Study On Seabed Terrian Matching Navigation With Multi-Sensors For AUV

The Seabed Terrain Matching Navigation(STMN)can realize the long time underwater navigation for Autonomous Underwater Vehicle(AUV),meet the "long-term,subtle,all-weather,high precision" navigation needs,and without external sensors and floating upward to correct accumulated errors.This paper took the submerging AUV's high precision navigation method for research background,and different acoustic sensors were treated as the topographic survey equipment to investigate the STMN method.The major work and conclusions of present paper were as follows.1.The STMN model and particularity were studied.Put forward the composition,principle and navigation model of STMN based on the terrain matching system of aircraft,and analyze the factors which will influence the matching precision.On this basis,the same and different points of terrain matching method with aircraft and AUV were discussed,and advantages and disadvantages of different sounding sensors for seabed terrain matching were analyzed,finally the STMN method was put forward.2.The method for multi-beam echo sounding data analyzing and terrain matching unit building were investigated.The filtering method for multi-beam echo sounding data was investigated started from analyzing the principle of multibeam echo sounding.The data model of real-time sounding data was built based on different sensors,and the building method of priori terrain database based on underwater digital terrain map and multi-beam echo sounding data was investigated.The storage structure which was applicable to priori terrain database for seabed terrain matching was investigated.The method for selection of searching area in terrain matching and eliminating errors between real-time sounding data and priori terrain data were investigated.3.The terrain matching positioning method based on multi-beam echo sounding data was investigated.The seabed terrain matching positioning model based on maximum a posteriori estimation was investigated using characteristics of multibeam echo sounder.For the problem that the fake positioning point in low relief areas and the increased of real-time sounding data will increase the computational complexity,the distinguishing method of fake positioning point based on Pulse Coupled Neural Network(PCNN)was put forward and the characteristics of this method was investigated using the data gotten from sea trial.Finally,the high precision positioning method based on Markov Random Field(MRF)aimed to the demand for high-precision underwater positioning.4.The STMN method based on low-cost sounding sensor was investigated.Some low-cost sounding sensor,such as acoustic altimeter and Doppler velocity log,were treated as seabed terrain measuring equipment.The path tracking model for STMN based on Bayesian filtering mode was put forward the approximate solution for bayesian filter problem was found using particle filter method.For the particles dilution problem in particle filter method,the improved particle filter STMN methods based on no-repeated-sampling and particle adjustment were put forward respectively,and the sounding data of acoustic altimeter and Doppler speed meter from multi-beam echo sounding data file was used to investigate the characteristic of both improved method.5.The multi-sensor STMN method was investigated.First the characteristic of seabed terrain was described,then the suitability was analyzed and next the path planning method for seabed terrain matching was investigated.Finally the multi-sensor terrain matching navigation method based on search positioning and path tracking was put forward according to the rich degree of terrain characteristic.On basic of the navigation model,the correction method based on terrain matching navigation for reference navigation system was put forward and the structure of AUV seabed terrain matching aided navigation system was designed.This paper investigated the STMN method,the data gotten from sea trail was used to make playback mode simulation experiment which is based on real sounding data.The results showed that the seabed terrain matching method can be used alone or combined with other method,and this method has good ability to adapt to the characteristic of seabed terrain.The navigation precision and position precision suit the needs of AUV underwater navigation and position.