| Abstract: | 本研究以「矯正鞋墊的舒適度」為研究重點，提出一項系統化分析法，針對個人挑選適合之鞋墊給予最佳的預測與評價，藉以提供鞋具設計者新的方向與方法，達成導入消費者最適化鞋楦選擇之目標。研究上，首先針對20位受測者與6雙樣本鞋墊款式，進行足底壓力實驗量測，找出受測者足壓之各項數據，包括：壓力峰值、壓力－時間積分值、接觸面積比、力量峰值、力量－時間積分值、改良式足弓壓力指標參數等壓值，再應用灰關聯分析，完成不同足弓之受測者在矯正鞋墊舒適度分析運算與評價，達成以人體足壓作為明確訊息數據，探討不明確之舒適度評價之成果預測。其次，本研究透過倒傳遞類神經網路做為學習過程之工具，進行足壓與鞋墊款式的學習辨識，建立一套自動評價鞋墊舒適度的系統，根據不同的足壓，系統有能力從資料庫中挑選出最適合的鞋墊款式，清楚辨識出各受測者足弓所適合之矯正鞋墊，而將舒適度數據匯入智慧分群系統，也可達到80％的最適鞋墊正確率預測，提供鞋具產業在產品開發的應用。 本研究具體成果與貢獻如下：1. 提供足部的力學壓力量測數據，作為矯正鞋墊之實例驗證。2. 探討受測者足壓、足弓指標與矯正鞋墊間之關聯性。3. 研究灰關聯分析於足底壓力分佈資料之舒適度預測成效。4. 應用倒傳遞類神經網路，進行矯正鞋墊足壓學習、驗證，達到分群效果，並找出適合各類足弓者之特徵鞋墊。關鍵詞：足底壓力、足弓指標、矯正鞋墊、灰關聯分析、倒傳遞類神經網路、舒適度評量。
In this research, the focus of the fitness of an orthotic insole is proposed. A systematic analysis method provides the best prediction and evaluation for the individual to choose the suitable insole for providing a new direction and method for the designer of the shoe to achieve the goal for consumers are the best choice for shoe insoles. In this research, firstly, 20-foot testers and 6 pairs of sample insole styles were used to measure the plantar pressure to find out the data of the measured foot pressure, including: pressure peak, pressure-time integral value, contact area ratio, force peak, force-time integral value, modified arch pressure index parameters, etc., and then apply grey correlation analysis to complete the analysis and evaluation of the insole comfort in the different arches. As a clear message data, we will explore the results of unclear comfort assessments. Secondly, this study uses the back propagation neural network as a tool for the learning process, learns the identification of foot pressure and insole style, and establishes a system to automatically evaluate the comfort of the insole. According to different foot pressures, the system has the ability to access data. The most suitable insole styles are selected in the database, and the corrective insoles suitable for the arch of each subject are clearly identified, and the comfort data is transferred into the intelligent grouping system, and the optimal insole correction rate up to 80% can also be achieved. With the application of industrial design in product development can be achieved.The specific results and contributions of this study are as follows:1. Provide the plantar pressure measurement data of the foot as an example to verify the corrective insole.2. To explore the relationship between the foot pressure, the arch index of the testers and the orthotic insole.3. Study the grey relational analysis to predict the fitness of the plantar pressure.4. Apply back propagation neural network to correct and verify the insole's foot pressure, achieve the clustering effect, and find out the characteristic insole suitable for all kinds of archers.KEYWORDS: Plantar pressure, Arch index, Orthotic insole, Grey relational analysis, Back propagation neural network, Fitness of evaluation. |
