Distribution room environment anomaly detection method and device, electronic equipment and storage medium

Abstract

The invention provides a power distribution room environment anomaly detection method and device, electronic equipment and a storage medium, and relates to the technical field of environment monitoring, the method comprises the following steps: obtaining environment data of a power distribution room for data preprocessing to obtain a standard data set, the environment data comprising various environment factor data; performing feature extraction on the standard data set to obtain a feature data set; respectively training an isolated forest model and a long short-term memory network model according to the standard data set and the feature data set; and utilizing the trained isolated forest model and the trained long-short-term memory network model to carry out environment anomaly detection according to the environment data, and sending anomaly early warning information. The abnormal environment of the distribution room is detected in combination with the isolated forest model and the long-short-term memory network model, instant information of current environment parameters is fully utilized, and the accuracy and response speed of abnormal detection are remarkably improved through deep time sequence analysis.

Description

Technical Field

[0001] The present invention relates to the technical field of environmental monitoring, and in particular to a method, device, electronic equipment and storage medium for detecting environmental anomalies in a power distribution room. Background Art

[0002] As an important part of the power system, the stability of the internal environment of the power distribution room is crucial to the reliable operation of the entire power grid, and is directly related to the normal order of urban operation and the safety and quality of residents' lives. Once an abnormality occurs in the power distribution room, whether it is a power outage caused by damage to power equipment, or more serious safety accidents such as fire and explosion, it will have an impact on society that cannot be ignored. Therefore, it is necessary to conduct reliable status monitoring of the power distribution room, accurately identify its abnormalities, and eliminate safety hazards in a timely manner to avoid the occurrence of safety accidents.

[0003] The environment of many power distribution rooms lacks effective supervision. If the temperature rises too fast, overloads and partial discharges are not handled in time, it is easy to cause property losses and safety accidents. In the existing technology, it usually relies on a single monitoring method to collect environmental information through specific gas sensors, temperature and humidity sensors and light intensity sensors, and perform a single threshold alarm. However, when facing a complex and changeable operating environment, the existing technology lacks a comprehensive abnormality detection method, which easily leads to omissions or false warnings. Summary of the invention

[0004] In view of the above-mentioned deficiencies in the prior art, the present invention provides a method, device, electronic device and storage medium for detecting abnormalities in a distribution room environment, which effectively solves the problem that the prior art is prone to omissions or false warnings when facing a complex and changeable distribution room operating environment.

[0005] In a first aspect, the present invention provides a method for detecting an abnormality in a distribution room environment, the method comprising: Acquire environmental data of the power distribution room for data preprocessing to obtain a standard data set, wherein the environmental data includes multiple environmental factor data; Performing feature extraction on the standard data set to obtain a feature data set; According to the standard data set and the feature data set, an isolation forest model and a long short-term memory network model are trained respectively; The trained isolation forest model and the trained long short-term memory network model are used to perform environmental anomaly detection according to the environmental data, and anomaly warning information is sent.

[0006] Furthermore, the acquisition of the environmental data of the power distribution room is performed on the data preprocessing to obtain a standard data set, including: Acquire the environmental factor data collected by multiple sensors in the power distribution room, wherein the environmental factor data at least includes temperature factor data, humidity factor data, noise factor data, vibration factor data, smoke factor data and electromagnetic intensity factor data; Performing data cleaning on the environmental factor data; The cleaned environmental factor data are standardized to obtain a standard data set.

[0007] Furthermore, extracting features from the standard data set to obtain a feature data set includes: Calculate the change rate of each environmental factor data according to the standard data set; The statistical characteristics of each environmental factor data are calculated according to the sliding window of the standard data set, and the statistical characteristics at least include mean, variance, maximum value and minimum value.

[0008] Furthermore, the training of the isolation forest model and the long short-term memory network model according to the standard data set and the feature data set respectively includes: Constructing the isolation forest model; Inputting the standard data set and the change rate into the isolation forest model to obtain an abnormality scoring result; The parameters of the isolation forest model are optimized according to the abnormality scoring result to obtain the trained isolation forest model.

