OPTIMIZATION OF VEHICLE POWER TO SUPPORT VEHICLE CONTROL

Abstract

System (60) for interacting with a vehicle user, comprising: an optimization module (70) configured to determine a control strategy for traversing a selected route based on a probabilistic prediction of actuator inputs; a monitoring module configured to monitor vehicle movement and vehicle user inputs during the operation of a vehicle (10) over the selected route; a prediction module (72) configured to predict an optimal local path for traversing a section of the selected route and to predict a set of optimal actuator control actions; an interface module configured to present a suggested set of actuator commands to the vehicle user based on the set of optimal actuator control actions; a control system configured to control the vehicle (10) to execute the proposed set of actuator commands, based on the vehicle user's acceptance of the proposed set of actuator commands; characterized by: a data fusion module (64) configured to select a detection method for detecting each of several emergency events, wherein the detection method specifies a combination of one or more types of detection data, wherein the detection method is selected based on a probability that an emergency event will be detected by the selected detection method and a reliability of the selected detection method.

Description

[0001] This disclosure relates to the field of vehicle control. In particular, this disclosure relates to systems and methods for optimizing the user control of a vehicle.

[0002] Vehicles are increasingly equipped with sensors and perception devices that enhance the cognitive abilities of vehicle control systems and drivers, thereby enabling autonomous control and / or driver assistance. In many cases, such as high-performance driving on a racetrack, drivers are consistently seeking ways to improve overall performance. Given the many variables that can affect performance, as well as numerous potential emergency events that can occur, it is desirable to create systems that assist drivers in improving their overall performance.

[0003] DE 10 2023 100 983 A1 discloses a system according to the preamble of claim 1. DE 10 2021 110 868 A1 describes a method and system for automatic vehicle control.

[0004] One of the problems underlying the invention is to provide an improved system for interaction with a vehicle user and a corresponding method.

[0005] This task is solved by the subject matter of the independent claims.

[0006] The invention relates to a system for interacting with a vehicle user, comprising: an optimization module configured to determine a control strategy for traversing a selected route based on a probabilistic prediction of actuator inputs, a monitoring module configured to monitor vehicle movement and vehicle user inputs during the operation of a vehicle over the selected route, and a prediction module configured to predict an optimal local path for traversing a section of a selected route and to predict a set of optimal actuator control actions.The system also includes an interface module configured to present a suggested set of actuator commands to the vehicle user based on the set of optimal actuator control actions, and a control system configured to control the vehicle to execute the suggested set of actuator commands, based on the vehicle user's acceptance of the suggested set of actuator commands. The system is characterized by a data fusion module configured to select a detection method for detecting each of several emergency events, wherein the detection method specifies a combination of one or more types of detection data, and wherein the detection method is selected based on the probability that an emergency event will be detected by the selected detection method and the reliability of the selected detection method.

[0007] In addition to one or more of the features described here, the control strategy is determined based on a driver model and a vehicle model and prescribes actuator input values ​​over time that maximize the vehicle's performance while driving along the selected route.

[0008] In addition to one or more of the features described here, the control strategy is determined based on a substitute model constructed from the driver model.

[0009] In addition to one or more of the features described here, the prediction module is configured to probabilistically predict the optimal local path over several successive time horizons based on the control strategy and a probability of encountering an emergency event during each successive time horizon.

[0010] In addition to one or more of the features described here, the selected route is a pre-planned route along a road with known parameters, and the control strategy is determined to optimize actuator inputs to minimize the amount of time required to travel the pre-planned route.

[0011] In another exemplary embodiment, a method for cooperating with a vehicle user comprises determining a control strategy for traversing a selected route based on a probabilistic prediction of actuator inputs by an optimization module, monitoring the vehicle movement and vehicle user inputs during the operation of a vehicle over the selected route, and predicting an optimal local path for traversing a section of the selected route and predicting a set of optimal actuator control actions by a prediction module.The method also includes presenting a proposed set of actuator commands to the vehicle user based on the set of optimal actuator control actions and controlling the vehicle to execute the proposed set of actuator commands, based on the vehicle user's acceptance of the proposed set of actuator commands. The method is further characterized by selecting a detection method for detecting each of several emergency events, wherein the detection method specifies a combination of one or more types of detection data, and wherein the detection method is selected based on the probability of an emergency event being detected by the selected detection method and the reliability of the selected detection method.

[0012] In addition to one or more of the features described here, the control strategy is determined based on a driver model and a vehicle model and prescribes actuator input values ​​over time that maximize the vehicle's performance while driving along the selected route.

[0013] In addition to one or more of the features described here, the control strategy is determined based on a substitute model constructed from the driver model.

[0014] In addition to one or more of the features described here, optimal local path prediction includes probabilistic prediction of the optimal local path over several successive time horizons based on the control strategy and a probability of encountering an emergency event over each successive time horizon.

[0015] In addition to one or more of the features described here, the selected route is a pre-planned route along a road with known parameters, and the control strategy is determined to optimize actuator inputs to minimize the amount of time required to travel the pre-planned route.

