Research On Demand Response Optimization And Behavior Of Power User Side

With the increasingly widespread use of kinds of distributed energy generations and new electrical appliances,and the improvement of individuation of customer demands,smart power utilization and its technology and architecture concepts emerged.In this context,advanced measurement and communication technology and intelligent load control technology is increasingly mature and being perfect.This led to the implementation and development of demand response technology for the goal of high efficiency bidirectional interaction electricity using.In order to obtain the economic and environmental benefits while improving the service quality of users,study the behavior of electricity consumption,actively integrate the resources of user side and optimally schedule energy becoming a positive and effective measure.This thesis constructs the building cluster energy system(BCES)which can realize the reciprocity,storage,control of energy flow and information interaction within the cluster,then proposes a grid-friendly building cluster energy optimization scheduling strategy according to the time-of-use(TOU)mechanism and power model of controllable elements in system.Considering that the uncertainty of wind and photovoltaic power generation,a coordinated economic scheduling mathematical model with the multi-form energy storage,combined cooling heating and power(CCHP)and refrigeration units as its control objective that based on the chance-constrained programming aiming at the lowest total operating cost of BCES is established,then the sequential Monte Carlo simulation(SMCS)algorithm and LINGO are used to solve the problem above.The results of example prove the model can guide BCES participating the peak load regulation,thus friendly integration can be implemented.Then,This thesis proposes a bottom-up model to subtly forecast household daily load profiles based on user behavior.With the analysis of user behavior,this paper sets up a state transition matrix and obtains the status of resident through non-homogeneous Markov chain.The residents' activities are associated with the corresponding electrical load,for establishing the joint probability distribution of appliance's using duration.Also,this paper models the household electrical appliances to get the daily load curve in different types of day,weather data,the number of families by sequential sampling method.The prediction simulation example shows a high accuracy of the load curve forecasting,also the validity of the model.In order to reveal the degree of users' response to electricity price,meteorology,type of day and other related factors,a comprehensive demand response model is proposed.This model realizes the simulation of user demand response behavior under TOU,and can provide basic guidance for electricity companies to develop Demand-side management(DSM)strategies.