Research On Visual Location System Of Intelligent Cleaning Robot For Large Condenser

Condenser is one of the key equipments in thermal power plant. It is the cold source of the thermodynamic cycle of the large steam turbine set. It can decrease the exhaust pressure and temperature of the steam turbine set, and enhance the cycle thermal efficiency of the turbo generator set. For a variety of reasons the fouling not favorable for heat exchange is accumulated in the inner wall of condenser tube. These fouling will bring about a series of problems. Based on current situation on method of cleaning of condenser, we investigate intelligent automated cleaning system for large condenser based on a mobile manipulator. Takes the condenser cleaning robot as main study object, designed a new scheme for condenser fouling cleaning in power plants, visual servo system were well designed and explanted.The necessity of fouling cleaning is explained by investigating the cause of fouling accumulation and the influence on heat transfer performance of condenser. Also, the advantages and disadvantages of the ordinary fouling cleaning methods are analyzed; on these bases we designed a new scheme for condenser fouling cleaning in power plants. According to the practical working environment of intelligent cleaning robot, we designed a new type cleaning robot which has two-joint manipulator arm and the track mobile robot body.System instructure including track mobile robot body,manipulator arm and high pressure water jet cleaning subsystem is discussed in detail.Visual servo system was analyzed in detail from two aspects of hardware and software.According to the robot works in the harsh conditions, visual position system is designed based on PC. Enhancing the contrast of images is obtained by several combined lamps.this paper discussed the problems that the designing of visual servo system often encountered, such as the light sources,the selecti-on of CCD camera,the selection of image acquisition card.Merits and shortcomings of Various Image Pre-processing methods are compared in the process of condenser tube position, and for getting the optimal effect of binarization arithmetics of condenser tube by the maximum variance between various classifications and then realize the edge detecting. The coordinates of all the tube are obtained by least square method, then transits the image pixel logistic coordinates to space coordinates,finally drives the spray gun to the position of the condensers tube to realize the automatic position.For realizing to recover the condenser tube 3d shape information from 2d images, the camera calibration process is introduced, including camera model,camera intrinsic matrix parameters and the camera external parameters. And realized the cleaning robot`s camera calibration.