Sensor data fusion system with noise reduction and fault protection

Abstract

Sensor data fusion system with noise reduction and fault protection. A sensor data fusion system (20) fuses data acquired by respective accelerometers having different properties. For example, when measuring a specified force, one accelerometer (1) has low noise and high bias, while another accelerometer (2) has high noise and low bias. The high noise, low bias accelerometer can be a gravimeter. The gravimeter measures the same physical variable as a conventional accelerometer, i.e., the specified force. By combining the expensive gravimeter with a low cost accelerometer, a composite sensor having low noise and low bias can be achieved. The composite sensor can be utilized in a reference navigation system (10) for gravity anomalies to achieve improved navigation performance.

Description

Technical Field

[0001] This disclosure relates in general to sensors used in geospatial positioning and navigation systems, and more specifically to synthetic sensors configured to produce fused sensor data representing measurements of the same physical variable (e.g., acceleration due to gravity). Background Technology

[0002] Gravity anomaly reference navigation systems provide navigation assistance by referencing the position of a platform relative to a gravity anomaly map / database. Typical navigation systems determine the navigation scheme based on measurements from an inertial measurement unit (IMU). However, IMUs introduce time-increasing errors due to integration within the inertial navigation algorithm. Gravity anomaly reference navigation systems are configured to combine navigation solutions based on IMU measurements with gravity anomaly measurements. Typically, gravity anomaly measurements are obtained using a gravimeter (an instrument used to measure a "specified force"). The specified force is defined as the non-gravity force per unit mass, expressed in meters per second squared (m / sec). 2 (That is, the unit of acceleration) measures a specified force (also known as "g-force"). Thus, the specified force is a type of acceleration.

[0003] Global Positioning System (GPS) and Global Navigation Satellite System (GNSS) are more widely used than gravity anomaly reference navigation systems. However, GPS and GNSS signals are more susceptible to external interference compared to gravity anomaly reference navigation systems. Specifically, GPS and GNSS signals have low power, making them vulnerable to interference, even at very low power levels. For example, GPS can be denial due to interference known as "congestion." In contrast, gravity anomaly reference navigation systems are resistant to interference. Therefore, when GPS and GNSS systems are unavailable (among various other "complementary" or "alternative" positioning, navigation, and tracking (PNT) schemes such as vision-based navigation, celestial navigation, terrain-referenced navigation, and opportunistic signal-based PNT), gravity anomaly reference navigation systems can be used as alternatives or supplementary solutions.

[0004] In a specific gravity anomaly reference navigation system, the gravimeter has an unacceptable noise level (e.g., 25 micro-g / sqrt(Hz)), but an acceptable bias level (e.g., 1 micro-g). (Unit: g equals 9.80665 m / sec) 2 。 ) Summary of the Invention

[0005] The subject matter disclosed in the following details relates to a sensor data fusion system that provides noise reduction and fault protection. According to some embodiments, the sensor data fusion system fuses data acquired from corresponding accelerometers with different properties. In one embodiment, when measuring a specified force, one accelerometer has low noise and high bias, while another accelerometer has high noise and low bias. In one proposed implementation, the high-noise, low-bias accelerometer is a gravimeter.

[0006] Gravimeters and conventional accelerometers measure the same physical variable. Conventional accelerometers typically have very low noise (referred to as rate random walk; e.g., 1 micro-g / sqrt(Hz)), but have a higher level of bias repeatability / stability (e.g., 30 micro-g) than the corresponding level of a gravimeter. By combining an expensive gravimeter with a low-cost accelerometer, a synthetic sensor system with low noise and low bias can be realized (hereinafter referred to as a "sensor data fusion system").

[0007] The subject matter disclosed in the following details further relates to a gravity anomaly reference navigation system that uses the sensor data fusion system to achieve improved navigation performance. More specifically, the measurement data from a noisy gravimeter and a low-noise accelerometer are mixed to achieve two benefits: (1) reduced noise and improved performance; and (2) fault protection.

[0008] As used herein, the term "sensor data fusion" refers to combining sensor data to create information with lower uncertainty than using sensors individually. In this context, "reduced uncertainty" means more accurate, more complete, or more reliable. As used herein, the terms "high" and "low" are relative, not absolute. For example, high noise is greater than low noise, and high bias is greater than low bias.

[0009] Although various embodiments of a sensor data fusion system for measuring a specified force applied to a mobile platform by means of navigation will be described in some details below, one or more of its embodiments may be characterized as one or more of the following embodiments.

[0010] One embodiment of the subject matter disclosed below is a sensor data fusion system, comprising: a first accelerometer having low noise and a high bias voltage; a second accelerometer having high noise and a low bias voltage (e.g., a gravimeter); a first subtractor connected to receive signals from the first accelerometer and signals from the second accelerometer and configured to output a first difference signal representing the difference between the signals from the first accelerometer and the signals from the second accelerometer; a first filter connected to receive the first difference signal output by the first subtractor and configured to output a signal representing an estimated relative bias voltage between the first accelerometer and the second accelerometer; and a second subtractor connected to receive signals from the first accelerometer and signals from the first filter and configured to output a second difference signal representing the difference between the signals from the first accelerometer and the signals from the first filter. The second difference signal has low noise and a low bias voltage.

[0011] According to some embodiments, the sensor data fusion system described in the preceding paragraph further includes: a fault detector connected to receive a first difference signal, and configured to output a true fault signal if the first difference signal persistently exceeds a specified difference threshold.

[0012] According to other embodiments, the sensor data fusion system further includes: a third accelerometer having low noise and high bias; a third subtractor connected to receive phenotypic signals from the second and third accelerometers and configured to output a third difference signal representing the difference between the signal from the second accelerometer and the signal from the third accelerometer; a second filter connected to receive the third difference signal output by the third subtractor and configured to output a signal representing an estimated relative bias between the second and third accelerometers; and a fourth subtractor connected to receive the signal from the third accelerometer and the signal from the second filter and configured to output a fourth difference signal representing the difference between the signal from the third accelerometer and the signal from the second filter. Furthermore, the sensor data fusion system may include: a first fault detector connected to receive a first difference signal, configured to output a true fault signal if the first difference signal persistently exceeds a specified difference threshold, otherwise outputting a false fault signal; a second fault detector connected to receive a third difference signal, configured to output a true fault signal if the third difference signal persistently exceeds a specified difference threshold, otherwise outputting a false fault signal; and a processing device connected to the first fault detector, the second fault detector, a third subtractor, and a fourth subtractor. The processing device is configured to: output the first difference signal if the first fault detector outputs a false fault signal and the second fault detector outputs a true fault signal; and output a second difference signal if the first fault detector outputs a true fault signal and the second fault detector outputs a false fault signal. In one proposed implementation, if the first fault detector and the second fault detector output corresponding false fault signals, the processing device is further configured to output a weighted sum of the first difference signal and the second difference signal. Furthermore, if the first fault detector and the second fault detector output corresponding true fault signals, the processing device can be further configured to output a weighted sum of the signals output by the first accelerometer and the third accelerometer with previously estimated bias voltages.

