Research On Energy Consumption Optimization Of Campus Building Based On Game Theory And Social-Technical System

The energy consumption of campus buildings includes internal energy consumption and external energy consumption.The main internal energy consumption equipment is heating,ventilation,and air conditioning system(HVAC),and the main external energy consumption equipment is charging station(CS).Energy consumption optimization can reduce the cost of electricity and provide an efficient and comfortable environment for campus buildings.However,there is a game between the most comfortable temperature and the most energy-saving temperature.This paper adjust the indoor temperature setting to balance human comfort and energy consumption cost.The existing campus building energy consumption optimization is based on the calculation of engineering system,which artificially separates the engineering system from the social factors.In fact,the coupling of building energy consumption and social factors is extremely close.It is a meaningful work to integrate social factors into the optimization of campus building energy consumption.This paper studies the optimization method of campus building energy consumption based on game theory and social technology system.The main research work is as follows:An energy consumption optimization method based on game theory and social technology system is proposed.The paper takes the building manager as the main object and the indoor temperature setting as the strategy,the game model of HVAC temperature control is established.The influence of campus building energy consumption on the productivity of indoor residents,indoor HVAC temperature setting and outdoor temperature is analyzed.The model aims at minimizing the energy consumption cost of campus buildings,and sets the indoor temperature as the optimal response strategy value,and finally achieves Nash equilibrium(NE),It can save energy cost and ensure the comfort and efficiency of indoor human activities.An energy consumption optimization method based on game theory and Multi-Agent Reinforcement Learning(MARL)is proposed.In this method,the Q-factor is updated by the reward obtained from the transformation of the Markov decision process(MDP),and the MARL algorithm is realized by continuously updating the Q-factor and reward.Compared with the traditional methods,the algorithm reduces the computational complexity of the game process and converges to the optimal result faster in time dimension.The simulation and empirical results show that the method can set the optimal indoor temperature according to the building energy consumption demand,real-time pricing(RTP),outdoor temperature and the number of indoor residents,so as to reduce the energy consumption cost of the whole campus building.An energy consumption optimization method of charging station(CS)based on game theory and maximum power point tracking(MPPT)algorithm is proposed.This paper presents a variable step MPPT algorithm to find the maximum power point of photovoltaic(PV)module and wind energy conversion system(WECS)through voltage and current deviation.Based on the historical charging information of plug in hybrid electric vehicles(PHEV),the future charging demand of PHEV is predicted by using artificial neural network(ANN).In the framework of game theory,the conversion power of PV and WECS,the predicted usage of PHEV and the energy consumption of indoor HVAC are gaming and learning,so as to reduce the energy consumption cost of campus buildings.The simulation and empirical results show that the proposed method can effectively meet the charging demand of PHEV and optimize the energy consumption cost of the whole campus building.Aiming at the campus power grid,a coalition non-cooperative game model is established to describe the energy consumption demand of campus buildings.Renewable energy generator(REG),utility company(UC)and their interaction are analyzed.This method can adapt to the dynamic energy consumption demand of campus buildings,REG,UC and real-time price(RTP).The simulation results show that the construction manager and REG form the final stable alliance through distributed hedonic rotation criterion in the coalition non-cooperation game model.UC form a stable power equilibrium distribution strategy through non-cooperative game which verifies the applicability of the coalition non-cooperation game method.