Improving the efficiency of reactive power compensation devices through neural analysis

Abstract

This article reflects the effectiveness of the use of neural analysis in control systems of powerful devices. The features of neural network modeling in the Matlab Simulink Neural Network Toolbox software package are presented. The state of the neural network in the field of error construction in event of a change in the operating mode of an electrical installation is clearly reflected. The application of the “Adaptive Linear Prediction” function in control systems of reactive power compensation devices is proposed. The positive aspects of the application are indicated, which in combination give a positive economic effect in the field of energy efficiency industrial enterprise.