Fault Diagnosis and Fault Tolerant Control of High Power Variable Frequency Speed Control System Based on Data Driven

Abstract

Compared with AC-DC-AC converter, the character of Cycloconverter is no intermediate DC link, small size, light weight and high efficiency of primary power conversion. It is widely used in the main drive speed control system of low-speed, large-capacity hot-rolling and cold-rolling mills. Its safety are very important. Once the Cycloconverter fails, it will result in huge economic losses. At present, there are many researches on fault diagnosis of AC-DC-AC inverter, but the fault diagnosis of Cycloconverter is still in the low stage. Therefore, it is necessary and practical to establish a fault diagnosis system and fault-tolerant control method for Cycloconverter. In this paper, a fault diagnosis method based on BP neural network and a fault-tolerant control method of Cycloconverter are designed. The effectiveness of this method is proved by simulation. This provides new methods and ideas for safety issues in industrial production processes.