基於模糊邏輯控制步態偵測之新型下肢復健外骨骼機器人輔助設計及多軸自我調變控制

Assistive Design and Multiaxis Self-Tuning Control of a Novel Exoskeleton Robot based on Fuzzy Logic Control in Gait Phase Detection for Rehabilitation of Lower Limb

李明展(Ming-Chan Lee)

指導教授 ： 李貫銘

國立臺灣大學/工學院/機械工程學研究所/博士(2022年)

https://doi.org/10.6342/NTU202201706

摘要

本論文之目的為開發創新下肢外骨骼機器人，包含其機構設計、控制策略、自製之感測器、及步態偵測之應用。輔助設計為其機構設計之概念，此輕量化機構之外骨骼具有16公分之無段式可調整長度之大腿及小腿腿管。此外骨骼重量低於16公斤，且每一腿具腰部、臀部、膝蓋及腳踝等四個自由度，以確保穿戴者之舒適，而馬達及諧和式減速機安裝於臀部及膝蓋之關節處以驅動外骨骼。為開發此下肢外骨骼機器人，首先建立包含上下位馬達驅動系統、控制器增益自我調變法及多軸控制之控制策略。其中，上下位馬達驅動系統之設計透過德州儀器生產之微控制器 (TMS320F28069)作為步態指令編程，及開發板 (TMS320F28069/DRV8301)作為關節指令編程。上位控制器透過控制器區域網路發送指令給下位馬達控制器以驅動臀部及膝蓋關節。本論文提出的下肢外骨骼機器人之控制器增益自我調變法整合多軸控制作為馬達控制策略以調整包含電流迴路、速度迴路及位置迴路的馬達控制系統之增益。此外，實驗室自製創新壓阻式觸覺感測器用於下肢外骨骼機器人之腳底壓力量測，並預測人體行走行為及建立步態偵測之應用。此具網格結構之壓阻式觸覺感測器使用多壁奈米碳管混合聚二甲基矽氧烷製作而成，並透過實驗優化此感測器之參雜率及網格結構設計以得到良好的感測範圍，經由感測器測試實驗，最穩定之參雜率為重量百分比7 %，及網格結構比例(寬度/線距/厚度)為1:1:1，測試完成之感測器安裝於外骨骼之腳部鞋墊中，抓取腳底反作用力並進行分析。而步態偵測系統主要由外骨骼、壓阻式觸覺感測器及資料擷取系統所構成，且將感測器量測之腳底反作用力經計算轉換為力量比例係數以改善模糊邏輯控制步態偵測性能，並經由步態偵測演算法驗證。實驗結果指出本論文提出之方法對於規律及不規律步態皆能偵測，軌跡追蹤誤差減少且改善均方根誤差。

關鍵字

外骨骼 ； 輔助設計 ； 多軸控制 ； 上下位控制 ； 增益調變 ； 觸覺感測器 ； 模糊邏輯 ； 步態

並列摘要

A novel lower limb exoskeleton robot, including the mechanical design, control strategy, lab-developed sensor and application to gait phase detection, were developed. The mechanical design of the lower limb exoskeleton robot was based on the concept of assistive design. The exoskeleton with a lightweight mechanism comprised a 16-cm stepless adjustable thigh and calf rod. The exoskeleton weighed less than 16 kg and had four degrees of freedom on each leg, including the waist, hip, knee, and ankle, which ensured fitted wear and comfort. Motors and harmonic drives (HDs) were installed on the joints of the hip and knee to operate the exoskeleton. A control strategy, including master and slave motor-driven system, controller gain self-tuning method, and multiaxis control, was established. The master and slave motor-driven system was programmed using a Texas Instruments microcontroller (TMS320F28069) for the walking gait commands and evaluation boards (TMS320F28069/DRV8301) of the joints. The master controller gave commands to direct the slave motor controller through the controller area network bus (CAN bus) to drive joints of hips and knees. The controller gain self-tuning method for the lower limb exoskeleton robot was proposed. The self-tuning method was proposed to adjust the control gains of the motor control system, including the current, speed, and position loops. The multiaxis control was integrated into the exoskeleton. Additionally, a novel lab-developed piezoresistive tactile sensor was developed and used to measure the foot pressure to predict human walking behavior with the lower limb exoskeleton robot. Based on the lab-developed sensor, the application to gait phase detection was established. The proposed piezoresistive tactile sensor with grid structures was fabricated using polymer composites of multi-walled carbon nanotubes combined with polydimethylsiloxane. The doping ratio and design of the grid structure of the sensors were optimized to achieve a good sensing range. Based on the sensor tests, the most stable doping ratio was fixed as 7 wt%, and the ratio of the grid structure (line width:line spacing:thickness) was 1:1:1. Sensors were set in the soles of the shoes on the foot pads of the exoskeleton, and the foot reactive forces were captured for analysis. The gait phase detection system mainly comprised the exoskeleton, the piezoresistive tactile sensor, and the data acquisition system. The foot reactive forces measured by sensors were calculated and converted to force ratio factors to improve the performance of gait phase detection by the fuzzy logic control. The performance of the proposed algorithm for gait phase detection was verified experimentally. The results indicated that both regular and irregular walking could be detected by the proposed method. Trajectory tracking errors were eliminated, and the root mean square errors reduced.

並列關鍵字

exoskeleton ； assistive design ； multiaxis control ； master–slave control ； gain tuning ； tactile sensor ； fuzzy logic ； gait phase

參考文獻

[1] United Nations Department of Economic and Social Affairs, Population Division (2019). World Population Prospects 2019. https://www.un.org/en/development/desa/population/publications/pdf/ageing/WorldPopulationAgeing2019-Report.pdf (accessed 29 May 2022).

[2] M. Cenciarini and A.M. Dollar, “Biomechanical Considerations in the Design of Lower Limb Exoskeletons,” in Proceedings of the 2011 IEEE International Conference on Rehabilitation Robotics, Zurich, Switzerland, 29 June 2011; p. 5975366. https://doi.org/10.1109/ICORR.2011.5975366.

[3] S. Jatsun, S. Savin, A. Yatsun, and A. Postolnyi, “Control system parameter optimization for lower limb exoskeleton with integrated elastic elements,” in Advances in Cooperative Robotics; World Scientific Publishing: Singapore, 2017; pp. 797–805.

[4] D. X. Liu, X. Wu, W. Du, C. Wang, C. Chen, and T. Xu, “Deep spatial-temporal model for rehabilitation gait: Optimal trajectory generation for knee joint of lower-limb exoskeleton,” Assembly Automation, vol. 37, pp. 369–378, 2017. https://doi.org/10.1108/AA-11-2016-155.

[5] C. Kirtley, “CGA Normative Gait Database,” Hong Kong Polytechnic University, Hong Kong, China, 2005.