Model predictive lane centering control with reference switching and disturbance suppression

Abstract

A system for model predictive lane centering (MPLC) control with reference switching and disturbance suppression includes an MPLC application having control logic that utilizes model predictive control (MPC) to selectively track static and dynamic references, and adjusts MPC weights and constraints for trajectory tracking when the system switches between the static and dynamic references. A Kalman filter estimates lateral force disturbances and yaw moment disturbances acting on the vehicle, as well as measured deviations that disrupt the measured yaw rate. The system detects anomalies in the MPC reference trajectory and adjusts actuator constraints in response to the detected anomalies. The system actively and continuously adjusts actuator outputs, causing the vehicle to track and follow a current lane center, and upon receiving a lane change command, smoothly changes lanes using the vehicle actuators to change position and change vehicle trajectory to enter an adjacent lane before returning to track and follow the center of the adjacent lane.

Description

Model-predictive lane centering control with reference switching and disturbance suppression Technical Field

[0001] This disclosure relates to vehicles, and more specifically, to vehicles that utilize advanced driver assistance systems (ADAS) to fully or partially control the movement of the vehicle. Background Technology

[0002] Intelligent vehicle technologies such as free-roaming grid navigation, parking guidance, and information systems help prevent human error when drivers operate the vehicle. These technologies have been used to improve road navigation and enhance the driver's parking ability when the driver is inside the vehicle. For example, rearview camera systems and collision warning systems have been developed to assist vehicle operators in avoiding collisions while parking. Furthermore, autonomous driving systems have been developed that allow vehicles to navigate roads automatically or semi-automatically.

[0003] While such systems are useful for their intended purpose, they often result in discontinuous or unstable maneuvering between lanes after receiving a lane change request initiated by the driver or the system. Therefore, there is still a need in the art for improved intelligent vehicle technologies that leverage pre-existing infrastructure to automatically and / or semi-automatically allow vehicles to smoothly navigate lane change requests with high precision, robustness, and redundancy, while maintaining or reducing computational resource utilization and minimizing the impact of disturbances on the smoothness of vehicle operation. Summary of the Invention

[0004] According to several aspects, a system for model predictive lane centering (MPLC) control with reference switching and disturbance suppression includes: a vehicle, one or more sensors mounted on the vehicle, the one or more sensors measuring real-time static and dynamic data about the vehicle, and one or more actuators mounted on the vehicle. The one or more actuators alter the static and dynamic characteristics of the vehicle. The system also includes a controller having a processor, memory, and input / output (I / O) ports communicating with the one or more sensors and the one or more actuators. The processor executes a portion of program code stored in memory. The program code portion includes an MPLC application program having at least first, second, third, fourth, and fifth control logic. The first control logic utilizes model predictive control (MPC) to selectively track static and dynamic references. The second control logic adjusts MPC weights and constraints for trajectory tracking as the system switches between static and dynamic references. The third control logic utilizes a Kalman filter to estimate lateral force disturbances and yaw moment disturbances acting on the vehicle, as well as measurement deviations that disrupt the measured yaw rate. The fourth control logic detects anomalies in the MPC reference trajectory. The fifth control logic adjusts actuator constraints in response to detected anomalies. The MPLC control system actively and continuously adjusts the actuator output commands to selectively make the vehicle track and follow the current lane center. When it receives a lane change command, it smoothly performs lane changes by instructing one or more actuators to change position and smoothly change the vehicle trajectory to enter the adjacent lane before returning to track and follow the center of the adjacent lane.

[0005] In another aspect of this disclosure, the first control logic further includes control logic for acquiring data from one or more sensors and one or more actuators, and control logic for performing static reference tracking. Static reference tracking causes the vehicle to follow and conform to the current lane center. The first control logic also includes control logic for deviating the vehicle from the current lane center while following and tracking it.

[0006] In another aspect of this disclosure, when a lane change command is received from one or more of the vehicle occupants and advanced driver assistance systems (ADAS), the MPLC application switches to tracking a dynamic reference. The dynamic reference is a smooth path extending between the current static reference and the future static reference.

[0007] In another aspect disclosed, the second control logic also includes control logic for adaptively adjusting the MPC weights based on the control objective. The weights change when the reference changes. The second control logic also includes control logic that adaptively adjusts the weights and constraints upon receiving a lane change command and then ramps them back to their nominal values ​​over a predetermined time period, thereby smoothing the transition between tracking static and dynamic references.

[0008] In another aspect of this disclosure, the second control logic also includes control logic to penalize rapid changes in the actuator position, and three different inputs to the MPC depending on whether the reference is static or dynamic, according to the following formula: initial lateral error (e y ), initial heading error (e ψ ) and reference curvature waypoints Control logic: Where: V x or V xr Define the longitudinal velocity of the vehicle; V y ω represents the lateral velocity of the vehicle. z Or ψ represents the vehicle's yaw rate; e y e represents the lateral error (distance) from the target path. ψ ψ represents the heading error from the target path. r The yaw rate is a reference value for the target path (related to the curvature of the target path); m is the vehicle mass; I zz The moment of inertia of the vehicle; l f l r C is the distance from the vehicle's center of gravity to the front and rear axles. f C r These are the front and rear tire lateral stiffness, respectively; δ f Forward steering angle, δ r Rear steering angle; A yd and R zd These are disturbances of lateral acceleration (force) and yaw acceleration (torque), respectively; x, x d These are the state vector and the perturbation / uncertainty state vector, respectively; and a hat "^" on top of any variable indicates that the capped variable is an estimate.

[0009] In another aspect of this disclosure, the third control logic further includes: control logic for compensating for the estimated lateral force disturbance and the estimated yaw moment disturbance by applying each of the estimated lateral force disturbance and the estimated yaw moment disturbance along the prediction range in the MPC; and control logic for compensating for the estimated measurement deviation by utilizing the yaw rate adjusted by the deviation in the initial state of the MPC.