[0016] In yet another exemplary embodiment, a vehicle system comprises a working memory with computer-readable instructions and a processing device for executing the computer-readable instructions.The computer-readable instructions control the processing device to determine a control strategy for traversing a selected route based on a probabilistic prediction of actuator inputs, to monitor vehicle movement and vehicle user inputs during the operation of the vehicle over the selected route, to predict an optimal local path for traversing a section of the selected route, to predict a set of optimal actuator control actions, to present a proposed set of actuator commands to the vehicle user based on the set of optimal actuator control actions, and to control the vehicle to execute the proposed set of actuator commands based on the vehicle user's acceptance of the proposed set of actuator commands.The vehicle system is further characterized by a data fusion module configured to select a detection method for detecting each of several emergency events, wherein the detection method specifies a combination of one or more types of detection data, and wherein the detection method is selected based on a probability that an emergency event will be detected by the selected detection method and a reliability of the selected detection method.

[0017] In addition to one or more of the features described here, the control strategy is determined based on a driver model and a vehicle model and prescribes actuator input values ​​over time that maximize vehicle performance while driving along the selected route.

[0018] In addition to one or more of the features described here, the control strategy is determined based on a substitute model constructed from a driver model.

[0019] In addition to one or more of the features described here, optimal local path prediction includes probabilistic prediction of the optimal local path over several successive time horizons based on the control strategy and a probability of encountering an emergency event over each successive time horizon.

[0020] The above features and advantages, and other features and advantages of the disclosure, are readily apparent from the following detailed description in conjunction with the accompanying drawings.

[0021] Other features, advantages, and details are shown only by way of example in the following detailed description, which refers to the drawings; they show: Fig. 1 a top view of a motor vehicle showing aspects of user interaction and a prediction system according to an exemplary embodiment; Fig. 2 a prediction and user interaction system according to an exemplary embodiment; Fig. 3 a flowchart illustrating aspects of a prediction, control and user interaction procedure according to an exemplary embodiment; Fig. 4 a data fusion process used to determine detection methods for emergency events, according to an exemplary embodiment; Fig. 5 an example of a decision tree according to an exemplary embodiment; Fig. 6 Aspects of a simplified vehicle model according to an exemplary embodiment; and Fig. 7 a computer system according to an exemplary embodiment.

[0022] The following description is by its very nature merely exemplary and is not intended to limit the present disclosure, its application, or uses. It should be noted that in all drawings, corresponding reference numerals indicate identical or corresponding parts and features.

[0023] According to one or more exemplary embodiments, methods and systems for supporting collaborative vehicle operation are provided. One embodiment of a system is configured to interact or cooperate with a driver by monitoring vehicle operation while traversing a route and determining an optimal set of controls (actuator inputs) predicted to optimize the vehicle's performance while traversing the route. The optimal set of controls is presented to the driver as a suggestion for improving performance. The driver has the option to accept the suggestion either by directly inputting actuator commands (e.g., controlling the steering and throttle) or by providing a signal to the system.

[0024] The embodiments described here offer a number of advantages. For example, they enhance vehicle responsiveness and improve driver performance. They provide drivers with improved real-time information, increasing their understanding of the vehicle's surroundings and creating opportunities for improved performance (e.g., faster lap times). These embodiments also enable predictive control for rapid and controlled responses to potential emergencies. The integration (data fusion) of various sources can be used to predict and assess the likelihood of emergencies, thereby improving the ability to respond to such situations.

[0025] For example, when driving on a racetrack, the driver may not be certain which path to follow (e.g., when entering a curve) and / or what speed to choose. These designs provide the driver with information that allows them to maximize lap times. Furthermore, they assist the driver in reacting to events, thus maximizing performance. Such information can be presented to the driver quickly and intuitively, enabling faster decision-making.

[0026] Fig. Figure 1 shows an embodiment of a motor vehicle 10 comprising a vehicle body 12 that at least partially defines a passenger compartment 14. The vehicle body 12 also supports various vehicle subsystems, including a drive system 16, and other subcomponents to support functions of the drive system 16, and other vehicle components, such as a brake subsystem, a suspension system, a steering subsystem, and, if the vehicle is a hybrid electric vehicle, a fuel injection subsystem, an exhaust subsystem, and others.

[0027] The vehicle 10 can be a combustion engine vehicle, an electric vehicle (EV), or a hybrid vehicle. In one embodiment, the vehicle 10 is a hybrid vehicle comprising a combustion engine system 18 and at least one electric motor arrangement. In another embodiment, the drive system 16 comprises an electric motor 20 and can include one or more additional motors positioned at different locations, such as a motor 32. The vehicle 10 can also be a fully electric vehicle with one or more electric motors.

[0028] The vehicle 10 also includes various control devices for controlling aspects of the vehicle's operation. Such devices include, for example, the power unit 18 and the motors 20 and 32, a steering wheel 34, an accelerator pedal 36, front brakes 38 connected to the front wheels 26, and rear brakes 40 connected to the rear wheels 30. Other actuators may include a front output actuator 42 and / or a rear output actuator 44 (e.g., if the vehicle 10 is a high-performance vehicle). The control devices and actuators are controllable via one or more control units, collectively represented by a controller 46. Various sensors are included for monitoring the vehicle's motion and operation, such as torque sensors, pedal position sensors, brake sensors, an inertial measurement unit (IMU), and others.

[0029] The vehicle also includes an environmental monitoring system for detecting and monitoring the area around the vehicle. This monitoring system includes, for example, one or more optical cameras 50 configured to capture images, which may be still images and / or video images. Additional devices or sensors may be included, such as one or more radar arrays 52, which are contained within the vehicle 10. The monitoring system is not limited to this and may include other types of sensors, such as lidar and infrared.

