Noise reduction method and system for time sequence power data

Abstract

The invention provides a noise reduction method and system for time sequence power data. The noise reduction method and system are used for realizing noise reduction on the time sequence power data acquired by power equipment. The method comprises the following steps: preprocessing a training set of original time sequence power data, wherein the preprocessing comprises random noise processing and normalization processing; constructing a deep convolutional auto-encoder model based on a deep convolutional neural network; inputting the pre-processed training set into the deep convolutional auto-encoder model for training, and performing parameter adjustment on the deep convolutional auto-encoder model in combination with the pre-processed training set to obtain an optimal deep learning model; testing the optimal deep learning model by using a test set of time sequence power data, and evaluating the performance of the model; and processing the time sequence power data by using the optimal deep learning model to obtain noise-reduced time sequence power data.

Description

technical field [0001] The present application relates to the technical field of power equipment, and in particular to a noise reduction method and system for time-series power data. Background technique [0002] Time-series power data refers to the business data of power equipment that has a certain time series and is collected, processed and analyzed in the links of generation, transmission, transformation, distribution, and use, such as load curves (or called load data, load curve data) , current curve (or called current data, current curve data), voltage curve (or called voltage data, voltage curve data), etc. Time-series power data can be used to provide support for user demand-side response and electricity price formulation. [0003] In the process of obtaining time-series power data of power equipment, it will be interfered by noise, especially in the environment of wireless transmission, and with the continuous advancement of smart grid, a large number of smart mete...

AI Technical Summary

Problems solved by technology

[0003] In the process of obtaining time-series power data of power equipment, it will be interfered by noise, especially in the environment of wireless transmission, and with the continuous advancement of smart grid, a large number of smart meters and advanced data acquisition equipment are installed and used. The equipment is susceptible to high-order harmonic interference, which makes the time-series power data acquired by the power equipment subject to serious noise interference

Images