[0013] Another embodiment of the subject matter disclosed below is a method for fusing sensor data, comprising: using a first measurement signal representing a first measured value of a specified force in the vertical direction of a platform's reference frame, outputting a first accelerometer having low noise and high bias; using a second measurement signal representing a second measured value of a specified force in the vertical direction of the platform's reference frame, outputting a second accelerometer having high noise and low bias; generating a first difference signal representing the difference between the first measurement signal and the second measurement signal; filtering the first difference signal to generate a bias estimation signal representing estimated relative biases of the first and second accelerometers; and generating a second difference signal representing the difference between the first measurement signal and the bias estimation signal. The method may further include: using the first difference signal during navigation if the first difference signal does not persistently exceed a specified difference threshold.

[0014] Another embodiment of the subject matter disclosed below is a system for a navigation platform, comprising: an inertial navigation system configured to generate a navigation scheme and including an accelerometer having low noise and high bias; a gravimeter having high noise and low bias; a guidance and control system communicatively coupled to the inertial navigation system and configured to control the platform according to the navigation scheme; a sensor data fusion module configured to generate a bias estimation signal representing a relative bias estimate from the outputs of the accelerometer and gravimeter, and then generate a signal representing a corrected output of the accelerometer derived by subtracting the bias estimation signal from the output of the accelerometer; and a time-matching buffer communicatively coupled to the inertial navigation system and configured to store information representing the platform's position, velocity, and... Attitude data; a gravity anomaly database, stored in a non-volatile tangible computer-readable storage medium; a gravimeter output predictor, communicatively coupled to receive time-stamped position, velocity, and attitude from a time-matched buffer and retrieve gravity anomaly data from the gravity anomaly database and configured to output a prediction signal representing a prediction of the gravimeter output; a residual and matrix calculation module, communicatively coupled to receive the prediction signal and the accelerometer correction output and configured to generate residuals, an H matrix, and an R matrix based on the difference between the prediction signal and the accelerometer correction output; and a Kalman filter, configured to generate a position correction based on the difference between the prediction signal and the accelerometer correction output and then send the position correction to the inertial navigation system.

[0015] Other embodiments of a sensor data fusion system that utilizes navigation measurements to apply a specified force to a mobile platform are disclosed below. Attached Figure Description

[0016] The features, functions, and advantages discussed in the foregoing sections can be implemented independently or in combination with other embodiments. For the purpose of illustrating the above and other embodiments, various implementations will now be described with reference to the accompanying drawings. Any figures briefly described in this section are not drawn to scale.

[0017] Figure 1 It is a block diagram identifying the components of a sensor data fusion system according to an embodiment including a low-noise, high-bias accelerometer, a high-noise, low-bias accelerometer (e.g., a gravimeter), and a fault detector.

[0018] Figure 2 It is used for identification and detection Figure 1 The flowchart shows the steps of a method for handling faults during the operation of a sensor data fusion system as described in the document.

[0019] Figure 3It is a block diagram identifying the components of a sensor data fusion system according to an alternative embodiment, including two low-noise, high-bias accelerometers, one high-noise, low-bias accelerometer (e.g., a gravimeter), two fault detectors, and a processing device (or computing device).

[0020] Figure 4 This is a diagram illustrating the architecture of a system that enhances inertial navigation using gravity anomaly-based navigation according to an exemplary embodiment.

[0021] The following reference will be made to the accompanying drawings, in which similar elements in different drawings have the same reference numerals. Detailed Implementation

[0022] An illustrative embodiment of a sensor data fusion system that uses navigation measurements to determine the force applied to a mobile platform is described below in some detail. However, this specification does not describe all features of the actual implementation. Those skilled in the art will recognize that in developing any such actual embodiment, several implementation-specific decisions must be made to achieve the developer's specific goals, such as compatibility with system-related and business-related constraints (variations from one implementation to another). Moreover, it should be recognized that such development efforts can be complex and time-consuming, but needless to say, remain routine work for those of ordinary skill in the art who will benefit from this disclosure.

[0023] Figure 1 This is a block diagram identifying the components of a sensor data fusion system 20 according to one embodiment, including a low-noise, high-bias accelerometer 1, a high-noise, low-bias accelerometer 2 (e.g., a gravimeter), and a fault detector 6. The output of the low-noise, high-bias accelerometer 1 is represented on the platform (…). Figure 1 A first measurement signal representing a first measured value of a force specified in the direction perpendicular to a reference frame (not shown). A second measurement signal representing a second measured value of a force specified in the direction perpendicular to the reference frame of the platform. The output of low-noise, high-bias accelerometer 1 is the primary output. The bias voltage of low-noise, high-bias accelerometer 1 is estimated using high-noise, high-bias accelerometer 2. (Here, low-noise, high-bias accelerometer 1 and high-noise, low-bias accelerometer 2 are collectively referred to as "accelerometer 1 and 2".)

[0024] Figure 1The sensor data fusion system 20 depicted further includes: a first subtractor 3, connected to receive a signal from a low-noise, high-bias accelerometer 1 and configured to output a first difference signal representing the difference between the signals output by accelerometers 1 and 2. Furthermore, the sensor data fusion system 20 includes: a filter 4 (e.g., a low-pass filter), connected to receive the first difference signal output by the first subtractor 3 and configured to output a bias estimation signal representing the estimated relative bias of accelerometers 1 and 2. The sensor data fusion system 20 further includes: a second subtractor 5, connected to receive the signal from the low-noise, high-bias accelerometer 1 and the bias estimation signal from the filter 4, and configured to output a second difference signal representing the difference between the signal from the low-noise, high-bias accelerometer 1 and the bias estimation signal. The second subtractor 5 corrects the output of the low-noise, high-bias accelerometer 1 using the estimated relative bias between accelerometers 1 and 2. The second difference signal has low noise and low bias.

[0025] In addition, if any sensor outputs a fault sample, simple fault protection can also be implemented. Figure 1 The sensor data fusion system 20 described further includes: a fault detector 6, connected to receive a first difference signal from the first subtractor 3. If the first difference signal persistently exceeds a specified difference threshold, the fault detector 6 is configured to output a true fault signal; otherwise, it outputs a false fault signal. In the latter case, via the navigation system ( Figure 1 (Not shown in the image) Approval to use the first difference signal.

