Improved Ant Colony Algorithm for Partner Selection

Partner selection is a classical combinatorial optimization problem. Its solution is a list of nodes which are the least value of each link. When the links and candidates are increased continuously, the complexity of partner selection grows exponentially. It is difficult to solve that problem in method of exhaustion. So this paper puts forward an improved algorithm, namely partner selection ant colony algorithm, to solve the problem of partner selection. The frame of this paper is as follow: First, set up the mathematics description. As known, the indexes of partner selection always have multiattribute, such as type of benefit, type of cost, type of interval, type of fixation. For obtaining the solution which is representative and objective, the indexes should be standardized differently according to their attributes. Second, set up a new algorithm- partner selection ant colony algorithm. For solving the problem of partner selection effectively, four aspects are modified to improve the basic ant colony algorithm. Third, a computing simulation example is shown. Partner selection ant colony algorithm can quicken the convergence rate, obtain optimal solution and decrease computing time. Computing simulation examples show it is validity.