An Improved Variable-Length Representation Approach for Knapsack Problem

One of the methods to solve knapsack problem is the variable-length representation approach based on genetic algorithm (GA). The original variable-length representation approach has the problem of low efficiency. In this paper we present a improved approach. In the early stage of the evolution, the heuristic searching method is applied to search better solutions rapidly, and in the latter part of evolution the random searching method is applied to improve the result. Through the simulation on some benchmark and two random knapsack problems, the results show that the improved variable-length representation approach has superiority in speed and accuracy over the standard method