[0026] Figure 2 It is used to identify and depict Figure 1The flowchart illustrates the steps of a fault detection method 40 during the operation of the sensor data fusion system 20, as depicted in the diagram. As previously described, the first subtractor 3 determines the difference between the accelerometer outputs (step 41). Then, the fault detector 6 compares this difference with a specified difference threshold indicating a fault (step 42). Next, it is determined whether the difference between the accelerometer outputs is greater than the specified difference threshold (step 43). Alternatively, if it is determined in step 43 that the difference between the accelerometer outputs is not greater than the specified difference threshold, the bias estimation is updated (step 44). In this case, the persistence counter is set to zero (step 45), and the process returns to step 41. Alternatively, if it is determined in step 43 that the difference between the accelerometer outputs is greater than the specified difference threshold, the bias estimation is not updated. Instead, the accelerometer output is rejected (step 46). Then, the persistence counter is incremented by 1 (step 47). Then, the fault detector 6 compares the persistence count with a specified count threshold (step 48). Next, it is determined whether the persistence count is greater than the specified count threshold (step 49). On the other hand, if it is determined in step 49 that the persistence count is not greater than the specified counting threshold, the process returns to step 41. On the other hand, if it is determined in step 49 that the persistence count is greater than the specified counting threshold, a fault is declared (step 50).

[0027] Figure 1 and Figure 2 The sensor data fusion method described herein results in the following: the bias voltage of the "synthetic sensor" is approximately the same as the bias voltage level of the high-noise, low-bias accelerometer 2 (which may be a gravimeter). The noise level of the "synthetic sensor" is approximately the same as the noise level of the low-noise, high-bias accelerometer 1.

[0028] The high-bias, low-noise accelerometer 1 can have its output modeled as a conventional navigation-grade accelerometer as follows:

[0029] a HL =a+b H +v L

[0030] Where a is the actual specified force, a HL It is the specified force being measured, b H It is a bias voltage, and v L It's noise.

[0031] The low-bias, high-noise accelerometer 2 can have its output modeled as a gravimeter as follows:

[0032] a LH =a+b L +v H

[0033] Among them, aLH It is the specified force being measured, b L It is a bias voltage, and v H It's noise.

[0034] Subtractor 3 is a differential circuit, where the two outputs a are... HL and a LH There are differences, which can be modeled as follows:

[0035] δa=a HL -a LH =(b H -b L )+(v L +v H )

[0036] Filter 4 can be a hold-up filter (b) H -b L A low-pass filter that slowly changes its value and filters out most of the noise. After low-pass filtering, the filter output can be modeled as follows:

[0037] δa F =LPF(δa)≈(b H -b L )

[0038] Subtractor 5 uses "estimate relative bias" to adjust the output a from the high-bias, low-noise accelerometer 1. HL The correction, resulting in an output with lower bias and lower noise, can be modeled as follows:

[0039] a LL =a HL -δa F ≈a+b H +v L -(b H -b L ) = a + b L +v L

[0040] Fault detector 6 is configured to perform fault detection by checking the relative difference δa between the upper and lower limits. If the value exceeds any limit, the persistence counter is incremented by one and the filter 4 is also instructed not to use samples as a "containment measure". If the persistence count exceeds a specified count threshold, a fault is detected and declared to have occurred.

[0041] For the purpose of simplifying the description below, the term "fusion scheme" will be defined as follows:

[0042] [fused_output, bias_est, fault] = fused_solution(accelerometer, gravimeter)

[0043] According to an alternative embodiment, filter 4 can be a Kalman filter instead of a low-pass filter. According to other embodiments, complementary filtering can also be used. For example, the output of the low-noise, high-bias accelerometer 1 can be filtered using a high-pass filter, while the output of the high-noise, low-bias accelerometer 2 can be filtered using a low-pass filter. The combination provides optimized performance.

[0044] Figure 3 This is a block diagram identifying the components of a sensor data fusion system 30, including two low-noise, high-bias accelerometers 1 and 1a according to alternative embodiments, a high-noise, low-bias accelerometer 2 (e.g., a gravimeter), two filters 4 and 4a, two fault detectors 6 and 6a, and a processing device 8 (or computing device). The inclusion of two low-noise, high-bias accelerometers 1 and 1a allows voting not only to detect faults but also to isolate them, and also allows for further noise reduction.

[0045] The output of the low-noise, high-bias accelerometer 1 is represented on the platform ( Figure 3 A first measurement signal representing a first measured value of a force specified in the vertical direction of a reference frame (not shown). A second measurement signal representing a second measured value of a force specified in the vertical direction of the platform's reference frame, output using a high-noise, low-bias accelerometer 2 (e.g., a gravimeter). A third measurement signal representing a third measured value of a force specified in the vertical direction of the platform's reference frame, output using a low-noise, high-bias accelerometer 1a.

[0046] Figure 3 The sensor data fusion system 30 depicted further includes: a first subtractor 3 connected to receive a signal from a low-noise, high-bias accelerometer 1. The first subtractor 3 is configured to output a first difference signal representing the difference between the signals output by accelerometers 1 and 2. A filter 4 (e.g., a low-pass filter) is connected to receive the first difference signal output by the first subtractor 3. The filter 4 is configured to output a bias estimate signal representing the estimated relative bias of accelerometers 1 and 2. The sensor data fusion system 30 further includes: a second subtractor 5 connected to receive a signal from the low-noise, high-bias accelerometer 1 and receive the bias estimate signal from the filter 4. The second subtractor 5 is configured to output a second difference signal representing the difference between the signal from the low-noise, high-bias accelerometer 1 and the bias estimate signal from the filter 4. The second subtractor 5 corrects the output of the low-noise, high-bias accelerometer 1 using the estimated relative bias between accelerometers 1 and 2. The second difference signal has low noise and low bias.

[0047] Furthermore, the sensor data fusion system 30 includes: a third subtractor 3a connected to receive a signal from a low-noise, high-bias accelerometer 1a. The third subtractor 3a is configured to output a third difference signal representing the difference between the signals output by accelerometers 1a and 2. A filter 4a (e.g., a low-pass filter) is connected to receive the third difference signal output by the third subtractor 3a. The filter 4a is configured to output a bias estimation signal representing the estimated relative bias of accelerometers 1a and 2. The sensor data fusion system 30 further includes: a fourth subtractor 5a connected to receive a signal from the low-noise, high-bias accelerometer 1a and the bias estimation signal from the filter 4a. The fourth subtractor 5a is configured to output a fourth difference signal representing the difference between the signal from the low-noise, high-bias accelerometer 1a and the bias estimation signal from the filter 4a. The fourth subtractor 5a corrects the output of the low-noise, high-bias accelerometer 1a using the estimated relative bias between accelerometers 1a and 2. The fourth difference signal has low noise and low bias.

