Method and apparatus for generating trajectory shape for autonomous vehicles

Abstract

An apparatus for controlling the direction and speed of travel of an autonomous vehicle or driver-assisted autonomous vehicle (AV). A GPS and map module receives a starting position and a destination position of the AV. A plurality of sensors identify a current lane and a proposed lane for the AV. A database of AV baseline maneuver profiles is provided for use in controlling one or more of the direction and speed of travel of the AV. A trajectory profile generation module generates a planned path for the AV with a lateral acceleration of 2 Hz or less based on the starting position, the destination position, the current lane and the proposed lane for the AV, and an AV baseline maneuver file selected from the database. A steering control module controls the direction of travel of the AV based on the generated AV planned path, and a supervisory control module controls the speed of the AV based on the generated AV planned path and cochlear constraints.

Description

Technical Field

[0001] (Related Application) This application is related to and claims priority from Provisional Patent Application No. 62 / 859,649, "Optimal Trajectory Shape Generation for the Minimization of Motion Sickness in Autonomous Vehicles," filed on June 10, 2019, the disclosure of which is incorporated herein by reference.

[0002] (Technical Field) Embodiments of the present invention relate to autonomous vehicles, and more particularly to controlling the trajectory of an autonomous vehicle to minimize motion sickness of passengers in the autonomous vehicle.

Background Art

[0003] An autonomous vehicle (AV), also referred to as a connected autonomous vehicle (CAV), driverless vehicle, or robotic vehicle, is a vehicle that can sense its surrounding environment and move safely with little or no human input. Autonomous vehicles combine various sensors such as cameras, radars, lidars, sonars, GPS, odometry, and inertial measurement units to recognize the surrounding environment. A control system interprets the sensed information to identify an appropriate navigation route as well as obstacles, relevant signs, and the like.

Summary of the Invention

Problems to be Solved by the Invention

[0004] Autonomous vehicles can cause passengers to experience motion sickness. Motion sickness is related to the frequency, level, and direction of external accelerations imparted to the passenger, as well as maneuvers performed by the autonomous vehicle that are unpredictable to the passenger. Part of the passenger unpredictability is based on how the autonomous vehicle performs the maneuver, i.e., the shape of the velocity-time graph that produces the acceleration of the autonomous vehicle, and the attitude (3D orientation) of the autonomous vehicle. The time-optimal paths and constant curvature turns employed by autonomous vehicles are not how people operate vehicles. What is needed is a "passenger comfort" optimal path, where reducing the probability of motion sickness is the property to optimize.

[0005] The embodiments are presented by way of example, and not by way of limitation, and will be more fully understood by reference to the following detailed description and when considered in conjunction with the following drawings, in which: [Brief explanation of the drawings]

[0006]

Figure 1

Figure 2

[0007] Embodiments of the present invention control the trajectory and speed of an autonomous vehicle to reduce the likelihood of passenger motion sickness. Specifically, embodiments of the present invention control the transition trajectory between dynamic states of position (orientation), velocity, and acceleration of the autonomous vehicle. In this context, the dynamic state of the autonomous vehicle is composed of the vehicle's embedded longitudinal axis (rear-to-forward) changes, controlled by propulsion and braking forces, and the vehicle's lateral or cross-axis changes, controlled by steering. Multiple interdependent controls of the vehicle's speed and direction are used to create a desired motion sickness-reducing trajectory. Control includes controlling the jerk (i.e., change in acceleration) components of the dynamic states, including, but not limited to, lateral and forward motion, within the path constraints of the desired direction of travel.

[0008] The speed and direction control is managed by a monitoring control module that suppresses time-dependent lateral acceleration and jerk in a frequency range that potentially interferes with the inner ear of the passengers of the autonomous vehicle. Referring to FIG. 2, an example of the maneuvering of the autonomous vehicle controlled in this regard includes operations such as a direction change from one road to another, or a change from one lane 220 to another (such as a lane change or a direction change from one road to another). The lane 220 is defined herein as the width of an area on a road or pavement on which the autonomous vehicle can travel. More broadly, a maneuver is any change in course or speed, and a trajectory is the path taken between one or both of these states. Embodiments are primarily for autonomous vehicles, but are also applicable to driver-assisted autonomous vehicles.

