System and method steps for correcting gyro drift of a motor vehicle

Abstract

A system of a motor vehicle includes a radar unit to generate a radar signal associated with a target located about the vehicle. The system also includes a tracker to generate a tracking signal associated with a radar heading and a Doppler effect based on the radar signal. The system also includes a wheel speed sensor to generate a wheel speed signal associated with a vehicle speed. The system also includes a gyroscope to generate a gyroscope signal associated with a measured yaw rate. The system also includes a computer having a processor and a computer readable medium. The processor is programmed to determine a gyroscope drift and a corrected yaw rate while the vehicle is in motion, wherein the gyroscope drift and the corrected yaw rate are based on at least the radar heading, the Doppler effect, and the vehicle speed and the measured yaw rate.

Description

Technical Field

[0001] This disclosure relates to motor vehicles having a system with a gyroscope, and more specifically, to a system and method steps for determining yaw rates for gyro drift and correction based at least on radar signals, wheel speed sensor signals, temperature signals, and gyroscope signals. Background Technology

[0002] Automotive systems equipped with inertial sensors such as gyroscopes are prone to both deterministic and random errors. Deterministic errors, such as bias error and scaling factor error, can be addressed through laboratory calibration. However, random errors, such as bias drift error, can only be modeled mathematically. Random errors are a major source of measurement error because uncompensated gyro bias error (expressed in degrees / hour or radians / second) produces angular errors proportional to the passage of time t. Therefore, the yaw axis is most sensitive to gyro drift, which reduces the accuracy of systems using gyroscopes. In particular, gyro drift limits yaw accuracy over increasingly longer time periods, and radar alignment may change over time, which in turn further reduces the accuracy of the relevant system.

[0003] Known systems that include gyroscopes may also include radar units for recalibrating the gyroscopes. However, these systems are used in batch mode architectures that require the vehicle to stop during recalibration. Furthermore, the accuracy of these systems may be so inaccurate that the system does not recalibrate the gyroscopes based on any data other than radar data.

[0004] An example of such systems could include an advanced driver assistance system (“ADAS”) with a gyroscope for determining attitude and bearing. ADAS could also include a global positioning system module (“GPS module”) for accurately determining the vehicle’s position and heading to detect gyroscope drift and recalibrate the gyroscope. However, the ability of the GPS module to perform these functions can be challenged when the vehicle is in tunnels, deep in urban areas, or other areas with weak or absent GPS signals.

[0005] Therefore, while the existing system has achieved its intended purpose, a new and improved system and methodology are needed to address these issues. Summary of the Invention

[0006] According to several aspects of this disclosure, a system for a motor vehicle includes a radar unit for generating radar signals associated with detected targets located around the motor vehicle. The system also includes a tracker coupled to the radar unit. The tracker generates a tracking signal based on the radar signals, associated with the current radar heading and current Doppler effect of a stationary target. The system also includes a wheel speed sensor for generating a wheel speed signal associated with the current speed of the motor vehicle. The system also includes a temperature sensor for generating a temperature signal associated with ambient temperature. The system also includes a gyroscope for generating a gyroscopic signal associated with a measured yaw rate at the current time step. The system also includes a computer having one or more processors coupled to the tracker, wheel speed sensor, temperature sensor, and gyroscope. The computer also includes a non-transitory computer-readable storage medium (“CRM”) storing instructions that cause the processor to be programmed to determine gyroscopic drift and corrected yaw rates during motor vehicle motion, based at least on the current radar heading, current Doppler effect, current speed of the motor vehicle, and measured yaw rate at the current time step.

[0007] In one aspect, the processor is also programmed to receive, at the current time step, a tracking signal from the tracker, a wheel speed signal from the wheel speed sensor, a temperature signal from the temperature sensor, and a gyroscope signal from the gyroscope. The processor is also programmed to use a Kalman filter to determine an innovation vector based on the differences between the tracking signal, wheel speed signal, temperature signal, gyroscope signal, and multiple associated predicted observations. The processor is further programmed to generate an adapted measurement covariance matrix based on the innovation vector and, more specifically, on the measurement covariance matrix to enhance robustness. The processor is also programmed to update the Kalman filter based on the tracking signal, wheel speed signal, temperature signal, and gyroscope signal. The processor is further programmed to predict at least one of the following based on the updated Kalman filter: predicted radar heading, predicted Doppler effect, estimated gyro scale factor from the gyroscope, wheel speed sensor scale factor from the wheel speed sensor, and gyroscope drift from the gyroscope.

[0008] On another front, the processor determines the whitening information based on the change in the maximum correlation entropy of the Kalman filter through the following calculation:

[0009]

[0010] Where e represents the current information, which is related to the error between the actual measurement vector z and the predicted measurement vector (i.e., e = z - Hx, where x is the predicted state vector and H is the linearized measurement covariance matrix). Furthermore, R... -1 / 2 This represents the inverse matrix of the Choleski decomposition of the measurement covariance matrix.

