Vehicle path prediction and target classification for autonomous vehicle operation
Self-driving and autonomous vehicles are very popular these days for scientific, technological, social, and economical reasons. In one aspect of this technology, path prediction and target classifications are main pillars and enablers for safety and mobility applications. Accurate Estimation of the lane assignment of the remote vehicle with respect to the host vehicle lane results in the accurate threat assessment. V2X and C-V2X are emerging technology for safety applications, and they cover relatively long range. Long range means more room for error in path prediction/target classification. Therefore, here, we offer an efficient path prediction and target classification algorithm to enable these technologies performing safety and mobility applications. The method fuses target vehicles data and host vehicle data in an efficient algorithm to perform this task. Other variations are also presented here.
1. A method for an autonomous vehicle path prediction and target classification, said method implemented by distributed processors, said method comprising:
receiving vehicle data of at least one of a location, a speed, and a heading for a vehicle and a set of other vehicles including a target vehicle;
performing cross track calculation using said vehicle data based on a plurality of forward combinations of a vehicle x and a vehicle y among said vehicle and said set of other vehicles, said cross track calculation including a plurality of terms dc xy respectively corresponding with the forward combinations;
determining in-lane or adjacent lane target classification of said target vehicle with respect to said vehicle based on said cross track calculation;
sending data or command for performing an autonomous vehicle operation of said vehicle to a controller based on determining said in-lane or adjacent lane target classification of said target vehicle with respect to said vehicle; and
performing said autonomous vehicle operation of said vehicle based on the data or command for performing the autonomous vehicle operation,
wherein performing said cross track calculation for a term dc xy , comprises calculating at least one of: a sine of an angle between the corresponding vehicle x and vehicle y in a road; and a lateral distance between the corresponding vehicle x and vehicle y at a curve.
2. The method of claim 1 , further comprising performing geofencing to correct said position of said vehicle.
3. The method of claim 1 , wherein performing said cross track calculation comprises finding an order or position of said vehicle and said set of other vehicles.
4. The method of claim 1 , further comprising generating a map for said vehicle using said vehicle data.
5. The method of claim 4 , wherein determining said map for said vehicle comprises generating lanes for said vehicle.
6. The method of claim 4 , wherein determining said map for said vehicle comprises determining one or more intersection regions for splitting said lanes.
7. The method of claim 1 , wherein determining said in-lane or adjacent lane target classification comprises removing an effect of a dc xy accumulation.
8. The method of claim 6 , wherein said one or more intersection regions are determined based on speed profiles of one or more vehicles.
9. The method of claim 6 , wherein said one or more intersection regions are determined using heading angles of one or more vehicles.
10. The method of claim 6 , wherein said one or more intersection regions are determined based on intersections of lanes generated from travel paths of one or more vehicles.
11. The method of claim 1 , further comprising performing curvature cleaning.
12. The method of claim 11 , wherein performing said curvature cleaning is based on determining previous samples of road curvature.
13. The method of claim 11 , wherein performing said curvature cleaning is based on determining previous time periods.
14. The method of claim 11 , wherein performing said curvature cleaning is based on determining an average value of road curvature.
15. The method of claim 1 , further comprising determining said position of said vehicle.
16. The method of claim 1 , further comprising determining global positioning system (GPS) coordinates for said vehicle.
17. The method of claim 1 , further comprising storing said vehicle data in a database.
18. The method of claim 1 , further comprising sending a warning to a central location.