Development of Resolver Circuit with Long Short Term Memory and Reinforcement Learning Algorithms

In our age, the usage areas of artificial intelligence have increased considerably. These areas were particularly concerned with the correct predictability of future data using available data. It has become necessary to work on various machine learning algorithms to be used in the calculations of the resolver circuit, which is a feedback element used for tracking the position and position information of the electric motor unit used in various vehicles. The use of machine learning algorithms in the design and implementation of the resolver circuit, which is one of the most important elements of electric motor designs, will shed light on future studies. In this study, it is focused on the use of machine learning algorithms in the calculation of the resolver circuit, position and position information and the performance differences between each other. In this study, LSTM (Long Short Term Memory) and Reinforcement Learning (RL) algorithms were compared. While comparing these algorithms, the types of LSTM and RL algorithms were also studied and compared. As a result of the results obtained, it was aimed that the motor designs would be less costly, and the results obtained in terms of more reliable motor position and position information to be used were promising. In addition, with this study, a basis was created for working on machine learning algorithms in the calculation of different parameters. With this study, a great way has been achieved in integrating algorithms used in electric vehicles, which are quite obsolete today, into AI-based algorithms.