I have seen the videos of robot contest in Asia Pacific, and I think what they have done is amazing. I don't know how they control such high precision.

It's not that accurate. If the robot is allowed to thread the needle, it will definitely not work with the existing technology. In fact, the accuracy you mentioned is not the same as that of industrial robots. Robots participating in Robocon Competition are all made by undergraduates in their spare time (or absent from class). Technically, it does not meet the accuracy requirements of tolerance fit in industrial production, so the accuracy is actually not at the same level.

The charm of Abu Robocon Competition lies in how to complete the electromechanical integration products that meet the use requirements according to the resources provided by the school and using the existing mechanical conditions and various sensors. Robocon's pit friends have an idea: electromechanical complementarity. The essence of electromechanical complementarity is to choose the appropriate mechanical or circuit method under the same requirements. The selection criteria are reliability, weight, accuracy and speed. In Robocon's competition, slow and steady, win the championship with speed. Only when the weight meets the requirements can you enter the competition, and the accuracy is enough to ensure the completion of each serial small task.

In terms of full-field positioning, the positioning mode of gyroscope+encoder (grating encoder+omni-directional wheel-McNamham wheel with small wheel at 90 degrees with big wheel) pioneered by Xi 'an Jiaotong University in 2007 is widely used. If you pay attention to the observation, there were many white lines in the field before 2007. However, in recent years, the area of the white line is gradually shrinking or even disappearing (for example, the site of Abobo Bokang 20 14 is basically distinguished by color). Hehe, give me publicity. . . Off topic. . . Anyway, robots have three degrees of freedom on the field. Assuming that the ground is regarded as the xoy plane, the robot has the freedom to rotate around the Z axis and translate along the X and Y axes. Return the phase of the robot relative to the initial position with the gyroscope (the initial phase is the phase of the starting area of the field). The speed information of the robot can be returned through at least two codewheels. Then, using the knowledge of high school physics supplemented by integration and so on, an algorithm can be made to get the position and posture of the robot on the whole field.

There are two methods for the action area of robot, because the position accuracy of the end effector (usually the gripper) is required to be high. One is mechanical limitation-such as guide wheels and devices with special shapes (such as rings can be positioned with V-grooves), and the other is using sensors-such as ultrasonic, infrared ranging and lidar to correct. Different methods have different meanings of the returned effective values and different correction items-but in general, most of them choose to correct the code disk values. Gyroscope used in the competition is still very expensive, so I won't mention the specific model (I can't remember clearly), and the accuracy is still guaranteed within the specified speed range. With these sensors, basically when the robot moves, it can have great flexibility. Through these sensors, the error caused by the slippage of the code wheel can be compensated, as well as the error between the actual working size and the marked size (including the error that the manufacturing error does not coincide with the benchmark).

In addition, in the process of full-field positioning, you can also choose a suitable path during the trip and take the markers on the site as the reference system to appropriately correct the path. In fact, this is closely related to the scheme of the competition-the successful defending power system corrected the path of the automatic robot (car A) with white lines in the green planet mission of Robocon 20 13. In other words, this is actually. . . It has something to do with the culture of each team. Path planning can profoundly reflect the guiding ideology of a team and the scientific research concept of a school student. Not much to say, that's all.

Salute HITCRT

Xi jiaotong university robot team retired idlers