Deep learning non-intrusive load monitoring method based on unsupervised optimization

Abstract

The invention discloses a deep learning non-intrusive load monitoring method based on unsupervised optimization. The first part is to establish a supervised neural network deep learning model; the second part is optimization of the model by using an unsupervised learning mode, and the first part comprises the following steps: monitoring all load information in a period of time from a target load cluster; preprocessing the data by using an algorithm, and normalizing the data; performing neural network training on the preprocessed data; and evaluating a network training result. The second part is optimization of the model by unsupervised learning, iteration is carried out on each target load clustering center by utilizing a K-means clustering algorithm, a training data training model is reconstructed, a supervised learning algorithm is optimized by utilizing an unsupervised algorithm, and then analysis is carried out on power consumption behaviors. According to the non-intrusive load monitoring method provided by the invention, the self-learning capability, universality, sensitivity and accuracy of processing a non-intrusive load monitoring problem by using a deep learning algorithm are greatly improved.

Description

technical field [0001] The invention relates to a non-intrusive load monitoring method, in particular to a non-intrusive non-intrusive load monitoring method based on non-supervised optimization based on deep learning. Background technique [0002] In the context of the extensive construction of the Internet of Things, the realization of power grid transparency, and the realization of the future development of smart grids where information networks and power grids coexist, knowing the operating characteristics of each load to provide users with load operating status changes and optimizing power consumption plans will be the future power grid. One of the important service models of operators, so load monitoring has become an important part of smart electricity consumption. At present, load monitoring is divided into intrusive load monitoring and non-invasive load monitoring. Intrusive load monitoring has the advantages of small error and fast response, but its high cost is no...

AI Technical Summary

Problems solved by technology

At present, load monitoring is divided into intrusive load monitoring and non-invasive load monitoring. Intrusive load monitoring has the advantages of small error and fast response, but its high cost is not suitable for large-scale promotion, so non-invasive load monitoring has advantages. Very obvious cost advantage

[0003] Non-intrusive load monitoring refers to the installation of monitoring equipment at the power entrance, through the collection and analysis of load cluster data there to obtain the type and operation status of each individual load, and the analysis of high-frequency characteristics also requires high requirements for monitoring equipment. Compared with low-frequency analysis, it shows a cost disadvantage, but the accuracy of low-frequency NILM is complex and inaccurate due to the lack of information.

The supervised learning itself has the disadvantages of poor self-learning ability, and the unsupervised learning itself has the disadvantages of poor accuracy of analysis results and inability to adapt to non-invasive load monitoring.

Images