Method and device for providing vehicle navigation simulation environment

1. A method for providing vehicle navigation simulation environment, comprising: training a model based on historical trips taken by historical passenger groups to determine a policy that maximizes a cumulative reward for a time period, wherein the trained policy determines a real action for a real vehicle in a real environment, and wherein the training the model further comprises: providing one or more states of a simulation environment to a simulated agent, wherein: the simulated agent comprises a simulated vehicle, and the states comprise a first current time and a first current location of the simulated vehicle; obtaining an action by the simulated vehicle when the simulated vehicle has reached a milestone, wherein: the action is selected from: waiting at the first current location of the simulated vehicle, picking up a passenger group A at an origin of passenger group A's transportation, and dropping off a passenger group B at a destination of passenger group B's transportation, the passenger groups A and B each comprise one or more passengers, and the milestone is an origin or a destination of any passenger group's transportation; determining a reward to the simulated vehicle for the action;

updating the one or more states based on the action to obtain one or more updated states for providing to the simulated vehicle, wherein the updated states comprise a second current time and a second current location of the simulated vehicle; and

wherein: if the action is waiting at the current location of the simulated vehicle, the second current time is a current time corresponding to the first current time plus a time segment, and the second current location is the same as the first current location; and if the action is picking up the passenger group A at the origin of the passenger group A's transportation, the second current time is a current time corresponding to picking up the passenger group A and the second current location is the origin of the passenger group A's transportation; and if the action is dropping off the passenger group B at the destination of the passenger group B's transportation, the second current time is a current time corresponding to dropping off the passenger group B and the second current location is the destination of the passenger group B's transportation.

2. The method of claim 1 , wherein:

the simulated vehicle performs the action of waiting at the first current location of the simulated vehicle only if the simulated vehicle has no passenger onboard; and

the simulated vehicle only performs the action of dropping off the passenger group B at the destination of passenger group B's when passengers onboard of the simulated vehicle meets a maximum capacity.

3. The method of claim 1 , wherein:

the origin of passenger group A's transportation is the closest to the first current location among all origins of all passenger group transportations received as requests by the simulated vehicle by the first current time; and

the destination of the passenger group B's transportation is the closest to the first current location among all destinations of all passenger groups onboard of the simulated vehicle at the first current time.

4. The method of claim 1 , wherein determining the reward to the simulated vehicle for the action comprises:

determining the reward to the simulated vehicle for waiting at the first current location of the simulated vehicle to be zero;

determining the reward to the simulated vehicle for picking up the passenger group A at the origin of the passenger group A's transportation to be zero; and

determining the reward to the simulated vehicle for dropping off the passenger group B at the destination of the passenger group B's transportation based on a theoretical travel distance from an origin of the passenger group B's transportation to the destination of the passenger group B's transportation.

5. The method of claim 1 , wherein:

the action of picking up the passenger group A at the origin of passenger group A's transportation comprises, when the simulated vehicle has a first passenger group onboard, picking up a second passenger group;

transporting the first and second passenger groups takes the least sum of: a total extra passenger travel time for routing option 1 and a total extra passenger travel time for routing option 2 ;

the routing option 1 comprises picking up the second passenger group, then dropping of the passenger group P, and then dropping of the second passenger group;

the routing option 2 comprises picking up the second passenger group, then dropping of the second passenger group, and then dropping of the first passenger group;

the total extra passenger travel time for the routing option 1 is a summation of extra time costing the first and second passenger groups when transported by the simulated vehicle following the routing option 1 as compared to being transported one-group-by-one-group without carpool; and

the total extra passenger travel time for the routing option 2 is a summation of extra time costing the first and second passenger groups when transported by the simulated vehicle following the routing option 2 as compared to being transported one-group-by-one-group without carpool.

