In Vivo Identification of Skeletal Muscle Dynamics with Nonlinear Kalman Filter: Comparison between EKF and SPKF

Table 3

Quantitative comparison on identification (SPKF versus EKF) in 7 different initial conditions.

Rabbit 1	SPKF	EKF1	EKF2
𝑃 for 𝐹𝑒	1.61×10^−4	1.61×10^−4	4.51×10^−4
𝑃 for 𝐿𝑐0	1.81×10^−4	2.47×10^−4	2.9×10^−3
Mean 𝐿𝑐0 (cm)	7.4	4.7	7.5
Deviation 𝐿𝑐0 (cm)	6.12×10^−2	7.67×10^−3	0.34
Average errors (%)	5.7	33.6	6.9
Rabbit 2	SPKF	EKF1	EKF2
𝑃 for 𝐹𝑒	1.72×10^−4	1.61×10^−4	4.53×10^−4
𝑃 for 𝐿𝑐0	8.97×10^−4	3.93×10^−4	2.5×10^−3
Mean 𝐿𝑐0 (cm)	6.8	3.9	10.4
Deviation 𝐿𝑐0 (cm)	8.92×10^−2	3.38×10^−2	6.49×10^−2
Average errors (%)	2.9	44.9	47.9

EKF1 and EKF2 represent EKF estimation with the observation noise covariance 𝑅=2.5×10⁻³ and 𝑅=2.5×10⁻², respectively. 𝑃 represents steady-state error covariance.