[0048] Furthermore, if any sensor outputs a fault sample, a fault protection algorithm is executed to determine the appropriate response. Figure 3 The sensor data fusion system 20 described herein further includes: a fault detector 6, connected to receive a first difference signal from a first subtractor 3; and a fault detector 6a, connected to receive a third difference signal from a third subtractor 3a. If the first difference signal persistently exceeds a specified difference threshold, the fault detector 6 is configured to output a true fault signal; otherwise, it outputs a false fault signal. Similarly, if the third difference signal persistently exceeds a specified difference threshold, the fault detector 6a is configured to output a true fault signal; otherwise, it outputs a false fault signal. (As used herein, the term "false fault signal" refers to a signal indicating a state of "no fault declared"; conversely, the term "true fault signal" refers to a signal indicating a state of "fault declared".)

[0049] According to one proposed implementation, the processing device 8 is configured to execute an algorithm that selects one of three available schemes based on the accelerometer's fault state detected by fault detectors 6 and 6a. Using the previously described notation, schemes 1, 2, and 3 are defined as follows:

[0050] [Solution 1, bias_est1, fault1] = fused_solution(accelerometer 1, gravimeter)

[0051] [Solution 2, bias_est2, fault2] = fused_solution(accelerometer 2, gravimeter)

[0052] Option 3 = (w) Option 1 + (1 – w) Option 2

[0053] Here, w is a weighting factor with values ​​in the range between 0 and 1.

[0054] When executed by processing device 8, the solution selection algorithm selects one of the following responses based on the fault conditions:

[0055] (1) If fault detectors 6 and 6a output corresponding false fault signals, the processing device 8 outputs a weighted sum of the first difference signal and the second difference signal for use by the navigation system (Scheme 3).

[0056] (2) If fault detector 6 outputs a true fault signal and fault detector 6a outputs a false fault signal, the processing device 8 outputs a fourth difference signal for use by the navigation system (Scheme 2).

[0057] (3) If fault detector 6 outputs a false fault signal and fault detector 6a outputs a true fault signal, the processing device 8 outputs a second difference signal for use by the navigation system (Scheme 1).

[0058] (4) If fault detectors 6 and 6a output the corresponding true fault signal, processing device 8 outputs a weighted sum of signals from the first and third accelerometers with previously estimated bias voltages, which are used by the navigation system. This fault state indicates that the high-noise, low-bias accelerometer 2 has failed. Therefore, fault detectors 6 and 6a are further configured to disable inputs from the high-noise, low-bias accelerometer 2 in schemes 1 and 2, and use scheme 3 as the output. In the absence of updates from the high-noise, low-bias accelerometer 2, schemes 1, 2, and 3 can be used with the previously estimated bias voltage.

[0059] Figure 4 This is a diagram illustrating the architecture of a navigation system 10 that enhances inertial navigation using gravity anomaly-based navigation according to an exemplary embodiment. The navigation system 10 is configured to determine the position of a platform 34 (e.g., an aircraft) in an Earth-based reference frame (such as the ECEF reference frame). All data processing steps can be performed by corresponding modules configured to provide the functions described below. Each module may include one or more processors programmed to execute instructions according to a corresponding software program.

[0060] The navigation system 10 includes a navigation computer system 7, an inertial measurement unit 11 (hereinafter referred to as "IMU 11"), and a gravimeter 21 (or other gravity anomaly detector), all mounted on the platform 34. The navigation computer system 7 communicates electronically with the IMU 11 via the IMU interface 12 and with the gravimeter 21 via the gravimeter interface 22. The navigation computer system 7 also communicates electronically with the motion control system on the platform 34 via the platform interface 33.

[0061] The navigation computer system 7 includes one or more devices selected from microprocessors, microcontrollers, digital signal processors, microcomputers, central processing units, application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), programmable logic devices, state machines, logic circuits, analog circuits, and digital circuits. When the computer device is under software control, it executes operable instructions stored in a non-transitory tangible computer-readable storage medium. According to one embodiment, the navigation computer system 7 includes software modules having code executable by a corresponding processor for performing the functions and processes described herein.

[0062] exist Figure 4 In the depicted embodiment, the navigation computer system 7 includes an IMU abstraction module 13 connected to an IMU interface 12 and an inertial navigation module 14 connected to the IMU abstraction module 13, thereby enabling inertial navigation. The navigation computer system 7 further includes: a platform guidance and control module 31 connected to receive output from the inertial navigation module 14; and a platform abstraction module 32 connected to receive output from the platform guidance and control module 31 and send input to the platform interface 33. The platform guidance and control module 31 is configured to use the inertial navigation output to control the platform 34 such that the platform 34 is approximately level (therefore, the second-order error assumption is true).

[0063] The navigation computer system 7 also includes a time-matching buffer 15 connected to the inertial navigation module 14, a gravimeter output predictor 16 connected to the time-matching buffer 15, and a gravity anomaly map / database 17 (hereinafter referred to as "gravity anomaly map / database 17"). The gravimeter output predictor 16 is a module configured to predict what the output of the gravimeter 21 should be using the time-matching state from the time-matching buffer 15 and gravity anomaly data from the gravity anomaly database 17, thus generating a "predicted sensor output".

[0064] The navigation computer system 7 further includes components that provide gravity anomaly-assisted navigation correction capabilities for correcting inertial navigation results. The gravity anomaly-assisted navigation correction subsystem includes a sensor data fusion module 23 connected to the IMU interface 12 and the gravimeter interface 22, a fused gravimeter abstraction module 24 connected to the sensor data fusion module 23, a residual and matrix calculation module 25 connected to the fused gravimeter abstraction module 24, a motion filter 26 connected to the residual and matrix calculation module 25, and an extended Kalman filter 27 connected to the motion filter 26.

[0065] IMU 11 includes a first plurality of sensors (e.g., three accelerometers with mutually orthogonal axes) for measuring acceleration and a second plurality of sensors (e.g., three gyroscopes with mutually orthogonal axes) for measuring the rotational rate of platform 34. IMU abstraction module 13 processes and encodes the signals output from the sensors of IMU 11 to form digital data representing measurements of the rotational rate (or Δ angle) and acceleration (or Δ velocity) of platform 34. This rotational and acceleration data is processed by inertial navigation module 14.