[0009] Furthermore, the training of the isolation forest model and the long short-term memory network model according to the standard data set and the feature data set respectively further includes: Constructing the long short-term memory network model; Inputting the standard data set, the change rate and the statistical features into the long short-term memory network model to obtain a prediction result; The long short-term memory network model is optimized according to the prediction result to obtain the trained long short-term memory network model.

[0010] Furthermore, the use of the trained isolation forest model and the long short-term memory network model to perform environmental anomaly detection according to the environmental data and send abnormal warning information includes: Inputting the environmental data into the trained isolation forest model to detect abnormal points in the environmental data; Inputting the environmental data into the trained long short-term memory network model to obtain predicted environmental parameters; If the isolation forest model detects the abnormal point, and / or the deviation value between the environmental data and the predicted environmental parameter exceeds a set threshold, the abnormal warning information is sent.

[0011] Furthermore, the method further comprises: The parameters of the isolation forest model and / or the long short-term memory network model are adjusted according to the abnormal warning information.

[0012] In a second aspect, the present invention provides a device for detecting abnormal environment in a power distribution room, the device comprising: A data processing module is used to obtain environmental data of the power distribution room for data preprocessing to obtain a standard data set, wherein the environmental data includes multiple environmental factor data; A feature extraction module, used to extract features from the standard data set to obtain a feature data set; A model training module, used for training an isolation forest model and a long short-term memory network model according to the standard data set and the feature data set respectively; The anomaly detection module is used to use the trained isolation forest model and the trained long short-term memory network model to perform environmental anomaly detection according to the environmental data and send abnormal warning information.

[0013] In a third aspect, the present invention provides an electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the method for detecting anomalies in a distribution room environment as described in the first aspect of the present invention.

[0014] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the method for detecting anomalies in a distribution room environment as described in the first aspect of the present invention.

[0015] The method, device, electronic device and storage medium for detecting abnormal environment in the distribution room provided by the present invention combine the isolation forest model with the long short-term memory network model to detect abnormal environment in the distribution room, which not only makes full use of the real-time information of the current environmental parameters, but also significantly enhances the accuracy and response speed of abnormal detection through in-depth time series analysis. By extracting multi-dimensional features from environmental data, rich input information is formed, which provides a solid foundation for the training and analysis of subsequent models. The present invention can monitor various environmental parameters in real time and detect potential abnormal situations in a timely manner. Through the comprehensive use of the two models, the false alarm rate is significantly reduced, the reliability of the early warning is improved, and the occurrence of equipment failures and safety accidents can be effectively prevented, ensuring that the equipment in the distribution room is in a safe and stable operating state. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0017] Figure 1 It is a first schematic diagram of the process of the method for detecting abnormal environment in a power distribution room provided by an embodiment of the present invention; Figure 2 It is a second schematic diagram of the process of the method for detecting abnormal environment in a power distribution room provided by an embodiment of the present invention; Figure 3 3 is a schematic diagram of the process of the method for detecting abnormal environment in a power distribution room provided by an embodiment of the present invention; Figure 4 It is a fourth schematic diagram of the process of the method for detecting abnormal environment in a power distribution room provided by an embodiment of the present invention; Figure 5 is a fifth schematic diagram of a method for detecting abnormal environment in a power distribution room provided by an embodiment of the present invention; Figure 6 6 is a schematic diagram of a method for detecting abnormal environment in a power distribution room provided by an embodiment of the present invention; Figure 7 7 is a schematic diagram of a method for detecting abnormal environment in a power distribution room provided by an embodiment of the present invention; Figure 8 It is a schematic diagram of the structure of a device for detecting abnormal environment in a power distribution room provided by an embodiment of the present invention; Fig. 9 It is a structural schematic diagram of an electronic device provided by an embodiment of the present invention.