[0011] In another aspect, the processor generates an adapted measurement covariance matrix according to the following formula:

[0012]

[0013]

[0014]

[0015] in The kernel function is represented by β; the positive tuning parameter is represented by β. Representing vectors In the matrix, element i; C represents the adaptive measurement covariance matrix; Let represent the updated measurement covariance matrix; and T denote the matrix transpose operator. After running the Kalman filter update step, the process can be iterated using the latest state estimate instead of the predicted state to determine the innovation.

[0016] In another aspect, the processor updates the Kalman filter using an implicit nonlinear observation model to estimate at least one gyroscope scaling factor and gyroscope drift according to the following equation:

[0017]

[0018] in The yaw rate, representing the corrected yaw rate of a motor vehicle; The X-coordinate of the radar unit on the vehicle; β0 represents the Y-coordinate of the radar unit on the vehicle; β0 represents the nominal installation azimuth angle of the radar unit; β represents the calibration error of the radar unit. Indicates the radar heading associated with the tracking signal; Indicates the lateral speed of a motor vehicle; Indicates the longitudinal speed of a motor vehicle; This indicates the estimated longitudinal speed of a motor vehicle; α represents the Doppler velocity associated with the tracking signal. gyro α represents the gyroscope scaling factor. wss Indicates the wheel speed sensor scaling factor; γ gyro This represents gyroscope drift; n represents iid noise; and This represents the measured yaw rate associated with the gyro signal generated by the gyroscope.

[0019] In another aspect, the processor, based on an updated Kalman filter, uses a process model to predict at least one of the estimated gyroscope scaling factor, wheel speed sensor scaling factor, and gyroscope drift according to the following equation:

[0020]

[0021] Among them G u This represents the predicted value of the gyroscope drift. The magnitude of the previous gyroscope zero bias error estimate is T. w The vector, t represents the current time step; and t+1 represents the next time step.

[0022] In another aspect, the processor further utilizes a newer Kalman filter to predict at least one of the estimated gyroscope scaling factor, wheel speed sensor scaling factor, and gyroscope drift based on process noise according to the following formula:

[0023]

[0024] On another front, the processor also predicts gyroscope drift based on the following formula:

[0025]

[0026]

[0027]

[0028] The gyroscope drift is modeled as a zero-mean Gaussian process with a known covariance kernel k(·,·).

[0029] On another front, the processor predicts the gyroscope's gyroscope drift based on the following formula:

[0030]

[0031]

[0032]

[0033] in The independent predictor of the gyroscope's gyroscope drift; t represents the current time step; Indicates longitudinal velocity; Indicates yaw rate; The predictor variable represents the gyroscope drift of the gyroscope at the previous time step; and The predictor variable represents the gyroscope drift of the gyroscope at the current time step.

[0034] On another front, the processor also predicts the gyroscope's gyroscope drift based on the spectral mixture kernel:

[0035]

[0036] Where C = C(ψ), is a spherically parameterized correlation matrix that correlates the predictors at previous and current time steps with... Model the correlation between them; and learn multiple parameters from the data. ψ. For computational reasons, it can be assumed that the prediction vector lies on a regular grid, thus allowing the use of the Kronecker structure.

[0037] According to several aspects of this disclosure, a computer for a motor vehicle system is provided. The system includes a radar unit for generating radar signals associated with detected targets located around the motor vehicle. The system also includes a tracker coupled to the radar unit. The tracker generates a tracking signal associated with the current radar heading and current Doppler effect of a stationary target based on the radar signals. The system also includes a wheel speed sensor for generating a wheel speed signal associated with the current speed of the motor vehicle. The system also includes a temperature sensor for generating a temperature signal associated with ambient temperature. The system also includes a gyroscope for generating a gyroscopic signal associated with a measured yaw rate at the current time step. The computer includes one or more processors coupled to the tracker, wheel speed sensor, and gyroscope. The computer also includes a non-transitory computer-readable storage medium (“CRM”) containing instructions. The processors are also programmed to determine gyroscopic drift and corrected yaw rates as the motor vehicle moves, wherein the gyroscopic drift and corrected yaw rates are based at least on the current radar heading, current Doppler effect, current speed of the motor vehicle, and measured yaw rates at the current time step.

