Stochastic Scheduling For A Single Batch Machine With Incompatible Job Families

In the manufacturing,there is a class of batch processing machines scheduling problems with incompatible job families.By incompatible job families,we mean that jobs from different families cannot be assigned together in the same batch.In this study,stochastic scheduling for a single batch processing machine with incompatible job families is considered,where the jobs have random arrival times and processing times.Firstly,the inventory of buffer is viewed as the system state and minimizing long-term average cost of job is viewed as the system optimization objective,then the stochastic scheduling problem of a single batch processing machine with incompatible job families is modeled as a semi-Markov decision process(SMDP)according to the running mechanism and characteristics of the system.Then,we optimize and control this problem by the policy iterative algorithm,and the influence of different buffer capacity on the job loss rate is analyzed.Finally,a Q learning algorithm combined with simulated annealing(SA-Q)is given for the curse problem of dimensionality which caused by the increase of the job families number.When the number of job families is two and three,the influence of different arrival rate and processing rate on the system performance is analyzed and the optimization effects of the two algorithms is compared.In the modernization of advanced production environment,in order to reduce inventory and waste,the traditional production mode of pursuing productive efficiency will be gradually replaced by just in time(JIT)production mode.For JIT production systems,which require producing only the necessary products in the necessary quantities at the necessary time.Therefore,in this study,stochastic scheduling for a single machine with incompatible job families under demand-driven production mode is considered.Firstly,the inventory of the buffer and bank is viewed as the system joint state and minimizing long-term average cost is viewed as the system optimization objective,then the problem is modeled as a SMDP.In the scheduling problem of demand-driven,the system has a large state space.In order to avoid the curse of dimensionality in the numerical optimization methods,the third chapter analyze the performance of the system under different demand by the SA-Q.Finally,the simulation results show that the model is reasonable and the algorithm is effective.