[0030] The vehicle 10, the monitoring system, the controller 46, and other vehicle systems comprise or are connected to an on-board computer system 54, which includes one or more processing devices 56 and a user interface 58. The user interface 58 may include a touchscreen, a voice recognition system, and / or various buttons to enable a user to interact with features of the vehicle. The user interface 58 may be configured to interact with a user or driver via visual communications (e.g., text and / or graphic displays), tactile communications or alarms (e.g., vibration), and / or audible communications. The on-board computer system 54 may also include or communicate with devices for monitoring the user or driver, such as interior cameras and image analysis components. Such devices may be integrated into a driver monitoring system (DMS).

[0031] The user interface 58 can include various types of displays and / or other devices that interact with a user and / or convey information to the user. For example, the vehicle 10 can include a display screen (e.g., console screen, full-view mirror, or FDM, etc.). In one embodiment, the vehicle 10 includes one or more head-up displays (HUDs). Other devices that may be included include indicator lights, haptic devices, interior lights, acoustic communication devices, and others. Haptic devices (tactile interfaces) include, for example, vibration devices in the vehicle's steering wheel and / or vehicle seat. The various displays, haptic devices, lights, and acoustic devices are configured to be used in different combinations to present information and recommendations to a user.

[0032] Fig. Figure 2 represents an embodiment of a prediction and user interaction system 60 configured to interact with a user or driver during vehicle operation to optimize or enhance performance. All or part of the system 60 may be integrated into the vehicle computer system 54, the controller 46, and / or any other suitable processing devices or control systems.

[0033] System 60 includes a perception and estimation module 62 configured to collect detection data from the environmental monitoring system. Examples of collected data include camera images, radar / LiDAR images, object and feature detection using image processing and / or machine learning, and others.

[0034] A data fusion module 64 is configured to evaluate the collected data and determine detection methods for detecting emergency conditions or events. In one embodiment, the data fusion module 64 includes a reliability assessment module 66 for determining the reliability of various detection modalities (e.g., cameras, radar, weather data, global positioning system (GPS) data, etc.). A probability calculation module 68 may be included for calculating probabilities of emergency events, as discussed further here.

[0035] An “emergency event” is an event, feature, object, or condition that could arise while the vehicle is traveling along a given route and that requires some real-time action or deviation to respond to. Examples of emergency events include objects such as vehicles or pedestrians in the vehicle's path, changes in road conditions, weather conditions, and others.

[0036] A global optimization module 70 is configured to generate an optimal control strategy for traversing a planned route, and a prediction module 72 is used to predict optimal actuator inputs and vehicle controls for responding to emergency conditions. The prediction module can predict an optimal local path for a segment of the planned route (block 74), brake control predictions (block 76), downforce command predictions (78), and / or other predictions for inputs or commands to actuators for optimal response to an emergency event.

[0037] A control device, such as the controller 46, is configured to operate one or more actuators based on user input. The one or more actuators can be operated based on direct user input, such as steering or depressing an accelerator pedal. Alternatively or additionally, the one or more actuators can be operated by the controller 46 in response to user interaction with a user interface 80. The controller 46 can, for example, operate the actuators 42 and 44 ( Fig. 1) for output control (e.g. via an output control module 47) and / or limited-slip differential control (LSD control) (e.g. via an LSD control module 49).

[0038] Fig. Section 3 presents a procedure 100, aspects of which can be performed offline (e.g., global path optimization) or online (i.e., during vehicle operation). Procedure 100 is discussed in conjunction with Blocks 101-106. Procedure 100 is not limited to the number or sequence of steps it contains, as some steps represented by Blocks 101-106 may be performed in a different order than described below, or fewer steps may be performed. For example, if a selected route is not a pre-planned route along a roadway with known parameters (e.g., a racetrack), global optimization, as discussed here, may be omitted.

[0039] Procedure 100 is used in conjunction with the vehicle of Fig. 1 and a processing system, which may be, for example, the computer system 54, the controller 46, or a combination thereof. The method 100 is also discussed with reference to the prediction and user interaction system 60 of Fig. 2. Aspects of the method 100 are discussed in connection with the vehicle 10 and the prediction and user interaction system 60 for explanatory purposes. It is noted that the method 100 is not so limited and can be carried out by any suitable processing device or processing system or combination of processing devices.

[0040] In block 101, the perception and estimation module 62 collects detection data. Detection data includes, for example, camera images, radar data, lidar data, GPS data, map data, temperature and / or any other suitable type of detection data.

[0041] In block 102, the data fusion module combines 64 different types of detection data and uses the combined data to determine detection procedures for various emergency events.

[0042] In general, different sources of detection data and different algorithms may be suitable for different events and conditions. For example, object detection may be most suitable using imagery and / or radar data, whereas weather condition detection may be most suitable using GPS or location data (optionally with imagery). Furthermore, conditions can affect the reliability of different modalities. Imagery data, for instance, is more reliable under sunny weather conditions. GPS data can also exhibit varying reliability, as location and / or weather conditions can affect signal reliability.

[0043] In one embodiment, the data fusion module 64 uses a methodology that considers the probability P of an emergency event occurring and the reliability R of different types of detection data when determining a suitable emergency detection algorithm and emergency detection procedure. Probabilities P can be determined by the probability calculation module 68, and reliabilities R can be determined by the reliability evaluation module 66.

