Traditional linear control strategies are usually straightforward and easy to implement, but they tend to be helpless when dealing with complex nonlinear problems, which drastically limit their scope of application. In order to find a way that can overcome their drawbacks while not giving up their main feature of simplicity, this paper introduces evolutionary programming as an optimization method, which tackles the difficulty by being integrated with conventional PID controllers to form self-adaptive control systems. This approach has both the advantages of flexibility and simplicity, without sacrificing accuracy. A CSTR (continuous stirred tank reactor) system is presented as a case in point to show its working principle and simulation results are also given to prove its efficiency. The last part of the paper gives some recommendations for future work and improvements.
[1]
Sam Kwong,et al.
Genetic algorithms: concepts and applications [in engineering design]
,
1996,
IEEE Trans. Ind. Electron..
[2]
D. B. Fogel,et al.
Applying evolutionary programming to selected control problems
,
1994
.
[3]
M. Silaghi,et al.
A new optimization algorithm
,
2000
.
[4]
Thomas Bäck,et al.
Evolutionary computation: comments on the history and current state
,
1997,
IEEE Trans. Evol. Comput..
[5]
Sam Kwong,et al.
Genetic algorithms: concepts and applications [in engineering design]
,
1996,
IEEE Trans. Ind. Electron..
[6]
Lawrence J. Fogel,et al.
Artificial Intelligence through Simulated Evolution
,
1966
.
[7]
Kenneth de Jong,et al.
Adaptive System Design: A Genetic Approach
,
1980,
IEEE Trans. Syst. Man Cybern..