Patent Application
20210129839
The present invention relates to method for controlling vehicle lane holding for a vehicle with an electric power assisted steering in accordance with the preamble of claim 1 and to a system for controlling the vehicle lane holding of a vehicle with an electric power assisted steering having the features of the first part of claim 17.
Lane Keeping Aid, Lane Keeping Support, or Auto Pilot, hereafter referred to as LKA, is a vehicle functionality aiming for helping the driver to keep a vehicle in a road lane. Depending on legislation and development maturity, the guiding principles may differ significantly, spanning from only helping the driver to follow the lane when the driver is detected to steer the vehicle, at least for a reasonable time span, to be holding the steering wheel and potentially also is active to different degree of autonomy, where the driver does not need to be active at all.
Other documented LKA behaviours are functionalities that do not guide the driver continuously, but only when the driver is about to drift out of lane.
Traditional LKA variants, as e.g. described above, are lacking a good interaction with the driver in a series of important situations:
The consequence of the two afore-mentioned LKA shortcomings are that the driver does not appreciate the LKA functionality.
It is therefore an object of the present invention to provide a method and a system respectively through which one or more of the above-mentioned problems and shortcomings are overcome.
According to one aspect of the present invention, it is an objective to provide a system and a method respectively allowing to control the vehicle both from the information from the traditional LKA as well as add the ability to add a driver input in order to make the driver being able to control the lane position.
It is a particular object to provide a system and a method respectively for the driver to be able to control the vehicle lane trajectory of a vehicle.
These objects are achieved through a method and a system respectively as initially referred to having the features of the respective independent claims.
Advantageous embodiments are given by the dependent claims, and are discussed in the description.
Particularly, in an advantageous embodiment, a mixed control as referred to above is achieved as follows:
With these two target vehicle paths, it is, in accordance to this aspect of the present invention, possible to add the two target vehicle paths so that the lane can be followed by the driver in a way that reduces the effort at the same time as the lane position within the lane or even outside the lane can be controlled by the driver.
Further embodiments are described in the detailed description as well as in the dependent claims.
It will be appreciated that features of the invention are susceptible to being combined in any combination without departing from the scope of the invention as defined by the accompanying claims.
Advantageous embodiments are given by the respective appended dependent claims.
The invention will in the following be further described by way of example only, in a non-limiting manner, and with reference to the accompanying drawings, in which:
Still other objects and features of the present invention will become apparent from the following detailed description considered in conjunction with the accompanying drawings. It is to be understood, however, that the drawings are designed solely for purposes of illustration and not as a definition of the limits of the invention, for which reference should be made to the appended claims and the description as a whole. It should be further understood that the drawings are not necessarily drawn to scale and that, unless otherwise indicated, they are merely intended to conceptually illustrate the structures and procedures described herein. The same reference numerals are used for illustrating corresponding features in the different drawings.
The LKA path A comprises a relative vehicle position calculation function, or means, 420, e.g. software in an electronic control unit, a microprocessor, where the on-board vision system sensor signals are used to calculate a number of vehicle position signals, or vehicle position indicating variables, at least a curvature, and in advantageous embodiments also a curvature derivative of the lane.
The driver torque control path B comprises a driver torque calculation function, or means, 440, e.g. comprising software of a control unit, e.g. a microprocessor, which can be programmed/made such that a steering-wheel torque in
The driver torque control path B also comprises a target relative vehicle state calculation function, means, 450 (software in a control unit, a microcomputer) which is a mathematical function that for a specific vehicle speed transforms a driver torque to a lateral vehicle state; at least one of the following states or a linear combination of one or more of the following states; vehicle yaw rate or acceleration, vehicle lateral speed or acceleration, vehicle curvature and vehicle body sideslip angle. The lateral vehicle state may, unless it is a position as discussed above, furthermore be integrated into a target relative vehicle lateral lane position (integrated once if it is proportional to the lateral velocity; integrated twice if it is proportional to a lateral acceleration).
The target lateral state vector obtained from the relative vehicle position calculation function 420 of control path A and the target delta lateral state vector obtained from the relative vehicle position function 450 of control path B are added together in an addition function or addition step (software) 460 forming a mixed control target lateral state vector. At least the lane curvature, or, in alternative, advantageous embodiments also the curvature derivative of the lane, is used in a vehicle state controller 470 to achieve the target lateral state of the vehicle 480 in a controlled manner.