[0044] Fig. Figure 4 presents an example of the methodology. In this example, the data fusion module 64 considers a number N of potential emergency events (events E1 - E). N ) and their associated probabilities (P1 - P N ).

[0045] In this example, "Method A" refers to the detection of an event using camera data C, and "Method B" refers to the detection of an event using lidar data L. Method A and / or Method B may include one or more additional types of detection data, such as map data M. In this example, P1 refers to the probability that an obstacle or object enters the road (event E1), and P2 refers to the probability that the vehicle encounters a patch of ice or a slippery condition (event E2).

[0046] For example, for event E1, P1 is calculated as follows: P1=(R1AcP1A+R1BcP1B)R1Ac+R1Bc, where c is a constant, P 1A the probability that event E1 is detected using method A, and P 1BThe probability that event E1 is detected using method B is R. 1A is the reliability of the data collected using method A (determined based on weather conditions, visibility, etc.), and R 1B is the reliability of the data collected using method B.

[0047] If P1 is sufficiently high, the emergency detection method for detecting event E1 is a combination of Method A and Method B (camera and lidar data). If P1 is not sufficiently high (i.e., it does not meet or exceeds a selected threshold), then Method A or Method B is selected based on the reliability R. 1A or R 1B is higher. Detection methods are also used for the remaining events E2 to E. Ndetermined. It is noted that real-time data can be collected and used during operation to update probability and reliability assessments as necessary.

[0048] With renewed reference to Fig. In block 103, the global optimization module 70 determines an optimal steering strategy for traversing a selected route. The route can be a pre-planned route (e.g., a planned route that specifies particular roads to be traveled, or a racetrack, or another included lane).

[0049] In the following description, the selected route is a pre-planned route, and a global optimization is performed to determine the vehicle dynamics and actuator inputs that result in an optimal vehicle path and actuator inputs. The optimization can be performed offline (e.g., before the vehicle is operated) or online during operation.

[0050] The global optimization module 70 is not limited to pre-planned or pre-selected routes. For example, if the vehicle is not following a pre-planned route, the optimization strategy can be performed in real time over relatively short time horizons to optimize performance locally.

[0051] In one embodiment, if the expected road characteristics (e.g. direction, cross slope, road surface) are known and the route is known (e.g. the route is a lap of a race track), a global optimization is performed to optimize vehicle operation over the route or a lap (lap time optimization).

[0052] Lap time optimization can be achieved by determining optimal vehicle dynamics and actuator controls for successive track segments or time windows. Optimal pedal and steering inputs (and other actuator commands such as downforce controls) are determined, for example, for each segment of a track or other known route. Examples of segments include straights and various curved sections (each of which may have its own parameters such as path, turning radius, crossfall, road gradient, etc.).

[0053] In one embodiment, the optimization is performed based on a driver model and a vehicle model with high fidelity. This optimization determines optimal inputs for a route with known profiles and attributes, including path design, cross slope and gradient characteristics, road boundaries, and others.

[0054] The driver model simulates actuator inputs from a human driver and can be represented by the following: f(x)=αx2+βx+γ, where x represents a vehicle position, vehicle speed, actuator state, or other variable related to position and / or dynamics, and α, β, γ are coefficients. For example, for brake speed control, the brake input A x,brake represented by: Ax,brake=αVx2+βVx+γ, where V x The braking speed is subject to maximum and minimum limitations. Any number of functions can exist for each actuator input. Furthermore, the function for a given input can be any suitable polynomial.

[0055] The vehicle model simulates the vehicle's response to driver inputs or actuator commands. Specifically, it simulates the vehicle's dynamic response (e.g., changes in speed, direction, yaw, etc.) to various inputs. In one example, the vehicle model includes elements representing the vehicle body, wheels, tires, drivetrain, suspension, brakes, steering, and other components.

[0056] The functions and coefficients for each actuator input are learned, for example, by a selected machine learning technique such as deep learning or reinforcement learning.

[0057] To reduce the computational costs associated with this optimization, one embodiment uses a substitute model to evaluate objectives in each optimization cycle (i.e., each round optimization or optimization for a section of the path).

[0058] A surrogate model mimics the behavior of the vehicle and driver models while being computationally less expensive. Surrogate optimization attempts to find a global minimum of an objective function (e.g., lap time) using a relatively small number of objective function evaluations.

[0059] The surrogate optimization balances two goals: exploration and speed. Exploration is conducted to find a global minimum. Speed ​​refers to the number of objective function evaluations.

[0060] The surrogate optimization alternates between surrogate construction and the search for a minimum. The surrogate model is constructed by evaluating an objective function (e.g., of the driver model) at a number of points and interpolating a radial basis function through these points.

[0061] The search for a minimum involves sampling random points (e.g., several thousand points) within constraints of an objective function f(x). The objective function f(x) is evaluated to derive a value of the objective function at one or more sampling points, and the values ​​are interpolated to generate a substitute model.

[0062] For a given point (a job holder), an evaluation function is performed based on a value s(x) of the substitute model at the job holder and a distance of each point. The distance, denoted as D(x), is defined as the difference between the substitute value at a point x and a value of the objective function at that point.

[0063] An evaluation function is determined by sampling points around the job holder and calculating the evaluation function based on the distances D(x) with reference to the points x.

[0064] The best point (i.e., the point with the highest evaluation function value) is selected, and the objective function is evaluated at the best point. The surrogate model is then set by updating it at the best point with the objective function evaluation result.