[0066] IMU interface 12 is a combination of electrical hardware and software interfaces that provide the navigation computing system 7 with IMU data output by IMU 11. IMU abstraction module 13 is software that converts IMU type / vendor-specified data into abstract, generic IMU data and interfaces. This module allows the inertial navigation module 14 to remain unchanged even when different types of IMUs are installed.

[0067] The inertial navigation module 14 is software that implements a typical "stripdown inertial navigation equation" (known in the art), which integrates gyroscope data (rotation rate or Δ angle) into the attitude and converts accelerometer data (source-specific or Δ rate) into a reference frame such as ECI, ECEF, or a local horizontal frame. The converted acceleration is then integrated into the rate and then into the position. In one proposed implementation, the inertial navigation module 14 includes executable code that integrates the rotation rate of platform 34 into the attitude of platform 34, taking into account the Earth's rotation. The platform acceleration is then projected into the ECEF reference frame using the platform attitude. Accordingly, the total platform acceleration (including gravity) due to the Earth's rotation can be calculated. The platform acceleration is then integrated into the rate of platform 34, and the platform rate is integrated to determine the position of platform 34 in the ECEF reference frame.

[0068] The time-matching buffer 15 stores the navigation state of a short historical period, including attitude, rate, position, and possibly other relevant variables. Once a measurement time stamp is assigned, the corresponding state estimate is calculated using interpolation based on the data available in the buffer. The time-matching buffer 15 includes executable code that uses time stamps to store the platform's position, rate, and attitude into a first-in-first-out buffer or a circular buffer. When image-based measurement time stamps are provided, the time-matching buffer 15 provides the time-matched position, rate, and attitude of the platform 34. The time-matched position, rate, and attitude of the platform 34 are then provided to the gravimeter output predictor 16.

[0069] The gravimeter output predictor 16 is a processor configured with software that processes platform position / attitude data received from the time-matching buffer 15 and retrieves gravity anomaly data from the gravity anomaly database 17. As previously mentioned, the gravimeter output predictor 16 generates a prediction of what the expected output of the gravimeter 21 will be. The gravity anomaly database 17 stores data representing specified forces as a function of position in the ECEF frame.

[0070] Gravity meter 21 measures a specified force along the vertical direction. Gravity meter 21 is positioned on platform 34 such that the z-axis of IMU 11 and gravity meter 21 is roughly aligned with the local vertical. Because any non-horizontal error of platform 34 only produces second-order effects on gravity meter measurements and accelerometer readings along the vertical direction, platform 34 does not require very high performance. Gravity meter interface 22 is an electrical hardware and software interface that allows gravity meter data to be provided to navigation computer system 7. Sensor data fusion module 23 performs local vertical accelerometer data fusion with the above-mentioned reference... Figures 1 to 3 The fusion of described gravimeter outputs. For example, sensor data fusion module 23 may include... Figure 1 The subtractors 3 and 5, filter 4, and fault detector 6 shown are... Figure 3 The subtractors 3, 3a, 5 and 5a, filters 4 and 4a, fault detectors 6 and 6a, and processing device 8 are shown.

[0071] exist Figure 4 In the depicted embodiment, the navigation computer system 7 includes a fusion gravimeter abstraction module 24 connected to the sensor data fusion module 23. The fusion gravimeter abstraction module 24 is software that converts gravimeter type / supplier-specified data into abstract, generic gravimeter data and interfaces. This module allows downstream processing blocks to remain unchanged even when using different gravimeters.

[0072] The navigation computer system 7 further includes a residual and matrix calculation module 25, communicatively coupled to receive the measured specified force (fusion scheme) generated by the sensor data fusion module 23 and the predicted sensor output generated by the gravimeter output predictor 16. The residual and matrix calculation module 25 compares the predicted specified force with the measured specified force to calculate the Kalman filter residual. The residual r represents the difference between the predicted specified force and the measured specified force. The residual and matrix calculation module 25 also calculates a sensitivity matrix (H-matrix) and a measurement noise covariance matrix (R-matrix). The H-matrix refers to the matrix that maps the state to the measurement in the linearized error equation, indicating how errors in the navigation scheme and the state in the Kalman filter 27 affect the predicted specified force. The R-matrix indicates the level and variables of uncertainty in the measurements acquired by the accelerometer.

[0073] The residual and matrix pass through motion filter 26 on their way to Kalman filter 27. Motion filter 26 is a low-pass filter that filters out the specified forces generated by the platform's motion. Typically, the filter's bandwidth is so low that only gravity remains. The primary force filtered out is acceleration generated by motion.

[0074] The navigation computer system 7 includes a Kalman filter 27 that receives motion-filtered data representing residuals and H and R matrices. The Kalman filter 27 is configured to generate position, rate, and attitude corrections based on the received residuals and matrices. Furthermore, the Kalman filter 27 estimates IMU corrections (such as bias, scaling factor, and misalignment correction) that are sent to and applied by the IMU abstraction module 13. The Kalman-filtered estimated position, rate, and attitude errors are sent to the inertial navigation module 14, and the inertial navigation module 14 applies these errors to generate a navigation scheme. In one proposed application, the navigation scheme is sent to a platform guidance and control module 31 (e.g., a flight controller or autopilot in the case of an aircraft), which is configured to control the movement of the platform 34 based at least in part on the received navigation scheme. The Kalman filter 27 maintains the covariance matrix of the state. The Kalman filter 27 is configured to calculate the gain using the H and R matrices, generate corrections using the gain residuals, and calculate the covariance matrix after an update using the gain matrix, the R matrix, and the current covariance matrix. The Kalman filter 27 also outputs data to the fused gravimeter abstraction module 24 to correct for gravimeter errors as needed.

[0075] In summary, the gravity anomaly-assisted navigation system includes a gravimeter 21 that generates time series of specified force measurements at different locations in the ECEF frame. Simultaneously with the gravity anomaly-assisted navigation correction capability operation, the IMU 11 generates a time series of inertial navigation information (hereinafter referred to as "IMU data"). An inertial navigation algorithm (executed in the inertial navigation module 14) is configured to analyze the IMU data to generate a time series of an estimated inertial navigation scheme representing the platform's changed position. This navigation scheme is calculated at least in part based on corrections derived by comparing the position determined by the gravity anomaly-based navigation system with the position determined by the inertial navigation system. The corresponding data from the inertial navigation system and the gravity anomaly-based navigation system are fused using a Kalman filter 27. The resulting navigation scheme incorporates corrections.

[0076] A particular system, apparatus, application, or process has been described herein as comprising multiple modules. A module can be an obviously functional unit implemented in software, hardware, or a combination thereof, except for those modules preferably implemented as hardware or firmware to enable the streaming computing disclosed herein. When the functionality of a module is performed by software in any component, the module can include a non-transitory tangible computer-readable storage medium.