[0038] In one aspect, the processor is also programmed to receive, at the current time step, a tracking signal from the tracker, a wheel speed signal from the wheel speed sensor, a temperature signal from the temperature sensor, and a gyroscope signal from the gyroscope. The processor is also programmed to determine an innovation vector based on the difference between the tracking signal, wheel speed signal, temperature signal, gyroscope signal, and multiple relevant predicted observations using a Kalman filter. The processor is further programmed to generate an adapted measurement covariance matrix based on the innovation vector and, more specifically, on the measurement covariance matrix to enhance robustness. The processor is also programmed to update the Kalman filter based on the tracking signal, wheel speed signal, temperature signal, and gyroscope signal. The processor is further programmed to predict at least one of the following based on the updated Kalman filter: predicted radar heading, predicted Doppler effect, estimated gyroscope scaling factor from the gyroscope, wheel speed sensor scaling factor from the wheel speed sensor, and gyroscope drift from the gyroscope.

[0039] On another front, the processor determines the whitened innovation based on the maximum correlation entropy change of the Kalman filter using the following formula.

[0040]

[0041] Where e represents the current information, which is related to the error between the actual measurement vector z and the predicted measurement vector (i.e., e = z - Hx, where x is the predicted state vector and H is the linearized measurement covariance matrix). Furthermore, R... -1 / 2 This represents the inverse of the Cholesky decomposition of the measurement covariance matrix.

[0042] In another aspect, the processor generates an adapted measurement covariance matrix according to the following formula:

[0043]

[0044]

[0045]

[0046] in The kernel function is represented by β; the positive tuning parameter is represented by β. Representing vectors In the matrix, element i; C represents the adaptive measurement covariance matrix; Let represent the updated measurement covariance matrix; T denotes the matrix transpose operator. After running the Kalman filter update step, the process can be iterated using the latest state estimate instead of the predicted state to determine the innovation.

[0047] In another aspect, the processor updates the Kalman filter using an implicit nonlinear observation model to estimate at least one gyroscope scaling factor and gyroscope drift according to the following equation:

[0048]

[0049] in The yaw rate, representing the corrected yaw rate of a motor vehicle; The X-coordinate of the radar unit on the vehicle; β0 represents the Y-coordinate of the radar unit on the vehicle; β0 represents the nominal installation azimuth angle of the radar unit; β represents the calibration error of the radar unit. Indicates the radar heading associated with the tracking signal; Indicates the lateral speed of a motor vehicle; Indicates the longitudinal speed of a motor vehicle; This indicates the estimated longitudinal speed of a motor vehicle; α represents the Doppler velocity associated with the tracking signal. gyro α represents the gyroscope scaling factor. wss Indicates the wheel speed sensor scaling factor; γ gyro This represents gyroscope drift; n represents iid noise; and This represents the measured yaw rate associated with the gyro signal generated by the gyroscope.

[0050] In another aspect, the processor further utilizes a process model based on an updated Kalman filter to predict at least one of the estimated gyroscope scaling factor, wheel speed sensor scaling factor, and gyroscope drift according to the following equation:

[0051]

[0052] in This represents the predicted value of the gyroscope drift. The magnitude of the previous gyroscope zero bias error estimate is T. w The vector, t represents the current time step; and t+1 represents the next time step.

[0053] In another aspect, the processor further utilizes a newer Kalman filter to predict at least one of the estimated gyroscope scaling factor, wheel speed sensor scaling factor, and gyroscope drift based on process noise according to the following formula:

[0054]

[0055] According to several aspects of this disclosure, a method for operating a motor vehicle system is provided. The system includes a radar unit, a tracker coupled to the radar unit, a wheel speed sensor, a temperature sensor, and a gyroscope. The system also includes a computer coupled to the tracker, wheel speed sensor, temperature sensor, and gyroscope. The computer includes one or more processors and a non-transitory computer-readable storage medium including instructions. The method steps include using the radar unit to generate a radar signal associated with a detected target located around the motor vehicle. The method steps also include using the tracker to generate a tracking signal associated with the current radar heading and current Doppler effect of a stationary target based on the radar signal. The method steps also include using the wheel speed sensor to generate a wheel speed signal associated with the current speed of the motor vehicle. The method steps also include using the temperature sensor to generate a temperature signal associated with the ambient temperature. The method steps also include using the gyroscope to generate a gyroscope signal associated with a measured yaw rate at the current time step. The method also includes using a processor to determine the gyro drift and corrected yaw rate of the vehicle in motion, wherein the gyro drift and corrected yaw rate are based at least on the current radar heading, the current Doppler effect, the current speed of the vehicle, and the yaw rate measured at the current time step.

