Research On The Positioning Method Of The Inspection Robot Based On The Offshore Oil Production Platform

In recent years,the sharp decline in international oil prices,resulting in a substantial drop in profits of CNOOC.Compressing investment,optimize resources,reduce costs become the priority measures to cope with the petroleum oil in cold winter.The unmanned oil production platform is a new attempt made by the Shanghai branch under the new situation.The design,construction and operation of the unmanned platform will help improve the utilization rate of production facilities and reduce unnecessary costs of design,construction and development,which can generate the best economic benefits and facilitate the sustainable development of the company.At the same time,there are many problems in the unmanned platform.For example,there are a number of blind spots in fixed monitoring points,which cannot accurately and clearly understand the specific conditions of the site.The existing central control system can only intervene in production systems,and cannot realize remote recovery for power systems.Combined with the existing control system and CCTV system,an independent inspection robot system is introduced to make up the above deficiencies and reduce the frequency of the personnel on the platform.The problem of locating the inspection robot on the oil production platform is the basis of the inspection robot’s successful completion of other inspection tasks.This paper focuses on the research on the location of the inspection robot using the environmental characteristics of the oil production platform.The main contents are as follows:1.According to the platform inspection robot system modeling and experimental modeling environment,based on the "Haibao" inspection robot system based on the experimental platform,establish movement model of oil production platform inspection robot system and determining oil production platform inspection robot system control input model,determine the expression form of offshore oil production platform unmanned environment characteristic,construction of sensor observation model.The uncertainties and errors arising from the modeling of platform inspection robot system are analyzed.The environmental noise model and sensor noise model are established,and the experimental need model is established,which lays the foundation for the positioning research of platform inspection robot.2.The study of oil production platform azimuth feature extraction and appropriate representation.When inspecting robots are inspecting,local maps should be set up for the features of the platform environment.In this experiment,we scanned the surrounding obstacles information by the laser radar installed by the inspection robot,and calculated the edges and corners of the surrounding equipment,and then set the edges and corners of these devices into the characteristic points of the local environment of the oil production platform.A feature map of the platform is established according to the characteristic points obtained.Finally,a systematic test is carried out through the "Haibao" project model to verify the feasibility of the method.3,the effective method to solve the local positioning of the inspection robot is studied.Theoretically,the use of the Untraced Kalman filter is better than the extended Kalman filter,that is,the error is smaller.The test platform combined with SLAM model and equipment inspection robot to pry the edges and corners as the platform of environmental characteristics,and no trace of Kalman filtering algorithm to detect inspection robot pose in the inspection process and different time periods in which the platform position information by using the extended Kalman filter.4,suppose two simple platform environment,based on a good mathematical model established in advance,extract the environmental features in a specific application environment,and expand the Kalman filter and the unscented Kalman filtering method to conduct the robot positioning simulation experiment.Through the experiment,verified the feasibility of using the extended Kalman filter algorithm in oil production platform and the unscented Kalman filtering algorithm to solve the problem of inspection robot localization,and got better precision accuracy effect by unscented Kalman filtering location than the application of the extended Kalman filter.