Aluminum Production Line Scheduling Problem Research

Production scheduling problems has been a research focus since recent years. The efficiency of production scheduling is good for enhancing the overall level of the manufacturing industry. The thesis will study the scheduling problem of Aluminum electrophoresis production line in Nanping Aluminum factory and try to solve the scheduling problems of the production line.The thesis firstly defines and classifies the scheduling problems and then summarizes the studying status and methods of the scheduling problems. On account of the status of the actual production line, the production line is summed up as a special class of hybrid flow scheduling problem and establishes the mathematical model of the production line according to the characteristics of the production process and various types of constraints. The thesis puts forward a kind of improved particle swarm algorithm and applies it to study the static scheduling problem of the production line. The simulation results confirm that the improved algorithm is better than elementary particle swarm algorithm.under the consider of dynamic random characteristics of the actual production line in the real process and the constraints of limited capacity of the cranes, the thesis puts forward a kind of Q learning algorithm to study the scheduling problem of the production line, using some representative rules as behaviors of Q learning algorithm, which realizes the combination of Q learning algorithm and scheduling rules. And then the thesis realizes the active scheduling of the production line based on the scrolled windows mechanism. Then the thesis compares the algorithm with particle swarm optimization algorithm and scheduling rules according to the result of simulation. The results of simulation confirms Q learning algorithm is better than other scheduling algorithms because Q learning algorithm can select different scheduling rules in different system states, optimizing the Scheduling Options. For Q learning algorithm, accurate mathematical model is unnecessary, and it has good timeliness, therefore, the thesis thinks it better than the other algorithms. The thesis studies the static scheduling problem of the Aluminum production line,which can provide reference to adjust scheduling task. The thesis studies the dynamic scheduling problem of the Aluminum production line,which can provide theoretical guidance for the design of the applied system.