Intelligent power distribution room monitoring method and system

Abstract

The invention provides an intelligent power distribution room monitoring method and system, and belongs to the technical field of data monitoring, and the method comprises the steps: determining a target prediction model from a plurality of standard prediction models based on historical data distribution characteristics; wherein the historical data distribution characteristics are data distribution characteristics of historical operation data of the target equipment on a time axis, the target equipment is equipment in a power distribution room, and the plurality of standard prediction models are models obtained by training different prediction models based on the historical operation data of the target equipment; inputting the collected operation data of the target equipment into the target prediction model to obtain predicted operation data of the target equipment in the prediction time period; and determining a monitoring strategy of the power distribution room based on the predicted operation data of the target equipment in the prediction time period. According to the intelligent power distribution room monitoring method and system provided by the invention, the accuracy and reliability of power distribution room monitoring can be improved.

Description

Technical Field

[0001] The present disclosure belongs to the field of data monitoring technology, and more specifically, relates to a method and system for monitoring an intelligent power distribution room. Background Art

[0002] With the rapid development of modern society, the stable operation of power systems is crucial to all walks of life and people's daily lives. As a key link in distributing electrical energy in power systems, the safe and reliable operation of the distribution room is directly related to the quality and stability of the entire power supply.

[0003] Existing intelligent monitoring systems have low accuracy and can only perform simple threshold judgments, failing to effectively predict and diagnose potential equipment failures. Therefore, an accurate and reliable distribution room monitoring method is needed. Summary of the Invention

[0004] The purpose of the present disclosure is to provide an intelligent power distribution room monitoring method and system to improve the accuracy and reliability of power distribution room monitoring.

[0005] A first aspect of the embodiments of the present disclosure provides a method for monitoring an intelligent power distribution room, comprising: Determining a target prediction model from multiple standard prediction models based on historical data distribution characteristics; wherein the historical data distribution characteristics are data distribution characteristics of historical operating data of a target device on a time axis, the target device is a device in a power distribution room, and the multiple standard prediction models are models obtained by training different prediction models based on the historical operating data of the target device; Input the collected operating data of the target device into the target prediction model to obtain the predicted operating data of the target device within the prediction period; The monitoring strategy for the distribution room is determined based on the predicted operating data of the target equipment within the prediction period.

[0006] A second aspect of the embodiments of the present disclosure provides an intelligent distribution room monitoring system, including: a model determination module for determining a target prediction model from a plurality of standard prediction models based on historical data distribution characteristics; wherein the historical data distribution characteristics are data distribution characteristics of historical operating data of a target device on a time axis, the target device is a device in a power distribution room, and the plurality of standard prediction models are models obtained by training different prediction models based on the historical operating data of the target device; A data prediction module is used to input the collected operating data of the target device into the target prediction model to obtain the predicted operating data of the target device within the prediction period; The strategy determination module is used to determine the monitoring strategy of the distribution room based on the predicted operating data of the target equipment within the predicted period.

[0007] According to a third aspect of an embodiment of the present disclosure, an electronic device is provided, comprising a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, the steps of the above-mentioned intelligent distribution room monitoring method are implemented.

[0008] According to a fourth aspect of the embodiments of the present disclosure, a computer-readable storage medium is provided, which stores a computer program. When the computer program is executed by a processor, the steps of the above-mentioned intelligent distribution room monitoring method are implemented.

[0009] The beneficial effects of the intelligent distribution room monitoring method and system provided by the embodiments of the present disclosure are: This paper uses historical data distribution characteristics to determine a target prediction model from multiple standard prediction models, enabling the present invention to select the prediction model most suitable for the current device. By inputting the collected operating data of the target device into the target prediction model, the device's operating data for the prediction period can be obtained. Predicting based on the appropriate model can provide a more accurate prediction of the operating status, improving the accuracy and reliability of distribution room monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] In order to more clearly illustrate the technical solutions in the embodiments of the present disclosure, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present disclosure. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0011] Figure 1 A flowchart of a method for monitoring an intelligent power distribution room provided in one embodiment of the present disclosure; Figure 2 This is a structural block diagram of an intelligent distribution room monitoring system provided by one embodiment of the present disclosure; Figure 3 A schematic block diagram of an electronic device provided in one embodiment of the present disclosure. DETAILED DESCRIPTION

[0012] In the following description, specific details such as specific system structures and techniques are provided for purposes of illustration rather than limitation to facilitate a thorough understanding of the embodiments of the present disclosure. However, it will be apparent to those skilled in the art that the present disclosure may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid obscuring the description of the present disclosure with unnecessary detail.

