Multi-working-condition power system performance prediction method and system based on Gaussian process regression

Abstract

The invention relates to a multi-working-condition power system performance prediction method based on Gaussian process regression, and the method comprises the following steps: 1), collecting conventional working condition data, preprocessing the data to acquire a training data set, and selecting a Gaussian regression process kernel function; 2) training a Gaussian process regression model, namely determining hyper-parameters of the Gaussian process regression model; 3) obtaining a next sampling point by adopting a sequential sampling algorithm, and adding the next sampling point into the conventional working condition data to obtain an augmented working condition data set; and 4) based on the prediction result and the augmented working condition data set, judging whether a stop condition is satisfied, if not, taking the augmented working condition data set as a new training data set and then returning to the step 2), and if so, outputting the prediction result. Compared with the prior art, the method has the advantages of low cost, multi-working condition and extreme working condition prediction and the like.

Description

technical field [0001] The invention relates to the field of power system performance evaluation, in particular to a multi-working-condition power system performance prediction method based on Gaussian process regression. Background technique [0002] In the field of industrial technology, the prediction of the performance of the power system is particularly important for evaluating the pros and cons of the power system. There are two traditional methods to evaluate the pros and cons of the power system, one is to conduct physical experiments to detect the performance of the power system in the experimental environment; the other is to establish a physical model of the power system and a computer simulation model, and then conduct a power system Simulation experiments to check its performance. The former needs to build a relatively complete experimental platform, which consumes a lot of manpower, material resources and time resources, and the working conditions that can be ...

AI Technical Summary

Problems solved by technology

However, separate machine learning and deep learning algorithm models often have poor generalization ability, that is, when only training data under one or more working conditions are obtained, the model cannot be trained to predict the power system under another new working condition. performance

Images