In flexible manufacturing systems, disordered parallel machine scheduling may cause interruptions in the production line; thus, a reasonable scheduling plan benefits profits of manufacturing. In the literature, the operation sequence, the operation assignment, the tool scheduling, and the tool switching problems have been studied with unlimited resources. However, in practical manufacturing process, a scheduling problem along with these problems with restrained resources should be studied to make scheduling solutions more performable. There are a number of operations processed on a group of parallel machines, and each operation requires a set of tools where the tool copy number and tool service life are limited. The objective of this problem is to minimize the make-span with restrained resources. This article studies a parallel machine scheduling problem combining operation scheduling, tool scheduling and restrained resources, which benefit the real industry. A Tabu-Genetic Algorithm is proposed to find the optimal solution. In small size problem, the proposed method can achieve similar results whereas the mixed integer programming approach can obtain the exact solution. However, in large size problem, the latter cannot obtain the solutions within acceptable times, whereas the proposed method can find solutions within reasonable times. This algorithm performs better when compared to Tabu search algorithm and genetic algorithm in selecting local and global optima.

To improve the modeling accuracy and efficiency of the tool wear monitoring system, a generalized regression neural network is adopted to build the tool wear prediction model because its excellent performance on learning speed and fast convergence to the optimal results whether the sample data are small or large. The low predictive accuracy and efficiency are caused by traditionally manual adjustment of the spread parameters in generalized regression neural network and then the improved fruit fly optimization algorithm is proposed to optimize the spread parameters of regression neural network automatically. Combining the improved fruit fly optimization and generalized regression neural network, the tool wear prediction method is proposed in the paper. Various experiments are carried out to validate the proposed method and the comparison results show a good agreement. In addition, the proposed method is compared to the tool wear prediction method in the literature, and the comparison results also show that the proposed method can achieve better performance.