Method and apparatus for controlling maneuvers of a vehicle

• US 5,870,303 A
• Filed: 10/14/1994
• Issued: 02/09/1999
• Est. Priority Date: 11/20/1987
• Status: Expired due to Fees

• Current Assignee: US Philips Corporation (Koninklijke Philips N.V.)
• Original Assignee: Philips Electronics North America Corporation (Koninklijke Philips N.V.)
• Inventors: Trovato, Karen I., Dorst, Leendert
• Primary Examiner(s): RUGGIERO, JOSEPH F
• Application Number:
  US08/323,189
• Time in Patent Office: 1,579 Days
• Field of Search: 364/148, 364/424.01, 364/424.02, 364/461, 364/449, 364/460, 364/167.01, 340/932.2, 701/23-28, 701/201, 701/202, 318/587, 318/568.12, 901/1, 901/50, 395/80, 395/90, 395/919, 395/902, 395/903, 395/904, 395/912
• US Class Current: 700/61
• CPC Class Codes: B62D 15/0285 Parking performed automatically
• CPC Class Codes: G05B 19/4061 Avoiding collision or forbidden zones
• CPC Class Codes: G05B 2219/35415 3-D three dimension, space input, spaceball
• CPC Class Codes: G05B 2219/40443 Conditional and iterative planning
• CPC Class Codes: G05B 2219/40448 Preprocess nodes with arm configurations, c-space and planning by connecting nodes
• CPC Class Codes: G05B 2219/40465 Criteria is lowest cost function, minimum work path
• CPC Class Codes: G05B 2219/40471 Using gradient method
• CPC Class Codes: G05B 2219/40476 Collision, planning for collision free path
• CPC Class Codes: G05B 2219/45131 Turret punch press
• CPC Class Codes: G05B 2219/49143 Obstacle, collision avoiding control, move so that no collision occurs
• CPC Class Codes: G05D 1/0217 in accordance with energy consumption, time reduction or distance reduction criteria
• CPC Class Codes: G05D 1/0246 using a video camera in combination with image processing means

Abstract

Maneuvers of a vehicle, in the presence of obstacles, are planned using a three-dimensional configuration space. Axes of the configuration space correspond to x and y coordinate locations of the rear differential of the vehicle and angle of the vehicle. The configuration space is filled with cost to goal and direction arrows values using an exhaustive search strategy. The direction arrows values point to a least cost path to a goal for the vehicle. The exhaustive search strategy involves searching a bow-tie shaped neighborhood of a goal state, and then a bow-tie shaped neighborhood of each state in the first neighborhood, iterating until all reachable states are searched. A precedence order is established so that states which are blocked are not searched.

Claims

1. Apparatus for maneuvering a vehicle comprising

means for storing a configuration space representation of a task space for the vehicle, which representation is referred to herein as a "configuration space";

means for propagating cost waves in the configuration space;

means for determining a maneuver path from a start state to a goal state in the configuration space based on the cost waves; and

means for controlling the vehicle to follow the maneuver path;

wherein the representation of the configuration space includes an array of states, each state having cost to goal and direction arrows fields.

2. The apparatus of claim 1 wherein the propagating means includes means for budding from a goal state, which budding means includes:

a) means for exploring all neighbors of a state in the configuration space;

b) means for improving neighbors of the state by assigning appropriate values of cost to goal and direction arrows fields to the state;

c) means for adding improved neighbors to a storage data structure; and

d) means for causing means 3a)-3c) to iterate for all states in the storing means.

3. The apparatus of claim 2 wherein said means for exploring searches states according to a precedence in which if a first neighbor state is blocked by an obstacle, no second neighbor state beyond the first neighbor state is searched.

4. The apparatus of claim 1 wherein the determining means includes computing means for following values in the direction arrows fields.

5. The apparatus of claim 1 wherein the array includes

a first dimension for representing a pose of the vehicle in a task space along a first Cartesian axis;

a second dimension for representing the pose of the vehicle in the task space along a second Cartesian axis; and

a third dimension for representing the pose of the vehicle in the task space according to an angular orientation of the vehicle.

