استفاده از شبكه عصبي براي تعيين محدوده مجاز كاري موتوري القايي با خطاي دور به دور

Using Neural Network To Determine The Permissible Domain For Operating Points Of An Induction Motor That Has Turn Faults

Practicability of driving an induction motor that has turn faults, and permissible operating points, at which no fault enlargement occurs, is shown in literature. In this paper, the permissible domain of operating points (PDOP) for a faulty motor are classified to steady and transient PDOP, which respectively means working points for arbitrary and short periods of times. As expected, for a certain size of fault, higher speed and torque as transient are tolerable, and this idea can be used for a more suitable control of a faulty motor. Dependency of PDOP on ambient temperature and capability of neural networks in shortening the time consuming process of determining steady PDOPs are shown. A thermal model for faulty motor is proposed which assists in following the temperature of fault location, and can be used for obtaining transient PDOPs and steady PDOPs at different ambient temperatures. The parameters of the model are obtained, with the aid of a GA, neural network and the results of some experiments. The validity and accuracy of results are shownʹ With the aid of many experimental tests.