Research On Initial Alignment Of Inertial Measuerement System For Real Estate Unit

Targeting real estate unit,the real estate field survey investigates and measures the ownership,boundaries,area and other aspects.Conventional real estate surveying method,such as total station and GPS,has the defects of high dependence on environment and low efficiency.Strapdown inertial measurement system which is independent and free from outside interference is proposed for real estate surveying in this thesis.The initial alignment is the premise and key of strapdown inertial measurement system,and the precision of initial alignment determines the positioning precision of system.The initial alignment methods used for real estate surveying are discussed in this thesis.The adaptive Kalman filtering algorithm is used to estimate and compensate the errors of attitude angles and inertial sensors in the initial alignment process,so that he precision and stability of initial alignment is improved,and the positioning precision of system is improved eventually.The main research work of this paper includes the following three aspects:1.The four position alignment technology used for strapdown inertial measurement system is investigated.The multi position alignment scheme in horizontal state is analyzed and compared,and the four-position alignment scheme in incline state is mainly studied with considerations of alignment time and accuracy.The improved four position alignment scheme and dual gyro alignment scheme in the plane is proposed.The relationship between impact factors,such as initial azimuth?lever pitch and latitude,and error influence coefficient of four position alignment accuracy,such as random drift of fiber optic gyro?Transposition error?Horizontal inclination measurement error and latitude error,is researched.2.The single position alignment technology used for strapdown inertial measurement system is investigated.The traditional and improved parsing coarse-alignment methods and these error characteristics are analysed.Based on the error equation of strapdown inertial system and external auxiliary information,the system state equation and observation equation of position known point and position unknown point are derived in detail.The degree of observability of system state variables on different special measurement points is analyzed by using the method of singular value decomposition.Simulation analysis of single position alignment is carried out based on Kalman filtering algorithm.3.The adaptive Kalman Filter used for single position alignment is investigated.Aimed at the problem of the reduced precision of system model and noise statistical characteristics are caused by random factors during the real estate surveying,the adaptive Kalman filtering algorithm is introduced.In order to avoid filtering divergence,the simplified Sage-Husa adaptive filtering algorithm is used to adapt observation noise covariance matrix.The current data is used more than before by using forgetting factors so that the single position alignment precision of stability can be improved.In order to adapt system noise covariance matrix,the fuzzy adaptive Kalman filtering algorithm is introduced.By using fuzzy rules to determine adaptive factors,the gain matrix and the system noise covariance matrix are adjusted adaptively,and the stability of single position alignment is improved.Through the Turntable experiment and positioning experiment,the effectiveness of the initial alignment scheme and filtering method proposed in this thesis are proved,and the positioning precision can meet the requirements of real estate survey.