A partheno-genetic algorithm for multidimensional knapsack problem

Abstract

The multidimensional knapsack problem is one of the most well known integer programming problems and applied in resource allocation widely. This paper proposes a partheno-genetic algorithm (PGA) for solving this problem. The PGA repeals crossover operators and implements the functions of crossover and mutation by partheno-genetic operators. In partheno-genetic operation, new heuristics and a worst-removed operator are designed for improving the profit of the solution. Simulation results show that the PGA achieves good performance, and it can restraints the immature convergence phenomenon efficiently and the worst-removed operator can improve the searching ability.