[0065] In one embodiment, the evaluation function (as f) merit (x) denotes a weighted combination of a scaled substitute model S(x) and distances D(x) as follows: fmerit(x)=wS(x)+(1−w)D(x), where w is a weight value. A larger value of w gives S(x) more importance, causing the surrogate optimization algorithm to minimize the surrogate model. A smaller value of w gives more importance to points with larger D(x) values, leading to the search for new regions of the surrogate model.

[0066] The scaled substitute model S(x) is based on a substitute value s(x) at point x as follows: S(x)=(s(x)−smin) / (smax−smin), where s max and s min Maximum and minimum values ​​are shown below the sampling points.

[0067] The distance D(x) is defined on the basis of a set of evaluated points, which are designated as x j are denoted by j = 1 ... k. d ij is defined as the distance from a sampling point i to an evaluated point k of the driver model. d min and d max are minimum and maximum values ​​of the distance d ij among all evaluated and scanned points.

[0068] D(x) can be represented by the following: D(x)=(dmax−d(x)) / (dmax−dmin), where d(x) is the minimum distance between a point x on the substitute model and an evaluated point k.

[0069] In block 104, environmental uncertainties are taken into account, and local path optimization is performed using a probabilistic prediction method. Prediction module 72 performs local prediction during vehicle operation (e.g., when the vehicle travels through a section of a racetrack).

[0070] In general, prediction module 72 determines the probability that a given emergency event will occur (i.e., an event or condition that the vehicle will encounter if no action is taken that deviates from a planned path).

[0071] With reference to Fig. In one embodiment, the prediction module 72 uses a prediction tree to evaluate the probability of an emergency event (or any one of several potential emergency events) occurring for each of several prediction windows. The prediction windows are continuously updated as the vehicle is operated.

[0072] The prediction tree shows nominal and emergency prediction branches with their associated probabilities. As in Fig. As shown in Figure 5, for example, a prediction tree comprises 90 nodes or time steps and branches representing nominal conditions (no emergency event, the vehicle can continue normally). The prediction tree 90 also includes branches for emergency conditions. Any number of prediction windows, branches, and nodes can be generated, and they are therefore not limited to those shown in Figure 5. Fig. Limited to the 5 shown.

[0073] As shown, the prediction tree comprises 90 nodes 0, 1, N1, N1+1, N2 and N3, each with an associated position state x i The subscript indices i comprise a subscript n, which indicates a nominal condition, and a subscript e, which indicates an emergency event. The number of subscript indices is based on a number of probable emergency events. In the example of Fig. 5 are a first emergency event in a first time horizon 91 with nodes N1 and N1+1 (e.g. a potential emerging obstacle) and a second emergency event (e.g. ice) in a second time horizon 92 with nodes N2 and N3.

[0074] For example, x n,n a positional state when no emergency events are predicted, and x e,nis a positional state when the first emergency event is predicted to occur (i.e., has a high probability or a probability above a threshold). x n,e is a positional state when the first emergency event is not predicted to occur, and the second emergency event is predicted to occur. x e,e This is a positional state when it is predicted that both emergency events will occur. Each node associated with an emergency event also includes an associated control action.

[0075] In the example of Fig. 5, if it is predicted that a position state at the beginning of the time horizon 91 will not lead to the occurrence of an emergency event (the position state is x n,n), the prediction tree 90 continues to node H1+1 without branching. However, if it is predicted that the position state at the beginning of time horizon 91 will lead to the occurrence of an emergency event (the position state is x) e,n ), the prediction tree 90 continues to node H1+1 with a branch. The prediction tree 90 can also branch in the time horizon 92, as in Fig. 5 shown.

[0076] During operation, the prediction module 72 determines the probability of an event occurring within a given time window. If the probability of an emergency event exceeds a threshold (or is more likely than the nominal condition), the prediction module 72 determines a proposed control action. When the prediction module 72 processes the decision tree 90, multiple emergency events may occur in more than one time window. In such a case, the proposed control action may combine several actions to address the emergency events.

[0077] In one embodiment, a vehicle prediction model is used to describe the vehicle dynamics (e.g., speed, direction of travel, yaw, etc.) and position states of the vehicle 10. The vehicle prediction model also optimizes control actions according to various constraints.

[0078] In one embodiment, the vehicle prediction model is a control-oriented model that uses a planar bicycle model that is in Fig. Figure 6 shows that the planar bicycle model comprises two velocity states (longitudinal velocity v). x and lateral velocity v y ), yaw rate r, position states 112 and 114, a direction deviation Δψ and a lateral deviation e.

[0079] The desired longitudinal speed v x The path along which the vehicle travels is given, and a longitudinal controller is used to calculate the required torque to maintain the desired longitudinal speed. The vehicle's lateral dynamics can be described using the following equations for lateral acceleration and changes in yaw rate: v˙y+Fyf+Fyrm−g(cos(θr)sin(ϕr)+sin(θr)Δψ)−vxr, where: v˙y=lfFyf−lrFyrm.

[0080] In the above, m is the vehicle mass, Iz is the vehicle's moment of inertia, g is the gravitational acceleration, and I f and I r These are distances from the front and rear axles to the vehicle's center of gravity (CG). θ r is the road gradient and Φ r is the crossfall angle of the road. F yf and F yr These are the forces acting on the front and rear tires.

