The Baldwin Multi-objective Particle Swarm Algorithm And Its Application In Purchase Quantity Allocation

The purchasing quantity allocation problem is the core of procurement management. In order to enable enterprises to purchase the total benefit to achieve optimal, the buyer supply corresponding to each supplier according to the index of cost, quality and delivery time allocation, which is the problem’s main task. PSO is a kind of heuristic evolutionary algorithm simulation of bird foraging behavior of flying cluster. In the iterative process, each particle in the solution space following the personal best particle and the global best particle optimization. Simple and easy to come true and strong versatility is the PSO’s advantage. However, PSO has its defects, such as convergence is slow and lack of diversity. In result, it is necessary to do further study on the improvement of the PSO.In view of the above problems, this paper conducts a preliminary study on the improvement of MOPSO and Baldwin learning strategies application issues. Then apply the improved algorithm to the purchasing quantity allocation problem with discount factors, and do research on the application of algorithm in multi-objective mixed integer problem, which in order to improve the level of enterprise supply chain management. The main contents of this paper include:(1) There are problems in slow convergence and solution set distribution of the algorithm. Based on the theory of multi-objective particle swarm optimization algorithm and Baldwin effect, this paper gives a feasible solution in the neighborhood of the particle and the learning strategies, and put forward the Baldwin hybrid multi-objective particle swarm optimization algorithm. To improve the convergence of the particle swarm algorithm the algorithm, and expand the search solution to improve the solution accuracy. Then prove the validity of the algorithm through the experimental comparison.(2) In case that the improved hybrid algorithm may exist the problem of poor diversity, Baldwin MOPSO use non-uniform mutation and elitist strategy to improve the diversity of particle swarm. Then the improved algorithm sets up a linear change of inertia weight to balance the global search and local search of particle swarm optimization, and prove its better performance through four classical test functions simulation experiment.(3) In practical application, this paper uses the Baldwin hybrid multi-objective PSO to solute the purchasing quantity allocation problem for increasing the number of candidate solutions. By adjusting the particle flight and increasing the equality constraint handling mechanism, the algorithm successfully solves the purchasing quantity allocation problem with discount factors. At last, this paper provides a more efficient and strongly interactive supplier selection and purchasing quantity allocation scheme for decision makers by introducing TOPSIS.