GB2558866A

GB2558866A - Predictive control system for autonomous driving vehicle

Info

Publication number: GB2558866A
Application number: GB1612281.4A
Authority: GB
Inventors: Ajanovic Zlatan; Stolz Michael
Original assignee: Kompetenzzentrum das Virtuelle Fahrzeug Forchungs GmbH
Current assignee: Kompetenzzentrum das Virtuelle Fahrzeug Forchungs GmbH
Priority date: 2016-07-15
Filing date: 2016-07-15
Publication date: 2018-07-25
Also published as: GB201612281D0

Abstract

This patent describes a method and a device for providing energy-optimal driving (ecodriving) of (semi)autonomous vehicle in presence of dynamically arising traffic situation dependent additional constraints (i.e other moving vehicles, traffic signs not known during initial planning step, etc.). The controlled vehicle is moving on an optimal speed trajectory (derived at least based on road gradient information) and approaching slowly the moving leading vehicle or the traffic sign not known during initial planning step. The proposed system generates an appropriate modified trajectory to overcome the newly arisen constraints and the decision whether to overtake or not (in case the constraint is a leading vehicle) aiming achieving global energy optimal motion. The trajectory and the decision are obtained based on previously derived optimal speed trajectory tree, cost-to-go map and leading vehicle motion prediction.

Description

(71) Applicant(s):

Kompetenzzentrum-Das Virtuelle Fahrzeug (Incorporated in Austria)

Forschungsgesellschaft mbH, Inffeldgasse 21/A/1, Graz 8010, Austria (72) Inventor(s):

Zlatan Ajanovic Michael Stolz (56) Documents Cited:

EP 2953110 A1 US 20150142207 A1

US 20160313133 A1 US 20140207325 A1 (58) Field of Search:

INT CL B60K, B60W, B62D, G08G Other: EPODOC, WPI, TXTA (74) Agent and/or Address for Service:

Kompetenzzentrum-Das Virtuelle Fahrzeug Forschungsgesellschaft mbH, Inffeldgasse 21/A/1, Graz 8010, Austria (54) Title of the Invention: Predictive control system for autonomous driving vehicle Abstract Title: Energy optimal control for semi-autonomous vehicle (57) This patent describes a method and a device for providing energy-optimal driving (ecodriving) of (semi) autonomous vehicle in presence of dynamically arising traffic situation dependent additional constraints (i.e other moving vehicles, traffic signs not known during initial planning step, etc.). The controlled vehicle is moving on an optimal speed trajectory (derived at least based on road gradient information) and approaching slowly the moving leading vehicle or the traffic sign not known during initial planning step. The proposed system generates an appropriate modified trajectory to overcome the newly arisen constraints and the decision whether to overtake or not (in case the constraint is a leading vehicle) aiming achieving global energy optimal motion. The trajectory and the decision are obtained based on previously derived optimal speed trajectory tree, cost-togo map and leading vehicle motion prediction.

At least one drawing originally filed was informal and the print reproduced here is taken from a later filed formal copy.

1306 18

1/10 co ο

Ο «7 Ο 5Λ Ο 'ς***' \**** [δ/Ιϋ] AliOOjGA

οοοοοο COCC^CM CM ι

Φ

Figure 2.

Figure 3.

Figure 4.

1306 18

Figure 5.

Figure 6

06 18

Figure 7.

co ο

ο

LC5

Ο

Ο 0)

CO X

Ο

C\j ο

ο «3

Figure 8.

ο	Ο	ο	ο	ο	ο	ο	ο	ο	ο
ο	ΙΌ	ο	LO	ο	ι;ο	ο	ι;ο	ο	U0
LO	'‘sf	’’sj’	co	co	OJ	OJ	Ί—	—

eoueisp co ο

ο

LO

Ο

JZS,

Ο 0) co χ ο

C\!

ο ο

«3

Figure 9.

ο	ο	ο	ο	ο	ο	ο	ο	ο	ο
ο	ΙΟ	ο	to	ο	ΙΟ	ο	to	ο	to
ΙΟ	’’’Sf'	’’sj’	co	co	OJ	OJ	Ί—	—

80UB]Sip

Getting route data

Generating optimal velocity tree and Cost-to-go map •v

Following optimal trajectory

Predicting future influence of constraints

MMMMN

Figure 10.

