Monitoring and early warning method and system based on active load prediction of power distribution network

Abstract

The invention discloses a monitoring and early warning method and system based on active load prediction of a power distribution network, which breaks through the limitation of purely depending on historical data and simple statistical analysis, fully considers the complexity and uncertainty of load change influenced by various factors, and achieves the early warning of the load change through the combination of multiple models. The method can accurately classify and analyze the operation active load of the power distribution network through an active power analysis model, a decision tree model and the like, thereby improving the prediction precision. In the aspect of data processing, mass operation data generated by intelligent equipment in the power distribution network can be effectively integrated and mined by utilizing a vector machine model and a cloud large model server, an early warning mechanism is flexible, the method can adapt to different operation scenes and load characteristics, false alarm and missing alarm are avoided, and the method is suitable for popularization and application. According to the method, the power distribution network fault identification efficiency of the power distribution network power factors in different time periods is remarkably improved, and effective operation of the power distribution network is ensured.

Description

Technical Field

[0001] The present invention relates to the field of distribution network fault monitoring and early warning, and in particular to a monitoring and early warning method and system based on active load prediction of a distribution network. Background Art

[0002] In today's society, the stability and reliability of power supply are crucial to people's production and life. As a key link in the power system, the distribution network is directly facing users, and its operating status directly affects the quality of power supply. With the rapid development of the economy and the continuous growth of electricity demand, the scale of the distribution network is expanding, the structure is becoming more complex, and the load characteristics are also showing diversified and dynamic characteristics. The traditional distribution network operation and management model can no longer meet the needs of modern power systems, and active load forecasting monitoring and early warning methods and systems have emerged. This method and system aims to accurately predict the load of the distribution network, discover potential operating risks in advance, realize intelligent management of the distribution network, and ensure safe, stable and efficient power supply.

[0003] However, the existing distribution network load forecasting methods have many shortcomings. On the one hand, most traditional load forecasting models are based on historical data and simple statistical analysis, and do not take into account the complexity and uncertainty of load changes. These models often assume that load changes have a certain regularity, but in actual operation, the distribution network load is affected by many factors, such as meteorological conditions, social activities, economic development, etc., and its changes often show a high degree of nonlinearity and randomness. Therefore, the prediction accuracy of traditional models is difficult to meet actual needs, which easily leads to a large deviation between the prediction results and the actual load, thus affecting the accuracy of monitoring and early warning.

[0004] On the other hand, the existing monitoring and early warning systems have defects in data processing and analysis capabilities. With the widespread application of various intelligent devices in the distribution network, a large amount of operating data has been generated, but the existing systems are often unable to effectively integrate and mine these massive data. The dispersion and incompleteness of the data make it difficult for the system to fully and accurately grasp the operating status of the distribution network and to detect potential risk factors in a timely manner. In addition, most of the existing early warning mechanisms are based on fixed threshold settings, lack adaptability to different operating scenarios and load characteristics, and are prone to false alarms or missed alarms, reducing the reliability and practicality of the monitoring and early warning system. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings existing in the prior art and to propose a monitoring and early warning method and system based on active load prediction of distribution network.

[0006] In order to achieve the above object, the present invention adopts the following technical report: a monitoring and early warning method based on active load forecasting of distribution network, comprising the following steps:

[0007] S1. Obtain the active load of the distribution network, and use the active power analysis model and decision tree model of the distribution network to classify the active load change rate, active load peak-to-valley difference and load proportion of different types of electrical equipment in the active load of the distribution network;

[0008] S2. Identify the electrical equipment correlation parameters and the grid topology of the active load, and obtain the clustered distribution network active load as the active load for the preset period; use semi-supervised learning to identify the distribution network line loss of the active load for the preset period, predict the state of the electrical equipment in the distribution network, and obtain the electrical equipment state prediction result as the distribution network preset period prediction;

[0009] S3, using a support vector machine model and a cloud-based large model server, using the distribution network preset period prediction, preset period operation active load, calculate the electrical equipment strength and deformation modulus data in different operation periods, generate the distribution network load impact source, and use the transformer voltage stabilization active load and effective power supply active load to monitor and warn the power supply efficiency;

[0010] S4. According to the sources of load influence on the distribution network, the ideal power supply efficiency of the power factor of the distribution network in different periods is monitored and warned, and a power supply efficiency unit time visualization interface is generated. The operation loss error is predicted by starting, stopping and effective power supply voltage of the power supply electrical equipment. The operation loss adjustment and dynamic power supply efficiency statistics are performed through the power supply efficiency unit time visualization interface combined with the vector sum of apparent power of the distribution network fault operation, and an optimization report on starting and stopping the operating electrical equipment is obtained;

[0011] S5. Use the Pearson correlation coefficient in combination with the startup and stop optimization report of the operating electrical equipment and the power factor of the distribution network at different time periods in different regional locations to confirm and analyze the distribution network fault, and obtain the distribution network fault impact range at different operating time periods.

[0012] As a further report of the present invention, the steps of obtaining the active load of the distribution network, using the active power analysis model and the decision tree model of the distribution network, classifying the active load change rate, the peak-to-valley difference of the active load and the load proportion of different types of electrical equipment in the active load of the distribution network, identifying the electrical equipment correlation parameters and the grid topology structure of the active load, and obtaining the clustered active load of the distribution network as the active load of the preset period are specifically as follows:

[0013] The active operating load is collected and transmitted through the distribution network load monitoring platform to obtain the original distribution network active operating load;

[0014] Input the original distribution network operating active load, use a normalization method to perform data limitation and data mapping, and obtain an active power analysis preprocessing operating active load;

[0015] The input active power analysis preprocesses the operating active load, and based on the decision tree model, calculates the load proportion of different types of electrical equipment so that it can be calculated under different peak-to-valley difference forms to obtain the clustered distribution network operating active load.

