Shipborne navigation system integrity monitoring and positioning optimization method

Abstract

The invention discloses an integrity monitoring and positioning optimization method for a shipborne navigation system. The method comprises the following steps: acquiring a GNSS (Global Navigation Satellite System) intermediate frequency signal; carrying out carrier separation and coherent integration by combining a carrier ring and a code ring; calculating a phase error by a loop phase discriminator, and obtaining the optimal estimation of the carrier and code phase error under the action of a cascaded non-coherent loop preprocessing filter MCKF; the phase error is converted into a pseudo-range error and a pseudo-range rate error through linear transformation, so that the pseudo-range and the pseudo-range rate of the GNSS are obtained; detecting whether the satellite signals have faults or not by adopting an RAIM technology based on a least square residual algorithm, and eliminating the fault satellite signals; obtaining an SINS navigation result according to the measurement of the IMU, and constructing an error model; the navigation result of the SINS is converted into a pseudo range and a pseudo range rate, combined navigation is carried out in a tight coupling mode with the pseudo range and the pseudo range rate of the GNSS, and the carrier NCO and the code generator are updated by utilizing the navigation result. The problems of signal tracking loop lock losing and navigation result divergence caused by an existing scalar tracking method in a severe signal environment are solved.

Description

Technical Field

[0001] The present invention relates to the field of satellite navigation technology, and in particular to a method for integrity monitoring and positioning optimization of a shipborne navigation system. Background Art

[0002] The Global Navigation Satellite System (GNSS) is an indispensable tool for ship navigation, but its performance in complex environments is limited, especially in ports and busy waters. In severe sea conditions, such as wind and wave conditions, GNSS signals will also produce multipath effects, interfering with the GNSS navigation system. Receivers with traditional scalar tracking loop (STL) architectures are difficult to meet the requirements of positioning continuity in these conditions, and the positioning accuracy cannot meet the specified standards. In addition, satellite signal failures can cause satellite positioning results to diverge. Therefore, in harsh signal environments, a more accurate and robust vector tracking loop method and integrity monitoring technology need to be applied to GNSS receivers to improve the navigation performance of ships in complex environments and enhance the reliability of ship positioning. Summary of the invention

[0003] The present invention is made to solve the above-mentioned problems, and aims to provide a method for monitoring the integrity and optimizing the positioning of a shipborne navigation system.

[0004] The present invention provides a method for monitoring the integrity and optimizing the positioning of a shipborne navigation system, which has the following characteristics and comprises the following steps: Step 1, using a GNSS antenna to receive satellite signals, and performing radio frequency front-end processing to convert the satellite signals into intermediate frequency signals; Step 2, using a carrier loop to perform carrier separation on the intermediate frequency signals to obtain sine signals and cosine signals; Step 3, bringing the sine signals and cosine signals after carrier separation into an integration-removal device in a code loop to obtain leading, immediate and lagging coherent integration values; Step 4, the coherent integration values ​​are output as carrier phase errors and code phase errors through a loop discriminator, and are used as the quantity measurements of a cascaded incoherent loop preprocessing filter to obtain code phase errors, carrier phase errors and carrier frequency errors. The carrier frequency change rate error is estimated by using the carrier phase error and the code phase error; in step 5, the carrier phase error and the code phase error are linearly transformed into the pseudorange error and the pseudorange rate error, so as to obtain the pseudorange and the pseudorange rate; in step 6, the integrity monitoring technology based on the least squares residual algorithm is used to detect whether the satellite signal is faulty, and the faulty satellite signal is eliminated; in step 7, the inertial navigation result is calculated, and a tightly coupled combined navigation processor is constructed with the pseudorange and the pseudorange rate to perform combined navigation and obtain the navigation result; in step 8, the navigation result is used, and the carrier phase error is used to adjust the carrier numerically controlled oscillator to complete the update of the carrier frequency. The navigation solution estimate value and the line-of-sight vector prediction value are used to predict the code rate of each tracking channel at the next moment to complete the update of the code rate.

[0005] In the method for integrity monitoring and positioning optimization of the shipborne navigation system provided by the present invention, it can also have the following characteristics: wherein, in step 4, based on the incoherent cascade vector tracking loop, the code phase error Δτ, the carrier phase error Δθ, the carrier frequency error Δf and the carrier frequency change rate error Δα are used as the state quantity X of the loop filter, and the discrete form of the system model is as follows:

[0006] X i =[Δτ i ,Δθ i ,Δf i ,Δα i ] Τ (1)

[0007]

[0008]

[0009] In the formula, the superscript i (i = 1, 2, ..., n) represents the number of the tracking channel, n is the number of all tracking channels, the subscript k represents the time, k / k-1 represents the estimation of time k at time k-1, and X i represents the state quantity of channel i, represents the state of channel i at time k, Φ kk-1 is the system state transition matrix, is the system noise vector, coefficient β = f C / A / f L1 Used to convert the unit of carrier phase to the unit of C / A code, f C / A is the code frequency, f L1 is the carrier frequency, and T is the update interval of the loop filter. The measurement Z consists of the code phase error and the carrier phase error output by the loop discriminator. The discrete form of the measurement model is as follows:

[0010]

[0011] In the formula, Z i Indicates the quantity measurement of channel i, and Respectively represent the code and carrier phase error measurement values ​​of channel i, represents the measurement noise vector of channel i at time k, H k is the measurement matrix at time k.

