Data fusion method of GBAS and ILS for precision approach

Abstract

The application relates to the field of aviation navigation and approach technology, and particularly discloses a GBAS and ILS data fusion method for precision approach, which comprises the following steps: acquiring navigation observation data of GBAS and ILS in the approach process of an airplane; determining an approach stage according to current flight parameters, adaptively switching a corresponding segmented motion model and constructing a filtering model comprising a state equation and an observation equation; performing state prediction based on the filtering model and calculating a filtering new information sequence; adopting a double-layer covariance correction mechanism to dynamically correct an observation noise covariance matrix; completing state updating on the corrected observation noise covariance matrix to obtain GBAS and ILS state estimation results; calculating dynamic weights of the GBAS and the ILS according to navigation source reliability, performing weighted fusion, and outputting airplane navigation positioning results. The application effectively suppresses ionospheric storm, multipath effect and other interferences, and improves navigation robustness and precision.

Description

Technical Field

[0001] This invention relates to the field of aviation navigation and approach technology, specifically to a GBAS and ILS data fusion method for precision approach. Background Technology

[0002] GBAS (Ground-Based Augmentation System) and ILS (Instrument Landing System) are two important precision approach and navigation systems in civil aviation, playing a crucial role in ensuring the safe and stable completion of aircraft approaches and landings.

[0003] GBAS provides positioning, navigation, and timing information based on the Global Navigation Satellite System (GNSS). It performs differential correction on satellite signals through ground reference stations and broadcasts them to aircraft via VHF Data Broadcast (VDB). It can achieve precise approach guidance with high accuracy and flexible tracks. However, its performance is highly sensitive to the space environment. Especially under conditions of strong ionospheric activity, it is easily affected by factors such as ionospheric delay anomalies, gradient abrupt changes, and scintillation effects, which can lead to increased positioning errors or even reduced availability.

[0004] In contrast, ILS uses ground-based radio beacons located on both sides and at the ends of the airport runway to provide aircraft with continuous and stable angular guidance information through localizers and glide slopes. The technology is mature and highly reliable, but its signal propagation characteristics are easily affected by factors such as complex terrain, buildings, vehicles and aircraft reflections around the airport, which can easily lead to multipath effects and introduce heading or glide slope deviations.

[0005] Because the two systems differ significantly in signal sources, error mechanisms, and environmental sensitivities—GBAS relies on GNSS signals and is susceptible to ionospheric disturbances, while ILS uses terrestrial radio beacons and is susceptible to site multipath effects—they are significantly complementary in terms of error sources and interference mechanisms. If data fusion methods can be used to achieve joint use of GBAS and ILS, it is possible to improve the overall performance of the joint navigation system, enhance the survivability of precision approach operations, and enable joint use of both systems. This would allow the joint navigation system to still provide guidance information that meets CATI accuracy requirements even under ionospheric storms and multipath interference, respectively.

[0006] In existing technologies, GBAS and ILS are typically used as independent navigation sources or combined within a simple weighted fusion framework. However, in complex interference environments: noise statistics exhibit significant non-stationarity, a single adaptive filtering method struggles to simultaneously suppress ionospheric anomalies and multipath disturbances, the aircraft's motion state undergoes phased changes during approach, fixed motion models are prone to filtering mismatch, and existing methods lack a dynamic perception mechanism for changes in the reliability of different navigation sources.

[0007] Therefore, there is an urgent need for a GBAS and ILS data fusion method that can simultaneously consider interference characteristics, changes in motion state, and the evolution of navigation source credibility. Summary of the Invention

[0008] This invention addresses the aforementioned problems in the prior art by providing a GBAS and ILS data fusion method for precision approach, effectively solving the problems of filter mismatch and lack of reliability perception under complex interference, and significantly improving navigation robustness and approach accuracy.

[0009] To achieve the above objectives, this invention proposes a GBAS and ILS data fusion method for precision approach, comprising:

[0010] S1. Acquire navigation observation data from GBAS and ILS during aircraft approach, perform time and space alignment on the navigation observation data, unify it to the airport local coordinate system, and form spatiotemporally consistent multi-source navigation input data.

[0011] S2. Determine the approach phase based on the current flight parameters of the aircraft, adaptively switch the corresponding segmented motion model, and construct a filter model containing state equations and observation equations based on the segmented motion model;

[0012] S3. Based on the filtering model, perform state prediction on spatiotemporally consistent multi-source navigation input data and calculate the filtered information sequence;

[0013] S4. A two-layer covariance correction mechanism based on innovation statistics and interference perception is adopted to dynamically correct the observation noise covariance matrix.

