Electromagnetic equipment multi-parameter non-intrusive identification method, system and equipment based on integration algorithm and storage medium

Abstract

The invention discloses an electromagnetic equipment multi-parameter non-intrusive identification method, system and device based on an integrated algorithm and a storage medium, belongs to the field of parameter identification, and aims to solve the problem that technicians need to understand complex algorithm ideas and adjust parameters and cannot concentrate on multi-parameter non-intrusive identification. According to the method, preprocessing and data division are performed on the electric signal data of the electromagnetic equipment so as to acquire different data sets, then the data sets are selected as required to construct a dynamic adjustment strategy auxiliary function to train the base learner, the optimal base learner is selected, and model evaluation is performed after the dynamic adjustment weight of the optimal base learner is calculated. The auxiliary functions for constructing the dynamic adjustment strategy comprise the MAPE function and the PCC function, and the weighting factor dynamic adjustment strategy based on the two functions can effectively reduce the multi-parameter non-intrusive identification error of the electromagnetic equipment and improve the model prediction performance. The method has important application value and guiding significance in the aspect of multi-parameter non-intrusive identification of the electromagnetic equipment.

Description

technical field [0001] The invention belongs to the field of parameter identification, and relates to an integrated algorithm-based multi-parameter non-invasive identification method, system, equipment and storage medium for electromagnetic equipment. Background technique [0002] In the operation of complex electromagnetic equipment, various parameters such as the operating frequency, voltage, power, and harmonic content of the equipment have large differences. For the learning and identification of multiple parameters of the equipment, equipment monitoring, management, and operation are carried out to realize user The two-way friendly interaction with the power grid is conducive to better planning and dispatching of power resources and intelligent management of complex electromagnetic equipment, ensuring the safe and normal operation of equipment, thereby improving economic benefits. Traditional machine learning algorithm models are simple, fast, and easy to display result...

AI Technical Summary

Problems solved by technology

[0004] In order to overcome the above-mentioned shortcomings of the prior art, the object of the present invention is to provide a multi-parameter non-intrusive identification method, system, equipment and storage medium for electromagnetic equipment based on an integrated algorithm, aiming at solving the problem that engineers and technicians need to understand in the prior art. Complicated algorithm thinking and meaningless parameter adjustment cannot concentrate on the flawed technical problems of multi-parameter non-intrusive identification of complex electromagnetic equipment

Images