Research On Information Processing Methods For Target Detection Of AUV In The Virtual Test System

Recently, AUV’s underwater detection technology is one of the hot researchtopicsin the area of Instrument Science and Technology. Target’s detection informationprocessing method is one of the key technologies in the area of AUV. In the process ofAUV navigate autonomously, multi-sensor information fusion technology caneffectively improve the accuracy of target detection, advanced path planning method isan important guarantee for AUV to reach the destination quickly and safely. In order toimprove the ability of AUV’s independent sailing, this paper focuses on two keytechnologies, which are the multi-sensor information fusion and path planning of AUV,to do some research. The main work of this paper is as follows.In the area of AUV’s target detection information and its applications, in-depthstudy of a multi-sensor information fusion technology and its application in the pathplanning is done. This paper research AUV’s typical multi-sensor information fusionlevel and commonly used algorithms firstly. Then, the pros and cons of the variousAVU’s path planning method are analyzed and compared. The theoretical foundation islaid for this paper to select an efficient, flexible and intelligent route planning method.In the AUV’s system platform, combined with the extended Kalman filter (EKF)tracking algorithm and unscented Kalman filter (UKF) tracking algorithm, a bearingsinformation fusion model based on the dual-sensor array is designed, which overcomethe shortage of single sensor can’t detect the target’s orientation. Practice shows thatUKF have a higher target positioning accuracy than EKF.In order to overcome the defects of traditional local convergence of the geneticalgorithm, adaptive crossover operator, adaptive mutation operator and optimizingoperator are proposed. In order to improve the speed of operation of the algorithm andadapt to the requirements of parallel processing of path planning, parallel structure ofthe traditional genetic algorithm which is combined with the structural features of themulti-core processors is improved. The practice shows that the improved algorithm canachieve the desired objectives and features high ability of global optimization andexpedite operation.Information fusion model and improved parallel genetic algorithm are applied tothe AUV path planning for comprehensive simulation in the virtual test system. Theresult shows that this model and algorithm can preferably achieve the goal ofautonomous navigation of AUV.