Improved Clonal Selection Algorithm And Its Application In Parameters Tuning Of Controller

Clonal Selection Algorithm (CSA) is a bionic intelligent computing method derivedfrom the clonal selection theory of biology immune system. The algorithm decides thedegree of reprodution and mutation of antibodies in the light of appetency and hasbeen widely used in many practical engineering fields with strong adaptive capacity,learning ability and the ability to maintain the diversity of the population. But it stillhas some shortcomings, such as slow convergence speed, low solving accuracy andeasy premature convergence and so on. So, it is a hot topic to improve the algorithmfor overcoming these defects.This paper proposes an improved clonal selection algorithm called CCCSA on thebasis of analyzing the principle of CSA and its improvement principles. In order toimprove the algorithm accuracy, chaos initialization method is used to generate theinitial population to improve the quality of initial population. Besides, in order toinhibit the antibodies falling into local optimum, an immune mutation operator basedon cloud model is introduced into the evolution process to ameliorate populationdiversity. The test on standard functions shows the effectiveness of improvedalgorithm with high calculating accuracy and fast convergence speed. However, theimproved algorithm is poor in solving high-dimension and complex function,therefore, an improved algorithm integrated with Cultural Algorithm is proposed inthis paper, which is on the basis of chaos cloud clonal selection algorithm (CCCSA)framework. The improved algorithm uses the knowledge of belief space to guide theantibody evolution process which enhanced direction and purpose of searching inorder to improve global searching ability and local refined search. The test on typicalfunctions shows that the improved algorithm can prevent premature convergence andhas strong global search ability.ADRC (Active-Disturbance-Rejection-Controller) has too many parameters toadjust, which limits its wide using in engineering. Therefore, the problem ofparameters optimization for ADRC is a key issue. It is important to adjust the fiveparameters the more prominent effect on the performance of controller. In the paper,CCCSA is applied to the parameters optimization for second-order ADRC, in whichfirst-order plus delay-time system is the control system and the objective evaluatingfunction is established referring to the performance index of ITAE. The simulation results show that adjusted ADRC has excellent control performance and stronganti-interference ability.