Impact Of Medium And Long Term Electricity Price Changes On Demonstration Information Release

With the implementation of demand side management,the value of response behavior of power users is obtaining more and more attention.More and more new technologies for exploring user side response potential and demonstration projects have been carried out in the demand side.In order to guide appropriate number users to participate in demonstration projects,it's necessary to predict the response behavior of users and decide how to publicize the demonstration project to the users.The purpose of this paper is to set up a model to determine optimal demonstration information release strategy in this background.Considering the power market reform,the promotion of volatility of the power price,and the transmission of electricity price fluctuation from the wholesale side to the demnad side,a model called particle swarm optimization-gradient descent hybrid neural network was proposed,which was helpful to researching influence of electricity price change on the optimal demonstration information release strategy model.based on the neural network to establish a long-term online electricity price forecasting model,and according to the gradient descent method to search the superior performance of the defect and obvious direction,combined with the particle swarm algorithm is proposed,which can search and can quickly search the solution model of particle swarm hybrid neural-gradient descent based on the network.Secondly,according to the transmission relationship between the online price and the selling price,the predicted single system price is converted into the sale side price.At first,a model of medium and long term generation price forecasting was built,and a model combining the globing searching of particle swarm optimization and directionality of gradient descent called particle swarm optimization-gradient descent hybrid neural network was proposed.Then herding behavior was used to study on the response behavior of users to the information,and a optimal release strategy model of information was established based on that.The role of electricity price in the user response behavior was discussed,and the influence of the price change on the information release strategy was also discussed.At last,an example was given to illustrate the feasibility of the neural network model in the electricity price forecasting and the superiority of the particle swarm gradient descent hybrid neural network.Aiming at a certain power price level,an optimal release strateg of ice storage air conditioning system demonstration information was made,which confirmed the necessity of the development of publishing strategy.The variation range of time-of-use price is roughly determined,and the influence of the change of time-of-use price on the optimal release strategy is explained through the calculation results.Examples showed that users would respond differently to the different information,different types of information could attract different users,and the change of power price would affect the users' response behavior.The particle swarm optimization-gradient descent neural network was also proved to be feasible.It provide a broad range of search and is with fast convergence rate.