Supersonic blowdown wind tunnel control using ABC optimized PID controller

Supersonic blowdown wind tunnels enable the testing of aircraft prototypes in the Mach 1.2 to 5 range, but these tunnels allow very limited test times due to their structure. In addition, once the wind tunnel system starts operating, the pressure in the tank where the air is stored changes constantly. This means that the parameters of the already nonlinear system dynamics are continually changing. To utilize this time effectively, fast-responding, stable and highly efficient controllers are needed. These controllers should be able to provide the required pressure in the settling chamber as fast as possible for the desired flow conditions in the test section. In these types of testbeds, Proportional-Integral-Derivative (PID) controllers are widely used because of their reliability, simplicity and ease of implementation. PID controllers can also provide fast and stable responses, as they can reduce the error, eliminate the steady-state error, and minimize the overshoot and oscillations. PID controllers only require the measurement of the error and the tuning of the coefficients, which can be performed manually or automatically. For a PID controller it is essential to optimize its coefficients to achieve the best performance and stability. There are different methods to tune a PID controller, such as trial and error, Ziegler-Nichols method, Cohen-Coon method, and optimization algorithms. This study proposes the use of an artificial bee colony in the optimization of PID coefficients used in the control of a supersonic blowdown wind tunnel. Because of complexity, an artificial bee colony is used to optimize PID coefficients with three different objective functions. The optimized coefficients are compared to gradient optimization results, and the best approach is determined.