[0018] Description of main component symbols: 800. Distribution room environment anomaly detection device; 810. Data processing module; 820. Feature extraction module; 830. Model training module; 840. Anomaly detection module; 900. Electronic device; 910. Processor; 920. Communication interface; 930. Memory; 940. Communication bus. DETAILED DESCRIPTION

[0019] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be further clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. It should be noted that the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0020] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0021] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application belongs. The terms used in this specification are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The term "and / or" used herein includes any and all combinations of one or more of the related listed items.

[0022] At present, the environment of many power distribution rooms lacks effective supervision. If the temperature rise, overload and partial discharge are not handled in time, it is easy to cause property loss and safety accidents. In the existing technology, it usually relies on a single monitoring method to collect environmental information through specific gas sensors, temperature and humidity sensors and light intensity sensors, and perform a single threshold alarm. However, when facing a complex and changeable operating environment, the existing technology lacks a comprehensive abnormality detection method, which easily leads to omissions or false warnings.

[0023] Example 1 The embodiment of the present invention provides a method for detecting abnormalities in a power distribution room environment, which effectively solves the problem that the prior art is prone to omissions or false warnings when faced with a complex and changeable power distribution room operating environment. Figure 1 FIG. 1 is a first schematic diagram of a method for detecting abnormal environment in a power distribution room according to an embodiment of the present invention. Figure 1 As shown, the method comprises the following steps: S100, obtaining environmental data of a power distribution room and performing data preprocessing to obtain a standard data set, wherein the environmental data includes data of multiple environmental factors.

[0024] Figure 2 2 is a second schematic diagram of the process of the method for detecting abnormal environment in a power distribution room provided by an embodiment of the present invention. Figure 2 As shown, data preprocessing specifically includes the following steps: S110. Obtain environmental factor data collected by various sensors in the power distribution room.

[0025] In an embodiment of the present invention, environmental factor data in a distribution room is collected by a variety of sensors, including but not limited to temperature sensors, humidity sensors, noise sensors, vibration sensors, smoke sensors and electromagnetic sensors, and the corresponding environmental factor data at least includes temperature factor data, humidity factor data, noise factor data, vibration factor data, smoke factor data and electromagnetic intensity factor data.

[0026] S120, performing data cleaning on environmental factor data.

[0027] In the embodiment of the present invention, data cleaning mainly includes removing missing values ​​and outliers, wherein the missing values ​​can be removed by filling methods, which include but are not limited to mean filling, median filling, mode filling, forward filling, backward filling and interpolation. Outliers can be removed by statistical-based methods, box plot-based methods and model-based methods. According to the characteristics of the data and analysis requirements, the appropriate method can be selected to clean the data more effectively.

[0028] S130, performing standardization processing on the cleaned environmental factor data to obtain a standard data set.

[0029] In the embodiment of the present invention, the standard data set is obtained by standardizing the environmental factor data after cleaning to the same range. Optionally, the embodiment of the present invention adopts standard deviation standardization to obtain the standard data set, and the conversion formula is as follows:

[0030] In the above formula, Z represents the standardized standard value, and X represents the original environmental factor data. μ represents the mean of the original environmental factor data set, σ Represents the standard deviation of the original environmental factor data set.

[0031] By cleaning and standardizing the collected environmental data, we can ensure the accuracy and reliability of the data, eliminate magnitude differences, unify data scales, improve algorithm performance, and maintain data distribution.

[0032] S200: Extract features from a standard data set to obtain a feature data set.

[0033] Figure 3 3 is a third schematic diagram of the process of the method for detecting abnormal environment in a power distribution room provided by an embodiment of the present invention. Figure 3 As shown, feature extraction specifically includes the following steps: S210. Calculate the change rate of each environmental factor data according to the standard data set.

[0034] The rate of change can intuitively show the degree of data fluctuation over time. By analyzing the rate of change, the future trend of data changes can be predicted. The calculation formula for the rate of change is:

[0035] In the above formula, X rate represents the rate of change, X t Represents the environmental factor data at the current moment, X t-1 Represents the environmental factor data at the previous moment.

[0036] S220. Calculate the statistical characteristics of each environmental factor data according to the sliding window of the standard data set.

