Nonlinear model-based control of pneumatic artificial muscle servo systems

Multiple Simultaneous Specification Control of Antagonistic Actuation by Pneumatic Artificial Muscles

“…PAM의 동적특성을 파악하기 위한 모델링연구에 있어 서 초기에는 단순한 형상변화만을 고려하여 인공근육에 가해지는 공압과 근육이 당기는 장력과의 관계를 유도하 였으나 [2], [5] 보다 정확한 운동을 예측하기 위해, 근육의 재 료특성 [6] , 재질사이의 마찰 [7] , 공기의 유체특성 [8], [9] 등을 포 함하고 있다. 인공근육 동특성 예측이 어려움으로 인하여 인공근육으로 이루어진 관절의 제어기법은 모델링오차로 인한 영향을 줄이는데 효과적인 가변구조제어기 [7], [9] , 신경 망제어기 [10] [11], [12] .…”

“…Ziegler-Nichols PID 게인 설정 방법 [13] ), ( [9], [10] , 그리고 압력제어밸브를 사용하는 방식 [7] 두가지로 크게 나누어 진다. 스풀 ( [14] .…”

MIMO Sliding Mode Controller for Gait Exoskeleton Driven by Pneumatic Muscles

“…It is assumed that pneumatic flow into and out of the PM is an adiabatic process [16]. Thus, the rate of change of pressure in PM can be described by the physics based model as:…”

“…With a single 3/5 analogue valve, the SISO control was made possible by merg-ing the models of four processes and, letting the valve voltage as input and the trajectory as output. Aschemann and Schindele [17] developed a cascade SM control algorithm with four major process models similar to [16]. In the outer trajectory SM control loop, the controller output P  together with desired average pressure ( ) was fed through (1) to provide references for the inner loop controllers.…”

“…This simplified the system models by ignoring the details of the pressure characteristics of the PM actuator, thus unpredictable errors or time delays could be introduced. Shen [16] eliminated these black boxes by modelling the entire control system with four major processes, which were flow dynamics of the valve, pressure and force dynamics of the antagonistic PM actuators and the load dynamics of a linear antagonistic mechanism. With a single 3/5 analogue valve, the SISO control was made possible by merg-ing the models of four processes and, letting the valve voltage as input and the trajectory as output.…”