Optimization of power system risk assessment method based on support vector machine using cross entropy theory

Abstract

The invention relates to a power system reliability evaluation technology, in particular to a power system risk evaluation method based on cross entropy theory optimization support vector machine, wherein the cross entropy method CEM is applied to a support vector machine SVM to carry out feature selection and SVM parameter optimization; the n-dimensional feature quantity is used as the input of SVM, and the training model is generated by off-line training. The risk level of the power system under the current state is predicted by learning the historical samples. The method includes generatingoff-line data, training the risk sample data by using cross-entropy optimization SVM method, and obtaining the optimized characteristics and parameters of power network risk assessment model; the on-line real-time power system risk assessment is carried out. Using cross-entropy algorithm to optimize support vector machine for risk prediction can effectively remove redundant and irrelevant features, reduce the number of features, combine the optimized parameters can have good convergence, and reduce the computational complexity and time-consuming. It has the characteristics of low error rate and short test time.

Description

technical field [0001] The invention belongs to the technical field of power system reliability assessment, and in particular relates to a power system risk assessment method based on a cross-entropy theory to optimize a support vector machine. Background technique [0002] The rapid development of my country's power system has made significant contributions to the social economy. However, there are still some problems in the power grid at this stage. There are limited means to deal with the sudden change of the power grid status under extreme external environment and other conditions, and the reliability of power supply needs to be further improved. Power system risk assessment can guide the differentiated operation and maintenance of power equipment, so as to solve the source of risk in a targeted manner and achieve the purpose of maintaining the safe and stable operation of the power system. [0003] At present, the application of risk assessment in power system is not p...

AI Technical Summary

Problems solved by technology

The training effect of existing machine learning is not very ideal. There are problems such as the need for a large number of training samples, the training results are easy to overfit, the training time is relatively long, and it is easy to fall into local optimum.

[0005] The above-mentioned machine learning methods all require accurate power grid structure and historical data of component reliability indicators for many years. As the distribution network structure becomes more and more complex, the amount of data continues to increase. Due to the uncertainty of the data, the prediction results obtained by using these traditional methods to evaluate the power supply reliability of the entire distribution network in the target year may deviate greatly from the true value

Images