[0014] S5. Based on the corrected observation noise covariance matrix, complete the state update and obtain the GBAS state estimation results and ILS state estimation results respectively.

[0015] S6. Calculate the dynamic weights of GBAS and ILS based on the navigation source credibility, perform weighted fusion on the state estimation results, and output the aircraft navigation and positioning results.

[0016] Therefore, this invention proposes a GBAS and ILS data fusion method for precision approach, which has the following advantages:

[0017] (1) The anti-interference ability is significantly enhanced, and it can simultaneously suppress the influence of ionospheric storms on GBAS and the influence of multipath effects on ILS. The motion model matching is higher, and the filtering stability and accuracy are improved through phased motion modeling and model switching mechanism.

[0018] (2) The reliability of the navigation source is dynamically adjustable, enabling adaptive switching between GBAS and ILS under different interference conditions. The algorithm structure is clear and the computational complexity is controllable, making it suitable for integration into airborne navigation systems.

[0019] (3) Improve the survivability of precision approach operations. It can still meet the accuracy requirements of Category I precision approach in complex electromagnetic environments and is applicable to practical engineering systems.

[0020] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0021] Figure 1 This is a schematic diagram of the GBAS and ILS data fusion method for precision approach according to the present invention;

[0022] Figure 2 This is a schematic diagram of the two-layer covariance correction logic of the GBAS and ILS data fusion method for precision approach according to the present invention. Detailed Implementation

[0023] To make the technical solutions, advantages, and objectives of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below. The described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the described embodiments of the present invention without creative effort are within the protection scope of this application.

[0024] Unless otherwise defined, the technical or scientific terms used in this invention shall have the ordinary meaning as understood by one of ordinary skill in the art to which this invention pertains.

[0025] like Figure 1 As shown, the GBAS and ILS data fusion method for precision approach provided by this invention includes:

[0026] S1. Acquire navigation observation data from GBAS and ILS during aircraft approach, perform time and space alignment on the navigation observation data, unify it to the local coordinate system of the airport, and form spatiotemporally consistent multi-source navigation input data.

[0027] The navigation observation data includes observation data superimposed with interference factors, namely the ionospheric delay error and delay gradient change corresponding to GBAS, and the modulation difference fluctuation and measurement bias corresponding to ILS.

[0028] S2. Determine the approach phase based on the aircraft's current flight parameters, adaptively switch the corresponding segmented motion model, and construct a filtered model containing state equations and observation equations based on the segmented motion model.

[0029] The approach phase includes the intermediate approach phase, the transition phase, and the final approach phase. The switching process of the segmented motion model includes:

[0030] S21. Define the stage discrimination factor, which is obtained by real-time calculation of the aircraft's altitude change rate, vertical velocity, acceleration, or glide angle; the stage discrimination factor is:

[0031] ;

[0032] In the formula, , , These are the weighting coefficients. Indicates the time of the aircraft Overall approach maneuver intensity. The vertical velocity component of the aircraft corresponds to the rate of descent. This represents the vertical acceleration component of the aircraft. This represents the angle between the aircraft's current flight path and the horizontal plane; it is usually a negative value, representing the glide slope angle. This represents the rate of change of the glide slope angle.

[0033] S22, When the stage discriminant factor satisfies At that time, a constant acceleration motion model is adopted; satisfying At that time, a constant velocity motion model is adopted; satisfying At that time, a weak maneuvering model is adopted; This is the first threshold parameter used to determine whether the aircraft is in the intermediate approach phase. This is the second threshold parameter used to determine whether an aircraft has entered the final approach phase.

[0034] The filtering models include:

[0035] The state transition equation for the constant acceleration motion model is:

[0036] ;

[0037] ;

[0038] The state transition equation for the constant velocity motion model is:

[0039] ;

[0040] ;

[0041] The state transition equation for the weak maneuvering model is:

[0042] ;

[0043] In the formula, Let be the aircraft state vector at the k-th sampling time. Let be the aircraft state vector at the (k-1)th sampling time. , , These are the state transition matrices for the corresponding motion models. This is the process noise vector. , These are the process noise input matrices for the corresponding motion model, and T is the system sampling period.