[0012] The method for integrity monitoring and positioning optimization of a shipborne navigation system provided by the present invention may also have the following features: wherein, in step 2, the preprocessing specifically includes the following steps:

[0013] Step 2-1, sort each ship information in the automatic identification system data according to the time column; Step 2-2, filter the rows and columns of the sorted data, remove useless features, and retain the ship number, longitude, latitude, ground speed, ground course, and time; Step 2-3, select a set of limited original data related to a specific water area according to longitude and latitude, and remove duplicate and useless data in the original data to obtain valid data; Step 2-4, interpolate the valid data, standardize the longitude, latitude, ground speed and ground course, and do not standardize the time and ship number.

[0014] The method for integrity monitoring and positioning optimization of the shipborne navigation system provided by the present invention may also have the following characteristics: wherein the calculation formulas for the code phase and carrier phase measurement values ​​are as follows:

[0015]

[0016] In the formula, I E , Q E ,I L , Q L They are the coherent integration outputs of the leading and lagging in-phase and quadrature branches of the correlator respectively.

[0017] The method for integrity monitoring and positioning optimization of the shipborne navigation system provided by the present invention may also have the following features: wherein the method for adaptively adjusting the measurement noise is as follows:

[0018]

[0019] In the formula, R i,k is the measurement noise covariance matrix, and is the variance of the discriminator output, d0 is the interval between the immediate code and the delayed code, T is the coherent integration time, and the CNR of each channel is calculated by the variance sum method at the output value of the 20ms coherent integration.

[0020] In the shipborne navigation system integrity monitoring and positioning optimization method provided by the present invention, the following features may also be provided: wherein, the generalized similarity between two random variables is measured by using the correlation entropy. Given two random variables XY, assuming that their joint distribution function is F XY (x, y), the related entropy between them can be expressed as:

[0021] V(X,Y)=E[κ(X,Y)]=∫κ(x,y)dF XY (x,y) (11)

[0022] Where E[·] is the expectation, and κ(·,·) represents the kernel function. The Gaussian kernel function is selected as the kernel function of the maximum correlation entropy. The Gaussian kernel function is expressed as follows:

[0023]

[0024] In the formula, e i =x i -y i ; σ represents the kernel function width, and σ>0. In actual situations, the joint distribution F XY (x, y) is often unknown, so the sample mean estimator is used to estimate the relevant entropy:

[0025]

[0026] In the formula, e(i)=x(i)-y(i), In order to satisfy the joint distribution of N samples, the Gaussian kernel function is expanded using Taylor series:

[0027]

[0028] According to the above analysis, the correlation entropy is essentially the weighted sum of all even-order moments of XY, which effectively integrates the information of high-order moments. Under the condition of limited error data, the cost function based on the maximum correlation entropy criterion can be further expressed as:

[0029]

[0030] The method for monitoring the integrity and optimizing the positioning of the shipborne navigation system provided by the present invention may also have the following characteristics: wherein, the optimization equation J based on the maximum correlation entropy criterion in the Kalman filter is m (X k ) is derived as follows:

[0031]

[0032] In the formula, G σ (||XY||) is the Gaussian kernel function, Z k is the measurement vector; H k is the measurement matrix of KF at time k; X k is the state vector of KF at time k, X k-1 is the state vector of KF at time k-1, F k-1 is the state transfer matrix of KF at time k-1, is the measurement noise covariance matrix R of KF at time k k The inverse matrix of is the prior estimated covariance matrix P of KF at time k k / k-1 The inverse matrix of . The Gaussian kernel function is Appropriate kernel bandwidth can introduce high-order moments into signal processing and solve the maximum entropy, let:

[0033]

[0034] Take the initial approximation And in Kalman filtering, You can get:

[0035]

[0036] in, In the basic KF,

[0037]

[0038] In the cascaded incoherent loop preprocessing filter, we can get:

[0039]

[0040] in, It is through L k The adjusted measurement noise covariance matrix. When measuring outliers, the value of the new information Deviation will occur. In the cascaded incoherent loop preprocessing filter, according to the Gaussian kernel function, when an abnormal value is measured, the entropy value is very small, then Lk tends to zero, The filter gain K of the pre-filter increases continuously. k Decreases, at this time, the filter will reduce the weight of the measured value and rely more on the prior estimate, thereby reducing the impact of outliers on the estimation results and overcoming the interference of non-Gaussian noise on state estimation.

[0041] In the method for monitoring the integrity and optimizing the positioning of the shipborne navigation system provided by the present invention, the following features may be provided: wherein the measurement of the KF tracking loop comes from the output of the PLL and DLL discriminator, and the carrier phase error and the code phase error are calculated using the two-quadrant inverse tangent function and the incoherent lead-minus-lag amplitude method, respectively. The measurement vector and measurement equation of the loop can be expressed as:

[0042]

[0043] Where V loop,k is the measurement noise, and the measurement noise covariance matrix In the tracking loop, it can be written as R loop,k =diag([R 1,k ,R 2,k ]). In the vector tracking loop,

[0044] According to the linear relationship between code phase error and pseudorange, carrier frequency error and pseudorange rate, the measurement result of the navigation filter can be calculated as follows:

[0045]

[0046] Where c is the speed of light (299792458 m / s); f CA refers to the code frequency, is the Doppler frequency shift (Hz); f L1 is the carrier frequency; v usr and are the speeds of the receiver and satellite i respectively; is the receiver clock error; is the satellite clock error.

