Bioinspired Algorithm Applied to Solve the Travelling Salesman Problem

The article is devoted to the study and development of modified bioinspired algorithm and the experimental studies of its characteristics to solve the travelling salesman problem. This algorithm is part of the swarm intelligence method, which is one of bioinspired approaches describing the collective behavior of a decentralized self-organizing system and consists of a multitude of agents (ants), locally interacting between each other and the environment. Ants belong to social insects living inside of a collective the colony. The self-organization includes the numerous mechanisms ensuring the achievement of the global target by a system as a result of insignificant interaction of elements in this system and is the basis of the “social” behavior of ants. Implementation of the local information by system’s elements is the principal feature of this interaction. This eliminates any centralized control and appeal to the global image representing the system in the external environment. Self-organization is a result of interaction between four components: randomness, multiplicity, positive and negative feedbacks. The computational experiment was carried out. The series of tests and experiments have specified the theoretical estimates of the time complexity of the proposed modified bioinspired algorithm. At the best, the time complexity of the algorithm O(nlogn) and in the worst case – o(n ). 3