[0044] S3. Based on the filtering model, perform state prediction on spatiotemporally consistent multi-source navigation input data and calculate the filtered information sequence.

[0045] The process of calculating the filtered innovation sequence for state prediction based on spatiotemporally consistent multi-source navigation input data using a filtering model includes:

[0046] S31. Based on the state transition matrix corresponding to the current approach phase, perform state prediction:

[0047] ;

[0048] In the formula, This represents the predicted state of the motion model at the previous moment. For the predicted state based on the motion model, This is the state transition matrix, the specific form of which is determined by the current flight phase, as described above. , , One of them.

[0049] S32. Calculate the state prediction error covariance matrix:

[0050] ;

[0051] In the formula, Here is the state prediction error covariance matrix. Let $\mathbf{a}$ be the state estimation error covariance matrix at the previous time step. Let be the process noise covariance matrix.

[0052] S33. Calculate the filtered innovation vector based on the observation equation:

[0053] ;

[0054] In the formula, Let be the information vector at time k. For the observation vector, This is the observation matrix.

[0055] S4. A two-layer covariance correction mechanism based on innovation statistics and interference perception is adopted to dynamically correct the observation noise covariance matrix.

[0056] like Figure 2 As shown, the first-level innovation-driven correction process of the two-level covariance correction mechanism includes:

[0057] S411. Define the forgetting factor:

[0058] ;

[0059] In the formula, The exponential decay coefficient is... For adaptive weights.

[0060] S412. First-level correction of observation noise covariance based on the innovation sequence:

[0061] ;

[0062] In the formula, The first-level corrected observation noise covariance at the current moment. Let be the current information vector at time k, and let be the information covariance component caused by the uncertainty of state prediction.

[0063] The second-level disturbance sensing correction process of the two-level covariance correction mechanism includes:

[0064] S421. Define the integrated interference sensing factor:

[0065] ;

[0066] In the formula, Let be the comprehensive interference sensing factor at time k. This represents the ionospheric delay gradient or ionospheric delay rate of change corresponding to GBAS. Let be the ILS modulation difference change, and be the filter innovation vector. , , The weighting coefficients are satisfied. .

[0067] S422. Constrain and correct the first-layer correction result based on the interference perception factor:

[0068] ;

[0069] In the formula, This is the observation noise covariance matrix ultimately used for filtering updates. This is the adjustment coefficient.

[0070] S423. When the comprehensive interference sensing factor is less than the preset interference discrimination threshold When the noise level is greater than or equal to the preset interference discrimination threshold, the corresponding observation noise covariance is frozen or amplified.

[0071] The two-level covariance correction mechanism also needs to satisfy the positive definiteness constraint of covariance:

[0072] ;

[0073] In the formula, The square of the energy or magnitude of the current innovation vector. The theoretical innovation covariance is determined by both the uncertainty of state prediction and the covariance of observation noise.

[0074] S5. Based on the corrected observation noise covariance matrix, complete the state update and obtain the GBAS state estimation results and ILS state estimation results respectively.

[0075] The state update process includes:

[0076] S51. Calculate the Kalman gain:

[0077] ;

[0078] In the formula, Update the weight matrix for the state.

[0079] S52, Performing state updates based on Kalman gain:

[0080] ;

[0081] In the formula, For the predicted state based on the motion model, The correction amount is provided by the observation information. This is the optimal state estimate after integrating GBAS and ILS.

[0082] S6. Calculate the dynamic weights of GBAS and ILS based on the reliability of the navigation source, perform weighted fusion on the state estimation results, and output the aircraft navigation and positioning results.

[0083] The formula for calculating the dynamic weight of GBAS based on navigation source credibility is as follows:

[0084] ;

[0085] In the formula, The GBAS observation noise covariance matrix is... This is the interference sensing factor corresponding to GBAS.

[0086] The formula for calculating the dynamic weight of ILS based on navigation source credibility is as follows:

[0087] ;

[0088] In the formula, The ILS observation noise covariance matrix, This is the interference sensing factor corresponding to ILS.

[0089] The weighted fusion process includes:

[0090] S61. Normalize the dynamic weights of GBAS and ILS calculated based on navigation source credibility:

[0091] ;

[0092] ;

[0093] In the formula, , These are the normalized GBAS weights and ILS weights, respectively. .