[0077] Although a sensor data fusion system for measuring a specified force applied to a mobile platform by means of navigation measurements has been described with reference to various embodiments, those skilled in the art will understand that various changes can be made and equivalents can be substituted for its elements without departing from the teachings herein. Furthermore, various modifications can be made to adapt the concepts and simplifications of the practices disclosed herein to specific situations. Accordingly, it is intended that the subject matter covered by the claims is not limited to the disclosed embodiments.

[0078] The embodiments disclosed above use one or more processing or computing devices. Such devices typically include processors or controllers, such as general-purpose central processing units, microcontrollers, reduced instruction set computer processors, application-specific integrated circuits (ASICs), programmable logic circuits, field-programmable gate arrays (FPGAs), digital signal processors (DSPs), and / or any other data processing circuitry capable of performing the functions described herein. The methods described herein can be encoded into executable instructions encapsulated in a non-transitory tangible computer-readable storage medium, including but not limited to storage devices and / or memory devices. When executed by a processing device, the instructions cause the processing device to perform at least a portion of the methods described herein.

[0079] The method claims set forth below should not be construed as requiring the steps set forth herein to be performed in alphabetical order (any alphabetical order in the claims is for reference only to the steps set forth previously) or in the order in which they are set forth, unless the language of the claims or the supporting description specifies a condition indicating the specific order in which some or all of these steps are performed. Nor should the method claims be construed as excluding any step from being performed in parallel or alternately, unless the language of the claims or the supporting description specifies a condition excluding such an interpretation.

[0080] This disclosure includes the subject matter described in the following items:

[0081] Item 1. A sensor data fusion system (20), comprising:

[0082] The first accelerometer (1) has low noise and high bias voltage;

[0083] The second accelerometer (2) has high noise and low bias voltage;

[0084] The first subtractor (3) is connected to receive a signal from the first accelerometer and a signal from the second accelerometer and is configured to output a first difference signal representing the difference between the signal from the first accelerometer and the signal from the second accelerometer;

[0085] A first filter (4) is connected to receive a first difference signal from a first subtractor and is configured to output a signal representing the estimated relative bias voltage of the first and second accelerometers; and

[0086] The second subtractor (5) is connected to receive a signal from the first accelerometer and a signal from the first filter and is configured to output a second difference signal representing the difference between the signal from the first accelerometer and the signal from the first filter.

[0087] Item 2. The sensor data fusion system according to Item 1, wherein the second accelerometer is a gravimeter.

[0088] Item 3. The sensor data fusion system according to Item 1 or Item 2, wherein the second difference signal has low noise and low bias voltage.

[0089] Item 4. A sensor data fusion system according to any one of the preceding items, wherein the filter is configured to filter out most of the noise while allowing a signal representing an estimated relative bias to pass through.

[0090] Item 5. The sensor data fusion system according to Item 4, wherein the filter is a low-pass filter.

[0091] Item 6. The sensor data fusion system according to Item 4, wherein the filter is a Kalman filter.

[0092] Item 7. The sensor data fusion system according to any one of the preceding items further includes: a fault detector (6) connected to receive a first difference signal, and configured to output a fault true signal if the first difference signal persists beyond a specified difference threshold.

[0093] Item 8. The sensor data fusion system according to any one of the preceding items further includes:

[0094] The third accelerometer (a) features low noise and high bias.

[0095] The third subtractor (3a) is connected to receive signals from the second accelerometer and signals from the third accelerometer and is configured to output a third difference signal representing the difference between the signals from the second accelerometer and the signals from the third accelerometer;

[0096] The second filter (4a) is connected to receive the third difference signal output by the third subtractor and is configured to output a signal representing the estimated relative bias voltage of the second and third accelerometers; and

[0097] The fourth subtractor (5a) is connected to receive a signal from the third accelerometer and a signal from the second filter and is configured to output a fourth difference signal representing the difference between the signal from the third accelerometer and the signal from the second filter.

[0098] Item 9. The sensor data fusion system according to Item 8, wherein the second accelerometer is a gravimeter.

[0099] Item 10. The sensor data fusion system according to Item 8 further includes:

[0100] The first fault detector (6) is connected to receive the first difference signal and is configured to output a true fault signal if the first difference signal persists beyond a specified difference threshold, and otherwise output a false fault signal.

[0101] A second fault detector (6a) is connected to receive a third difference signal and is configured to output a true fault signal if the third difference signal persistently exceeds a specified difference threshold, and otherwise output a false fault signal; and

[0102] Processing device (8) is connected to a first fault detector, a second fault detector, a third subtractor, and a fourth subtractor, wherein the processing device is configured to:

[0103] If the first fault detector outputs a false fault signal while the second fault detector outputs a true fault signal, then the first difference signal is output; and

[0104] If the first fault detector outputs a true fault signal while the second fault detector outputs a false fault signal, then the second difference signal is output.

[0105] Item 11. The sensor data fusion system according to Item 10, wherein if the first fault detector and the second fault detector output corresponding false fault signals, the processing device is further configured to output a weighted sum of the first difference signal and the second difference signal.

[0106] Item 12. The sensor data fusion system according to Item 10, wherein if the first fault detector and the second fault detector output corresponding true fault signals, the processing device is further configured to output a weighted sum of the signals output by the first accelerometer and the third accelerometer having previously estimated bias voltages.

[0107] Item 13. A method for fusing sensor data, comprising:

[0108] A first measurement signal is output using a first accelerometer (1) with low noise and high bias, the first measurement signal representing a first measurement value of a specified force in the direction perpendicular to the reference frame of the platform;

[0109] A second measurement signal is output using a second accelerometer (2) with high noise and low bias, the second measurement signal representing a second measurement value of a specified force in the direction perpendicular to the reference frame of the platform;

[0110] Generate a first difference signal representing the difference between the first measurement signal and the second measurement signal;

[0111] The first difference signal is filtered to generate a first bias estimation signal representing the estimated relative bias of the first and second accelerometers; and

[0112] A second difference signal is generated, representing the difference between the first measurement signal and the first bias estimation signal.

[0113] Item 14. The method according to Item 13 further includes: using the first difference signal during navigation if the first difference signal does not persistently exceed a specified difference threshold.

[0114] Item 15. The method according to Item 13 or Item 14 further includes: generating a fault true signal if the first difference signal persistently exceeds a specified difference threshold.

