Research On Load Forecast Of Linqing Electric Power System

In the process of gradually building urban power supply network into intelligent network,the factors such as its cost and stability have attracted more and more attention.According to relevant statistical data,the electricity consumption of medium-sized cities is 29 billion kwh/year.If the accuracy of electricity load forecasting can be increased by one percentage,there will be hundreds of millions of benefits.Therefore,the test speed and precision of the irregular load forecasting technology need to be improved urgently.Load forecasting technology which is widely used at present is not comprehensive enough to deal with the typical characteristics of load at the beginning of technical formulation.On this basis,the effect of improving test accuracy and speed is negligible,and the generality can not withstand the test.The wind speed factor in the test system has great uncertainty and can not be well added to the test technology.Intelligent algorithm can solve this problem,but at the same time,it also brings the problem of speed decline caused by the reduction of operation efficiency.Based on cloud computing and cloud model,this paper optimizes the precision and speed of the new load forecasting method,and verifies the success of the optimization by simulation.In this paper,the cumulative effect of air temperature and the Least Square Support State Machine(LSSM)optimized by cloud model are studied in depth.By extracting the details of historical load sample data and considering the characteristics of short-term load forecasting,the correlation between sample data and various factors of forecasting method is calculated.Cloud computing is considered as the influence.The biggest factor is used as training sample,and the concept of dislocation sample is introduced into training sample,which improves the accuracy of the method.The simulation verifies the feasibility.The precision is also improved by 2%after using cloud model optimization.In dealing with the problem of uncertainty,but compared with the traditional method,the processing of deterministic factors.The effect has declined.This paper divides the influencing factors of power supply network into determinacy and uncertainty,and combines the traditional prediction model with the model based on cloud computing.The least squares support state machine under cloud model optimization and particle swarm optimization respectively pins the uncertainties and determinacy factors,and gets the prediction results by weighted calculation.The prediction accuracy is improved comprehensively.In this paper,the existing models are applied to the historical load data and meteorological data of Linqing City.The results show that the use of cloud models not only improves the accuracy,but also brings the problems of increasing the running time and flying.The introduction of cloud computing technology has improved the speed of prediction.The combined forecasting simulation based on cloud model presented in this paper proves that compared with traditional load forecasting,this method improves both forecasting time and accuracy.