Design of artificial intelligence based speed estimator for DC drives / Pauziah Saleh

This thesis describes the design of Artificial Intelligence Based speed estimator for separately excited DC motor using feedforward backpropagation method. The design is created using the MATLAB Toolbox. The training applied to the open loop and closed loop system. A comparison analysis of behavior was performed. The data from the closed loop DC motor with PID controller is used. The variable input data of armature voltage, armature current and output speed were collected by using simulation of the system. The training took only few minutes on a PC for the 30000 input-output training data . For this purpose, the Lavenberg-Marquardt back propagation algorithm was used. A standard three layer feed-forward neural network with tan-sigmoid (tansig) activation functions in the hidden layer and purelin at the output layer is used for this test. The result shows that by using only one hidden layer, minimum error can be obtained as what is needed and also excellent in result. It is satisfied that the application of ANN feed-forward back-propagation method in closed loop system, the speed obtained the excellent result. A comparison between the output of the motor using conventional method that ANN system is able together with PID controller . This was tested by training the system using minimum hidden nodes until reach at the optimum results for the closed loop step and also variable step function. The solutions to the parameter estimated speed for DC motor and without using the tancho generator, the speed of the DC motor can be measured. It is also increasing the realibility for the whole drive.