[0056] In one aspect, the method steps further include utilizing a processor to receive a tracking signal from a tracker, a wheel speed signal from a wheel speed sensor, a temperature signal from a temperature sensor, and a gyroscope signal from a gyroscope at the current time step. The method steps further include utilizing a Kalman filter to determine innovation based on the difference between the tracking signal, wheel speed signal, temperature signal, gyroscope signal, and multiple associated predicted observations. The method steps further include using a processor to generate an adapted measurement covariance matrix based on the innovation vector and further based on the measurement covariance matrix to enhance robustness. The method steps further include using a processor to update the Kalman filter based on the tracking signal, wheel speed signal, temperature signal, and gyroscope signal. The method steps further include using a processor to predict at least one of the following items based on the updated Kalman filter: predicted radar heading, predicted Doppler effect, estimated gyroscope scaling factor of the gyroscope, wheel speed sensor scaling factor of the wheel speed sensor, and gyroscope drift of the gyroscope.

[0057] In another aspect, the method steps also include the processor determining the gyroscope drift according to the following formula:

[0058]

[0059]

[0060]

[0061]

[0062]

[0063]

[0064] Where e represents the current information, which is related to the error between the actual measurement vector z and the predicted measurement vector (i.e., e = z - Hx, where x is the predicted state vector and H is the linearized measurement covariance matrix); R -1 / 2 denoted as the inverse of the Chollisky decomposition of the measurement covariance matrix. e denotes the previous information determined at the previous time step; The kernel function is represented by β; the positive tuning parameter is represented by β. Representing vectors In the matrix, element i; C represents the adaptive measurement covariance matrix; The updated measurement covariance matrix is ​​represented by T; T represents the matrix transpose operator. The yaw rate, representing the corrected yaw rate of a motor vehicle; The X-coordinate of the radar unit on the vehicle; β0 represents the Y-coordinate of the radar unit on the vehicle; β0 represents the nominal installation azimuth angle of the radar unit; β represents the calibration error of the radar unit. Indicates the radar heading associated with the tracking signal; Indicates the lateral speed of a motor vehicle; Indicates the longitudinal speed of a motor vehicle; This indicates the estimated longitudinal speed of a motor vehicle; α represents the Doppler velocity associated with the tracking signal. gyro α represents the gyroscope scaling factor. wss Indicates the wheel speed sensor scaling factor; γ gyro This indicates gyroscope drift; n represents iid noise; This represents the measured yaw rate associated with the gyro signal generated by the gyroscope. The value represents the predicted value of the gyroscope drift; t represents the current time step; and t+1 represents the next time step.

[0065] Further applicability will become apparent from the description provided herein. It should be understood that the specification and specific examples are for illustrative purposes only and are not intended to limit the scope of this disclosure. Attached Figure Description

[0066] The accompanying drawings described herein are for illustrative purposes only and are not intended to limit the scope of this disclosure in any way.

[0067] Figure 1This is a schematic diagram of an example of a motor vehicle with a system including a radar unit, a tracker, wheel speed sensors, a temperature sensor, a gyroscope, and a computer for recalibrating the gyroscope.

[0068] Figure 2 It is used for operation Figure 1 A flowchart illustrating an example of the system's method steps. Detailed Implementation

[0069] The following description is exemplary in nature and is not intended to limit this disclosure, application, or use.

[0070] This disclosure describes an example of a motor vehicle having a system including a gyroscope, a radar unit, a tracker, a temperature sensor, and a computer for determining gyro drift and correcting yaw rates. More specifically, in a non-limiting example described in detail below, the system is implemented in an advanced driver assistance system (“ADAS”) having a gyroscope for determining the vehicle’s attitude and orientation and integrating the vehicle’s angular rates, such as pitch rate, roll rate, and yaw rate. However, it is contemplated that the system could be implemented in any other suitable automotive or non-automotive system having a gyroscope. The computer is programmed to estimate the vehicle’s kinematics using the radar unit, tracker, wheel speed sensor, temperature sensor, and gyroscope without requiring the vehicle to stop. As described in detail below, the tracker identifies targets based on the lack of motion of one or more static targets. The processor uses predictions from a Kalman filter to adjust the provided covariance to reduce the effects of unmodeled noise. The processor also utilizes a Kalman filter to fuse radar signals, WSS information, ambient temperature data, and gyroscope signals to estimate corrected or unbiased yaw rates.

