Design Of Magnetostrictive Tactile Sensor And Surface Texture Recognition Of Object

Touch plays an important role in human perception system.It can perceive the shape,texture and other important attributes of objects.However,tactile technology is relatively backward compared with visual technology,so tactile technology has become a hot research direction.Especially in robotic perception,surface texture detection and minimally invasive surgery,researchers have tried to use a variety of technologies and materials to produce sensitive,reliable,flexible,small and low-cost tactile sensors.In this paper,aiming at the problem of robot texture identification,based on the inverse magnetostrictive effect of the New Magnetostrictive material,the tactile sensor is designed,analyzed and experimentally studied,and texture tactile sensor,texture and friction tactile sensor are made.Firstly,based on the principle of biological perception,combined with the biological perception mechanism model and the characteristics of magnetostrictive materials,texture tactile sensor and texture and friction tactile sensor are designed.Based on the linear constitutive model of magnetostrictive materials,the texture tactile sensor is analyzed and the relationship between texture features and output signals is obtained.Based on the Armstrong model,the model of the beam which detects the friction force by texture and friction tactile sensor is analyzed,and the relationship between the friction force and the output signal is obtained.Secondly,a prototype of texture tactile sensor is fabricated,and the whole texture tactile sensor is assembled in a knuckle using 3D printing technology.Five kinds of fabrics which are most well known in our life are selected as experimental samples.Subjective tactile perception experiments are carried out on each sample.Roughness-smoothness and sparseness-fineness of each sample are scored according to the results of subjective perception.An experimental platform with the function of texture signal extraction was built,and the output characteristics of the sensor were tested.The common fabric samples in daily life were tested,and the performance of the sensor was verified.Finally,four eigenvalues representing texture rough-smooth and sparse-fine attributes are extracted from the output signal.The rough-smooth and sparse-fine attributes of each sample are identified by the relationship of eigenvalues,and compared with the results of subjective tactile perception.PSD eigenvalues are extracted from texture vibration signals obtained by tactile sensors,and different experimental samples are classified and recognized by ELM algorithm.