Research And Control Of Cable-Parallel Mechanism For Antenna Surface Profile Measurement

The observation efficiency of large antennas is directly affected by the accuracy of its surface profile.Therefore,the measurement and compensation of antenna profile is particularly important.But the existing methods of surface profile measurement can't meet the requirements,such as full-attitude and all-time.From the perspective of “relative motion and passive measurement”,this paper proposes a method that uses the cable-parallel mechanism,as the carrier of the near-field probe,to do the surface profile measurement,and finishes relevant theoretical and experimental research.Firstly,this paper analyzes the similarities between the 1.1m model and the 110 m model based on the similarity theory,for early laboratory researches.And the similarity model and distortion model,being verified by simulations,are established to obtain the parameters' mapping relationship.Secondly,this paper establishes the kinematics and dynamics model of the cable-parallel mechanism,and analyzes its workspace.The cable force distribution optimization is performed using minimum variance and least square norm to avoid the fake-traction form of cables.And the evaluation parameter is proposed to find the optimal working plane.For the 1.1m model,the optimal plane is 810 mm high,with the coverage rate50.08%.Also,this paper makes the mechanism tilted to compensate the angle of antenna,ensuring the mechanism is vertical.The measurable angle ranges from 59.103° to 120.897° with this method.Thirdly,this paper proposes a pre-slack control strategy based on error prediction and compensation.The error model is established,and the neural network is used to predict and compensate the error because of its non-linear characteristics.In order to avoid the impact and discontinuity caused by the error,a pre-slack control strategy is proposed,and with this strategy,you can trade the continuity of the motion by sacrificing some position accuracy.Finally,the pre-slack control system,based on error prediction and compensation,is established with the feedback of binocular vision.The tests on the 1.1m model show that this control strategy can reduce the position error by 80%,and also ensure the continuity of the movement.The experiment of the space trajectory shows that the maximum position error is 2.9730 mm,and the average position error is 1.0158 mm.According to the similarity theory,the average position error of 110 m model is about10.158 cm,reaching the required accuracy.