[0071] Now for reference Figure 1Non-limiting examples of motor vehicle 100 include system 102 implemented as an advanced driver assistance system 102 (“ADAS”). While motor vehicle 100 is illustrated as a passenger car, it can be any other suitable type of transportation, including trucks, sports utility vehicles (SUVs), motorhomes, buses, aircraft, ships, etc., with system 102. In at least one example, motor vehicle 100 is configured to operate in at least one of several autonomous driving modes, as defined by the Society of Automotive Engineers (SAE) (SAE defines operation at levels 0-5). For example, motor vehicle 100 can receive and process two-dimensional and / or three-dimensional data about its surroundings and can also be programmed and / or configured to store and execute logical instructions contained in hardware, software, firmware, combinations thereof, etc., enabling it to operate with some user assistance (partial automation) or without any user assistance (fully autonomous driving). For example, at levels 0-2, a human driver monitors or controls most driving tasks, typically without assistance from motor vehicle 100. For example, at Level 0 (“No Automation”), the human driver is responsible for all vehicle operations. At Level 1 (“Driver Assistance”), vehicle 100 sometimes assists with steering, acceleration, or braking, but the driver remains responsible for the vast majority of vehicle control. At Level 2 (“Partial Automation”), vehicle 100 can control steering, acceleration, and braking in specific situations without human interaction. At Levels 3-5, vehicle 100 undertakes more driving-related tasks. At Level 3 (“Conditional Automation”), vehicle 100 can handle steering, acceleration, and braking in certain situations, as well as monitor the driving environment. However, Level 3 may require occasional driver intervention. At Level 4 (“High Automation”), vehicle 100 can handle the same tasks as at Level 3, but without relying on driver intervention in certain driving modes. At Level 5 (“Full Automation”), vehicle 100 can handle all tasks without any driver intervention. In at least one example, vehicle 100 is configured to operate according to one of Levels 2-5.

[0072] ADAS 102 may include one or more sensors 104 and one or more computers 106. In this non-limiting example, sensor 104 may include a gyroscope 108 for generating a gyroscopic signal associated with a measured yaw rate at the current time step. Sensor 104 may also include a pair of wheel speed sensors 110, 112 coupled to one of the two wheels 118, 120. Each wheel speed sensor 110, 112 generates a wheel speed signal associated with the current speed of the vehicle 100. Sensor 104 may also include a radar unit 118 for generating radar signals associated with detected targets located around the vehicle 100. Sensor 104 may also include a tracker 120 coupled to radar unit 118 and generating a tracking signal associated with the current radar heading and current Doppler effect of a stationary target based on the radar signal. More specifically, tracker 120 determines the associated untracked heading, Doppler effect (along with the corresponding error covariance), and identification of the stationary target. Sensor 104 may also include temperature sensor 122 for generating a temperature signal associated with ambient temperature.

[0073] It is conceivable that ADAS can have any number of these sensors and / or other suitable sensors. A non-exhaustive and non-limiting list of other vehicle sensors may include one or more of the following: vehicle pitch sensors, vehicle roll sensors, motion sensors, proximity sensors, laser identification detection and ranging (LIDAR) sensors, imaging sensors (e.g., complementary metal oxide semiconductor (CMOS) sensors, charge-coupled sensors (CCD), image enhancement sensors, etc.), infrared sensors, thermal sensors, short-range, medium-range, or long-range wireless signal sensors, vehicle position sensors (e.g., Global Positioning System (GPS) and Global Navigation Satellite System (GLONASS) sensors), vehicle acceleration sensors, vehicle braking sensors, and vehicle steering sensors, to name just a few.

[0074] System 102 also includes one or more computers 106. Each computer 106 includes one or more processors 124 coupled to gyroscope 108, wheel speed sensors 110, 112, tracker 120, and temperature sensor 122. Each computer 106 also includes a non-transitory computer-readable storage medium (“CRM”) 126 containing instructions that cause processor 124 to be programmed to determine gyroscopic drift and corrected yaw rate as the vehicle 100 moves, wherein the gyroscopic drift and corrected yaw rate are based at least on the current radar heading, the current Doppler effect, the current speed of the vehicle 100, and the measured yaw rate at the current time step. More specifically, as described in the following non-limiting detailed example, processor 124 is programmed to receive, at the current time step, a tracking signal from tracker 120, wheel speed signals from wheel speed sensors 110 and 112, a temperature signal from temperature sensor 122, and a gyroscope signal from gyroscope 108, determine gyroscope drift, and determine a corrected yaw rate based on the gyroscope drift.

[0075] Processor 124 is programmed to determine the innovation vector using a Kalman filter based on the difference between the tracking signal, wheel speed signal, temperature signal, gyroscope signal, and multiple associated predicted observations. Processor 124 determines the whitening innovation vector according to the following formula:

[0076]

[0077] Where e represents the current information, which is related to the error between the actual measurement vector z and the predicted measurement vector (i.e., e = z - Hx, where x is the predicted state vector and H is the linearized measurement covariance matrix); where R -1 / 2 This represents the inverse of the Choleski decomposition of the linearized measurement covariance matrix.

[0078] Processor 124 is programmed to generate an adapted measurement covariance matrix based on the innovation vector and further based on the measurement covariance matrix to enhance robustness. More specifically, processor 124 generates the adapted measurement covariance matrix according to the following formula:

[0079]

[0080]

[0081]

[0082] in The kernel function is represented by β; the positive tuning parameter is represented by β. Representing vectors In the matrix, element i; C represents the adaptive measurement covariance matrix; Let represent the updated measurement covariance matrix; and T denote the matrix transpose operator. After running the Kalman filter update step, the process can be iterated using the latest state estimate instead of the predicted state to determine the innovation.