[0115] Item 16. The method according to any one of items 13 to 15, further comprising:

[0116] A third measurement signal is output using a third accelerometer (1a) with low noise and high bias, the third measurement signal representing a third measurement value of a specified force in the direction perpendicular to the reference frame of the platform;

[0117] Generate a third difference signal representing the difference between the second and third measurement signals;

[0118] The third difference signal is filtered to generate a second bias estimation signal representing the estimated relative bias of the second and third accelerometers; and

[0119] A fourth difference signal is generated, representing the difference between the third measurement signal and the second bias estimation signal.

[0120] Item 17. The method according to Item 16 further includes:

[0121] If the first difference signal persists beyond the specified difference threshold, a true fault signal is generated; otherwise, a false fault signal is generated.

[0122] If the third difference signal persists beyond the specified difference threshold, a true fault signal is generated; otherwise, a false fault signal is generated.

[0123] If one of the difference signals between the first difference signal and the second difference signal does not persistently exceed the specified difference threshold, and the other difference signal between the first difference signal and the second difference signal persistently exceeds the specified difference threshold, then one of the difference signals between the first difference signal and the second difference signal is used during navigation.

[0124] Item 18. A system (10) for a navigation platform, comprising:

[0125] An inertial navigation system (11) is configured to generate a navigation scheme and includes an accelerometer (1) with low noise and high bias.

[0126] Gravity gauge (2) with high noise and low bias voltage;

[0127] The guidance and control system (31) is communicatively coupled to the inertial navigation system and configured to control the platform according to the navigation scheme;

[0128] The sensor data fusion module (23) is configured to generate a bias estimation signal representing a relative bias estimation from the outputs of the accelerometer and gravimeter, and then generate a signal representing a corrected output of the accelerometer derived by subtracting the bias estimation signal from the output of the accelerometer.

[0129] A time-matching buffer (15) is communicatively coupled to the inertial navigation system and configured to store data representing the platform’s position, velocity, and attitude using time stamps;

[0130] A gravity anomaly database (17) is stored in a non-transitory tangible computer-readable storage medium;

[0131] Gravity meter output predictor (16) is communicatively coupled to receive time-stamped position, velocity and attitude of the platform from a time-matching buffer and to retrieve gravity anomaly data from a gravity anomaly database and is configured to output a prediction signal representing a prediction of the gravimeter output.

[0132] The residual and matrix calculation module (25) is communicatively coupled to receive the prediction signal and the accelerometer's correction output and is configured to generate residuals, an H matrix, and an R matrix based on the difference between the prediction signal and the accelerometer's correction output; and

[0133] The Kalman filter (27) is configured to generate a position correction based on the difference between the predicted signal and the accelerometer's correction output and then send the position correction to the inertial navigation system.

[0134] Item 19. The system according to Item 18, wherein the sensor data fusion module includes:

[0135] The first subtractor (3) is connected to receive a signal from the accelerometer and a signal from the gravimeter and is configured to output a first difference signal representing the difference between the signals from the accelerometer and the gravimeter;

[0136] Filter (4), connected to receive the first difference signal output by the first subtractor and configured to output a bias estimation signal; and

[0137] The second subtractor (5) is connected to receive the signal from the accelerometer and the bias estimation signal and is configured to output a second difference signal representing the difference between the signal from the accelerometer and the bias estimation signal.

[0138] Item 20. The system according to Item 19, wherein the sensor data fusion module further includes: a fault detector (6) connected to receive a first difference signal, and the fault detector is configured to output a fault true signal if the first difference signal persists beyond a specified difference threshold.

[0139] Item 21. A method for a navigation platform, comprising:

[0140] The first accelerometer (1) outputs a first measurement signal, which represents a first measurement value of a specified force in the direction perpendicular to the reference frame of the platform.

[0141] An inertial navigation scheme is generated using the first measurement signal;

[0142] The second accelerometer (2) outputs a second measurement signal, which represents a second measurement value of a specified force in the direction perpendicular to the reference frame of the platform;

[0143] A signal representing the specified force being measured is generated based on the difference between the first measurement signal and the second measurement signal;

[0144] Generate a signal representing the predicted specified force using a historical navigation state and a gravity anomaly database;

[0145] Inertial navigation scheme correction is generated based on the difference between the predicted specified force and the measured specified force;

[0146] A corrected inertial navigation scheme is generated by applying inertial navigation scheme correction to the existing inertial navigation scheme; and

[0147] The control platform is based on the corrected inertial navigation scheme.

[0148] Item 22. The method according to Item 21, wherein the first accelerometer has low noise and high bias and the second accelerometer has high noise and low bias.

[0149] Item 23. The method according to Item 22, wherein generating a signal representing the specified force measured based on the difference between the first measurement signal and the second measurement signal comprises:

[0150] The difference signal is filtered to produce a bias estimate signal representing the estimated relative bias of the first and second accelerometers; and

[0151] Generate a difference signal representing the difference between the first measurement signal and the bias estimation signal.

Claims

1. A sensor data fusion system, comprising: The first accelerometer features low noise and high bias. The second accelerometer features high noise and low bias. A first subtractor is connected to receive signals from the first accelerometer and signals from the second accelerometer and is configured to output a first difference signal representing the difference between the signals from the first accelerometer and the signals from the second accelerometer; A first filter is connected to receive the first difference signal from the first subtractor and is configured to output a signal representing the estimated relative bias of the first accelerometer and the second accelerometer; as well as The second subtractor, connected to receive a signal from the first accelerometer and a signal from the first filter, is configured to output a second difference signal representing the difference between the signal from the first accelerometer and the signal from the first filter. The second difference signal has low noise and low bias voltage.

2. The sensor data fusion system according to claim 1, wherein, The second accelerometer is a gravimeter.

3. The sensor data fusion system according to claim 1 or 2, wherein, The first filter is configured to filter out most of the noise while allowing the signal representing the estimated relative bias to pass through.

4. The sensor data fusion system according to claim 3, wherein, The first filter is a low-pass filter.

5. The sensor data fusion system according to claim 3, wherein, The first filter is a Kalman filter.

6. The sensor data fusion system according to claim 1 or 2, further comprising: A fault detector is connected to receive the first difference signal, and if the first difference signal persists beyond a specified difference threshold, the fault detector is configured to output a true fault signal.

7. The sensor data fusion system according to claim 1, further comprising: The third accelerometer features low noise and high bias. A third subtractor is connected to receive signals from the second accelerometer and signals from the third accelerometer and is configured to output a third difference signal representing the difference between the signals from the second accelerometer and the signals from the third accelerometer; The second filter is connected to receive the third difference signal output by the third subtractor and is configured to output a signal representing the estimated relative bias of the second accelerometer and the third accelerometer. as well as A fourth subtractor is connected to receive a signal from the third accelerometer and a signal from the second filter and is configured to output a fourth difference signal representing the difference between the signal from the third accelerometer and the signal from the second filter.

