Intelligent Sensing System Design Based On Flexible Tactile Sensor

With the development and popularization of flexible electronics,flexible tactile sensors have gradually become important components in wearable electronic devices and intelligent robots.By converting external stimuli into measurable signals,flexible tactile sensors can enable friendly interactions between machines and humans,as well as between machines themselves.As electronic devices that can simulate the human skin’s ability to perceive external stimuli,flexible tactile sensors can perform similar functions,such as monitoring pressure,temperature,and other signals,with the expectation of demonstrating capabilities that are comparable to or even surpass human perception.However,at present,there are still many challenges in achieving non-interfering collection of multimodal signals and high-performance flexible tactile sensors.Furthermore,with the continuous development of artificial intelligence(AI),combining flexible tactile sensors with AI to build more intelligent sensing systems will strongly promote the development of future robot technologies and intelligent terminals.This dissertation achieves the construction of biomimetic flexible single mode capacitive tactile sensors and dual mode flexible temperature pressure tactile sensors without signal crosstalk through reasonable structural design,and combines artificial intelligence learning algorithms to construct an intelligent perception system for wearable devices.The main work includes:(1)Supported by a simple and economical three-step low-temperature hydrothermal growth method,a biomimetic flexible tactile sensor inspired by the multi-layered structure of the thorn of bluegrass is prepared,and its application prospects in intelligent sensing are demonstrated.Thanks to the 3D layered structure and biomimetic flexibility of the capacitive sensing,this sensor exhibited highly integrated sensing characteristics,including high sensitivity,wide pressure sensing range,fast response/recovery time,high pressure resolution,and excellent long-term durability,all of which are essential for meeting the evolving demands of intelligent sensing technology.In addition,the application of 3D layered biomimetic flexible capacitive pressure sensors in intelligent sensing was explored,and a robot hand with a 3D layered biomimetic flexible capacitive pressure sensor was constructed to imitate the pulse diagnosis of traditional Chinese medicine doctors.Furthermore,a high-resolution flexible sensing array was established for spatial pressure mapping on flat/curved surfaces and effective perception of Braille characters.(2)In order to realize the true application of flexible tactile sensors in daily life and production,a dual-mode flexible tactile sensor with temperature-pressure decoupling capability is designed to address the issue of signal crosstalk between temperature and pressure during monitoring.For the temperature sensing functional layer,it is fabricated by coating polyurethane acrylate doped with PS hedgehog-shaped microspheres,on the interdigital electrode and curing in situ using light,which prevents the temperature sensing part from being affected by pressure.A supercapacitive ionic pressure sensing functional layer,is obtained by doping Poly(vinylidene fluoride-co-hexafluoropropylene)(PVDF-HFP)with an ionic liquid.It is noteworthy that this pressure-sensitive material is not sensitive to temperature,thus avoiding crosstalk caused by temperature on the pressure sensing part.Differing from the simple process of stacking structures used in the past,the stacking structure of the temperature sensing functional layer and the pressure sensing functional layer is implemented using a flexible printed circuit board(FPCB),which strongly supports the stability of the dual-mode sensor.It is shown that the dual-mode flexible tactile sensor efficiently achieves temperature and pressure crosstalk-free detection,which provides guidance for the development of low-cost and structurally simple crosstalk-free multimodal tactile sensors in the future.Additionally,due to the electronic tunneling effect induced by the addition of PS micro/nano-particles with a hedgehog-shaped structure and the formation of the supercapacitive ionic capacitor effect caused by doping with the ionic liquid,the dual-mode tactile sensor has high sensitivity,a wide operating range,short response time,and excellent durability while achieving crosstalk-free signal acquisition.(3)Based on the 3D layered structure of biomimetic flexible capacitive pressure sensors and the multilayer perceptron neural network algorithm,an intelligent glove for sign language gesture recognition and robot interaction system was designed.The system consists of a frontend module and a back-end module.The front-end module includes a 3D layered structure of biomimetic flexible capacitive pressure sensors for finger motion capture and a printed circuit board for signal conditioning,processing,and wireless transmission.The back-end module includes a server for signal judgment,inference,and recognition.At the start of the system,the capacitance data corresponding to the maximum and minimum bending is collected and normalized.During system operation,the 5-channel capacitance signals are non-linearly converted into voltage signals by a capacitance-to-digital converter,and then filtered to remove interference signals and environmental noise.Subsequently,the signals are further processed by an analog-to-digital converter,and then transmitted to the server via WIFI communication and TCP.The server analyzes and processes the received signals to achieve accurate recognition and interaction of 7 sign language gestures.The experimental results show that the average accuracy of the 7 sign language gestures can reach 98.33%.(4)By combining the flexible dual-mode temperature-pressure sensor,mechanical hand is endowed with the ability to recognize pressure and temperature in space.A 16-channel capacitive and resistive high-speed synchronous acquisition system is designed,which to detect the capacitance and resistance changes of 16 flexible dual-mode temperature-pressure sensors in real time.At the same time,the algorithm of convolutional neural network is applied to achieve the intelligent perception ability of the shape and softness of objects by the mechanical hand.Finally,by imitating the human hand’s retracting reaction,the mechanical hand is given the ability to protect itself,like a human body,from damage caused by high temperature.Specifically,when the mechanical hand grasps an object with a temperature that is too high,it automatically retracts to protect the mechanical hand and the flexible sensor from damage caused by high temperature.