Research On Integrated Monitoring System Of User Side Power Information Based On OneNET Platform

With the strategic goal of "carbon neutrality,carbon peak",China’s traditional energy system has begun to transform to a low-carbon,safe and efficient energy system,among which photovoltaic power generation has taken a leading position in the renewable energy generation.The connection of photovoltaic power generation to the grid will affect the power quality of the grid,and the photovoltaic array itself is vulnerable to malfunction,which will not only affect the power quality and safety of power users,but also reduced the efficiency and life of the photovoltaic system.The rapid development of Internet of Things technology made it possible to comprehensively monitor power quality and photovoltaic operation status.This thesis has designed a set of user-side electricity information comprehensive monitoring system to realize comprehensive monitoring of power quality and photovoltaic array operation status,reduced maintenance costs,and realized more efficient and safer energy utilization.Specific work contents are as follows:(1)The two-window full-phase FFT bispectral algorithm is proposed to detect the noisy harmonic signal,and the method of empirical mode decomposition and wavelet threshold combined denoising is adopted to denoise the signal.The Blackman window function is used to suppress the spectrum leakage phenomenon,and the simulation data is used to verify that the two-window full-phase FFT bispectral algorithm has a better harmonic detection effect.By combining the noise reduction method with the harmonic detection method,MATLAB simulation proves that the method can accurately detect each harmonic content in the noisy signal.(2)A photovoltaic fault diagnosis method based on SAW-Transformer neural network is proposed.Based on the typical Seq2 seq model,Transformer is improved and applied to photovoltaic fault diagnosis and classification.A PV array model was built in Simulink to simulate different types of faults,obtain the output characteristic curve of PV array,and extract the fault characteristics of PV array.Adam W algorithm is used to optimize the neural network parameters and sparse the attention weights,so a SAW-Transformer based photovoltaic fault diagnosis model is established,which improves the accuracy of photovoltaic array fault diagnosis.(3)Designed a user-side electricity consumption information comprehensive monitoring system based on the One NET platform.The device connected to the Internet of Things has the network communication function,completed the hardware and software design and prototype development of the intelligent power terminal,and conducted basic functional tests on the system,verifying that the user-side electricity consumption information comprehensive monitoring system has good real-time performance and stability,and the measurement error meets the design requirements.