| Semiconductor Wafer Fabrication System(SWFS)is the most complicate production processing system in semiconductor manufacturing processing.The requirements of wafer production and handling process are becoming increasingly stringent as the development of wafer manufacturing equipment,processes and technology expand and the need of size and weight increase.As the primary tool for wafer processing,the Automated Material Handling System(AMHS)has evolved into an integral part of SWFS.So as to effectively improve wafers handling efficiency and equipment utilization,unified AMHS has gradually become the first choice for wafer manufacturers.However,due to the polycyclic coupling effect caused by the direct connection of Interbay and Intrabay in the unified AMHS,the number of requests for the wafer handling in the system and the handling task of the Overhead Hoist Transport(OHT)are greatly increased.Considering the temporary congestion and conflict of OHTs would occur more frequently,the temporary congestion and conflict of OHTs would occur more frequently at the same time,the handling efficiency and reliability of the AMHS have become an urgent problem for the efficient development of semiconductor wafer manufacturing systems.As stated above,unified AMHS was taken as the research object,and the scheduling strategy of unified AMHS was studied deeply considering the structural and operational characteristics of the unified AMHS.Eventually the multi-objective optimization of AMHS operational performance was achieved.The detailed descriptions are as follows:1)Studied the hot lot’s handling performances in unified automated material handling system(AMHS)of wafer fab.A sequential-search heuristic rule based multi-objective scheduling approach was proposed considering the order of handling between emergency and ordinary.A unified AMHS hot lot’s multi-objective scheduling optimization model was established by analyzing the constraint relationship between different wafer card processing time,processing equipment and other factors in the system,which realizes the comprehensive optimization performances in term of lot’s average waiting time,average transport time and average cycle time and effectively improves hot lot’s handling performance while reducing the impact on normal lot’s processing comprehensive performances,and then achieved an unified AMHS emergency optimized scheduling of hot lots.2)Studied the temporary congestion in the handling of OHT in unified AMHS,a multi-objective scheduling method based on the modified Hungarian algorithm was proposed to improve the handling efficiency of material handling systems.A no-wait orbit,which considers the form of temporary congestion in unified AMHS,was built to reduce the probability of system’s temporary congestion to some extent.And the best real-time scheduling of OHTs and wafer was realized by Hungarian algorithm.The optimized route of OHT was obtained by the shortest route rules and the material handling performances was more fast and efficient.3)According to the distinguished characteristics of the real-time optimization scheduling of unified AMHS,aiming to improve the performance indicators which was taken by OHT’s average utilization ratio,delivery time,waiting time and due date satisfaction rate of wafer lots and considering the comprehensive performances of the unified automated material handling system and semiconductor wafer processing system,a dynamic scheduling strategyl based fuzzy logic control was proposed.The scheduling collects wafer lot’s due date,wafer lot’s waiting time and AMHS’s loading ratio in real-time and dynamically adjusts the current scheduling rules of system to realize real-time optimal scheduling of the unified AMHS.Based on the above contents,the effectiveness of dynamic scheduling strategy was verified by simulation experiments through the actual production data of a wafer manufacturing enterprise,and compared with the widely used scheduling rules.All in all,the scheduling strategy proposed could achieve multi-objective optimization,which provided a new method and technique for complex SWFS and AMHS scheduling problems and has a certain reference value for formulating some of innovative and efficient scheduling methods. |