[0047] The method for integrity monitoring and positioning optimization of a shipborne navigation system provided by the present invention may also have the following characteristics: wherein, in step 6, the integrity monitoring method based on characteristic slope weighting is used to monitor satellite failure conditions, and the satellite characteristic slope is calculated as follows:

[0048]

[0049] In the formula, A LSR(1,i) With A LSR(2,i) Represents the matrix A LSRThe first row and the i-th column element and the second row and the i-th column element of . Based on the analysis of the impact of satellite characteristic slope on positioning accuracy, a differentiated weighting strategy is adopted. During the test process, satellites with larger characteristic slopes are given higher weights; conversely, satellites with smaller characteristic slopes have smaller weights, which reduces their test statistics. The weight matrix P after the characteristic slope improvement is slope The calculation is as follows:

[0050] P slope =diag(Slope1 / K * ,Slope2 / K * ,…,Slope n / K * ) (29)

[0051] In the formula, K * is the proportionality coefficient, K * =min(Slope i ), then the pseudorange residual vector is calculated as follows:

[0052]

[0053] The fault detection quantity of the characteristic slope weighted least squares residual algorithm is calculated as follows:

[0054]

[0055] In the shipborne navigation system integrity monitoring and positioning optimization method provided by the present invention, it can also have the following characteristics: wherein, in step 7, the SINS navigation result error is calculated according to the differential equation, and the formula is as follows:

[0056]

[0057] In the formula, φ=[φ E ,φ N ,φ U ] Τ is the ship's attitude, v n =[v N ,v E ,v U ] Τ and δv n =[δv N ,δv E ,δv U ] Τ are the speed and error of the ship to the north, east and upward respectively, (L, λ, h) and (δL, δλ, δh) are the latitude, longitude, altitude and error of the ship’s current position respectively, ω i n nis the angular velocity of the navigation coordinate system relative to the inertial coordinate system in the navigation coordinate system, is the angular velocity error of the Earth's rotation, is the angular velocity error of the navigation system, is the attitude rotation matrix from the carrier coordinate system to the navigation coordinate system, f b is the accelerometer output result of IMU, is the measurement error of the gyroscope, δf b is the measurement error of the accelerometer, δg n is the gravity error.

[0058] In the method for monitoring the integrity and optimizing the positioning of the shipborne navigation system provided by the present invention, the following features may be further provided: wherein, in step 7, the GNSS / SINS combined navigation processor based on the tightly coupled mode is used to calculate the attitude, velocity error, position error, and IMU gyroscope error ε of the ship. b , IMU accelerometer error Clock drift clk and the clock drift rate d clk As the state quantity X of the integrated navigation filter NAV , the discrete form of the system model is as follows:

[0059] In the formula, F SINS (t) and F GNSS (t) are the state transfer matrices of SINS and GNSS systems, G SINS (t) and G GNSS (t) are the SINS and GNSS noise transformation matrices, respectively, W SINS (t) and W GNSS (t) are SINS and GNSS noise variables respectively. NAV The discrete form of the measurement model is composed of the interpolation of pseudorange and pseudorange rate of SINS and GNSS as follows:

[0060]

[0061] In the formula, H NAV,ρ (t) and are the measurement matrices of pseudorange and pseudorange rate, V NAV,ρ (t) and The measurement noise of pseudorange and pseudorange rate respectively. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Figure 1 is a flow chart of a method for integrity monitoring and positioning optimization of a shipborne navigation system in an embodiment of the present invention;

[0063] Figure 2is a diagram of a cascaded incoherent vector tracking structure used in an embodiment of the present invention;

[0064] Figure 3 is a flow chart of the RAIM algorithm used in the embodiment of the present invention. DETAILED DESCRIPTION

[0065] In order to make the technical means, creative features, objectives and effects achieved by the present invention easy to understand, the following embodiments and accompanying drawings specifically illustrate a method for monitoring the integrity and optimizing positioning of a shipborne navigation system of the present invention.

[0066] In this embodiment, a method for integrity monitoring and positioning optimization of a shipborne navigation system is provided, and an integrity monitoring technology is added, which can improve the signal tracking capability and reliability of the GNSS receiver in a harsh signal environment, and is particularly suitable for ship navigation in an environment with limited signal quality.

[0067] This embodiment provides a cascaded incoherent vector tracking loop method based on a maximum correlation entropy Kalman filter. The method uses the output of the carrier and code discriminator as the measurement vector, uses the carrier-to-noise ratio adaptive calculation and re-adjusts the measurement noise covariance matrix of the loop preprocessing filter to effectively smooth the output of the discriminator to enhance the robustness of the vector tracking loop algorithm.

[0068] The receiver autonomous integrity monitoring (RAIM) technology is introduced into the vector tracking loop (VTL), and the GNSS fault is detected by the weighted least squares residual method. Then, the improved GNSS tracking loop based on RAIM is combined with the strapdown inertial navigation system (SINS) to construct a GNSS / INS integrated navigation system. In this process, the integrated navigation results are fed back to the tracking loop to correct the error of SINS. In addition, in the integrated navigation system, RAIM is used to detect GNSS faults to improve the tracking sensitivity of the GNSS tracking loop, the anti-interference ability of the receiver, the navigation positioning accuracy, and the interference from the faulty satellite signal. It is an effective solution to improve the performance of the navigation system without increasing the hardware cost of the navigation system.

