Design Of ROV Navigation System Based On Multi-Source Information Fusion

With the increasing demand for resources,more and more countries have paid attention to abundant marine resources.As an important tool for marine exploration and development,Remotely Operated Vehicle(ROV)can replace human beings for underwater operation.Moreover,it has an irreplaceable position in the field of seabed survey,underwater inspection and marine scientific research.However,as the marine environment is complex and diverse,it is necessary to assemble a high realiable navigation system for ROV in order to satisfy the operation requirements and provide reference information for operator.Therefore,it is important to develop and research the high reliable navigation system for ROV.Nowadays,with the development and maturity of information fusion technology,the research of multi-sensor information fusion is the main trend of the development of ROV navigation technology.In this paper,firstly,the ROV navigation system platform is constructed according to the design requirement.What's more,combined with the actual demand,the hardware system is improved and optimized.Besides,the types of key sensors are determined.Meanwhile,the related electric circuit for navigation system,such as,the power circuit,data interface and communication circuit of the,are designed in detail.Then,the hardware platform of the navigation system is built.Secondly,we establish the basic model of ROV navigation system.Mainly,the transformation relationship between each coordinate frame of the navigation system is introduced.Moreover,the principle of strapdown inertial navigation system,including the updating equations of navigation system are derived.Besides,we estabilish the error model of electronic compass(EC),doppler velocity log(DVL)and depth meter by analysising the working principle and error mechanism.Thirdly,the error sources of the micromachined inertial measurement unit are analyzed,and the calibration experiments of gyroscopes and accelerometers are carried out,then the related parameters are obtained.In view of the actual situation that magnetometer needs to be calibrated,a field calibration scheme based on three-dimensional rotation is designed,and the calibration experiment is designed and carried out to verify the effectiveness of the design method.Then,the ROV multi-sensor information fusion scheme is studied.The filter algorithm is used to fuse the measured data of different sensors and the output data of the inertial navigation system to obtain more reliable data information.This paper uses extended Kalman filter for data fusion.At the same time,the design of the theoretical derivation and digital simulation analysis to verify the performance of the designed system.At the same time,the implementation of the algorithm is based on the embedded software STM32.Finally,the onshore,pool,lake,and marine experiments of the navigation system for ROV are carried out,respectively.And the experimental results demonstrate that the designed system satisfies the requirements of ROV operation.