Prediction And Analysis Of Shaking Of Outdoor Platform And Optimization Of Feedforward Compensation Trajectory Of Manipulator

The live operation in the distribution network is an important way to improve the reliability of power supply,reduce the loss of power outage,and ensure the safety of the power grid.At present,the live operation in China is still dominated by the manual live operation.The use of live robots to replace part of the manual live operation can keep the operator away from the dangerous working environment and has become a new trend in the development of the modern operation of the smart grid.In live operation,the connection of the drainage line is an important part.At present,due to the limitation of its mechanical structure and environmental interference,the working platform shakes irregularly during the movement of the manipulator,which makes the positioning error of the manipulator end reach centimeter level.That does not meet the safety standard of the robot.At the same time,the robot has the problems of small working space,unstructured operating environment,and difficult trajectory planning.How to realize the trajectory planning of the manipulator under the shaking of the aerial work platform is of great significance to the popularization and application of the live working robot to complete the work of connecting the drainage line.Based on the state grid science and technology project,aiming at the problem that the positioning error of the manipulator end does not meet the safe operation standard due to the shaking of the operation platform during the outdoor 10 k V high-altitude live connection drainage line operation,this paper studies the trajectory prediction of the operation platform shaking and the trajectory optimization of the manipulator and designs an experimental system to verify the accuracy of the prediction model and the feasibility of the trajectory optimization.This paper mainly completed the following work:(1)Shaking analysis and trajectory prediction of outdoor aerial platformFirstly,a qualitative analysis of the site environment and the working platform of a substation is made.The scheme of the working platform shaking test is designed and the data is collected.Then,the collected data are cleaned and de-noised and transformed into the shaking displacement track data of the working platform.Finally,the LSTM(Long Short Term Memory)neural network model is trained by using the time series data of platform shaking track,and the accuracy of the model position prediction is verified by the test data.(2)Feedforward compensation trajectory optimization of manipulator based on shaking prediction.The D-H parameter method is used to establish the kinematic model of the TR910 manipulator,and the algebraic analytic method is used to complete the derivation of the inverse kinematics formula.By the analysis of the influence of the shaking of the working platform on the end positioning error of the manipulator,the feedforward compensation is used to correct the target point.Then the linear trajectory planning with parabola transition in Cartesian space and the cubic spline interpolation in joint space are combined to achieve the trajectory planning of the manipulator.At the same time,in order to reduce the joint loss of the manipulator,this paper takes the jerk-minimum as the optimization goal and uses a genetic algorithm to optimize the trajectory of the manipulator.(3)The design of the whole experimental system and the verification of the track compensation optimization method based on the prediction of shaking.According to the field investigation,the design requirements of the experiment system are summarized.Then,the whole experiment system is built and the operation process of the simulated drain line is established.Then,the shaking simulation of the working platform and the prediction of track position are completed on the experimental system,which verifies the accuracy of the prediction model.Finally,the simulated operation of the connecting drain line is completed,which verifies that the optimized trajectory after feedforward compensation correction under the shaking of the aerial platform can meet the operation requirements.In this paper,the prediction of the shaking trajectory of the aerial platform and the feedforward compensation trajectory optimization based on the model prediction are completed.And a set of the experimental system is built to simulate the shaking of the platform and the operation of connecting the drainage line,which verifies the accuracy of the prediction model and the effectiveness of the trajectory optimization of the manipulator.These works provide a theoretical basis for the practical engineering application.