10/10

Predictive control system for autonomous driving vehicle

BACKGROUND OF INNOVATION (State-of-the-art)

Knowledge about upcoming driving route and road conditions enables optimization of the speed trajectory of a vehicle with respect to energy consumption. Various approaches using heuristics or optimization methods are presented in many papers and patents [1] [2] [3] [4j. Optimized speed trajectory is used as a reference value for low-level controllers such as cruise control, or as advice to a human driver with appropriate Human- Machine Interface.

The problem arises when a vehicle moving on an optimal trajectory approaches another vehicle moving in front or traffic signs not known during initial planning step. It is obvious that the vehicle cannot execute the planned motion. The vehicle has either to slow down and follow the vehicle or to overtake it. In publications dealing with the execution of optimal overtaking speed, trajectory planning is done in a way that the modified trajectory has the smallest deviation from the desired speed when the vehicle is overtaking. Various additional criteria such as safety and comfort are considered [5] [6] [7] [8j.

There are many patents dealing with overtaking and avoiding collision with moving and nonmoving obstacles on a road, providing suggestions for the driver to overtake or not to overtake; providing information about time benefits of overtake for the driver; generating collision-free trajectories; etc. The German patent [9] is probably one of the earliest dealing with the topic of overtaking. It presents a method to provide information on a possibility to overtake the vehicle in front without additional acceleration. Patent [10] describes the overtaking when multiple vehicles are involved. The simple decision making mechanism is based on a speed differences between the vehicles. The method estimates possible time gain if the overtaking is performed, considering the vehicle in front and other vehicle in a close distance in front (V2X). The goal is to avoid overtaking if the time benefits are not significant. Patent [11] uses an optimization method to generate collision-free reference trajectory for a fixed time horizon future driving. The trajectory is determined repeatedly, taking into account convenience/comfort and safety constraints. Patent [12] describes the assisting of an overtaking by taking into account overtaking time and catchup time between multiple vehicles. Based on this, an assessment if overtaking should be initiated and what speed should be used. Patent [13] uses prediction of the motion of other vehicles and estimates the gap in the neighboring lane. The decision for overtaking is made based on safety criteria. Patent [14] describes a method and an apparatus which provides information about estimated time loss if the driver doesn't overtake the vehicle in front. This was primarily developed for truck drivers who need to arrive at a destination at a specific time. Patent [15] describes a framework in which the optimal path/trajectory for a vehicle is generated in threat situations. As optimization criteria, the scalar threat assessment based at least on predicted vehicle stability is used. Optimal trajectories are generated repeatedly for a time horizon of future driving.

None of mentioned publications and patents deals with energy efficient overtaking or reusing of the calculated speed trajectory tree and cost-to-go map to generate modified speed trajectory for overcoming dynamically arising traffic situation dependent additional constraints with global optimal solution.

SUMMARY OF INNOVATION (Difference from the state-of-the-art)

Papers and patents dealing with ecodriving do not consider dynamically arising traffic situation dependent additional constraints (i.e other moving vehicles, traffic signs not known during initial planning step). The integration of constraints into optimal ecodriving calculation is necessary to achieve valid real driving benefits. Without the integration, it is possible that the optimal ecodriving produces even worse results and increase energy consumption. Papers dealing with optimal overtaking consider only deviation from a desired speed which leads to a locally optimal solution only. And simply limiting speed to the speed limit if planned speed is higher would result in nonoptimal solution.

The proposed invention presents system acting in a way that upon detection of leading vehicle moving in front or traffic signs not known during initial planning step an eventual conflict is predicted and the speed trajectory is changed to satisfy newly detected constraints. In contrary to existing approaches, this method aims to achieve global optimal solution by generating a new optimal speed trajectory. The new optimal speed trajectory has to be generated because the newly arisen constraints (leading vehicles moving in front or traffic signs not known during initial planning step) changed the initial optimization problem and therefore the previously determined optimal speed trajectory is not optimal anymore, so there is no benefit in persisting to still follow it.