[0083] Processor 124 is programmed to update the Kalman filter using an implicit nonlinear observation model to further estimate at least one gyroscope scaling factor and gyroscope drift according to the following equation:

[0084]

[0085] in The yaw rate represents the corrected yaw rate of a motor vehicle; where This represents the X-coordinate of the radar unit on the vehicle; where Let represent the Y-coordinate of the radar unit on the vehicle; where β0 represents the nominal installation azimuth angle of the radar unit; where β represents the calibration error of the radar unit; and where β represents the calibration error of the radar unit. Indicates the radar heading associated with the tracking signal; where This represents the lateral speed of a motor vehicle; where This represents the longitudinal speed of a motor vehicle; where This represents the estimated longitudinal speed of the motor vehicle; where α represents the Doppler velocity associated with the tracked signal; where α gyro This represents the gyroscope scaling factor; where α wss This represents the wheel speed sensor scaling factor; where γ gyro This represents gyroscope drift; where n represents iid noise; and where This represents the measured yaw rate associated with the gyro signal generated by the gyroscope.

[0086] Processor 124 is programmed to predict at least one of the following based on an updated Kalman filter: predicted radar heading, predicted Doppler effect, estimated gyro scaling factor, wheel speed sensor scaling factor of wheel speed sensor, and gyro drift of gyroscope 108. More specifically, processor 124 is programmed to predict at least one of the estimated gyro scaling factor, wheel speed sensor scaling factor, and gyro drift using a process model based on an updated Kalman filter according to the following equation:

[0087]

[0088] in Let represent the predicted value of the gyroscope drift; where t represents the current time step; and t+1 represents the next time step.

[0089] Processor 124 is also programmed to predict, based on an updated Kalman filter, at least one of the estimated gyroscope scaling factor, wheel speed sensor scaling factor, and gyroscope drift according to the following formula:

[0090]

[0091] Processor 124 is also programmed to predict gyroscope drift based on the following formula:

[0092]

[0093]

[0094]

[0095] The gyroscope drift is modeled as a zero-mean Gaussian process with a known covariance kernel k(·,·).

[0096] Processor 124 is also programmed to predict the gyroscope drift of the gyroscope according to the following formula:

[0097]

[0098]

[0099]

[0100] in This represents the independent predictor of the gyroscope's gyroscope drift; where t represents the current time step; where Represents longitudinal velocity; where Indicates the yaw rate; where The predictor variable represents the gyroscope drift of the gyroscope at the previous time step; and where The predictor variable represents the gyroscope drift of the gyroscope at the current time step.

[0101] Processor 124 is also programmed to predict gyroscope drift based on the spectral mixing kernel:

[0102]

[0103] Where C = C(ψ) is a spherically parameterized correlation matrix that modifies the predictors at previous and current time steps. Modeling the correlation between them; where multiple parameters are learned from the data. For computational reasons, it can be assumed that the prediction vector lies on a regular grid, thus allowing the use of the Kronecker structure.

[0104] Now for reference Figure 2 It shows the operation Figure 1 The process 200 of the system 102 shown is a non-limiting example. The process 200 begins at block 202, where radar unit 118 generates a radar signal associated with a detected target 128 located around the motor vehicle 100.

[0105] In box 204, tracker 120 generates a tracking signal based on radar signals that is associated with the current radar heading and current Doppler effect of the stationary target.

[0106] In frame 206, wheel speed sensors 110 and 112 generate wheel speed signals associated with the current speed of the motor vehicle 100.

[0107] In box 208, temperature sensor 120 generates a temperature signal that is correlated with the ambient temperature.

[0108] In box 210, gyroscope 108 generates a gyroscope signal associated with the measured yaw rate at the current time step.

[0109] In block 212, as the vehicle 100 moves, the processor 124 determines the gyro drift and corrected yaw rate of the vehicle 100 during its movement, wherein the gyro drift and corrected yaw rate are based at least on the current radar heading, the current Doppler effect, the vehicle's current speed, and the yaw rate measured at the current time step. More specifically, the processor 124 receives gyro signals from gyroscope 108, wheel speed signals from wheel speed sensors 110 and 112, a tracking signal from tracker 120, and a temperature signal from temperature sensor 122 at the current time step. The processor 124 uses a Kalman filter to determine innovation based on the difference between the tracking signal, wheel speed signal, temperature signal, gyro signal, and multiple associated predicted observations according to Equation 1 above. The processor also generates an adapted measurement covariance matrix based on the whitening innovation determined at the current time step and further based on the measurement covariance matrix according to Equations 2 to 4 above to enhance robustness. Processor 124 also updates the Kalman filter based on the tracking signal, wheel speed signal, temperature signal, and gyroscope signal according to Equation 5 above.

