Optimal process parameters design for a wire bonding of ultra-thin CSP package based on hybrid methods of artificial intelligence

Purpose – The aim of this research is to combine the Taguchi method and hybrid methods of artificial intelligence, to use them as the optimal tool in wire bond designing parameters for an ultra-thin chip scale package (CSP) package, and then construct a set of the optimal parameter analysis flow and steps. Design/methodology/approach – The hybrid methodology of artificial Intelligence was used in order to identify the optimum parameters design for a wire bonding of ultra-thin CSP package. This paper employed desirability function to integrate two quality characteristics (loop height and wire pull strength) into a single quality indicator to construct a well-trained neural network prediction system with hybrid genetic algorithm. Findings – The processes parameters of low-loop of micro HDD driver IC were optimized with GA, thereby achieving the objective of improving process yield and robustness design of micro HDD driver IC. Practical implications – The engineers could quickly obtain the optimal production process parameter with the demand of multi-quality characteristics, and enhance the assembly quality and yield of driver IC of micro HDD. Originality/value – This paper applies the design of experiments approach to a lower wire loop processes parameters design, and improves the process yield and robustness design of micro HDD driver IC.