[0010] In another aspect of this disclosure, the Kalman filter estimation of lateral force disturbances and estimated yaw moment disturbances uses measured vehicle speed and vehicle motion relative to a measured static reference. The measured static reference defines a static reference path.

[0011] In another aspect disclosed herein, the fourth control logic further includes control logic for detecting anomalies in the MPC reference trajectory by comparing the trajectory offset, heading, and curvature with moving averages of the trajectory offset, heading, and curvature. The fourth control logic also includes control logic for detecting anomalies when the difference between one or more of the trajectory offset, heading, and curvature and the moving average of the trajectory offset, heading, and curvature exceeds a predetermined threshold.

[0012] In another aspect of this disclosure, the fourth control logic also includes control logic that, under predefined conditions, includes disabling anomaly detection using semantics from the trajectory planner when a lane change has been requested.

[0013] In another aspect of this disclosure, the fifth control logic further includes control logic that, upon detecting an anomaly, adjusts the actuator constraints by reducing the actuator amplitude and rate limits from a first level to a second level less than the first level for a predetermined time period. The actuator amplitude limit is adjusted to limit the vehicle's lateral acceleration to within the predicted range based on the curvature of the last anomaly-free reference trajectory. The fifth control logic also includes control logic that progressively decreases the actuator rate limit to a calibrable fraction of the nominal actuator rate limit during the predetermined time period. Both the amplitude and rate limits are temporarily held at the reduced values ​​before linearly increasing back to the nominal values.

[0014] In another aspect of this disclosure, a method for Model Predictive Lane Centering (MPLC) control with reference switching and disturbance suppression includes: measuring real-time static and dynamic data about the vehicle using one or more sensors mounted on the vehicle. The method also includes altering the static and dynamic characteristics of the vehicle using one or more actuators mounted on the vehicle, and executing a portion of program code stored in the memory of a controller, which includes input / output (I / O) ports communicating with the one or more sensors and the one or more actuators, by a processor of the vehicle's controller. The program code portion includes an MPLC application having control logic for: selectively tracking static and dynamic references using model predictive control (MPC), adjusting MPC weights and constraints for trajectory tracking when switching between static and dynamic references, and estimating lateral force disturbances and yaw moment disturbances acting on the vehicle using a Kalman filter, as well as measurement deviations from the measured yaw rate. The MPLC application also includes control logic for detecting anomalies in the MPC reference trajectory and for adjusting actuator constraints in response to detected anomalies. The MPLC application actively and continuously adjusts the actuator output commands to selectively make the vehicle track and follow the current lane center, and when it receives a lane change command, it smoothly performs lane changes by instructing one or more actuators to change position and smoothly change the vehicle trajectory to enter the adjacent lane before returning to track and follow the center of the adjacent lane.

[0015] In another aspect of this disclosure, the method further includes acquiring data from one or more sensors and one or more actuators, performing static reference tracking, wherein static reference tracking causes the vehicle to follow and conform to the current lane center; and causing the vehicle to deviate from the current lane center while following and conforming to it.

[0016] In another aspect of this disclosure, when a lane change command is received from one or more of the vehicle occupants and advanced driver assistance systems (ADAS), the method executes control logic to switch to an MPLC application that tracks dynamic references. Dynamic references are smooth paths extending between current static references and future static references.

[0017] In another aspect of this disclosure, the method also includes adaptively adjusting MPC weights based on control objectives, wherein the weights change when the reference changes, and adaptively adjusting the weights and constraints upon receiving a lane change command and subsequently ramping the weights and constraints back to nominal values ​​over a predetermined time period, thereby smoothing the transition between tracking static and dynamic references.

[0018] In another aspect of this disclosure, the method also includes penalizing rapid changes in the actuator position, and changing three different inputs to the MPC according to the following formula, depending on whether the reference is a static or dynamic reference: the initial lateral error (e y ), initial heading error (e ψ ) and reference curvature waypoints Where: V x or V xr Define the longitudinal velocity of the vehicle; V y ω represents the lateral velocity of the vehicle. z Or ψ represents the vehicle's yaw rate; e y e represents the lateral error (distance) from the target path. ψ ψ represents the heading error from the target path. r The yaw rate is a reference value for the target path (related to the curvature of the target path); m is the vehicle mass; I zz The moment of inertia of the vehicle; l f l r C is the distance from the vehicle's center of gravity to the front and rear axles. f C r These are the front and rear tire lateral stiffness, respectively; δ f Forward steering angle, δ r Rear steering angle; A yd and R zd These are disturbances of lateral acceleration (force) and yaw acceleration (torque), respectively; x, x dThese are the state vector and the perturbation / uncertainty state vector, respectively; and a hat "^" on top of any variable indicates that the capped variable is an estimate.

[0019] In another aspect of this disclosure, the method further includes compensating for the estimated lateral force disturbance and the estimated yaw moment disturbance by applying each of the estimated lateral force disturbance and the estimated yaw moment disturbance along the prediction range in the MPC; and compensating for the estimated measurement bias by utilizing the bias-adjusted yaw rate in the initial state of the MPC.

[0020] In another aspect of this disclosure, the method further includes using measured vehicle speed and vehicle motion relative to a measured static reference. The measured static reference defines a static reference path to generate Kalman filter estimates of lateral force disturbances and estimated yaw moment disturbances.

[0021] In another aspect of this disclosure, the method further includes detecting anomalies in the MPC reference trajectory by comparing trajectory offset, heading, and curvature with moving averages of trajectory offset, heading, and curvature. The method also includes detecting anomalies when the difference between one or more of trajectory offset, heading, and curvature and the moving average of trajectory offset, heading, and curvature exceeds a predetermined threshold; and, in predefined cases, including when a lane change has been requested, using semantics from the trajectory planner to disable anomaly detection.

[0022] In another aspect of this disclosure, the method further includes adjusting the actuator constraints upon detecting an anomaly by reducing the actuator amplitude and rate limits from a first level to a second level less than the first level for a predetermined time period, wherein the actuator amplitude limit is adjusted to limit the vehicle's lateral acceleration to within the range of the predicted lateral acceleration based on the curvature of the last anomaly-free reference trajectory; and the actuator rate limit is progressively reduced to a calibrable fraction of the nominal actuator rate limit during the predetermined time period. Both the amplitude and rate limits are temporarily held at the reduced values ​​before linearly increasing back to the nominal values.