[0069] In traditional STL-based receivers, all tracking channels are independent of each other and have poor adaptability to the environment. The Vector Tracking Loop (VTL) realizes the effective integration of the originally separated signal processing and information processing in the STL. It uses the received good signal data to predict the carrier frequency and code phase of the obscured satellite signals. The carrier tracking loop usually uses a phase locked loop (PLL) for carrier tracking to obtain the Doppler frequency related to the pseudo-range rate measurement. The delay locked loop (DLL) is used for the code loop to track the pseudo-random noise (PRN) code to calculate the pseudo-range measurement. Then, the position, velocity, and time of the receiver are calculated by the obtained pseudo-range measurement and pseudo-range rate measurement. However, in some harsh signal scenarios, such as low carrier-to-noise ratio (CNR), signal obstruction, and multipath, the scalar tracking loop is prone to lose lock, resulting in increased errors in the navigation solution.

[0070] Different from the scalar tracking method, the vector tracking navigation method in this embodiment uses an integrated navigation processor to obtain navigation results. The output value carrier and code phase error of the carrier / code discriminator are linearly transformed into pseudorange rate error and pseudorange error, and then the pseudorange and pseudorange rate of the navigation receiver are obtained. The measurement value is constructed by combining the pseudorange and pseudorange rate converted from the results of the strapdown inertial navigation system to update the integrated navigation filter. At the same time, the new navigation solution is fed back to the tracking loop to update the frequency of the local carrier / code replica. The combination of the tracking loop and the integrated navigation processor realizes mutual assistance between tracking channels.

[0071] At the same time, when measuring outliers, the value of the new information Deviation will occur. In the cascaded incoherent loop preprocessing filter (Maximum Correntropy Kalman filter, MCKF) based on the maximum entropy criterion, when an abnormal value of measurement occurs, the entropy value is very small. Then L k tends to zero, The filter gain K of the pre-filter increases continuously. k Decreases, at this time, the filter will reduce the weight of the measurement value and rely more on the prior estimate, thereby reducing the impact of outliers on the estimation results, overcoming the interference of non-Gaussian noise on state estimation, and improving the measurement accuracy, thereby improving the navigation performance of the navigation receiver in poor signal scenarios. Figure 1 It is a flow chart of the shipborne navigation system integrity monitoring and positioning optimization method in an embodiment of the present invention.

[0072] like Figure 1As shown, the method for monitoring integrity and optimizing positioning of a shipborne navigation system in this embodiment includes the following steps:

[0073] Step S1, using a GNSS antenna to receive satellite signals, and performing RF front-end processing to convert the satellite signals into intermediate frequency signals.

[0074] Step S2: performing carrier separation on the intermediate frequency signal using a carrier loop to obtain a sine signal and a cosine signal.

[0075] Step S3, the sine (I) signal and cosine (Q) signal after carrier separation are brought into the integration-removal device in the code loop to obtain the coherent integration values ​​I of the leading (E), immediate (P) and lagging (L) E ,I P ,I L ,Q E ,Q P ,Q L .

[0076] Step S4, the coherent integration value is output as carrier phase error and code phase error through the loop discriminator, which are used as the quantity measurement of MCKF to estimate the code phase error, carrier phase error, carrier frequency error, and carrier frequency change rate error.

[0077] Figure 2 4 is a diagram of a cascaded incoherent vector tracking structure used in an embodiment of the present invention.

[0078] like Figure 2 As shown, the GNSS signal processing unit of the cascaded incoherent vector tracking loop based on the maximum correlation entropy Kalman filter includes: a mixer / correlator, an integrator-remover, a carrier / code phase detector, a cascaded incoherent loop preprocessing filter MCKF, a carrier / code NCO, a signal generator and an integrated navigation processor. The carrier and code phase errors output by the carrier / code phase detector are input into the cascaded incoherent loop preprocessing filter MCKF, and after smoothing, the optimal estimated phase error is obtained.

[0079] In this embodiment, based on the incoherent cascade vector tracking loop, the code phase error Δτ (chip), the carrier phase error Δθ (radian), the carrier frequency error Δf (Hz) and the carrier frequency change rate error Δα (Hz / s) are used as the state quantity X of the loop filter, and the discrete system model is as follows:

[0080] X i =[Δτ i ,Δθ i ,Δf i ,Δα i ] Τ (1)

[0081]

[0082] Where superscript i (i = 1, 2, ..., n) represents the number of the tracking channel, n is the number of all tracking channels, subscript k represents the time, k / k-1 represents the estimation of time k at time k-1, and X i represents the state quantity of channel i, Represents the state of channel i at time k. Φ kk-1 is the system state transition matrix, is the system noise vector. Coefficient β = f C / A / f L1 It is used to convert the unit of carrier phase (cycle) into the unit of C / A code (chip), f C / A is the code frequency (taking GPS L1 as an example, it is 1.023MHz), f L1 is the carrier frequency (taking GPS L1 as an example, it is 1575.42MHz). T is the update interval of the loop filter.