The method and the apparatus of proposed invention also gives valid conclusion if it is more beneficial to overtake the vehicle or to slow down based on predicted energy usage for the trip and actual overtaking maneuver. Upon detection of a leading vehicle moving in front, the method described is used to properly re-plan optimal speed trajectory (i.e it can recognize that it is better to speed up and overtake the vehicle sooner). Re-planning is based on the information about the leading vehicle moving in front (i.e velocity and distance) and information about upcoming road (i.e topology, curves, etc.). Presented method has integrated decision making and overtaking execution part such that it also assesses overtaking execution cost in the decision making process. In this way if the execution of overtaking is too costly from an energy point of view, the decision for overtaking will not be positive.

Upon detection of traffic signs not known during initial planning step, the method predicts the influence of the traffic sign. If the traffic sign is time-invariant (i.e speed limit) the method predicts the range in operational space this traffic sign is valid. This can be fixed estimated distance, until the next intersection or until the next known traffic sign canceling it. If the traffic sign is time-varying (i.e traffic light) the method predicts it's states as long in future as it influences the optimization problem. The method then predicts the conflict between the newly arisen constraint and the current planned speed trajectory and if necessary re-plans trajectory.

This approach does not significantly increase the problem complexity as we can reuse already derived optimal trajectories (optimal trajectory tree) and cost-to-go map from every system state point to the end of the trip as a side product of using backward calculation method based on state of the art methods such as Bellman principle (Dynamic Programming). Re-planning is required only for a transition period from current system state in operational state point to a new system state in the operational state point where influence of leading vehicle moving in front is eliminated (i.e overtaking or slowing down).

DETAILED DESCRIPTION (Inventive step and non-obviousness)

Method and device

This invention comprises a method and a device for energy efficient driving in a dynamic environment (i.e. other vehicles in traffic or previously unknown traffic signs). The device comprises a navigation unit, an constraint detection unit, a predictive control unit, a memory and the interfaces to communicate and interact with the vehicle and the driver.

The method for deriving optimal vehicle speed trajectory consists of seven steps, which are described below. Their relation is represented on a flowchart in Figure 10.

1) Getting route data

During this step, the route data for the desired trip is either acquired from HMI or other source of information (i.e V2X, cloud service). This data can be current position, goal and other parameters of a drive (i.e preferred route, traveling time constraints). According to a selected route, information about upcoming road inclination, road curvature, traffic conditions, speed limitations, weather conditions and other relevant information is acquired i.e from an onboard navigation system, cloud service, communicated with other vehicles or other source of information.

2) Generating optimal velocity tree and cost-to-go map

During this step, the optimization procedure for the desired trip aiming minimum cost of traveling is executed. The cost can be represented by energy (fuel) used for trip with additional factors (i.e time, comfort, safety) in a weighted sum. To calculate the cost dynamics model of the vehicle, the influence of gravity component (calculated from road gradient), air drag resistance, roll resistance, boardnet power consumption etc. is considered. For the calculation, efficiency maps for powertrain components can also be used. Route information such as speed limits and curvature speed constraints are also considered. The optimization procedure is based on Bellman's principle of optimality (Dynamic Programming) for the desired horizon. It runs backwards from the final point to the start point in operational space (i.e distance or time domain).

During the optimization step a cost-to-go map is generated. The cost-to-go map represents optimal costs to finish the trip from each discretized system state variable (i.e velocity, battery State of Charge, etc.) in each discretized operational space point to the end of the trip. Beside this map, information on trajectories from each point to end, which have this optimal cost-to-go, is stored. These trajectories build an optimal velocity tree.

3) Following the optimal trajectory

During this step, according to the information on current position (acquired from the navigation system), the predictive control system uses actual optimal velocity trajectory to set desired velocity for lower level controller. This lower level controller can be either directly vehicle traction power controller (i.e ECU, E-motor controller, transmission controller) or driver who is informed by means of proper HMI how to achieve optimal driving behavior.

4) Detecting newly arisen traffic situation dependent additional constraints

During this step, the detection of the vehicle moving in front or traffic signs not known during initial planning step is analyzed. Detecting of the constraints can be achieved with a state of the art onboard sensor (i.e camera, radar, ultrasonic sensor, lidar sensor etc.) or other means of sensing (i.e cloud service, V2X communication, etc.). If the constraint is detected, the impact on a planned vehicle motion is analyzed and proper re-planning in an efficient manner is carried out. Otherwise step 3 is applied.