[0013] In order to make the purpose, technical solutions and advantages of the present disclosure more clear, specific embodiments will be described below with reference to the accompanying drawings.

[0014] Please refer to Figure 1 , Figure 1 A flow chart of a method for monitoring an intelligent power distribution room provided in one embodiment of the present disclosure includes: S101: Determine a target prediction model from multiple standard prediction models based on historical data distribution characteristics; wherein the historical data distribution characteristics are data distribution characteristics of the historical operating data of the target device on the time axis, the target device is the device in the distribution room, and the multiple standard prediction models are models obtained by training different prediction models based on the historical operating data of the target device.

[0015] In this embodiment, historical data distribution features refer to various characteristics of the target device's historical operating data on a timeline, such as whether the data is continuous or concentrated within a single time period. Target devices refer to various devices in a power distribution room, such as transformers, distribution cabinets, and circuit breakers. A standard prediction model is a model obtained by training different types of prediction models using the target device's historical operating data. These models can be used to predict the target device's future operating data, such as the autoregressive integrated moving average model (ARIMA), the grey prediction model, the long short-term memory network model, and the vector autoregression model.

[0016] Considering that different models are applicable to different data characteristics, taking the aforementioned ARIMA model and gray prediction model as examples, the ARIMA model, as a statistical model based on time series, requires a certain amount of data to accurately identify trends in the data in order to estimate the model parameters and determine the order. If the amount of data is too small, it will not be able to accurately capture the inherent laws of the data, resulting in poor model fitting and prediction results. The gray prediction model, on the other hand, has relatively low data requirements. By performing generative processing such as accumulation on the raw data, it weakens the randomness of the data and can effectively predict with the inherent connections between the data when the data volume is small.

[0017] Therefore, in the present disclosure, a target prediction model can be determined from multiple standard prediction models based on the distribution characteristics of historical data.

[0018] S102: Input the collected operating data of the target device into a target prediction model to obtain predicted operating data of the target device within a prediction period.

[0019] In this embodiment, various sensors installed on the target equipment in the distribution room, such as current sensors, voltage sensors, temperature sensors, etc., can be used to obtain the operating parameters of the equipment in real time, such as current size, voltage value, equipment temperature, working status and other data.

[0020] By inputting the collected operating data of the target device into the established target prediction model, you can obtain the predicted operating data of the target device within the prediction period. The prediction period can be one hour, one day, etc., and the specific duration is set according to actual needs.

[0021] S103: Determine a monitoring strategy for the power distribution room based on the predicted operating data of the target equipment within the predicted period.

[0022] In this embodiment, the monitoring strategy for the power distribution room is determined based on the predicted operating data of the target equipment within the predicted period, including: determining a failure risk of the target device based on operating data of the target device within a prediction period; In response to the failure risk of the target device being greater than or equal to the first risk, determining the first monitoring strategy as the monitoring strategy for the power distribution room; In response to the failure risk of the target device being less than the first risk, determining the second monitoring strategy as the monitoring strategy for the power distribution room; The monitoring frequency of the first monitoring strategy is greater than the monitoring frequency of the second monitoring strategy.

[0023] In this embodiment, the operating data of the target device within the prediction period may be analyzed and processed by a threshold judgment method, a trend analysis method, or a comparative analysis method to obtain the failure risk of the target device.

[0024] Taking the threshold judgment method as an example, a normal range threshold and multiple abnormal range thresholds are set for each operating parameter of the target device, and each abnormal range corresponds to a fault risk. The operating data of the target device during the prediction period is compared with these thresholds. If a parameter exceeds the normal range and is within one of the abnormal range thresholds, it indicates that the target device may have a certain fault risk. The greater the magnitude of the excess and the longer the duration, the higher the fault risk. The fault risk corresponding to the abnormal range is used as the fault risk of the target device. The first risk can be pre-set according to actual needs or the importance level of the target device. It should be noted that in this embodiment, a monitoring method for the target device is given. In actual use, S101-S103 can be executed separately for the equipment in the distribution room or other scenarios to predict the operating parameters of the electrical equipment in the entire distribution room.

[0025] The first monitoring strategy may be "It is recommended to troubleshoot the power distribution room every day," and the second monitoring strategy may be "It is recommended to troubleshoot the power distribution room every two days." The recipients of the first monitoring strategy may be management personnel of the power distribution room, etc. The monitoring frequencies of the first monitoring strategy and the second monitoring strategy may be preset.

[0026] As can be seen from the above, the present disclosure determines a target prediction model from multiple standard prediction models based on the distribution characteristics of historical data, enabling the present disclosure to select the prediction model that best suits the current device. By inputting the collected operating data of the target device into the target prediction model, the device's operating data for the prediction period can be obtained. Predicting based on the appropriate model can yield a more accurate prediction of the operating status, improving the accuracy and reliability of distribution room monitoring.