[0023] In another aspect of this disclosure, a method for Model Predictive Lane Centering (MPLC) control with reference switching and disturbance suppression includes: measuring real-time static and dynamic data about the vehicle using one or more sensors mounted on the vehicle; altering the static and dynamic characteristics of the vehicle using one or more actuators mounted on the vehicle; and executing a portion of program code stored in the memory of the controller by a processor of the vehicle's controller. The controller also includes input / output (I / O) ports communicating with one or more sensors and one or more actuators. The program code portion includes an MPLC application having control logic for selectively tracking static and dynamic references using model predictive control (MPC), including: acquiring data from one or more sensors and one or more actuators; and performing static reference tracking. Static reference tracking causes the vehicle to follow and conform to the current lane center. The MPLC application also includes control logic for deviating the vehicle from the current lane center while following and tracking it. When a lane change command is received from one or more of the vehicle occupants and advanced driver assistance systems (ADAS), control logic for switching to tracking a dynamic reference in the MPLC application is executed. The dynamic reference is a smooth path extending between the current static reference and a future static reference. The MPLC application also includes control logic for adjusting MPC weights and constraints for trajectory tracking when switching between static and dynamic references. This includes: adaptively adjusting MPC weights based on control objectives, where weights change as the reference changes; and adaptively adjusting weights and constraints upon receiving a lane change command and subsequently ramping them back to nominal values ​​over a predetermined time period, thus smoothing the transition between tracking static and dynamic references. The MPLC application also includes control logic for penalizing rapid changes in actuator position; depending on whether the reference is static or dynamic, the three different inputs to the MPC are changed according to the following formula: initial lateral error (e... y ), initial heading error (e ψ ) and reference curvature waypoints Where: V x or V xr Define the longitudinal velocity of the vehicle; V y ω represents the lateral velocity of the vehicle. z or ψ e represents the vehicle's yaw rate. y e represents the lateral error (distance) from the target path. ψ ψ represents the heading error from the target path. r yaw rate reference for the target path (related to the curvature of the target path); m is the mass of the vehicle; I zz The moment of inertia of the vehicle; l f lr C is the distance from the vehicle's center of gravity to the front and rear axles. f C r These are the front and rear tire lateral stiffness, respectively; δ f Forward steering angle, δ r Rear steering angle; A yd and R zd These are disturbances of lateral acceleration (force) and yaw acceleration (torque), respectively; x, x d These are the state vector and the disturbance / uncertainty state vector, respectively; and a hat "^" on top of any variable indicates that the hatted variable is an estimate. The MPLC application also includes control logic for using a Kalman filter to estimate lateral force disturbances and yaw moment disturbances acting on the vehicle, as well as measurement biases that disrupt the measured yaw rate, including: compensating for the estimated lateral force disturbances and estimated yaw moment disturbances by applying each of the estimated lateral force disturbances and estimated yaw moment disturbances along the prediction range in the MPC; and compensating for the estimated measurement biases by using a bias-adjusted yaw rate in the initial state of the MPC. The MPLC application also includes control logic for using measured vehicle speed and vehicle motion relative to a measured static reference, which defines a static reference path to generate Kalman filter estimates of lateral force disturbances and estimated yaw moment disturbances, and for detecting anomalies in the MPC reference trajectory, including: detecting anomalies in the MPC reference trajectory by comparing trajectory offset, heading, and curvature with moving averages of trajectory offset, heading, and curvature, and detecting anomalies when the difference between one or more of trajectory offset, heading, and curvature and the moving averages of trajectory offset, heading, and curvature exceeds a predetermined threshold. The MPLC application also includes control logic for disabling anomaly detection using semantics from the trajectory planner under predefined conditions, including when a lane change has been requested; and for adjusting actuator constraints in response to detected anomalies, including: adjusting actuator constraints by reducing actuator amplitude and rate limits from a first level to a second level less than the first level for a predetermined time period upon detection of an anomaly, wherein the actuator amplitude limit is adjusted to limit the vehicle's lateral acceleration to within the predicted range based on the curvature of the last anomaly-free reference trajectory; and progressively decreasing the actuator rate limit to a calibrable fraction of the nominal actuator rate limit during the predetermined time period. Both amplitude and rate limits are temporarily held at the reduced values ​​before linearly increasing back to the nominal values. The MPLC application proactively and continuously adjusts actuator output commands to selectively cause the vehicle to track and follow the current lane center, and upon receiving a lane change command, smoothly performs a lane change by instructing one or more actuators to change position and smoothly alter the vehicle trajectory to enter the adjacent lane before returning to tracking and following the center of the adjacent lane.

[0024] 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

[0025] The accompanying drawings described herein are for illustrative purposes only and are not intended to limit the scope of this disclosure in any way.

[0026] Figure 1 is a schematic diagram of a system for Model Predictive Lane Centering (MPLC) control with reference switching and disturbance suppression according to an exemplary embodiment;

[0027] Figure 2A is a partial schematic diagram of a vehicle equipped with the MPLC control system of Figure 1 that performs static reference tracking, according to an exemplary embodiment.

[0028] Figure 2B is a partial schematic diagram of a vehicle equipped with the MPLC control system of Figure 1 that performs lane changes, according to an exemplary embodiment.

[0029] Figure 3 is a flowchart depicting how the output of the MPLC control system of Figure 1 is adjusted to overcome disturbances using a Kalman filter and actuator constraints according to an exemplary embodiment.

[0030] Figure 4 is a flowchart illustrating zero-offset control using a Kalman filter to adjust the output of the MPLC control system of Figure 1 to overcome disturbances, according to an exemplary embodiment; and

[0031] Figure 5 is a top perspective view of a vehicle equipped with the MPLC control system of Figure 1, wherein the vehicle according to an exemplary embodiment is shown to be subjected to forces, torques and moments under several different conditions. Detailed Implementation

[0032] The following description is merely exemplary in nature and is not intended to limit this disclosure, its application, or its uses.