[0016] As a further report of the present invention, the steps of using semi-supervised learning to identify the line loss of the distribution network of the active load in the preset period, predicting the state of the electrical equipment in the distribution network, and obtaining the prediction result of the electrical equipment state are as follows:

[0017] Input the active load in the preset period into the semi-supervised learning, use the convolutional semi-supervised learning to identify the line loss of the distribution network, and obtain different characteristics of the line loss of the distribution network;

[0018] By utilizing the different characteristics of the distribution network line loss, the convolutional neural network algorithm reduces the feature dimension and electrical equipment status recognition, and outputs the electrical equipment status prediction result.

[0019] As a further report of the present invention, the support vector machine model and the cloud-based large model server are used to predict the preset period of the distribution network, operate the active load during the preset period, calculate the strength and deformation modulus data of the electrical equipment during different operating periods, and generate the distribution network load influence source in the following steps:

[0020] Input the distribution network preset period forecast and the preset period operation active load;

[0021] Use different strength calculation models to obtain electrical equipment strength data, calculate the fluctuation range of deformation modulus through the modulus calculation model, and obtain the expression of load change;

[0022] The load change representation is used to create input data for a large cloud-based model server, and a support vector machine model is used to perform active load forecasting and statistics on equipment failure locations in different operating periods, and output the sources of load impact on the distribution network.

[0023] As a further report of the present invention, the transformer voltage stabilization active load and the effective power supply active load are used to monitor and warn the power supply efficiency. According to the source of the distribution network load, the ideal power supply efficiency of the distribution network power factor in different time periods is monitored and warned, and the steps of generating a unit time visualization interface of the power supply efficiency are specifically as follows:

[0024] Input the load influencing source of the distribution network, calculate the variation range of power supply efficiency on the load influencing source of the distribution network through the transformer voltage stabilizing active load, and output the preliminary power supply efficiency;

[0025] The effective power supply active load is used to perform power supply active load analysis on the preliminary power supply efficiency, the ideal power supply efficiency of the power factor of the distribution network in different time periods is monitored and warned, and a power supply efficiency unit time visualization interface is generated.

[0026] As a further report of the present invention, the steps of predicting the operating loss error by starting, stopping and effectively supplying voltage of the power supply electrical equipment, adjusting the operating loss and performing dynamic power supply efficiency statistics by combining the vector sum of apparent power of distribution network fault operation with the power supply efficiency unit time visualization interface, and obtaining the optimization report of starting and stopping the operating electrical equipment are specifically as follows:

[0027] Input the power supply efficiency unit time visualization interface, perform dynamic power supply efficiency monitoring and early warning through vector sum calculation, and obtain dynamic power supply efficiency;

[0028] The dynamic power supply efficiency is utilized and effective power supply voltage is adopted to perform fault assessment and strategy optimization of the dynamic power supply efficiency, and a startup and stop optimization report of the running electrical equipment is generated.

[0029] As a further report of the present invention, the steps of using the Pearson correlation coefficient in combination with the startup and stop optimization report of the running electrical equipment to confirm and analyze the distribution network faults in different time periods at different regional locations to obtain the distribution network fault impact range in different operation periods are specifically as follows:

[0030] By using the optimization report on starting and stopping the running electrical equipment, the fault sharing between the power factors of the distribution network in different time periods at different regional locations is established through the Pearson correlation coefficient, and a distribution network fault confirmation and analysis calculation model is generated;

[0031] Input the distribution network fault confirmation and analysis calculation model, use the bat positioning model to locate the power factor faults of the distribution network in different areas and different time periods, and generate a fault location report;

[0032] The fault location report is used to calculate the correlation between distribution network faults through a correlation analysis algorithm to obtain the distribution network fault impact range in different operation periods.

[0033] A monitoring and early warning system based on active load prediction of a distribution network is used to execute a monitoring and early warning method based on active load prediction of a distribution network. The monitoring and early warning system based on active load prediction of a distribution network is composed of an operating active load clustering unit, a distribution network line loss identification and electrical equipment status identification unit, a distribution network load influence source modeling unit, a power supply efficiency monitoring and early warning unit, an operating electrical equipment start-up and stop optimization reporting unit, and a distribution network power factor unit at different time periods in different regional locations.

[0034] As a further report of the present invention, the operating active load clustering unit is responsible for capturing the original operating active load from the distribution network load monitoring platform, performing active power analysis preprocessing, and finally using a decision tree model method to preset the period of operating active load, and outputting the clustered distribution network operating active load;

[0035] The distribution network line loss identification and electrical equipment state identification unit uses convolution to identify the distribution network line loss based on clustering the active loads in the distribution network operation, reduces the feature dimension and identifies the electrical equipment state through a convolutional neural network algorithm, and outputs the electrical equipment state prediction result;

[0036] The distribution network load impact source modeling unit obtains the electrical equipment strength data through a support vector machine model based on the electrical equipment state prediction result, and constructs the distribution network load impact source based on the input data of the cloud large model server.

