Non-intrusive Monitoring Online System Of Household Electric Loads

Load monitoring,as a necessary way to obtain detailed information of the user's User power consumption,is the basis of intelligent power consumption and energy saving and emission reduction.It can provide power companies with electrical data for power system development planning and load forecasting,and help power companies formulate reasonable incentives for demand response and detect the electricity stealing behavior.Besides,users can reasonably adjust and optimize their own electrical behaviors by understanding their own electricity consumption in order to achieve the purpose of saving electricity bills.The non-intrusive load monitoring method can obtain the power information of each load within the detected user by installation of information acquisition devices at the power entrance of the monitored user,extraction of electrical characteristics of electric load and then processing and analysis of electrical data using appropriate algorithms.This method effectively,which solves the problems of traditional intrusive monitoring technology,such as difficult installation,maintenance cost and complex management,is the main trend of future load monitoring.According to the characteristics of intrusive and non-intrusive load monitoring technology,the advantages and disadvantages of the two technologies are analyzed from six aspects,including hardware level,user acceptance and so on.The algorithm of noninvasive load identification based on active and reactive power,current waveform and harmonic,voltage current trajectory and transient power four different load characteristics is analyzed.The principles,existing problems and improved methods of the non-intrusive load identification algorithm based on the characteristics of different loads are studied.According to the actual operating conditions of the household electric equipment,the model and parameters of the voltage and current sensor are reasonably selected.A signal conditioning circuit consisting of three level circuits with amplification,filtering and migration is built to preprocess the collected voltage and current signals so that the signals meet the working requirements of USB 6009 acquisition card.The working interface of the monitoring system is designed to realize the human-computer friendly interactive function.By completing the configuration of the USB 6009 working mode,sampling channel and sampling rate,the continuous digital sampling of voltage and current waveforms is realized.The extraction algorithm of load feature is designed by using LabVIEW graphical language,and the function of recording the signal in the form of file is realized.A waveform display module is designed to monitor the voltage and current signals on the front panel in real time.In this paper,a noninvasive load identification algorithm is designed.The waveform of load current is selected as the load characteristic.Dynamic time warping technology is applied to match the extracted steady-state current signal with the template library built by feature-additive criterion.In detail,The first experiment shows that the most characteristics of current waveform do not satisfy the feature-additive criterion.Therefore,it is impossible to realize the decomposition and recognition of the household electric loads from the entrance of the electric power.By analyzing the steady-state characteristics of current waveform,the extensed criterion based on feature-additive criterion and the condition of extracting steady-state current waveform are presented.Finally,the dynamic time warping technique is used to match the extracted steady-state current waveform with the waveforms in the template library.