

[0185] An acquisition module 301 is configured to acquire multiple pieces of environmental information time series data of a local area where the IoT feeder terminal is located, collected by multiple sensors;

[0186] The selection module 302 is configured to select target data from each piece of environmental information time series data;

[0187] A fusion module 303 is configured to input the plurality of selected target data into an online sequential extreme learning machine (OS-ELM) model to obtain data fusion features output by the OS-ELM model;

[0188] The prediction module 304 is configured to input the data fusion features into a pre-stored DS evidence theory model to obtain a wildfire risk prediction result output by the DS evidence theory model.

[0189] In a possible implementation, the selection module is configured to:

[0190] The particle swarm optimization (PSO) algorithm is used to select target data from each environmental feature time series data to be input into the OS-ELM model.

[0191] In a possible implementation, the selection module is configured to select the target data from each piece of environmental feature time series data using the PSO algorithm:

[0192] For each piece of environmental characteristic time series data, the environmental characteristic time series data is simulated as a particle swarm, and the speed and position of each particle in the particle swarm are determined according to the following formula:

[0193]

[0194] in, represents the velocity of the i-th particle at the k+1th update, represents the velocity of the i-th particle at the k-th update, represents the position of the i-th particle at the k+1th update, represents the position of the i-th particle at the k-th update, represents the individual extreme value of the i-th particle at the k-th update, represents the group extreme value at the kth update, c1 is the individual learning factor in the PSO algorithm, c2 is the social learning factor in the PSO algorithm, and w k is the inertia weight factor of the PSO algorithm at the kth update, and rand(0,1) represents a random value between 0 and 1;

[0195] When the termination condition of the PSO algorithm is reached, the updating is stopped and the group extreme value at the last update is obtained as the target data selected from the environmental feature time series data.

[0196] In a possible implementation, the device further includes:

[0197] The parameter determination module is configured to calculate the inertia weight factor w in the PSO algorithm according to the following formula k :

[0198]

[0199] Among them, w min and w max is the minimum and maximum value of the preset inertia weight factor, n is the sliding window of the inertia weight factor, w j is the inertia weight factor at the jth update, β is the preset descent control parameter, T max is the maximum evolutionary generation;

[0200] The individual learning factor c1 and the social learning factor c2 in the PSO algorithm are calculated according to the following formula:

[0201]

[0202] Among them, cmin and c max The minimum and maximum values ​​of the preset learning factor.

[0203] In a possible implementation, the device further includes:

[0204] A receiving module is configured to receive a handling strategy table issued by a master station device, wherein the handling strategy table records handling strategies corresponding to wildfire risk prediction results;

[0205] The strategy execution module is configured to execute the handling strategy corresponding to the wildfire risk prediction result based on the handling strategy table.

[0206] In a possible implementation, the device further includes:

[0207] A data reporting module is configured to send the data fusion feature and its corresponding wildfire risk prediction result to the master station device so that the master station device can re-predict a new wildfire risk prediction result and send the new wildfire risk prediction result to the IoT feeder terminal;

[0208] A result receiving module is configured to receive new wildfire risk prediction results issued by the master station device;

[0209] The new strategy execution module is configured to execute the handling strategy corresponding to the new wildfire risk prediction result based on the handling strategy table.

[0210] In a possible implementation, the device further includes:

[0211] The model structure module is configured to receive an updated DS evidence theory model issued by a master station device, wherein the updated DS evidence theory model is trained by the master station device based on the data fusion features uploaded by the IoT feeder terminal and the actual wildfire results.

[0212] The present disclosure also provides an environmental information processing device for a power distribution line, Figure 4 The following is a block diagram showing the structure of the environment information processing device for the power distribution line of the master station device provided by the embodiment of the present disclosure. The device can be implemented as part or all of the electronic device through software, hardware or a combination of both. Figure 4 The environmental information processing device of the power distribution line includes:

[0213] The data receiving module 401 is configured to receive data fusion features uploaded by each IoT feeder terminal and its corresponding wildfire risk prediction result, wherein the data fusion features are obtained by the IoT feeder terminal performing data selection and feature extraction fusion of an online sequential extreme learning machine (OS-ELM) model on multiple environmental information time series data of the local area where the IoT feeder terminal is located collected by multiple sensors; the wildfire risk prediction result is obtained by the IoT feeder terminal inputting the data fusion features into a pre-stored DS evidence theory model and executing the DS evidence theory model;

[0214] The result determination module 402 is configured to determine whether the wildfire risk prediction result of each IoT feeder terminal is accurate based on the data fusion characteristics uploaded by each IoT feeder terminal and its corresponding wildfire risk prediction result;

[0215] The result sending module 403 is configured to obtain a new wildfire risk prediction result and send the new wildfire risk prediction result to the IoT feeder terminal when the wildfire risk prediction result of the IoT feeder terminal is inaccurate.