5) Predicting future influence

If detected constraint is leading vehicle moving in front during this step, the motion of leading vehicle moving in front is predicted. The prediction can be based on simple assumption that the vehicle will continue to drive with the same speed or advanced prediction by using historical information of the vehicle, known information of incoming route, traffic conditions and the leading vehicle behavior model or even V2X communication. Vehicle behavior models can be also adjusted to a detected type of vehicle moving in front.

If the detected constraint is a traffic sign, its influence is predicted differently: If the traffic sign is time-invariant (i.e speed limit) the method predicts the range in the operational space this traffic sign is valid. This can be fixed estimated distance, until the next intersection or until the next known traffic sign canceling it. This provides the speed limit and the space range where it is valid. If the traffic sign is time-varying (i.e traffic light) the method predicts it's states as long in future as it influences the optimization problem. This provides the speed limit and the time intervals in future when the speed limit is valid. These are needed to predict the conflict in future.

6) Predicting conflict

During this step, the conflict between the planned motion of the controlled vehicle and dynamically arising traffic situation dependent additional constraints is predicted.

If the detected constraint is leading vehicle moving in front based on the predicted motion of the leading vehicle, the planned motion of the controlled vehicle and environment information, possible conflicts according to the defined criteria such as minimum distance and speed difference are detected. If the distance between the vehicles is less than some predefined value (or time to collide) and the speed difference is not big enough to overtake (less than the predefined value), a conflict is predicted.

If the detected constraint is the traffic sign, the conflict is predicted differently: If the traffic sign is time-invariant (i.e speed limit) the method checks if the planned speed is higher than the speed limit on the predicted range. If this is true, the conflict is predicted. If the traffic sign is time-varying (i.e traffic light) the method checks if the planned speed us complying with state of the traffic sign in the planned time when the vehicle reaches the point in space where the traffic signal is active. If the planned speed is not complying the conflict is predicted.

If the conflict is predicted, proper adjusting of optimal trajectory is needed, otherwise the vehicle can continue following the planned speed trajectory.

7) Adjusting optimal trajectory

During this step, the optimal trajectory is adjusted by taking into consideration dynamically arisen constraints. The adjustment is done by re-planning the optimal trajectory in an efficient way by reusing cost-to-go map and an optimal speed trajectory tree. The re-planning is done with forward trajectory building method (i.e forward DP, A*, heuristics, etc.) starting from current system state in operational space to the system state and operational space point where the dynamically arisen constraints have no influence on the optimal speed trajectory (i.e overtaking performed, traffic light passed, planned speed trajectory slowed under speed limit) and is merged with the optimal trajectory tree previously derived. The re-planning is considering several safety factors such as maximum time of overtaking execution (related to the minimum relative speed to the leading vehicle) in case of overtaking leading vehicle and minimum distance (or time to collision) from the leading vehicle in case of following the leading vehicle.

This procedure can be used for an entire trip or a finite moving horizon which is repeated during driving.

References [1] G. D. N. L. 0. A Sciarretta, Optimal Ecodriving Control: Energy-Efficient Driving of Road Vehicles as an Optimal Control Problem, IEEE Control Systems Magazine, pp. 71-90, October 2015.

[2] M. S. F. R. Oskar Johansson, Method and module for controlling a vehicle's speed. US Patent US8744718B2, 21 June 2011.

[3] J. Η. Η. P. Oskar Johansson, Method and module for deretmining of velocity reference values for a vehicle control system. US Patent US20120083984 Al, 31 May 2010.

[4] J. Ο. M. S. J. S. D. Y. Dimitar P. Filev, Efficiencz-based Speed control with traffic-compatible speed offsets. US Patent US8930115B2, 26 February 2013.

[5] S. H. W. D. B. A. R. M.Wang, Optimal Lane Change Times and Accelerations of Autonomous and Connected Vehicles, TRB 95th Annual Meeting Compendium of Papers, 1 January 2016.

[6] J. S. N Murgovski, Predictive cruise control with autonomous overtaking, in 54th IEEE Conference on Decision and Control (CDC), Osaka, December 2015.

[7] S. T. T. M.A.S. Kamal, Efficient Vehicle Driving on Multi-lane Roads Using Model Predictive Control Under a Connected Vehicle Environment, in IEEE Intelligent Vehicles Symposium (IV), Seoul, Korea, July 2015.

[8] T. Shamir, How should an Autonomous Vehicle Overtake a Slower Moving Vehicle: Design and Analysis of an Optimal Trajectory, IEEE Transactions on Automatic Control, vol. 49, pp. 607-610, April 2004.

