111 人工知能を用いたハイポイドギヤの歯当り評価システムの開発

概要

論文の詳細を見る
Tooth Bearing Inspection is one of the most common inspections for hypoid gear pairs. Usually, in the inspection, hypoid gear pairs are qualitatively evaluated by OK/NG. However, in order to conduct a follow-up survey on bad quality of products, quantitative evaluation of tooth bearing could be effective. Change in machine setting parameters for lapping process enables many kinds of tooth bearing to be purposely given on hypoid gear pairs. Using this method, a certain evaluation value could be given to each tooth bearing. Considering such evaluation values, experienced technicians for the inspection could evaluate each result of the tooth bearing inspection. The present paper describes a computer evaluation system for tooth bearing of hypoid gear pair which can evaluate the results of the inspections instead of the experienced technician. The evaluation performance of the developed system was discussed, and quality of normative tooth bearings which are used for training the developed system greatly affects the performance. The comparison of system evaluations with the technician's ones reveals that correct answer ratio of the developed system within evaluation allowance reached 90% in the best case.
2002-04-18