[0110] In block 214, processor 124 predicts at least one of the following items based on the updated Kalman filter according to Equations 6 to 15 above: predicted radar heading, predicted Doppler effect, estimated gyroscope scaling factor of gyroscope, wheel speed sensor scaling factor of wheel speed sensor, and gyroscope drift of gyroscope.

[0111] Therefore, machine learning systems incorporating fuzzy controllers have been described, which receive datasets, such as multiple state-action values, from a reinforcement learning agent (RLA) controller. The controller can be software, hardware, or a combination thereof. By using two cascaded controllers, where the output (data) of the RLA controller is received as input to the fuzzy controller, a machine learning system can be created where the output data can be interpreted using the fuzzy controller. In this way, engineers, system designers, and others can gain a more comprehensive understanding and / or solve their implementation problems.

[0112] According to a non-limiting example, computer 106 may include a vehicle engine control computer, a vehicle braking system computer, and a vehicle steering control computer, wherein each computer 106 executes instructions to perform at least one automated or partially automated vehicle operation (e.g., adaptive cruise control (ACC), lane keeping assist, lane departure warning, forward collision warning, automatic emergency braking, pedestrian detection, and blind spot warning, to name just a few). It should be understood that system 102 does not require multiple computers. For example, the system may have more computers. The vehicle engine control computer, vehicle braking system computer, and vehicle steering control computer are merely examples. A non-exhaustive and non-limiting list of vehicle computers may include a body control module (BCM), a powertrain control module (PCM), a power transfer unit (PTU), and a suspension control module, to name just a few. As will be described in more detail below, by providing output to one or more computers 106, system 102 can activate vehicle functions (e.g., control vehicle acceleration, control vehicle braking, and / or control vehicle steering).

[0113] Generally, the described computing system and / or device may employ any of a variety of computer operating systems, including, but not limited to, versions and / or variations of the MYLINK or INTELLILINK applications. Examples of computing devices include, but are not limited to, in-vehicle computers, computer workstations, servers, desktop computers, laptops, handheld computers, or other computing systems and / or devices.

[0114] Computing devices typically include computer-executable instructions that can be executed by one or more computing devices (such as those listed above). Computer-executable instructions can be compiled or interpreted from computer programs created using various programming languages ​​and / or technologies, including but not limited to JAVA, C, C++, VISUAL BASIC, JAVASCRIPT, PERL, etc., alone or in combination. Some of these applications can be compiled and executed on virtual machines, such as the Java Virtual Machine, Dalvik Virtual Machine, etc. Typically, a processor (e.g., a microprocessor) receives instructions from memory, computer-readable media, etc., and executes those instructions to perform one or more processes, including one or more processes described herein. These instructions and other data can be stored and transferred using various computer-readable media.

[0115] Computer-readable media (also known as processor-readable media) include any non-transitory (e.g., tangible) medium that contributes to providing data (e.g., instructions) that can be read by a computer (e.g., by the computer's processor). Such media can take many forms, including, but not limited to, non-volatile and volatile media. Non-volatile media can include, for example, optical discs or magnetic disks, and other permanent storage devices. Volatile media can include, for example, dynamic random-access memory (DRAM), which typically constitutes main memory. Such instructions can be transmitted via one or more transmission media, including coaxial cables, copper wires, and optical fibers, including wires containing a system bus coupled to the computer processor. Common forms of computer-readable media include, for example, floppy disks, floppy disks, hard disks, magnetic tape, any other magnetic media, CD-ROMs, DVDs, any other optical media, punched cards, paper tape, any other physical media with a perforated pattern, RAM, PROM, EPROM, FLASH-EEPROM, any other memory chips or cassette tapes, or any other computer-readable medium.

[0116] The databases, data repositories, or other data storage areas described in this document can include various mechanisms for storing, accessing, and retrieving a wide variety of data, including hierarchical databases, a set of files in a file system, application databases in proprietary formats, relational database management systems (RDBMS), and so on. Each such data storage area is typically contained within a computing device employing a computer operating system such as one of those described above, and is accessed via a network in any one or more of various ways. File systems can be accessed from the computer operating system and can include files stored in various formats. In addition to languages ​​used for creating, storing, editing, and executing stored procedures, RDBMS typically use a Structured Query Language (SQL), such as the PL / SQL language mentioned above.

[0117] In some examples, system elements may be implemented as computer-readable instructions (e.g., software) on one or more computing devices (e.g., servers, personal computers, etc.) and stored on an associated computer-readable medium (e.g., disks, storage, etc.). A computer program product may include instructions stored on a computer-readable medium for performing the functions described herein.