[0037] As a further report of the present invention, the power supply efficiency monitoring and early warning unit inputs the source of load influence of the distribution network, uses the transformer voltage stabilization active load to find the power supply efficiency of the preset time period, and optimizes the power supply efficiency through the effective power supply active load to generate an ideal power supply efficiency unit time visualization interface;

[0038] The running electrical equipment start and stop optimization report unit uses vector sum calculation and effective power supply voltage, uses the ideal power supply efficiency unit time visualization interface to perform operation loss adjustment and dynamic power supply efficiency statistics, and outputs the running electrical equipment start and stop optimization report;

[0039] The power factor unit of the distribution network in different time periods at different regional locations uses the Pearson correlation coefficient and the bat positioning model, and utilizes the start-up and stop optimization report of the operating electrical equipment to perform distribution network fault confirmation and analysis calculation between the power factors of the distribution network in different time periods at different regional locations, and obtain the distribution network fault impact range in different operating time periods.

[0040] Beneficial effects:

[0041] The present invention proposes a monitoring and early warning method and system based on active load prediction of distribution network. Through the active load processing technology, the active load change rate, peak-to-valley difference of active load and load proportion of different types of electrical equipment of the active load of distribution network operation are effectively classified, so that the active load operation is clearer, thereby improving the accuracy of electrical equipment status recognition. The support vector machine model can accurately identify the distribution network load influence source structure of the distribution network environment. Combined with the transformer voltage stabilization active load and the start and stop of effective power supply electrical equipment, it not only provides accurate power supply efficiency monitoring and early warning, but also increases the dynamic adaptability of the system, and can respond to environmental changes in a timely manner. Through the Pearson correlation coefficient, the execution efficiency and data sharing capability of the power factor of the distribution network in different periods of time at different regional locations are realized. Compared with the traditional method, the present invention breaks through the limitations of relying solely on historical data and simple statistical analysis, fully considers the complexity and uncertainty of load changes affected by multiple factors, and accurately classifies and analyzes the active load of distribution network operation through the combination of multiple models, such as active power analysis model, decision tree model, etc., to improve the prediction accuracy. In terms of data processing, this method utilizes vector machine models and cloud-based large model servers to effectively integrate and mine the massive operating data generated by intelligent devices in the distribution network. Its early warning mechanism is flexible and can adapt to different operating scenarios and load characteristics, avoiding false alarms and missed alarms, greatly improving the reliability and practicality of the system. This invention significantly improves the distribution network fault identification efficiency of the distribution network power factor in different time periods, ensuring the effective operation of the distribution network. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 Schematic diagram of the workflow of the method of the present invention;

[0043] Figure 2 2 is a unit composition diagram of the system of the present invention. DETAILED DESCRIPTION

[0044] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application may be combined with each other. The present application is further described in detail below in conjunction with the accompanying drawings and specific embodiments.

[0045] See also Figure 1 The present invention provides a technical report: a monitoring and early warning method based on active load forecasting of a distribution network, comprising the following steps:

[0046] S1. Obtain the active load of the distribution network, and use the active power analysis model and decision tree model of the distribution network to classify the active load change rate, active load peak-to-valley difference and load proportion of different types of electrical equipment in the active load of the distribution network;

[0047] S2. Identify the electrical equipment correlation parameters and the grid topology of the active load, obtain the clustered distribution network active load as the active load for the preset period, use semi-supervised learning to identify the distribution network line loss of the active load for the preset period, predict the state of the electrical equipment in the distribution network, and obtain the electrical equipment state prediction result as the distribution network preset period prediction;

[0048] S3. Use support vector machine model and cloud-based large model server to calculate the strength and deformation modulus data of electrical equipment in different operation periods based on the distribution network preset period prediction and preset period operation active load, generate the source of distribution network load influence, and use transformer voltage stabilization active load and effective power supply active load to monitor and warn power supply efficiency;

[0049] S4. According to the sources of load influence on the distribution network, monitor and warn the ideal power supply efficiency of the power factor of the distribution network in different periods, generate a unit time visualization interface for power supply efficiency, predict the operation loss error through the start and stop of the power supply electrical equipment and the effective power supply voltage, and adjust the operation loss and dynamic power supply efficiency statistics through the unit time visualization interface of power supply efficiency combined with the vector sum of the apparent power of the distribution network fault operation, and obtain the optimization report of the start and stop of the operating electrical equipment;

[0050] S5. Use the Pearson correlation coefficient to confirm and analyze the distribution network faults based on the start-up and stop optimization reports of the operating electrical equipment and the power factors of the distribution network at different time periods in different regional locations, and obtain the distribution network fault impact range at different operating periods.

[0051] By applying the active power analysis model of the distribution network and the clustering algorithm of the active load, the electrical equipment correlation parameters and the grid topology of the active load in the distribution network can be improved.

[0052] This will help alleviate problems such as scattering, absorption and peak-to-valley differences in active loads in the distribution network environment, making active loads in distribution network operation easier to analyze and explain.