[0009] Referring to FIG. 1, an embodiment 100 of the present invention receives inputs from the following three sources, namely, a GPS and map module 101 that provides the current position of the autonomous vehicle, or a selected or chosen starting position, and a selected destination position with reference to a digital map, a database of baseline maneuver profiles 103 for the autonomous vehicle, which provides basic information for controlling the direction, speed, and acceleration of the autonomous vehicle, such as information for lane changes, left or right turns, or entering or exiting a highway, sensor inputs 104 from sensors such as front sensors, and optionally side sensors, that identify, for example, the current lane and turning angle environment of the autonomous vehicle, and the lane proposed by the autonomous vehicle if it is different from the current lane. and receives inputs therefrom.

[0010] Inputs to the GPS and map module 101 (e.g., provided by or for a passenger) include the starting or current position (Position A) and the destination position (Position B) of the autonomous vehicle. The GPS and map module 101 identifies a plan to get the autonomous vehicle from Position A to Position B and provides it to the trajectory profile generation module 105. As the autonomous vehicle travels, sensors 104, e.g., one or more monocular or binocular cameras, radar, lidar, GPS transceiver, or a combination of each sensor, sense approaching turning angles or identify possible, proposed, or planned routes or lanes 220 (in accordance with the inputs received from the GPS and map module 101), incorporate parameters of the turning angles and lanes 220, including the road conditions, any traffic jams, or other types of road congestion (pedestrians, bicycles, roadworks, obstacles, signs), and the approaching turning angles determined based on the GPS and map inputs, and send this information to the trajectory profile generation module 105. The lane-keeping requirements from the database of the baseline maneuver profile 103 are also utilized by the trajectory profile generation module 105. The trajectory profile generation module 105 selects a maneuver profile from the database of the baseline maneuver profile 103 and calculates a path with a low lateral acceleration, e.g., less than 2 Hz, to achieve the maneuver. (With reference to the passenger's body structure, maintaining the main acceleration force in the vertical rather than the lateral direction is achieved in an autonomous vehicle that uses coordinated turning like a two-wheeler or rotates the passenger to adjust the sum of gravity and centrifugal force so that the passenger experiences only the vertical sensation.)

[0011] The calculated path with low lateral acceleration is sent to the steering control module 107 and the supervisory control module 106. The supervisory control module 106 uses the path geometry, the cochlear constraints 102, the speed limit obtained from the GPS and map module 101, and environmental information (e.g., road conditions, traffic congestion) to select a speed for the autonomous vehicle. This speed is sent to the trajectory profile generation module 105 to set the lane margin 205 in subsequent iterations of generating a trajectory for the autonomous vehicle, as well as to the speed control module 108. The cochlear constraints 102 consist of the frequency level of the cochlear crossover spectrum and the decay time of the disturbance history.

[0012] According to one embodiment, the predictive trajectory generation module utilizes a set of tables for locating points between lane boundaries for the full range of turning angles. These data are generated based on the autonomous vehicle design during the development phase of the system design and include the rate of progression through turns when the number of defining points is changed from three to two.

[0013] According to Embodiment 200, the parameters of the moving direction or turning direction (DOT) 225 and the lane 220 shown in FIG. 2 include the lane 220 and, if there is a planned area, the planned area 210 or the trajectory of the center of the desired route. The planned area 210 is defined herein as the width of the lane 220 minus the width of the lane margin 205 (both if one or both left and right lane margins exist) and the width of the autonomous driving vehicle (vehicle width 215). The lane margin 205 is defined herein as a safety barrier between the outermost maneuver of the vehicle and the edge of the lane 220. The lane margin 205 is a function of the vehicle speed and the lane 220. The planned area 210 is a route range that can be used by the trajectory profile generation module 105 to design a route for the autonomous driving vehicle to travel with the lowest inner ear disorder, or at least to maintain the inner ear disorder below a threshold understood to cause motion sickness. The inner ear disorder is a function of the route design and the vehicle speed. When maneuvering an autonomous driving vehicle on a narrow road, the planned area 210 may decrease to zero, and the only parameter available for controlling the lateral acceleration is the speed of the autonomous driving vehicle.

