Research on genetic algorithms for the discounted {0-1} knapsack problem

被引：0
|
作者：
He Y.-C. [1 ]
Wang X.-Z. [2 ]
Li W.-B. [3 ]
Zhang X.-L. [4 ]
Chen Y.-Y. [1 ]
机构：
[1] College of Information Engineering, Shijiazhuang University of Economics, Shijiazhuang
[2] College of Computer Science and Software Engineering, Shenzhen University, Shenzhen, 518060, Guangdong
[3] Laboratory of Network and Information Security, Shijiazhuang University of Economics, Shijiazhuang
[4] College of Mathematics and Information Science, Hebei Normal University, Shijiazhuang
来源：
| 2016年 / Science Press卷 / 39期
关键词：
Discounted {0-1} knapsack problem; Genetic algorithm; Greedy strategy; Non-normal coding individual; Repair and optimization;
D O I：
10.11897/SP.J.1016.2016.02614
中图分类号：
学科分类号：
摘要：
At present, the deterministic algorithm based on dynamic programming is the main method for solving the discounted {0-1} knapsack problem (D{0-1}KP). But its complexity is pseudo polynomial time, and when the value coefficients and the weight coefficients of the D{0-1}KP instance are in a large range, the deterministic algorithm is no longer practical. In this paper, we use genetic algorithm with elitist reservation strategy (EGA) to solve the D{0-1}KP. Firstly, we establish two new mathematical models of the D{0-1}KP. Secondly, in order to use EGA to solve the D{0-1}KP based on the first mathematical model, we propose a greedy repair and optimization algorithm (GROA) to deal with the non-normal coding individual. Combining EGA with GROA, we give the first genetic algorithm (FirEGA) for solving the D{0-1}KP. Thirdly, for solving the D{0-1}KP by EGA and the second mathematical model, we propose another algorithm named NROA, which is an greedy repair and optimization algorithm too, to deal with the non-normal coding individual, and give the second genetic algorithm (SecEGA) for solving the D{0-1}KP based on EGA and NROA. Finally, we ascertain the reasonable values of the crossover probability and the mutation probability of the FirEGA and SecEGA on the basis of the computational results of four kinds instances of D{0-1}KP. The computational results show that FirEGA and SecEGA are fit for solving the large instance of the hard D{0-1}KP, and they can obtain an approximation solution whose approximation rate is close to 1. Furthermore, the average performance of FirEGA is more efficient than SecEGA. © 2016, Science Press. All right reserved.
引用
收藏
页码：2614 / 2630
页数：16
相关论文
共 30 条
  • [21] Schmitt L.M., Theory of genetic algorithms, Theoretical Computer Science, 259, 1-2, pp. 1-61, (2001)
  • [22] Mitchell M., An Introduction to Genetic Algorithms, (1996)
  • [23] Chen G.-L., Wang X.-F., Zhuang Z.-Q., Et al., Genetic Algorithm and Its Applications, (2003)
  • [24] Liu Y., Kang L.-S., Chen Y.-P., Non-Numerical Parallel Algorithms (2)-Genetic Algorithm, (2003)
  • [25] Michalewicz Z., Et al., A modified genetic algorithm for optimal control problems, Computers & Mathematics with Applications, 23, 12, pp. 83-94, (1992)
  • [26] Runarsson T.P., Yao X., Stochastic ranking for constrained evolutionary optimization, IEEE Transactions on Evolutionary Computation, 4, 3, pp. 284-294, (2000)
  • [27] Coello C.A.C., Theoretial and numerical constraint-handling techniques used with evolutionary algorithm-A survey of the state of art, Computer Methods in Applied Mechanics & Engineering, 191, pp. 1245-1287, (2002)
  • [28] Michalewicz Z., Schoenauer M., Evolutionary algorithms for constrained pararmeter optimization problems, Evolutionary Computation, 4, 1, pp. 1-32, (1996)
  • [29] He Y.-C., Song J.-M., Zhang J.-M., Et al., Research on genetic algorithms for solving static and dynamic knapsack problems, Application Research of Computers, 32, 4, pp. 1011-1015, (2015)
  • [30] Pisinger D., Where are the hard knapsack problems?, Computers & Operations Research, 32, 9, pp. 2271-2284, (2005)
123