Hybridization of evolutionary algorithm and deep reinforcement learning for multi-objective orienteering optimization

Liu，Wei1; Wang，Rui1; Zhang，Tao1; Li，Kaiwen1; Li，Wenhua1; Ishibuchi，Hisao2

2022

10.1109/TEVC.2022.3199045

IEEE TRANSACTIONS ON EVOLUTIONARY COMPUTATION   Impact Factor & Quartile

1089-778X

1941-0026

Multi-objective orienteering problems (MO-OPs) are classical multi-objective routing problems and have received much attention in recent decades. This study seeks to solve MO-OPs through a problem-decomposition framework, that is, an MO-OP is decomposed into a multi-objective knapsack problem (MOKP) and a traveling salesman problem (TSP). The MOKP and TSP are then solved by a multi-objective evolutionary algorithm (MOEA) and a deep reinforcement learning (DRL) method, respectively. While the MOEA module is for selecting cities, the DRL module is for planning a Hamiltonian path for these cities. An iterative use of these two modules drives the population towards the Pareto front of MO-OPs. The effectiveness of the proposed method is compared against NSGA-II and NSGA-III on various types of MO-OP instances. Experimental results show that our method performs best on almost all the test instances, and has shown strong generalization ability.

Computational modeling decomposition deep reinforcement learning evolutionary algorithms Multi-objective optimization Optimization orienteering problems pointer networks Reinforcement learning Sociology Statistics Training Urban areas

EI

English

Others

20223512672209

Combinatorial optimization ; Deep learning ; Evolutionary algorithms ; Learning algorithms ; Multiobjective optimization ; Personnel training ; Reinforcement learning ; Traveling salesman problem ; Urban planning

Urban Planning and Development:403.1 ; Ergonomics and Human Factors Engineering:461.4 ; Artificial Intelligence:723.4 ; Machine Learning:723.4.2 ; Operations Research:912.3 ; Personnel:912.4 ; Combinatorial Mathematics, Includes Graph Theory, Set Theory:921.4 ; Optimization Techniques:921.5 ; Numerical Methods:921.6

COMPUTER SCIENCE

2-s2.0-85136875413

Scopus

1.College of System Engineering, National University of Defense Technology, Changsha, PR China
2.Guangdong Provincial Key Laboratory of Braininspired Intelligent Computation, and Department of Computer Science and Engineering, Southern University of Science and Technology, Shenzhen, China

Liu，Wei,Wang，Rui,Zhang，Tao,et al. Hybridization of evolutionary algorithm and deep reinforcement learning for multi-objective orienteering optimization[J]. IEEE TRANSACTIONS ON EVOLUTIONARY COMPUTATION,2022,PP(99):1-1.

Liu，Wei,Wang，Rui,Zhang，Tao,Li，Kaiwen,Li，Wenhua,&Ishibuchi，Hisao.(2022).Hybridization of evolutionary algorithm and deep reinforcement learning for multi-objective orienteering optimization.IEEE TRANSACTIONS ON EVOLUTIONARY COMPUTATION,PP(99),1-1.

Liu，Wei,et al."Hybridization of evolutionary algorithm and deep reinforcement learning for multi-objective orienteering optimization".IEEE TRANSACTIONS ON EVOLUTIONARY COMPUTATION PP.99(2022):1-1.