Multi-objective mission planning problem of agile Earth observing satellites

In this paper, we concern multi-objective mission planning problem of a constellation of the next generation agile Earth-observing satellites. As usual, the goal of Earth-observing satellite mission planning is to maximize the sum weights of selected tasks. But for agile satellites, based on its stronger ability, single criterion is often a poor measure of planning results. For agile satellite, the starting times of the observations are free, different starting time to observe brings different image quality. Meanwhile, for a constellation of satellites, some of them are used frequently and the others are used scarcely may be bad way of usage. The good planning solution is to arrange the satellites reasonably to take images as many as possible with good quality and satisfy different users. We give four criteria to evaluate scheduling results: the sum of weights, average image quality, balance usage of satellites and selected task numbers. A multi-objective evolutionary algorithm is designed to solve the problem. Experimental results suggest that multi-objective mission planning produce good quality of planning result and our algorithm works well for the multi-objective mission planning problem of agile satellites.