[0033] Referring to Figure 1, a Model Predictive Lane Centering Control (MPLC) system 10 is schematically illustrated. System 10 includes a vehicle 12. Vehicle 12 is shown as a passenger car; however, it should be understood that without departing from the scope or intent of this disclosure, vehicle 12 can be any type of vehicle, including but not limited to: automobiles, trucks, sport utility vehicles (SUVs), vans, motorhomes, semi-trailers, tractor-trailers, delivery vehicles (including vehicles used in warehouses), tricycles, motorcycles, aircraft, amphibious vehicles, or any other such vehicle 12 that comes into contact with the ground. Vehicle 12 is equipped with at least one Human-Machine Interface (HMI) 13, which allows occupants of vehicle 12 to interact with various systems of vehicle 12. In some examples, HMI 13 may include a steering wheel and / or buttons and / or a dial and / or haptic interface disposed thereon, an accelerator pedal, a brake pedal, a transmission shift device, a clutch pedal, and one or more buttons or screens for the vehicle's entertainment system and / or heating, ventilation, and air conditioning (HVAC) system. In one non-limiting example, HMI 13 may include a touchscreen interface installed in the passenger compartment of vehicle 12. The touchscreen interface provides prompts and / or information to the occupants of vehicle 12, who can interact with the prompts and / or information to change functions of vehicle 12, including changing HVAC temperature control functions and advanced driver assistance system (ADAS) functions, including lane keeping and lane changing preferences. In some additional non-limiting examples, HMI 13 may include one or more turn signal indicator stalks or buttons, windshield wiper control stalks or buttons, etc.

[0034] Without departing from the scope of this disclosure, the MPLC system 10 may have various configurations, but typically includes a sensor subsystem 14 and a vehicle motion control (VMC) system 16. The sensor subsystem 14 includes a plurality of electromagnetic (EM) sensors 18A-18D mounted along the periphery of the vehicle 12. In the example provided in FIG. 1, the EM sensors 18A-18D are located at the front, left, right, and rear of the vehicle 12, respectively, to provide 360-degree overlapping coverage of the EM sensors 18A-18D. However, it should be understood that without departing from the scope or intent of this disclosure, the sensor subsystem 14 may have any number of sensors 18, and the EM sensors 18A-18D may cover, recover, or have overlapping fields of view 20. Each of the EM sensors 18A-18D is operable to collect or sense information in a predefined area around the vehicle 12. Information from the EM sensors 18A-18D is transmitted to a controller 22. The controller 22 communicates with both the sensor subsystem 14 and the VMC system 16. In one example, EM sensors 18A-18D include EM sensors such as cameras, infrared cameras, video cameras, light detection and ranging (LiDAR) sensors, radio detection and ranging (RADAR) sensors, and sound navigation and ranging (SONAR) sensors. Additional sensors 18 detect static and dynamic information about the vehicle 12. In several examples, additional sensors 18 can measure and record various static and dynamic state information or data about the vehicle 12. Additional sensors 18 may include, but are not limited to: an inertial measurement unit (IMU) 18E, a suspension control unit such as a semi-active damped suspension (SADS) sensor 18F, a global positioning system (GPS) sensor 18G, a wheel speed sensor 18H capable of measuring the rotational speed of one or more wheels 22 of the vehicle, a throttle position sensor 18I, an accelerator pedal position sensor 18J, a brake pedal position sensor 18K, a steering position sensor 18L capable of measuring the position, steering rate, and steering speed of the steering system 24, a tire pressure monitoring system 18M, an aerodynamic component position sensor 18N, and so on. The IMU 18E can measure movement, acceleration, and so on in multiple degrees of freedom. In a specific example, the IMU 18E can measure position, movement, acceleration, and so on in at least three degrees of freedom. Similarly, the SADS sensor 18F can be an IMU 18E capable of measuring in three or more degrees of freedom. In some examples, the SADS sensor 18F can be a suspension hub accelerometer, etc. Therefore, the vehicle 12 state information can include any of a variety of data, including but not limited to: wheel 22 speed data, SADS sensor 18F, GPS sensor 18G, and IMU 18E data, including attitude, acceleration, position, and so on.

[0035] Additionally, the MPLC system 10 is operable to move, manipulate, or otherwise drive the vehicle 12 automatically or semi-automatically via a plurality of actuators 26, which are capable of altering the motion of the vehicle 12 when it is operating in manual, semi-automatic, or fully automatic mode using an advanced driver assistance system (ADAS) 28. The vehicle 12 includes one or more wheels 30 connected to a drivetrain 32. The drivetrain 32 may include various components, such as an internal combustion engine (ICE) 34 and / or an electric motor 36, and a transmission 38 capable of transmitting the prime mover generated by the ICE 34 and / or the electric motor 36 to the wheels 30. In a non-limiting example, the vehicle 12 may include an ICE 34 operating on the rear axle 40 and one or more electric motors 36 operating on the front axle 42 of the vehicle 12. However, it should be understood that the vehicle 12 may use one or more ICE 34s and / or one or more electric motors 36 in other configurations without departing from the scope or intent of this disclosure. For example, the vehicle 12 may have an ICE 34 operating only on the front axle 42, while the electric motors 36 operate only on the rear axle 40. In a further example, the ICE 34 can act on both the front axle 42 and the rear axle 40, and the electric motor 36 can also act on both the front axle 42 and the rear axle 40. In several aspects, the drivetrain 32 includes one or more actuators, such as an all-wheel drive (AWD) system including an electronically controlled or electric AWD (eAWD) system, and a limited-slip differential (LSD) system including an electronically controlled limited-slip differential (eLSD) system, wherein each of the AWD, eAWD, LSD, and eLSD systems is capable of generating and / or modifying forces generated in the X and / or Y directions at the road contact surface with the tires 43 disposed on the wheels 30. The AWD and / or eAWD systems can transmit torque generated by the ICE 34 and / or the electric motor 36 to and from the front axle 42 and rear axle 40 of the vehicle 12, and / or from one side of the vehicle 12 to the other. Similarly, the LSD or eLSD system can transmit torque from one side of the vehicle 12 to the other. In some examples, AWD / LSD and / or eAWD / eLSD can directly alter or manage torque transmission from ICE 34 and / or electric motor 36, and / or AWD / LSD and / or eAWD / eLSD can act on the braking system 44 of vehicle 12 to adjust the amount of torque transmitted to each of the wheels 30 of vehicle 12.

