Research On The Interaction Between Comprehensive Energy Supply And Demand In The Zone Considering The Demand Response Of Buildings

Nowadays,smart buildings are concerned by more and more experts and scholars.However,in terms of load modeling of different types of smart buildings,considering the thermal effect and dynamic electricity price mechanism,the complexity of load modeling increases,and the traditional modeling methods and existing methods are difficult to meet the requirements of modeling accuracy of different types of smart buildings.Secondly,in terms of load forecasting of different types of industrial zone with smart buildings,after considering the influence of weather factors and dynamic electricity price game mechanism,the forecasting of smart building load and electric vehicle load in the industrial zone is often ignored,resulting in low satisfaction of residents in the building.Finally,in terms of energy optimal scheduling and multi-objective optimization of smart buildings,after considering the thermal effect,dynamic electricity price game mechanism,the influence of weather factors and electrical connection,the multi-objective optimization problem becomes more and more complex.This paper puts forward corresponding solutions to the above problems.In terms of load modeling of different types of smart buildings,the improved WGAN-GP based on a priori condition verification,namely conditional WGAN-GP,is introduced to model different types of smart buildings considering the influence of thermal effect and dynamic electricity price mechanism.The example analysis shows that the modeling accuracy of office buildings,residential buildings and commercial buildings is improved by 3.6%,4% and 3.8% compared with the original algorithm.In terms of different types of load forecasting of industrial zone with smart buildings,the attention long short term memory（Attention LSTM）algorithm based on the attention mechanism of dynamic Bayesian network（DBN）structure is constructed.Based on the correlation theory and structure,the algorithm improves the forecasting accuracy of all kinds of loads,especially fractional loads.The example analysis shows that the accuracy of this algorithm is 93.2%,93.2%,94.2% and 93.9% respectively.In terms of energy optimal scheduling and multi-objective optimization of smart buildings,a new method of multi-energy optimization in smart buildings is proposed.Firstly,the transfer retention ratio（TRR）is added.Secondly,the improved normal boundary intersection（INBI）algorithm is formed by improving the normal boundary intersection（NBI）algorithm through adaptive weighted sum,adjust uniform axes method and Mahalanobis double base point distance.The parameter TRR and INBI algorithm are used to solve the multi-objective optimization problem in smart buildings and improve the regulation efficiency.In order to meet the needs of decision-makers for evaluation indicators,the average deviation of the parameter TRR is reduced by 60%compared with the case without TRR.The example analysis shows that this method is superior to the existing algorithms in three optimization objectives.Among the three optimization objectives,the equipment cost,power supply cost and residents’ comfort are reduced by 8.2%,7.6% and 1.6% respectively.Based on the demand response of smart buildings,this paper analyzes the supply-demand relationship of smart buildings and industrial zone containing smart buildings,and improves the corresponding algorithms from the aspects of load modeling,load forecasting and multi-objective optimization,which improves their accuracy respectively,so as to make the regulation of smart buildings based on supply-demand relationship more effective.