Research On Improvement Based On MOEA/D

Multi-objective optimization evolutionary algorithms based on Decomposition(MOEA/D)is an important branch of multi-objective optimization evolutionary algorithms(MOEAs).Though MOEA/D is a promising algorithm which has solved many multi-objective optimization problems(MOPs),it has many improvable parts which we can research on it.This thesis aims at exploting some improving algorithms researches on solving many-objective optimization problems(MaOPs)which are based on MOEA/D.The main work of this thesis can be concluded as following:1.To improve the capability of MOEA/D in solving MaOPs,a version of improving MOEA/D(I-MOEA/D)is proposed.We redesigned the weight vectors used in the sub-problems,making the distribution of the weight vectors broader and more effective to ensure the diversity and convergence of solutions in the objective space.Moreover,a new decomposition approach,called the weighted mixture-style method,which combines the advantages of the weighted sum decomposition and the Tchebycheff decomposition approach,is adopted in I-MOEA/D to improve the effectiveness of algorithm.Experimental results verify the effectiveness of I-MOEA/D in improving MOEA/D on solving MaOPs.2.To improve the capability of MOEA/D in solving MOPs and MaOPs,another version of improving MOEA/D(mMOEA/D)is proposed.In mMOEA/D,a novel elastic weight vectors design method is introduced and adopted to make those weight vectors spread more widely.On the other hand,a flexible and efficient trail DE operator is designed and used in mMOEA/D for further enhancing the performance of MOEA/D.In the part of experiment,there are many experimental results can certificate the effectiveness of mMOEA/D in improving MOEA/D on solving MOPs and MaOPs.3.To improve the capability of MOEA/D in solving Ma OPs,a version of improving MOEA/D which based on the optimal DE schemes(MOEA/D-oDE)is proposed.For one thing,MOEA/D-oDE adopts a newly-introduced decomposition approach to decompose the many-objective optimization problems,which combines the advantages of the weighted sum approach and the Tchebycheff approach.For another thing,a kind of combination mechanism for DE operators is designed for finding the best child solution so as to do the posteriori computing.We also have done many experiments to prove the effectiveness of MOEA/D-oDE in improving MOEA/D on solving the MaOPs.