Spatial Position-pose Analysis And Experimental Validation Of Rigid-flexible Coupling Robot For Steel Structure Building Inspection

To overcome the problems of blind zone and incomplete detection in traditional health monitoring of steel structure building,a magnetic adsorption wheel robot was used for steel structure health monitoring.Flexible structure detection robot adopts deformable and bendable flexible steel belt to connect the front and rear bodies,so it can over all kinds of obstacles flexibly.But this rigid-flexible coupling structure also brings great difficulties to real-time pose estimation and control.How to obtain high precision pose parameters is an important premise and foundation for flexible robot control.In order to obtain more accurate estimates of robot pose parameters,this paper takes a magnetic adsorption type flexible structure detection robot as an example,based on two improved solution filtering algorithms to solve spatial attitude of robot with rigid-flexible coupling structure under different conditions,and select the optimal algorithm for solving the flexible robot attitude as the input parameters of pose estimation.The work and results of this study are as follows:First of all,through the installation of the body before and after the robot with the encoder inertial measurement unit of multi sensor information fusion,access to flexible inspection robot before and after the vehicle real-time dynamic posture parameters before and after the establishment of the rigid flexible coupling structure of the robot body displacement and attitude kinematics model,perception due to the flexible steel belt deformation leads to the relative change of body position before and after the robot.On this basis,according to the motion characteristics of the flexible robot,the four element algorithm is selected as the updating method of vehicle body posture,and the vehicle attitude update and the initial attitude angle estimation are realized.Secondly,preprocessing the original data of MEMS inertial measurement unit before and after the robot is processed,and the preprocessed data are input into the explicit complementary filter as input parameters,and the three sensor data are fused to calculate the front and rear body posture of the robot.In order to improve the estimation accuracy of robot's attitude parameters,an improved complementary filtering algorithm based on explicit complementary filtering algorithm and first-order complementary filtering algorithm is proposed.In the Matlab environment,the algorithm of the two kinds of complementary filtering algorithm for the flexible robot body's attitude is simulated in two cases of slow and fast speed.The algorithm has the advantages of fast convergence and high accuracy.Then,according to the attitude of flexible robot system actual movement process is generally nonlinear problem,using the extended Kalman filter algorithm of robot flexible rigid flexible coupling structure of space attitude,the algorithm uses the Taylor series linearization of nonlinear system,can provide high precision robot pose estimation.On this basis,puts forward a new method for solving the extended attitude fusion filtering algorithm Kalman filtering algorithm and explicit complementary filter,the algorithm will be four yuan of attitude accelerometer and magnetometer data obtained by compensating the gyro as state variables,four complementary dominant element solution is to filter the attitude as the measuring value of the extended Kalman filter model the.The dynamic tracking performance of the improved algorithm is better,and the attitude angle estimation precision is higher.Finally,the modified explicit complementary filtering algorithm and the improved extended Kalman filter algorithm are used to solve the static and dynamic attitudes of the front and rear vehicle body in different working conditions of the flexible robot respectively.The position of the robot is determined by using the path estimation algorithm and obtained through data fusion.Space Position of Rigid and Flexible Coupled Structures for Flexible Robots.Analyze and compare the attitude angle errors of the two improved filtering algorithms,compare the accuracy of the algorithm's attitude determination,calculation time and dynamic tracking performance and other indicators.The experimental results show that the angle parameters obtained by the extended Kalman filter algorithm have higher accuracy and better dynamic tracking performance.Finally,the extended Kalman filter algorithm is used to solve the posture of the flexible detector,which provides the precise spatial pose parameters for the flexible detection robot in the obstacle-resilient movement of the complex building structure,so as to ensure that the flexible robot successfully completes the damage detection task of the steel structure.