[0027] In one embodiment of the present disclosure, the plurality of standard prediction models include: a target autoregressive integrated moving average model and a target grey prediction model; Determine the target forecast model from multiple standard forecast models based on the distribution characteristics of historical data, including: In response to the historical data distribution characteristics satisfying the first distribution condition or the second distribution condition, determining the target autoregressive integrated moving average model as the target prediction model; In response to the historical data distribution characteristics satisfying the third distribution condition or the fourth distribution condition, determining the target grey prediction model as the target prediction model; Among them, the first distribution condition is that the overall data is continuous and the duration exceeds the first duration, the second distribution condition is that the overall data is discontinuous but the data within the second duration before the first moment is continuous, the third distribution condition is that the overall data is continuous but the duration does not exceed the first duration, and the fourth distribution condition is that the overall data is discontinuous and the data within the second duration before the first moment is discontinuous. The first moment is the moment for collecting the operating data of the target device.

[0028] In this embodiment, the multiple standard prediction models include: a target autoregressive integral moving average model and a target grey prediction model. Both the target autoregressive integral moving average model and the target grey prediction model are trained based on historical operation data of the target device.

[0029] From the description in S101, we can know that different models are applicable to different data characteristics, and the autoregressive integrated moving average model is more suitable for data prediction when the amount of data is sufficient, and the grey prediction model is more suitable for data prediction when the amount of data is small.

[0030] Therefore, in this embodiment, four distribution conditions are set: the first distribution condition, the second distribution condition, the third distribution condition, and the fourth distribution condition. Considering that when the overall data is continuous and lasts for a long time, it means that there are enough data points to reflect the inherent laws of the data. Alternatively, although the overall data is discontinuous, there is a period of continuous data before the first moment. This continuous data can also provide a certain amount of effective information for the ARIMA model. In this case, a prediction model whose basic model is the autoregressive integrated moving average model (i.e., the target autoregressive integrated moving average model) can be used for prediction. The first and second durations can be set based on experience or actual needs.

[0031] For new equipment in the distribution room or equipment recently put into production, when the overall data is continuous but the duration is short, the data volume is relatively limited and cannot meet the data requirements of the ARIMA model. Or, when the data is extremely discontinuous due to monitoring equipment failure, data loss, etc., the ARIMA model cannot effectively process and establish an accurate model, because the discontinuity of the data will make it difficult for the model to capture the inherent relationship and trend of the data. The grey prediction model has a strong processing ability for incomplete data. It can perform operations such as whitening on the original discontinuous data and convert it into usable information. Through methods such as correlation analysis in grey system theory, it can find the implicit relationship between the data and make predictions.

[0032] Therefore, in this embodiment, four different distribution conditions are set and their corresponding target prediction models are set.

[0033] From the above, it can be concluded that the present disclosure sets four different historical data distribution conditions and flexibly selects the target autoregressive integrated moving average model (ARIMA) or the target grey prediction model as the target prediction model according to these conditions, so that the selection of the model in the present disclosure is more targeted and flexible, and the most appropriate prediction model can be selected according to the historical data characteristics of different equipment, thereby improving the accuracy and reliability of distribution room monitoring.

[0034] As can be seen from the foregoing, both the target autoregressive integral moving average model and the target grey prediction model are trained based on the historical operating data of the target device. Therefore, this embodiment explains the training processes of the target autoregressive integral moving average model and the target grey prediction model respectively. First, the training process of the target autoregressive integral moving average model is explained: In one embodiment of the present disclosure, a process of training a target autoregressive integrated moving average model based on historical operating data of a target device includes: In response to the historical data distribution characteristic satisfying the second distribution condition, filling discontinuous time period data in the historical operation data of the target device based on a trend fitting method to obtain second historical operation data; An autoregressive integrated moving average model is trained based on the second historical operating data to obtain a target autoregressive integrated moving average model.

[0035] In this embodiment, if the distribution characteristics of historical data meet the second distribution condition, that is, the overall data is discontinuous but the data within the second time period before the first moment is continuous, then since the data within the second time period is continuous, the inherent logic is that there are fewer discontinuous data in the middle, that is, there are fewer missing data. At this time, the trend fitting method can be used to fill the missing data in the middle to obtain the second historical operation data. It should be noted that the second historical operation data not only includes the expanded missing data in the middle, but also includes the historical operation data of the target device that is actually collected. Among them, the trend fitting method is specifically explained as estimating the missing data based on the data trend before and after the missing data segment using methods such as linear fitting or polynomial fitting. Taking linear fitting as an example, by calculating the slope and intercept of the data before and after, the equation of the fitting line is obtained, and then the filling value is determined based on the position of the time point of the missing data on the line.