[0083] The measurement Z consists of the code phase error and carrier phase error output by the loop discriminator. The discrete form of the measurement model is as follows:

[0084]

[0085] Among them, Z i Indicates the quantity measurement of channel i, and Represent the code and carrier phase error measurement values ​​of channel i respectively. represents the measurement noise vector of channel i at time k, H k is the measurement matrix at time k.

[0086] The calculation formula for the code and carrier phase measurements is as follows:

[0087]

[0088] in, I E , Q E ,I L , Q L They are the coherent integration outputs of the leading and lagging in-phase and quadrature branches of the correlator respectively.

[0089] The adaptive adjustment method of the measurement noise is as follows:

[0090]

[0091]

[0092] Among them, R i,kis the measurement noise covariance matrix, and is the variance of the discriminator output, d0 is the interval between the immediate code and the delayed code, and T is the coherent integration time. The CNR of each channel is calculated by the variance sum method on the output value of the 20ms coherent integration.

[0093] Correlation entropy is used to measure the generalized similarity between two random variables. Given two random variables XY, assume that their joint distribution function is F XY (x,y). The correlation entropy between them can be expressed as:

[0094] V(X,Y)=E[κ(X,Y)]=∫κ(x,y)dF XY (x,y) (11)

[0095] Among them, E[·] is the expectation and κ(·,·) represents the kernel function.

[0096] Maximum Correntropy Criterion (MCC) is widely used in signal analysis that is interfered by diverse noises. Its effectiveness lies in its ability to effectively deal with signal distortion problems in different noise environments and improve the robustness and accuracy of signal processing.

[0097] The Gaussian kernel function is selected as the kernel function of the maximum correlation entropy. The Gaussian kernel function is expressed as follows:

[0098]

[0099] Among them, e i =x i -y i ; σ represents the kernel function width, and σ>0. In actual situations, the joint distribution F XY (x,y) is often unknown. Therefore, the sample mean estimator is used to estimate the relevant entropy:

[0100]

[0101] Among them, e(i)=x(i)-y(i), To satisfy the N samples of the joint distribution. Use Taylor series to expand the Gaussian kernel function:

[0102]

[0103] According to the above analysis, the correlation entropy is essentially the weighted sum of all even-order moments of XY, which effectively integrates the information of high-order moments. Under the condition of limited error data, the cost function based on MCC can be further expressed as:

[0104]

[0105] In order to effectively reduce the adverse effects of non-Gaussian noise on filter performance, MCC is introduced into the Kalman filter of the GNSS receiver vector tracking loop to replace the traditional minimum mean square error criterion. Under this improved framework, the optimization equation J based on MCC in the Kalman filter is m (X k ) is derived as follows.

[0106]

[0107] Among them, G σ (||XY||) is the Gaussian kernel function, Z k is the measurement vector; H k is the measurement matrix of KF at time k; X k is the state vector of KF at time k, X k-1 is the state vector of KF at time k-1, F k-1 is the state transfer matrix of KF at time k-1, is the measurement noise covariance matrix R of KF at time k k The inverse matrix of is the prior estimated covariance matrix P of KF at time k k / k-1 The inverse matrix of .

[0108] The Gaussian kernel function is Appropriate kernel bandwidth can introduce high-order moments into signal processing. To solve the maximum entropy, let:

[0109]

[0110] Take the initial approximation And in Kalman filtering, You can get:

[0111]

[0112] in,

[0113] In the basic KF,

[0114]

[0115] In MCKF, we can get:

[0116]

[0117] in, It is through L k Adjusted measurement noise covariance matrix.

[0118] When measuring outliers, the value of the innovation In MCKF, according to the Gaussian kernel function, when there is an abnormal value in the measurement, the entropy value is very small. Then L k tends to zero, The filter gain K of the pre-filter increases continuously. k Decreases, at this time, the filter will reduce the weight of the measured value and rely more on the prior estimate, thereby reducing the impact of outliers on the estimation results and overcoming the interference of non-Gaussian noise on state estimation.

[0119] The KF tracking loop is measured from the output of the PLL and DLL discriminator. The carrier phase error and code phase error are calculated using the two-quadrant inverse tangent function and the incoherent lead-minus-lag amplitude method, respectively.

[0120] The measurement vector and measurement equation of the loop can be expressed as:

[0121]

[0122]

[0123] Among them, V loop,k is the measurement noise, and the measurement noise covariance matrix In the tracking loop, it can be written as R loop,k =diag([R 1,k ,R 2,k ]).

[0124] In VTL, based on the linear relationship between code phase error and pseudorange, carrier frequency error and pseudorange rate, the measurement result of the navigation filter can be calculated as follows:

[0125]

[0126] Where, c is the speed of light (299792458m / s); f CA Refers to the code frequency (GPS L1 C / A: 1.023MHz). is the Doppler frequency shift (Hz); f L1 is the carrier frequency (GPS L1: 1575.42MHz); v usr and are the speeds of the receiver and satellite i respectively; is the receiver clock error; is the satellite clock error.

[0127] Step S5: The carrier phase error and the code phase error are linearly transformed into a pseudorange error and a pseudorange rate error, thereby obtaining a pseudorange and a pseudorange rate.

[0128] Step S6: Use the integrity monitoring technology based on the least squares residual algorithm to detect whether the satellite signal is faulty and eliminate the faulty satellite signal.

