Wyniki wyszukiwania - BazTech

Ograniczanie wyników

1 Metrology and Measurement Systems

1 2025

Znaleziono wyników: 1

Liczba wyników na stronie

Wyniki wyszukiwania

Sortuj według:

Ogranicz wyniki do:

Kernel extreme learning machine combined with gray wolf optimization for temperature compensation in pressure sensors

Wang Huan, Wu Ting, Liu Pan, Zou Yijun, Zeng Qinghua

Metrology and Measurement Systems

2025

Vol. 32, nr 1

1--14

In order to improve the measurement accuracy of pressure sensors, a method based on gray wolf optimization (GWO) to optimize kernel extreme learning machine (KELM) is proposed to address the problem of nonlinear drift that can be easily affected by temperature in the working environment. Firstly, the fast search capability of the GWO algorithm is used to find optimal regularization coefficients and kernel function parameters of the KELM algorithm; secondly, the random mapping of the traditional ELM algorithm is replaced by the kernel mapping of the KELM algorithm to improve the generalization and stability degradation brought by the random assignments. Finally, the voltage signal values under different temperature and pressure environments are obtained through calibration experiments and compensated by the GWO-KELM algorithm. The results show that the GWO-KELM method has a better compensation effect compared with the traditional BP neural network with a full-scale error of 0.13% (FS), the ELM algorithm with a full-scale error of 0.12%FS, and the KELM algorithm with a full-scale error of 0.12% in the range of 0 to 700 kPa absolute pressure and -40˚ to 70˚. The full-scale error is only 0.07% and the maximum absolute error is as low as 0.5446 kPa, which improves the accuracy index by one order of magnitude.