查詢結果分析
來源資料
相關文獻
- 整合類神經網路與田口式穩健設計法於止滑紋路靜摩擦係數預測之可行性研究
- Neural Network Procedures for Taguchi's Dynamic Problems
- A Fast and Efficient Competitive Learning Design Algorithm Based on Weight Vector Training in Transform Domain
- 專家系統振動訊號圖型判別之研究
- 反傳遞模糊類神經網路於流量推估之應用
- 類神經網路(Neural Networks)的種類及其在影像處理上的應用
- C++Fuzzy類神經網路物件導向發展系統之建立
- 臺灣汽保費率之估計--對數線性費率模式與類神經網路之比較
- 運用類神經網路於股價指數之套利--以日經225指數為例
- 使用類神經網路預估碳化鎢材料放電加工性能
頁籤選單縮合
| 題 名 | 整合類神經網路與田口式穩健設計法於止滑紋路靜摩擦係數預測之可行性研究=Feasibility Study of Integrating Neural Network and Taguchi Robust Design Method for Predicting Static Coefficient of Friction of Non-slip Texture |
|---|---|
| 作 者 | 顏端黎; 王則眾; | 書刊名 | 實踐設計學報 |
| 卷 期 | 22 2025.09[民114.09] |
| 頁 次 | 頁2-25 |
| 分類號 | 440.8 |
| 關鍵詞 | 靜摩擦力; 田口式穩健設計方法; 類神經網路; Static friction; Taguchi robust method; Artificial neural network; |
| 語 文 | 中文(Chinese) |
| 中文摘要 | 隨著臺灣老齡化社會的發展,滑倒事故頻傳,防滑產品設計的重要性愈發凸顯。本研究致力於防滑產品設計,嘗試將類神經網路與田口式穩健設計法相互結合,旨在實現對止滑紋路乾式靜摩擦係數的精準預測。研究初始,通過對市售防滑產品止滑紋路的深入剖析,獲取研究切入點;接續,運用田口式穩健設計法,選定止滑紋路的陰刻、陽刻類型、排列方式以及接觸面積等作為關鍵控制因素,藉助L_(18)直交表規劃實驗,製作實驗樣本,並依據ASTM D1894標準進行摩擦係數測試。將測試所得數據輸入類神經網路進行多輪分析與對比,完成對類神經網路模型的訓練;與此同時,基於田口式穩健設計法生成的各因素優化配比以及參數反應圖的變化趨勢,再次微調控關鍵控制因素之條件,並預測止滑紋路的靜摩擦係數。基於上述研究,針對不同使用環境提出了適配的優化防滑紋路設計方案,並制定了所構建神經網路模型在實際場景中的應用指導準則。本研究提出的整合類神經網路與田口式穩健設計法,有效降低了防滑產品的開發成本,顯著縮短了開發週期,為防滑產品的設計與優化提供了系統且可靠的技術支撐,有助於對推動防滑技術領域的發展。 |
| 英文摘要 | In light of Taiwan's aging population and the prevalence of slip and fall incidents, the significance of anti-slip product design has escalated. This study is dedicated to the design of anti-slip products, and it attempts to combine the neural network and Taguchi's robust design method. The objective of this study is to accurately predict the coefficient of static friction of the non-slip texture dry type. The study commenced with a comprehensive examination of commercially available non-slip textures, which was utilized to establish a foundation for the study. Subsequently, Taguchi's robust design method was employed to select the shading, anodizing type, arrangement, and contact area of the non-slip textures as the pivotal controlling factors. The experimental samples were meticulously planned using an L_(18) orthogonal array, and the friction coefficients were measured in accordance with the ASTM D1894 standard. Subsequently, the data obtained from the tests were entered into a neural network, which underwent several rounds of analysis and comparison to complete the training of the neural network model. Concurrently, the conditions of the key control factors were fine-tuned based on the optimized ratios of the factors and the trend of the parameter response diagrams generated by the Taguchi's robust design method. The static coefficient of friction of the stop-slip texture was predicted. The proposed optimized anti-slip texture design scheme is applicable to diverse usage environments, and the guidelines for the application of the constructed neural network model in real-world scenarios have been formulated. The integration of a neural network with Taguchi's robust design method, as outlined in this study, has been shown to effectively reduce the development cost of anti-slip products, significantly shorten the development cycle, and provide a systematic and reliable technical support framework for the design and optimization of anti-slip products. This approach is expected to contribute to the enhancement of safety and efficiency in various operational settings. |
本系統中英文摘要資訊取自各篇刊載內容。