[9] A. D. Wiesmeier, Device for displaying overtaking recommendations for the driver of a vehicle. DE Patent DE3622447C1, 4 July 1986.

[10] E. I. Peter Harda, Overtake assessment arrangement and system and autonomous vehicle with an overtake assessment arrangement. US Patent US2015/0353094A1, 3 June 2015.

[11] M. A. Julia Nilsson, Manoeuver generation for automated driving. US Patent US2015/0073663A1, 5 September 2014.

[12] A. Martin, Method and system for assisting overtaking. EU Patent EP3018029A1, 4 November 2014.

[13] J. S. Μ. K. T. W. Sven Rebhan, Method and system for predictive lane change assistance, program software product and vehicle. US Patent US20150321699A1, 15 April 2015.

[14] T. Sandberg, Method and apparatus for providing a driver with support for taking decisions before overtaking. World Patent WO2015108474A1, 15 January 2015.

[15] S. C. P. K. D. I. Sterling J. Anderson, Integrated framework for vehicle operator assistance based on a trajectory prediction and threat assessment. US Patent US8744658B2, 9 April 2013.

[16] L. Z. B. R. Y. P. a. P. H. Da Yang, Modeling and Analysis of the Lane-Changing Execution in Longitudinal Direction, IEEE Transactions on Intelligent Transportation Systems, March 11, 2016.

BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 shows a system block diagram of the device illustrating relations of predictive control system with other parts of system.

Figure 2 shows an optimal velocity tree represented with blue lines and optimal trajectory from initial trip start at standstill represented with green line. On the lower graph the red line represents the road gradient in percentage and green line represents road altitude (multiplied by 10 to be comparable to the gradient). The trip length is limited to only 500 m in order to make influence of the road gradient visible clearly. Velocity discretization step is 1 m/s in order to keep tree simple. In a practical usage this step should be smaller (0.1 m/s).

Figure 3 shows a cost-to-go map in 3D representation with colors representing optimal energy needed to finish the trip from current state and operational space point. It can be noted that the overall shape depends on the road gradient and that there is a valley around 300 m which is a top of the hill. Also costs are smaller at higher speeds which results from higher kinetic energy of the vehicle.

Figure 4 shows a cost-to-go map in 3D representation with colors from other angle.

Figure 5 shows a cost-to-go map in 2D representation with contours.

Figure 6 shows a cost-to-go map in 2D representation with colors.

Figure 7 shows a combined cost-to-go map in 2D representation with optimal speed trajectory tree. This illustrates the relation between each trajectory branch from optimal trajectory tree and corresponding cost from cost-to-go map. Minimal costs represented on cost-to-go map will only be achieved if vehicle continues to move on the trajectory starting from that point to the end of the trip. Figure 8 shows controlled and leading vehicle trajectories. The leading vehicle is 30 m in front of the controlled vehicle at the beginning of the trip and moving with 8 m/s. The controlled vehicle will reach the leading vehicle around 25 seconds after a start of the trip. With planned speed overtaking would last too long (around 20 seconds), to avoid that trajectory has to be modified. The controlled vehicle should either speed up (pass to the faster trajectory) and overtake or slow down (pass to the slower trajectory).

Figure 9 shows controlled and leading vehicle trajectories in case of no conflict. The leading vehicle is 100 m in front of the controlled vehicle at the beginning of the trip and moving with 5 m/s. The controlled vehicle reaches the leading vehicle around 25 seconds after a start of a trip. As speed difference is enough to execute overtaking in a predefined maximum time, overtaking will be executed and the speed trajectory doesn't have to be modified to avoid collision. The speed difference can be concluded from the angle between trajectories.

Figure 10 shows a flowchart representing the method used by predictive control system to adapt to the newly arisen constraints.

Claims

Claims What is claimed is:

1. A method for providing energy optimal control and automated decision for overtaking of a semi autonomous vehicle comprising in a first step the incorporation of route data, in a second step the calculation and storing of cost-to-go map and optimal trajectory tree in an usable manner in a third step the following of the calculated optimal trajectory, in a fourth step the detection of newly arisen constraints, in a fifth step the prediction of the influence of the constraint, in a sixth step the prediction of conflicts, in an seventh step the adjusting of the optimal trajectory

2. A device for providing at least one of the values reference speed and lane for motored vehicle comprising a navigation unit, an constraints detection unit, a predictive motion planning unit, a memory and the interfaces to communicate and interact with the vehicle and the driver.

