Combining convergence and diversity in evolutionary multiobjective optimization
Evolutionary Computation
A fast and elitist multiobjective genetic algorithm: NSGA-II
IEEE Transactions on Evolutionary Computation
Hi-index | 0.00 |
The multiple Unmanned Aerial Vehicles (UAVs) reconnaissance problem with stochastic observation time (MURSOT) is modeled by modifying the typical vehicle routing problem with stochastic demand (VRPSD). The objective consists in optimizing mission duration, total time and the quantity of UAVs. This multi-objective optimization problem is solved using a steady-state multi-objective evolutionary algorithm MOEA with e - dominance conception. In this paper, we propose a heuristic evolutionary operation (HEO) using Insert-to-Nearest Method (INM). Route Simulation Method (RSM) is presented in details to estimate the expected cost of each route and this method is designed especially for our MURSOT. The developed algorithm is further validated on a series of test problems adapted from Solomon's vehicle routing problems. Experimental results show that the INM is capable of finding better solutions in contrast and its advantage is more remarkable as the size of the problem become larger.