Research On Multi-Objective Optimal Dispatching Strategy Of Active Distribution Network Considering New Energy Consumption

In order to cope with the impact of large-scale distributed energy access such as wind power generation,photovoltaic power generation,electric vehicle charging and swapping facilities,distributed energy storage,etc.,the distribution network is moving towards an active distribution network with active control and operation capabilities Step forward.The active distribution network can actively adjust according to the actual operating state of the power grid,participate in the operation and control of the power grid,and realize the passive consumption of distributed renewable energy to active guidance.In order to give full play to the advantages of the active distribution network,improve the ability to absorb distributed energy,improve the power quality and safe and reliable operation of the distribution network,and provide better user services,this article fully considers the uncertainties of distributed energy and load.,Sequential characteristics,research and implementation of optimal dispatching strategies for active distribution networks.The main research contents include:(1)Considering the complexity,timing,heterogeneity and other characteristics of multi-terminal data,the fast prediction method of multi-terminal data combined with sliding time window and convolutional neural network is studied..A high quality sample input matrix for the convolutional neural network is constructed by analyzing the historical data through data cleaning and time series correlation.Based on the architecture of convolutional neural network,a fast prediction model of multi-terminal data is constructed.On this basis,a parallel strategy is designed for the convolutional layer and pooling layer in the process of forward propagation and back propagation of network training according to the structure of convolutional neural network for document number terminals.For multi-data terminals,a fast data parallel prediction method based on task averaging and priority ordering is proposed.A large number of simulation examples verify the rationality and effectiveness of the proposed method.(2)Considering the timing and uncertainty of distributed power and load,the multi-objective rapid optimization dispatching model of active distribution network is studied.Based on the obtained high-quality data,combined with the uncertainty of distributed power generation and load and time series model,the interval number time series expression of source load is established.Considering the localized consumption of distributed power supply and based on the superposition and correlation analysis of multiple interval number time series,the morphological similarity measurement expression of power supply and multiple loads is established to realize the maximum energy consumption.Based on the interval number theory,an interval multi-objective optimal scheduling model is constructed from three aspects:operating cost,reliability cost and maximum consumption of distributed power.For the proposed mathematical model,on the one hand,the interval radius and midpoint are used to express the interval number,and the interval is converted into a certain value for solution;on the other hand,a dynamic interval multi-objective optimization is proposed directly for the interval interval multi-objective model based on the improved firework algorithm method.In the optimization process,the improved algorithm is designed to solve the model parallelization from two aspects of population segmentation and multi-objective computation based on population segmentation.The effectiveness of the proposed multi-objective optimization method is analyzed by simulation of a typical IEEE example.(3)Considering the diversity of multi-objective Pareto solutions,a multi-objective decision analysis method with uncertainty is studied.On the one hand,aiming at the Pareto solution which is transformed into definite value,the influence of uncertainty is considered based on the set pair analysis theory,and the influence of uncertain factors on the ranking results is analyzed by sorting stability.On the other hand,for the Pareto solution represented by interval numbers,based on interval distance decision-making theory,decision analysis is made by comparing multiple interval numbers.The simulation example verifies and analyzes the two decision-making methods,which provide references for the multi-objective decision-making of different schemes.(4)Designed and realized the multi-objective optimization dispatching decision system of the active distribution network.Use pycharm2018.2.4 software Python language to program the proposed algorithm,and complete the system design and development through its own HTML,CSS and JQuery editor.The main functions of the system include:user information management,data information management,data prediction,optimal scheduling,and multi-objective decision-making.