Real-time collision early warning system and method for mechanical arm in tunnel environment

Abstract

The invention provides a tunnel environment mechanical arm real-time collision early warning system and method.The tunnel environment mechanical arm real-time collision early warning method comprises the steps that firstly, tunnel three-dimensional space positioning information and tunnel circumferential contour surface information are obtained through a positioning system and a digital tunnel design module, and then joint variables when a mechanical arm reaches an expected pose are obtained through calculation of an inverse kinematics module; then, real-time poses of all joints of the mechanical arm relative to a reference coordinate system are obtained through calculation of a forward kinematics module, and then in a collision detection and early warning module, a model is established, and all parameters are set; a hierarchical bounding box detection algorithm and a separation axis theorem are adopted to realize collision detection and multi-level early warning between the mechanical arms, between the mechanical arms and a hydraulic pipeline and between the mechanical arms and a tunnel environment, and finally an alarm module receives an early warning instruction and sends an early warning signal to the outside. The method is suitable for real-time collision early warning of engineering equipment, the detection precision is high, the consideration condition is comprehensive, and safe and efficient operation in the tunnel construction process can be achieved.

Description

technical field [0001] The invention relates to the technical field of engineering machinery, in particular to a real-time collision warning system for a manipulator in a tunnel environment, and a real-time collision warning method for a manipulator in a tunnel environment based on the system. Background technique [0002] When industrial robots perform multi-machine and multi-arm collaborative operations in the limited space of the tunnel, collisions are prone to occur between the robotic arm and the robotic arm, the hydraulic pipelines suspended from the robotic arm and the robotic arm, and the robotic arm and the tunnel environment, seriously affecting construction safety and construction safety. Efficiency, so real-time collision warning is required. [0003] The collision warning of existing industrial robots is mainly realized by adding external detection devices such as torque sensors or vision systems to realize hardware design, and realize software design through hi...

AI Technical Summary

Problems solved by technology

[0004] 1. In terms of hardware, the implementation method is too complicated, the cost investment increases, and the collision detection method based on the torque needs to estimate the joint friction force, which is affected by many factors such as the robot pose, speed, temperature, and grease condition. It is difficult to accurately model and identify, and the accuracy is poor, so it is not suitable for hydraulic heavy-duty manipulators in tunnel operations

[0005] 2. In terms of software, the division of the scheduled early warning area must be reasonable. If the division is rough, the detection accuracy will be too low. If the division is too fine, the complexity of the algorithm will be increased. The current division of the scheduled early warning area is not very good. Taking both into consideration, the dimensionality reduction processing method adopted not only simplifies the complexity of the algorithm, but also reduces the detection accuracy.

[0006] 3. The existing collision warning system or anti-collision system can only artificially evaluate the possibility of collision accidents in different areas in advance by dividing the predetermined warning area, and then detect the specific collision accident that is about to occur in real time, which cannot be achieved in real time. Early warning, not to mention the level of early warning

[0007] In addition, the existing collision warning system only provides early warning for the collision between the robot arm and the robot arm or the collision between the robot arm and the obstacles in the work site, and lacks the early warning for the collision between the robot arm and the hydraulic pipeline suspended under the robot arm, and the comprehensive consideration is not comprehensive enough.

Images