Research On Wind Power Acceptable Capacity Of Power Grid Based On Multi-objective Algorithm

Wind power,characterized by its pollution-free superiority,has become one of the most prominent and competitive renewable energy.Recently,the exploration and utilization of wind energy have been rapidly developed around the world.However,wind power is difficult to be accurately predicted because its strong uncertainty creates great challenges for system operations and may even lead it un-dispatching.Therefore,it makes the reasonable evaluation of wind power acceptable capacity a significant role for a power system.This is especially the case for some decision-making areas such as economic dispatch or optimal acceptance of wind power,i.e..This passage makes the research based on the application of wind power uncertainty and acceptance of wind power in the multi-objective optimization model,which is summarized as follows:(1)The research of improved multi-objective INSGA2 algorithm.Based on the extensive research of multi-objective optimization method and problems with constraints of uncertainties,NSGA2 is determined as the multi-objective optimization algorithm of assessing the acceptable capacity of wind power with random-coupled characteristics.As the traditional NSGA2 algorithm easily traps into prematurity while maintaining the overall distribution,and lacks of diversity and convergence,NDX operator is adapted to replace SBX crossover operator,besides,the stochastic simulation technology,accumulated fitness strategy,filled threshold jump selection method are introduced,so that the INSGA2 algorithm is proposed.The passage gives practicability and validity comparative analysis on the Pareto solutions of NSGA2,INSGA2 and PSO algorithms based on multiple standards high-dimensional test functions.The GD(Generational short)and SP(Spacing)indexes are adopted to evaluate,the results show that INSGA2 algorithm have significant improvements on integrity of Pareto front and uniform distribution,as well as algorithm convergence.(2)The research of wind power uncertainty model considering the statistical characteristics.First,this paper analyzes the distribution model of wind power prediction errors containing uncertain information,including Gaussian,Laplace,Cauchy model and the Versatile distribution based on the data of actual wind field.As the comparison and analysis has shown,the Versatile distribution model has more precise fitting precision on differernt wind farms and differernt forecasting time-scales,it can exhibit significant summit skewed characters well and there is no heavy tail.Then,a partitioning method which can adaptively divide the forecasted power range according to its historical data is firstly proposed in order to improve the availability and efficiency of the model which exhibits the uncertainty with a maximum precision.Qualitative(PDF and CDF curve-fitting)and quantitative(error evaluation indexes)comparative analysis indicates that the adaptive partitioning strategy can improve the model fitting effect of Versatile distribution.Last,the wind power forecast value with corresponding confidence intervals can be gained based on the shape parameters look-up tables,which laid a foundation for introducing the wind power uncertainty information into the evaluation of wind power acceptable capacity.(3)The evaluation model of wind power acceptable capacity based on day-ahead,day-in and real-time.The evaluation model of wind power acceptable capacity considering large-scale wind power inserting is put forward,and three objective functions are proposed according to the actual demand,which are the wind power acceptable capacity,operation cost of power system considering environmental benefits,and environment pollution.Then,the uncertainty information of wind power prediction is introduced into the model to establish constrained programming model.A case is introduced into the rolling model,and is solved by both INSGA2 and NSGA2 multi-objective optimization algorithms,the case study verified that the proposed INSGA2 algorithm was superior to get the Pareto optimal solutions.(4)Taking the universally accepted IEEE 118 bus system as an example,the multi-objective optimization rolling model of evaluating the wind power acceptable capacity is solved using INSGA2 multi-objective optimization algorithm.The Pareto frontier solution is gained by introducing different conditions,evaluation parameters,such as the average power generation cost,the rate of wind abandoning and the unit pollution emission of evaluation system.The results indicates that the proposed scheduling model is able to give an optimal dispatching scheme compared with other scheduling models,which gives important reference to the grid-connected operation and the planning of wind farm.