Research On Non-intrusive Load Dispatching Method Of Electric Vehicles Based On Demand Response

At present,with the continuous development of China’s social and economic level,fossil energy is gradually in short supply,and environmental pollution is becoming more and more serious.In this context,electric vehicles have gradually replaced traditional fuel vehicles as an important means of transportation for residents.At the same time,the annual growth rate of China’s power generation can no longer meet the electricity demand of various regions.If the power load can be dispatched from the demand side,the smooth operation of the power grid can be guaranteed to a certain extent.With the rapid increase in the number of electric vehicles,they have become an important scheduling target in demand response.The existing demand response methods have problems such as single scheduling method object,lack of user trust,and difficulty in obtaining user electricity data.With the construction of the ubiquitous power Internet of Things and smart grids,it is possible to collect,communicate and transmit residential electricity data.Through residents’ electricity consumption data,electric power operators can analyze the potential of different electric equipment in terms of demand response,and formulate reasonable dispatch plans for residents’ electricity load.Aiming at the residential electricity load including electric vehicles,this paper proposes a non-intrusive demand response dispatching method,combining non-intrusive load monitoring technology and demand response dispatching strategies.The main work and results of this paper are as follows:1.Research on event detection algorithm based on sliding window.On the basis of the use of statistical features,the use of sliding window and threshold determination for event detection can determine the start and stop time of electrical equipment based on the total active power of residential houses,and extract its load events.In the stage of simulation experiment and result analysis,the performance of the event detection algorithm proposed in this paper is evaluated through evaluation indicators such as missed detection rate and false detection rate.The experimental results show that the algorithm can effectively complete the task of extracting load events.2.Research the load decomposition model based on deep learning.The mathematical principles of deep learning in load decomposition are introduced,and the applications of three kinds of neural networks in load decomposition are introduced.A load decomposition method based on long-and short-term memory neural network is proposed.The detected load events are used as input to learn the relevant characteristics of electrical equipment during the operation,so as to realize load decomposition and monitoring.The results of the calculation example show that compared with other neural network models,the proposed method achieves the best results in indicators such as mean square error and absolute average error,and can complete the task of load decomposition.3.Research on demand response scheduling methods.Analyze the potential and elasticity of different loads in demand response through the residential electricity load data extracted by non-intrusive load monitoring technology.A system framework combining non-intrusive monitoring technology and demand response dispatching methods is established.Taking electric vehicles as an example,a demand response dispatching strategy based on time-of-use electricity prices is proposed.The simulation experiment results show that the non-intrusive load monitoring technology can solve the problem of obtaining user electricity consumption data,and through the demand response scheduling method,effectively reduce the total load during the peak electricity consumption period.