Research On Optimal Scheduling Of Microgrid Based On Short-Term Power Load Forecasting

With the high-quality development of our economy,people’s demand for electric energy is increasing day by day.In order to ensure the quality of electricity and reduce the problems of excessive energy consumption and environmental pollution,clean energy generation has been rapidly developed.The efficient utilization of clean energy can not only save the optimal scheduling time of microgrid,but also ensure the reliability of power supply.One of the key technologies to realize power grid dispatching is to predict short-term power load quickly and accurately.Therefore,in order to achieve the optimal economic operation of microgrid,it is of great significance to carry out research on shortterm power load prediction and optimal scheduling of microgrid.The main research contents and conclusions of this paper are as follows:First of all,in order to solve the problem of poor accuracy of load prediction,the temperature,humidity,precipitation,time and historical load data were determined as the input of the prediction model through data processing and analysis of influencing factors,and the BP neural network short-term load prediction model optimized based on Cuckoo algorithm was established.The cuckoo algorithm is used to optimize the weight threshold of BP neural network to omit the process of subjective parameter selection.Combined with practical examples,the prediction effect of BP network and CS-BP model is compared and analyzed.The results show that the prediction accuracy of BP network optimized by Cuckoo algorithm is improved,but the prediction time is longer.Then,aiming at the problems of long learning time and insensitivity to time series data in the BP network optimized by Cuckoo algorithm,the least square support vector machine model with faster learning speed is introduced for prediction.In order to further realize accurate prediction,the improved Cuckoo algorithm was used to optimize the LSSVM model and obtain the optimal prediction parameters.Meanwhile,the main influencing factors of the load analyzed above were taken as the prediction input to construct the combined load prediction models based on CS-LSSVM and ICS-LSSVM respectively.Combined with practical examples,the prediction results,prediction errors and evaluation indexes of ICS-LSSVM model,CS-LSSVM model and CS-BP model mentioned above are compared and analyzed.The results show that the prediction accuracy and prediction speed of ICS-LSSVM model are better improved.Finally,in order to achieve the optimal economic operation of the microgrid system,the improved cuckoo algorithm proposed above is used to optimize the scheduling.Based on the accurate short-term power load prediction model established above,and on the basis of determining the objective function of optimal scheduling of the micro-grid,the minimum sum of the normal operation and maintenance cost and environmental cost of the system is selected to consider the constraint problem in the actual operation of the micro-grid.A micro-grid economic dispatching model including photovoltaic array,fan,micro gas turbine and storage battery is constructed.The running cost and scheduling results of Cuckoo algorithm before and after the improvement are compared and analyzed with the example.The results show that the improved Cuckoo algorithm realizes the optimal economic operation of microgrid.Figure[49] Table[15] Reference[81]...