Bio-inspired Model Of Haptic Perception And Its Application In Surgical Robot

The haptic perception model is a mathematical expression that describes the quantitative relationship between subjective perception and objective stimuli.Haptic feedback is a technique that convey tactile sensations of contacting object to human in virtual or teleoperational environments using some specific devices.An accurate model of haptic perception provide a valuable guidance to researching the mechanism of haptic perception and improving the realism of the haptic feedback.In previous work of modelling haptic perception,researchers generally established a direct model to describe the relationship between tactile sensation and parameters of stimuli without considering the inherent neural mechanism of tactile sensation.Previous researches indicate that the tactile sensation in the brain is the result of processing population responses of tactile receptors.Therefore we studied the characteristics of human tactile perception and the population responses of tactile receptors behind tactile perception,and established a bio-inspired model of haptic perception,which includes simulating population responses of tactile receptors to mechanical stimuli and converting neural impulses of population responses to tactile sensations.Then we applied the bio-inspired model of haptic perception in controlling haptic feedback for the surgical robot.In view of the problem that the previous simulation models are low computation efficient in simulating population responses of tactile receptors,we proposed a rapid simulation model consisted of resistance network model and firing model of single receptor to replicate population responses of tactile receptors.Since the human tactile sensations are not only influenced by surface properties of the contacted object,but also affected by the interaction mode(pressing or sliding)between skin and object,we used the height image of the object surface properties and the data(pressure and position signal)of interaction between the fingertip and the object surface to describe the input stimuli of the simulation model.In order to test whether the neural responses predicted by our simulation model are accurate,we acquired height image and interaction data as the inputs according to the stimuli adopted in neurophysiological experiments.The test results indicate that the predicted neural responses to stimuli such as scanning dotted texture,pressing spherical surface match well with neural data measured in neurophysiological experiments,and achieve correlation of more than 0.85 in most of conditions.Besides the simulation model runs real-time with 3000 receptors at sampling rate of 0.3 k Hz.In view of the problem that the previous haptic perception models rarely consider inherent neural mechanism of tactile sensation,we proposed a bio-inspired haptic perception model based on simulating population responses of tactile receptors.The bio-inspired haptic perception model is consisted of simulating the population responses of tactile receptors and converting the population responses into tactile sensations.For any inputs,the bio-inspired perception model can output quantitative values of roughness sensation,pressure sensation and vibration sensation.In order to test whether the tactile sensations predicted by our perception model are accurate,we acquired height image and interaction data as the inputs according to the stimuli adopted in psychophysical experiments.The test results indicate that the haptic perception model achieve correlation of about 0.9 in predicting the roughness sensation to dotted and grating textures,pressure sensation to curvature surface,and vibration sensation to sinusoidal and diharmonic vibration stimuli.In order to improve the realism of haptic feedback for surgical robot,we proposed a bioinspired haptic feedback algorithm to control haptic device of surgical robot based on the model of haptic perception.In the bio-inspired haptic feedback algorithm,the actuator in the haptic device is not controlled by the sensor signal,but the control amount reversed by tactile sensation outputted from our perception model,thereby conveys realistic haptic feedback to the users.In order to evaluate the performance of the bio-inspired haptic feedback algorithm,we constructed an experimental platform of artificial palpation using the Da Vinci surgical robot,and recruited25 subjects to participate this experiment.The experimental results indicate that bio-inspired haptic feedback algorithm increase the localizing accuracy by about 15%,decrease the average applied force by about a half,compared to the linear algorithm and piecewise discrete algorithm.