Industrial Control Engineering

Permanent URI for this collection

http://etd.aau.edu.et/handle/12345678/192

Browse

Browsing Industrial Control Engineering by Author "Amare, Meseret"

Now showing 1 - 1 of 1
  • No Thumbnail Available
    Item
    Linear Quadratic Gaussian Controller Tuned With Particle Swarm Optimization For Speed Tracking Control of Brushless DC Motor
    (Addis Ababa University, 2022-02) Amare, Meseret; Dereje, Sheferaw (PhD)
    Brushless DC(BLDC)motors are widely used in Electric Vehicle applications because of their high starting torque, high efficiency, long operating life and better speed versus torque characteristics. The major problem in the brushless dc motor drive system is that some disturbances originate in the drive which will result in errors and reduce the stability of the system. This problem can be fixed by using good modeling approach and high performance controllers like linear quadratic gaussian. But the selection or tuning of the parameters of the linear quadratic gaussian controller is a tedious process. Therefore it is important to use artificial intelligence based optimization methods to select the parameters of the linear quadratic gaussian controllers to achieve the high performance of linear quadratic gaussian controller. Hence, in this thesis, a linear quadratic gaussian controller tuned with PSO is designed and it’s performance in speed control for a Brushless DC motor is analyzed. The performance of the proposed controller of brushless dc motor was analyzed in terms of speed tracking capability, back emf and hall sensor response, high and low-speed behavior, and speed reversal conditions using MATLAB /SIMULINK. The linear quadratic gaussian controller performance has been compared with proportional integral control strategies in terms of the four quadrants operation and braking system response.The simulation results show that the linear quadratic gaussian-particle swarm optimization controller has a more significant overshoot reduction compared to PI controllers,and a good transient response with a rise time , settling time,and the minimum steady-state speed error percentage of has been achieved for linear quadratic gaussian-particle swarm optimization. Then particle swarm optimization tuned linear quadratic gaussian controller methods are better when compared with PI conventional methods.