[0216] In a possible implementation, the result determination module is configured to:

[0217] For each IoT feeder terminal, obtaining wildfire risk prediction results of other IoT feeder terminals within a predetermined distance range from the IoT feeder terminal;

[0218] Determining whether the wildfire risk prediction result of the IoT feeder terminal is correct based on the wildfire risk prediction results of the other IoT feeder terminals;

[0219] The part of the result sending module that obtains a new wildfire risk prediction result when the wildfire risk prediction result of the IoT feeder terminal is inaccurate is configured as follows:

[0220] When the wildfire risk prediction result of the Internet of Things feeder terminal is inaccurate, a new wildfire risk prediction result of the Internet of Things feeder terminal is determined based on the wildfire risk prediction results of the other Internet of Things feeder terminals.

[0221] In a possible implementation, the result determination module is configured to:

[0222] For each IoT feeder terminal, obtaining data fusion features of other IoT feeder terminals within a predetermined distance range from the IoT feeder terminal;

[0223] Inputting the data fusion features of the other IoT feeder terminals and the data fusion features of the IoT feeder terminal into the DS evidence theory model to obtain a new wildfire risk prediction result output by the DS evidence theory model;

[0224] Compare the wildfire risk prediction result of the IoT feeder terminal with the new wildfire risk prediction result to determine whether the wildfire risk prediction result of the IoT feeder terminal is accurate.

[0225] In a possible implementation, the device further includes:

[0226] A model training module is configured to train the DS evidence theory model based on the data fusion features uploaded by each IoT feeder terminal and its corresponding real wildfire results to obtain an updated DS evidence theory model;

[0227] The model delivery module is configured to deliver the updated DS evidence theory model to each IoT feeder terminal.

[0228] The technical terms and technical features mentioned in the implementation of this device are the same as or similar to those mentioned in the implementation of the above method. For the interpretation and description of the technical terms and technical features involved in this device, please refer to the explanation of the implementation of the above method, and no further details will be given here.

[0229] The present disclosure also discloses an electronic device, Figure 5 A structural block diagram of an electronic device according to an embodiment of the present disclosure is shown.

[0230] like Figure 5 As shown, the electronic device 500 includes a memory 501 and a processor 502, wherein the memory 501 is used to store one or more computer instructions, wherein the one or more computer instructions are executed by the processor 502 to implement the method according to the embodiment of the present disclosure.

[0231] Figure 6 A schematic diagram showing the structure of a computer system suitable for implementing the method of the embodiment of the present disclosure is shown.

[0232] like Figure 6 As shown, the computer system 600 includes a processing unit 601, which can execute various processes in the above-mentioned embodiments according to a program stored in a read-only memory (ROM) 602 or a program loaded from a storage unit 608 into a random access memory (RAM) 603. Various programs and data required for the operation of the computer system 600 are also stored in the RAM 603. The processing unit 601, the ROM 602, and the RAM 603 are connected to each other via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0233] The following components are connected to the I / O interface 605: an input section 606 including a keyboard, a mouse, etc.; an output section 607 including a cathode ray tube (CRT), a liquid crystal display (LCD), a speaker, etc.; a storage section 608 including a hard disk, etc.; and a communication section 609 including a network interface card such as a LAN card, a modem, etc. The communication section 609 performs communication processing via a network such as the Internet. A drive 610 is also connected to the I / O interface 605 as needed. A removable medium 611, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 610 as needed so that a computer program read therefrom can be installed into the storage section 608 as needed. Among them, the processing unit 601 can be implemented as a processing unit such as a CPU, a GPU, a TPU, an FPGA, or an NPU.

[0234] In particular, according to embodiments of the present disclosure, the methods described above can be implemented as computer software programs. For example, embodiments of the present disclosure include a computer program product comprising computer instructions that, when executed by a processor, implement the method steps described above. In such embodiments, the computer program product can be downloaded and installed from a network via the communication portion 609 and / or installed from a removable medium 611.

[0235] The flowcharts and block diagrams in the accompanying drawings illustrate the possible implementation architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each box in the flowchart or block diagram can represent a module, program segment or part of code, and the module, program segment or part of code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in an order different from that marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flowchart, and the combination of boxes in the block diagram and / or flowchart, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or can be implemented using a combination of dedicated hardware and computer instructions.