[0081] The tire forces can be determined using a brushed tire model that captures the lateral force drop resulting from an applied longitudinal force. The tire force F y This is represented on a tire by: Fy={−Cαtan α+Cα23ξuFz|tan α|tan α−Cα327ξ2μ2Fz2tan2α|α|≺αsl−ξuFzsgn α|α|≥αsl, where: αsl=arctan3ξuFzCα.

[0082] C α is the transverse stiffness, F z is the nominal load on the tire, α is the slip angle, µ is the road friction and ξ is a performance reduction factor.

[0083] The power reduction factor represents the remaining shear force capacity based on a friction boundary circle, as follows: ξ=1−(FxμFz)2, where F x the axial force.

[0084] The front tire slip angle α f and the rear tire slip angle α r are determined on the basis of: αf=vy+lfrvx−δ, αr=vy−lrrvx.

[0085] The non-linear tire force of the rear axle (F yr ) is linearized around the current operating point of the vehicle at each time step. The front lateral force (F) yf The input signal is treated as a control input instead of the steering angle δ and is then mapped to the steering angle. The lateral dynamics are then written as affine functions of the velocity and yaw states: v˙y=Fyf+F¯yr+C¯αr(α−α¯r)m−g sin(ϕr)cos(θr)−vxr, r˙=lfFyf−lr(F¯yr+C¯αr(α−α¯r))Iz, where F yr, F yr and F yr The lateral force of the rear axle, the rear slip angle, and the lateral stiffness are involved. The positional states of the deviation in direction of travel Δψ and the lateral deviation e are local to a path with a given curvature κ(s) and road gradient θ. r (s) and road cross slope angle ϕ(s), which define the equations of motion as: Δψ=r−uκ(s)cos(ϕr)cos(θr), e˙=uΔψ+ν.

[0086] The resulting continuous-time vehicle model includes vehicle dynamics and position states and can be represented as follows: x˙=Acx+Bcu+dc, where x is a state vector with a number of vehicle states and u is a control action. x and u are represented as follows, for example: x=[vyrΔψe], u=Fyf.

[0087] The coefficients of the vehicle model are: Ac=[C¯αrmvx−lrC¯αrmvx−vx−gsin(θr)0−lrC¯αrlzvxlr2C¯αrlzvx00010010vx0], Bc=[1mlfIz00], dc=[F¯yr−C¯αrα¯rm−gsin(ϕr)cos(θr)−lr(F¯yr−C¯αrα¯r)Iz−vxκ(s)cos(ϕr)cos(θr)0]

[0088] The continuous model is then discretized as: x(k+1)=Akx(k)+Bku(k)+dk.

[0089] The overall system model (i.e., driver model and vehicle model) is extended to encompass all prediction horizons. For example, for three potential emergency events within a prediction window, the variable position states are represented as x. i,i,i The discretized system model can be represented for k time steps as: x(k+1)=[xn,n,n(k+1)xn,n,e(k+1)⋮xe,e,e(k+1)]=[An,n,nk0000An,n,ek0000⋱00 00Ae,e,ek][xn,n,n(k)xn,n,e(k)⋮xe,e,e(k)]+[Bn,n,nk0000Bn,n,ek0000⋱0000B e,e,ek][un,n,n(k)un,n,e(k)⋮ue,e,e(k)]+[Cn,n,nk0000Cn,n,ek0000⋱0000Ce,e ,ek][un,n,n(k+1)un,n,e(k+1)⋮ue,e,e(k+1)]+[dn,n,n(k)dn,n,e(k)⋮de,e,e(k)]

[0090] The optimization of tax actions can be achieved under the use of an objective function J e be carried out, which is defined as: Je=∑k=1N[xn,n,nkxn,n,ek⋮xe,e,ek]T[pn,n,nQn,n,n0000pn,n,eQn,n,e0000⋱0000pe,e,eQe,e][xn,n,nkxn,n,n, ek⋮xe,e,ek]+[Δun,n,nkΔun,n,ek⋮Δue,e,ek]T[pn,n,nMn,n,n0000pn,n,eMn,n,e0000⋱0000pe,e,eMe,e][Δun,n ,nkΔun,n,ek⋮Δue,e,ek]+[Wrn,n,nWrn,n,e…Wre,e,e][srn,n,n(k)srn,n,e(k)⋮sre,e,e(k)]+[Wβn,n,nWβn,n,e…W βe,e,e][sβn,n,n(k)sβn,n,e(k)⋮sβe,e,e(k)]+[Wen,n,nWen,n,e…Wee,e,e][sen,n,n(k)sen,n,e(k)⋮see,e,e(k)]

[0091] In the Target Function J eThe subscript indices i include a subscript n, which specifies a nominal condition at a given time step, and a subscript e, which specifies an emergency event or a condition for which a tax action is assigned.

[0092] In the objective function J e is sr i,i,i a slack variable for a yaw rate constraint for each prediction path, sβ i,i,i is a slip variable for a drift constraint for each prediction path and se i,i,i is a slack variable for a lateral deviation constraint for each prediction path. P i,i,i is the probability of each prediction path, Q i,i,i is a tracking weight of each prediction path and M i,i,i is a weight for the proximity of tax actions for each prediction path. Wri,i,i is a weight of the yaw rate constraint slip variable for each prediction path, Wβi,i,i is a weight of the drift limitation slip variable and Wei,i,i is a weight of the lateral deviation restriction slip variable.