[0036] After obtaining the second historical operating data, an autoregressive integral moving average model can be trained based on the second historical operating data. The training process will not be described in detail in this application. Similarly, if the distribution characteristics of the historical data meet the first distribution condition, the autoregressive integral moving average model can be directly trained based on the historical operating data of the target device to obtain a target autoregressive integral moving average model. That is, in response to the historical data distribution characteristics meeting the first distribution condition, the autoregressive integral moving average model is trained based on the historical operating data of the target device to obtain a target autoregressive integral moving average model.

[0037] Next, this embodiment explains the training process of the grey prediction model: The process of training the target grey prediction model based on the historical operating data of the target equipment includes: In response to a correlation between the first data and the second data being greater than a first correlation threshold, training a grey prediction model based on the first data and the second data to obtain a first target grey prediction model; In response to a correlation between the first data and the second data being less than or equal to a first correlation threshold, training a grey prediction model based on the first data to obtain a second target grey prediction model; The first data is historical operation data of the target device, the second data is historical operation data related to the first data and does not belong to the target device, and the model parameters of the first target grey prediction model and the second target grey prediction model are different.

[0038] In this embodiment, the first data is the historical operating data of the target device, such as historical operating voltage, operating current, and operating temperature. The second data can be environmental factors such as the temperature and humidity of the environment in which the target device is located, or voltage fluctuations and frequency changes of the power grid. The first target grey prediction model is the GM (1, N) model, and the second target grey prediction model is the GM (1, 1) model.

[0039] The GM (1, 1) model and the GM (1, N) model are both prediction models in grey system theory, where "GM" represents a grey model, "1" indicates a first-order model, "1" indicates a single variable in the model, and "N" indicates N variables in the model. The GM (1, 1) model performs a single accumulation and other processing on the raw data to establish a first-order single-variable grey differential equation, thereby predicting the development trend of the data. The GM (1, N) model can utilize the relationships between multiple variables to more comprehensively capture the patterns and trends in the data, thereby improving the prediction ability of the target equipment operation data.

[0040] When the correlation between the first and second data is greater than the first correlation threshold, it indicates that the second data has a significant impact on the first data and cannot be ignored. Therefore, the GM (1, N) model is selected and both the first and second data are used as input variables for model training. Conversely, when the correlation between the first and second data is less than or equal to the first correlation threshold, it indicates that the second data has a relatively small impact on the first data, or in other words, from a data relationship perspective, the additional information provided by the second data is not very helpful in predicting the target device operating data. In this case, the GM (1, 1) model can be selected and trained based on the first data.

[0041] The correlation between the first data and the second data can be determined by calculating the Pearson correlation coefficient between the two, and the calculated Pearson correlation coefficient is used as the correlation between the first data and the second data. The first correlation threshold can be determined based on experience.

[0042] From the above, it can be concluded that the present disclosure determines the distribution characteristics of historical data and, when the data meets the second distribution condition, uses a trend fitting method to fill in the discontinuous time period data. This effectively addresses the gaps in the data and enables the ARIMA model to be applied to scenarios that would otherwise be inaccessible due to discontinuous data. Secondly, based on the correlation between the first and second data, the present disclosure selects either the GM(1,1) model or the GM(1,N) model for training, enabling the gray prediction model to better adapt to different data scenarios and improving the accuracy of equipment operating parameter predictions.

[0043] The training process of the GM (1, 1) model is not described in detail in this application, but the training process of the GM (1, N) model is different and is specifically explained as follows: In one embodiment of the present disclosure, the training process of the first target grey prediction model includes: Determine the initial development coefficient and initial grey action of grey prediction model based on least square method; Determine the parameter vector to be estimated based on the initial development coefficient and the initial grey action; The parameter vector to be estimated is iteratively optimized based on the particle swarm algorithm until the difference between multiple consecutive fitness functions is less than the first fitness threshold or the number of iterations reaches the first number of iterations, thereby obtaining the first target grey prediction model.

[0044] The first target grey prediction model is the GM (1, N) model, where N is the number of variables included in the first target grey prediction model; Iteratively optimize the estimated parameter vector based on the particle swarm algorithm, including: Initialize particles in the particle swarm algorithm; The particle swarm algorithm after initialization iteratively optimizes the parameter vector to be estimated. The dimension of each particle in the particle swarm algorithm is N+1.

