According to the characteristics of the traveling salesman problem, a discrete chimp optimization algorithm based on a nearest neighbour traction operator is proposed. Firstly, we introduce the concept of high-quality fragment, and design a retrieval method combining with the optimal individuals of each group to improve the efficiency of the learning strategy with intra-group, and discretize the hunting process of chimpanzee groups in combination with the characteristics of combinatorial optimization problems. Then, through an inter-group communication mechanism to eliminate path crossing of some individuals. Finally, in order to overcome the shortcomings of slow convergence and low search efficiency of the traditional neighborhood search operators, a new neighborhood search method, the neighbour traction operator, is proposed, which has clearer search purpose and more efficient convergence, and an adaptive probabilistic disturbance control strategy is designed to effectively balance the exploitation and exploration of the algorithm. Experiments are conducted on 30 instances from the TSP standard datasets. The experimental results show that the designed discrete chimp optimization algorithm has high solution quality and fast convergence speed, and can be applied to the solution of combinatorial optimization problems.