[0037] In an embodiment of the present invention, the statistical features include at least mean, variance, maximum and minimum values, and these statistical features can be calculated by a sliding window based on a standard data set. The length of the sliding window determines the number of time steps contained in the sliding window, which is usually set according to the periodicity and short-term trend of the data. The sliding window step is usually set to 1, and a larger step can also be set as needed. The sliding window technology is used to effectively analyze the environmental factor data in the standard data set and extract useful statistical features, which can provide valuable input for further data analysis and modeling, and improve the accuracy of detection.

[0038] S300, training the isolation forest model and the long short-term memory network model according to the standard data set and the feature data set respectively.

[0039] Figure 4 4 is a schematic diagram of the process of the method for detecting abnormal environment in a power distribution room provided by an embodiment of the present invention. Figure 4 As shown in Figure 1, the training of the isolation forest model specifically includes the following steps: S310. Construct an isolation forest model.

[0040] The isolation forest model is mainly used to detect abnormal behavior in multidimensional data. It divides the data space by randomly selecting a feature and a feature value until all data points are isolated. Compared with normal samples, abnormal samples are less in number and have large differences in feature values, so they are more easily isolated, thereby using the characteristic that abnormal points are easily isolated for anomaly detection. In an embodiment of the present invention, a portion of samples are randomly selected from a standard data set as a training set, and multiple isolated trees are constructed iteratively to form an isolation forest, where each tree is constructed independently, and samples and features are randomly selected each time they are constructed.

[0041] S320, input the standard data set and the change rate into the isolation forest model to obtain anomaly scoring results.

[0042] The original value and change rate in the standard data set are used as the input features of the isolation forest model to train the isolation forest model. The normal and abnormal samples are distinguished by constructing random trees. Each tree in the isolation forest divides the sample space by random subsampling. The calculation formula of the anomaly score of the isolation forest is as follows:

[0043] In the above formula, S(x) represents the abnormality score, E(x) Representation sample x The expected path length, c(n) Represents the number of normal samples.

[0044] S330. Optimize the parameters of the isolation forest model according to the anomaly scoring result to obtain a trained isolation forest model.

[0045] In the embodiment of the present invention, the number of isolation forest model trees, the size of subsamples and the anomaly score threshold are mainly optimized, wherein: The generalization ability of the model can be improved by increasing the number of trees. For example, assuming that 100 trees are used in the initial training, but in the evaluation of the isolation forest model, it is found that fewer anomalies are detected, indicating that the isolation forest model is not sensitive enough to data changes. By increasing the number of trees to 300, the isolation forest model can better cover the diversity of the data and detect more potential abnormal samples. Increasing the number of trees can make the isolation forest model have higher detection accuracy, because more trees can better explore the different characteristics of the data and reduce the possibility of missing anomalies.

[0046] A reasonable subsample size can improve the ability to identify outliers. For example, the initial setting is to use all samples for training, but the training time is too long. By setting the subsample size to 200, the isolation forest model only uses part of the samples for training, which maintains the detection effect while greatly shortening the training time. Adjusting the subsample size helps to reduce the training time of the model, and when the amount of data is large, a reasonable choice of subsample size will not significantly affect the accuracy of the model.

[0047] The accuracy of the model can be improved by dynamically adjusting the anomaly scoring threshold. For example, if the anomaly scoring threshold is initially set to 0.1, an anomaly warning will be issued if 10% of the data is considered abnormal, which may result in a high false alarm rate. By lowering the anomaly scoring threshold to 0.05, false alarms are reduced, ensuring that anomaly warnings are issued only in truly abnormal situations, making anomaly detection more accurate and improving the practicality of the model.

[0048] The number of trees, subsample size, and anomaly score threshold are adjusted to the optimal values ​​according to the actual situation to obtain a trained isolation forest model.

[0049] Figure 5 5 is a fifth schematic diagram of the process of the method for detecting abnormal environment in a power distribution room provided by an embodiment of the present invention. Figure 5 As shown in Figure 1, the training of the long short-term memory network model includes the following steps: S340. Construct a long short-term memory network model.