[0236] The units or modules involved in the embodiments described in this disclosure may be implemented by software or programmable hardware. The units or modules described may also be provided in a processor, and the names of these units or modules do not, in certain circumstances, constitute limitations on the units or modules themselves.

[0237] As another aspect, the present disclosure further provides a computer-readable storage medium. This computer-readable storage medium may be included in the electronic device or computer system described in the above embodiments, or may be a standalone computer-readable storage medium not incorporated into the device. The computer-readable storage medium stores one or more programs, which are used by one or more processors to execute the methods described in the present disclosure.

[0238] The above description is merely a preferred embodiment of the present disclosure and an illustration of the technical principles employed. Those skilled in the art should understand that the scope of the invention herein is not limited to the technical solutions formed by the specific combination of the above-mentioned technical features, but also encompasses other technical solutions formed by any combination of the above-mentioned technical features or their equivalents without departing from the inventive concept. For example, a technical solution formed by replacing the above-mentioned features with (but not limited to) technical features having similar functions disclosed in this disclosure.

Claims

1. A method for processing environmental information of a power distribution line, characterized in that: Applied to IoT feeder terminals, including: Acquire multiple pieces of environmental information time series data of the local area where the IoT feeder terminal is located, collected by multiple sensors; Select target data from each piece of environmental information time series data; Inputting a plurality of selected target data into an online sequential extreme learning machine (OS-ELM) model to obtain data fusion features output by the OS-ELM model; Inputting the data fusion features into a pre-stored DS evidence theory model to obtain a wildfire risk prediction result output by the DS evidence theory model; The data fusion feature and its corresponding wildfire risk prediction result are sent to the master station device, so that the master station device obtains the wildfire risk prediction results of other Internet of Things feeder terminals within a predetermined distance range from the Internet of Things feeder terminal; based on the wildfire risk prediction results of the other Internet of Things feeder terminals, it is determined whether the wildfire risk prediction result of the Internet of Things feeder terminal is correct; and when the wildfire risk prediction result of the Internet of Things feeder terminal is inaccurate, based on the wildfire risk prediction results of the other Internet of Things feeder terminals, a new wildfire risk prediction result of the Internet of Things feeder terminal is determined, and the new wildfire risk prediction result is sent to the Internet of Things feeder terminal; Receive the new wildfire risk prediction result.

2. The method according to claim 1, characterized in that The step of selecting target data from each piece of environmental information time series data includes: Preprocessing the plurality of environmental information time series data to obtain a plurality of environmental feature time series data that meet prediction requirements; The particle swarm optimization (PSO) algorithm is used to select the target data from each environmental feature time series data.

3. The method according to claim 2, characterized in that The method of selecting the target data from each environmental feature time series data using the particle swarm optimization (PSO) algorithm includes: For each piece of environmental characteristic time series data, the environmental characteristic time series data is simulated as a particle swarm, and the speed and position of each particle in the particle swarm are determined according to the following formula: in, represents the velocity of the i-th particle at the k+1th update, represents the velocity of the i-th particle at the k-th update, represents the position of the i-th particle at the k+1th update, represents the position of the i-th particle at the k-th update, represents the individual extreme value of the i-th particle at the k-th update, represents the group extreme value at the kth update, c1 is the individual learning factor in the PSO algorithm, c2 is the social learning factor in the PSO algorithm, and w k is the inertia weight factor of the PSO algorithm at the kth update, and rand(0,1) represents a random value between 0 and 1; When the termination condition of the PSO algorithm is reached, the updating is stopped and the group extreme value at the last update is obtained as the target data selected from the environmental feature time series data.

4. The method according to claim 3, characterized in that The method further comprises: The inertia weight factor w in the PSO algorithm is calculated according to the following formula k : Among them, w min and w max is the minimum and maximum value of the preset inertia weight factor, n is the sliding window of the inertia weight factor, w j is the inertia weight factor at the jth update, β is the preset descent control parameter, T max is the maximum evolutionary generation; The individual learning factor c1 and the social learning factor c2 in the PSO algorithm are calculated according to the following formula: Among them, c min and c max The minimum and maximum values ​​of the preset learning factor.

5. The method according to claim 1, wherein The method further comprises: Receive a disposal strategy table issued by a master station device, wherein the disposal strategy table records disposal strategies corresponding to wildfire risk prediction results; Based on the handling strategy table, the handling strategy corresponding to the wildfire risk prediction result is executed.

