最佳化策略應用於無人搬運載具穩定性設計－－以不同貨叉材質設計為例

Optimization in design of the Auto Guided Vehicle with different typical fork materials

本論文提出新的最佳化策略應用於無人搬運載具設計，主要結合多目標反應曲面法、田口方法、灰色關聯分析及無因次數分析探討具堆高機功能之無人搬運載具在不同貨叉材質下之最長搬運距離及最小振動值之最佳化關係。首先，以樂高EV3機器人模組模擬具堆高機功能之無人搬運載具，並與無線傳輸振動器結合進行自動進行貨物搬運，選定三個設計參數因子（舉升旋轉度數、車輛重量與馬達轉速）分別針對塑膠與鋁擠型之不同貨叉材質在相同地面環境條件下，應用多目標反應曲面法找出在搬運過程中振動值最小及較佳搬運距離之組合條件。經由最佳化分析得知塑膠材質之貨叉最佳化組合為舉升高度3186度、車輛重量2.4 kg、車輪馬達轉速30 rpm，搬運距離3296 cm、振動值為0.0212；鋁擠型材質之貨叉最佳化組合為舉升高度3000度、車輛重量2.4 kg、車輪馬達轉速22 rpm，搬運距離為2632 cm、振動值為0.0201。研究結果得知鋁擠型材質之貨叉有較佳的穩定性，但塑膠材質之貨叉有較佳的搬運距離。然後將各品質特性，藉由因次分析出的各因子進行變異數分析，結果顯示塑膠貨叉相較鋁擠型有較佳的振動值與搬運距離。最後利用灰關聯分析找出整體最佳因子水準組合並以無因次數分析得出馬達功率及振動值函數關係，本論文提出之最佳化策略可以廣泛應用於評估工業上不同類型無人搬運載具的可靠度分析。 The paper presents a new optimization strategy applied in design of the Auto Guided Vehicle (AGV), mainly with multi-target response surface methodology, Taguchi method, grey relational analysis and dimensionless number to explore the optimize relationships in the maximum better working distance and the minimum vibrations with plastic fork and extruded aluminum fork in the forklift function of AGV. First, LEGO EV3 robot module is designed to forklift function and conduct automatic goods handling in combination with wireless transmission vibrations data logger. Finally, three design parameters (i.e., Motor rotation degrees, vehicle weight and motor revolution) are used to find the minimum vibrations and better working distance on the goods handling via multi-target response surface methodology for the different fork materials (plastic and extruded aluminum) in the same ground environments. The results show that the optimum operation parameters are the lift height: 3186 degrees, the vehicle weight: 2.4 kg, the moves revolution: 30 rpm, the moving distance: 3296 cm and the average minimum vibration value: 0.0212 for the plastic fork. The optimum operation parameters of the other aluminum extrusion fork are the lift height: 3000 degrees, the vehicle weight: 2.4 kg, the moves revolution: 22 rpm, the moving distance: 2632 cm and the vibration value: 0.0201. It is found that the extruded aluminum material of fork is more stability moving than the plastic material of fork, however, the plastic one is the better handling distance than the extruded aluminum one. The dimensionless of each quality is carried out by the analysis of variance (ANOVA) of the design factors, and the results show that the plastic fork has a better vibration value and the moving distance than the extruded aluminum one. We can predict the vibration values of the AGV via the motor power in combination the grey relational analysis with dimensionless number. The optimization strategy is also applied to the other function of AGV.