For facilitating the reading of the description of advantageous embodiments, below a section containing definitions used in the following description as well as some functional explanations.
An on-board vision system is a vehicle position identification functionality that identifies the vehicle position in relation to the actual lane of the road by information achieved from an on-board camera system, an on-board lidar system, an on-board radar system, an on-board telecommunication system and/or an on-board GPS system that also can be linked to map data so that the lane curvature and the lane curvature derivative can be achieved also in that way and possibly sensor fusion of them. The data achieved is in the form of e.g. a lane curvature, and optionally e.g. one or more of a lane curvature spatial derivative with respect to the forward distance from the current position, a heading angle and a lateral lane position. The LKA functionality further is used to calculate a target vehicle path in relation to the afore-mentioned vehicle position in relation to the actual lane. Thus, the target vehicle path needed to control the vehicle according to the afore-mentioned traditional LKA is available. The lane curvature and the lane curvature derivative can be used to calculate a target lateral state vector consisting of the elements for the look-ahead distance comprising at least;
A steering position actuator is an actuator which can be used to influence one or more of the steering actuator states, such as the rear wheel steering angle, the individual steering angles of the wheels, the axle braking torque or force, the wheel braking torque or force, the driving torque or force on the individual axles, the driving torque or force on the individual wheels, the camber angle on each axle, or the camber angle on each wheel.
A state is defined as a translational or rotational position, a velocity or an acceleration, or from one or more of these states derived states, such as e.g. a vehicle slip angle, which is the angle between the vehicle local x-axis and the vehicle speed vector.
A signal bus is a transmission path on which signals can be read and/or transmitted.
Steering feel is built of the sum of at least some of the following building blocks (one or more of):
A lateral acceleration feedback torque is a torque felt by the driver that corresponds to the lateral acceleration of the vehicle.
A tyre friction torque is the friction between the tyres and the road or a model of this friction.
A steering system friction or a friction torque is the friction of the parts of the linkage of the steering system or a model of this friction.
A damping torque occurs owing to damping of the tyres and the steering system or a model of this damping.
A returnability torque comes from the geometry of the steering system or a model of the steering system.
These torque contributions can be vehicle speed dependent. The torque contributions can also be calculated via mathematical models or sensed via sensors in the vehicle or steering system.
A compensation torque is one of, or the sum of one or more of, the above-mentioned tyre friction torque, the friction torque, the damping torque and the returnability torque. The parts of the compensation torque are calculated from mathematical models of the different torque parts.
The lateral acceleration torque is calculated from a bicycle model, which uses vehicle speed and steering angle as input, and give the lateral acceleration as output. The lateral acceleration feedback is a function of the lateral acceleration calculated from the vehicle model.
The mathematical model of the tyre friction torque is a model of an angle or angular speed driven hysteresis. The mathematical model of the tyre also contains a relaxation part such that as the tyre rolls, the torque of the hysteresis will have a relaxation length so that the hysteresis torque decreases with the rolling length of the tyre. The relaxation can preferably be the well-known half-life exponential decay function. The model of the tyre friction is the combination of the hysteresis and the relaxation so that there e.g. can be an increase owing to the hysteresis torque taking place at the same time as the torque decreases owing to the relaxation. The resulting torque of the model is then the sum of the two parts.
The mathematical model of the friction torque is a model of an angle or angular speed driven hysteresis. The maximum torque in the hysteresis can be shaped by a function so that the maximum torque is different on centre compared to off centre.
The mathematical model of the damping torque consists of a damping constant times an angular speed or translational speed, such as e.g. the rack velocity, measured somewhere in the linkage between the road wheels and the steering wheel. The damping constant can be such that the damping has a blow-off, such that the damping constant decreases for great angular or translational speeds. The damping constant can be vehicle speed dependent as well as different for steering outwards compared to inwards. The damping constant can also be a function of the steering-wheel or torsion-bar torque.
A returnability torque is a vehicle speed dependent and steering-wheel angle dependent torque.
