Design of low order quantitative feedback theory and H-infinity-based controllers using particle swarm optimisation for a pneumatic actuator system

The quantitative feedback theory (QFT) and are the two most popular, well defined and powerful robust control techniques. These techniques are used to achieve high performance of the system in the presence of system uncertainties and disturbances. This work is concerned with the design of simplified structure and low order robust control algorithms based on QFT and/or techniques for pneumatic servo actuator system. The particle swarm optimization (PSO) method is used to tune the parameters of the controller and weighting functions subject to QFT and/or constraints. These algorithms are designed to achieve the robustness over a wide range of system parameters change and disturbances. The pneumatic system parameters uncertainty is the main problem in the design of a desired control algorithm for this plant. Where the variation in thermodynamic conditions causes an uncertainty in a number of the model’s parameters and the large change of pneumatic actuator load leads to variation in the actuator dynamics. Therefore, two ranges of pneumatic actuator load variation are considered. The first one is a small range variation when the load mass, M=0.1 to 5 kg which is widely required in many industrial applications of pneumatic actuators for control of machines such as robots. The second one is a wide range variation when the load mass M=0.1 to 100 kg which is required for the pneumatic servo actuators when they are used in missiles applications. PSO based QFT control algorithm is proposed for pneumatic servo actuator system. The PSO algorithm is used to optimize the loop-shaping step (subject to QFT constraints), which is performed manually in the standard QFT control design. The obtained controller is simpler structure and lower order than the standard QFT controller and the same robustness of the standard QFT control is achieved. On the other hand, the design of conventional control with structured and unstructured uncertainties is also presented. The PSO algorithm is used to minimize the infinity norm of the transfer function of the nominal closed loop system to obtain the optimal parameters of the weighting functions. This method simplifies the design procedure and leads to obtain a sub-optimal controller can achieve the position control of pneumatic actuator. A method for robust controllers design using restricted structure controllers such as PID, Lag-Lead and deadbeat control algorithms is presented. The method uses particle swarm optimization (PSO) to tune the controller and performance weighting function parameters by minimizing a cost function subject to -norm specifications. It is shown that the designed controllers present robustness over a wide range of parameters change. Also, it can be shown that the proposed deadbeat control algorithm can achieve more desirable time and frequency responses specifications with simpler structure and lower order controller. Since there is a lot of parallelism between QFT and control techniques and they may complement one other, a robust hybrid /QFT controller is designed to assure robust stability and robust performance of the uncertain pneumatic servo actuator system. This controller achieves in the same time the design requirements that arise from both QFT and control techniques. The PSO algorithm is used to obtain the controller parameters by minimizing a new proposed cost function subject to QFT and constraints. The PSO based hybrid /QFT controller can give automatically better performance (in terms of rise time and settling time) than the previous works that used only one of them. Finally, a comparison between all the designed controllers in this work shows that the superiority of the PSO based deadbeat controller. It can achieve the same robustness of conventional methods with simple structure and low order controller and with more desirable time and frequency response specifications.

Preview PDF FK 2010 54R.pdf Download (926kB) | Preview

Actions (login required)

View Item