Flexible Job Shop Scheduling Problem Solving Based on Genetic Algorithm with Model Constraints

An improved genetic algorithm (GA) is presented to solve Flexible Job Shop scheduling (FJSS) problem. This algorithm combines CA with constraint model based on polychromatic sets theory (PST). According to the characteristic of FJSS problem, this algorithm uses contour matrix to formalize the restrictions between workpiece and operations, and between operations and machines. During the process of encoding, decoding, crossover and mutation, the contour matrixes guarantee the genetic searching in the valid solution space. At the same time, the calculation of fitness value is simplified and the computer programming is easy developed. The disadvantage of traditional GA that is the premature convergence and slow convergent speed are improved. Experimental results demonstrate that this algorithm is very stable and its efficiency is improved. It is fit for solving FJSS problem and dynamic scheduling problem.