Study Of The Self-alignment Method Under Latitude Unknown In Dynamic Disturbance Condition

The initial alignment is one of the most important techniques in the inertial navigation system. The accuracy and time required of initial alignment directly affects the start-up time of the navigation system and the subsequent navigation precision. Nowadays, the method of static base alignment is relatively mature, and has been successfully applied to all kinds of SINS. However, how to achieve high precision self-alignment in dynamic environment is still the key problem, especially in the complex dynamic environment, such as underwater, underground, forest, city and so on, where GPS equipment cannot be used to obtain the information of latitude and where latitude must be calculated by inertial sensor's measurement information in real-time so that self-alignment can be completed. This paper studies on achieving self-alignment of SINS under unknown latitude in dynamic disturbance conditions. The initial alignment is completed using the feature that the gravity acceleration stays unchanged in the inertial coordinate system to calculate the latitude information of high accuracy in real time. This paper carried out studies on the following aspects:Firstly, a new method for calculating the latitude of the inertial coordinate system based on the inertial coordinate system is presented. The latitude of certain precision is calculated by geometric relationship using the feature that the gravity acceleration stays unchanged in the inertial coordinate system. However, the direct use of gravity acceleration to calculate latitude information can lead to large errors because of interference acceleration or even worse the latitude cannot be calculated. In order to reduce the interference caused by linear vibration the acceleration of gravity was integrated to get the velocity information and the smoothing velocity information was used to calculate latitude, which improved the accuracy and effectiveness of latitude. The error analysis and experimental results proved the effectiveness of the algorithm. On the basis of this calculation, the coarse alignment of SINS is realized by using the characteristics of the gravity acceleration in the inertial system, which contains the north information. The experiments showed that this method can realize coarse alignment under latitude unknown condition.Secondly, based on finished the coarse alignment, the fine alignment error model is established under unknown latitude in dynamic disturbance conditions, where treats linear vibration, lever-arm interference and latitude error were treated as uncertain disturbance. Then, the self-alignment of SINS is realized by using innovation adaptive filter. The first method is using the exponential function adaptive filtering based on the strictest criterion of convergence to replace fuzzy reasoning and solve the problem of unsuitability of fuzzy adaptive filtering for improving the accuracy. The validity of the proposed method is proved by the result of experiments. The other method is computing weighted calculation method, which can be used to solve the problem of the possibility of having evident deviation of real innovation covariance in the innovation adaptive algorithm. Therefore, the proportion of the innovation with evident deviation in the actual covariance was reduced, which ensured the performance of the innovation filtering algorithm and improved the stability of the system at the same time. The experimental results show that the improved adaptive filtering algorithm can effectively realize the self-alignment of SINS.Finally, to solve the problem of the process noise uncertainty of the fine alignment model caused by the random error of the sensor, such as the gyro and accelerometer, an adaptive filtering algorithm that adjust the measurement noise and process noise at the same time was proposed. Based on the basic principle of the innovation adaptive filtering algorithm, the Mahalanobis distance between the theoretical value and the real value of the innovation is used to determine the error size and weight of the new innovation. The impact of the larger errors was reduced by adjusting the proportion of the innovation in the estimation of measurement noise and process noise, and the robustness of the system is improved. The experimental results show that the new adaptive method can effectively solve the problem of the self-alignment of SINS under unknown latitude in dynamic disturbance conditions.