Trajectory control of DC servo using OS-ELM based controller

In this work implementation of On Line Sequential Extreme Learning Machine (OS-ELM) based trajectory control strategy of DC servo motor is investigated. Extreme Learning Machine (ELM) is a novel method for training of the forward neural networks with single hidden layer is extremely fast and accurate compared to gradient based learning methods. Further to minimize the effect of machine parameters variations, ELM is trained on line using sequential recursive least square method. Comparative study of artificial neural network based controller, ELM based controller, and OS-ELM based controller has been done for DC servo motor for non linear fan or impeller type of load.