A review of parameter estimators and controllers for induction motors based on artificial neural networks

• We examine several artificial neuronal network schemes and their performances to estimate induction motor parameters.
• By means of this methodology, one can know where the induction motor motion control using neuronal networks is going and its trend.
• We present a comparison between several approaches to estimate parameters and control induction motors and report them with some of their results.

Induction motors (IMs) are the most used electromechanic machines in industrial applications. Their control has become the subject of many studies since the 70 s, and there have been several approaches to achieve high-performance adjustable speed drivers (ASDs). The review presented in this article can support the state of some related researches, since it deals with current state-of-the-art of Artificial Neural Networks (ANNs) oriented to experiments that perform motion control with induction motors. It summarizes many previous works focused on IM and can help the reader to have a starting point to begin their own research on choosing a correct type of Neural Network (NN). The paper provides a list of ANNs used to improve the ASD-control, extending the IM-driver life and achieving proper motor operation, their size and performance. A good match between IM parameter values and the data that the controller needs for the induction machine is imperative. Artificial Intelligence (AI) is a helpful tool to achieve this. The summary will also present an overview of different ANN-based drive approaches.