Research On Optimal Scheduling Of Silicon Single Crystal Production Process

Silicon single crystal production has the characteristics of complex process,multiple procedures,long cycle,and high energy consumption.In the large-scale silicon single crystal production process,how to improve factory production efficiency,shorten production cycles,and reduce production costs while ensuring production safety has become the most concerned issue of enterprises.Therefore,in view of the above-mentioned problems,this paper considers multiple constraints at the same time,and conducts optimization scheduling research on the silicon single crystal production process.The main work content is as follows:(1)Through understanding the process flow of silicon single crystal growth and the characteristics of silicon single crystal production,the type of silicon single crystal production scheduling is clarified,and the current existing scheduling methods and research results are reviewed.(2)In order to improve the production efficiency of silicon single crystal under the premise of ensuring the safe production of the factory,this paper considers the maximum power load constraint allowed by the factory,established a silicon single crystal production process scheduling model with the goal of minimizing the maximum completion time,and adopted GPSO is used to solve the problem.In this algorithm,individuals perform crossover operations and mutation operations to improve the algorithm’s global search capabilities,and follow the global optimal particles and individual optimal particles to accelerate the convergence speed of the algorithm.Through case simulations of scheduling problems of different scales,and comparing with the optimization results of the GA and the PSO,it is shown that the algorithm is better than the GA and the PSO in solving large-scale problems.Finally,the task scheduling scheme is given.(3)In order to provide a certain basis for the selection of the transformer capacity of the factory while ensuring the production efficiency of silicon single crystals,this paper establishes a silicon single crystal production process scheduling model with the goal of minimizing the maximum completion time and minimizing the maximum power load and then proposes a Disturbed Non-dominated Sorting Particle Swarm Optimization combined with Genetic Algorithm(RDNSGPSO).The algorithm combines the NSGA-Ⅱ and GPSO algorithm,evaluates the pros and cons of particles through fast non-dominated criteria,uses GPSO update strategy to update the particles,and selects the Pareto frontier particle closest to the individual as the perturbation particle to improve the population diversity.Finally,through the standard function test and example simulation,and compared with the solution results of the NSGA-Ⅱ,NSPSO and Non-dominated Sorting Particle Improved Swarm Optimization(NSGPSO)algorithm,showing the effectiveness of the algorithm superiority.(4)In order to reduce production costs under the premise of ensuring safe production in the factory,this paper combines the time-of-use electricity price policy and considers the maximum power load constraints allowed by the factory,and a silicon single crystal production process scheduling model with the goal of minimizing tardiness/advance cost and minimizing energy cost is established,and the RDNSGPSO algorithm based on real number coding strategy is used to solve the problem to reduce the complexity and running time of the algorithm.In addition,in order to improve the global search capability of the algorithm,the parameters of real number coding have been adjusted.Finally,a case simulation is compared with the result of not using time-of-use electricity prices,which shows that the model proposed in this paper can reasonably adjust and transfer processes,achieve the purpose of reducing energy consumption costs,and have a certain guiding effect on actual production.