[0036] In a further example, vehicle 12 may include means for altering the normal force on each wheel 30 of vehicle 12 via one or more active aerodynamic actuators 46 and / or active suspension actuators 48. Active aerodynamic actuators 46 may actively or passively alter the aerodynamic profile of one or more active aerodynamic elements 50 of vehicle 12, such as wings, spoilers, fans or other suction devices, actively managed venturi tubes, etc. Without departing from the scope or intent of this disclosure, active suspension actuators 48, such as active dampers 52, may be magnetorheological dampers or other such electro-, hydraulic, mechanical, or pneumatically adjustable dampers. For simplicity in the following description, ICE 34, electric motor 36, eAWD, eLSD, LSD, AWD, braking system 44, aerodynamic control system, active aerodynamic element 50, active damper 52, etc., will be simply and more broadly referred to as "actuator 26".

[0037] The terms “forward,” “rear,” “inner,” “inward,” “outer,” “outer,” “top,” and “bottom” are used relative to the orientation of the vehicle 12 as shown in the figures of this application. Thus, “forward” refers to the direction toward the front of the vehicle 12, and “rear” refers to the direction toward the rear of the vehicle 12. “Left” refers to the direction toward the left side of the vehicle 12 relative to the front of the vehicle 12. Similarly, “right” refers to the direction toward the right side of the vehicle 12 relative to the front of the vehicle 12. “Inner” and “inward” refer to the direction toward the interior of the vehicle 12, “outer” and “outer” refer to the direction toward the exterior of the vehicle 12, “bottom” refers to the direction toward the bottom of the vehicle 12, and “top” refers to the direction toward the top of the vehicle 12. Furthermore, the terms “top,” “at the top,” “bottom,” “side,” and “top” are used relative to the orientation of the actuator 26 and the vehicle 12, which are shown more broadly in the figures of this application. Therefore, although the orientation of actuator 52 or vehicle 12 may vary relative to a given purpose, these terms are intended to still be applied relative to the orientation of the components of system 10 and vehicle 12 shown in the accompanying drawings.

[0038] Each of the plurality of actuators 26 and sensors 18 equipped to vehicle 12 communicates with controller 22. Controller 22 is a non-general-purpose electronic control device having a pre-programmed digital computer or processor 54, non-transitory computer-readable medium or memory 56 for storing data such as control logic, software applications, instructions, computer code, data, lookup tables, etc., and input / output (I / O) ports 58. Computer-readable medium or memory 56 includes any type of media capable of being accessed by a computer, such as read-only memory (ROM), random access memory (RAM), hard disk drive, optical disc (CD), digital video disc (DVD), or any other type of memory. “Non-transitory” computer-readable memory 56 does not include wired, wireless, optical, or other communication links for transmitting transient electrical or other signals. Non-transitory computer-readable memory 56 includes media where data can be permanently stored and media where data can be stored and subsequently rewritten, such as rewritable optical discs or erasable memory devices. Computer code includes any type of program code, including source code, object code, and executable code. Processor 54 is configured to execute code or instructions. Vehicle 12 may have a controller 14, which may include a dedicated Wi-Fi controller or an engine control module, a transmission control module, a body control module, an infotainment control module, etc. Without departing from the scope or intent of this disclosure, I / O port 58 may be configured to communicate via wired communication, wireless communication via Wi-Fi protocols under IEEE 802.11x, etc.

[0039] Controller 14 also includes one or more applications 60. An application 60 is a software program configured to perform a specific function or set of functions. Application 60 may include one or more computer programs, software components, instruction sets, processes, functions, objects, classes, instances, associated data, or portions thereof suitable for implementation in appropriate computer-readable program code. Application 60 may be stored in memory 56 or in additional or separate memory. Examples of applications 60 include audio or video streaming services, games, browsers, social media, etc. In other examples, application 60 is used to manage body control system functions, suspension control system functions, aerodynamic control system functions, etc., in the exemplary vehicle 12.

[0040] Referring now to Figures 2A, 2B, and 3 and continuing to refer to Figure 1, system 10 utilizes one or more applications 60 stored in memory 56, particularly MPLC application 62, to manage and coordinate the chassis and drivetrain actuators 26 of vehicle 12 in order to optimally maintain the position of lane 100 on road surface 102.

