Differential Evolution Quantum Particle Swarm Optimization for Solving Job-shop Scheduling Problem

摘要

Job-shop Scheduling Problem (JSP) has become a hotspot for production management and combinatorial optimization research due to its high computational complexity. For the more complicated condition of JSP, how to achieve a short time to obtain the JSP optimal solution by designing the real-time and efficient strategy is still a problem in recent years. To solve this problem, a particle swarm optimization algorithm based on differential quantum properties is proposed. On the basis of quantum-behaved particle swarm optimization, variation, crossover, and selection operation are utilized by particles, which can better keep the diversity of the particles in the population, avoiding the local optimum in the later phase of the iteration. Multi neighborhood search is used for particles' local search to improve search precision. The operation-based encoding is firstly used to allow the research optimization algorithm for JSP solving. The results show that the research algorithm could achieve a short time to obtain the JSP optimal solution, the search precision and speed are improved obviously comparing with other algorithms such as the classical particle swarm optimization.