Short-term Load Forecasting Method Of Distribution Network Considering Heat Island Effect Factor And Study On Optimum Reconfiguration Of Network

Load forecasting is not only an important part of power network dispatching,but also the basis of optimal operation of power grid.The effect of load forecasting directly affects the safety,stability and economic benefit of power grid.The existing load forecasting methods can meet certain precision requirements.However,due to the complexity of distribution network structure and the development trend of load diversity,the existing load forecasting methods need to be improved to meet the development needs of power grid.With the aggravation of urbanization process in Zhengzhou,the urban heat island effect caused by it becomes more and more serious which becomes a factor that cannot be ignored in load prediction.The optimization and reconstruction of distribution network can reduce the loss of power network,realize economic operation,and achieve the purpose of energy saving and environmental protection.This paper mainly studies the short-term load forecasting model of distribution network considering heat island effect and the optimization and reconstruction of network.Firstly,a load forecasting method based on improved genetic algorithm is proposed to optimize the extreme learning machine which generates input layer weights and hidden layer thresholds randomly leading to network instability.An improved genetic algorithm based on hill-climbing method is used to optimize the weight and thresholds of extreme learning machines,so as to obtain the optimization model with strong stability and high prediction accuracy.The prediction results of the optimization model are compared with BP network and extreme learning machine to verify the effectiveness of the proposed method.Secondly,the intensity of urban heat island effect is calculated by using the temperature data of Zhengzhou urban and suburban areas in recent two years,and the seasonal and diurnal variation characteristics of urban heat island effect in Zhengzhou are analyzed.The influence of heat island effect on distribution network load in Zhengzhou city is studied and their correlation is analyzed.The short-term load forecasting model with the heat island effect is established,and the model is verified by MATLAB.Finally,the improved genetic algorithm is used to optimize the extreme learning machine prediction model considering heat island effect,and the load of Zhengzhou railway station distribution network is predicted.The distribution network power flow is calculated based on the predicted results,which lays a foundation for network optimization and reconstruction.By improving the genetic algorithm to search the distribution network structure with minimum network loss,the optimal operation mode is obtained,and the optimal reconstruction of Zhengzhou distribution network is realized.The research results show that load forecasting model can effectively improve the short-term load forecasting accuracy of distribution network.In addition,the operation mode of the minimum network loss is obtained through the optimization and reconstruction of the network,and the optimal operation of the network is realized.It can be applied to the optimal dispatching of power system with strong practicability.