[0041] Specifically, the MPLC application 62 includes multiple subroutines performing different but interrelated functions that enable the vehicle 12 to operate predictably in automatic or semi-automatic mode while maintaining the position of lane 100. More specifically, the MPLC application 62 performs model predictive control (MPC) for lane 100 centered on reference switching, and arbitrates between the physical static reference path 104 and the dynamic reference path 106 for automated or on-demand lane 100 changes. In several ways, the dynamic reference path 106 is essentially a smooth path between the current static reference path 104 and the future static reference path 104'. To maintain a smooth transition between lanes 100, the weights 108 within the MPC calculation can be or may be temporarily modified such that the input weights penalize rapid changes in the actuator 26 position, such as rapid changes in the steering system 24 position or steering system 24 angle, and the steering system 24 angle and rate are limited by predefined input constraints. That is, the MPC weights depend on the control objective. The weights also change when the reference changes. More specifically, when the reference changes, some weights decrease and then gradually rise to their nominal values ​​over time, thus smoothing the transition. This depends on whether the reference is static (reference path 104) or dynamic (reference path 106) changing to the three different inputs to the MPC: the initial lateral error (e... y ), initial heading error (e ψ ) and reference curvature waypoints In an exemplary, non-limiting embodiment, MPC computation can be visualized using the following equation: Among them: ·V x or V xr Define the vehicle's longitudinal velocity; ·V y The vehicle's lateral velocity is 12; ·ω z Or ψ represents the yaw rate of vehicle 12; ·e y The lateral error (distance) is 104 units from the target path; ·e ψ The heading error is 104 units from the target path; ·ψ r The yaw rate reference for target path 104 (related to the curvature of target path 104); ·m is the mass of vehicle 12; ·I zz Let l be the moment of inertia of vehicle 12; f l r C is the distance from the center of gravity of vehicle 12 to the front axle 42 and rear axle 40 of vehicle 12; f C r The lateral stiffness of the front and rear tires is 43, respectively; ·δ f Forward steering angle, δ r For the rear steering angle (if equipped, both are controlled by controller 22); A yd and Rzd These are disturbances of lateral acceleration (force) and yaw acceleration (torque), respectively; ·x, x d These are the state vector and the perturbation / uncertainty state vector, respectively; and the hat "^" on top of any of the above variables indicates that the hatted variable is an estimate.

[0042] MPLC application 62 uses model predictive control (MPC) to track the trajectory of vehicle 12 based on static reference path 104. Static reference path 104 can be described as a target path, which is a fixed path in space on road surface 102. In several examples, static reference path 104 may be lane center 112; however, it should be understood that in some other non-limiting examples, static reference path 104 may be a path parallel to lane center 112 but not actually lane center 112. For example, in some cases, when adjacent vehicle 12' is in or near lane line 114 near the main vehicle 12, static reference path 104 may be automatically adjusted by MPLC application 62 to remain within the current lane 100, but to provide an additional gap or distance between the main vehicle 12 and adjacent vehicle 12', thus making static reference path 104 parallel to but not identical to lane center 112. That is, static reference path 104 may be offset 116 from lane center 112. In several aspects, when the ADAS system 28 of vehicle 12 is only tasked with centering lane 100, relevant references are generated based on the perceived lane center 112 and the expected offset 116 from the desired center. Conversely, when the goal of ADAS system 28 is lane changing, evasive maneuvers, etc., relevant references are generated to create a path from the current position of vehicle 12 to the target lane 101 with a predetermined look-ahead distance.

[0043] Turning more specifically to Figure 3 and continuing to refer to Figures 1 and 2, the lane 100 changing process using the MPLC application 62 is shown in more detail as a series of logical steps in flowchart form. Input information 200 from various sensors 18 and actuators 26 of the vehicle 12 is received by the MPLC application 62. In several aspects, the input information 200 includes measurements in the X and / or Y directions of the vehicle 12's travel, including but not limited to: the lateral error e of the vehicle 12. y Heading error e ψ and yaw rate ω z Or ψ. Input information 200 is fed forward to Kalman filter 202 and reference anomaly detection subroutine 204.

[0044] Kalman filter 202 at least estimates the acceleration disturbance acting on vehicle 12. Lateral force disturbance and yaw moment disturbance And the yaw rate ω measured during destruction zOr the measurement deviation of ψ. It should be understood that the yaw rate may include the measurement deviation adapted by the MPLC application 62. As will be described below, the estimated disturbance is applied along the prediction range in the MPC, thereby compensating for the bias. By applying the estimated disturbance along the prediction range of the MPC, the estimated disturbance is compensated, and thus the effect of the disturbance on the vehicle's dynamics is reduced. Furthermore, the yaw rate ω adjusted for bias is used in the initial state of the MPC. z Or ψ, thus allowing the MPLC application 62 to compensate for any deviations via a Kalman filter. In several non-limiting examples, the Kalman filter 202 is based on the vehicle 12 speed, the vehicle 12's forward and / or rearward steering angle δ f、 δ r and yaw rate ω z Or ψ, and the motion estimation of vehicle 12 relative to the measured static reference path 104 (e.g., lane center 112) to determine the state of vehicle 12.

[0045] Anomaly detection subroutine 204 detects anomalies in the MPC reference trajectory or path 104. More specifically, anomalies are detected by comparing trajectory offset 116, vehicle 12 heading, and vehicle 12 current path curvature with corresponding moving averages. Anomalies are detected when one or more differences between vehicle 12 current path curvature and vehicle 12 heading and the reference trajectory or path 104 exceed predetermined and / or variable thresholds. For example, when the vehicle 12 operator or ADAS 28, which controls the vehicle 12 path, requests a lane 100 change, semantic data from the trajectory planner portion of the MPLC application 62 is used to disable anomaly detection. That is, anomaly detection is disabled, allowing ADAS 28 and MPLC application 62 to correctly and effectively execute lane 100 changes when such manipulation is required.

[0046] The outputs of Kalman filter 202 and reference anomaly detection subroutine 204 are then passed to MPC subroutine 206. More specifically, the perturbation... This is reported to MPC subroutine 206. It should be understood that the actuators 26 of vehicle 12 (such as those described above) are constrained to operate within predictable parameters and for one or more specific purposes. For example, the position of steering system 24 is limited by mechanical and / or electronic constraints of a given steering rack, shafts 42, 40, and wheels 30. Therefore, the actuators 26 of steering system 24 are constrained by one or more of physical and / or electronic constraints to prevent the steering angle δ from being adjusted. f δ rExceeding the physical and / or electronic limitations of the steering system 24. Without departing from the scope or intent of this disclosure, such constraints may include hardware and software-imposed limitations, such as physical stop and physical rate limits, and electronically or software-imposed articulation limits or articulation rate limits. In a further aspect, constraints may be adaptable and dependent on any of a variety of different factors, including but not limited to vehicle 12 position, speed, attitude, tilt, etc. It should also be understood that the exemplary steering system 24 actuator 26 is intended only as a non-limiting example of the type of constraint that may be applied to the actuator 26 of the vehicle 12. Therefore, the steering angle δ f δ r Minimum and maximum values ​​δ f,min δ f,max δ r,min δ r,max and the minimum and maximum steering angular rates Actuator 26 constraints are defined and applied to the output of reference anomaly detection subroutine 204 before being received by MPC subroutine 206. The actuator 26 constraints have a variety of effects, but in particular, they have the practical advantage of limiting the actuator 26 output to a practical, feasible output that can be correctly and appropriately implemented by the vehicle 12 system.

