A two-phase decoding genetic algorithm for TFT-LCD array photolithography stage scheduling problem with constrained waiting time

• Published in: Computers & industrial engineering Vol. 125; pp. 200 - 211
• Main Authors: Hong, Tzu-Yen, Chien, Chen-Fu, Wang, Hung-Kai, Guo, Hong-Zhi
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.11.2018
• Subjects: Genetic algorithm; Photolithography stage scheduling; Smart production; TFT-LCD; TFT-LCD; Smart production; Genetic algorithm; Photolithography stage scheduling
• ISSN: 0360-8352; 1879-0550
• DOI: 10.1016/j.cie.2018.08.024

A two-phase decoding genetic algorithm (TDGA) is developed for TFT-LCD array photolithography stage scheduling problem (PSSP). The proposed TDGA is able to effectively minimize the total idle time as well as control the job exceeded limited waiting time for PSSP. [Display omitted] •A two phase decoding GA is developed for TFT array photolithography scheduling.•MINLP model is developed and shows the TDGA can obtain solution effectively.•An empirical study was conducted in a leading TFT-LCD company.•The proposed TDGA can improve the solution of PSO and GA by 30% and 26%.•The results have validated the proposed approach in real settings. The thin-film-transistor liquid crystal display (TFT-LCD) array process is usually the longest process in TFT-LCD production. During the TFT-LCD array process, the scheduling problem of the photolithography stage entails complicated constraints such as photo mask availability, available machines for jobs, and limited waiting time. This study proposes a two-phase decoding genetic algorithm (TDGA) to maximize utilization in the photolithography stage, which is usually the bottleneck. An empirical study was conducted at a leading TFT-LCD manufacturing company. To compare the performance of the TDGA with other metaheuristics, eight scenarios were simulated based on the empirical data. The experimental results show the practical viability of the proposed TDGA.