[0050] The LSTM network model is used to capture long-term dependencies in time series and obtain the predicted values ​​of environmental data. It can capture the trends and patterns of data changes over time, which is crucial for the dynamic monitoring of environmental data. The LSTM network model can more accurately understand the evolution of environmental parameters and thus identify anomalies more accurately. The construction of the LSTM network model mainly includes defining the model and compiling the model. Defining the model is to define the hierarchical structure of the network, add the LSTM network layer and specify the number of neurons in the hidden layer, and add an appropriate output layer. Compiling the model is to select the optimizer, loss function and evaluation indicator.

[0051] In the embodiment of the present invention, a multi-layer long short-term memory network model is constructed. The long short-term memory network unit has an input gate, a forget gate, and an output gate for processing time-dependent data. Finally, the output of the long short-term memory network is mapped to the future environmental data prediction through a fully connected layer. The output function of the long short-term memory network model is as follows:

[0052] In the above formula, y t represents the predicted value of environmental data, h t represents the hidden state of the LSTM network, FullyConnected Represents the fully connected layer, which is a linear transformation layer in the neural network. It transforms the hidden state of the long short-term memory network h t Mapped to the output space through linear transformation, the specific formula is:

[0053] In the above formula, W represents the weight matrix of the fully connected layer, b Represents the bias vector, which is optimized W and b , the model can obtain the predicted value of environmental data more accurately.

[0054] Optionally, use mean squared error MSE As the loss function measures the error between the predicted value and the true value, the formula of the loss function is as follows:

[0055] In the above formula, MSE represents the mean square error, n represents the number of samples, y i represents the true value, Represents the predicted value.

[0056] In an embodiment of the present invention, the Adam optimization algorithm is used to adjust the weights and biases of the long short-term memory network model, and the optimization process minimizes the value of the loss function.

[0057] S350, input the standard data set, change rate and statistical features into the long short-term memory network model to obtain the prediction result.

[0058] In the embodiment of the present invention, the original value, change rate and statistical features in the standard data set are used as input features, wherein the change rate of the time series can reflect the trend and fluctuation characteristics, which helps the long short-term memory network model to capture the dynamic changes of the data, and the statistical features are used as auxiliary inputs. The input features are input into the long short-term memory network model in the form of a sliding window in the time series to ensure the capture of the time dependency of the data and finally obtain the prediction results.

[0059] S360. Optimize the long short-term memory network model according to the prediction results to obtain a trained long short-term memory network model.

[0060] In an embodiment of the present invention, the number of hidden layer units, learning rate, batch size and time step of the long short-term memory network model are optimized according to the prediction results, wherein: The more hidden units there are, the stronger the model's memory capacity is. For example, the LSTM network model was initially trained with 50 hidden units, but failed to accurately capture complex time dependencies. By increasing the number of hidden units to 100 through cross-validation, the LSTM network model was able to better learn the long-term dependencies of time series and improve its ability to predict abnormal data. Increasing the number of hidden units allows the LSTM network model to learn more features and improve its adaptability to complex environmental changes and prediction accuracy.

[0061] The learning rate affects the speed and stability of LSTM network model training. A high learning rate will lead to model instability, while a low learning rate will make the model converge slowly. Therefore, it is necessary to find the optimal value by adjusting the learning rate. For example, if the initial learning rate is 0.01, the training results will fluctuate greatly. By reducing the learning rate to 0.001, the model training will be more stable, the convergence speed will also be faster, and finally more accurate prediction results will be obtained. Appropriate learning rate adjustment can prevent the LSTM network model from converging too slowly or falling into a local optimal solution, ensuring that the LSTM network model converges faster and more accurately.

[0062] Reasonable batch size setting can ensure that the appropriate amount of data is available each time the LSTM model is updated, which can improve computational efficiency and maintain training stability. For example, an initial batch size of 32 results in a long training time and the LSTM model does not perform well on the test data. By increasing the batch size to 64, the training speed can be increased, while the generalization ability of the LSTM model can be improved and fluctuations in training can be reduced.