[0047] Within MPC subroutine 206, MPC constraints are adjusted in response to detected anomalies. After an anomaly is detected, output commands are generated in 208 (e.g., steering system angle δ). fThe amplitude and rate limits of actuator 26, previously used by MPC subroutine 206, are reduced from a first level to a second level less than the first level for a predetermined time period. The amplitude limit of actuator 26 is adjusted to limit lateral acceleration within the range of predicted lateral acceleration values ​​based on the curvature of the last non-abnormal target reference path 104 or dynamic reference path 106 trajectory. The rate limit of actuator 26 is then progressively reduced to a calibrable fraction of the nominal limit of actuator 26. Both amplitude and rate limits are held at the reduced values ​​for a predetermined and / or variable time period before linearly increasing back to their respective nominal values. The predetermined and / or variable time period is a recovery period, empirically adjusted to be long enough to mitigate the event and allow sufficient time for the vehicle 12 operator to respond, while also being short enough not to cause vehicle 12 to drift out of lane 100 during a brief event. In some non-limiting examples, there are two times for maintaining the amplitude and rate limits or weights and constraints at the adjusted values, followed by the time for each ramp back to the nominal values. For each time the amplitude and rate limits are maintained at the adjusted values, the range of the predetermined and / or variable time periods can be between about two (2) and three (3) seconds, totaling about four (4) to six (6) seconds. In another practical example, the absolute minimum and absolute maximum durations of the predetermined and / or variable time periods extend from about 0.5 seconds to about five (5) seconds, or in total from about one (1) second to about ten (10) seconds. The two times for which the amplitude and rate limits or weights and constraints are maintained at the adjusted values ​​are taken into account. The output command 208 generated by the MPC subroutine 206 causes the vehicle 12 to actively and adaptively adjust the position and rate of the individual actuators 26 to allow the vehicle 12 to selectively and accurately follow the target reference path 104 and / or the dynamic reference path 106, including any applicable offsets 116.

[0048] Turning now to Figures 4 and 5 and continuing to refer to Figures 1 through 3, in some examples, the MPLC application 62 can be used for zero-offset control. The Kalman filter 202 can provide zero-lateral-offset control on a case-by-case basis in the presence of simultaneous force and moment disturbances and / or simultaneous heading and / or curvature errors. To implement zero-offset control using the Kalman filter 202, the MPLC application 62 utilizes measurements, such as: and status information, Generate initial state: It should be at least related to lateral and yaw moment disturbances. When combining and utilizing MPC calculations for processing within MPC subroutine 206: Generate command 208, such as the steering angle command δ fThis allows vehicle 12 to maintain a zero-offset position during operation. In another example, upon receiving one or more of an occupant command or ADAS 28 command to change from the current lane 100 to the adjacent lane 101, the MPLC application 62, which performs lane 100 centering during steady-state operation, switches from static reference 104 tracking to dynamic reference 106 tracking. Dynamic reference 106 tracking allows vehicle 12 to smoothly change its route to transition from the current lane 100 to the adjacent lane 101. The MPLC application 62 generates output commands to various actuators 26 of vehicle 12, including initiating a deviation from static reference 104 tracking to dynamic reference 106 tracking and subsequently blending the dynamic reference 106 tracking with a steering system 24 angle δ to track a second static reference 104' in the adjacent lane 101. f Order.

[0049] Turning now to Figure 5 and continuing to refer to Figures 1 through 4, a visual representation of the MPLC application 62 performing anomaly mitigation relative to reference trajectories 104, 106 is shown schematically with additional detail. In summary, the MPLC application 62 calculates the external force F currently applied to or exerted by vehicle 12. xd F yd and torque M zd In some examples, as shown in Figure 4, crosswind force 300 can be applied unevenly to the outer surface of vehicle 12, and trailer moment 301 or torque force 302 can be applied unevenly longitudinally and laterally, and as torque to vehicle 12. Model mismatch 304 may result in a force F that differs between expected and actual values. xd F yd and torque M zd Mismatches can also occur with road surface defects such as inclination angle 306 and slope 308. Each such externally applied force F... xd F yd Or torque M zdThe force and / or torque equivalent 310 is modeled to generate a force and / or torque equivalent, which is then used by the MPLC application 62 to maintain the position of lane 100 in a zero-offset manner. In some examples, the reference trajectories 104, 106 may be discontinuous or may “jump” according to road surface markings 102 (e.g., lane lines 114 indicating entrance or exit ramps, turning lanes, etc.). In response to the detection of such abnormal road markings, such as jumps in the reference heading, using multiple onboard sensors 18, the MPLC application 62 tightens the amplitude and rate limits of the steering system 24 in response to the detected jumps in the reference heading. After a predetermined and / or variable amount of time, the amplitude and rate limits of the steering system 24 ramp back to their nominal values. The tightening of the amplitude and rate limits the peak lateral acceleration and jerks that the occupants of vehicle 12 may experience in response to trajectory anomalies. Thus, by attenuating or otherwise remedying peak lateral acceleration and jerks, the comfort of the vehicle occupants and their confidence in the functionality of the ADAS system 28 are ensured.