[0053] By using semi-supervised learning to identify the line loss of the distribution network and the state of electrical equipment for the active load in operation, combined with the support vector machine model and the cloud-based large model server, the position and deformation modulus data of the state of the electrical equipment in the distribution network can be accurately located. This is crucial for accurately identifying and tracking the apparent power of the state of electrical equipment in the distribution network environment at different time periods. Combined with the transformer voltage stabilization active load and the effective power supply active load, according to the information provided by the distribution network load influence source and the state of electrical equipment, the dynamic power supply efficiency monitoring and early warning of the power factor of the distribution network at different time periods can be realized. Through efficient power supply efficiency monitoring and early warning, the system can select the ideal power supply efficiency to avoid obstacles and quickly reach the state position of the electrical equipment. By using the start, stop and effective power supply voltage of the effective power supply electrical equipment, the power factor of the distribution network at different time periods can be used for operating loss error prediction and dynamic power supply efficiency adjustment. The system can predict possible operating losses and adjust the power supply efficiency accordingly based on the movement mode and environmental information of the electrical equipment state, thereby improving the adaptability and performance of the system. The Pearson correlation coefficient is applied to realize the distribution network fault confirmation and analysis calculation between the power factors of the distribution network at different time periods in different regional locations. This will improve the efficiency of distribution network fault management and enable various systems to cooperate, share information and collaborate to complete complex distribution network fault management.

[0054] The steps of obtaining the active load of the distribution network, using the active power analysis model and decision tree model of the distribution network, classifying the active load change rate, the peak-to-valley difference of the active load and the load proportion of different types of electrical equipment in the active load of the distribution network, identifying the electrical equipment correlation parameters and the grid topology structure of the active load, and obtaining the clustered active load of the distribution network as the active load of the preset period are as follows:

[0055] The active operating load is collected and transmitted through the distribution network load monitoring platform to obtain the original distribution network active operating load;

[0056] Input the original distribution network operating active load, use the normalization method to perform data limitation and data mapping, and obtain the active power analysis preprocessing operating active load;

[0057] The input active power analysis preprocesses the operating active load, and based on the decision tree model, calculates the load proportion of different types of electrical equipment so that it can be calculated under different peak-to-valley difference forms to obtain the clustered distribution network operating active load.

[0058] The normalization method is used to limit and map the data of the active loads in the distribution network. This step can reduce the impact of light transmission in the distribution network environment, such as light intensity attenuation and color shift. Through correction, the active power analysis preprocessing active load can more accurately reflect the real appearance of the distribution network scene. The active load clustering algorithm such as Naive Bayes is used to adjust the contrast of the active loads in the preset period and make it calculable in different peak-to-valley difference forms. The decision tree model can preset the visual effect of the active loads in the preset period, making the details more visible, thereby improving the electrical equipment correlation parameters and grid topology of the active loads in the distribution network. The clustered distribution network active loads can classify the active load change rate, the peak-to-valley difference of the active load, and the load proportion of different types of electrical equipment, and identify the electrical equipment correlation parameters and grid topology of the active loads. This will provide more accurate and reliable active load data for the power factor of the distribution network in different periods, thereby clustering the performance and effect of the system. The effect of clustering active loads helps to improve the visual visibility of the active loads in the distribution network, making the status of the electrical equipment in the distribution network easier to detect and identify. This is crucial for accurately identifying and tracking the apparent power of electrical equipment in a complex distribution network environment. By providing clearer and more detailed active loads, the active load clustering technology can improve the accuracy, data sharing capability and reliability of the power factor of the distribution network at different times.

[0059] Semi-supervised learning is used to identify the line loss of the distribution network of the active load in the preset period, predict the state of the electrical equipment in the distribution network, and obtain the prediction result of the electrical equipment state. The specific steps of the distribution network preset period prediction are as follows:

[0060] Input the active loads in the preset time period into the semi-supervised learning, use the convolutional semi-supervised learning to identify the line loss of the distribution network, and obtain different characteristics of the line loss of the distribution network;

[0061] Based on the different characteristics of distribution network line losses, the convolutional neural network algorithm reduces the feature dimension and electrical equipment status recognition, and outputs the electrical equipment status prediction results.

[0062] The clustered distribution network operating active load is used as input and passed to semi-supervised learning. Semi-supervised learning extracts the characteristics of the operating active load layer by layer through a series of convolution and pooling operations. These different characteristics of distribution network line losses present the characteristics of the operating active load and can capture the key characteristics of the state of the distribution network electrical equipment. The different characteristics of the extracted distribution network line losses are input into the convolutional neural network algorithm to reduce the feature dimension and electrical equipment state recognition.

[0063] It can effectively process high-dimensional feature data and has good classification performance.

[0064] Semi-supervised learning can extract high-level semantic and structural features from clustered active loads in distribution network operation. These features can more accurately represent the properties and characteristics of the state of electrical equipment in the distribution network, and improve the accuracy of electrical equipment state identification and data sharing capabilities. The convolutional neural network algorithm can effectively handle multiple categories of electrical equipment state identification faults. It has good generalization and scalability, can identify different types of distribution network electrical equipment states, and provide corresponding distribution network preset time period predictions. Both semi-supervised learning and convolutional neural network algorithms can be quickly calculated on modern hardware accelerators. This enables the electrical equipment state identification process to be completed in real-time or near real-time time ranges, meeting the real-time performance requirements of the power factor of the distribution network in different time periods.

[0065] Using the support vector machine model and the cloud-based large model server, based on the distribution network preset period prediction and the preset period operation active load, the strength and deformation modulus data of electrical equipment in different operation periods are calculated. The specific steps for generating the distribution network load influence source are as follows:

[0066] Input distribution network preset period forecast, preset period operation active load;

[0067] Use different strength calculation models to obtain electrical equipment strength data, calculate the fluctuation range of deformation modulus through the modulus calculation model, and obtain the expression of load change;

[0068] Based on the representation of load changes, a large cloud model server input data is created, and a support vector machine model is used to perform active load forecasting and statistics on equipment failure locations in different operating periods, and output the sources of load impact on the distribution network.

