台灣地區地狹人稠’機車已成爲我們日常生活中不可或缺的交通工具’機 車在行駛時除了雙眼可看見的前方視野，傳統機車均以固定式後視鏡來提供駕 駛者後方視野，但是當機車行駛於彎道或路面變化時，容易造成機車於行進間 可能會瞬間產生視野死角，無法提供駕駛者後方視野的路況資訊，雖然在這個 空窗期並不長，但是某些緊急狀況下卻可能造成交通事故。有鑑於此，本論文 發展出一套具適路性之智慧型機車後視鏡系統，並透過各式感測器來偵測路況 及機車姿態之變化作爲系統之輸入訊號，再運用灰色預測（grey prediction)預 先估算後視鏡應調整角度，並藉由控制器來控制後視鏡馬達，使機車後視鏡可 隨機車俯仰（pitch)、轉向（steer)、側傾（roll)之不同動態與不同道路環境而 改變其角度，改善傳統後視鏡視野死角的缺點，以提供機車駕駛者更寬廣的後 視鏡視野，亦能提升駕駛者安全，減少交通事故的發生。

In Taiwan, the motorcycle has played an important role in our daily transportation. The headlights and rear-view mirrors in a motorcycle act as eyes to provide front and rear views for the rider. Since the motorcycle’s mirrors are fixed, if a motorcycle makes turns or runs on various road surfaces, it is easy to have an instant blind area for a rider, thus causing a traffic accident sometimes.
In this paper, an adaptive rear-view mirror system for motorcycles has been developed. According to the input signals from sensors, this system adaptively adjusts the motorcycle rear-view mirror to follow the changes of vehicle dynamics in pitch, steering and roll angles under different road
conditions. A grey predictor was also developed in this paper to estimate the changes in vehicle dynamics in advance. This system will improve on the shortcomings of the conventional rear-view mirror, and provide wide and appropriate rear views for the motorcycle riders. It can also enhance the riders’ safety and reduce the occurrence of traffic accidents.