US-12617433-B2 - Algorithm operation management apparatus and method

US12617433B2US 12617433 B2US12617433 B2US 12617433B2US-12617433-B2

Abstract

Disclosed herein is an algorithm operation management apparatus and method, in which the algorithm operation management apparatus includes a learning part configured to learn by classifying an accident situation and a safety situation for each driving condition; a determination part configured to adjust an operating parameter for at least one algorithm related to driving control of a vehicle on the basis of a determination result of the learning part; and an operation part configured to perform the driving control of the vehicle by applying the adjusted operating parameter on the basis of the at least one algorithm.

Inventors

Byeong Hwan Jeon
Young Hun NA

Assignees

HYUNDAI MOBIS CO., LTD.

Dates

Publication Date: 20260505
Application Date: 20240529
Priority Date: 20230922

Claims (10)

1 . An algorithm operation management apparatus, comprising: one or more processors; and a memory storing logic commands, when executed by the one or more processors, causing the one or more processors to: perform driving control of a vehicle by performing at least one algorithm, wherein performance of the at least one algorithm is determined based on an operating parameter applied to the at least one algorithm; learn by classifying an accident situation and a safety situation according to a plurality of driving conditions; adjust the operating parameter for the at least one algorithm on the basis of the learning result according to a current driving condition; and perform the driving control of the vehicle by performing the at least one algorithm to which the adjusted operating parameter is applied to.
2 . The algorithm operation management apparatus of claim 1 , wherein, when the learning result corresponds to the accident situation, the one or more processors are configured to adjust the operating parameter to increase or decrease operation performance of at least one algorithm related to the accident situation.
3 . The algorithm operation management apparatus of claim 1 , wherein the one or more processors are configured to adjust an operating parameter for at least one of the remaining algorithms of the at least one algorithm according to adjustment of an operating parameter for any one of the at least one algorithm.
4 . The algorithm operation management apparatus of claim 1 , wherein the operating parameter includes at least one among a processor occupation time, an occupied resource, an operating cycle, a processing resolution, a complexity level of the at least one algorithm and an operation priority between the at least one algorithm.
5 . The algorithm operation management apparatus of claim 1 , wherein the one or more processors are configured to determine the plurality of driving conditions on the basis of vehicle information for a vehicle location, an external environment, an internal environment, and a driving status.
6 . The algorithm operation management apparatus of claim 1 , wherein the at least one algorithm includes at least one among a driving lane detection algorithm, a surrounding object detection algorithm, a parking line detection algorithm, a surrounding vehicle detection algorithm, and a space detection algorithm.
7 . The algorithm operation management apparatus of claim 1 , wherein, in a state before the operating parameter for the at least one algorithm is set, the one or more processors are configured to set the operating parameter for the at least one algorithm on the basis of the current driving condition and the learning result.
8 . The algorithm operation management apparatus of claim 1 , wherein the one or more processors are configured to perform the learning by receiving information corresponding to an accident situation and a safety situation in the same driving condition.
9 . The algorithm operation management apparatus of claim 1 , wherein the one or more processors are configured to derive a determination result on the basis of a verification result of the learning result.
10 . An algorithm operation management method, comprising: performing driving control of a vehicle by performing at least one algorithm, wherein performance of the at least one algorithm is determined based on an operating parameter applied to the at least one algorithm; learning, by the one or more processors, by classifying an accident situation and a safety situation according to a plurality of driving conditions; adjusting, by the one or more processors, the operating parameter for the at least one algorithm on the basis of the learning result according to a current driving condition; and performing, by the one or more processors, the driving control of the vehicle by performing the at least one algorithm to which the adjusted operating parameter is applied to.

Description

CROSS REFERENCE TO RELATED APPLICATION The present application claims priority of Korean Patent Application No. 10-2023-0127203 filed on Sep. 22, 2023, the entire contents of which is incorporated herein for all purposes by this reference. BACKGROUND OF THE DISCLOSURE Field of the Disclosure The present disclosure relates to an algorithm operation management apparatus and method that allows algorithms used to assist vehicle driving to have optimal reliability and performance. Description of the Related Art Recently, among vehicle-related technologies, there is a growing interest on autonomous driving vehicles with unmanned driving functions that drive automatically without driver intervention. In autonomous driving vehicles, information necessary for driving, such as a shape of a road and obstacles, can be identified by analyzing data obtained through various sensors such as satellite navigation devices, inertial navigation devices, radio detecting and ranging (RADAR), ultrasonic measuring devices, laser scanners, and cameras, which are mounted in the vehicles. In addition, in autonomous driving vehicles, driving control can be performed, such as avoiding obstacles, by controlling a steering and a speed of the vehicle according to the information such as the identified road shape and obstacles. Various algorithms can be applied to such autonomous driving vehicles to acquire information necessary for driving and perform driving control. In such autonomous driving vehicles, scheduling between algorithms can be performed according to priorities set between the algorithms and performance of the algorithms. The foregoing is intended merely to aid in the understanding of the background of the present disclosure, and is not intended to mean that the present disclosure falls within the purview of the related art that is already known to those skilled in the art. SUMMARY OF THE DISCLOSURE Accordingly, the present disclosure has been made keeping in mind the above problems occurring in the related art, and the present disclosure is intended to provide an algorithm operation management apparatus and method that allows algorithms for vehicle driving to have optimal performance and reliability by adjusting operating parameters of the algorithms through learning results for a plurality of driving conditions. It should be noted that objects of the present disclosure are not limited to the above-described objects, and other objects of the present disclosure will be apparent to those skilled in the art from the following descriptions. According to one aspect, there is provided an algorithm operation management apparatus including: a learning part configured to learn by classifying an accident situation and a safety situation according to a plurality of driving conditions; a determination part configured to adjust an operating parameter for at least one algorithm related to driving control of a vehicle on the basis of a determination result of the learning part according to a current driving condition; and an operation part configured to perform the driving control of the vehicle by applying the adjusted operating parameter on the basis of the at least one algorithm. As one example, when the determination result of the learning part corresponds to the accident situation, the determination part may adjust the operating parameter to increase or decrease operation performance of at least one algorithm related to the accident situation. As one example, the determination part may adjust an operating parameter for at least one of the remaining algorithms of the at least one algorithm according to adjustment of an operating parameter for any one of the at least one algorithm. As one example, the operating parameter may include at least one among a processor occupation time, an occupied resource, an operating cycle, a processing resolution, a complexity level of the at least one algorithm and an operation priority between the at least one algorithm. As one example, the plurality of driving conditions may be determined on the basis of vehicle information for a vehicle location, an external environment, an internal environment, and a driving status. As one example, the at least one algorithm may include at least one among a driving lane detection algorithm, a surrounding object detection algorithm, a parking line detection algorithm, a surrounding vehicle detection algorithm, and a space detection algorithm. As one example, in a state before the operating parameter for the at least one algorithm is set, the determination part may set the operating parameter for the at least one algorithm on the basis of the current driving condition and the determination result of the learning part. As one example, the learning part may perform the learning by receiving information corresponding to an accident situation and a safety situation in the same driving condition. As one example, the learning part may derive the determination result on the ba