[0045] In this embodiment, the development coefficient and the gray action are important parameters of the model. In this embodiment, the first data and the second data are processed and calculated by the least square method to determine a set of initial development coefficients and initial gray action. and initial ash action After that, the initial ash action The essence of is a vector, that is , and combine them into a parameter vector to be estimated ,in represents transpose, Indicates the influence of the first related factor on the dominant factor, Indicates the degree of influence of the second related factor on the dominant factor, Indicates the degree of influence of the Nth related factor on the dominant factor.

[0046] Considering that the parameter vector space to be estimated of the GM(1, N) model is relatively complex, the particle swarm algorithm has a strong global search capability. Through random search and information sharing of particles in the solution space, it can start from different initial points and explore in the entire parameter space, with a greater chance of finding the global optimal parameter combination, thereby improving the accuracy of parameter estimation and enabling the model to fit data and make predictions more accurately. Therefore, this embodiment optimizes the parameter vector to be estimated through the particle swarm algorithm. Each parameter in the parameter vector to be estimated determines the quantitative relationship between the system characteristic variables and the related factor variables in the GM(1, N) model. The development coefficient reflects the development trend and inertia of the system characteristic variables themselves. It reflects the degree and direction of the influence of relevant factor variables on the system characteristic variables. Their combined effect enables the GM(1, N) model to capture the state changes of the system at different times, forming a dynamic model structure that can reflect the inherent laws of the system.

[0047] The process of initializing particles in the particle swarm algorithm is specifically explained as randomly generating a group of particles within a certain range of values ​​to form the initial particle swarm. This range can be determined based on prior knowledge of the operating parameters of the distribution room equipment and the general range of values ​​for the grey prediction model parameters. For example, the development coefficient is within the range of [-1, 1], and the grey action is within the range of [-10, 10].

[0048] The dimension of particles in the particle swarm algorithm is N+1, that is, each particle can be expressed as The number of particles in the particle swarm can be set based on experience or the number of particles used to solve similar problems. The fitness function of the particle swarm algorithm can be the mean square error or the mean absolute error. If the mean absolute error is used as the fitness function, for the Model parameters represented by particles , calculate its predicted value for the operating parameters of the distribution room equipment Compared with the actual monitoring value The mean absolute error between ,in, is the number of data samples used for validation, Indicates the The actual monitoring value corresponding to each particle, Indicates the The smaller the fitness function value, the better the model parameters represented by the particle.

[0049] Until the difference between multiple consecutive fitness functions is less than the first fitness threshold or the number of iterations reaches the first number of iterations, the first target grey prediction model is obtained. The plurality referred to can be 4, which is a preset value. The first fitness threshold and the first number of iterations can be set based on data from the experiment or based on reference values ​​provided by the particle swarm algorithm.

[0050] More specifically, in the particle swarm algorithm, inertia weight is used to control the degree to which particles inherit their historical velocity. A larger inertia weight is beneficial for global search, allowing particles to explore the search space more extensively; a smaller inertia weight is beneficial for local search, allowing particles to focus more on the area near the current optimal solution. Therefore, in this embodiment, the inertia weight of the particle swarm algorithm can also be determined based on the first formula.

[0051] The first formula is: ,in, is the inertia weight, is the minimum value of inertia weight, which can be set to 0.4. is the maximum value of the inertia weight, which can be set to 0.9. is the variance of the current population fitness, which is used to measure the dispersion of the fitness of particles in the population. It is the maximum value of the fitness variance, which can be determined by the maximum fitness variance recorded in the initial population or during the algorithm operation. is the current iteration number, is the maximum number of iterations, that is, the first iteration number.

[0052] ,in, represents the number of particles, It is The fitness value of a particle, is the average fitness value of the particle swarm.

[0053] The logic of the first formula is: the variance of the current population fitness This reflects the degree of dispersion of the fitness values ​​of individual particles in the population relative to the average fitness value. A larger variance indicates greater disparity in the fitness values ​​of the particles, and the presence of multiple solutions of varying quality within the population. This indicates that the algorithm has not yet found a relatively concentrated, optimal region in the solution space, and a larger inertia weight is needed to conduct a more extensive global search to explore more possible solutions. Represents the ratio of the current fitness variance to the maximum fitness variance recorded. When it is close to 1, it means that the fitness variance of the current population is close to the maximum, that is, the fitness of particles in the population varies greatly, and the algorithm needs to maintain a large inertia weight for global search. When it is close to 0, it means that the fitness variance is very small, the fitness values ​​of the particles are relatively concentrated, and the algorithm has found a better solution area to a certain extent. At this time, it is necessary to reduce the inertia weight in order to perform a more refined local search.

