Research Of Navigation System Integrity Monitoring And Information Fusion Based On Multi-Sensors Network

Multi-sensors network navigation system is a type of system with new design concept. Thesensors systems are installed in different locations of an aircraft or a ship to provide redundantmeasuring information for the navigation system as well as the inertial information used for thecalculation of local motion compensation. In order to improve the performance of the navigationsystem, some key technologies about integrity monitoring and information fusion based onmulti-sensors system are researched in this paper. The main innovation points are as follows:Firstly, this paper studied the design method of multi-sensors navigation system and proposed thesystem model based on sensors networks, analyzed the structure of multi-sensors data fusion, and thenput forward the optimal configuration rule after researching the optimal configuration method ofmulti-sensors network, using the parameter of mean time between failures to analyze the reliability ofthe configuration. Finally, a single sensor contribution rate parameter was adopted to optimize thenumber of the redundant sensors. From the above research, some commonly used optimal schemes ofsensors configuration were put forward.Secondly, this paper analyzed the integrity monitoring method of multi-sensors, and adopted theMoving-Window Generalized Likelihood Ratio Test method to solve the problem of GeneralizedLikelihood Ratio Test method that the test sensitivity became weak as time went by, which greatlyimproved the reliability of the navigation system. To solve the problem that when the redundantinformation was not enough we could only diagnose the system fault but could not isolate the faultedsensor, a synthesized method combine wavelet transform with parity check was adopted. Choosing theproper wavelet basis and the decomposition level of wavelet transform to make the detection functionhas different sensitivity from the faults to the maneuvers, simultaneously, took the parity function as asafety insurance device. The simulation results of step and ramp faults proved the effectiveness of thealgorithms proposed in this paper.Thirdly, some multi-sensors information fusion theories and some commonly used Filterstructures were analyzed, based on which the virtual sensors fusion algorithm was put forward, withthis method the order of sensors node was reduced before information fusion. The simulation resultindicated that by using this method we can get high-precision navigation information relative toorthogonal axes with the outputs from single-node multi-sensors. Then a distributedinformation fusion algorithm applied to multi-nodes navigation system was researched, this method studied the distributed system structure, measurement model and the integration algorithm. Thesimulation result suggested that this method can make full use of the redundant information, whichimproved the performance of the estimate of navigation states effectively.Fourthly, a digital simulation model based on multi-sensors network navigation system wasestablished which was comprised of trajectory generation module, integrity monitoring module andinformation fusion module. Moreover, a semi-physical simulation platform was designed, and itsfunction was realized preliminarily.This paper analyzed the new methods of integrity monitoring and information fusion based onsensors network navigation system, and put forward some effective methods aimed at the keytechnologies which had achieved some innovation fruit, all those studied above will contribute to thedesign of high performance navigation systems in the near future.