MEMS 6-axis force sensor chip for spike pins in athletic sports shoes

Takahashi, Hidetoshi; Ishido, Hiroshi; Nakai, Akihito; Takahata, Tomoyuki; Matsumoto, Kiyoshi; Shimoyama, Isao

doi:10.1016/j.sna.2023.114702

Cited by 1 publication

(1 citation statement)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, diaphragm measurements must be performed to avoid the interference of optical signals between different dimensions during structural deformation. Takahashi et al [27] proposed a spike-pin-type MEMS 6-axis force sensor for athletic sports shoes with good application prospects. Therefore, additional types of six-axis force sensors for wearable devices must be developed.…”

Section: Introductionmentioning

confidence: 99%

Development and calibration of large deformation-compliant six-axis force sensor

Huang,

Yin,

et al. 2024

Meas. Sci. Technol.

View full text Add to dashboard Cite

Improving the measurement accuracy and minimising the coupling between directions are the keys to researching the compliant six-axis force sensors.The use of a six-axis force sensor to accurately monitor the ground reaction force (GRF) and centre of pressure (COP) during human motion is of great significance in the fields of biomechanics and pathological gait diagnosis. Although complete force information can be obtained using a commercial six-axis force sensor, its high stiffness affects the natural gait and easily leads to human fatigue. A compliant six-axis force sensor based on a flexible optical waveguide is proposed, in which the force and torque of six dimensions are detected by reasonably arranging six modular sensing units, and the mechanical decoupling of some dimensions is realised in theory. For the interdimensional coupling and error caused by machining process factors, as well as the nonlinear relationship between the input and output of the proposed compliant six-axis force sensor, a DE-RBF decoupling algorithm is proposed to decouple the calibration data. Compared with the least squares method (LSM) and the radial basis function (RBF) neural network decoupling algorithm, the obtained type-I errors were reduced by 87.7629%, 43.6265%, respectively, and type-II errors by 35.3312%, 56.9162%, respectively. The decoupling result's maximum type-I and type-II errors were reduced from 7.7125% and 2.7382% in LSM and 3.1029% and 2.8917% in RBF to 0.5916% and 0.9558%, respectively. The measurement accuracy of the compliant six-axis force sensor was significantly higher; however, the time effectiveness of the proposed DE-RBF decoupling algorithm was slightly lower than that of the RBF neural network by 2.47%. In conclusion, the decoupling accuracy and timeliness of the proposed DE-RBF decoupling algorithm can satisfy the requirements of compliant six-axis force sensors to monitor low-frequency biomechanical signals, such as human motion.

show abstract

Section: Introductionmentioning

confidence: 99%