[0054] Reflects the current search stage of the algorithm. As the iteration proceeds, The value gradually increases from 0 to 1. This will make the inertia weight tend to take a larger value, which is conducive to the algorithm to conduct a large-scale global search at the beginning and quickly locate the area where a better solution may exist. The denominator plays a regulatory role, which ensures that the value range of the inertia weight is within and This avoids the extreme situation where the inertia weight is too large or too small due to the change of the two factors in the numerator.

[0055] From the above, it can be concluded that the present disclosure uses the particle swarm algorithm to iteratively optimize the parameter vector to be estimated, and can start from different initial points and explore the entire parameter space to find the globally optimal parameter combination, thereby improving the accuracy of parameter estimation, and then improving the accuracy and reliability of the prediction of the target device operating parameters, thereby improving the accuracy and reliability of the distribution room monitoring. The present disclosure dynamically adjusts the inertia weight of the particle swarm algorithm through the first formula, and balances the global search and local search according to the variance of the current population fitness and the number of iterations. As the iteration proceeds, the gradually decreasing inertia weight is conducive to local search, so that the particles are more focused on the area near the current optimal solution, which improves the accuracy and reliability of the particle swarm optimization, and then improves the accuracy of parameter estimation, thereby improving the accuracy and reliability of the prediction of the target device operating parameters, and achieving the effect of improving the accuracy and reliability of the distribution room monitoring.

[0056] Corresponding to an intelligent distribution room monitoring method in the above embodiment, Figure 2 This is a structural block diagram of an intelligent distribution room monitoring system provided by an embodiment of the present disclosure. For ease of explanation, only the parts related to the embodiment of the present disclosure are shown. Figure 2 The intelligent distribution room monitoring system 20 includes: a model determination module 21, a data prediction module 22 and a strategy determination module 23.

[0057] The model determination module 21 is configured to determine a target prediction model from a plurality of standard prediction models based on historical data distribution characteristics; wherein the historical data distribution characteristics are data distribution characteristics of historical operating data of a target device on a time axis, the target device is a device in a power distribution room, and the plurality of standard prediction models are models obtained by training different prediction models based on the historical operating data of the target device; The data prediction module 22 is used to input the collected operating data of the target device into the target prediction model to obtain the predicted operating data of the target device within the prediction period; The strategy determination module 23 is used to determine the monitoring strategy of the distribution room based on the predicted operating data of the target equipment within the predicted time period.

[0058] In one embodiment of the present disclosure, the plurality of standard prediction models include: a target autoregressive integrated moving average model and a target grey prediction model; The model determination module 21 is specifically configured to determine the target autoregressive integrated moving average model as the target prediction model in response to the historical data distribution characteristics satisfying the first distribution condition or the second distribution condition; In response to the historical data distribution characteristics satisfying the third distribution condition or the fourth distribution condition, determining the target grey prediction model as the target prediction model; Among them, the first distribution condition is that the overall data is continuous and the duration exceeds the first duration, the second distribution condition is that the overall data is discontinuous but the data within the second duration before the first moment is continuous, the third distribution condition is that the overall data is continuous but the duration does not exceed the first duration, and the fourth distribution condition is that the overall data is discontinuous and the data within the second duration before the first moment is discontinuous. The first moment is the moment for collecting the operating data of the target device.

[0059] In one embodiment of the present disclosure, an intelligent distribution room monitoring system 20 further includes: a first model training module, configured to fill discontinuous time period data in the historical operation data of the target device based on a trend fitting method in response to the historical data distribution characteristics satisfying a second distribution condition, to obtain second historical operation data; An autoregressive integrated moving average model is trained based on the second historical operating data to obtain a target autoregressive integrated moving average model.

[0060] In one embodiment of the present disclosure, an intelligent distribution room monitoring system 20 further includes: a second model training module, configured to, in response to a correlation between the first data and the second data being greater than a first correlation threshold, train a grey prediction model based on the first data and the second data to obtain a first target grey prediction model; In response to a correlation between the first data and the second data being less than or equal to a first correlation threshold, training a grey prediction model based on the first data to obtain a second target grey prediction model; The first data is historical operation data of the target device, the second data is historical operation data related to the first data and does not belong to the target device, and the model parameters of the first target grey prediction model and the second target grey prediction model are different.

[0061] In one embodiment of the present disclosure, the second model training module is specifically used to determine the initial development coefficient and initial grey action amount of the grey prediction model based on the least squares method; Determine the parameter vector to be estimated based on the initial development coefficient and the initial grey action; The parameter vector to be estimated is iteratively optimized based on the particle swarm algorithm until the difference between multiple consecutive fitness functions is less than the first fitness threshold or the number of iterations reaches the first number of iterations, thereby obtaining the first target grey prediction model.