[0129] Figure 3 is a flow chart of the RAIM algorithm used in the embodiment of the present invention.

[0130] like Figure 3 As shown, in this embodiment, the integrity monitoring method based on characteristic slope weighting is used to monitor satellite failure conditions, and the satellite characteristic slope is calculated as follows:

[0131]

[0132] in, A LSR(1,i) With A LSR(2,i) Represents the matrix A LSR The first row, i-th column element and the second row, i-th column element of .

[0133] Based on the analysis of the impact of satellite characteristic slope on positioning accuracy, a differentiated weighting strategy is adopted. During the test process, satellites with larger characteristic slopes are given higher weights; conversely, satellites with smaller characteristic slopes are given smaller weights, reducing their test statistics. Based on the improved weight matrix P after characteristic slope slope The calculation is as follows:

[0134] P slope =diag(Slope1 / K * ,Slope2 / K * ,…,Slope n / K * ) (29)

[0135] in, K * is the proportionality coefficient. In this study, K * =min(Slope i ). Then the pseudorange residual vector is calculated as follows:

[0136]

[0137] The fault detection quantity of the characteristic slope weighted least squares residual algorithm is calculated as follows:

[0138]

[0139] Step S7, calculating the inertial navigation result, and constructing a tightly coupled integrated navigation processor with the pseudorange and pseudorange rate, performing integrated navigation, and obtaining a navigation result.

[0140] In this embodiment, the SINS navigation result error is calculated according to the differential equation, and the formula is as follows:

[0141]

[0142] Where φ=[φ E ,φ N ,φ U ] Τ is the ship's attitude, v n =[v N ,v E ,v U ] Τ and δv n =[δv N ,δv E ,δv U ] Τ are the speed and error of the ship to the north, east and upward respectively. (L, λ, h) and (δL, δλ, δh) are the latitude, longitude, altitude and error of the ship’s current position respectively. is the angular velocity of the navigation coordinate system (n system) relative to the inertial coordinate system (i system) in the navigation coordinate system, is the angular velocity error of the Earth's rotation, is the angular velocity error of the navigation system, is the attitude rotation matrix from the carrier coordinate system (b system) to the navigation coordinate system (n system). b is the accelerometer output result of IMU, is the measurement error of the gyroscope, δf b is the measurement error of the accelerometer, δg n is the gravity error.

[0143] Based on the tightly coupled GNSS / SINS integrated navigation processor, the ship's attitude, velocity error, position error, and IMU gyroscope error ε b , IMU accelerometer error Clock drift clk and the clock drift rate d clk As the state quantity X of the integrated navigation filter NAV , the discrete form of the system model is as follows:

[0144]

[0145] Among them, F SINS (t) and F GNSS (t) are the state transfer matrices of SINS and GNSS systems, G SINS (t) and G GNSS (t) are the SINS and GNSS noise transformation matrices, respectively, W SINS (t) and W GNSS(t) are SINS and GNSS noise variables, respectively.

[0146] Quantity measurement Z NAV The discrete form of the measurement model is composed of the interpolation of pseudorange and pseudorange rate of SINS and GNSS as follows:

[0147]

[0148]

[0149] Among them, H NAV,ρ (t) and are the measurement matrices of pseudorange and pseudorange rate, V NAV,ρ (t) and The measurement noise of pseudorange and pseudorange rate respectively.

[0150] In the pre-filter, the CNR and MCC of each channel are used to adaptively calculate and adjust the measurement noise covariance matrix of the pre-filter. That is, the improved VTL can adjust the measurement noise covariance matrix by detecting the measurement anomalies of the pre-filter, and further adjust the Kalman gain. The influence of noise and nonlinearity is reduced to a certain extent. At the same time, the phase error is linearly changed into pseudorange error and pseudorange rate error. Combined with the pseudorange prediction value and the pseudorange rate prediction value, the measurement of pseudorange and pseudorange rate is obtained and used as the measurement value of the combined navigation filter, which can improve the estimation accuracy of the navigation filter and the robustness of the navigation system. The integrity monitoring algorithm calculates the fault identification test quantity, sorts the test quantities of different satellites to eliminate the satellite corresponding to the maximum test quantity, and repeats until all faults are eliminated at the current moment. It can suppress the adverse effects of faulty satellite signals on navigation results and improve the reliability of navigation performance.

[0151] Step S8, using the navigation results and the carrier phase error to adjust the carrier numerically controlled oscillator (NCO), complete the update of the carrier frequency, the navigation solution estimate and the line of sight vector (LOS) prediction value are used to predict the bit rate of each tracking channel at the next moment, and complete the bit rate update.

[0152] In summary, the above are only preferred embodiments of the present invention and are not intended to limit the protection scope of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

[0153] Functions and Effects of the Embodiments

[0154] The present embodiment involves a method for monitoring the integrity of a shipborne navigation system and optimizing positioning, which creatively applies the cascaded incoherent vector tracking loop and the receiver autonomous integrity monitoring technology to the shipborne GNSS tracking and navigation method. It solves the problem of satellite signal failure and signal tracking loop loss of lock and navigation result divergence caused by the existing scalar tracking method in a harsh signal environment. Through the implementation method of the loop, the receiver's signal tracking capability and navigation accuracy in a harsh signal environment can be improved, which is particularly suitable for ship navigation needs in situations where the GNSS receiver frequently experiences signal interruptions and satellite signal failures.