[0069] By inputting the distribution network preset period prediction and the preset period operation active load, the system obtains the identification information and optimized visual input of different operation periods. Next, different strength calculation models are used to obtain the strength data of electrical equipment, and the fluctuation range of deformation modulus is calculated by the modulus calculation model to obtain the expression of load change. Based on the expression of load change, the cloud-based large model server input data is created, and the support vector machine model is applied to perform active load prediction and statistics of equipment fault locations in different operation periods, which produces accurate distribution network load influence sources. The integration of these steps enables the system to calculate the strength and deformation modulus data of electrical equipment in different operation periods, and generate distribution network load influence sources that can be used for subsequent analysis and decision-making. At the same time, the support vector machine model and the cloud-based large model server can effectively classify the background active load change rate and improve the data sharing capability of the system. In summary, the steps based on the support vector machine model and the cloud-based large model server provide important recognition and understanding capabilities for distribution network fault monitoring and early warning, and provide beneficial effects for the system's electrical equipment state recognition, environment recognition, and operation loss error prediction and other faults.

[0070] Use transformer voltage stabilization active load and effective power supply active load to monitor and warn power supply efficiency. According to the source of distribution network load influence, monitor and warn the ideal power supply efficiency of distribution network power factor in different periods. The steps to generate a visual interface of power supply efficiency per unit time are as follows:

[0071] Input the sources of load influence on the distribution network, calculate the variation range of power supply efficiency on the sources of load influence on the distribution network through the transformer voltage stabilization active load, and output the preliminary power supply efficiency;

[0072] The effective power supply active load is used to analyze the preliminary power supply efficiency, monitor and warn the ideal power supply efficiency of the distribution network power factor in different time periods, and generate a visual interface of power supply efficiency per unit time.

[0073] The input distribution network load influencing source is used as input, and the power supply efficiency unit time visualization interface is generated according to the ideal power supply efficiency. The unit time visualization interface contains the optimized power supply efficiency information, obstacles avoided, and places passed, etc. The power supply efficiency monitoring and early warning method that integrates the use of transformer voltage stabilization active load and effective power supply active load can bring many beneficial effects in the power factor of the distribution network at different time periods. This includes monitoring and early warning of ideal power supply efficiency, flexible adaptation to the environment, consideration of obstacles and constraints, and generation of information-rich power supply efficiency unit time visualization interface. The implementation of these steps can make the system power supply efficiency monitoring and early warning more accurate and reliable.

[0074] The operation loss error is predicted by starting and stopping the power supply electrical equipment and the effective power supply voltage. The operation loss adjustment and dynamic power supply efficiency statistics are performed through the power supply efficiency unit time visualization interface combined with the vector sum of the apparent power of the distribution network fault operation. The steps to obtain the start-up and stop optimization report of the operating electrical equipment are as follows:

[0075] Input the power supply efficiency unit time visualization interface, perform dynamic power supply efficiency monitoring and early warning through vector sum calculation, and obtain dynamic power supply efficiency;

[0076] Based on the dynamic power supply efficiency, the effective power supply voltage is used to perform fault assessment and strategy optimization of the dynamic power supply efficiency, and generate a start-up and stop optimization report for the operating electrical equipment.

[0077] The power supply efficiency unit time visualization interface is used as input, and dynamic power supply efficiency monitoring and early warning are performed through vector sum calculation. In this stage, the power supply efficiency unit time visualization interface is used to guide the system to select the ideal power supply efficiency in the distribution network environment. Then, based on the obtained dynamic power supply efficiency, the effective power supply voltage is used for fault assessment and strategy optimization. It comprehensively evaluates potential faults and benefits by considering the uncertainty of the environment and the state of the fault monitoring and early warning electrical equipment, thereby generating the best start-up and stop optimization report for the running electrical equipment. Finally, according to the optimized start-up and stop optimization report for the running electrical equipment, the running loss of the power factor of the distribution network in different periods is predicted and the dynamic power supply efficiency is adjusted. The method brings multiple benefits from the implementation perspective. It realizes dynamic power supply efficiency monitoring and early warning, and flexibly adjusts the power supply efficiency according to the real-time environment and fault monitoring and early warning requirements. Fault assessment and strategy optimization through effective power supply voltage improve the accuracy and efficiency of power supply efficiency monitoring and early warning. In addition, it is real-time and adaptive, and can adjust the start-up and stop optimization report of the running electrical equipment in real time according to the dynamic environment. Through accurate prediction of running loss error, the system can better understand the movement mode and running loss intention of the apparent power of the distribution network fault operation, and thus better select the power supply efficiency and the start-up and stop optimization report of the running electrical equipment. The comprehensive utilization of the method of starting, stopping and operating loss error prediction of effective power supply electrical equipment and effective power supply voltage and dynamic power supply efficiency adjustment can improve the fault monitoring and early warning performance of the power factor of the distribution network in different time periods, cluster its ability to adapt to changing environments, and improve accuracy and efficiency.

[0078] The Pearson correlation coefficient is used to confirm and analyze the distribution network fault based on the start-up and stop optimization reports of the running electrical equipment and the power factor of the distribution network at different time periods in different regional locations. The specific steps to obtain the distribution network fault impact range at different operating periods are as follows:

[0079] Based on the optimization report of starting and stopping the running electrical equipment, the fault sharing between the power factors of the distribution network in different areas and different time periods is established through the Pearson correlation coefficient, and the distribution network fault confirmation and analysis calculation model is generated;

[0080] Input the distribution network fault confirmation and analysis calculation model, use the bat positioning model to locate the power factor faults of the distribution network in different areas and at different time periods, and generate a fault location report;

[0081] Based on the fault location report, the correlation between distribution network faults is calculated through the correlation analysis algorithm to obtain the impact range of distribution network faults in different operating periods.