[0050] The Model Predictive Lane Centering Control (MPLC) system 10 disclosed herein offers several advantages. These advantages include, but are not limited to: the ability to automatically and / or semi-automatically allow the vehicle to smoothly navigate lane change requests with high accuracy, robustness, and redundancy, utilizing pre-existing infrastructure, while maintaining or reducing computational resource utilization and minimizing the impact of disturbances on the smoothness of vehicle operation. Furthermore, the MPC-based control of system 10 reduces calibration work, as the vehicle dynamics 12 are inherent to controller 22. Actuator constraints prevent actuator 26 from being commanded to perform actions beyond its capabilities, while reference switching provides complete flexibility to adjust the control response for automatic lane change 100 transitions, whereas smooth adjustment of lane change 100 transitions might be difficult or impractical using only static reference 104. Using dynamic reference 106 for lane changes provides the flexibility to achieve improved lane change performance, whereas traditionally there has been insufficient flexibility to suppress the initialization of lane change manipulation without overly numbing the end of the lane change. Finally, zero-offset tracking is ensured even in the presence of unknown disturbances, biases, or model mismatch 304 by utilizing Kalman filter 202.

[0051] The description in this disclosure is merely exemplary in nature, and variations thereof without departing from the spirit and scope of this disclosure are intended to fall within its scope. Such variations should not be considered as departing from the spirit and scope of this disclosure.

Claims

1. A Model Predictive Lane Centering (MPLC) control system with reference switching and disturbance suppression, the system comprising: vehicle; One or more sensors are mounted on the vehicle, the one or more sensors measuring real-time static and dynamic data about the vehicle; One or more actuators are disposed on the vehicle, which alter the static and dynamic characteristics of the vehicle; a controller has a processor, memory, and input / output (I / O) ports communicating with the one or more sensors and the one or more actuators; the processor executes a portion of program code stored in the memory, the portion of program code including an MPLC application, the MPLC application including: first control logic for selectively tracking static and dynamic references using model predictive control (MPC); and second control logic for adjusting MPC weights and constraints as the system switches between static and dynamic references. The system includes: a third control logic for tracking the vehicle's trajectory; a fourth control logic for detecting anomalies in the MPC reference trajectory; and a fifth control logic for adjusting actuator constraints in response to detected anomalies. The MPLC control system actively and continuously adjusts actuator output commands to selectively cause the vehicle to track and follow the current lane center, and upon receiving a lane change command, smoothly performs a lane change by instructing one or more actuators to change position and smoothly alter the vehicle's trajectory to enter an adjacent lane before returning to tracking and following the center of the adjacent lane.

2. The system according to claim 1, wherein, The first control logic further includes: control logic for obtaining data from the one or more sensors and the one or more actuators; control logic for performing static reference tracking, wherein the static reference tracking causes the vehicle to follow and conform to the current lane center; and control logic for causing the vehicle to deviate from the current lane center while following and tracking the current lane center.

3. The system according to claim 1, wherein, When the lane change command is received from one or more of the vehicle occupants and advanced driver assistance systems (ADAS), the MPLC application switches to tracking the dynamic reference, which is a smooth path extending between the current static reference and the future static reference.

4. The system according to claim 1, wherein, The second control logic further includes: control logic for adaptively adjusting MPC weights based on control objectives, wherein the weights also change when the reference changes; and control logic for adaptively adjusting the weights and constraints upon receiving the lane change command and subsequently ramping the weights and constraints back to nominal values ​​over a predetermined time period, thereby smoothing the transition between tracking static and dynamic references.

5. The system according to claim 4, further comprising: Control logic that penalizes rapid changes in actuator position; Depending on whether the reference is static or dynamic, the three different inputs to the MPC are changed according to the following formula: the initial lateral error (e y ), initial heading error (e ψ and reference curvature waypoints Control logic: Where: V x or V xr Define the longitudinal velocity of the vehicle; V y ω represents the lateral velocity of the vehicle. z Or ψ represents the yaw rate of the vehicle; e y e represents the lateral error (distance) from the target path. ψ ψ represents the heading error from the target path. r The yaw rate is a reference value for the target path (related to the curvature of the target path); m is the mass of the vehicle; I zz Let l be the moment of inertia of the vehicle; f l r C is the distance from the vehicle's center of gravity to the front and rear axles of the vehicle. f C r These are the front and rear tire lateral stiffness, respectively; δ f Forward steering angle, δ r Rear steering angle; A yd and R zd These are disturbances of lateral acceleration (force) and yaw acceleration (torque), respectively; x, x d These are the state vector and the perturbation / uncertainty state vector, respectively; and the hat "^" on top of any variable indicates that the hatted variable is an estimate.

6. The system according to claim 1, wherein, The third control logic further includes: control logic for compensating for the estimated lateral force disturbance and the estimated yaw moment disturbance by applying each of the estimated lateral force disturbance and the estimated yaw moment disturbance along the prediction range in the MPC; and control logic for compensating for the estimated measurement deviation by utilizing the yaw rate adjusted by the deviation in the initial state of the MPC.

7. The system according to claim 6, wherein, The Kalman filter estimation of lateral force disturbances and estimated yaw moment disturbances uses measured vehicle speed and vehicle motion relative to a measured static reference, wherein the measured static reference defines a static reference path.

8. The system according to claim 1, wherein, The fourth control logic further includes: control logic for detecting anomalies in the MPC reference trajectory by comparing trajectory offset, heading, and curvature with a moving average of the trajectory offset, heading, and curvature; and control logic for detecting anomalies when the difference between one or more of the trajectory offset, heading, and curvature and the moving average of the trajectory offset, heading, and curvature exceeds a predetermined threshold.

9. The system according to claim 8, further comprising: In predefined cases, including when a lane change has already been requested, control logic that uses semantics from the trajectory planner to disable anomaly detection is employed.

10. The system according to claim 1, wherein, The fifth control logic further includes: control logic for adjusting actuator constraints by reducing actuator amplitude and rate limits from a first level to a second level less than the first level for a predetermined time period when an anomaly is detected, wherein the actuator amplitude limit is adjusted to limit the vehicle lateral acceleration to within the range of the predicted lateral acceleration based on the curvature of the last anomaly-free reference trajectory; and control logic for gradually reducing the actuator rate limit to a calibrable fraction of the nominal actuator rate limit during the predetermined time period; and wherein the amplitude and rate limits are temporarily held at the reduced values ​​before linearly increasing back to the nominal values.

Images