[0063] The time step can ensure that environmental data is captured for a long enough period of time. For example, the initial time step was set to 10, but the model failed to accurately predict the fluctuations in environmental data. By increasing the time step to 20, the LSTM network model can capture data changes over a longer time span and improve the ability to predict future environmental data.

[0064] According to the actual situation, the number of hidden layer units, learning rate, batch size and time step can be adjusted to the optimal value to obtain a trained long short-term memory network model.

[0065] S400, using the trained isolation forest model and the trained long short-term memory network model to perform environmental anomaly detection according to environmental data, and send abnormal warning information.

[0066] Figure 6 6 is a schematic diagram of the method for detecting abnormal environment in a power distribution room according to an embodiment of the present invention. Figure 6 As shown, environmental anomaly detection includes the following steps: S410, input the environmental data into the trained isolation forest model to detect abnormal points in the environmental data.

[0067] In an embodiment of the present invention, based on the feature values ​​of the input environmental data, the trained isolation forest model detects abnormal points in the environmental data.

[0068] S420, input the environmental data into the trained long short-term memory network model to obtain predicted environmental parameters.

[0069] In an embodiment of the present invention, the current actual environmental data is input into a trained long short-term memory network model, and the predicted environmental parameters for the next time step are output.

[0070] S430: If the isolation forest model detects an abnormal point, and / or the deviation value between the environmental data and the predicted environmental parameters exceeds a set threshold, an abnormal warning message is sent.

[0071] The detection results of the isolation forest model are combined with the prediction results of the long short-term memory network model to perform anomaly judgment. When the isolation forest model detects an abnormal point, and / or the deviation value between the current actual environmental data and the predicted environmental parameters exceeds the set threshold, an abnormal warning is triggered and an abnormal warning message is sent. The set threshold can be set according to the actual situation.

[0072] As a preferred implementation of the present invention, Figure 7 7 is a seventh schematic diagram of the process of the method for detecting abnormal environment in a power distribution room provided by an embodiment of the present invention. Figure 7 As shown, the method also includes the following steps: S500: Adjust parameters of the isolation forest model and / or the long short-term memory network model according to the abnormal warning information.

[0073] In an embodiment of the present invention, the user can ignore or confirm the abnormal warning information according to the actual situation, and dynamically adjust the parameters of the model according to the processing feedback of the abnormal warning information, such as the abnormal scoring threshold of the isolation forest model or the learning rate of the long short-term memory network model. The optimized parameters enable the model to better capture environmental changes and improve the detection accuracy and prediction ability of the model. At the same time, through reasonable parameter settings, unnecessary calculations are avoided, the model training time is reduced, and the response speed of the model is improved. The optimized model can better adapt to new data, reduce false positives and false negatives, and improve performance in practical applications.

[0074] The power distribution room environment anomaly detection provided by the embodiment of the present invention combines the isolation forest model with the long short-term memory network model to detect the abnormal environment in the power distribution room, which not only makes full use of the real-time information of the current environmental parameters, but also significantly enhances the accuracy and response speed of anomaly detection through in-depth time series analysis. By extracting multi-dimensional features from environmental data, rich input information is formed, which provides a solid foundation for the training and analysis of subsequent models.

[0075] Example 2 Based on the same technical concept, an embodiment of the present invention provides a device for detecting abnormal environment in a power distribution room. Figure 8 Schematic diagram of the structure of the device for detecting abnormal environment in a power distribution room provided by an embodiment of the present invention. Figure 8 As shown, the power distribution room environment abnormality detection device 800 includes: The data processing module 810 is used to obtain the environmental data of the power distribution room for data preprocessing to obtain a standard data set, where the environmental data includes multiple environmental factor data.

[0076] The feature extraction module 820 is used to extract features from the standard data set to obtain a feature data set.

[0077] The model training module 830 is used to train the isolation forest model and the long short-term memory network model according to the standard data set and the feature data set respectively.

