本子計畫三之目標在於完成一組穿戴式復健輔具，分為硬體與軟體兩大部分，硬體方面為機構、控制器與電路設計，軟體方面則為控制演算法的設計。硬體方面，機構部分使用SolidWorks來建構穿戴式復健輔具之模型，而採用ANSYS分析各部分物件受力情況並進行修正且改良。材料部份選用能達到高肌肉強度和可伸縮性的液壓放大自修復靜電致動器(Hydraulically amplified self-healing electrostatic, HASEL actuators)來製作人工肌肉，此材料受到電損具有自行修復的能力，使穿戴式復健輔具能夠達到輕量且高適應性，無須仰賴其他較為厚重的剛性架構；電路部分使用德州儀器的DSP系列做為控制核心，搭配自行設計之電路板，並以電池做為動力來源，配合高壓放大器，驅動穿戴式復健輔具，並使用DAQ訊號擷取電流訊號，以計算電極之電容值。軟體方面，本計畫擬利用Python程式語言進行撰寫，並採用適應性類神經網路模糊推論系統(Adaptive Network-Based Fuzzy Inference System, ANFIS)訓練HASEL驅動電壓、電容及機構角度之關係之data driven模型，當利用HASEL人工肌肉進行人體肢體復健時，利用該data driven模型將電容值轉成角度之回饋訊號，以進行人體肢體之復健運動控制。此計畫之研究期許能將軟硬體整合，實際進行實驗並根據測試結果進行改良，最終達到實用化、舒適化和人性化並符合輔助使用者的復健及生活需求。 ;The main objective of the third subproject is to build a set of wearable rehabilitation assistive device. This involves two major aspects: Hardware and software. For hardware aspect, main focus will be on mechanical design, controller and circuit structuring. Software aspect will be focused on designing and composing control algorithm.The hardware involves different components. For mechanism structuring, SolidWorks will be used to construct the model of these wearable rehabilitation assistive device. ANSYS will be used to analyze the reaction of each individual mechanical part under different force and pressure, providing directions for correction and improvement. In regards to materials, HASEL(Hydraulically Amplified Self-healing Electrostatic) actuator, with its known high muscle strength and scalability, will be used to make artificial muscles and construct these wearable rehabilitation assistive devices. This material can be self-perceived its movement and self-healing electrical damage, enabling wearable assistive tools to be lightweight and highly adaptable without relying on heavy rigid structure. As for structuring circuit, DSP will be used as core control and will be combined with our designed circuit board. The battery will be used as the power source to activate and drive these wearable subsidiary tools with high voltage amplifier. The DAQ device is used to capture the signal from the current sensor to calculate the capacitance of the electrode.As for software aspect, Python will be employed in this research to compose necessary algorithm. In addition, ANFIS will be used to train the data driven model. Using the database of angle and capacitance values to control the voltage of HASEL actuator to change its action. When using HASEL artificial muscles for human limb rehabilitation, the data driven model is used to convert the capacitance value into an angle feedback signal to perform the rehabilitation motion control of the human limb.During the research period of this project, a full integration between hardware and software will be implemented and practical improvements will be made based on the actual experiments and the testing results. The final goal of this research is to achieve comfortable rehabilitation assistive device in line with the needs of the corresponding users.
