Intelligent Vehicle Integrated Navigation Data Fusion Algorithm

Intelligent vehicle technology is got increasingly affected attention by variousdeveloped countries, has became a research hotspot. In Our our country, governmentmakemakes it as a priority development advanced technology. Intelligent vehicletechnology is the background of this this paperpaper. The key content is intelligentvehicle integrated navigation data fusion algorithm.After studing the basic theory of integrated navigation system,In this paper wechoose BDcompass navigation satellite system and/INSinertial navigation system ischosen be constitute the integrated navigation system, In this paper, introducingintroduce the component of this system and Kalman filter model in detail. In addition,test and verify this method improves improveing the accuracy of navigation system.Thekey point of this paper is data fusion algorithm Using tThe Kalman filter in integratednavigation system has two problems, one is system model isn’t nonlinear, the other isuncertain noise model is uncertain. They all might affect the accuracy of navigation,even would lead to volatilization. In this paper IThe study puts emphasis up on thesetwo problems.The UKF(Unscented Kalman Filter) is a method for dealing with the modelnonlinear problem. The algorithm uses UT(unscented transform) producing a series ofsampling sites approach the state vectorX k1. We could make let the sampling sitesthough the nonlinear equation. After weighting the sampling sites, it could got get stateestimation and update the data. The innovation of new algorithm is addings the intervaloptimal smoothing method it is named as interval smoothing UKF. The algorithmsmoothes the state in a immovable areas after finishing normal Kalman. After rectifyingthe computational data,It it could further enhance the accuracy of navigation. Thesimulation result shows the effectiveness of the approach. Not only enhancing theaccuracy of the system, but also making the algorithm more robust.The adaptive Kalman is a good method for dealing with noise model uncertainproblem. In this paper, improve the traditional Sage-Husa adaptive algorithm.Theimproved adaptive Kalman adds determinant conditions, contrasting the variance ofinnovation to theoretic vaiance, if it satisfies the determinant conditions the improvedalgorithm would adjust statistics of the measurement noise, that will decrease thecalculated amount. The robust theory is added in this improved adaptive Kalmanalgorithm, it could decrease gross error disturbance and enhance stability of the system.The simulation result shows the effectiveness of the approach, the improved adaptiveKalman fixes the noise model uncertain problem and has less calculation amount, it get a better real time property. The new algorithm adds robust theory further enhances theaccuracy.In summary, integrated navigation is used in intelligent vehicle could enhance theaccuracy of navigation and make the system more robust. In this paper, intervalsmoothing UKF fixes system model nonlinear problem, improved Sage-Husa adaptiverobust Kalman algorithm fixes the noise model uncertain problem. These all enhancethe location accuracy efficiently, make the system more stable.