Research On Automatic Docking Connector Signal Detection System

In the process of propellant filling,the tower structure is opened,so the rocket would be swing under the wind.This would make some troubles for the joint and follow-up of the connector and destroy the structure of the connector.Therefore,in order to reduce the acting force of rocket and enhance the reliability,this paper studies a signal detecting system based on monocular vision that achieves the automatic joint and initiative follow-up of the connector by detecting the specific location information of rocket filler.The hardware part and software part of signal detection system are designed and studied in this paper.In the part of hardware,set up a platform,connect vidicon with DSP by data collection channel,and load programs written in the upper computer to DSP by emulator.Three-dimensional informations of rocket filler can be gotten after image processe by DSP,and images are showed on the displayer.In the part of software,CCS3.3 is used as the exploitation environment,Zhang calibration method is used as calibrate vidicon.Firstly,make pretreatment for images and get centroid by centroiding algorithm,and then gain the two-dimensional coordinate of launcher vehicle in the image plan.Finally,use the theory of pin-hole imaging to recover depth,and get the distance between camera and rocket,then the coordinate in another direction can be defined.There is a signal detecting experiment in the laboratory.The experimental results indicate that the signal detecting system based on monocular vision can detect the specific location of rocket filler betterly.It is practical and feasible,and provides technical support for automatic docking of connector.