Translated Abstract
Delta robot is a kind of high-speed, light payload parallel robot. They are usually guided by their vision systems and use their end effectors which are determined by three servo axes to locate and transport the target objects. Because of the characteristics, like light weight, high speed, positioning accuracy, low cost and high efficiency, Delta robots have been widely used in electrical, food and medical areas to do processing and assembly work. The Delta mechanism was invented by Dr. Clavel. And after its protection patent expiring in 2009, people in this sector began their development in Delta robot. The enterprises have pushed out their own Delta robots. Colleges and institutes have started their research and achieved some pleasant results. However, these robots are designed for light payload and for heavy payload, there is no complete solution published. Our research group has done much research work on mechanism design and kinematics analysis. The newly-designed mechanims is capable of holding objects up to 50 kg. The main topic of this thesis is about control algorithms. A complete set of control algorithms is developed. The feasibility and effectiveness of the control algorithms were proved by our Delta robot system built in our lab. The main contents and innovative points are as follows:
1. Firstly, the trajectory planning algorithm is completed which is composed of three parts: moving modes, velocity profile and door-like path. In this dissertation, the definitions and algorithms of the three parts are developed and presented. The moving modes include continuous path (CP) mode and point to point (PTP) mode. The two moving modes ensure that Delta robot can maximize its high speed advantage. For the velocity profile, a trapezoidal velocity profile (TVP) is used. The velocity is calculated at every moment. The door-like path is used for the robot to do repetitive pick and place objects operations. In this dissertation, we interpolate some points in the door-like path to make it move more stably and faster.
2. Secondly, the stability control system is designed. The control system includes an inner layer and an outer one. The inner layer is actually a PI controller which adjusts the angular velocities of the motors, while the outer layer is a PID controller which is for the angles of the motors. For the inner layer, a set of methods are deveopled to adjust motor parameters. For the PID controller, a nonlinear integrator is used to make the controller more accurate. Then a feedforward controller is designed to reduce the lag effect. A gravity compensator is added to reduce the positioning error.
3. In addition, to make the communication between the host computer and the Delta robot possible, a data transmission system based on UDP protocol was designed. In order to guarantee the reliability of communication, the CRC, BCD code and the cyclic sequence number are applied. Besides, we define the transmission data records. There are two kinds of transmission data records, one is data requirement record and the other is data downloading record. The former is used for the host computer to real-time monitoring the motion state of the robot. The latter is used for the host computer to control the robot to complete some specific tasks.
In this dissertation, a feasible solution for the design and development of Delta robot is presented. It is the basis of the future work in vision feedback system and is of some reference value for other parallel mechanisms.
Corresponding authors email