Extracting Amplitude Characteristics Of Dynamic Power Signals And Their Impact On Power Measurement

Realizing"peak carbon"and"carbon neutrality"are major strategic decisions of the Party Central Committee,involving extensive and profound social,economic,and technological changes.Building a novel power system is one of the"Energy Green and Low-Carbon Transformation Actions"proposed in the State Council’s"Action Plan for Peak Carbon Emissions by2030"（Guo Fa[2021]No.23）.It is also an important task for solidly promoting the"dual carbon"goals.In recent years,with technological changes and advancements,the novel power system has presented"dual high"and"dual stochastic"characteristics:（1）the proportion of wind and photovoltaic renewable energy sources connected to the grid continues to increase,and their power output exhibits strong random fluctuation characteristics;（2）the proportion of high-power nonlinear dynamic loads connected to the grid continues to increase,and the dynamic load current exhibits rapid random fluctuation characteristics.In recent years,domestic and foreign research has found that the random fluctuation characteristics of wind and photovoltaic power generation and high-power dynamic loads can cause serious over-charge in electric energy meter,affecting fair and just electricity trade settlement.For example,Chinese scholars found that the dynamic load electric energy meter of a certain electric arc furnace steelmaking process had an over-charge rate of 16%;Dutch and British scholars found that the dynamic load electric energy meter of a certain water pump had an over-charge rate of-91%;Dutch scholars found that the electricity metering of photovoltaic power generation in a certain farm’s had an over-charge rate of 40%.How to analyze the amplitude characteristics of complex dynamic electric energy signals,extract important（sensitive）features that affect accurate electric energy metering,is a new scientific issue that must be solved in the construction of novel power systems at home and abroad.Solving this problem is of great theoretical significance and application value in ensuring fair and just settlement of new energy power generation and helping the green and low-carbon energy action.In response to the over-charge problem of electric energy metering equipment（smart electricity meter）based on digital signal processing technology caused by rapid and random fluctuation of complex dynamic electric energy signals,this paper actually collected electric energy signals of three typical electricity usage scenarios,studied and analyzed the amplitude characteristics of complex dynamic electric energy signals,explored important feature extraction methods of mapping the amplitude domain to the run-length domain,and analyzed the impact of important feature parameters on the dynamic error of electric energy metering.The main contents of the paper are as follows:Firstly,using the short-time Fourier transform method,the fundamental wave amplitude signal of complex dynamic electric energy signals was extracted,and the amplitude characteristics of complex dynamic electric energy signals in three typical scenarios were analyzed using amplitude domain and run-length domain analysis methods,including electric arc furnace steelmaking,spot welding machine machining,and high-speed rail traction substation.Secondly,the bimodal decomposition and principal component analysis methods were used to extract 16 run-length waveform modes of 8 types of complex dynamic currents;an important feature parameter and algorithm of mapping the current amplitude signal to the run-length waveform mode in the amplitude domain were proposed to characterize the rapid change characteristics of complex dynamic current;based on this method,the important feature parameters of the 16 run-length waveform modes were analyzed.A feature waveform library of complex dynamic current was established to solve the problem of insufficient samples of dynamic current feature waveforms for many years.Then,the wavelet transform run-length waveform mode modeling method was studied to weaken the influence of noise on electric energy metering in the signal.Two evaluation indices of Y andV_rmse for selecting wavelet basis functions and decomposition levels were constructed,solving the problem of selecting wavelet basis functions and decomposition levels for different run-length feature waveform modes.Finally,based on the above research results,the experimental points characterizing important features in the amplitude domain were added to international and domestic standards,and the experimental analysis method was used to verify that the above important feature parameters in the amplitude domain are sensitive feature parameters that affect the over-charge of electric energy meters,which provides a basis for solving the problem of insufficient coverage of scenario experimental points in current international and domestic standards.