Research On Short-term Load Forecasting Based On Variational Mode Decomposition And IWOA-LSSVM

Short-term load forecast of the power system is a necessary condition to ensure the safety and stability of the power system,and it is also the inevitable result of the cost control of the power system.The production and management department of the power system can carry out a series of production preparations based on its forecast results.For instance,it can guide the power generation plan in advance to cope with the peak power consumption,or predict the underestimation of power consumption in advance,and then reduce the power output to reduce the production cost.The operating constraints of power production and transportation equipment can also be set based on the predicted results to ensure the safe operation of related equipment and delay the life of the equipment.These measures play an extremely important role in the safety,stability and reliability of the power system.However,the accuracy of the prediction results obtained by traditional prediction methods is generally poor.Therefore,more accurate power system load prediction has always been a hot topic in the power field.Least Squares Support Vector Machine(LSSVM)has excellent generalization ability and has a good performance in the field of load forecasting.However,LSSVM itself also has the problem of blindness for parameter selection.For solving the blindness of the setting of the relevant parameters of the LSSVM model and the shortcomings that the prediction accuracy cannot meet the requirements,the improved whale optimization algorithm(IWOA)is used to optimize its parameters to improve the prediction accuracy of the entire prediction model.This paper introduces population mutation strategy and search extension strategy to improve the whale optimization algorithm(WOA).In the first step of WOA operation,population mutation operation is added to increase the diversity of whale populations,and then the search extension strategy is used to expand the WOA space Search scope.The improved whale optimization algorithm(IWOA)and the least square support vector machine(LSSVM)model are combined to propose an improved IWOA-LSSVM model.Simulation experiments show that the model can effectively improve compared with the WOA-LSSVM model.Forecast accuracy of short-term load forecasting.VMD is adopted because the power load will be affected by various factors such as society and environment,resulting in a large amount of noise pollution in the load data,which will lead to non-linearity and instability of the data.VMD can effectively extract the local features of load data and eliminate noise pollution.In order to improve the calculation efficiency of the overall model,this paper proposes a self-reconstruction strategy,which first sets the search range of the modal number,and then introduces Fast Fourier Transform(FFT)to determine the initial value,and then uses sample entropy to reconstruct.To obtain the final,and finally determine the specificity of the penalty factor through the principle of mutual information correlation,these tasks can effectively reduce the workload of the overall prediction model to achieve the optimal effect.In order to verify the validity and reliability of the directions proposed in this article,a set of foreign and domestic data are selected for simulation experiments.The simulation experiment results show that the prediction results of the prediction model proposed in this article have been significantly improved on different prediction days.This has a good reference value for the short-term load forecasting methods and theoretical research of the power system.