A driver torque is the torsion-bar torque compensated with a compensation torque as discussed above.
Controllability describes the ability of an external input to move the internal state of a system from any initial state to any other final state in a finite time interval.
A vehicle state controller is here defined as a dynamic function for achieving a target state in a vehicle in a controlled manner.
A vehicle state actuator, is an actuator that when actuated influences one or several vehicle states. Examples of vehicle state actuators are brakes, engine, controllable four-wheel-drive clutches, controllable differentials, active dampers, electric or hydraulic wheel motors and electrically or hydraulically driven axles.
An actuator is a mechanism or system that is operated by an ECU and converts a source of energy, typically electric current, hydraulic fluid pressure, or pneumatic pressure, into a motion, force or torque.
A target value, reference value or request is a set point for the actuator that is achieved by the use of either a closed loop controller and/or a feed-forward controller.
A vehicle model is a mathematical model that transforms a road-wheel angle or steering angle and a vehicle speed to a number of vehicle yaw and/or lateral states, e.g. vehicle yaw rate and acceleration, vehicle lateral speed and acceleration, vehicle curvature and vehicle body sideslip angle.
A transformation is defined as a mathematical function or lookup table with one input value used to produce one output value. That means that a transformation can be used, with its tuneable parameters, to create a relation between the input value and the output value with arbitrary tuneable shape. A transformation can have time-varying parameters that are even dependent on other values, a so-called gain scheduling, so that the transformation is a function with parameters that themselves are functions. An example of such a transformation is a vehicle state to driver torque relation where the relation is a vehicle speed dependent continuously rising, degressive shaped function.
A transfer function is the relation of the outputs of a system to the inputs of said system, in the Laplace domain, with the variable s, considering its initial conditions. If we, as an example of a single input, single output system, have an input function of X(s), and an output function Y(s), the transfer function G(s) is here defined to be Y(s)/X(s).
A steering-wheel torque measurement is a torque measured in the steering column or steering wheel or a force measured in the steering rack times the torque ratio between the steering rack and the steering wheel.
A steering-wheel angle is here referred to as any angle between the steering wheel and the road wheel times the ratio between the angular degree of freedom and the steering-wheel angular degree of freedom. It can also be a rack position times its ratio between the rack translational degree of freedom to the steering-wheel angular degree of freedom.
A trailer arrangement is defined as a passenger car trailer or caravan, or for a heavy truck a full-trailer, supported by front and rear axle or axles and pulled by a drawbar, a semi-trailer, or a dolly with a semi-trailer.
The steering angle, which is here shown for one wheel, but if the wheels are steered differently, as in the case for e.g. Ackermann steering, the steering angle is defined as the mean value of the angles of the two wheels.
A natural coordinate system or natural coordinates is another way of representing direction. It is based on the relative motion of the object of interest, the vehicle, rather than a fixed coordinate plane (x, y). The unit vectors (t, n) are:
A vector is an array of one or more elements.
A linear function f(x) is a function that satisfies additivity, f(x+y)=f(x)+f(y) and homogeneity of degree 1, f(αx)=αf(x) for all α. The homogeneity and additivity properties together are called the superposition principle, which implies that the order of linear transformations is not important.
The purpose of this first embodiment is purely to increase the convenience of the driver by making the effort to steer the vehicle lower. In this embodiment, the driver steers the vehicle, but the steering feel is built around the curvature of the road, i.e. the natural coordinates of the road. This is achieved by using the lane curvature and curvature derivative only from control path A of
According to one aspect of the present invention, as also referred to earlier in this application, it is an objective to control the vehicle both from the information from a traditional LKA as well as add the ability to add a driver input in order to make the driver being able to control both the lane position as well as being able to change lane with reduced effort, but the driver has to steer himself in order to stay in lane, here called a mixed control. Such a mixed control is in a first advantageous embodiment achieved by a series of steps according to
In control path A, the following method steps are taken:
In the relative vehicle position calculation function (also called means) 420, the sensor signals from the on-board vision system are, according to the definition above, used for:
The output of the relative vehicle position calculation function 420 is here a target lateral state vector consisting of the elements for the look-ahead distance comprising at least;
In control path B, to have normal steering feel around the states described by the target lateral state vector, the following method steps are taken:
In the driver torque calculation function (also called means) 440, the sensor signals from the steering system are, according to the definition above, used for:
In the target relative vehicle state calculation function (also called means) 450, the driver torque is, according to the definition above, used for a transformation of the driver torque, according to
The output of the target relative vehicle state calculation function 450 is a target delta lateral state vector consisting of the target delta curvature and a derivative the target delta curvature.
