根據量測雜訊及衛星分佈情形,本論文提出GPS/INS(全球衛星定位系統/慣性導航系統)資料模糊融合系統,以決定GPS與INS測量值權重大小,並由此估測自走車之狀態。對於平行停車之整個控制系統包含:自走車之動態模式;以Kalman濾波器為主之狀態估測器;資料取樣器;模糊決策機制以判定所屬區開;增益排程器以協調停車之程多;以及採用滑模控制之路徑追蹤控制器。 A GPS and INS data fusion system using fuzzy rules is proposed based on the geometry of satellite distributions and measurement noise. Then the outputs of the fuzzy fusion system are used as weights of the GPS and INS signals which in tum is used as a state estimator for para11el parking a car-like mobile robot. The overall control architecture consists of the car-like mobile robot model; a state estimator which is a Kalman filter driven by the fuzzy fusion system; a data sampler; a fuzzy decision maker to c1assifythe parking workspace; a gain scheduler which coordinates the whole parking process; and a path tracking controller which uses fuzzy sliding mode control.