| With the advancement of the national modernization process,large manipulators are applied in all aspects.This paper takes establishing the dynamic model of the manipulator by using the parameter identification method as the task,and mainly completes the dynamic model of the large manipulator,dynamic parameter identification and six-axis force / torque sensor noise reduction three tasks.Firstly,the dynamic model of the large manipulator is completed.The linear torsion spring is used to describe the joint flexibility.The Stribeck model is used to describes the friction.The assumption mode method is used to describe the flexibility of the arm.The modified DH parameter method and the floating coordinate system are used to complete analysis of kinematics.And then the Kane method was used to establish the dynamic model of the large manipulator at no load,and the established model was applied to the case with the end load by means of the end six-axis force/torque sensor.The forward and inverse dynamics of the manipulator are introduced.Since the dynamic model of the manipulator is a stiff differential equation,this paper uses the 5th-order Gear method to solve it.The numerical simulation of the established model was completed by MATLAB programming method.The sevendegree-of-freedom large manipulator simulation model was established by SolidWorks,ANSYS and Adams software.Comparing the established dynamic model with the simulation model,the established dynamic model was verified successful.Secondly,the model parameters are obtained by means of system identification.The Lagrange method is used to illustrate that the flexible arm dynamics model is a linear model of its dynamic parameters.Then the linear model of the parameters is derived based on the Kane method,and its minimum inertia parameter set is discussed.For robotic arms with joint torque sensors,the joint and arm parameters are identified separately.The identification track selects two tracks of all joints moving at high speed or low speed at the same time.The parameters of each joint are independently identified,and the mathematical model created by MATLAB is fitted.There are two options for the parameter identification of the arm: 1)Identify the stiffness parameters at low speed,and identify the flexible parameters at high speed;2)Identify the stiffness and flexibility parameters at high speed.In order to reduce the parameters of one identification,the identification is made by the method of recursing from the end arm to the base,and the linear method and the nonlinear method are used respectively.The identification effects of different schemes and different methods are compared.Then,in order to improve the accuracy of the model when the robot arm is loaded,a noise reduction method for the six-axis force/torque sensor at the end of the manipulator is proposed,including a search method for finding matching signals and a method for adaptive decomposition level.Since the number of decomposition level directly affects the noise reduction result of the wavelet threshold denoising method,this paper based on the correlation analysis between the six-axis force/torque sensor multi-path signals,and obtains a set of matching signals by means of the proposed search method.Then,by detecting the approximation coefficient of different decomposition level,whether the retention trend of the matching signal is retained or not,and determining the appropriate decomposition level.Finally,related simulations and experiments are completed.Using the simulation model established in the software to generate the parameter identification simulation data,the identification effect of the joint parameters of the robot arm at low speed and high speed running is completed and analyzed,and the identification effect of the different parameters is determined by different identification methods.Experiments are carried out on the proposed six-axis force/torque sensor noise reduction method.The results show that the proposed de-noising method has certain effects. |
PDF Full Text Request