摘要
本文主要針對二維夾持系統於夾持規劃階段時，利用均方根統計定位變異量與定位精度矩陣兩種分析方法，探討工件定位精度的特性。首先進行幾何夾持拘束分析，找出定位包括直線及圓弧夾持接觸邊之稜柱形工件的定位裝置定位位置。再針對不同型式的夾持接觸邊進行夾持接觸軌跡分析，找出三點接觸的夾持必要條件。當夾持接觸點有定位變異量時，使用均方根統計定位變異量分析，探討工件直線位移變異量或旋轉角變異量的變化情形。並使用蒙地卡羅模擬程式來加以驗證。最後，由所定義之定位精度矩陣，發展出整體定位精度、直線位移定位精度與旋轉角定位精度的評估準則，並找出評估結果為最佳定位精度的定位裝置定位位置。 由統計定位變異量分析的結果可知，夾持接觸點法線方向定位變異量對工件變異量的影響遠大於切線方向。而定位精度矩陣分析的評估準則可以快速決定稜柱形工件最佳定位精度的定位裝置定位位置。定位精度最佳的定位裝置定位位置，宜採用工件直線位移變異量與旋轉角變異量最小的定位位置。若定位位置接近奇異定位點時，工件變異量會變得非常大，其定位精度最差。由分析的結果可知，均方根統計定位變異量、定位精度矩陣與蒙地卡羅模擬三種分析方法，對工件之最小與最大直線位移變異量與旋轉角變異量的定位位置之結果，具有相當好的一致性。
並列摘要
This paper presents two analyses, Root Sum of Square (RSS) statistical localization variance analysis and localization accuracy matrix analysis, for the characteristics of localization accuracy of a 2D fixturing system in fixturing planning stage. By using geometric fixturing constraint theory, prismatic workpiece with line or arc segments is fixtured by locators in a unique and accurate position and orientation. The fixturing contact locus analysis is next employed to verify the necessary fixturing condition of 3-point contact. According to the localization variances of the fixturing contact points and the shapes of prismatic workpiece, the translational and rotational position variances of a workpiece can be expressed by the variances of the position errors of locators. RSS statistical localization variance analysis is then utilized to calculate the translational and rotational position variances of a workpiece. Monte Carlo simulation is carried out to verify the obtained results. By using localization accuracy matrix analysis, three principles for evaluating different localization accuracies are obtained, and the position of locators with the best localization accuracy can be found. From the results of these analyses, some important conclusions are obtained. The normal positional variances of the fixturing contact points for the positional variance of the workpiece are more important than the tangential ones. The localization accuracy matrix analysis provides a quick way to find the position of locators with a minimum localization variance and the best localization accuracy. The translational and rotational position variance of a workpiece are extreme large when the position of locators is near singular point. Different examples are given to justify these methods. The results of these three analyses are consistent with each other.