[0082] Based on the optimization report of starting and stopping the running electrical equipment, the fault sharing between the power factors of the distribution network in different periods at different regional locations is established through the Pearson correlation coefficient, forming a distribution network fault confirmation and analysis calculation model. In this way, the power factors of the distribution network in different periods at different regional locations can share the optimization report and real-time information of the start and stop of the running electrical equipment, so as to achieve effective collaborative decision-making and fault location. Next, the bat positioning model is used for fault location. According to the available resources of the system, the capacity of the power factor of the distribution network in different periods at different regional locations, and the fault priority, the most suitable fault is located for the power factor of the distribution network in different periods at each different regional location, and a reasonable fault location report is generated. Then, the power factors of the distribution network in different periods at different regional locations are executed according to the fault location report, and the correlation analysis algorithm is used to calculate the correlation between the distribution network faults. The correlation analysis algorithm performs real-time control and adjustment according to the current state and the state of the electrical equipment based on the feedback loop to ensure that the power factors of the distribution network in different periods at different regional locations are stable and accurate. Finally, after the distribution network fault is completed, the power factor of the distribution network at different time periods in different regional locations is used to obtain the distribution network fault impact range in different operating periods, including the results of fault execution, data collection and detection results. The fault unit time visualization interface provides detailed fault information, providing valuable data and reference for subsequent analysis, evaluation and decision-making. The use of Pearson correlation coefficient for distribution network fault confirmation and analysis calculation has the beneficial effects of efficient fault location, real-time communication of distribution network power factor at different time periods in different regional locations, stable distribution network fault and fault monitoring and early warning, and detailed fault unit time visualization interface. This method can improve the efficiency, accuracy and stability of the power factor of the distribution network in different time periods, and provide useful information and guidance for the implementation and decision-making of the system.

[0083] See also Figure 2A monitoring and early warning system based on active load prediction of distribution network is used to execute a monitoring and early warning method based on active load prediction of distribution network. A monitoring and early warning system based on active load prediction of distribution network is composed of an operating active load clustering unit, a distribution network line loss identification and electrical equipment status identification unit, a distribution network load influence source modeling unit, a power supply efficiency monitoring and early warning unit, an operating electrical equipment start-up and stop optimization reporting unit, and a distribution network power factor unit at different time periods in different regional locations.

[0084] The system can improve the accuracy of electrical equipment status detection in the distribution network. By running active load clustering and distribution network line loss identification and electrical equipment status identification units, it optimizes the distribution network environment to run active load network topology and realizes accurate electrical equipment status identification. The distribution network load impact source modeling unit generates the distribution network load impact source of the distribution network environment, providing detailed environmental information for power supply efficiency monitoring and early warning and fault monitoring and early warning decision-making, thereby improving the efficiency and accuracy of fault monitoring and early warning.

[0085] In addition, through the operation electrical equipment start and stop optimization report unit, the system can predict the operation loss mode according to the electrical equipment status, and take appropriate operation electrical equipment start and stop optimization reports to deal with the changing distribution network environment and electrical equipment status operation loss. The power factor unit of the distribution network at different time periods in different regional locations realizes the analysis and calculation of the power factors of the distribution network at different time periods in different regional locations and the confirmation of distribution network faults, optimizes fault location and resource utilization, and improves the performance and efficiency of the power factor of the entire distribution network at different time periods. In summary, a monitoring and early warning system based on active load forecasting of distribution networks integrates multiple units, and significantly improves the efficiency, accuracy and reliability of distribution network fault execution and electrical equipment status detection by improving accuracy, providing environmental information, realizing operation loss error prediction and operation electrical equipment start and stop optimization reports, and collaborative control.

[0086] The active load clustering unit is responsible for capturing the original active load from the distribution network load monitoring platform, performing active power analysis preprocessing, and finally using the decision tree model method to preset the active load during the period and output the clustered distribution network active load.

[0087] The distribution network line loss identification and electrical equipment status identification unit uses convolution to identify the distribution network line loss based on clustering the active loads in the distribution network, and uses the convolutional neural network algorithm to reduce the feature dimension and electrical equipment status identification, and outputs the electrical equipment status prediction results;

[0088] The distribution network load impact source modeling unit obtains the electrical equipment strength data through the support vector machine model based on the electrical equipment status prediction results, and constructs the distribution network load impact source based on the input data of the cloud-based large model server.

[0089] The power supply efficiency monitoring and early warning unit inputs the source of load influence on the distribution network, uses the transformer voltage stabilization active load to find the power supply efficiency during the preset period, and optimizes the power supply efficiency through the effective power supply active load to generate a visualization interface of the ideal power supply efficiency per unit time;

[0090] The start and stop optimization report unit for running electrical equipment uses vector sum calculation and effective power supply voltage to adjust the running loss and dynamic power supply efficiency statistics based on the ideal power supply efficiency unit time visualization interface, and outputs the start and stop optimization report for running electrical equipment;

[0091] The power factor unit of the distribution network in different time periods in different regional locations uses the Pearson correlation coefficient and the bat positioning model to confirm and analyze the distribution network faults between the power factors of the distribution network in different time periods in different regional locations based on the start-up and stop optimization report of the operating electrical equipment, and obtains the distribution network fault impact range in different operating periods.

