A learnable ant colony optimization to the mission planning of multiple satellites

Mission planning plays a very important role in the management process of imaging satellites, and it is the hardcore of imaging satellites application systems.A learnable ant colony optimization (LACO) is proposed to the mission planning of multiple satellites.Before each iterative loop,the LACO randomly selects an appropriate parameter combination via dynamic parameter decision model according to the performance knowledge of parameters.Different than standard ant colony optimization,the LACO extracts some available component knowledge,and applies the obtained component knowledge to guide artificial ants to construct feasible solutions in the subsequent optimization process.Under the effective cooperation of ant colony optimization,dynamic parameter decision model and component knowledge,the performance of LACO was largely improved.Twenty-one testing instances were applied to compare the performance of different approaches.Experimental results suggest that the LACO outperforms other two approaches.