In order to synthesize adaptive control systems over asynchronous electric drive that includes a motor with defects or degradation, we proposed a structure and developed an algorithm for training a PID-neurocontroller based on a multilayer feedforward neural network. Such an approach makes it possible to operatively respond to a change in the characteristics of control object that occurs as a result of the emergence and development of damage and degradation of the motor. This, in turn, makes it possible to improve controllability of the motor, and, consequently, to prolong its operation life cycle and to enhance the energy efficiency of its operation. The proposed solutions, in contrast to the traditional, do not require the use of additional equipment for implementation. It is only needed to change a control program for the frequency converter based on the constructed algorithm. To implement the proposed solutions in practice, we synthesized an algorithm for training a neural network of the PID-neurocontroller with self-tuning. It enables the calculation of weights of the neurons that could be in the future used as the basis of software for a physical control system with the PID-neurocontroller. We mathematically modeled the operation of IM with breaks of the rotor bars and the short-circuited turns in the stator windings when using the proposed controller. An analysis of modeling results showed that the proposed approach to control the electric drive with a damaged IM makes it possible to decrease the amplitude and the number of non-basic harmonics of current and power signals of IM while maintaining the preset parameters of the technological process. Thus, our paper demonstrates the effectiveness of applying the proposed approach to the tasks on maintaining the predefined parameters of the technological process for the case of a stochastic change in the characteristics of control object Author Biographies Dmytro Mamchur, Kremenchuk Mykhailo Ostrohradskyi National University Pershotravneva str., 20, Kremenchuk, Ukraine, 39600 PhD, Associate Professor Department of Automation and computer-integrated technologies
Alan : Fen Bilimleri ve Matematik
Dergi Türü : Uluslararası
Benzer Makaleler | Yazar | # |
---|
Makale | Yazar | # |
---|