[0093] The objective function J e It is subject to the discrete-time model as well as various constraints. Such constraints can include steering action constraints, rate-of-slew constraints, a soft yaw rate stability constraint, a soft drift stability constraint, and a soft lateral deviation constraint (collision avoidance).

[0094] The slack variables are, for example, restricted to non-negative constraints: [srn,n,nksrn,n,ek ⋮sre,e,ek]≥0,[sβn,n,nksβn,n,ek ⋮sβe,e,ek]≥0,[sen,n,nksen,n,ek ⋮see,e,ek]≥0.

[0095] The objective function can also be subject to equality restrictions for position states as follows: xn,n,nk=xn,n,ek…=xe,e,ek for rk=1:N1 xn,n,nk=xn,n,ek=xn,e,nk=xn,e,ek for rk=N1+1:N2, xe,n,nk=xe,n,ek=xe,e,nk=xe,e,ek for rk=N1+1:N2, xn,n,nk=xn,n,ek for rk=N2+1:N3, xn,e,nk=xn,e,ek for rk=N2+1:N3, xe,n,nk=xe,n,ek for rk=N2+1:N3, xe,e,nk=xe,e,ek fork=N2+1:N3.

[0096] The objective function may also be subject to restrictions on tax equality as follows: un,n,nN1+1=ue,n,nN1+1, un,n,nN2+1=un,e,nN2+1, ue,n,nN2+1=ue,e,nN2+1, un,n,nN3+1=un,n,eN3+1, un,e,nN3+1=un,e,eN3+1, ue,n,nN3+1=ue,n,eN3+1, ue,e,nN3+1=ue,e,eN3+1.

[0097] Additional constraints can be applied, including a yaw rate stability constraint, a drift stability constraint, and / or a lateral deviation constraint. The yaw rate stability constraint is as follows: rmin≤r≤rmax, where: rmin=−μξr(gcos(φr)cos(θr)+κvx2sin(φr)cos(θr))−gsin(φr)cos(θr)vx, rmax=cos(θr)vxμξr(gcos(φr)+κvx2sin(φr))−gsin(φr).

[0098] The drift stability limitation is as follows: lrrvx−αr,sat≤β≤lrrvx+αr,sat. where α r,sat A saturated (maximum) rear slip angle is present. The lateral deviation limit is as follows: emin(k)≤e(k)≤emax(k).

[0099] With renewed reference to Fig. In section 3, block 105 presents the optimized control actions for the driver as suggested actuator controls or actuator inputs. The suggestion can take any suitable form, such as visual, acoustic, haptic, or a combination thereof.

[0100] Interface 80, for example, can be a real-time advisory module that provides information and suggestions to the driver. The module can be configured to provide an augmented reality display or screen showing obstacles and the surrounding environment, a predicted optimal path (e.g., from global optimization), and predicted optimal actuator commands (suggested steering actions). A haptic system can provide directional information and / or warnings. The intensity of a haptic signal can be controlled, for example, based on the urgency or threat of a predicted event.

[0101] An augmented reality display projected onto the windshield can, for example, provide information about obstacles or other emergency events and their probabilities, the local position of the vehicle, current actuator commands, potential changes in road gradient, cross slope and curvature, and other information.

[0102] The display can provide text and / or visual suggestions for the optimal path (e.g., route and speed profile). These suggestions assist the driver in making decisions and finding the best way to navigate a section of road or a turn, and / or in responding to an emergency. Depending on the situation, the driver may be presented with different options. For example, if an obstacle is predicted, several suggestions may be displayed. The driver can then decide which option to take, for example, based on their preferred driving style. Predictions regarding the level of downforce to be applied can also be displayed to inform the driver and reduce any conflicts between the driver and the downforce actuators.

[0103] In block 106, the vehicle 10 is controlled when the driver accepts the suggested actuator controls. The driver can, for example, interact with the interface by activating a touchscreen, giving verbal or nonverbal consent, or otherwise indicating acceptance of the suggested actuator controls. In another example, the driver can accept the suggestion by directly controlling one or more actuators as proposed (e.g., by applying the brakes, depressing the accelerator pedal, steering, etc.).

[0104] Fig.Section 7 presents aspects of an embodiment of a computer system 140 that can perform various aspects of the embodiments described herein. The computer system 140 comprises at least one processing device 142, which generally includes one or more processors for performing aspects of the image acquisition and image analysis procedures described herein.

[0105] Components of the computer system 140 include the processing device 142 (such as one or more processors or processing units), a working memory 144, and a bus 146 that connects various system components, including the system working memory 144, to the processing device 142. The system working memory 144 can be a non-transient, computer-readable medium and can comprise a variety of media readable by a computer system. Such media can be any available media accessible to the processing device 142 and include both volatile and non-volatile media, as well as removable and non-removable media.

[0106] System memory 144, for example, includes non-volatile memory 148 such as a hard disk drive and may also include volatile memory 150 such as random access memory (RAM) and / or cache memory. The computer system 140 may further include other removable / non-removable, volatile / non-volatile computer system storage media.