[0014] Note that it should be noted that the planned area 210 may decrease to a line if the width of the lane 220 decreases to less than twice the width of the lane margin 205 and the width of the autonomous driving vehicle width 215. Furthermore, the planned area 210 may reach an endpoint if the width of the lane 220 decreases to less than the vehicle width 215. The vehicle width 215 is defined herein as the cross-section of the vehicle orthogonal to the velocity vector with respect to the lane 220 and is a parameter of the vehicle's structure and posture. The vehicle width 215 determines the limit distance at which the center of the generated trajectory may approach the inner edge of the lane margin 205. The generated trajectory is defined herein as a route that satisfies geometric planning constraints and reduces or minimizes the probability of motion sickness.

[0015] The width 215 of a two-wheeled autonomous vehicle may vary with pose unless it is spherical, i.e., a leaned two-wheeled vehicle is wider than an upright one. The lane 220, along with the width 215 and lane margin 205 requirements, are used to calculate the planning region 210. The lane margin 205 is defined herein as the portion of the lane 220 at one or more edges of the lane 220. The lane 220 minus the lane margin 205 (on one or both sides of the lane 220) defines the planning region 210. The planning region 210 provides a range of degrees of freedom for varying the trajectory of the autonomous vehicle from the center of the lane 220.

[0016] The lane margin 205 requirement is the distance a vehicle must remain from the edge or boundary of the lane 220 (either the left or right edge, one lane boundary, or both lane boundaries). According to an embodiment, the lane margin 205 requirement varies with the width of the lane 220 and the vehicle speed. A wider lane 220 results in or allows for a larger lane margin 205 requirement. The higher the autonomous vehicle speed, the greater the lane margin 205 requirement because corrective actions require a larger margin at higher speeds, resulting in higher accelerations.

[0017] According to an embodiment of the present invention, there are two basic types of maneuvers: changing targets, including lane changes, direction selection at intersections, transitions from one highway to another, and exiting or entering a highway, and following a curved road. Because motion sickness is a cumulative symptom, roads with accidental turns, curves, or lane changes can be addressed more aggressively. The history of induced acceleration fades over time. However, when driving on a winding mountain road, it is necessary to reduce speed to prolong acceleration events, and also to reduce the amplitude of lateral acceleration (i.e., the maximum lateral acceleration of the autonomous vehicle).

[0018] These factors require a two - stage path planning method as follows. Each lateral acceleration event that results in a low - frequency lateral acceleration (2 Hz or less) and has an amplitude of lateral acceleration exceeding human sensitivity is recorded along with its severity. The sum of the time - weighted severities is used to set the lateral acceleration level for future maneuvers or planned trajectories.

[0019] The planned trajectory is designed to minimize and smooth the curvature of the autonomous vehicle traveling along the trajectory. In one embodiment, a curve with C3 characteristics or higher is generated between the existing position and orientation of the autonomous vehicle and the desired position and orientation of the autonomous vehicle. That is, at least, the first derivative of the curvature is a continuous function. Further, the second derivative of the curvature may be limited in absolute magnitude. The continuity of the first derivative of the curvature of the path (C3 position curve) reduces the frequency components of the acceleration felt by the passengers and further reduces motion sickness. The smoothness of the lateral acceleration is achieved by using at least a C3 (continuous third - derivative) curve of the path.

[0020] An example of a method for generating a curve with a continuous second derivative (C2) is to use the following cubic spline.