[0062] In one embodiment of the present disclosure, the first target grey prediction model is a GM (1, N) model, where N is the number of variables included in the first target grey prediction model; The second model training module is specifically used to initialize particles in the particle swarm algorithm; The particle swarm algorithm after initialization iteratively optimizes the parameter vector to be estimated. The dimension of each particle in the particle swarm algorithm is N+1.

[0063] In one embodiment of the present disclosure, the strategy determination module 23 is specifically configured to determine the failure risk of the target device based on the operating data of the target device within the prediction period; In response to the failure risk of the target device being greater than or equal to the first risk, determining the first monitoring strategy as the monitoring strategy for the power distribution room; In response to the failure risk of the target device being less than the first risk, determining the second monitoring strategy as the monitoring strategy for the power distribution room; The monitoring frequency of the first monitoring strategy is greater than the monitoring frequency of the second monitoring strategy.

[0064] See also Figure 3 , Figure 3 This is a schematic block diagram of an electronic device provided by an embodiment of the present disclosure. Figure 3 The electronic device 300 in the embodiment shown may include: one or more processors 301, one or more input devices 302, one or more output devices 303, and one or more memories 304. The processors 301, input devices 302, output devices 303, and memories 304 communicate with each other via a communication bus 305. The memory 304 is used to store computer programs, which include program instructions. The processor 301 is used to execute the program instructions stored in the memory 304. The processor 301 is configured to call the program instructions to execute the functions of the modules / units in the above-mentioned system embodiments, such as Figure 2 The functions of modules 21 to 23 are shown.

[0065] It should be understood that in the embodiments of the present disclosure, the processor 301 may be a central processing unit (CPU), or may 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, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0066] The input device 302 may include a touchpad, a fingerprint collection sensor (for collecting user fingerprint information and fingerprint direction information), a microphone, etc. The output device 303 may include a display (LCD, etc.), a speaker, etc.

[0067] The memory 304 may include a read-only memory and a random access memory, and provides instructions and data to the processor 301. A portion of the memory 304 may also include a non-volatile random access memory. For example, the memory 304 may also store device type information.

[0068] In a specific implementation, the processor 301, input device 302, and output device 303 described in the embodiments of the present disclosure can execute the implementation methods described in the first and second embodiments of an intelligent distribution room monitoring method provided in the embodiments of the present disclosure, and can also execute the implementation methods of the electronic device described in the embodiments of the present disclosure, which will not be repeated here.

[0069] In another embodiment of the present disclosure, a computer-readable storage medium is provided. The computer-readable storage medium stores a computer program. The computer program includes program instructions. When the program instructions are executed by a processor, all or part of the process of the method in the above embodiment is implemented. The computer program can also be used to instruct related hardware to complete the process. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by the processor, the steps of each of the above method embodiments are implemented. The computer program includes computer program code, which can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium can include: any entity or device capable of carrying computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal and software distribution medium.

[0070] The computer-readable storage medium can be an internal storage unit of the electronic device in any of the aforementioned embodiments, such as a hard disk or memory of the electronic device. The computer-readable storage medium can also be an external storage device of the electronic device, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) card, a flash memory card, etc. Furthermore, the computer-readable storage medium can include both an internal storage unit of the electronic device and an external storage device. The computer-readable storage medium is used to store computer programs and other programs and data required by the electronic device. The computer-readable storage medium can also be used to temporarily store data that has been output or is about to be output.

[0071] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the above description has generally described the composition and steps of each example according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this disclosure.

[0072] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the electronic devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0073] In the several embodiments provided in this application, it should be understood that the disclosed electronic devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of units is only a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces or units, or can be an electrical, mechanical or other form of connection.

[0074] Units described as separate components may or may not be physically separate, and components shown as units may or may not be physical units, i.e., they may be located in one place or distributed across multiple network units. Some or all of these units may be selected based on actual needs to achieve the objectives of the embodiments of the present disclosure.

[0075] In addition, the functional units in the various embodiments of the present disclosure may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0076] The above are only specific embodiments of the present disclosure, but the scope of protection of the present disclosure is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or replacements within the technical scope disclosed in this disclosure, and such modifications or replacements should be included in the scope of protection of the present disclosure. Therefore, the scope of protection of the present disclosure should be based on the scope of protection of the claims.

