Research On Demand Response Of Intelligent Building Based On Improved MOEA/D And Master-slave Game Strategy

Based on the background of smart buildings,this paper conducts demand response research on the electricity consumption of building users based on price and incentives.The framework of the smart building home energy management system system is proposed,and the mathematical model of load optimization dispatching is constructed with the goal of the lowest user electricity cost,the largest new energy consumption and the smallest power peak-valley difference.The MOEA/D algorithm is improved from several aspects of weight vector,crossover operator selection,offspring population correction and external archive update,so that it can adapt to the current load scheduling model,speed up the convergence speed and improve the robustness of the algorithm.In addition,taking into account the conflict of interest between e-commerce sellers and users,game theory is introduced to construct an incentive demand response model based on master-slave game,which increases users’ energy experience while reducing electricity bills,and promotes the economic development of e-commerce sellers.The main work of this article is divided into the following parts:(1)Household energy management system is Established and the intelligent building load dispatch model is constructed.The role of HEMS is reviewed and its theoretical system and structural framework are constructed.By analyzing user behavior and habits,users are divided into four categories,and user loads are divided into interruptible loads,transferable loads,and charging according to the characteristics of electricity consumption.Energy storage load and new energy power generation load are four types of loads,each type of load model and its constraints are constructed separately,and the intelligent building load dispatching model is constructed with the goals of user electricity charges,new energy consumption rate and power peak-valley difference.(2)Intelligent building load optimizes dispatching by improving MOEA/D.Because MOEA/D has better performance when dealing with continuous multi-objective problems,this paper adopts MOEA/D as the optimization algorithm to solve the Pareto frontier.For the problem that the discrete weight vector may lead to local optimality,the idea of generalized decomposition is introduced to generate an adaptive weight vector combined with uniform weight distribution.In order to improve the robustness of the algorithm,a strategy for adaptive selection of crossover operators based on historical experience is designed for crossover operators,which counts the crossover operators used by non-dominated solutions stored in external archives,and uses roulette.Adaptive selection strategy;in order to avoid the algorithm from falling into the local optimum,the idea of characteristic constraint mapping is introduced to modify the offspring population points,which indirectly increases the search range of the algorithm and improves the diversity of the algorithm population;finally,the calculation based on Euclidean distance The sparsity of non-dominated solutions is used to update the external population to improve the quality of stored non-dominated solutions.(3)Users obtain price rewards from e-commerce vendors by transferring load and electricity time periods,thereby constructing an incentive demand response model based on a master-slave game.Through the explanation of the concept of non-cooperative game and master-slave game and the method of solving the elements,an incentive-based demand response model of e-commerce seller and user master-slave game is proposed.Through the transfer of transferable loads by users,the retail e-commerce company provides a price reward mechanism to encourage more users to participate in demand response and choose their electricity price to use,analyze the demand relationship between the retail e-commerce company and the user to construct the utility function of both parties.In the solution method,the electricity retailer announces the electricity price for users to choose in each electricity time period.The user makes the optimal electricity consumption behavior strategy under the condition that the user meets his own electricity demand according to the choice and does not greatly affect the satisfaction of electricity consumption.,And then feed back the user’s power strategy to the power supplier.The power supplier uses this to make corresponding changes to the power supply price for each power usage period to increase its maximum profit,and finally find a balanced optimal scheduling method and power price strategy.(4)The building load collaborative optimization scheduling system and experimental platform are established.In order to better implement the user-side demand response of intelligent buildings,price-based DR and incentive-based DR should be combined to guide intelligent building load dispatch.Due to the large changes in users’ electricity consumption habits during holidays,this article will switch user types between office users during work days and holidays,use a price-based optimal scheduling model to guide scheduling during holidays,and use an incentive game model to solve users during work days.Best electricity behavior.In order to increase the user’s sense of participation and interaction,a human-computer front-end interactive GUI interface is constructed for dispatchers and users to use.