Research On Load Characteristic Identification Technology Of Resident Users' Switching Power Supply

The intelligence of residential electricity usage is an important field of smart grid research.Load of residential electricity usage has the characteristics of a large scale,a great variety and different features of electricity usage.As the key technology of realizing the intelligence of electricity usage in an all-round way,load monitoring could provide detailed information of household energy efficiency,as well as analyze the structure of household power consumption,therefore making users acknowledged of the impacts on household energy efficiency of their own actions,and guiding them to adopt energy saving measures consciously.However,it is difficult to popularize the method of embedding monitoring model into every single load.Therefore,non-intrusive load monitoring method is adopted,and only at the entrance of power will the electricity usage data be collected.The identification of load will be implemented by signal analyzing and processing,and finally we will get the situation of all the loads in the whole electricity usage network.The identification technology of residential electricity usage is of great researching importance.The paper mainly discusses the load characteristic identification technology of resident users' switching power supply.Regarding the physical topology of loads as the starting point,the paper divides common residential loads into two categories,general load and switch power supply load.Since the owning rate of switch power supply load is considerably high among residential users,the paper focuses on switch power supply load and does research on the extraction of load characteristics and identification of load,as well as finishes the design of hardware system for switch power supply load characteristic identification.The specific work is as follows:1)First of all,the paper analyzes the researching situation of domestic and foreign load monitoring technology as well as load identification algorithm,studies the fundamental structure of switch power supply load,designs the experimental program for electricity usage signal acquisition of switch power supply load,collects the data sample of load operation in both single and mixed way,and further analyzes the electricity using characteristic.2)An extraction method for switch power supply load characteristic based on second-order statistics is proposed.The paper uses the envelope characteristic of a single load current wave form to capture the load transient data window.On this basis,the autocorrelation and covariance characteristics of switch power supply load are obtained.Then we focus on short-time Fourier transform theory and its application in the extraction of switch power supply load characteristic,and obtain time-frequency characteristics of the parameters of switch power supply load.3)We need obtain the three-dimensional characteristic parameter space of switch power supply load and design an optimized scheme of K-means load clustering center and inertial weight value combined with particle swarm algorithm,therefore realizing the optimal clustering of switch power supply load for residential users.The load clustering results are compared with the prior values of the characteristic parameters to obtain the identification results of switch power supply load characteristics.4)Ultimately,we use DSP to construct a hardware system for switch power supply load characteristic identification and design the collection access module and the conditioning module circuit to realize the hardware encapsulation of the algorithm for switch power supply load characteristic identification.The paper proposes a identification technology for characteristic of residential users' switch power supply load,which could extract switch power supply load effectively and judge the switch power supply load exactly.The algorithm is of great efficiency and is convenient for hardware to actualize.