A Memetic Algorithm Applied To The Allocation Problem Of The Concrete Mixing Plants

The allocation problem of the concrete mixing plants (APCMP) belongs to the combinatorial optimization problem, it shows the more important practical significance with the rapid development of the construction industry. Due to the increasing of the concrete demand, the supply of concrete also begins to scale and normalize, and the reasonable allocation of the mixing plants appears important. The reasonable allocation of the mixing plants can reduce the number of the sites, and then reduce the cost of setting stations; it also can shorten the distance of the mixing plants and the construction areas, ensure the quality of concrete, reduce the transportation cost, and have more benefits to transport vehicle distribution. Therefore, the resolution of this problem has certain theoretical and engineering significance.Because the original known concrete mixing stations can not meet the needs of all key construction areas, now we will choose some one from the given concrete mixing stations to meet the needs. The chosen mixing plants and the original concrete mixing plants will supply the concrete for the all key construction areas together. The problem requires the number of the increased mixing plants should be less as far as far as possible, and the total distance between the mixing stations and the building areas will minimum. Firstly, the paper makes analysis to the problem and sets a model according to the impact of problem conditions; we use the greedy algorithm based on the distance sorting to solve the problem on the basis of specific data. We introduce the steps of this method, and it can get distribution strategy, but there are some deficiencies.In this paper, according the greedy algorithm based on the distance sorting, always can’t get the global optimal solution, we introduce the memetic algorithm to solve the problem after analyzing the concrete mixing station location problem. Memetic algorithm (MA) is a mixture of computational intelligence technology. It is a combination of the global search based on the population and the local search based on the individual and we can constitute a different memetic algorithm using the different search strategy. In a combination of the two search operation, the algorithm can obtain the optimal solution of the problem. This paper introduces memetic algorithm for solving the allocation problem.. We designed the implementation steps of the memetic algorithm according to the specific restriction conditions and the optimized objective of the location problems, including the coding for the problem’s solution, initial population, evolution operation, local search operation and the terminated conditions. The algorithm uses the selection, crossover and mutation in the global search and the greedy algorithm as the local search strategy. Finally, we get the optimum solution to the allocation problem, and get the set of the selected stations and supply distribution solutions for the concrete mixing plants to the key construction areas. The memetic algorithm for the allocation problem can overcome the fault of getting into the local optimum easily of the greedy algorithm and genetic algorithm, according designing the global and local search strategy. The total transportation length of the solution can get the minimum, it proved the memetic algorithm applied to the allocation problem is effective.