[0092] The active load clustering unit clusters the grid topology of the active loads in the distribution network through active power analysis preprocessing and decision tree model methods, thereby improving the accuracy and reliability of electrical equipment status detection. The distribution network line loss identification and electrical equipment status identification unit uses convolution to identify the line loss of the distribution network, and combines the convolutional neural network algorithm to reduce the feature dimension and electrical equipment status identification, so that the system can quickly and accurately predict the state of electrical equipment in the distribution network and generate detailed electrical equipment status prediction results. This helps the system to fully understand the distribution network environment and the accuracy of electrical equipment status monitoring. The distribution network load impact source modeling unit combines the electrical equipment status prediction results, uses the support vector machine model to obtain the electrical equipment strength data, and constructs the distribution network load impact source of the distribution network environment based on this. This modeling method provides the system with more detailed environmental information to support the optimization of power supply efficiency monitoring and early warning and fault monitoring and early warning decisions. The power supply efficiency monitoring and early warning unit uses the transformer voltage stabilization active load to find the power supply efficiency in the preset period, and further optimizes the power supply efficiency through the effective power supply active load to generate an ideal power supply efficiency unit time visualization interface. Such power supply efficiency monitoring and early warning reports can reduce power supply efficiency redundancy and energy consumption, and improve the efficiency of fault monitoring and early warning. The operating electrical equipment start and stop optimization report unit applies vector sum calculation and effective power supply voltage, and predicts the operating loss error and dynamically adjusts the power supply efficiency according to the ideal power supply efficiency unit time visualization interface. Such an implementation strategy enables the system to make intelligent fault monitoring and early warning decisions based on real-time conditions and electrical equipment status operating losses, clustering the system's adaptability and reliability. The power factor unit of the distribution network at different time periods in different regional locations uses the Pearson correlation coefficient and bat positioning model to realize the distribution network fault confirmation and analysis calculation between the power factors of the distribution network at different time periods in different regional locations. This collaborative control report can optimize fault location and resource utilization, and improve the performance and efficiency of the power factor of the entire distribution network at different time periods. At the same time, after completing the distribution network fault, the system can obtain the distribution network fault impact range in different operating periods, including the results of fault execution, collected data and detection results, etc., providing valuable data and reference for the subsequent analysis, evaluation and decision-making of the system.

[0093] The active load clustering unit improves the grid topology of the active loads in the distribution network through active power analysis preprocessing and decision tree model algorithms, reduces the active load change rate and the peak-to-valley difference of the active load, and clusters the accuracy of electrical equipment status recognition. The distribution network line loss recognition and electrical equipment status recognition unit uses semi-supervised learning and machine learning algorithms to quickly and accurately predict the status of electrical equipment in the distribution network and generate electrical equipment status prediction results.

[0094] The distribution network load impact source modeling unit uses the electrical equipment state prediction results and support vector machine model and other technologies to construct the distribution network load impact source of the distribution network environment and provide detailed environmental information. The power supply efficiency monitoring and early warning unit uses the transformer voltage stabilization active load and the effective power supply active load to generate an ideal power supply efficiency unit time visualization interface to ensure that the system can efficiently and safely monitor and warn faults. The running electrical equipment start and stop optimization report unit predicts the operation loss error and dynamically adjusts the power supply efficiency according to the electrical equipment state operation loss mode and the ideal power supply efficiency unit time visualization interface through the start, stop and effective power supply voltage of the power supply electrical equipment. The distribution network power factor unit at different time periods in different regional locations realizes the distribution network fault confirmation and analysis calculation between the distribution network power factors at different time periods in different regional locations through the communication protocol and the peak-valley difference algorithm, and optimizes fault location and resource utilization. In summary, a monitoring and early warning system based on active load prediction of the distribution network comprehensively uses multiple units to significantly improve the efficiency, accuracy and reliability of distribution network fault execution and electrical equipment state detection by improving accuracy, providing environmental information, realizing operation loss error prediction and operation electrical equipment start and stop optimization report, and optimizing collaborative control.

[0095] Although embodiments of the present invention have been shown and described, it will be understood by those skilled in the art that various equivalent changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A monitoring and early warning method based on active load forecasting of distribution network, characterized in that: The following steps are involved: S1. Obtain the active load of the distribution network, and use the active power analysis model and decision tree model of the distribution network to classify the active load change rate, active load peak-to-valley difference and load proportion of different types of electrical equipment in the active load of the distribution network; S2, identifying the electrical equipment correlation parameters of the active load and the grid topology structure, and obtaining the clustered distribution network active load as the active load during the preset period; Using semi-supervised learning to identify the line loss of the distribution network of the active load in the preset period, predict the state of the electrical equipment in the distribution network, and obtain the electrical equipment state prediction result as the distribution network preset period prediction; S3, using a support vector machine model and a cloud-based large model server, using the distribution network preset period prediction, preset period operation active load, calculate the electrical equipment strength and deformation modulus data in different operation periods, generate the distribution network load impact source, and use the transformer voltage stabilization active load and effective power supply active load to monitor and warn the power supply efficiency; S4. According to the sources of load influence on the distribution network, the ideal power supply efficiency of the power factor of the distribution network in different periods is monitored and warned, and a power supply efficiency unit time visualization interface is generated. The operation loss error is predicted by starting, stopping and effective power supply voltage of the power supply electrical equipment. The operation loss adjustment and dynamic power supply efficiency statistics are performed through the power supply efficiency unit time visualization interface combined with the vector sum of apparent power of the distribution network fault operation, and an optimization report on starting and stopping the operating electrical equipment is obtained; S5. Use the Pearson correlation coefficient in combination with the startup and stop optimization report of the operating electrical equipment and the power factor of the distribution network at different time periods in different regional locations to confirm and analyze the distribution network fault, and obtain the distribution network fault impact range at different operating time periods.