6. The method according to claim 5, characterized in that The method further comprises: When receiving a new wildfire risk prediction result sent by the master station device, the handling strategy corresponding to the new wildfire risk prediction result is executed based on the handling strategy table.

7. The method according to claim 5, characterized in that The method further comprises: An updated DS evidence theory model is received from a master station device, where the updated DS evidence theory model is trained by the master station device based on data fusion features uploaded by the IoT feeder terminal and actual wildfire results.

8. A method for processing environmental information of a power distribution line, characterized in that: Applicable to master station equipment, including: Receive data fusion features uploaded by each IoT feeder terminal and its corresponding wildfire risk prediction result, wherein the data fusion features are obtained by the IoT feeder terminal performing data selection and feature extraction fusion of an online sequential extreme learning machine (OS-ELM) model on multiple environmental information time series data of the local area where the IoT feeder terminal is located collected by multiple sensors; the wildfire risk prediction result is obtained by the IoT feeder terminal inputting the data fusion features into a pre-stored DS evidence theory model and executing the DS evidence theory model; For each IoT feeder terminal, obtaining wildfire risk prediction results of other IoT feeder terminals within a predetermined distance range from the IoT feeder terminal; Determining whether the wildfire risk prediction result of the IoT feeder terminal is correct based on the wildfire risk prediction results of the other IoT feeder terminals; When the wildfire risk prediction result of the Internet of Things feeder terminal is inaccurate, a new wildfire risk prediction result of the Internet of Things feeder terminal is determined based on the wildfire risk prediction results of the other Internet of Things feeder terminals.

9. The method according to claim 8, characterized in that The determining whether the wildfire risk prediction result of the IoT feeder terminal is accurate based on the data fusion characteristics uploaded by each IoT feeder terminal and its corresponding wildfire risk prediction result includes: For each IoT feeder terminal, obtaining data fusion features of other IoT feeder terminals within a predetermined distance range from the IoT feeder terminal; Inputting the data fusion features of the other IoT feeder terminals and the data fusion features of the IoT feeder terminal into the DS evidence theory model to obtain a new wildfire risk prediction result output by the DS evidence theory model; Compare the wildfire risk prediction result of the IoT feeder terminal with the new wildfire risk prediction result to determine whether the wildfire risk prediction result of the IoT feeder terminal is accurate.

10. The method according to claim 8, characterized in that The method further comprises: Based on the data fusion features uploaded by each IoT feeder terminal and its corresponding real wildfire results, the DS evidence theory model is trained to obtain an updated DS evidence theory model; The updated DS evidence theory model is distributed to each IoT feeder terminal.

11. An environmental information processing device for a power distribution line, characterized in that: Applied to IoT feeder terminals, including: An acquisition module is configured to acquire a plurality of environmental information time series data of a local area where the IoT feeder terminal is located, collected by a plurality of sensors; A selection module is configured to select target data from each piece of environmental information time series data; a fusion module configured to input a plurality of selected target data into an online sequential extreme learning machine (OS-ELM) model to obtain data fusion features output by the OS-ELM model; A prediction module is configured to input the data fusion features into a pre-stored DS evidence theory model to obtain a wildfire risk prediction result output by the DS evidence theory model; The data reporting module is configured to send the data fusion feature and its corresponding wildfire risk prediction result to the master station device, so that the master station device obtains the wildfire risk prediction results of other Internet of Things feeder terminals within a predetermined distance range from the Internet of Things feeder terminal; based on the wildfire risk prediction results of the other Internet of Things feeder terminals, determine whether the wildfire risk prediction result of the Internet of Things feeder terminal is correct; and when the wildfire risk prediction result of the Internet of Things feeder terminal is inaccurate, determine a new wildfire risk prediction result of the Internet of Things feeder terminal based on the wildfire risk prediction results of the other Internet of Things feeder terminals, and send the new wildfire risk prediction result to the Internet of Things feeder terminal; The result receiving module is configured to receive the new wildfire risk prediction result sent by the master station device.

12. The device according to claim 11, characterized in that The selection module is configured to: Preprocessing the plurality of environmental information time series data to obtain a plurality of environmental feature time series data that meet prediction requirements; The particle swarm optimization (PSO) algorithm is used to select the target data from each environmental feature time series data.

