摘要

Defective blocks, which are easily produced at the edges of wafers during the solar cell wafer production process, can affect wafer usability. The traditional approach to prevent this problem from occurring is to cut out and remove the wafer edge up to a certain fixed width. Nevertheless, the removal of such a defective area often leads to usable areas to be cut out, resulting in unnecessary waste. Therefore, this study aims to design an automated cutting path planning algorithm that can calculate a viable cutting path so that after having passed through such a path, wafers can preserve a relatively maximum amount of usable area. The algorithm uses machine vision to detect defective areas in each wafer image, and combines with fuzzy potential function to generate a design for the relatively optimal cutting path. Finally, defective areas at the wafer edge are cut out along this path, resulting in greater usable area being preserved for each wafer as compared to the traditional cutting method, thus allowing each wafer deliver more benefits.