[0078] The anomaly detection module 840 is used to use the trained isolation forest model and the trained long short-term memory network model to perform environmental anomaly detection based on environmental data and send abnormal warning information.

[0079] The power distribution room environment anomaly detection device provided by the embodiment of the present invention can monitor various environmental parameters in real time and detect potential abnormal situations in a timely manner. Through the comprehensive use of the two models, the false alarm rate is significantly reduced, the reliability of the early warning is improved, and the occurrence of equipment failures and safety accidents can be effectively prevented, ensuring that the equipment in the power distribution room is in a safe and stable operating state.

[0080] It can be understood that the implementation method of the distribution room environment abnormality detection method described in the above-mentioned embodiment 1 is also applicable to this embodiment and can achieve the same technical effect, so it will not be repeated here.

[0081] Example 3 Based on the same concept, an embodiment of the present invention further provides an electronic device, Fig. 9 is a schematic diagram of the structure of an electronic device provided by an embodiment of the present invention, such as Fig. 9 As shown, the electronic device 900 may include: a processor 910, a communication interface 920, a memory 930 and a communication bus 940, wherein the processor 910, the communication interface 920 and the memory 930 communicate with each other through the communication bus 940. The processor 910 may call the logic instructions in the memory 930 to execute the steps of the method for detecting abnormal environment in the power distribution room as described in the above embodiments. For example, it includes: S100, obtaining environmental data of the power distribution room for data preprocessing to obtain a standard data set, where the environmental data includes data of multiple environmental factors; S200, extracting features from a standard data set to obtain a feature data set; S300, training the isolation forest model and the long short-term memory network model according to the standard data set and the feature data set respectively; S400, using the trained isolation forest model and the trained long short-term memory network model to perform environmental anomaly detection according to environmental data, and send abnormal warning information.

[0082] The processor 910 may be a central processing unit (CPU). The processor may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or a combination of the above chips.

[0083] In addition, the logic instructions in the above-mentioned memory 930 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0084] The memory 930 may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function; the data storage area may store data created by the processor, etc. In addition, the memory may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some embodiments, the memory may optionally include a memory remotely disposed relative to the processor, and these remote memories may be connected to the processor via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0085] Example 4 Based on the same concept, an embodiment of the present invention further provides a computer-readable storage medium, which stores a computer program, and the computer program includes at least one code segment, which can be executed by a main control device to control the main control device to implement the steps of the power distribution room environment abnormality detection method described in the above embodiments. For example, it includes: S100, obtaining environmental data of the power distribution room for data preprocessing to obtain a standard data set, where the environmental data includes data of multiple environmental factors; S200, extracting features from a standard data set to obtain a feature data set; S300, training the isolation forest model and the long short-term memory network model according to the standard data set and the feature data set respectively; S400, using the trained isolation forest model and the trained long short-term memory network model to perform environmental anomaly detection according to environmental data, and send abnormal warning information.

[0086] Based on the same technical concept, an embodiment of the present invention further provides a computer program, which is used to implement the above method embodiment when executed by a main control device.

[0087] The computer program may be stored in whole or in part on a computer-readable storage medium packaged with the processor, or may be stored in whole or in part on a memory not packaged with the processor.

[0088] Based on the same technical concept, an embodiment of the present invention further provides a processor, which is used to implement the above method embodiment. The above processor can be a chip.

[0089] In summary, the method, device, electronic device and storage medium for detecting abnormal environment in the distribution room provided by the present invention, combined with the isolation forest model and the long short-term memory network model, perform abnormal environment detection in the distribution room, which not only makes full use of the real-time information of the current environmental parameters, but also significantly enhances the accuracy and response speed of abnormal detection through in-depth time series analysis. By extracting multi-dimensional features from environmental data, rich input information is formed, which provides a solid foundation for the training and analysis of subsequent models. The present invention can monitor various environmental parameters in real time and detect potential abnormal situations in a timely manner. Through the comprehensive use of the two models, the false alarm rate is significantly reduced, the reliability of the early warning is improved, and the occurrence of equipment failures and safety accidents can be effectively prevented, ensuring that the equipment in the distribution room is in a safe and stable operating state.