6. The method of claim 5 , further comprising:

if the total extra passenger travel time for the routing option 1 is less than the total extra passenger travel time for the routing option 2 , assigning the simulated vehicle to follow the routing option 1 ; and

if the total extra passenger travel time for the routing option 1 is more than the total extra passenger travel time for the routing option 2 , assigning the simulated vehicle to follow the routing option 2 .

7. The method of claim 1 , wherein the action for the real vehicle in the real environment is selected from:

(action 1 ) waiting at a current location of the real vehicle,

(action 2 ) picking up a passenger group, and

(action 3 ) dropping off a passenger group.

8. The method of claim 1 , wherein the trained policy further determines an order of dropping passenger groups onboard of the real vehicle, if two or more passenger groups are onboard.

9. A non-transitory computer-readable storage medium storing instructions that, when executed by one or more processors, cause the one or more processors to perform a method for providing vehicle navigation simulation environment, the method comprising: training a model based on historical trips taken by historical passenger groups to determine a policy that maximizes a cumulative reward for a time period, wherein the trained policy determines a real action for a real vehicle in a real environment, and wherein the training the model further comprises: providing one or more states of a simulation environment to a simulated agent, wherein: the simulated agent comprises a simulated vehicle, and the states comprise a first current time and a first current location of the simulated vehicle; obtaining an action by the simulated vehicle when the simulated vehicle has reached a milestone, wherein: the action is selected from: waiting at the first current location of the simulated vehicle, picking up a passenger group A at an origin of passenger group A's transportation, and dropping off a passenger group B at a destination of passenger group B's transportation, the passenger groups A and B each comprise one or more passengers, and the milestone is an origin or a destination of any passenger group's transportation; determining a reward to the simulated vehicle for the action;

updating the one or more states based on the action to obtain one or more updated states for providing to the simulated vehicle, wherein the updated states comprise a second current time and a second current location of the simulated vehicle; and

wherein: if the action is waiting at the current location of the simulated vehicle, the second current time is a current time corresponding to the first current time plus a time segment, and the second current location is the same as the first current location; and if the action is picking up the passenger group A at the origin of the passenger group A's transportation, the second current time is a current time corresponding to picking up the passenger group A and the second current location is the origin of the passenger group A's transportation; and if the action is dropping off the passenger group B at the destination of the passenger group B's transportation, the second current time is a current time corresponding to dropping off the passenger group Band the second current location is the destination of the passenger group B's transportation.

10. The non-transitory computer-readable storage medium of claim 9 , wherein:

the simulated vehicle performs the action of waiting at the first current location of the simulated vehicle only if the simulated vehicle has no passenger onboard; and

the simulated vehicle only performs the action of dropping off the passenger group B at the destination of passenger group B's when passengers onboard of the simulated vehicle meets a maximum capacity.

11. The non-transitory computer-readable storage medium of claim 9 , wherein:

the origin of passenger group A's transportation is the closest to the first current location among all origins of all passenger group transportations received as requests by the simulated vehicle by the first current time; and

the destination of the passenger group B's transportation is the closest to the first current location among all destinations of all passenger groups onboard of the simulated vehicle at the first current time.

12. The non-transitory computer-readable storage medium of claim 9 , wherein determining the reward to the simulated vehicle for the action comprises:

determining the reward to the simulated vehicle for waiting at the first current location of the simulated vehicle to be zero;

determining the reward to the simulated vehicle for picking up the passenger group A at the origin of the passenger group A's transportation to be zero; and

determining the reward to the simulated vehicle for dropping off the passenger group B at the destination of the passenger group B's transportation based on a theoretical travel distance from an origin of the passenger group B's transportation to the destination of the passenger group B's transportation.