Claims

1. A method for monitoring an intelligent power distribution room, characterized in that: include: Determining a target prediction model from a plurality of standard prediction models based on historical data distribution characteristics; wherein the historical data distribution characteristics are data distribution characteristics of historical operating data of a target device on a time axis, the target device is a device in a power distribution room, and the plurality of standard prediction models are models obtained by respectively training different prediction models based on the historical operating data of the target device; Inputting the collected operating data of the target device into the target prediction model to obtain the predicted operating data of the target device within the prediction period; A monitoring strategy for the power distribution room is determined based on the predicted operating data of the target equipment within the predicted time period.

2. The intelligent distribution room monitoring method according to claim 1, characterized in that: The multiple standard prediction models include: a target autoregressive integrated moving average model and a target grey prediction model; The step of determining a target prediction model from a plurality of standard prediction models based on the distribution characteristics of historical data includes: In response to the historical data distribution characteristic satisfying the first distribution condition or the second distribution condition, determining the target autoregressive integrated moving average model as the target prediction model; In response to the historical data distribution characteristic satisfying the third distribution condition or the fourth distribution condition, determining the target grey prediction model as the target prediction model; Among them, the first distribution condition is that the overall data is continuous and the duration exceeds the first duration, the second distribution condition is that the overall data is discontinuous but the data within the second duration before the first moment is continuous, the third distribution condition is that the overall data is continuous but the duration does not exceed the first duration, the fourth distribution condition is that the overall data is discontinuous and the data within the second duration before the first moment is discontinuous, and the first moment is the moment when the operating data of the target device is collected.

3. The intelligent distribution room monitoring method according to claim 2, characterized in that: The process of training a target autoregressive integrated moving average model based on the historical operating data of the target device includes: In response to the historical data distribution characteristic satisfying a second distribution condition, filling discontinuous time period data in the historical operation data of the target device based on a trend fitting method to obtain second historical operation data; An autoregressive integrated moving average model is trained based on the second historical operation data to obtain a target autoregressive integrated moving average model.

4. The intelligent distribution room monitoring method according to claim 2, characterized in that: The process of training a target grey prediction model based on the historical operation data of the target device includes: In response to a correlation between the first data and the second data being greater than a first correlation threshold, training a grey prediction model based on the first data and the second data to obtain a first target grey prediction model; In response to a correlation between the first data and the second data being less than or equal to the first correlation threshold, training a grey prediction model based on the first data to obtain a second target grey prediction model; Among them, the first data belongs to the historical operation data of the target device, the second data is historical operation data related to the first data and does not belong to the target device, and the model parameters of the first target grey prediction model and the second target grey prediction model are different.

5. The intelligent distribution room monitoring method according to claim 4, characterized in that: The training process of the first target grey prediction model includes: Determining the initial development coefficient and initial grey action of the grey prediction model based on the least squares method; Determining a parameter vector to be estimated based on the initial development coefficient and the initial grey action; The parameter vector to be estimated is iteratively optimized based on the particle swarm algorithm until the difference between multiple consecutive fitness functions is less than a first fitness threshold or the number of iterations reaches a first number of iterations, thereby obtaining a first target grey prediction model.

6. The intelligent distribution room monitoring method according to claim 5, characterized in that: The first target grey prediction model is a GM (1, N) model, where N is the number of variables included in the first target grey prediction model; The iterative optimization of the parameter vector to be estimated based on the particle swarm algorithm includes: Initializing particles in the particle swarm algorithm; The parameter vector to be estimated is iteratively optimized based on the initialized particle swarm algorithm, where the dimension of each particle in the particle swarm algorithm is N+1.

7. The intelligent distribution room monitoring method according to claim 1, characterized in that: The determining of the monitoring strategy of the power distribution room based on the predicted operating data of the target equipment within the predicted period includes: determining a failure risk of the target device based on operating data of the target device within the prediction period; In response to the failure risk of the target device being greater than or equal to the first risk, determining the first monitoring strategy as the monitoring strategy for the power distribution room; In response to the failure risk of the target device being less than the first risk, determining the second monitoring strategy as the monitoring strategy for the power distribution room; The monitoring frequency of the first monitoring strategy is greater than the monitoring frequency of the second monitoring strategy.

8. An intelligent distribution room monitoring system, characterized in that: include: a model determination module, configured to determine a target prediction model from a plurality of standard prediction models based on historical data distribution characteristics; wherein the historical data distribution characteristics are data distribution characteristics of historical operating data of a target device on a time axis, the target device is a device in a power distribution room, and the plurality of standard prediction models are models obtained by training different prediction models based on the historical operating data of the target device; a data prediction module, configured to input the collected operating data of the target device into the target prediction model to obtain predicted operating data of the target device within a prediction period; A strategy determination module is used to determine the monitoring strategy of the distribution room based on the predicted operating data of the target equipment within the predicted time period.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images