基于L型铁钴的三维力触觉传感器设计及输出特性

Design and Output Characteristics of Three-Dimensional Force Tactile Sensor Based on L-Type Iron and Cobalt

该文提出了一种基于L型铁钴丝的三维力触觉传感器,利用磁致伸缩材料铁钴合金设计制作了三维力触觉传感单元,并测试了该结构的输出特性。基于电磁原理、逆磁致伸缩效应和欧拉伯努利动力学原理,推导了输出电压方程。搭建触觉传感单元实验平台,对传感器进行法向和切向标定,设计的传感器法向测力范围为0~13 N,切向测力范围为0~6 N,输出特性具有良好的线性关系。其中,0~8 N法向力范围内灵敏度约为23.997 mV/N,9~13 N法向力范围内灵敏度约为17.537 mV/N;0~4 N切向力范围内灵敏度约为11.599 mV/N,5~6 N切向力范围内灵敏度约为6.281mV/N。在4Hz频率、2N法向力条件下,该传感器动态响应的输出电压幅值变化范围为46.5~49.8mV,与静态力输出电压幅值的相对误差低于5%,响应时间和恢复时间分别为25 ms和29ms,表明该传感器具有良好的动态特性。分别采用广义逆矩阵和BP神经网络对传感器测得的三维力进行解耦,其中,BP神经网络解耦方法平均误差仅有2.90%,可有效提高测量精度。该传感器体积小、输出线性度好,可同时实现静态力和动态力的测量,在智能机器人等领域具有良好的应用前景。

In today's era,manipulators,as a replacement and function extension for human hands,have been widely used in various industrial fields.In order to meet more complex perception needs,more and more three-dimensional force tactile sensors have emerged.According to the different sensing mechanisms,three-dimensional force sensors can be divided into piezoelectric,piezoresistive,capacitive,electromagnetic,photoelectric and other types.Piezoresistive sensors are measured by the change in resistance of resistance,which has a certain hysteresis and cannot measure the physical quantity in dynamic scenarios well;In the measurement of capacitive sensors,the influence of parasitic capacitance cannot be ignored,and its measurement circuit is complex;Electromagnetic sensors are susceptible to interference from the external environment.As a piezomagnetic material,magnetostrictive material has the characteristics of changing with the magnetization state of the applied material under stress,can respond instantaneously,and the output signal is stable.This gives magnetostrictive materials a great advantage in tactile sensing,enabling fast measurement of dynamic and static forces.Among many magnetic materials,new sensitive magnetostrictive materials represented by iron gallium are widely used in sensors that measure small forces such as displacement,texture and touch due to their high sensitivity and strong linearity.However,iron-gallium alloy has excellent electromagnetic properties in a single fixed orientation,and has certain limitations in multi-dimensional force measurement.In this paper,a three-dimensional force tactile sensor based on L-type iron-cobalt wire is proposed,and a three-dimensional force tactile sensing unit is designed and fabricated by using magnetostrictive material iron-cobalt alloy,and the output characteristics of the structure are tested.Based on the electromagnetic principle,the inverse magnetostrictive effect and the Euler Bernoulli kinetic principle,the output voltage equation is derived.The experimental platform of the tactile sensing unit was built,the normal and tangential calibration of the sensor was carried out,and the normal force measurement range of the designed sensor was 0~13 N,the tangential force measurement range was 0~6 N,and the output characteristics had a good linear relationship.Among them,the sensitivity in the range of 0~8 N normal force is about 23.997 mV/N,and the sensitivity in the range of 9~13 N normal force is about 17.537 mV/N;The sensitivity in the range of 0~4 N tangential force is about 11.599 mV/N,and the sensitivity in the range of 5~6 N tangential force is about 6.281 mV/N.Under the condition of 4 Hz frequency and 2 N normal force,the output voltage amplitude change range of the dynamic response of the sensor is 46.5~49.8 mV,the relative error with the amplitude of the static force output voltage is less than 5%,and the response time and recovery time are 25 ms and 29 ms,respectively,indicating that the sensor has good dynamic characteristics.The generalized inverse matrix and BP neural network are used to decouple the three-dimensional force sensed by the sensor,and the average error of the BP neural network decoupling method is only 2.90%,which can effectively improve the measurement accuracy.The sensor has small size and good output linearity,which can realize the measurement of static force and dynamic force at the same time,and has a good application prospect in intelligent robots and other fields.