Large-area coverage electronic skin system with long-distance approach sense

Abstract

The invention relates to a large-area coverage electronic skin system with a long-distance approach sense. The large-area coverage electronic skin system comprises a basic perception unit and a central unit, wherein the basic perception unit comprises a plurality of perception modules and a processing module; the perception modules are in mutual expansion couplet; each perception module comprises a contact force sense detection module and a long-distance proximity sense perception module; and the central unit is connected with the basic perception unit and is used for acquiring perception data of the basic perception unit in real time, monitoring the working state of the basic perception unit in real time and processing the acquired perception data. By fusing remote proximity sense and contact force sense perception information for activating the perception unit in the electronic skin system and fusing the robot body sense, the proximity object space position, the proximity speed and direction, and the shape and size are estimated; by fusing robot motion planning path information, a collision probability is predicted; and by performing calculation through a collision identification deep learning neural network model, the collision scene category and the collision risk are estimated.

Description

technical field [0001] The invention specifically relates to a large-area coverage electronic skin system with long-distance proximity sense, which is mainly aimed at the collection and processing of tactile information and long-distance proximity information of electronic skin parts that need large-area coverage by a new type of human-machine collaborative robot. Background technique [0002] In the industrial field, many tasks that require complex cognition and flexible handling of emergencies can only be completed by robots and humans, which requires robots and humans to coexist in the same workspace. In the field of life services, it is even more difficult to completely isolate robots from humans. Therefore, how to enable robots and humans to cooperate safely and harmoniously in the same space is an important direction for the development of robots. Among them, the detection of possible collisions between robots and humans, as well as between robots and the surrounding ...

AI Technical Summary

Problems solved by technology

At present, a lot of research has been done in this area, and a lot of results have been put forward, but none of them are mature, and few of them have been practically applied.

The main related technologies and defects include: (1) Using robot joint torque force or joint motor current detection to identify collisions, the disadvantages are poor reliability, poor scene adaptability, and high cost of modeling complex sensors; (2) Using machine vision or laser Radar is used to detect collisions, the field of view is small, and it is easily affected by changes in ambient light and occlusion. The algorithm processing is complex, the reliability is not good, the detection accuracy is poor, and it cannot handle micro-distance collisions; (3) With the development of electronic skin technology in recent years, A scheme to detect collisions using electronic skin with a force tactile sensor and a short-range proximity sensor array is proposed. However, because the detection distance is too short, it is often detected only after the collision has occurred, or it is detected when the proximity occurs, but the distance is too large. Short, insufficient response time, unable to achieve timely response to collision avoidance actions

Images