Assembly Line Balancing Using Carnivorous Plant Algorithm

Abstract

Assembly line balancing (ALB) involves minimising workstations, eliminating bottlenecks, distributing workloads evenly, and maximising line efficiency. Finding a reliable solution for ALB problems, which are often complex, is crucial. Among the available approaches (exact solution, priority-based heuristics, and meta-heuristics), meta-heuristics specifically, have shown promise in tackling complex real-world problems. The chosen approach for this project is the meta-heuristics approach, specifically the carnivorous plant algorithm (CPA), which has shown promising results in solving mechanical engineering design problems, however, it has not been applied to ALB. This project aims to identify fundamental elements of CPA, implement fundamental elements of CPA to solve ALB problem and evaluate the effectiveness of CPA in solving ALB. The implementation of CPA to solve ALB begins with a thorough review of CPA to identify the fundamental elements of CPA. Subsequently, the identified fundamental elements of CPA are adjusted slightly to ensure the successful implementation of CPA in solving ALB problem. The evaluated performance of the result for ALB solution by using CPA is 0.725 and 5.196 for line efficiency and smoothness index, respectively. The performance of the CPA solution is then compared to the performance of the selected common priority rules solutions namely, largest candidate rule (LCR), smallest candidate rule (SCR), and ranked positional weight (RPW). The result of the comparison showed that the CPA solution is similar to the performance of LCR and RPW solution in terms of line efficiency at 0.725. However, when compared to SCR solution, it showed that CPA solution has 16.7% better performance. Additionally, in terms of smoothness index CPA solution emerged as the victor when compared against the selected priority rules by having 10.6% better performance than both LCR and RPW solution, and 52.8% better performance than SCR solution. According to these results, it is shown that ALB solution by using CPA performed better than the selected common priority rules.