	US 11,680,863 B2
	Method for reducing the hysteresis error and the high frequency noise error of capacitive tactile sensors
Aiguo Song, Nanjing (CN); Shuyan Yang, Nanjing (CN); Baoguo Xu, Nanjing (CN); Huijun Li, Nanjing (CN); Hong Zeng, Nanjing (CN); and Lifeng Zhu, Nanjing (CN)
Assigned to SOUTHEAST UNIVERSITY, Nanjing (CN)
Appl. No. 17/288,534
Filed by SOUTHEAST UNIVERSITY, Nanjing (CN)
PCT Filed Jul. 1, 2020, PCT No. PCT/CN2020/099648 § 371(c)(1), (2) Date Apr. 25, 2021, PCT Pub. No. WO2021/253505, PCT Pub. Date Dec. 23, 2021.
Claims priority of application No. 202010564182.0 (CN), filed on Jun. 19, 2020.
Prior Publication US 2022/0107237 A1, Apr. 7, 2022
Int. Cl. G01L 25/00 (2006.01); G01L 1/14 (2006.01); G06F 3/041 (2006.01); G06F 3/044 (2006.01); G01L 9/12 (2006.01)

CPC G01L 25/00 (2013.01) [G01L 1/142 (2013.01); G01L 9/12 (2013.01); G06F 3/044 (2013.01); G06F 3/0418 (2013.01)]

5 Claims

1. A method for reducing a hysteresis error and a high frequency noise error of capacitive tactile sensors, comprising the following steps:

step 1: a calibration, specifically comprising the following steps:

step 1-1: a positive stroke calibration: successively applying m standard loads to a capacitive tactile sensor, wherein the m standard loads gradually increase at a specified step size; forming a loading measurement point each time a standard load is applied; recording a first corresponding standard load value and a first capacitance value at each loading measurement point; and after a loading is completed, fitting first standard load values and first capacitance values of m loading measurement points to form a positive stroke curve;

step 1-2: a negative stroke calibration: successively applying the m standard loads to the capacitive tactile sensor, wherein the m standard loads gradually decrease at the specified step size; forming an unloading measurement point each time the standard load is applied; recording a second corresponding standard load value and a second capacitance value at each unloading measurement point; and after an unloading is completed, fitting second standard load values and second capacitance values of the m unloading measurement points to form a negative stroke curve; and

step 1-3: repeating step 1-1 and step 1-2 n times, wherein n≥3, and n positive stroke curves and n negative stroke curves are formed;

step 2: an averaging, comprising the following steps:

step 2-1: a positive stroke averaging: calculating a first arithmetic average value of n first capacitance values corresponding to the n positive stroke curves formed in step 1 at the each loading measurement point, and recording the first arithmetic average value of each capacitance as a positive stroke average capacitance; obtaining m positive stroke average capacitances at the m loading measurement points; and forming an average positive stroke curve according to the first standard load values of the m loading measurement points and the m positive stroke average capacitances;

step 2-2: a negative stroke averaging: calculating a second arithmetic average value of n second capacitance values corresponding to the n negative stroke curves formed in step 1 at the each unloading measurement point, and recording the second arithmetic average value of each capacitance as a negative stroke average capacitance; obtaining m negative stroke average capacitances at the m unloading measurement points; and forming an average negative stroke curve according to the second standard load values of the m unloading measurement points and the m negative stroke average capacitances;

step 2-3: a comprehensive averaging: calculating a third arithmetic average value of the n positive stroke capacitance values and the n negative stroke capacitance values at a certain standard loading measurement point, and recording the third arithmetic average value as a comprehensive stroke average capacitance; obtaining m comprehensive stroke average capacitances at m standard loading measurement points; and forming a comprehensive stroke curve according to the m standard loads and the m comprehensive stroke average capacitances;

step 3: a fitting modeling: separately performing a least square fitting on the average positive stroke curve, the average negative stroke curve, and the comprehensive stroke curve formed in step 2, to obtain a positive stroke fitting function F_p, a negative stroke fitting function F_n, and a comprehensive fitting function F_ave;

step 4: a measurement: using the capacitive tactile sensor calibrated in step 1 to perform an actual measurement, to obtain several measurement capacitance data;

step 5: a noise filtering: performing a weighted average filtering on the measurement capacitance data obtained in step 4, to obtain a filtered capacitance value;

step 6: a stroke direction discrimination: if it is a continuous measurement mode in step 4, comparing the filtered capacitance value at a current time with the filtered capacitance value at a previous time, to determine whether a current measurement mode is a loading stroke or an unloading stroke; or if it is a single measurement mode in step 4, directly performing step 7; and

step 7: a resolving: if a discrimination result in step 6 is the loading stroke, substituting the filtered capacitance value at the current time into the positive stroke fitting function F_pin step 3 to obtain the a force at the current time; if the discrimination result is the unloading stroke, substituting the filtered capacitance value at the current time into the negative stroke fitting function F_n, in step 3 to obtain the force at the current time; or if the discrimination result is the single measurement mode, substituting the filtered capacitance value at the current time into the comprehensive fitting function F_ave; in step 3 to obtain the force at the current time.