In this article a new version of the ant colony optimisation algorithm with a desirability function for the triple matching problem is described. The problem is modelled by means of two 2-dimensional arrays. The new version of the ant algorithm was compared with the previous version of the ant algorithm and tested for different values of ant algorithm parameters; the results of these tests are presented and discussed.
The subject of this paper is the inventory-production problem, which is a one of the optimization problems in a decision area in which inventory volume and production volume are considered together. There are many approaches to this problem but for the first time, this problem is modelled by means of a capacitated graph network and a solution to the problem is proposed on the basis of this model which consists of finding the maximum flow with the minimum sum of production and inventory cost. In this article, only a solution for one kind of product for the deterministic inventory-production optimisation problem is presented and for this one kind of product, a maximum flow with a minimum cost for each considered demand period is calculated. The maximum flow with minimum cost is a solution to the homogenous inventory-production optimisation problem. The solution to the one kind of product for the inventory-production problem consist of maximum flow with minimum cost for a total demand from all periods, which has been taken into consideration.
In this article, ant colony optimisation algorithms for the triple matching problem are described. This is the first elaborated ant algorithm for this problem. The problem is modeled by means of a 3-dimensional array. The ant algorithm was compared with the Apx3Dmatchnig-F algorithm and tested for different values of ant algorithm parameters. The results of these tests were presented and discussed.
This article describes a new ant colony optimisation algorithm for the facility localisation problem with a new heuristic pattern proposed by the author, which consists of three parts: the function of the average cost of client servicing; the total minimum cost of servicing from a site, which is selected and included into the solution; the function of improving the cost of already serviced clients. In this comparison, simulations were presented, and two parameters were observed: the number of sites and the cost of client servicing. The new algorithm allowed to improve the solution in both of these parameters.