References [1] A. Sciarretta, G. D. Nunzio and L. Ojeda, Optimal Ecodriving Control: Energy-Efficient Driving of Road Vehicles as an Optimal Control Problem, IEEE Control Systems Magazine, pp. 71-90, October 2015.

[2] 0. Johansson, M. Sodergren and F. Roos, Method and module for controlling a vehicle's speed. US Patent US8744718B2, 21 June 2011.

[3] 0. Johansson, J. Hansson and H. Pettersson, Method and module for deretmining of velocity reference values for a vehicle control system. US Patent US20120083984 Al; 31 May 2010.

[4] D. P. Filev, M. J. 0., S. J. Sywabowski and D. Yanakiev, Efficiency-based Speed control with traffic-compatible speed offsets. US Patent US8930115B2, 26 February 2013.

[5] M.Wang, P. H. S, W. Daamen, B. Arem and R.Happee, Optimal Lane Change Times and Accelerations of Autonomous and Connected Vehicles, TRB 95th Annual Meeting Compendium of Papers, 1 January 2016.

[6] N. Murgovski and J. Sjoberg, Predictive cruise control with autonomous overtaking, in 54th IEEE Conference on Decision and Control (CDC), Osaka, December 2015.

[7] M. Kamal, S. Taguchi and T.Yoshimura, Efficient Vehicle Driving on Multi-lane Roads Using Model Predictive Control Under a Connected Vehicle Environment, in IEEE Intelligent Vehicles Symposium (IV), Seoul, Korea, July 2015.

[10] P. Harda and E. Israelsson, Overtake assessment arrangement and system and autonomous vehicle with an overtake assessment arrangement. US Patent US2015/0353094A1, 3 June 2015.

[11] J. Nilsson and M. Ali, Manoeuver generation for automated driving. US Patent - US2015/0073663A1, 5 September 2014.

[12] A. Martin, Method and system for assisting overtaking. EU Patent EP3018029A1,4 November 2014.

[13] S. Rebhan, J. Schmudderich, M. Kleinehagenbrock and T. WeiBwange, Method and system for predictive lane change assistance, program software product and vehicle. US Patent US20150321699A1,15 April 2015.

8/8

[14] T. Sandberg, Method and apparatus for providing a driver with support for taking decisions before overtaking. World Patent WO2015108474A1,15 January 2015.

[15] S. J. Anderson, S. C. Peters and K. D. laegnemma, Integrated framework for vehicle operator assistance based on a trajectory prediction and threat assessment. US Patent US8744658B2, 9 April 2013.

[16] F. Flehmig and U. Schulz, Method and driver assistance device for supporting lane changes or passing maneuvers of a motor vehicle. US Patent US 2015/0142207 Al, 21 May 2015.

[17] U. P. Mudalige and M. Losh, Efficient Data flow algorithm for autonomous lane changing, passing and overtaking behaviors. US Patent US 2014/0207325,24 July 2014.

[18] T. Yuya and T. Taisetsu, Travel control device and travel control system. EU Patent EP 2 953 110 Al, 09 December 2015.

8/8

Intellectual

Property

Office

Application No: GB1612281.4 Examiner: Mr Vivek Raghavan

2. A device for providing energy optimal control and automated decision for overtaking of a semiautonomous vehicle comprising a navigation unit, an constraint detection unit, a predictive control unit, a memory and the interfaces to communicate and interact with the vehicle and the driver.

Amendments to the claims have been filed as follows:

Claims

What is claimed is:

1. A method for providing energy optimal control and automated decision for overtaking of a semiautonomous vehicle comprising :

• In a first step acquiring of route data for the whole trip or at least one segment with defined final desired velocity, • Ina second step calculating and storing of a cost-to-go map and an optimal trajectory tree for the route without dynamic constraints, • In a third step repeatedly acquiring actual information about other traffic participants, traffic light states or other dynamic constraints, • In a fourth step adjusting of optimal reference speed trajectory and reference lane trajectory using information from step two and step three.

• Ina fifth step providing at least one of the values reference speed or reference lane to driver or vehicle motion controller