FIELD: self-driving cars.
SUBSTANCE: invention relates to a method and system for calculating data for controlling the operation of a self-driving car. The self-driving car (hereinafter – SDC) is moving along a road segment that has a traffic lane. The method contains stages at which the predicted trajectory of an object associated with an object in the road segment is obtained, a set of anchor points along the traffic lane is obtained. The predicted trajectory of the object is based on object movement data, and the set of anchor points is a default vehicle path for SDC along the traffic lane. One of the set of anchor points indicates a potential future position of SDC along the default vehicle path. For each one of the set of anchor points, a sequence of future moments of time is determined when SDC is potentially located in the corresponding future position corresponding to one of the set of anchor points, thereby forming a matrix structure that includes future position-time pairs for SDC. The future position-time pair indicates when and where SDC should be potentially located in the future along the default vehicle path. For each future position-time pair in the matrix structure, the predicted trajectory of the object is used to determine the distance to the nearest object to SDC, as if SDC is located in the corresponding future position-time pair, and the distance to the nearest object to SDC is stored in association with the corresponding future position-time pair in the matrix structure. The distance is stored in the matrix structure for the corresponding future position-time pair, which is a safe corridor for SDC, if SDC is located in the corresponding future position-time pair.
EFFECT: increase in the safety of controlling the self-driving car is achieved.
27 cl, 10 dwg