[0155] This embodiment improves the GNSS navigation performance by improving the algorithm, and is an effective way to improve the GNSS navigation capability without increasing the hardware cost of the ship navigation system.

[0156] The above-mentioned embodiments are preferred examples of the present invention and are not intended to limit the protection scope of the present invention.

Claims

1. A method for monitoring integrity and optimizing positioning of a shipborne navigation system, characterized in that: The following steps are involved: Step 1: Receive satellite signals using a GNSS antenna and perform radio frequency front-end processing to convert the satellite signals into intermediate frequency signals; Step 2, using a carrier loop to perform carrier separation on the intermediate frequency signal to obtain a sine signal and a cosine signal; Step 3, bringing the sine signal and the cosine signal after carrier separation into an integration-removal device in a code loop to obtain leading, immediate and lagging coherent integration values; Step 4, the coherent integral value is output as a carrier phase error and a code phase error through a loop discriminator, which is used as a quantity measurement of a cascaded incoherent loop preprocessing filter to estimate the code phase error, carrier phase error, carrier frequency error, and carrier frequency change rate error; Step 5, the carrier phase error and the code phase error are linearly transformed into a pseudorange error and a pseudorange rate error, thereby obtaining a pseudorange and a pseudorange rate; Step 6, using the integrity monitoring technology based on the least squares residual algorithm to detect whether the satellite signal is faulty, and eliminate the faulty satellite signal; Step 7, calculating the inertial navigation result, and constructing a tightly coupled integrated navigation processor with the pseudorange and the pseudorange rate, performing integrated navigation, and obtaining a navigation result; Step 8, using the navigation result and the carrier phase error to adjust the carrier numerically controlled oscillator to complete the update of the carrier frequency, the navigation solution estimate and the line-of-sight vector prediction value are used to predict the bit rate of each tracking channel at the next moment to complete the bit rate update.

2. The method for integrity monitoring and positioning optimization of a shipborne navigation system according to claim 1, characterized in that: in, In step 4, based on the incoherent cascade vector tracking loop, the code phase error Δτ, carrier phase error Δθ, carrier frequency error Δf and carrier frequency change rate error Δα are used as the state quantity X of the loop filter. The discrete system model is as follows: X i =[Δτ i ,Dth i ,Δf i ,Da i ] Τ (1) In the formula, the superscript i (i = 1, 2, ..., n) represents the number of the tracking channel, n is the number of all tracking channels, the subscript k represents the time, k / k-1 represents the estimation of time k at time k-1, and X i represents the state quantity of channel i, represents the state of channel i at time k, Φ k / k-1 is the system state transition matrix, is the system noise vector, coefficient β = f C / A / f L1 Used to convert the unit of carrier phase to the unit of C / A code, f C / A is the code frequency, f L1 is the carrier frequency, T is the update interval of the loop filter, The measurement Z is composed of the code phase error and the carrier phase error output by the loop discriminator. The discrete form of the measurement model is as follows: In the formula, Z i Indicates the quantity measurement of channel i, and Respectively represent the code and carrier phase error measurement values ​​of channel i, represents the measurement noise vector of channel i at time k, H k is the measurement matrix at time k.

3. The method for integrity monitoring and positioning optimization of a shipborne navigation system according to claim 1 or 2, characterized in that: in, The calculation formulas for the code phase and carrier phase measurements are as follows: In the formula, I E , Q E ,I L , Q L They are the coherent integration outputs of the leading and lagging in-phase and quadrature branches of the correlator respectively.

4. The method for integrity monitoring and positioning optimization of a shipborne navigation system according to any one of claims 1 to 3, characterized in that: in, The adaptive adjustment method of the measurement noise is as follows: In the formula, R i,k is the measurement noise covariance matrix, and is the variance of the discriminator output, d0 is the interval between the immediate code and the delayed code, T is the coherent integration time, and the CNR of each channel is calculated by the variance sum method at the output value of the 20ms coherent integration.

5. The method for monitoring integrity and optimizing positioning of a shipborne navigation system according to claim 4, characterized in that: in, Use the correlation entropy to measure the generalized similarity between two random variables. Given two random variables XY, assume that their joint distribution function is F XY (x, y), the related entropy between them can be expressed as: V(X,Y)=E[κ(X,Y)]=∫κ(x,y)dF XY (x,y) (11) Among them, E[·] is the expectation, κ(·,·) represents the kernel function, The Gaussian kernel function is selected as the kernel function of the maximum correlation entropy. The Gaussian kernel function is expressed as follows: In the formula, e i =x i -y i ; σ represents the kernel function width, and σ>0. In actual situations, the joint distribution F XY (x, y) is often unknown, so the sample mean estimator is used to estimate the relevant entropy: In the formula, e(i)=x(i)-y(i), In order to satisfy the joint distribution of N samples, the Gaussian kernel function is expanded using Taylor series: According to the above analysis, the correlation entropy is essentially the weighted sum of all even-order moments of XY, which effectively integrates the information of high-order moments. Under the condition of limited error data, the cost function based on the maximum correlation entropy criterion can be further expressed as:

