Research On Tactile Sensing System For Robotic Soft Grasping

Manipulators are robotic devices that can simulate human arm movements and accomplish object grasping,placing and other actions.With the continuous development of automation technology and artificial intelligence technology,current robotics have a broad application prospect in the medical,warehousing,logistics and manufacturing,which can help people achieve richer work,improve production efficiency and work security,and decrease costs.The traditional manipulator usually grasps objects through position feedback,which cannot precisely control the grasping force and can easily cause object damage or failure when grasping soft and fragile objects.The soft grasping technology allows the manipulator to adapt to more complex environments,such as grasping irregular shapes,smooth surfaces or fragile objects.The key to achieve robotic soft grasping is making the manipulator with tactile sensing function.Currently,researchers have tried to apply the sensing system based on flexible tactile sensors to the manipulator for simulating the soft grasping function of human hands,which can be used to realize the grasping operation of fragile and sliding objects.The paper focuses on a tactile sensing system for robotic soft grasping,designing a flexible piezoelectric tactile sensor with biomimetic cilia structure,proposing a soft grasping strategy based on time-domain standard deviation analysis,and successfully carrying out grasping experiments for fragile and sliding objects and grasping judgment experiments for high temperature objects.The main research components are as follows:（1）From improving the piezoelectric properties of piezoelectric material polyvinylidene fluoride（Poly（vinylidene fluoride）,PVDF）,composite nanofiber membranes were prepared based on electrospinning technology and composite Fe₃O₄nanoparticles.Through characterization tests and piezoelectric performance test experiments,it was confirmed that the composite nanoparticles effectively improved the piezoelectric performance of PVDF.（2）A flexible PVDF tactile sensor with biomimetic cilia structure was prepared to effectively enhance the sensor performance.A five-stage unit signal conditioning circuit was designed for the characteristics of the charge signal from the sensor.Through comparative testing experiments,it was demonstrated that the designed cilia-type sensor had the highest piezoelectric output,while stability tests were conducted to show that the sensor can output stable electrical signals under at least 30,000 cycles of vibration tests,with strong robustness.Experiments were conducted to identify the surface texture roughness and thermal sensing of objects,and the results showed that the sensor can effectively identify the surface property information of objects.（3）A two-finger soft grasping system was designed for the demand of robotic soft grasping.The soft grasping strategies based on time-domain standard deviation analysis and frequency-domain FFT analysis were proposed by analyzing the characteristics of tactile,slip and thermal signals.With the practical,based on the soft grasping strategy of time-domain standard deviation analysis,the soft grasping experiments for different types of objects were successfully conducted.The results indicated that the soft grasping strategy based on time-domain standard deviation analysis was feasible.This paper investigates the tactile sensing system for robotic soft grasping,which provides reference value for the research of realizing machine tactile perception and tactile information recognition,and also has certain engineering practical significance for the research of robotic soft grasping system.