Implementation of voting-based model with Whales Optimization Algorithm for solving path planning of printed circuit board’s hole drilling process
|
Full Text |
Pdf
|
|
Author |
Kok Chi Zhao, Sharifah Najwa Syed Mustafar, Amar Faiz Zainal Abidin, Ili Najaa Aimi Mohd Nordin, Ismail Mohd Khairuddin, Ahmad Fariz Hasan, Sigit Widiyanto and Masmaria Abdul Majid
|
|
e-ISSN |
1819-6608 |
|
On Pages
|
1671-1676
|
|
Volume No. |
20
|
|
Issue No. |
19
|
|
Issue Date |
January 10, 2026
|
|
DOI |
https://doi.org/10.59018/1025189
|
|
Keywords |
computational intelligence, whale optimization algorithm, printed circuit board, path planning, travelling salesman problem.
|
Abstract
Printed Circuit Boards are fundamental components in almost all electrical devices, and reducing their production cost remains a key priority. One effective approach to minimising the cost per unit is to enhance the efficiency of the drilling process by optimising the travel distance of the drill bit. This paper explores the application of the Whale Optimization Algorithm, a continuous-based optimisation method, in improving path planning, which is state-based in nature. The main difficulty is to model the state-based problem in order to fit a continuous-based algorithm that is different in nature. In this paper, the voting-based model is proposed to ease the conversion between these different spaces, where each dimension of the search space corresponds to a number of votes representing the sequence in which the drill bit visits each hole. The higher the vote, indicates earlier the drilling sequence or turn. The proposed approach is benchmarked against one of the popular case studies from existing literature, with results showing that the proposed model might solve the difficulty of converting two different types of optmisation models but indirectly creates additional ‘local minima’ where the model creates multiple proposed solution that leads to the same path. This creates additional unwanted computation by the algorithm to escape from the local minima. This can be clearly seen from the result, where the proposed approach has the largest average number of iterations before obtaining the global best value compared to other algorithms.
Back