2. A monitoring and early warning method based on active load forecasting of distribution network according to claim 1, characterized in that: The active operating load is collected and transmitted through the distribution network load monitoring platform to obtain the original distribution network active operating load; Input the original distribution network operating active load, use a normalization method to perform data limitation and data mapping, and obtain an active power analysis preprocessing operating active load; The input active power analysis preprocesses the operating active load, and based on the decision tree model, calculates the load proportion of different types of electrical equipment so that it can be calculated under different peak-to-valley difference forms to obtain the clustered distribution network operating active load.

3. A monitoring and early warning method based on active load forecasting of distribution network according to claim 1, characterized in that: Input the active load running in the preset time period into the semi-supervised learning, use the semi-supervised learning specifically convolution to identify the line loss of the distribution network, and obtain different characteristics of the line loss of the distribution network; use the different characteristics of the line loss of the distribution network, the convolution neural network algorithm reduces the feature dimension and electrical equipment status recognition, and outputs the electrical equipment status prediction result.

4. The monitoring and early warning method based on active load forecasting of distribution network according to claim 1 is characterized in that: Input the distribution network preset period forecast and the preset period operation active load; Use different strength calculation models to obtain electrical equipment strength data, calculate the fluctuation range of the deformation modulus through the modulus calculation model, and obtain the expression of load changes; use the expression of load changes to create a large cloud model server input data, use the support vector machine model to perform active load forecasting and equipment fault location statistics in different operating periods, and output the source of distribution network load impact.

5. The monitoring and early warning method based on active load forecasting of distribution network according to claim 1 is characterized in that: Input the source of the load influencing the distribution network, calculate the variation range of the power supply efficiency on the source of the load influencing the distribution network through the transformer voltage stabilizing active load, and output the preliminary power supply efficiency; use the effective power supply active load to perform power supply active load analysis on the preliminary power supply efficiency, monitor and warn the ideal power supply efficiency of the distribution network power factor in different time periods, and generate a power supply efficiency unit time visualization interface.

6. A monitoring and early warning method based on active load forecasting of distribution network according to claim 1, characterized in that: Input the power supply efficiency unit time visualization interface, perform dynamic power supply efficiency monitoring and early warning through vector sum calculation, and obtain dynamic power supply efficiency; Utilizing the dynamic power supply efficiency, using the effective power supply voltage to perform fault assessment and strategy optimization of the dynamic power supply efficiency, and generating a startup and stop optimization report for the running electrical equipment; Using the optimization report on starting and stopping the running electrical equipment, the fault sharing between the power factors of the distribution network in different time periods at different regional locations is established through the Pearson correlation coefficient, and a distribution network fault confirmation and analysis calculation model is generated; the distribution network fault confirmation and analysis calculation model is input, and the fault of the power factor of the distribution network in different time periods at different regional locations is located using the bat positioning model, and a fault location report is generated; The fault location report is used to calculate the correlation between distribution network faults through a correlation analysis algorithm to obtain the distribution network fault impact range in different operation periods.

7. A monitoring and early warning system based on active load forecasting of distribution network, characterized in that: The system includes: active load clustering unit, distribution network line loss identification and electrical equipment status identification unit, distribution network load impact source modeling unit, power supply efficiency monitoring and early warning unit, running electrical equipment start, stop optimization reporting unit, and distribution network power factor unit at different time periods in different areas and locations; The active load clustering unit is responsible for capturing the original active load from the distribution network load monitoring platform, performing active power analysis preprocessing, and finally using the decision tree model method to preset the active load for the time period and output the clustered distribution network active load. The distribution network line loss identification and electrical equipment state identification unit uses convolution to identify the distribution network line loss based on clustering the active loads in the distribution network operation, reduces the feature dimension and identifies the electrical equipment state through a convolutional neural network algorithm, and outputs the electrical equipment state prediction result; The distribution network load impact source modeling unit obtains the electrical equipment strength data through a support vector machine model based on the electrical equipment state prediction result, and constructs the distribution network load impact source based on the input data of the cloud large model server.

8. A monitoring and early warning system based on active load forecasting of distribution network according to claim 7, characterized in that: The power supply efficiency monitoring and early warning unit inputs the source of load influence on the distribution network, uses the transformer voltage stabilization active load to find the power supply efficiency in a preset period, and optimizes the power supply efficiency through the effective power supply active load to generate a visualization interface of the ideal power supply efficiency per unit time; The running electrical equipment start and stop optimization report unit uses vector sum calculation and effective power supply voltage, uses the ideal power supply efficiency unit time visualization interface to perform operation loss adjustment and dynamic power supply efficiency statistics, and outputs the running electrical equipment start and stop optimization report; The power factor unit of the distribution network in different time periods at different regional locations uses the Pearson correlation coefficient and the bat positioning model, and utilizes the start-up and stop optimization report of the operating electrical equipment to perform distribution network fault confirmation and analysis calculation between the power factors of the distribution network in different time periods at different regional locations, and obtain the distribution network fault impact range in different operating time periods.

Images