umagazine issue 18 64 用，他們的手臂肌肉與亞洲人不同， 故相同的手勢，系統識別出來的準確 性有所差異。為此，研究團隊研究基 於Akaike 信息準則的自回歸方法，以 及基於強度分析的小波方法和HilbertHuang 變換方法，對從肌肉表面肌電圖 提取的信號進行分析，並且使用機器 學習技術預測手勢，包括遺傳算法— 反向傳播神經網絡算法，粒子群優化— 反向傳播神經網絡算法，支持向量機 方法和極限學習機算法，以達至Myo 智能手環適合本澳殘障人士及長者使 用需要。 智能輪椅控制系統 這以人因工程為設計理念，結合人工 智能算法，設計出來的創新多功能的 智能電動輪椅，會配合壓力分佈分析 系統及肌電信號，收集本澳殘障人士 及長者在出行及生活上的壓力分佈數 據及肌電數據，以便給予政府及公共 設施機構在設計相關的建築物及設施， 如巴士站、公園等，有著更科學的數 據得以優化，配合澳門智慧城市落地 使用在本澳殘障人士及長者生活上。 未來我與研究團隊將以失能老人長時 間使用輪椅後的壓瘡問題為選題依據， 通過研究失能長者的壓瘡成因，結合 人工智能學習算法，擬合人體壓瘡形 成的“壓力- 時間”函數F(t)。並基於 該函數，建立智慧輪椅非線性控制系 統模型。最終目標是設計一種可以通 過深度學習演算法，預防壓瘡併發症 的智慧輪椅控制系統。 Multifunction Intelligent Electric Wheelchair Designed with human factors engineering and ergonomic concepts in combination with artificial intelligence (AI) algorithms, our innovative multifunction Intelligent Electric Wheelchair can satisfy the needs of the disadvantaged and the elderly with mobility impairments. The chair can be controlled via three methods: muscle signal control (EMG), brain wave control (EEG), and mobile application. Our research study has three core components: the multi-control methods of the Intelligent Electric Wheelchair, innovative function design, and appearance design. The design of the chair is ergonomic and takes into consideration the human anatomy of Macao residents. The smart wheelchair allows a variety of seating positions so users can adopt different postures. This function helps to improve users’ health by preventing them from staying in the same posture for too long. Different control methods, including Myo armband, EEG, and mobile applications, can benefit different users according to their individual needs in different environments. Adjust Gesture Armbands to Meet Local Needs Gestural control methods enabled by Myo armband equipment may be used to control the Intelligent Electric Wheelchair. However, this feature is currently limited to five different gestures, including wave in/out, spread, double tap, and fist. In the future, we plan to develop more gestures to fulfill specific needs of different users. In addition, the armband was originally designed for Western operators, whose arm muscles are different than those of Asians. Thus, the accuracy of gesture recognition on the armband require further adjustment to meet the needs of Macao residents. To improve 優化重型車輛監測路面情況技術以降低交通意外 分析於第 22 屆馮如杯中獲特等獎 Dr Wong receives an Outstanding Award at the 22nd ‘Feng Ru Cup’ College Students Academic and Scientific Works Competition
RkJQdWJsaXNoZXIy MTQ1NDU2Ng==