具自動化量測能力之機器人整體性精密度校正系統

Abstract

本論文提出一具自動化量測能力之整體性精密度校正系統，來解決計算機輔助設計系統與機器人之座標等效（coordinate's equivalence）問題，目前機器人精密度校正的研究大都著重於某些局部區域，也就是說僅要求機器人在某些工作區域的誤差參數（CEPs）能達到所需的精密度要求。這主要是由於不精確的誤差結果不能完全被模式化（model）和限制辨別誤差參數所需量測的數目。為了克服此僅是區域性準確的問題，我們首先提出利用可調式量測空間分割（adjustable measurement space division），根據CEPs實際分佈情形將工作區間合理分割為數個小區域（local region），同時選擇在此小區域的具代表性的誤差能數；並且利用此有限的誤差參數經模糊小腦模型算數計數器（Fuzzy Cerebellar Model Arithmetic Computer，簡稱FCMAC）神經網路的學習法則來產生整個工作區域的合理的誤差參數。我們並完成一機器人位置自動化量測系統。最後，我們透過模擬及實驗來驗證此具自動化量測能力之整體性精密度校正系統的可行。

An automated calibration system is proposed to resolve the coordinates' equivalence problem in integrating the CAD system and robot manipulators. Current robot calibration schemes are inevitably with certain locality, i.e., the calibrated error parameters (CEPs) will produce the desired accuracy only in certain region of the robot workspace. This is mainlybecause of the errors resulting in the imprecision are not completely modeled and only limited number of measured data are available for identifying the CEPs. To overcome this locality problem, we propose first performing the adjustable measursement space division to appropriately divide the workspace into local regions and select the representing set of CEPs from each local region. An automatic pose measurement system is implemented to improve th time-consuming task of measuring the pose of a robot manipulator. Learning algorithms based on the FCMAC neural networks are then employed to generate appropriate sets of CEPs for the whole workspace based on the derived finite sets of CEPs above. Simulations and experiments are excuted to verify the proposed scheme with a variable D-H parameter model.