本研究以人因「舒適度」為探討方向，並以足部產品的「鞋墊」作為對象，設計一個以足底壓力（簡稱足壓）為前提的灰關聯（Grey Relational）評價模式，找出該足型與鞋墊樣本中關聯度最高的鞋墊款式，當成最適合的鞋墊選項，作為後續類神經網路（Artificial Neural Network, ANN）的「足壓－鞋墊」之訓練樣本對。藉由多次的疊代計算，類神經網路將擁有足夠的引申能力，能自動針對所輸入的足壓資料進行分群，從現有的鞋墊樣本中找出適合測試者之鞋墊款式，達成本研究最終目的。因此，本研究使用倒傳遞類神經網路（Back-Propagation Neural Network, BPNN）技術，將專家經驗與灰關聯分析技術轉換成數學模式，使不了解相關領域的工業設計師，參考類神經網路的計算所得之分群結果，正確進行設計決策，以縮短設計週期，並滿足顧客的個別差異性需求。
本論文的具體研究成果與貢獻如下：
1. 提供足部的力學實驗作為實例之驗證對象。
2. 探討足壓與鞋墊的舒適度關係。
3. 運用灰關聯於足壓資料的舒適度評價計算。
4. 驗證類神經網路的學習與分群效果。
5. 以類神經網路進行舒適度資料的學習，並作最適鞋墊的預測。 The purpose of the research is to estimate the comfort of a foot with different insoles by using the grey-relational approach based on the plantar pressure. When we find the most-related sample between foot shapes and insoles, we can put them into the artificial neural network (ANN) as the training pair (pressure-insole) for network training. After training iterations, the network will have enough generalizing capability to classify the pattern of the plantar pressure. Back-Propagation neural network (BPNN) is used to convert expertise and to classify insoles.Referring to the classified results estimated by the network, designer who does not master the related domain can make correct decisions when design project is proceeding. Furthermore, this approach can efficiently reduce the design-cycle time and meets the customers’ demands. Results and contributions in this paper are shown in the following:1. To conduct a foot experiment to verify research assumptions.2. To discuss the related comfort factors between the foot shape and insoles 3. To investigate validity of the gray-relational approach to estimate the comfort of foot based on the plantar pressure data.4. To verify the validity of ANN’s learning and classifying5. To use ANN which learning from the comfortable data to predict the most appropriate insole.Keywords: Plantar Pressure, Insole, Grey-Relational, Back-Propagation Neural Network, Comfort Evaluation