[0094] S62. Perform linear fusion based on normalized weights to obtain the final localization result:

[0095] ;

[0096] In the formula, For the final navigation and positioning results, The output is the position calculated by GBAS at time k. The output is the position or equivalent deviation calculated by ILS at time k.

[0097] This invention uses real flight state data during aircraft approach and landing phases acquired by the flight simulation platform X-Plane as the true reference for navigation, and generates GBAS and ILS navigation observation data containing ionospheric storm and multipath effect interference. This is suitable for verifying the effectiveness of the method under complex interference. The specific implementation process is as follows:

[0098] Data preparation: The aircraft position, velocity and other parameters are obtained through X-Plane as the ground truth, and GBAS and ILS observation data are generated, and interference such as ionospheric delay error and modulation difference fluctuation are superimposed respectively.

[0099] Spatiotemporal alignment: Aligning the interfered and undisturbed navigation data in time and space, and unifying them to the local coordinate system of the airport.

[0100] Phase discrimination: Calculate parameters such as the rate of change of aircraft altitude to obtain the phase discrimination factor, determine the intermediate approach, transition or final approach phase, and switch the corresponding motion model.

[0101] Filtered prediction: Construct a filtered model based on the selected motion model, perform state prediction, and calculate the innovation sequence.

[0102] Covariance correction: First, the first layer of correction is completed through the statistical properties of the innovation. Then, the perception factor is constructed by combining interference indicators such as the ionospheric delay gradient to carry out the second layer of constraint correction.

[0103] State update and fusion: The state update is completed by calculating the Kalman gain, the dynamic weights are calculated based on the observation noise covariance and the interference sensing factor, and the final navigation result is obtained by linear fusion after normalization.

[0104] Results verification: Using mean square error as an indicator, the fusion results were verified to meet the Type I precision approach accuracy requirements in both the horizontal and vertical directions.

[0105] Therefore, this invention provides a GBAS and ILS data fusion method for precision approach. By aligning multi-source data in time and space, performing phased motion modeling, correcting bi-layer covariance, and dynamically weighting fusion, it accurately suppresses ionospheric storms and multipath interference, dynamically matches changes in flight state, and evaluates the reliability of navigation sources in real time. Even in complex environments, it can still output navigation results that meet the requirements of Category I precision approach, improve the robustness and survivability of the system, and is suitable for integration of airborne navigation systems.

[0106] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for fusing GBAS and ILS data for precision approach, characterized in that, include: S1. Acquire navigation observation data from GBAS and ILS during aircraft approach, perform time and space alignment on the navigation observation data, unify it to the local coordinate system of the airport, and form spatiotemporally consistent multi-source navigation input data. S2. Determine the approach phase based on the current flight parameters of the aircraft, adaptively switch the corresponding segmented motion model, and construct a filter model containing state equations and observation equations based on the segmented motion model; S3. Based on the filtering model, perform state prediction on spatiotemporally consistent multi-source navigation input data and calculate the filtered information sequence; S4. A two-layer covariance correction mechanism based on innovation statistics and interference perception is adopted to dynamically correct the observation noise covariance matrix. S5. Based on the corrected observation noise covariance matrix, complete the state update and obtain the GBAS state estimation results and ILS state estimation results respectively. S6. Calculate the dynamic weights of GBAS and ILS based on the navigation source credibility, perform weighted fusion on the state estimation results, and output the aircraft navigation and positioning results.

2. The GBAS and ILS data fusion method for precision approach according to claim 1, characterized in that, In S1, the navigation observation data includes observation data superimposed with interference factors, namely the ionospheric delay error and delay gradient change corresponding to GBAS, and the modulation difference fluctuation and measurement deviation corresponding to ILS.