Number

[0021] However, the path is generated using two curves of x and y. When a curve is plotted on the x - y lane, a turning path is obtained. The direction is generated by the first derivative of the composite curve, and the curvature is obtained by the first derivative of the direction and the wheelbase of the vehicle. The passengers of the autonomous vehicle desire a smooth curvature, that is, at least the first derivative of the curvature is smooth. This requires a path with at least a continuous third - derivative, that is, a C3 curve. This can be generated using a quartic spline as follows.

Number

number

number

[0022] It should be further appreciated that embodiments of the present invention may utilize C3 or higher curves other than splines, including cosine generating curves and the like.

[0023] According to an embodiment, a linear quartic spline can be used, but the process of starting with a 3-point quartic spline and using the slopes of two parametric curves with slight modifications to generate a 4-point quartic spline reduces the overall curvature and lateral acceleration experienced by the passengers.

[0024] The two curves generated are Px(t) and Py(t), where the parameter t is the distance along the path. Calculation of the coefficients is within the skill of the art. The direction and curvature are derived from Px(t) and Py(t) as follows:

number

number

[0025] According to an embodiment of the present invention, changes in velocity are controlled in a manner similar to that described above for changes in curvature of an autonomous vehicle. Embodiments further generate a continuous jerk (first derivative of lateral acceleration) function.

[0026] When the self-driving vehicle has exactly two wheels lined up in a row, there is a further reduction in motion sickness. People are less affected by vertical acceleration than by lateral acceleration. A two-wheeled vehicle has coordinated turning to maintain balance. This reduces or eliminates the lateral acceleration acting on the human sensory receptors, and the two-wheeled vehicle can move faster along the desired path with the same level of susceptibility to motion sickness. Above, it was stated that the width of the lane 220, together with the requirements of the vehicle width 215 and the lane margin 205, is used to calculate the planned area 210, i.e., the range of degrees of freedom for changing the trajectory from the center of the lane 220. When an embodiment of the present invention is applied to a two-wheeled vehicle, since the vehicle width 215 is smaller than that of a four-wheeled vehicle, the planned area 210 becomes larger.

[0027] When the self-driving vehicle is used as a self-driving taxi, the motion sickness level needs to be set to the 10th percentile level (in a given environment, a certain percentage of passengers will experience motion sickness). When it is used as a personal self-driving vehicle, the motion sickness level can be set to the percentile level of the passengers using the vehicle, for example, by turning a knob or by touch screen, or alternatively by selecting between comfort, sports, or race track driving modes. The higher the passenger's tolerance to motion sickness, the faster the passenger can move along the route to the destination.

Claims

1. An apparatus for controlling the moving direction and speed of an autonomous vehicle or a driver assistance type autonomous vehicle (AV), a GPS and map module that receives the starting position and destination position of the AV and the speed limit, and generates a plan to move the AV from the starting position to the destination position; a trajectory profile generation module that receives the plan and calculates in real time a route including a driving procedure for moving the AV from the starting position to the destination position, wherein the step of calculating the route in real time includes: tracking a sensitivity corresponding to each of lateral acceleration events having an intensity greater than a human sensitivity level; calculating a total of time-weighted severities based on the corresponding severities of the tracked lateral accelerations; setting a lateral acceleration level for the route or one of the future driving procedures based on the total of the time-weighted severities, the trajectory profile generation module including these steps; a monitoring control module (106) that receives the calculated route and selects a speed for the AV based on the geometric shape of the calculated route, inner ear constraints including the frequency level of the inner ear crossover spectrum and the decay time of the disturbance history, the speed limit, and environmental information; a speed control module that receives the calculated route and the selected speed, and controls the lateral acceleration and the change in the lateral acceleration of the AV based on the calculated route and the selected speed, characterized by the apparatus.

2. The trajectory profile generation module that receives the plan and calculates in real time a route including a driving procedure for moving the AV from the starting position to the destination position is a trajectory profile generation module that generates a curve of a continuous third derivative or a derivative greater than a continuous third derivative. The apparatus according to claim 1.

Images