Research On Intelligent Scheduling Teconology Of Interbay System In Semiconductor Wafer Fabrication System

Semiconductor Wafer Fabrication System (SWFS) is one of the most complicate discrete processing systems in this world, which has characteristics such as reentrancy, mass and mixed products in process, complicated processing operations, coexistence of single-processing and batch-processing equipments, long fabrication time and high investments. As the wafer size grows from 200mm to 300mm and then to 450mm in recent years, the automated material handling system (AMHS) has been widely adopted, and how to improve the overall equipment efficiency of SWFS using AMHS has become a critical and urgent problem to be solved. AMHS includes the Interbay and Intrabay material handling system, and the Interbay system is a very important sub-system. Through the Interbay optimal scheduling, we can reduce the cassette delievery time, promote the vehicles transport efficiency, increase wafer throughput, shorten the cycle time and promote the overall equipment efficiency. Therefore, the study of optimal scheduling methods of Interbay material handling system has great significance to SWFS. The detailed research descriptions of this thesis are as follows.1) The real-time dispatching method of Interbay material handling system. The temporary blockings of OHTs during transporting and the workload balancing constraint are emphatically studied, due to the features of the Interbay material handling system. A cost model of dispatching is proposed considering four factors in real-time dispatching, and an improved Hungarian algorithm based dispatching strategy combined with a fuzzy logic based weight-adjusting model is then presented to solve the target model.2) The optimization method of Interbay material handling system based on compositional rules. Aiming to meet the dynamic adjusting and global optimization demands, a Genetic programming based compositional rule generator is proposed to prioritizing the lots.3) The combined lot scheduling and vehicles transporting method. So far the research of Interbay scheduling has been focusing mostly on the transporting sub-problem and neglecting other sub-problems: the OHT routing problem and lot scheduling before entering the target bay. These three phases constitutes a complete scheduling period and should be integrated into the scheduling model simultaneously. Therefore in this part of the thesis an combined lot scheduling and transporting model is established, and a parallel multi-objective genetic algorithm (PMOGA) is proposed to solve the model.At last, an intelligent scheduling prototype system of the Interbay material handling system is developed based on the actual production data of a wafer fab located in Shanghai China. This prototype is originally designed and developed to validate the effectiveness of the proposed approaches. The experimental results of this thesis provide some novel methods which can improve the overall performance of the wafer fab and eventually promote the profit and reduce production cost of the enterprise.