3. The GBAS and ILS data fusion method for precision approach according to claim 1, characterized in that, In S2, the approach phase includes the intermediate approach phase, the transition phase, and the final approach phase. The switching process of the segmented motion model includes: S21. Define the stage discrimination factor, wherein the stage discrimination factor is: ； In the formula, , , These are the weighting coefficients. Indicates the time of the aircraft Overall approach maneuver strength, This represents the aircraft's velocity component in the vertical direction. This represents the vertical acceleration component of the aircraft. This indicates the angle between the aircraft's current flight path and the horizontal plane. Indicates the rate of change of the glide angle; S22, when the stage discrimination factor satisfies At that time, a constant acceleration motion model is adopted; satisfying At that time, a constant velocity motion model is adopted; satisfying At that time, a weak maneuvering model was adopted; This is the first threshold parameter used to determine whether an aircraft is in the intermediate approach phase. This is a second threshold parameter used to determine whether an aircraft has entered the final approach phase; The filtering models include: The state transition equation for the constant acceleration motion model is: ； ； The state transition equation for the constant velocity motion model is: ； ； The state transition equation for the weak maneuvering model is: ； In the formula, Let be the aircraft state vector at the k-th sampling time. Let be the aircraft state vector at the (k-1)th sampling time. , , These are the state transition matrices for the corresponding motion models. This is the process noise vector. , These are the process noise input matrices for the corresponding motion model, and T is the system sampling period.

4. The GBAS and ILS data fusion method for precision approach according to claim 3, characterized in that, In S3, the process of predicting the state of spatiotemporally consistent multi-source navigation input data based on the filtering model and calculating the filtered innovation sequence includes: S31. Based on the state transition matrix corresponding to the current approach phase, perform state prediction: ； In the formula, This represents the predicted state of the motion model at the previous moment. For the predicted state based on the motion model, The state transition matrix is ​​as described above. , , One of them; S32. Calculate the state prediction error covariance matrix: ； In the formula, Here is the state prediction error covariance matrix. Let $\mathbf{a}$ be the state estimation error covariance matrix at the previous time step. The process noise covariance matrix; S33. Calculate the filtered innovation vector based on the observation equation: ； In the formula, Let be the information vector at time k. For the observation vector, This is the observation matrix.

5. The GBAS and ILS data fusion method for precision approach according to claim 4, characterized in that, In S4, the first-level innovation statistically driven correction process of the two-level covariance correction mechanism includes: S411. Define the forgetting factor: ； In the formula, The exponential decay coefficient is... For adaptive weights; S412. First-level correction of the observation noise covariance matrix based on the innovation sequence: ； In the formula, This is the observation noise covariance matrix after the first layer correction at the current time.

6. The GBAS and ILS data fusion method for precision approach according to claim 5, characterized in that, In S4, the second-layer interference sensing correction process of the two-layer covariance correction mechanism includes: S421. Define the integrated interference sensing factor: ； In the formula, Let be the comprehensive interference sensing factor at time k. This represents the ionospheric delay gradient or ionospheric delay rate of change corresponding to GBAS. Let be the ILS modulation difference change, and be the filter innovation vector. , , These are the weighting coefficients; S422. Constrain and correct the first-layer correction result based on the interference perception factor: ； In the formula, This is the final observation noise covariance matrix used for filtering updates. This is the adjustment coefficient; S423. When the comprehensive interference sensing factor is less than the preset interference discrimination threshold When the noise level is greater than or equal to the preset interference discrimination threshold, the corresponding observation noise covariance matrix is ​​frozen or amplified.

7. The GBAS and ILS data fusion method for precision approach according to claim 6, characterized in that, In S4, the two-level covariance correction mechanism must also satisfy the positive definiteness constraint of covariance: 。 8. The GBAS and ILS data fusion method for precision approach according to claim 7, characterized in that, In S5, the state update process includes: S51. Calculate the Kalman gain: ； In the formula, Update the weight matrix for the state; S52. Perform state update based on the Kalman gain: ； In the formula, This is the optimal state estimate after integrating GBAS and ILS.

9. The GBAS and ILS data fusion method for precision approach according to claim 8, characterized in that, In S6, the formula for calculating the dynamic weight of GBAS based on the navigation source credibility is as follows: ； In the formula, The GBAS observation noise covariance matrix is... The interference sensing factor corresponding to GBAS; The formula for calculating the dynamic weight of ILS based on navigation source credibility is as follows: ； In the formula, The ILS observation noise covariance matrix, This is the interference sensing factor corresponding to ILS.

10. The GBAS and ILS data fusion method for precision approach according to claim 9, characterized in that, In S6, the weighted fusion process includes: S61. Normalize the dynamic weights of GBAS and ILS calculated based on navigation source credibility: ； ； In the formula, , These are the normalized GBAS weights and ILS weights, respectively. ; S62. Perform linear fusion based on normalized weights to obtain the final localization result: ； In the formula, For the final navigation and positioning results, The output is the position calculated by GBAS at time k. The output is the position or equivalent deviation calculated by ILS at time k.

Images