Research On Residential Air Conditioners Load Forecasting Based On Statistical Model

Load forecasting plays a crucial role in the power grid planning and scheduling and the development of demand response strategies.In recent years,as the world continues to warm up,the proportion of air conditioners during the peak of summer remains high,resulting in a contradiction between the supply and demand of the power grid and the increasing peak-to-valley difference of the system.To alleviate this situation,the application of demand response technology to load optimization is strongly advocated all over the world.How to establish a acceptable forecasting model,forecasting of residential air conditioners,so as to analyze the ability of residential air conditioning to participate in demand response has become an urgent problem to be solved.This paper has conducted in-depth research on the load forecast of residential air conditioners.Firstly,on the basis of considering the electrothermal conversion process of air conditioners and the heat transfer process inside and outside the building,the working state model of single air conditioner based on physical principle is analyzed,and the influence of parameters in the equivalent thermal model on the operating characteristics of air conditioners is studied.Then,considering the probability and statistical characteristics of each parameter,a statistical model of air conditioning is established.Based on the above research,a load forecasting method based on the statistical model of air conditioners is proposed,and the in-day and day-ahead loads of workdays and weekdays of high-rise and multi-storey buildings are forecasted.Finally,the demand response capability assessment process is proposed,and the residential air conditioner scheduling strategy based on changing user set temperature and load translation is compared.The main work includes:(1)Based on the description of single residential air-conditioner electrothermal conversion and house thermodynamics process,the second-order equivalent thermal parameter model is analyzed,and the calculation of building parameters in the equivalent model is deduced.The local sensitivity analysis method is used to study the parameters of air conditioning.The influence of building parameters and meteorological parameters on air conditioner operating characteristics(including single run time,minimum cycle,air conditioning energy consumption).(2)According to the randomness of residential users' air conditioners,the probability distribution function of each parameter in the working state model of an air conditioner,and a statistical model of residential air conditioners is established.The Monte Carlo method is used to obtain the cluster power of the air conditioners,so as to forecast load of the workdays and weekdays of the high-rise residences and multi-storey dwellings,and analyze the results.A single-sample K-S test analysis of the peak load results shows that the peaks of multiple forecasted loads follow a normal distribution.(3)The residential building demand response capability assessment method is proposed,and the demand response capability evaluation index is defined.For the demand response regulation strategy based on changing the set temperature and the translation load,the dayahead and in-day load forecasting of high-rise residential buildings and multi-storey dwellings is carried out respectively,and the simulation calculation results are obtained.The demand response strategy,which indicates load shifting,is more conducive to reducing load peaks and making the load curve more gradual.