[0118] The processor is implemented through circuits, chips, or other electronic components and may include one or more microcontrollers, one or more field-programmable gate arrays (FPGAs), one or more application-specific circuits (ASICs), one or more digital signal processors (DSPs), one or more user-defined integrated circuits, etc. The processor can be programmed to process sensor data. Data processing may include processing video feeds or other data streams captured by sensors to determine the main vehicle's road lane and the presence of any target vehicles. As described below, the processor instructs vehicle components to initiate actions based on the sensor data. The processor may be integrated into a controller, such as an autonomous mode controller.

[0119] Memory (or data storage device) is implemented through circuits, chips, or other electronic components, and may include one or more of the following: read-only memory (ROM), random access memory (RAM), flash memory, electrically programmable memory (EPROM), electrically programmable and erasable memory (EEPROM), embedded multimedia card (eMMC), hard disk drive, or any volatile or non-volatile media. Memory can store data collected from sensors.

[0120] Regarding the media, processes, systems, methods, and inspirations described herein, it should be understood that although the steps of these processes are described as occurring according to an ordered sequence, these processes can be practiced with the described steps performed in a different order than those described herein. It should also be understood that some steps may be performed simultaneously, other steps may be added, or some steps described herein may be omitted. In other words, the process descriptions herein are provided for the purpose of illustrating certain embodiments and should not be construed as limiting the claims in any way.

[0121] Therefore, it should be understood that the above description is illustrative and not restrictive. Many embodiments and applications beyond the examples provided will be apparent to those skilled in the art upon reading the above description. The scope of the invention should not be determined by reference to the foregoing description, but rather by reference to the appended claims and the full scope of their equivalents. Future developments are anticipated and intended in the art discussed herein, and the disclosed systems and processes will be incorporated into such future embodiments. In summary, it should be understood that the invention is capable of modifications and alterations and is limited only by the appended claims.

[0122] All terms used in the claims are intended to be given the simple and common meaning as understood by those skilled in the art, unless expressly indicated otherwise herein. In particular, the use of singular articles such as “a,” “the,” “the,” etc., should be interpreted as listing one or more of the indicated elements, unless the claims expressly limit them to the contrary.

Claims

1. A system for a motor vehicle, the system comprising: A radar unit for generating radar signals that are associated with detected targets located around the vehicle; A tracker, coupled to the radar unit, generates a tracking signal based on the radar signal, the tracking signal being correlated with the current radar heading and current Doppler effect of the stationary target; A wheel speed sensor is used to generate a wheel speed signal, which is associated with the current speed of the motor vehicle; A temperature sensor is used to generate a temperature signal that is correlated with the ambient temperature; A gyroscope is used to generate a gyroscope signal that is correlated with the measured yaw rate at the current time step. and Computers, including: At least one processor is coupled to the tracker, the wheel speed sensor, the temperature sensor, and the gyroscope; and A non-transitory computer-readable storage medium, including instructions, such that the at least one processor is programmed to: At the current time step, the system receives a tracking signal from the tracker, a wheel speed signal from the wheel speed sensor, a temperature signal from the temperature sensor, and a gyroscope signal from the gyroscope. The Kalman filter is used to determine the innovation based on the difference between these observed signals and the predicted observations; Based on the whitening innovation determined at the current time step and also based on the measurement covariance matrix, an adapted measurement covariance matrix is ​​generated. The Kalman filter is updated based on the tracking signal, the wheel speed signal, the temperature signal, and the gyroscope signal. Based on the updated Kalman filter, at least one of the following items is predicted: the predicted radar heading, the predicted Doppler effect, the estimated gyroscope scaling factor of the gyroscope, the wheel speed sensor scaling factor of the wheel speed sensor, and the gyroscope drift. as well as Determine the gyro drift and corrected yaw rate while the vehicle is in motion, wherein the gyro drift and the corrected yaw rate are based at least on the current radar heading, the current Doppler effect, the current speed of the vehicle, and the yaw rate measured at the current time step.

2. The system according to claim 1, wherein the at least one processor determines the whitening innovation based on the maximum correlation entropy change of the Kalman filter by calculating the following formula. : in This represents the current information, compared to the actual measurement vector. The error is related to the predicted measurement vector; in This represents the inverse of the Choleski decomposition of the measurement covariance matrix.

3. The system of claim 2, wherein the at least one processor generates the adapted measurement covariance matrix according to the following formula: in Represents the kernel function; in Indicates the positive tuning parameter; in Representing vectors elements in ; in The measurement covariance matrix representing the adaptation; in This represents the updated measurement covariance matrix; and in This represents the matrix transpose operator.

Images