High-precision inertial navigation information fusion method based on confidence analysis

Abstract

The application aims to provide a high-precision inertial navigation information fusion method based on confidence analysis, comprising the following steps: designing the comprehensive error state of a pure inertial navigation system and a high-precision inertial navigation system, constructing the comprehensive error state variance; establishing an observation equation, building the equation of discrete Kalman filtering; performing Kalman filtering on the pure inertial navigation system and each high-precision system respectively; calculating the confidence index of each system, thereby calculating the normalized confidence index through the confidence index, and realizing the weighted information distribution of multiple high-precision systems. The application aims at the problem that multiple high-precision inertial navigation systems lack information fusion and effective cooperative positioning, establishes the comprehensive error state and state equation by means of the pure inertial navigation system and performs Kalman filtering, calculates the confidence index and the normalized confidence index of each high-precision system according to the second-order center distance of the comprehensive velocity error state, thereby realizing the high-precision inertial navigation information fusion based on the confidence analysis, and improving the navigation precision and reliability of the system.

Description

TECHNICAL FIELD

[0001] The application relates to an inertial navigation processing method, in particular to a high-precision inertial navigation processing method. BACKGROUND

[0002] In recent years, with the increasing demand for marine navigation, many navigation methods are widely used in practice, such as celestial navigation, inertial navigation, terrain aided navigation, satellite navigation and the like. Compared with other navigation methods, inertial navigation can completely autonomously navigate. It can output the attitude, speed and position information of a carrier without information exchange with the external environment, and thus has very strong autonomy and anti-interference capability.

[0003] In high-safety occasions such as large ships and airplanes, two or even multiple sets of high-precision inertial navigation systems are often redundantly configured to improve the precision and reliability, and the redundant information is used for fault detection and isolation of each part of the system to ensure safe navigation of the carrier.

[0004] However, the conventional redundant configuration often adopts a master-slave backup mode, and only uses the navigation information of one inertial navigation system, resulting in resource waste. Therefore, how to perform online evaluation on the performance of the high-precision inertial navigation system, realize information fusion, and fully utilize the navigation information of the high-precision inertial navigation system to improve the navigation precision and reliability has become a problem to be solved.

[0005] However, in the published articles, such as a paper entitled 'Carrier Phase Time Difference Auxiliary SINS / GNSS Tight Combination Navigation Method' by Dong Yi, Wang Dingjie and the like in the 29th volume of the Journal of China Inertial Technology, No. 4, August 21, the authors propose an improved inertial / satellite tight combination navigation method based on carrier phase time difference, which directly uses high-precision carrier phase time difference observation information to improve the online compensation precision of inertial navigation error. However, the method needs to obtain external reference position information. In the patent 'A Method for Selecting Navigation Information of Multiple Inertial Navigation Devices' by Xu Qihua and Liu Bing, the authors evaluate the effectiveness of the key monitoring information of each inertial navigation device and judge the self-checking information, and when the number of effective inertial navigation devices is more than one, the navigation information of the inertial navigation device with relatively higher precision or the default setting is selected for external output. However, the method can only evaluate and select the inertial navigation devices, and cannot realize information fusion of multiple inertial navigation systems. Therefore, it is innovative and has practical engineering value to research a method for realizing weighted information fusion of multiple high-precision inertial navigation systems without information exchange with the external environment. SUMMARY

[0006] The application aims to provide a high-precision inertial navigation information fusion method based on confidence analysis, which can realize high-precision inertial navigation information fusion based on confidence analysis and improve the navigation precision and reliability of the system.

[0007] The object of the present application is achieved in that:

[0008] The present application is based on a high-precision inertial navigation information fusion method with confidence analysis, characterized in that:

[0009] (1) The comprehensive error state of a pure inertial navigation system and a high-precision inertial navigation system is designed, and a comprehensive error state variance is constructed;

[0010] (2) The eastward and northward position differences of the pure inertial navigation system and the high-precision inertial navigation system are taken as observations, an observation equation is established, and a discrete Kalman filter equation is built;

[0011] (3) The pure inertial navigation system and each high-precision system are subjected to Kalman filtering respectively;

[0012] (4) The confidence index of each high-precision system is calculated according to the second-order central moment of the comprehensive velocity error state estimation of each high-precision system, so as to calculate the normalized confidence index through the confidence index, realize the weighted information distribution of multiple high-precision systems, and the position output obtained through the information weighted distribution is:

[0013]

[0014]

[0015] wherein, Q i is the normalized confidence index of the i-th high-precision system, L i is the eastward velocity output of the i-th high-precision system, λ i is the northward velocity output of the i-th high-precision system, is the eastward velocity output obtained through the information weighted distribution, is the northward velocity output obtained through the information weighted distribution.

