| With the improvement of living standards,the popularity of high-power electrical appliances in households and residential electricity consumption has increased significantly,meanwhile because residential electricity consumption takes a big share in total electric load,especially during the peak hours in grid,the grid suffers lager peak regulating pressure.With the development of smart grid technology,the ability of resident to participate in demand response is increasing due to the installation of Advanced Metering Infrastructure in households and the popularity of smart appliances which enables interactions between smart grid and electricity customers.So that the residential customers can actively respond to the incentives through the demand response project,adjust their consumption mode to participate in the peak load shifting and flexible the grid load features.Thus it's necessary to evaluate residential peak regulating potentiality reasonably.Based on the above mentioned background,this paper is constructed as follows,mainly focusing on the characteristics and evaluation methods of residents' peak shaving potential considering the demand response.Firstly,considering the demand response,the evaluation indexes of residential users'peak shaving potential are put forward.Based on the concept of generator peak shaving ability,in this paper,the residential peak regulating potentiality is defined and quantified from three aspects:peak regulating capability,peak regulating probability,and peak regulating rate.Then,this paper studies the character extraction method of residents' peak regulating potential under the environment of big data,in order to quantify the potential of peak regulating.To extract the information that characterizes the peak regulating potential of customers from the massive data,in the meantime,to improve the efficiency and security of data transmission and reduce the pressure of data communication and computing,sparse coding approach,which is commonly used in signal processing,is firstly adopted to extract features of residents'peak regulating potential.During the process of feature extraction,a non-negative K-SVD algorithm is adopted to solve the orientation optimization problem.And later,a multi-layer perceptron neural network model is applied to establish the mapping relation between extracted features and peak regulating potentiality,so that the evaluation is realized.Lastly,this paper takes 50000 residents'daily load curves as an example to carry out the evaluating process,then the effectiveness study of feature extraction,the performance study of data compression and comparative study with K-means clustering and principal component analysis are conducted to verify its effectiveness and superiority.Lastly,the paper studies the impact of time-of-use price on residents'peak regulating potential.Firstly,the concept of elastic coefficient of demand response is introduced to analyze the change of residents' load profiles before and after time-of-use electricity price.And the changes of the load profiles reflect the users'response to the time-of-use price incentive,that is,the users' willingness and acceptability.Therefore,in order to quantify the impact of time-of-use price mechanism on peak regulating potential of users,the influencing factor of time-of-use electricity price is introduced to quantify the impact on peak regulating capability,peak regulating probability,and peak regulating rate of residential users.Finally,taking the electricity consumption data of 50,000 residents as an example,this paper studies the impact of time-of-use price mechanism on peak regulating potential,and further evaluates residents'peak regulating potential in this area.At the same time,under the circumstance of different price setting and peak period division,the distributions of residents' peak shaving potential are shown,which proves to be a guiding role in formulating resident time-of-use electricity price. |
PDF Full Text Request