Technology Efficacy in Active Prosthetic Knees for Transfemoral Amputees: A Quantitative Evaluation

Table 2

Normalized knee gait parameters, 𝑄, of individual components in active prosthetic systems.

	Knee angle				Knee torque			Knee power
	𝜃max.stance	𝜃min.stance	𝜃max.swing	𝜃min.swing	𝑇max.stanceflex.	𝑇min.stanceext.	𝑇Area	𝑃max.propulsion	𝑃max.swing
Sensors
S1p [37, 39, 41–43]	0.68𝜃^𝑛	2.38𝜃^𝑛	1.03𝜃^𝑛	3.2𝜃^𝑛	7.69𝑇^𝑛	1.14𝑇^𝑛	0.93𝑇^𝑛∗	1.41𝑃^𝑛	0.2𝑃^𝑛
S2f [6]	0.5𝜃^𝑛	0.5𝜃^𝑛∗	0.98𝜃^𝑛∗	0.95𝜃^𝑛∗	5𝑇^𝑛	0.75𝑇^𝑛	0.2𝑇^𝑛	0.44𝑃^𝑛	0.12𝑃^𝑛
S2s [42]	0.35𝜃^𝑛	NV	0.9𝜃^𝑛	2𝜃^𝑛	NV	NV	NV	NV	NV
S3 [39, 41, 43]	0.72𝜃^𝑛∗	1.83𝜃^𝑛	1.08𝜃^𝑛	4.67𝜃^𝑛	4.33𝑇^𝑛∗	1.18𝑇^𝑛	1.07𝑇^𝑛∗	1.06𝑃^𝑛∗	0.18𝑃^𝑛
S4 [6]	0.5𝜃^𝑛	0.5𝜃^𝑛∗	1.03𝜃^𝑛	0.95𝜃^𝑛∗	11.38𝑇^𝑛	0.89𝑇^𝑛∗	0.31𝑇^𝑛	1.27𝑃^𝑛	0.18𝑃^𝑛
Actuators
A1e [6]	0.5𝜃^𝑛	0.5𝜃^𝑛∗	0.95𝜃^𝑛	0.95𝜃^𝑛∗	5𝑇^𝑛	0.75𝑇^𝑛∗	0.2𝑇^𝑛	0.44𝑃^𝑛	0.12𝑃^𝑛
A1b [39]	0.9𝜃^𝑛∗	1.5𝜃^𝑛∗	1.13𝜃^𝑛	2𝜃^𝑛	8.75𝑇^𝑛	1.8𝑇^𝑛	1.95𝑇^𝑛	NV	NV
A1s [39, 41]	0.9𝜃^𝑛∗	3.75𝜃^𝑛	1.02𝜃^𝑛∗	4𝜃^𝑛	9.75𝑇^𝑛	0.69𝑇^𝑛	0.83𝑇^𝑛∗	1.63𝑃^𝑛	0.23𝑃^𝑛
A2s [44]	NV	NV	1.15𝜃^𝑛	NV	NV	NV	NV	NV	NV
A2v [41]	0.35𝜃^𝑛	0.5𝜃^𝑛∗	0.98𝜃^𝑛∗	6𝜃^𝑛	2.5𝑇^𝑛∗	1.4𝑇^𝑛	0.12𝑇^𝑛	1.1𝑃^𝑛∗	0.135𝑃^𝑛
A3 [42]	0.35𝜃^𝑛	NV	0.9𝜃^𝑛	2𝜃^𝑛	NV	NV	NV	NV	NV
Control method
C1 [6]	0.5𝜃^𝑛	0.5𝜃^𝑛	0.95𝜃^𝑛∗	0.95𝜃^𝑛∗	5𝑇^𝑛	0.75𝑇^𝑛	0.1𝑇^𝑛	0.44𝑃^𝑛	0.12𝑃^𝑛
C2 [37]	0.9𝜃^𝑛∗	4𝜃^𝑛	0.93𝜃^𝑛	2𝜃^𝑛	NV	NV	NV	NV	NV
C3 [39, 41, 43]	0.72𝜃^𝑛	1.83𝜃^𝑛∗	1.06𝜃^𝑛	4.67𝜃^𝑛	7.33𝑇^𝑛	0.92𝑇^𝑛∗	0.59𝑇^𝑛	1.41𝑃^𝑛∗	0.2𝑃^𝑛
C4 [44]	NV	NV	1.15𝜃^𝑛	NV	1.75𝑇^𝑛∗	0.35𝑇^𝑛	1.15𝑇^𝑛∗	NV	NV
C5 [42]	0.35𝜃^𝑛	NV	0.9𝜃^𝑛	2𝜃^𝑛	NV	NV	NV	NV	NV

S1p: angle sensor, S2f: FSR (force sensitive resistive), S2s: strain gauge, S3: moment sensor or torque sensor, S4: position sensor, and NV: no value reported from study. A1e: electric motor with series elastic springs, A1b: brushless motor, A1s: motor connected to ball screw, A2s: pneumatic cylinder controlled by stepper motor, A2v: pneumatic cylinder controlled by servovalve, and A3: magnetorheological actuator. C1: finite state control, C2: proportional, integral, and derivative (PID), C3: finite-state-based impedance control approach, C4: iterative learning control (ILC), and C5: adaptive control scheme. Values with asterisk indicate the closest to normal values for each of the nine gait parameters of the knee, that is, 𝑄 closest to 1. Based on all nine values for each sensor, actuator, and controller, it can be deduced that most sensors (S1p, S2f, S3, and S4) are all acceptable sensors, motor connected to ball screw (A1s) is the most appropriate actuator, and finite-state-based impedance control approach (C3) would most likely perform as the best controller when combined to other sensor and actuating components.