[0016] The present application can also include:

[0017] 1. The comprehensive error state and the state equation in step (1) are as follows:

[0018]

[0019] wherein the comprehensive error state is specifically:

[0020] δL 12 : difference between latitude error outputs of the pure inertial navigation system and the high-precision inertial navigation system;

[0021] δλ 12 : difference between longitude error outputs of the pure inertial navigation system and the high-precision inertial navigation system;

[0022] The difference between the output velocity errors of the pure inertial navigation system and the high-precision inertial navigation system;

[0023] The difference between the output attitude errors of the pure inertial navigation system and the high-precision inertial navigation system;

[0024] The error of the earth rotation angular velocity caused by the latitude error of the pure inertial navigation system and the high-precision inertial navigation system;

[0025] The transfer angular velocity error caused by the latitude error and the velocity error of the pure inertial navigation system and the high-precision inertial navigation system;

[0026] The gyro error of the pure inertial navigation system;

[0027] The gyro error of the high-precision inertial navigation system;

[0028] The accelerometer error of the pure inertial navigation system;

[0029] The accelerometer error of the high-precision inertial navigation system;

[0030] The gyro noise of the pure inertial navigation system;

[0031] The gyro noise of the high-precision inertial navigation system;

[0032] The accelerometer noise of the pure inertial navigation system;

[0033] The accelerometer noise of the high-precision inertial navigation system.

[0034] 2. The confidence index in step (4) is P1, P2…P i …P n , which is specifically:

[0035]

[0036] wherein i represents the i-th high-precision system, n represents the total number of high-precision systems, wherein i≤n, represents the comprehensive velocity error state estimation of the i-th high-precision system, D[] represents the second-order central moment calculation, P i is the confidence index of the i-th high-precision system.

[0037] 3. The normalized confidence index Q1, Q2…Q i …Q n , which is specifically:

[0038]

[0039] and satisfies:

[0040]

[0041] wherein i represents the i-th high-precision system, n represents the total number of high-precision systems, wherein i≤n, P i is the confidence index of the i-th high-precision system, Q i is the normalized confidence index of the i-th high-precision system.

[0042] The present application has the advantages that: the present application aims at the problem of lacking information fusion and effective cooperative positioning among multiple high-precision inertial navigation systems, establishes comprehensive error states and state equations by means of a pure inertial navigation system and performs Kalman filtering, calculates the confidence index and the normalized confidence index of each high-precision system according to the second-order center distance of the comprehensive velocity error state, thereby realizing high-precision inertial navigation information fusion based on confidence analysis, and improving the navigation precision and reliability of the system. BRIEF DESCRIPTION OF DRAWINGS

[0043] Figure 1 is the flow chart of the present application;

[0044] Figure 2 is the system position error comparison before and after the high-precision inertial navigation information fusion based on confidence analysis. DETAILED DESCRIPTION

[0045] The present application will be described in more detail below with examples combined with the drawings:

[0046] Combined Figures 1-2 , the high-precision inertial navigation information fusion method based on confidence analysis of the present application comprises the following steps:

[0047] Step one: design the comprehensive error states of the pure inertial navigation system and the high-precision inertial navigation system, and construct the comprehensive error state variance. The comprehensive error states and state equations are as follows:

[0048]

[0049]

[0050] wherein the comprehensive error states are specifically:

[0051] δL 12 : the difference between the latitude error output by the pure inertial navigation system and the high-precision inertial navigation system;

[0052] δλ 12 : the difference between the longitude error output by the pure inertial navigation system and the high-precision inertial navigation system;

[0053] The difference between the output velocity errors of the pure inertial navigation system and the high-precision inertial navigation system;

[0054] The difference between the output attitude errors of the pure inertial navigation system and the high-precision inertial navigation system;

[0055] The error of the earth rotation angular velocity caused by the latitude error of the pure inertial navigation system and the high-precision inertial navigation system;

[0056] The transfer angular velocity error caused by the latitude error and the velocity error of the pure inertial navigation system and the high-precision inertial navigation system;

[0057] The gyro error of the pure inertial navigation system;

[0058] The gyro error of the high-precision inertial navigation system;

[0059] The accelerometer error of the pure inertial navigation system;

[0060] The accelerometer error of the high-precision inertial navigation system;

[0061] The gyro noise of the pure inertial navigation system;

[0062] The gyro noise of the high-precision inertial navigation system;

[0063] The accelerometer noise of the pure inertial navigation system;

[0064] The accelerometer noise of the high-precision inertial navigation system;

[0065] Step two: take the eastward and northward position differences of the pure inertial navigation system and the high-precision inertial navigation system as observations, and establish the observation equation. And build the equation of discrete Kalman filter.