[0107] System memory 144 can comprise at least one program product with a set (e.g., at least one) of program modules configured to perform functions of the embodiments described herein. For example, system memory 144 stores various program modules that generally perform the functions and / or methodologies of embodiments described herein. It may contain one or more modules 152 to perform functions discussed herein. System 140 is not so limited, as it may contain other modules.As used here, the term "module" refers to a processing circuit arrangement that may include an application-specific integrated circuit (ASIC), an electronic circuit, a processor (shared, dedicated or grouped) and working memory running one or more software or firmware programs, a combinational logic circuit and / or other suitable components providing the described functionality.

[0108] The processing device 142 can also communicate with one or more external devices 156, such as a keyboard, a pointing device, and / or any other devices (e.g., network card, modem, etc.), enabling the processing device 142 to communicate with one or more other computing devices. Communication with various devices can take place via input / output interfaces (I / O interfaces) 164 and 165.

[0109] The processing device 142 can also communicate with one or more networks 166, such as a local area network (LAN), a wide area network (WAN), a bus network, and / or a public network (e.g., the Internet), via a network adapter 168. It should be understood that, although not shown, other hardware and / or software components can be used in conjunction with the computer system 140. Examples include, but are not limited to: microcode, device drivers, redundant processing units, external disk drive arrangements, RAID systems, and data archive storage systems, etc.

[0110] The terms "a" and "an" do not denote a limitation of quantity, but rather indicate the presence of at least one of the cited subject matter. The term "or" means "and / or" unless clearly indicated otherwise by the context. Reference throughout the patent description to "an aspect" means that a specific element (e.g., feature, structure, step, or property) described in connection with that aspect is contained in at least one aspect described therein and may or may not be present in other aspects. Furthermore, it should be understood that the described elements can be combined in any suitable manner across the various aspects.

[0111] When an element such as a layer, film, area, or substrate is described as being "on" another element, it can be directly on top of that element, or intermediate elements may be present. Conversely, when an element is described as being "directly on" another element, no intermediate elements are present.

[0112] Unless otherwise stated herein, all test standards are the most recent applicable standard since the filing date of this application, or, if priority is claimed, the filing date of the earliest priority application in which the test standard appears.

[0113] Unless otherwise defined, technical and scientific terms used herein have the same meanings as they would normally be understood by a person skilled in the art in the field to which this disclosure belongs.

[0114] Although the above disclosure has been described with reference to exemplary embodiments, it is understandable to the person skilled in the art that various modifications can be made and elements thereof can be exchanged for equivalents without deviating from its scope of protection. Furthermore, many modifications can be made to adapt a particular situation or material to the teachings of the disclosure without deviating from its essential scope of protection. Therefore, it is intended that the present disclosure is not limited to the specific embodiments disclosed, but encompasses all embodiments that fall within its scope of protection.

Claims

[1] System (60) for interacting with a vehicle user, comprising: an optimization module (70) configured to determine a control strategy for traversing a selected route based on a probabilistic prediction of actuator inputs; a monitoring module configured to monitor vehicle movement and vehicle user inputs during the operation of a vehicle (10) over the selected route; a prediction module (72) configured to predict an optimal local path for traversing a section of the selected route and to predict a set of optimal actuator control actions; an interface module configured to present a suggested set of actuator commands to the vehicle user based on the set of optimal actuator control actions; a control system configured to control the vehicle (10) to execute the proposed set of actuator commands, based on the vehicle user's acceptance of the proposed set of actuator commands; characterized by : a data fusion module (64) configured to select a detection method for detecting each of several emergency events, wherein the detection method specifies a combination of one or more types of detection data, wherein the detection method is selected based on a probability that an emergency event will be detected by the selected detection method and a reliability of the selected detection method. [2] System (60) according to claim 1, wherein the control strategy is determined on the basis of a driver model and a vehicle model and prescribes actuator input values ​​over time that maximize the performance of the vehicle (10) when driving along the selected route. [3] System (60) according to claim 2, wherein the control strategy is determined on the basis of a substitute model constructed from the driver model. [4] System (60) according to claim 1, wherein the prediction module (72) is configured to predict the optimal local path over several successive time horizons based on the control strategy and a probability of encountering an emergency event during each successive time horizon. [5] System (60) according to claim 1, wherein the selected route is a pre-planned route along a road with known parameters, and the control strategy is determined to optimize actuator inputs to minimize the amount of time required to travel the pre-planned route. [6] Methods of cooperating with a vehicle user, comprising: Determining a control strategy for driving along a selected route based on a probabilistic prediction of actuator inputs by an optimization module (70); Monitoring vehicle movement and vehicle user inputs during the operation of a vehicle (10) over the selected route; Predictions of an optimal local path for traversing a section of the selected route and predictions of a set of optimal actuator control actions by a prediction module (72); Representing a proposed set of actuator commands for the vehicle user based on the set of optimal actuator control actions; Controlling the vehicle (10) to execute the proposed set of actuator commands, based on the vehicle user's acceptance of the proposed set of actuator commands; characterized by : Selecting a detection method to detect each of several emergency events, wherein the detection method specifies a combination of one or more types of detection data, and wherein the detection method is selected based on a probability that an emergency event will be detected by the selected detection method and a reliability of the selected detection method. [7] Method according to claim 6, wherein the prediction of the optimal local path comprises the probabilistic prediction of the optimal local path over several successive time horizons based on the control strategy and a probability of encountering an emergency event over each successive time horizon. [8] Method according to claim 6, wherein the selected route is a pre-planned route along a road with known parameters, and the control strategy is determined to optimize actuator inputs to minimize the amount of time required to travel the pre-planned route.

Images