Constrained Multi-objective Differential Evolutionary Algorithm with Adaptive Constraint Handling Technique

Caixia Cui ； Qinqin Fan

《World Scientific Research Journal》 7卷3期 (2021/03) Pp. 322-339

https://doi.org/10.6911/WSRJ.202103_7(3).0042

摘要

Finding feasible solutions and a good approximate Pareto front (PF) are two important tasks in the constrained multi-objective optimization (CMO). Various constraint handling techniques (CHTs) have a significant impact on these two tasks. To realize the adaptive adjustment of CHTs, a constrained multi-objective differential evolution algorithm with adaptive constraint handling technique (ACHT-CMODE) is proposed in the current study. In the ACHT-CMODE, three state-of-the-art constraint handling methods are integrated and an improved reverse generation distance is used to evaluate their performances. Also, the Q-learning method is utilized to guide the evolution of CHTs. The performance of the ACHT-CMODE is compared with that of the other five constraint multi-objective evolutionary algorithms on 18 test functions. Experimental results show that the overall performance of the ACHT-CMODE is the best among all compared algorithms, and the proposed algorithm is capable of selecting a suitable constraint handling method to solve a particular type of constrained multi-objective optimization problems (CMOPs).

關鍵字

Constrained multi-objective optimization ； Constraint handling technique ； Adaptive ； Q-learning

參考文獻

R. Char, A. Savvaris, A. Tsourdos, et al. Solving multiobjective constrained trajectory optimization problem by an extended evolutionary algorithm [J]. IEEE Transactions on Cybernetics, Vol.50 (2020) No.4, p.1630-1643.

A. Maminov, M. Posypkin. Constrained multi-objective robot's design optimization. 2020 IEEE Conference of Russian Young Researchers in Electrical and Electronic Engineering. (St. Petersburg and Moscow, Russia, 2020), p.1992-1995.

N.G. Paterakis, M. Gibescu, A.G. Bakirtzis, et al. A multi-objective optimization approach to risk-constrained energy and reserve procurement using demand response [J]. IEEE Transactions on Power Systems, Vol.33 (2018) No.4, p.940-3954.

Z.Z. Liu, Y. Wang. Handling constrained multiobjective optimization problems with constraints in both the decision and objective spaces [J]. IEEE Transactions on Evolutionary Computation, Vol.23 (2019) No.5, p.870-884.

M.A. Jan, Q.F. Zhang. MOEA/D for constrained multiobjective optimization: Some preliminary experimental results, 2010 UK Workshop on Computational Intelligence (UKCI), (2010), p.1-6.