[0066] Step three: Kalman filter the pure inertial navigation system and each set of high-precision system respectively.

[0067] Step four: calculate the confidence index of each system according to the second-order central moment of the comprehensive velocity error state estimation of each set of high-precision system, and then calculate the normalized confidence index through the confidence index calculation to realize the weighted information distribution of multiple high-precision systems.

[0068] The confidence index calculation process is as follows:

[0069]

[0070] Where i represents the i-th set of high-precision system, n represents the total number of high-precision systems, and i≤n. represents the integrated velocity error state estimation of the i-th high-precision system, D[] represents the second-order central moment calculation, P i is the confidence index of the i-th high-precision system.

[0071] The normalized confidence index is calculated as follows:

[0072]

[0073] where i represents the i-th high-precision system, n represents the total number of high-precision systems, and i≤n. i is the confidence index of the i-th high-precision system, Q i is the normalized confidence index of the i-th high-precision system, and satisfies:

[0074]

[0075] The position output obtained by information weighted distribution is:

[0076]

[0077]

[0078] where Q i is the normalized confidence index of the i-th high-precision system, L i is the eastward velocity output of the i-th high-precision system, λ i is the northward velocity output of the i-th high-precision system, is the eastward velocity output obtained by information weighted distribution, is the northward velocity output obtained by information weighted distribution. Figure 1 The flow of high-precision inertial navigation information fusion based on confidence analysis is illustrated in the form of a block diagram.

Claims

1. A high-precision inertial navigation information fusion method based on confidence analysis is characterized by: (1) Design the comprehensive error states of pure inertial navigation systems and high-precision inertial navigation systems, and construct the comprehensive error state variance; (2) Using the east and north position differences between the pure inertial navigation system and the high-precision inertial navigation system as observation quantities, the observation equation is established, and the discrete Kalman filter equation is constructed; (3) Perform Kalman filtering on the pure inertial navigation system and each high-precision system; (4) According to the second-order central moment of the comprehensive velocity error state estimation corresponding to each high-precision system, the confidence index of each system is calculated, and then the normalized confidence index is calculated through the confidence index to realize the weighted information distribution of multiple high-precision systems. The position output obtained by the weighted information distribution is: Among them, Q i is the normalized confidence index of the i-th high-precision system, L i is the eastward velocity output of the i-th high-precision system, λ i is the northbound velocity output of the i-th high-precision system, is the eastward velocity output obtained by weighted distribution of information, The northbound velocity output is obtained by weighted distribution of information.

2. The high-precision inertial navigation information fusion method based on confidence analysis according to claim 1 is characterized by: The comprehensive error state and state equation in step (1) are as follows: The comprehensive error state is specifically: δL 12 : The difference between the latitude error output by pure inertial navigation system and high-precision inertial navigation system; δλ 12 : The difference between the longitude errors output by pure inertial navigation and high-precision inertial navigation systems; Pure inertial navigation system gyro noise; High-precision inertial navigation system gyro noise; Pure inertial navigation system accelerometer noise; High-precision inertial navigation system accelerometer noise.

3. The high-precision inertial navigation information fusion method based on confidence analysis according to claim 1 is characterized by: The confidence indicators in step (4) are P1, P2…P i …P n , which are specifically: Where i represents the i-th high-precision system, n represents the total number of high-precision systems, where i≤n, represents the comprehensive velocity error state estimation of the i-th set of high-precision systems, D[] represents the second-order central moment calculation, P i is the confidence index of the i-th high-precision system.

4. The high-precision inertial navigation information fusion method based on confidence analysis according to claim 1 is characterized by: The normalized confidence indicators Q1, Q2...Q i …Q n Specifically: And satisfy: Where i represents the i-th high-precision system, n represents the total number of high-precision systems, where i≤n, P i is the confidence index of the i-th high-precision system, Q i is the normalized confidence index of the i-th high-precision system.

Images