13. The non-transitory computer-readable storage medium of claim 9 , wherein:

the action of picking up the passenger group A at the origin of passenger group A's transportation comprises, when the simulated vehicle has a first passenger group onboard, picking up a second passenger group;

transporting the first and second passenger groups takes the least sum of: a total extra passenger travel time for routing option 1 and a total extra passenger travel time for routing option 2 ;

the routing option 1 comprises picking up the second passenger group, then dropping of the passenger group P, and then dropping of the second passenger group;

the routing option 2 comprises picking up the second passenger group, then dropping of the second passenger group, and then dropping of the first passenger group;

the total extra passenger travel time for the routing option 1 is a summation of extra time costing the first and second passenger groups when transported by the simulated vehicle following the routing option 1 as compared to being transported one-group-by-one-group without carpool; and

the total extra passenger travel time for the routing option 2 is a summation of extra time costing the first and second passenger groups when transported by the simulated vehicle following the routing option 2 as compared to being transported one-group-by-one-group without carpool.

14. The non-transitory computer-readable storage medium of claim 13 , further comprising:

if the total extra passenger travel time for the routing option 1 is less than the total extra passenger travel time for the routing option 2 , assigning the simulated vehicle to follow the routing option 1 ; and

if the total extra passenger travel time for the routing option 1 is more than the total extra passenger travel time for the routing option 2 , assigning the simulated vehicle to follow the routing option 2 .

15. The non-transitory computer-readable storage medium of claim 9 ,

wherein the action for the real vehicle in the real environment is selected from:

(action 1 ) waiting at a current location of the real vehicle,

(action 2 ) picking up a passenger group, and

(action 3 ) dropping off a passenger group.

16. The non-transitory computer-readable storage medium of claim 9 , wherein the trained policy further determines an order of dropping passenger groups onboard of the real vehicle, if two or more passenger groups are onboard.

17. A method for providing vehicle navigation, comprising: determining a current location a real vehicle; and in response to determining that the real vehicle has reached a milestone, the milestone being an origin or a destination of a passenger group's transportation, providing an instruction based at least on a policy that maximizes a cumulative reward for the real vehicle, wherein: the policy is determined using a model trained based on historical trips taken by historical passenger groups; the provided instruction comprises: waiting at the current location, picking up a passenger group A at an origin of passenger group A's transportation, or dropping off a passenger group B at a destination of passenger group B's transportation; the passenger groups A and B each comprise one or more passengers; the instruction of picking up the passenger group A at the origin of passenger group A's transportation comprises, when the real vehicle has a first passenger group onboard, picking up a second passenger group; transporting the first and second passenger groups takes the least sum of: a total extra passenger travel time for routing option 1 and a total extra passenger travel time for routing option 2 ; the routing option 1 comprises picking up the second passenger group, then dropping of the passenger group P, and then dropping of the second passenger group; the routing option 2 comprises picking up the second passenger group, then dropping of the second passenger group, and then dropping of the first passenger group; the total extra passenger travel time for the routing option 1 is a summation of extra time costing the first and second passenger groups when transported by the simulated vehicle following the routing option 1 as compared to being transported one-group-by-one-group without carpool;

the total extra passenger travel time for the routing option 2 is a summation of extra time costing the first and second passenger groups when transported by the simulated vehicle following the routing option 2 as compared to being transported one-group-by-one-group without carpool; and

wherein: if the action is waiting at the current location of the simulated vehicle, the second current time is a current time corresponding to the first current time plus a time segment, and the second current location is the same as the first current location; and if the action is picking up the passenger group A at the origin of the passenger group A's transportation, the second current time is a current time corresponding to picking up the passenger group A and the second current location is the origin of the passenger group A's transportation; and if the action is dropping off the passenger group B at the destination of the passenger group B's transportation, the second current time is a current time corresponding to dropping off the passenger group Band the second current location is the destination of the passenger group B's transportation.

18. The method of claim 17 , wherein:

if the total extra passenger travel time for the routing option 1 is less than the total extra passenger travel time for the routing option 2 , the instruction comprises following the routing option 1 ; and

if the total extra passenger travel time for the routing option 1 is more than the total extra passenger travel time for the routing option 2 , the instruction comprises following the routing option 2 .