After the two control paths, A and B, there is an addition performed in addition function or addition step 460 where the target lateral state vector (e.g. κ, cf.
In the vehicle state controller 470, the vehicle path is controlled in the following method steps:
In a first embodiment of the first method step of 470 for the calculation of a target yaw and/or lateral vehicle state, the mixed control target lateral state vector is used to calculate e.g. a target lateral acceleration. This calculation is made by first calculating the curvature function of the distance, which is the sum of the curvature and the spatial curvature derivative with respect to the forward distance from the current position times the distance in the look-ahead frame of interest. With the vehicle speed, this function is converted to a function of time rather than of distance. The target lateral acceleration is the curvature function of the distance times the square of the vehicle speed.
Besides the before-mentioned target vehicle lateral acceleration, other possible target vehicle states are curvature, road-wheel angle, steering-wheel angle, vehicle lateral velocity, vehicle slip angle and vehicle yaw rate or linear combinations thereof, or in other words, the target can be generalized to target yaw and/or lateral vehicle states.
In a first embodiment of the second method step of the step 470 for the calculation of a target yaw and/or lateral vehicle state, which comprises controlling the vehicle actuators towards this target yaw and/or lateral vehicle state, the target yaw and/or lateral vehicle state function of time is fed through an inverse vehicle model. That means that, ideally, with vehicle speed and vehicle model, the target vehicle actuator states will be such that the vehicle path will follow the target vehicle path.
In the case of front-wheel steering only, the target vehicle actuator state is the front-wheel angle. This angle is controlled in a typical front steering gear of the vehicle 480 with a road-wheel angle interface or a steering-wheel angle interface times the ratio to the road wheels.
In the case of front-wheel and rear-wheel steering, the target vehicle actuator states are the front-wheel and rear-wheel angles. These angles are controlled in typical front and rear steering gears of the vehicle 480 with a road-wheel angle interface or a steering-wheel angle interface times the ratio to the road wheels and a rear-wheel angle interface.
Or alternatively, the target lateral states as well as the target delta lateral states are transformed to any of the in the definition listed vehicle yaw and/or lateral states.
Note specifically that the transformation between different yaw and/or lateral vehicle states from the target lateral states that is the result of the relative vehicle position calculation 420, the target delta lateral states that is the result of the target relative vehicle state calculation 450 to the actuator states in the vehicle state controller 470 can be in any order, meaning that:
The target lateral state vector that is the result of the relative vehicle position calculation 420, and the target delta lateral state vector that is the result of the target relative vehicle state calculation 450 can be in the form of steering angles or any of the in the definition listed vehicle yaw and/or lateral states, i.e. they can be generalized to vehicle yaw and/or lateral vehicle states. Depending of the choice of states, the transformation needed in the form of a quasi-static vehicle model, which for a person skilled in the art is the same as an invertible function of steering angle and vehicle speed. That means that for every vehicle speed, there is a function between the steering angle and any of the yaw and/or lateral vehicle states. And hence, as the equations described in the different embodiments are linear or can be linearized, the order of transformation is not of importance, which lies within the definition of the term linear.
With the two target vehicle paths, control path A and control path B, all combinations of alternatives, it is, in accordance to this aspect of the present invention, possible to add the two target vehicle paths to achieve a mixed control of the vehicle path so that the lane can be followed at the same time as the lane position within the lane or even outside the lane can be controlled by the driver. The result is that the drivers steering efforts will be as low as if the road were straight.
It should be clear that the invention is not limited to the specifically illustrated embodiments but that it can be varied in a number of ways within the scope of the appended claims.
| Number | Date | Country | Kind |
|---|---|---|---|
| 1751179-1 | Sep 2017 | SE | national |
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/SE2018/050966 | 9/21/2018 | WO | 00 |
