基于支持向量机的热轧主电机风冷变频策略

Variable Frequency Strategy of Air-Cooled Motor for Hot Rolling Main Motor Based on Support Vector Machine

对热轧主电机温升及其风冷变化机制进行分析,基于大量的热轧生产线工况数据建立了热轧主电机定子绕组的暂态温升模型,采用基于支持向量机(SVM)的回归算法对模型中系统参数进行辨识,并与传统最小二乘法的辨识结果进行对比.结果表明:所建立的暂态温升模型更加灵活且具有更好的预测效果;所提出的风冷电机变频策略在确保生产线安全运行的同时能够有效降低风冷电机的耗能,4台风冷电机的平均节能率超过50%.

Based on the analysis of the mechanism of the temperature rise of the main motor, the air volume of the cooling fan and the data of hot rolling production lines, the transient temperature rise model of the stator winding was established. The regression model based on support vector machine （SVM） was used to identify the system parameters in the model, and compare with the traditional the least square method. The statistics showed that the SVM method was more flexible and had better prediction effect. The fre- quency conversion strategy of the air cooled motor based on this modeling method could effectively reduce the energy consumption of the air-cooled motor, and simultaneously ensure the safe operation of the pro- duction line. The energy saving average rate of the four cooling motors was more than 50%.