Инновационный метод диагностики приводов прокатных станов на основе динамических нейронных сетей

Abstract

We present a methodof diagnostics of technicalstate ofelectricrolling millsin the steelindustry,on the basis ofthe analysis of parametersof current, voltage, speed, and controltasksusinga recurrentartificial neuralnetwork. The methodenhances theaccuracy and reliabilityof diagnosingfault conditionsof the electricequipment is in operationin the earlystages ofthe non-observedandtheir occurrence,which preventssuddenemergency stopmotorandcan significant-lyreduce repair costs.