Research On Flexible Control Method Of Mechanical Claw Based On Non-Contact Sensing

Grasping objects is one of the most common tasks in robot applications.When robots grip precise and tiny electronic components,the accuracy of force control is extremely high.Traditional mechanical claws are constrained by limited local space,which makes it difficult to install force sensors,resulting in high cost and insufficient precision of force control.This paper focuses on the force control of grabbing objects using a mechanical claw and proposes a flexible control method based on non-contact sensing.By finely perceiving and processing the control current,the precise control of the acting force is achieved.The main research contents are as follows.Firstly,To ensure the reliable operation of the system,it is necessary to identify,eliminate,or tolerate measurement information.This paper proposes an LMS adaptive filtering algorithm with variable step size factor updated weights,which filters current data and makes the signal smoother and more continuous.By conducting experiments with the mechanical claw grasping system and measuring the variation data of current and force,the function relationship model between current and force is fitted,which achieves non-contact precise perception of gripping force.Secondly,due to the diversity of physical properties of grasped objects,such as small size,easy deformation,and fragility,dynamic and precise control of gripping force is necessary during the grasping process.The force control algorithm is adopted for object grasping,however,PID control suffers from problems like slow response speed,poor stability,and large overshoot.This study proposes a method of gray prediction-based PID control,which improves gray prediction by updating effective values through an isometric resurgent model.Compared with the PID control method,gray prediction-based PID control has better stability,speed,and accuracy.Finally,to realize the design and verification of the above-mentioned algorithms in this study,a mechanical claw grasping system is designed,consisting of an actuator,controller,and host computer.The controller is designed using the ARM series MCU with high-precision closed-loop control function,which can communicate with the host computer or other controllers through interfaces to achieve remote control and data exchange.When communicating data with the controller,the host computer can configure the control parameters and draw data into graphs.The experimental results demonstrate that the algorithm proposed in this paper can achieve high response and high precision control of force.The maximum gripping force is 18.74 N,with a force gripping error within 0.1N and position accuracy within 1 millimeter.Additionally,the device has the characteristics of small size,easy carrying,and high cost-effectiveness,which makes it significant for actual applications of mechanical claws in the industrial automation field.