[0090] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0091] The above-mentioned embodiments only express several implementation methods of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the scope of the patent of the present invention. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

[0092] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for detecting abnormal environment in a power distribution room, characterized in that: The method comprises: Acquire environmental data of the power distribution room for data preprocessing to obtain a standard data set, wherein the environmental data includes data of multiple environmental factors; Performing feature extraction on the standard data set to obtain a feature data set; According to the standard data set and the feature data set, an isolation forest model and a long short-term memory network model are trained respectively; The trained isolation forest model and the trained long short-term memory network model are used to perform environmental anomaly detection according to the environmental data, and anomaly warning information is sent.

2. The method for detecting abnormal environment in a power distribution room according to claim 1, characterized in that: The step of obtaining the environmental data of the power distribution room and performing data preprocessing to obtain a standard data set includes: Acquire the environmental factor data collected by multiple sensors in the power distribution room, wherein the environmental factor data at least includes temperature factor data, humidity factor data, noise factor data, vibration factor data, smoke factor data and electromagnetic intensity factor data; Performing data cleaning on the environmental factor data; The cleaned environmental factor data are standardized to obtain a standard data set.

3. The method for detecting abnormal environment in a power distribution room according to claim 1, characterized in that: The step of extracting features from the standard data set to obtain a feature data set includes: Calculate the change rate of each environmental factor data according to the standard data set; The statistical characteristics of each environmental factor data are calculated according to the sliding window of the standard data set, and the statistical characteristics at least include mean, variance, maximum value and minimum value.

4. The method for detecting abnormal environment in a power distribution room according to claim 3, characterized in that: The step of training the isolation forest model and the long short-term memory network model respectively according to the standard data set and the feature data set includes: Constructing the isolation forest model; Inputting the standard data set and the change rate into the isolation forest model to obtain an abnormality scoring result; The parameters of the isolation forest model are optimized according to the abnormality scoring result to obtain the trained isolation forest model.

5. The method for detecting abnormal environment in a power distribution room according to claim 3, characterized in that: The method of training the isolation forest model and the long short-term memory network model according to the standard data set and the feature data set respectively further includes: Constructing the long short-term memory network model; Inputting the standard data set, the change rate and the statistical features into the long short-term memory network model to obtain a prediction result; The long short-term memory network model is optimized according to the prediction result to obtain the trained long short-term memory network model.

6. The method for detecting abnormal environment in a power distribution room according to claim 1, characterized in that: The using the trained isolation forest model and the trained long short-term memory network model to perform environmental anomaly detection according to the environmental data and send abnormal warning information includes: Inputting the environmental data into the trained isolation forest model to detect abnormal points in the environmental data; Inputting the environmental data into the trained long short-term memory network model to obtain predicted environmental parameters; If the isolation forest model detects the abnormal point, and / or the deviation value between the environmental data and the predicted environmental parameter exceeds a set threshold, the abnormal warning information is sent.

7. The method for detecting abnormal environment in a power distribution room according to any one of claims 1 to 6, characterized in that: The method further comprises: The parameters of the isolation forest model and / or the long short-term memory network model are adjusted according to the abnormal warning information.

8. A device for detecting abnormal environment in a power distribution room, characterized in that: The device comprises: A data processing module is used to obtain environmental data of the power distribution room for data preprocessing to obtain a standard data set, wherein the environmental data includes data of multiple environmental factors; A feature extraction module, used to extract features from the standard data set to obtain a feature data set; A model training module, used for training an isolation forest model and a long short-term memory network model according to the standard data set and the feature data set respectively; The anomaly detection module is used to use the trained isolation forest model and the trained long short-term memory network model to perform environmental anomaly detection according to the environmental data and send abnormal warning information.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: The processor executes the computer program to implement the method for detecting abnormal environment in a power distribution room as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for detecting an abnormality in a distribution room environment as described in any one of claims 1 to 7 is implemented.

Images