基于神经网络的电机智能保护器研究

介绍了一种电机智能保护装置系统，分析了对电动机进行各类保护的原理，并且针对过载保护．提出一种基于人工神经网络的电动机长期过载运行温升预测的算法，将这种算法应用于实际的保护中．有效地弥补了不能用数学模型精确描述电机运行中发热和散热过程的缺陷。实验证明，在样本足够多的情况下，这种方法是切实可行的。

Study of Electromotor Protector Based on Neural Network

The system of electromotor protector is presented by analyzing a series of protecting principles of the electromotor. To solve the problem of overload,an algorithm of predicting the temperature-rise of the eleetromotor based on neural network is brought forward. The use of the algorithm will effectively make up for the defect that can not describe exactly the heat given out and absorbed during electromotor operation. Experiments have proved that the method is effective with enough samples.