System Design And Anti-collision Algorithm Research Of SCARA Robot

With the advancement of automation,robot is being incresingly applied to the industrial production.The dissertation analyzes the current development of industrial robots,design a flip SCARA robot with internal wiring.Based on robot dynamics model,a non-sensor safety collision avoidance strategy is presented to solve the security problem of robot and human cooperative work.The parameters identification and signal processing technology are studied in detail.The robot hardware is composed of mechanical system and electric control system.The structure of robot arm and ball screw,the type of reducer in mechanical system are identified.The type of motor,driver and controller in electrical control system are choosed.Starting and protection circuit and circuit layout in electric control box is designed.The model of SCARA robot joint coordinate system is established using D-H model to get inverse and forward kinematics.The dynamics equations are established by Lagrange method and the equation deduced process of the SCARA robot are presented.As joint friction force is an important factors about affecting the robot's motion state,several common joint friction models are disscussed in the thesis.Based on the robot dynamics model,the parameter identification method is used to solve the dynamic parameters and the friction coefficient.The trajectory of the robot's motion is given,the parameters of the robot's motion state and the joint torque signal are obtained by sampling.Therefore,the parameter identification problem is transformed into the mathematical problem about solving the optimal solution of the dynamic parameters.the problem is solved through Commonly used numerical analysis methods,such as,the least square method,neural network method,genetic algorithm and so on.The least square method is considered as an ideal method in terms of calculation precision and convergence speed.When the system parameters and the given trajectory are known,theoretical value of the joint torque can be obtained through the dynamical equation.The actual motor output force is obtained by sampling.When the difference between the theoretical value and sampled value is large,the robot can be considered as in a abnormal state.It may be due to equipment failure or collision.Whichever its reason,measures should be taken to avoid the situation.In the actual situation,the model error,parameter identification error and change of the load will cause calculated error of the joint torque.Therefore,it is needed to eliminate the influence of these factors by the adaptive algorithm.Based on the above theoretical analysis,experimental test is performed.The experimental results suggests that the robot controller can accurately detect the collision of robot with the external environment and the appropriate emergency measures will be taken to avoid the occurrence of the failures.Error of the model will be controled within ?5% and its reaction cycle is 10 ms.Whilst,the joint torque is obtained through sampling current of the motor,which assures personnel safety and the safe operation of the equipment without increasing the hardware cost.