13. The device according to claim 12, characterized in that The selection module adopts the particle swarm optimization (PSO) algorithm to select the target data from each environmental feature time series data, and is configured as follows: For each piece of environmental characteristic time series data, the environmental characteristic time series data is simulated as a particle swarm, and the speed and position of each particle in the particle swarm are determined according to the following formula: in, represents the velocity of the i-th particle at the k+1th update, represents the velocity of the i-th particle at the k-th update, represents the position of the i-th particle at the k+1th update, represents the position of the i-th particle at the k-th update, represents the individual extreme value of the i-th particle at the k-th update, represents the group extreme value at the kth update, c1 is the individual learning factor in the PSO algorithm, c2 is the social learning factor in the PSO algorithm, and w k is the inertia weight factor of the PSO algorithm at the kth update, and rand(0,1) represents a random value between 0 and 1; When the termination condition of the PSO algorithm is reached, the updating is stopped and the group extreme value at the last update is obtained as the target data selected from the environmental feature time series data.

14. The device according to claim 13, characterized in that The device further comprises: The parameter determination module is configured to calculate the inertia weight factor w in the PSO algorithm according to the following formula k : Among them, w min and w max is the minimum and maximum value of the preset inertia weight factor, n is the sliding window of the inertia weight factor, w j is the inertia weight factor at the jth update, β is the preset descent control parameter, T max is the maximum evolutionary generation; The individual learning factor c1 and the social learning factor c2 in the PSO algorithm are calculated according to the following formula: Among them, c min and c max The minimum and maximum values ​​of the preset learning factor.

15. The device according to claim 11, characterized in that The device further comprises: A receiving module is configured to receive a handling strategy table issued by a master station device, wherein the handling strategy table records handling strategies corresponding to wildfire risk prediction results; The strategy execution module is configured to execute the handling strategy corresponding to the wildfire risk prediction result based on the handling strategy table.

16. The device according to claim 15, characterized in that The device further comprises: The new strategy execution module is configured to execute the handling strategy corresponding to the new wildfire risk prediction result based on the handling strategy table.

17. The device according to claim 16, characterized in that The device further comprises: The model structure module is configured to receive an updated DS evidence theory model issued by a master station device, wherein the updated DS evidence theory model is trained by the master station device based on the data fusion features uploaded by the IoT feeder terminal and the actual wildfire results.

18. An environmental information processing device for a power distribution line, characterized in that: Applicable to master station equipment, including: The data receiving module is configured to receive data fusion features uploaded by each IoT feeder terminal and its corresponding wildfire risk prediction result, wherein the data fusion features are obtained by the IoT feeder terminal performing data selection and feature extraction fusion of an online sequential extreme learning machine (OS-ELM) model on multiple environmental information time series data of the local area where the IoT feeder terminal is located collected by multiple sensors; the wildfire risk prediction result is obtained by the IoT feeder terminal inputting the data fusion features into a pre-stored DS evidence theory model and executing the DS evidence theory model; The result determination module is configured to obtain, for each IoT feeder terminal, wildfire risk prediction results of other IoT feeder terminals within a predetermined distance range from the IoT feeder terminal; and determine whether the wildfire risk prediction result of the IoT feeder terminal is correct based on the wildfire risk prediction results of the other IoT feeder terminals; The result sending module is configured to determine a new wildfire risk prediction result of the Internet of Things feeder terminal based on the wildfire risk prediction results of other Internet of Things feeder terminals when the wildfire risk prediction result of the Internet of Things feeder terminal is inaccurate, and send the new wildfire risk prediction result to the Internet of Things feeder terminal.

19. The device according to claim 18, characterized in that The result determination module is configured as follows: For each IoT feeder terminal, obtaining data fusion features of other IoT feeder terminals within a predetermined distance range from the IoT feeder terminal; Inputting the data fusion features of the other IoT feeder terminals and the data fusion features of the IoT feeder terminal into the DS evidence theory model to obtain a new wildfire risk prediction result output by the DS evidence theory model; Compare the wildfire risk prediction result of the IoT feeder terminal with the new wildfire risk prediction result to determine whether the wildfire risk prediction result of the IoT feeder terminal is accurate.

20. The device according to claim 18, characterized in that The device further comprises: A model training module is configured to train the DS evidence theory model based on the data fusion features uploaded by each IoT feeder terminal and its corresponding real wildfire results to obtain an updated DS evidence theory model; The model delivery module is configured to deliver the updated DS evidence theory model to each IoT feeder terminal.

21. An electronic device, characterized in that: The method comprises a memory and a processor, wherein the memory is used to store one or more computer instructions, wherein the one or more computer instructions are executed by the processor to implement the method according to any one of claims 1 to 10.

22. A readable storage medium, characterized in that Computer instructions are stored thereon, and when the computer instructions are executed by a processor, the method according to any one of claims 1 to 10 is implemented.

Images