Speed self-tuning adaptive control of DC motor

Abstract

This thesis presents a self tuning pole placement adaptive controller which updates its parameters automatically in order to conserve the desired characteristics of the controlled system. PMDC Motor is used as plant (process) response of this adaptive controller. Two methods are used for offline system identification (Pattern Search Genetic Algorithm, Non-Linear Least Squares Estimation), data acquisition (NI-PCI 6221-37pin) is used to acquire data that will be used to identify the parameter of the motor (armature resistance, inductance, and inertia), armature current, input voltage and rotational speed of the PMDC Motor are the data acquired and used in system identification. DAQ tool box in MTLAB/SIMULINK is used to perform the analysis on the data acquired based on Non-Linear Least Squares Estimation and Pattern Search Genetic Algorithm. PID controller is designed based on the model estimated, and is implemented and tested using the motor in order to validate the obtained model. The Recursive Least Squares Estimator block in SIMULINK is used to perform the online estimation. The controller is implemented and simulated in SIMULINK (SIMULINK Desktop Real-Time), the behavior of the PMDC motor is close to the desired response. Output response of the PMDC Motor can be changed due to a perturbation, environment changes or parameters changes. STC is able to track the system (plant or process) for maintaining the desired behavior.