6. The apparatus of claim 5 wherein

a. the vehicle is a car;

b. the propagating means includes;

i) means for exploring all neighbors of a state in the configuration space;

ii means for improving neighbors of the state by assigning appropriate values of cost to goal and direction arrows fields to the state;

iii) means for adding improved neighbors to a storage data structure; and

iv) means for causing means 7b)i)-7b)iii) to iterate for all states in the storing means;

c. the determining means includes computing means for following values in the direction arrows fields; and

d. the pose is measured along the first and second Cartesian axes according to a position of the rear differential of the car.

7. The apparatus of claim 6 wherein each state in the representation of the configuration space has a set of neighbor states which fall within a bow-tie shape in task space.

8. A device for maneuvering a vehicle comprising

a) means for storing an internal representation of objects surrounding the vehicle, which objects are relevant to a desired maneuver;

b) means for selecting a movement trajectory based on the internal representation; and

c) means for controlling the vehicle to follow the trajectory;

wherein the internal representation includes a set of possible configurations of the vehicle and the selecting step includes selecting a sequence of configurations from the set of configurations.

9. The apparatus of claim 8 wherein the vehicle is a car.

10. The apparatus of claim 9 wherein the path is a path for parallel parking between two parked cars.

11. The apparatus of claim 10 wherein the path is a path for making a three point turn.

12. The device of claim 8, wherein

a. the internal representation is a configuration space representation of a task space for the vehicle, which representation is referred to herein as a "configuration space";

b. the means for selecting comprises

i. means for propagating cost waves in the configuration space; and

ii. means for determining a maneuver path from a start state to a goal state in the configuration space based on the cost waves.

13. A method for planning a path for an object to follow and for controlling the object to follow the path in a task space comprising the steps of:

a) propagating cost waves in a configuration space representation of the task space, which propagating step comprises the steps of;

i) exploring all neighbors of a state in the configuration space using a precedence in which if a first neighbor state is blocked by an obstacle, no second neighbor state beyond the first neighbor state is searched;

ii) improving neighbors of the state by assigning appropriate values of cost to goal and direction arrows fields to the state;

iii) adding improved neighbors to a storage data structure; and

iv) causing iteration through steps 13a)i)-13a)iii) for all states in the storing means; and

b) controlling the object to follow a path indicated by the direction arrows fields.

14. A method for planning a path for an object to follow in a task space and for controlling the object to follow the path comprising the steps of:

a) propagating cost waves in a configuration space representation of the task space, which propagating step comprises the steps of;

i) exploring all neighbors of a state in the configuration space, wherein neighbors are defined according to a neighborhood which varies from state to state;

ii) improving neighbors of the state by assigning appropriate values of cost to goal and direction arrows fields to the state;

iii) adding improved neighbors to a storage data structure; and

iv) causing iteration of steps 13a)i)-13a)iii) for all states in the storing means; and

b) controlling the object to follow a path indicated by the direction arrows fields.

15. The method of claim 14 wherein a shape of the neighborhoods varies by state of configuration space.

16. The method of claim 14 wherein neighbors correspond to respective numbers of transitions along respective parameter axes of configuration space and the respective numbers vary by state of the configuration space.

17. A device for maneuvering an autonomous vehicle comprising

a) means for storing an internal representation of

i) objects surrounding the vehicle, which objects are relevant to a desired maneuver;

and

ii) a non-holonomic nature of the vehicle;

b) means for selecting a movement trajectory based on the internal representation according to some criterion for optimizing movement; and

c) means for controlling the vehicle to follow the trajectory autonomously.

18. The device of claim 17 wherein the movement is parallel parking.

19. The device of claim 17 wherein the movement is a three point turn.

20. The device of claim 17 wherein the representation of the non-holonomic nature of the vehicle includes a representation of a neighborhood of permissible motions of the vehicle.

21. A device for maneuvering an autonomous vehicle comprising

a) means for storing an internal representation of

i) objects surrounding the vehicle, which objects are relevant to a desired maneuver;

and

ii) a non-holonomic nature of the vehicle;

b) means for selecting a movement trajectory based on the internal representation according to some criterion for optimizing movement; and

c) means for controlling the vehicle to follow the trajectory autonomously, wherein

the representation of the objects around the vehicle has more than two dimensions;

the representation includes a representation of a goal or obstacle position; and

the goal or obstacle position is represented differently in different two dimensional layers of the representation.

22. The device of claim 21 wherein the representation of the non-holonomic nature of the vehicle includes a representation of a neighborhood of permissible motions of the vehicle, which permissible motions include motions between different two dimensional layers of the representation.