Research On The Analysis Method Of Time-frequency Domain Global Characteristics Of Complex Dynamic Energy Signals And Its Influence On Energy Measurement

In recent years,in order to thoroughly implement the CPC Central Committee’s major strategic decision of"carbon peaking"and"carbon neutrality",China has steadily promoted clean energy substitution in the steel industry and firmly promoted electric energy substitution in the transportation field.With the implementation of the dual carbon strategy,the proportion of high-power dynamic loads such as intermediate frequency furnaces and high-speed railway electric locomotive in China’s new power system is increasing rapidly,and the dynamic load power presents complex characteristics such as large fluctuations,strong randomness,fast time-varying and multi harmonics.In recent years,research at home and abroad has shown that these complex changing characteristics can lead to inaccurate electricity metering in electricity trading,which in turn has adverse effects on promoting electricity substitution.However,due to the short implementation time of the dual carbon strategy,relevant research at home and abroad is still in the initial stage.It is not clear which complex change characteristics the new power system still has,and which sensitive characteristics are causing electricity metering to exceed the tolerance are still unclear.Especially because the impact of complex change characteristics on electricity metering is a long-term dynamic process,the frequency domain characteristics are the key characteristics that reflect the impact of complex change characteristics on electricity metering.Currently,research results at home and abroad have not provided relevant answers.In this regard,this article focuses on the analysis of the time-frequency domain characteristics of complex dynamic electrical energy signals in new power systems,and which time-frequency domain features can lead to inaccurate electrical energy measurement.It conducts in-depth research on the time-frequency domain characteristics analysis and feature extraction methods of complex dynamic electrical energy signals.The specific work is described below.Firstly,a ZS-STFT（Zero Search-Short Time Fourier Transform）global feature analysis algorithm in time-frequency domain is proposed.The algorithm utilizes the zero crossing feature of the signal waveform,adaptively adjusts the STFT time domain window width,and solves the impact of phase accumulation effects caused by grid frequency fluctuations on harmonic analysis.The amplitude analysis error of the algorithm is reduced to 10^-3orders of magnitude,and the phase analysis error is less than 1°.Secondly,a time-frequency domain data representation model for complex dynamic power signals and its time-frequency domain globalization are established.The feature matrix extraction algorithms are:harmonic amplitude and phase probability density distribution feature matrix,harmonic amplitude and phase joint distribution feature matrix,fundamental and major harmonic amplitude impact intensity distribution feature matrix,which solves the problem of characterizing and extracting time-frequency domain characteristics.For two typical nonlinear dynamic loads:intermediate frequency furnaces and high-speed railway electric locomotives,the analysis results of their various time-frequency domain global characteristics are presented.Then,the device hardware architecture of ARM+PIC microcontroller+FPGA is constructed,and a synchronous testing method of dynamic error of electric energy meters is designed and developed.An experimental device for dynamic error testing of electric energy meters is developed.Completed the communication protocol design,PIC microcontroller control function development,and device functions testing.Finally,a distorted waveform combined power reference test signal model that can reflect the time-frequency domain global characteristics of two types of typical loads is established,and a dual current loop power meter dynamic error synchronization test experimental system is constructed.The experimental results verify that the extracted harmonic time-frequency domain global characteristics are sensitive to out-of-tolerance metering of electric energy meters.