Many optimization problems in the industrial engineering world, particularly in manufacturing systems, are very complex in nature and difficult to solve by conventional optimization techniques. The use of formal approaches is frequently limited owing to the complexity of the models or algorithms. In this paper, a novel approach is proposed for large scale single machine weighted tardiness problems, which combine heuristic algorithm and divide-and-merge strategy. Given an instance of a scheduling problem, our method divides the problem into subproblems using a feasible solution as a basis; then ach subproblem is optimized and the solutions of subproblems are then merged to form a complete solution to the original problem. The aim of proposed method is to find good quality solution within a reasonable amount of computational time. We test our algorithm on big single machine total weighted tardiness problem, which is an NP-hard problem. The results show that the proposed approach have promising application to some traditional hard problems, especially to single machine weighted tardiness problem with big scale.