Research Of The Information Fusion On Ship’s SINS/GPS/CNS Integrated Navigation System

Strapdown Inertial Navigation System(SINS), Global Position System(GPS) andCelestial Navigation System(CNS) are three modern primary navigation systems. SINSis a self-independent and high-reliability navigation system who can provide navigationinformation such as acceleration, speed, location, course, attitude, etc. The accumulationof errors by time is its major drawback. GPS supplies high-precision navigationinformation with a low updating frequency and the system is unreliable. CNS is unableto provide real-time information of speed and location but offers high-precision attitudeat a low updating frequency. It is obvious that one simple navigation system is notcomprehensive and integrated navigation system combined with SINS, GPS and CNS isan ideal solution. The combination takes advantage of the navigation information fromSINS, GPS and CNS synthetically, in which the SINS plays the most important rolewhile the GPS and CNS are supplementary. SINS/GPS/CNS integrated navigationsystem is the development trend of navigation and measurement technics.Based on the output characteristics of the SINS, GPS and CNS, this article studiedthe methods to correct the accumulated measuring errors by time utilizinghigh-precision information of speed and location from GPS and attitude from CNSrespectively. And a deep research of the information-fusion methodologies to realizehigh-precision SINS/GPS/CNS integrated navigation is made.Introduction of basic theory for SINS/GPS/CNS integrated navigation system isrevealed in the article firstly. Basic principles and navigation algorithms of SINS,system structure and measuring mechanism of GPS and operating principle of CNS arestudied. Characteristics of the noise from the three systems are analyzed respectivelyand an error module is established based on the stimulated algorithm.How to fuse the measuring information from different sources is the key point torealize integrated navigation. Kalman Filtering-the basic filtering method ofinformation-fusion technic is researched. Three information-fusion methodologies aredesigned according to the operating mode of integrated navigation and the merits anddemerits are compared and analyzed on the basis of system precision, fault-tolerance,stability and the practical demand of the laboratory.Three information-fusion methodologies are developed according to the fusionmethodology which are descripted and analyzed in detail. Problems like the informationasynchronization of three measuring elements and the fault-tolerance of the system areconsidered sufficiently. Finally, the flow and software stimulation of federated filteringand cascaded filtering is designed based on the algorithm. Stimulation of SINS/GPSintegrated navigation is finished in order to comparing the measuring precision. Thestimulated results showed that enhanced high-precision attitude information is obtained when the Kalman Filter in SINS/GPS integrated navigation is supplemented by thehigh-precision attitude from CNS and improved system fault-tolerance ability isachieved through the selection of the methodology.