8. The sensor data fusion system according to claim 7, wherein, The second accelerometer is a gravimeter.

9. The sensor data fusion system according to claim 7, further comprising: A first fault detector is connected to receive the first difference signal and is configured to output a true fault signal if the first difference signal persists beyond a specified difference threshold, and otherwise output a false fault signal. A second fault detector is connected to receive the third difference signal and is configured to output a true fault signal if the third difference signal persists beyond the specified difference threshold, and otherwise output a false fault signal. as well as A processing device is connected to the first fault detector, the second fault detector, the third subtractor, and the fourth subtractor, wherein the processing device is configured to: If the first fault detector outputs a false fault signal and the second fault detector outputs a true fault signal, then the first difference signal is output. and If the first fault detector outputs a true fault signal while the second fault detector outputs a false fault signal, then the second difference signal is output.

10. The sensor data fusion system according to claim 9, wherein, If the first fault detector and the second fault detector output corresponding false fault signals, the processing device is further configured to output a weighted sum of the first difference signal and the second difference signal.

11. The sensor data fusion system according to claim 9, wherein, If the first fault detector and the second fault detector output corresponding true fault signals, the processing device is further configured to output a weighted sum of the signals output by the first accelerometer and the third accelerometer with previously estimated bias voltages.

12. A method for fusing sensor data, comprising: A first measurement signal is output using a first accelerometer with low noise and high bias, the first measurement signal representing a first measurement value of a specified force in the direction perpendicular to the reference frame of the platform; A second measurement signal is output using a second accelerometer with high noise and low bias, the second measurement signal representing a second measured value of a specified force in the vertical direction of the reference frame of the platform; Generate a first difference signal representing the difference between the first measurement signal and the second measurement signal; The first difference signal is filtered to generate a first bias estimation signal representing the estimated relative bias of the first accelerometer and the second accelerometer; and Generate a second difference signal representing the difference between the first measured signal and the first bias estimation signal. The second difference signal has low noise and low bias voltage.

13. The method of claim 12, further comprising: If the first difference signal does not exceed the specified difference threshold for an extended period, the first difference signal is used during navigation.

14. The method according to claim 12 or 13, further comprising: If the first difference signal persists beyond the specified difference threshold, a true fault signal is generated.

15. The method according to claim 12 or 13, further comprising: A third measurement signal is output using a third accelerometer with low noise and high bias, the third measurement signal representing a third measurement value of a specified force in the vertical direction of the reference frame of the platform; Generate a third difference signal representing the difference between the second measurement signal and the third measurement signal; The third difference signal is filtered to generate a second bias estimation signal representing the estimated relative bias of the second accelerometer and the third accelerometer; and A fourth difference signal is generated, representing the difference between the third measurement signal and the second bias estimation signal.

16. The method of claim 15, further comprising: If the first difference signal persists beyond the specified difference threshold, a true fault signal is generated; otherwise, a false fault signal is generated. If the third difference signal persists beyond the specified difference threshold, a true fault signal is generated; otherwise, a false fault signal is generated. and If one of the first difference signal and the second difference signal does not persistently exceed the specified difference threshold, and the other of the first difference signal and the second difference signal persistently exceeds the specified difference threshold, then the first difference signal and the second difference signal are used during navigation.

17. A system for a navigation platform, comprising: An inertial navigation system is configured to generate navigation schemes and includes an accelerometer with low noise and high bias. Gravimeter with high noise and low bias voltage; A guidance and control system is communicatively coupled to the inertial navigation system and configured to control the platform according to the navigation scheme; The sensor data fusion module is configured to generate a bias estimation signal representing a relative bias estimation from the outputs of the accelerometer and the gravimeter, and then generate a signal representing a corrected output of the accelerometer derived by subtracting the bias estimation signal from the output of the accelerometer. A time-matching buffer is communicatively coupled to the inertial navigation system and configured to store data representing the platform's position, velocity, and attitude using time stamps; The gravity anomaly database is stored in a non-transitory, tangible, computer-readable storage medium; A gravimeter output predictor is communicatively coupled to receive time-stamped position, velocity, and attitude of the platform from the time-matching buffer and to retrieve gravity anomaly data from the gravity anomaly database and is configured to output a prediction signal representing the prediction of the gravimeter's output. A residual and matrix calculation module, communicatively coupled to receive the prediction signal and the correction output of the accelerometer, and configured to generate a residual, an H matrix, and an R matrix based on the difference between the prediction signal and the correction output of the accelerometer; and A Kalman filter is configured to generate a position correction based on the difference between the predicted signal and the correction output of the accelerometer, and then send the position correction to the inertial navigation system.

18. The system according to claim 17, wherein, The sensor data fusion module includes: A first subtractor is connected to receive signals from the accelerometer and signals from the gravimeter and is configured to output a first difference signal representing the difference between the signals from the accelerometer and the gravimeter; A filter, connected to receive the first difference signal output from the first subtractor and configured to output the bias estimation signal; and The second subtractor is connected to receive the signal from the accelerometer and the bias estimation signal and is configured to output a second difference signal representing the difference between the signal from the accelerometer and the bias estimation signal.

19. The system according to claim 18, wherein, The sensor data fusion module further includes: a fault detector, connected to receive the first difference signal, and configured to output a true fault signal if the first difference signal persistently exceeds a specified difference threshold.

20. A method for a navigation platform, comprising: The first accelerometer outputs a first measurement signal, which represents a first measured value of a specified force in the direction perpendicular to the reference frame of the platform. The first measurement signal is used to generate an inertial navigation scheme; A second measurement signal is output using a second accelerometer, the second measurement signal representing a second measured value of a specified force in the vertical direction of the reference frame of the platform; A signal representing the specified force being measured is generated based on the difference between the first measurement signal and the second measurement signal; Generate a signal representing the predicted specified force using a historical navigation state and a gravity anomaly database; Inertial navigation scheme correction is generated based on the difference between the predicted specified force and the measured specified force; A corrected inertial navigation scheme is generated by applying inertial navigation scheme correction to the existing inertial navigation scheme; and The platform is controlled according to the corrected inertial navigation scheme.

21. The method according to claim 20, wherein, The first accelerometer has low noise and high bias, and the second accelerometer has high noise and low bias.

22. The method according to claim 21, wherein, Generating a signal representing the measured specified force based on the difference between the first measurement signal and the second measurement signal includes: The difference signal is filtered to generate a bias estimation signal representing the estimated relative bias of the first accelerometer and the second accelerometer; and Generate a difference signal representing the difference between the first measurement signal and the bias estimation signal.

Images