模糊樹突狀神經元模型控制應用於下肢復健機器人系統
No Thumbnail Available
Date
2020
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
本論文目的在於完成下肢復健機器人系統，幫助受傷後無法行走的患者進行復健。為了達到降低成本及安全的目標，我們採用氣體作為整體系統的主要動力源。整體氣動式下肢復健機器人機構主要是透過各關節一單桿氣壓缸，藉由活塞兩側氣室壓力不同達到活塞往復運動或定位控制，並且透過桿長的伸縮推拉腿部機構達到如同關節轉動的效果。在系統機構確定後，本論文也進行正逆運動學的模擬驗證，瞭解步態行走時，各關節角度與端點空間座標的關係。控制活塞兩側氣室壓力的部分，採用開關閥進行氣流的控制，透過軟體撰寫脈衝寬幅調變的系統控制程式來決定開關閥單位週期的開關時間。本論文提出模糊樹突狀神經元模型演算法，透過模糊邏輯與樹突狀神經元模型演算法加強細微神經傳導訊號的特性，加強整體平台移動時的精確度。確定脈衝寬幅調變的系統能使開關閥有效控制氣體流量後，使用差動脈衝寬幅調變法，同時搭配傳統比例積分微分、智慧型的模糊類神經網路和本論文提出的模糊樹突狀神經元模型等控制法運算開關閥單位週期開關時間進行氣動式下肢復健平台控制。透過不同演算法計算的數值達到不同步態軌跡關節角度追蹤響應，並計算在單關節、雙關節同動及雙腳同動時，各控制法針對各關節控制上的誤差性能指標進行控制法優劣評估及能否有效抑制氣動造成的同動震盪。
The thesis aims to develop a lower limb rehabilitation robot system to help patients recover from the injuries and have an ability to walk by themselves. To create the safety and low cost system, we use the high pressure air to be the main power of the system. Each joint of the system has a pneumatic cylinder, and the reciprocating motion or positioning control of pneumatic cylinder piston is achieved through the different pressure of the air chamber on both sides of a piston. The movements of a piston change the length of the rod. Because of the lower limb mechanism, the joint will be rotated through the rod push or pull the leg. Also, the thesis derives the kinematics analysis of the lower limb to obtain the relationship between the angles and the coordinates of the end point. The thesis used the solenoid valve to control the pressure of chambers. The thesis uses the program of the pulse width modulation (PWM) to control the time of the air into the chambers. The thesis proposes the algorithm Fuzzy Dendritic Neuron Model which combines Fuzzy logic with Dendritic Neuron Model to enhance the characteristic of subtle nerve signal and improve the accuracy of the platform control. After confirming the PWM signal enable solenoid valve to control the air flow, using the differential algorithm with the PID, FNN and Fuzzy Dendritic Neuron Model proposed in the thesis to calculate the PWM values given to the solenoid valve to track the trajectory of the human being lower limb gait. According to the command of the gait, the algorithms are developed to control the lower limb rehabilitation robot. Finally, the experiment will conduct the analysis of the performing results to evaluate the accuracy and robustness of the designed.
The thesis aims to develop a lower limb rehabilitation robot system to help patients recover from the injuries and have an ability to walk by themselves. To create the safety and low cost system, we use the high pressure air to be the main power of the system. Each joint of the system has a pneumatic cylinder, and the reciprocating motion or positioning control of pneumatic cylinder piston is achieved through the different pressure of the air chamber on both sides of a piston. The movements of a piston change the length of the rod. Because of the lower limb mechanism, the joint will be rotated through the rod push or pull the leg. Also, the thesis derives the kinematics analysis of the lower limb to obtain the relationship between the angles and the coordinates of the end point. The thesis used the solenoid valve to control the pressure of chambers. The thesis uses the program of the pulse width modulation (PWM) to control the time of the air into the chambers. The thesis proposes the algorithm Fuzzy Dendritic Neuron Model which combines Fuzzy logic with Dendritic Neuron Model to enhance the characteristic of subtle nerve signal and improve the accuracy of the platform control. After confirming the PWM signal enable solenoid valve to control the air flow, using the differential algorithm with the PID, FNN and Fuzzy Dendritic Neuron Model proposed in the thesis to calculate the PWM values given to the solenoid valve to track the trajectory of the human being lower limb gait. According to the command of the gait, the algorithms are developed to control the lower limb rehabilitation robot. Finally, the experiment will conduct the analysis of the performing results to evaluate the accuracy and robustness of the designed.
Description
Keywords
單桿氣壓缸, 開關閥, 比例積分微分(PID)控制器, 模糊類神經網路(FNN), 模糊樹突狀神經元模型(Fuzzy-DNM), pneumatic cylinder, solenoid valve, Proportional-integral-derivative (PID) controller, Dendritic Neuron Model (DNM), Dendritic Neuron Model (Fuzzy-DNM) controller, Proportional-integral-derivative controller