Research On Fault Risk Assessment And Safety Early Warning Technology Of Active Distribution Network

With the continuous implementation of the "double carbon" policy,the new power system with renewable energy as the main body,China’s energy and electricity show the characteristics of energy clean,and along with the deep integration of intelligent control technology and low-carbon energy technology,renewable energy access to the distribution network number grows year by year,the network structure is increasingly complex,the passive recipient of the distribution network into active and controllable power consumption and output side,that is,active distribution network.In addition,the distributed power supply and its intermittent nature bring new risks and challenges to the safe and sound operation of the distribution network.Therefore,in order to cope with the poor applicability and low efficiency of traditional assessment and warning methods,it is necessary to consider different types of influencing factors in the distribution network,analyze the distribution network operation accurately in advance,and bring into play the active control capability of the active distribution network to reduce risks.Therefore,it is necessary to conduct a comprehensive and accurate assessment and early warning of the active distribution network.This paper studies the problem of risk "assessment and early warning" of active distribution network with distributed generation.In order to mine limited valuable information from massive and high latitude data of distribution network and improve the accuracy of risk research and judgment of active distribution network.This paper proposes an integrated approach to active distribution network fault risk assessment and security warning,and the main research works are as follows:(1)For the risk assessment of active distribution network,the risk base of distribution network operation under multiple risk factors of wind power,photovoltaic power fluctuation and load power fluctuation is analyzed from the perspective of both operational safety and reliability,and the evaluation system of multi-dimensional distribution network is constructed,using the hierarchical analysis method based on entropy correction to assign weights to the factors of safe and stable operation,calculate various specific indicators,and then study and judge the situation of distribution network.(2)Analyze multi-source risk factors,this paper focuses on risk assessment methods using improved Monte Carlo-Latin hypercube sampling for risk events caused by multiple randomness causes.The focus of this paper is to simulate the risk factor environment based on a probabilistic model for analyzing risk sources,simulate real out-of-power scenarios,and thus validate the evaluation of the system.This includes the use of an improved scenario reduction technique to reasonably reduce the output conditions of each generation in the simulated operation to improve the accuracy of the model calculation.The risk indicator values are calculated by the new risk assessment method,thus enabling quantitative evaluation of the operational risk of the main distribution network under a variety of stochastic factors.(3)In terms of risk early warning.In this paper,a failure risk warning model based on PSO-BP network is proposed,and a prediction error probability distribution model is established for the risk causes arising from the prediction errors.Feature extraction is performed on the processed samples,and the weights are adjusted and optimized by the particle swarm algorithm for different input influencing features.At the same time,establish early warning standards and release alarm information in time to identify the category and level of faults.The simulation results show that the depth neural network combination algorithm model has achieved very good results in both prediction accuracy and operation speed.This paper interacts between the two modules of assessment and early warning.This paper proves the efficiency of mechanism-data combination in the research system.On the one hand,the scenic load uncertainty is modeled and the uncertainty of generation output and load demand is analyzed,solves the problems of poor simulation accuracy and slow calculation of various random variables by the improved scenario method,and provides theoretical support for risk assessment by establishing a perfect risk assessment index system,On the other hand,the deep neural network is used to decouple multi scene and multi-source information,solve the difficult problem of input analysis of various influencing factors,simplify the early warning work,and establish the fault risk assessment and early warning system of the overall active distribution network,so as to achieve "fast","accurate" and"stable" in the research and judgment of active distribution network faults,and make the distribution of risk levels more intuitive.The problems existing in the active distribution network can be detected in time,which provides effective help for its stable operation.