6. The method for monitoring integrity and optimizing positioning of a shipborne navigation system according to claim 5, characterized in that: in, The optimization equation J based on the maximum correlation entropy criterion in Kalman filtering m (X k ) is derived as follows: In the formula, G σ (||XY||) is the Gaussian kernel function, Z k is the measurement vector; H k is the measurement matrix of KF at time k; X k is the state vector of KF at time k, X k-1 is the state vector of KF at time k-1, F k-1 is the state transfer matrix of KF at time k-1, is the measurement noise covariance matrix R of KF at time k k The inverse matrix of is the prior estimated covariance matrix P of KF at time k k / k-1 The inverse matrix of The Gaussian kernel function is Appropriate kernel bandwidth can introduce high-order moments into signal processing and solve the maximum entropy, let: Take the initial approximation And in Kalman filtering, You can get: in, In the basic KF, In the cascaded incoherent loop preprocessing filter, we can get: in, It is through L k The adjusted measurement noise covariance matrix, When measuring outliers, the value of the innovation Deviation will occur. In the cascaded incoherent loop preprocessing filter, according to the Gaussian kernel function, when an abnormal value is measured, the entropy value is very small, then L k tends to zero, The filter gain K of the pre-filter increases continuously. k Decreases, at this time, the filter will reduce the weight of the measured value and rely more on the prior estimate, thereby reducing the impact of outliers on the estimation results and overcoming the interference of non-Gaussian noise on state estimation.

7. The method for monitoring integrity and optimizing positioning of a shipborne navigation system according to claim 6, characterized in that: in, The measurement of the KF tracking loop comes from the output of the PLL and DLL discriminator. The carrier phase error and code phase error are calculated using the two-quadrant inverse tangent function and the incoherent lead-minus-lag amplitude method, respectively. The measurement vector and measurement equation of the loop can be expressed as: Where V loop,k is the measurement noise, and the measurement noise covariance matrix In the tracking loop, it can be written as R loop,k =diag([R 1,k ,R 2,k ]), In the vector tracking loop, based on the linear relationship between the code phase error and pseudorange, and the carrier frequency error and pseudorange rate, the measurement result of the navigation filter can be calculated as follows: Where c is the speed of light (299792458 m / s); f CA refers to the code frequency, is the Doppler frequency shift (Hz); f L1 is the carrier frequency; v usr and are the speeds of the receiver and satellite i respectively; is the receiver clock error; is the satellite clock error.

8. The method for monitoring integrity and optimizing positioning of a shipborne navigation system according to claim 1, Features: in, In step 6, the integrity monitoring method based on characteristic slope weighting is used to monitor satellite failure conditions. The satellite characteristic slope is calculated as follows: In the formula, A LSR(1,i) With A LSR(2,i) Represents the matrix A LSR The first row and the i-th column element and the second row and the i-th column element of Based on the analysis of the impact of satellite characteristic slope on positioning accuracy, a differentiated weighting strategy is adopted. During the inspection process, satellites with larger characteristic slopes are given higher weights. On the contrary, the weight of the satellite with a small characteristic slope is also small, reducing its test statistic. The weight matrix P after the characteristic slope improvement is slope The calculation is as follows: In the formula, K * is the proportionality coefficient, K * =min(Slope i ), then the pseudorange residual vector is calculated as follows: The fault detection quantity of the characteristic slope weighted least squares residual algorithm is calculated as follows:

9. The method for integrity monitoring and positioning optimization of a shipborne navigation system according to claim 1, characterized in that: in, In step 7, the SINS navigation result error is calculated according to the differential equation, and the formula is as follows: In the formula, φ=[φ E ,φ N ,φ U ] Τ is the ship's attitude, v n =[v N ,v E ,v U ] Τ and δv n =[δv N ,δv E ,δv U ] Τ are the speed and error of the ship to the north, east and upward respectively. (L,λ,h) and (δL,δλ,δh) are the latitude, longitude, altitude and error of the ship’s current position respectively. is the angular velocity of the navigation coordinate system relative to the inertial coordinate system in the navigation coordinate system, is the angular velocity error of the Earth's rotation, is the angular velocity error of the navigation system, is the attitude rotation matrix from the carrier coordinate system to the navigation coordinate system, f b is the accelerometer output result of IMU, is the measurement error of the gyroscope, δf b is the measurement error of the accelerometer, δg n is the gravity error.

10. The method for integrity monitoring and positioning optimization of a shipborne navigation system according to claim 1, characterized in that: in, In step 7, the tightly coupled GNSS / SINS integrated navigation processor uses the ship’s attitude, velocity error, position error, and IMU gyroscope error ε b , IMU accelerometer error Clock drift clk and the clock drift rate d clk As the state quantity X of the integrated navigation filter NAV , the discrete form of the system model is as follows: In the formula, F SINS (t) and F GNSS (t) are the state transfer matrices of SINS and GNSS systems, G SINS (t) and G GNSS (t) are the SINS and GNSS noise transformation matrices, respectively, W SINS (t) and W GNSS (t) are SINS and GNSS noise variables, Quantity measurement Z NAV The discrete form of the measurement model is composed of the interpolation of pseudorange and pseudorange rate of SINS and GNSS as follows: In the formula, H NAV,ρ (t) and are the measurement matrices of pseudorange and pseudorange rate, V NAV,ρ (t) and The measurement noise of pseudorange and pseudorange rate respectively.

Images