Research On Initial Alignment Method Based On Bio-inspired Polarized Skylight Sensor

In recent years,navigation has been widely and deeply applied in many fields such as aerospace,deep-sea exploration,and ground transportation.At present,navigation technology is developing towards miniaturization,intelligence,strong autonomy,high accuracy,fast response,and strong anti-interference.As a typical navigation method,strapdown inertial navigation system(SINS)has become the most core technology in the navigation field due to its small size,strong autonomy,full navigation parameters,and high output frequency.As a kind of navigation method for dead reckoning,it is essentially an integral process for navigation.The navigation errors can accumulate over time,and the imprecise initial state will cause the navigation to fail.Therefore,a high-precision initial state is required before navigation.The initial alignment technique can obtain a high-precision initial attitude and parameters.The most critical technique is to determine the initial attitude matrix between the body coordinate system and the navigation coordinate system.However,the existing initial alignment methods spend long alignment time with low accuracy.Besides,some initial alignment methods cannot reach autonomous alignment requirements.It has been found that a kind of sand ants named by Cataglyphis can return to their nests along a nearly straight path after longrange random foraging,which provides a biological perspective for a high-precision autonomous alignment technology.Inspired by this,the bio-inspired polarized skylight sensor is developed,which can meet the requirements of autonomous initial alignment with high precision,no accumulate error and strong anti-interference ability.In this paper,a bio-inspired polarized skylight sensor(PSS)and SINS are combined to establish a SINS/polarized skylight sensor integrated navigation system(SINS/PSS).The initial alignment includes two stages,namely coarse alignment and fine alignment.In coarse alignment phase,an analytical coarse alignment method based on polarized skylight is proposed.In fine alignment phase,based on different typical misalignment angles,the linear and nonlinear models of SINS/PSS are established.The observability is analyzed for different models,respectively.Kalman filter(KF)and unscented Kalman filter(UKF)are designed for different models to estimate unknown states,respectively.The main research contents are as follows:1)For the problem that the traditional analytical coarse alignment method is affected by the accuracy of the gyroscope,a novel analytical coarse alignment method based on polarization information is proposed,and the error characteristics of the proposed analytical coarse alignment method are analyzed.The simulations is verified that when the accuracy of gyroscope is imprecise,the polarization information can be used for the analytical coarse alignment to meet the requirements.2)For the problem that the traditional fine alignment method has low alignment accuracy and long accuracy time with small misalignment angles,a novel linear model of SINS/PSS integrated navigation system is established.The geometric observability and stochastic observability analysis methods are used to analyze the observability and the degree of observability of the SINS/PSS linear model.KF algorithm is designed to achieve information fusion.The simulations and experiments are used to compare the traditional method and SINS/PSS,which proves the effectiveness of the proposed method3)For the problem that the initial alignment accuracy is affected by different typical misalignment angles,a novel SINS/PSS nonlinear model is established.The observability analysis method based on Li derivative is used to analyze the observability of the SINS/PSS nonlinear model.The UKF algorithm is adopted to achieve information fusion.The simulations and experiments verifies that the initial alignment time and accuracy of the proposed SINS/PSS nonlinear model are hardly affected by different typical misalignment angles,and the performance of alignment is significantly better than the SINS nonlinear model.