Design of low cost universal artificial neuron controller for chopper fed embedded DC drives

Artificial neural network (ANN) has become very popular in many control applications due to their high computation rate and ability to handle nonlinear functions. This paper proposes an artificial neuron controller for closed loop speed control of DC drive fed by DC chopper. Neuron control is used to reduce the steady state error, overshoot and settling time. The signal corresponding to the motor speed error and change in speed error are used as inputs to ANN Controller. The controller outputs the required change in duty cycle of pulse width modulated gating signal applied to DC chopper. Thus the voltage fed to the armature of the DC motor is adjusted for achieving the desired speed response. The training patterns for the neuron controller are obtained from the conventional PI controller and the effectiveness of the proposed neuron controller is studied using simulation studies.The designed controller was implemented in a low cost 8051-based embedded system and the results are documented. Two-loop control system was implemented with an inner ON/OFF current controller and an outer ANN speed controller.A conventional controller has heavy computation burden whereas a trained neural network requires less computation time. The artificial neural network has the ability to generalize and can interpolate in between the training data. This advantage of ANN makes the ANN controller universal. The ANN controller designed was tested on two different motors and found to work effectively on driving both of them.     

A Numerical Computation of Heat and Fluid Flow in L-Shaped Curved Channels

A numerical analysis of steady, hydrodynamically and thermally fully developed, incompressible laminar flow with constant physical properties is presented. Two different geometrical cases of L-shaped channels are considered: right-curved and left-curved. The solution of the discretized continuity, momentum, and energy equations was obtained by using an elliptic Fortran program based on the SIMPLE algorithm. Solutions are obtained for Dean number ranges from 4.1 to 210.6 with dimensionless radius of curvature of 100, and Prandtl number of 0.7. The secondary flow streamlines, the isotherms, average Nusselt numbers, and friction factors are presented depending on Dean number and L-shaped channel orientation. As a result, it is observed that the secondary flows resulting from centrifugal forces change the distribution of the velocity and the temperature fields. In addition, it is determined that channel orientation has a profound effect on the flow and temperature fields.