CN-116108612-B - Method and device for identifying lane change behavior, storage medium and computer equipment

CN116108612BCN 116108612 BCN116108612 BCN 116108612BCN-116108612-B

Abstract

The application provides a lane change behavior recognition method, a lane change behavior recognition device, a storage medium and computer equipment, according to the position and lane change state of a host vehicle corresponding to each automatic driving moment, A first confidence level of a first lane to which the host vehicle belongs at each automatic driving moment is determined to reflect the possibility that the host vehicle changes lanes from the first lane to other lanes. And determining a second degree of confidence of the second lane at each automatic driving moment according to the position and the lane change state of the host vehicle corresponding to each automatic driving moment so as to determine the possibility that the host vehicle changes lanes from other lanes to the second lane. The computer equipment can identify lane changing behaviors of the host vehicle in the automatic driving process according to the first confidence level and the second confidence level of each lane at each automatic driving time. Therefore, misjudgment caused by excessive dependence on the internal state of the planning control algorithm module can be corrected, so that lane changing behavior in the automatic driving process can be more accurately identified, and the identification accuracy is improved.

Inventors

Pan Huangkai
WU JIACHEN
ZHENG ZIWEI

Assignees

广州文远知行科技有限公司

Dates

Publication Date: 20260508
Application Date: 20221031

Claims (11)

1. A method of identifying a lane change behavior, the method comprising: acquiring a main vehicle position corresponding to each automatic driving moment and a first lane to which the main vehicle belongs; for each automatic driving moment, determining a first confidence level corresponding to the first lane at the automatic driving moment according to the position of the host vehicle corresponding to the automatic driving moment and the lane changing state of the host vehicle at the automatic driving moment, wherein the first confidence level is used for reflecting the possibility that the host vehicle changes lanes from the first lane to other lanes except the first lane; For each automatic driving moment, if at least part of the host vehicle runs in a second lane at the automatic driving moment, determining a second confidence level corresponding to the second lane at the automatic driving moment according to the host vehicle position corresponding to the automatic driving moment and the lane changing state of the host vehicle at the automatic driving moment, wherein the second confidence level is used for reflecting the possibility that the host vehicle changes lanes from other lanes except the second lane to the second lane; For each automatic driving moment, if the first self-confidence level of the first lane corresponding to the automatic driving moment is greater than or equal to a first preset self-confidence level threshold value, and the second self-confidence level of the second lane adjacent to the first lane corresponding to the automatic driving moment is greater than or equal to a second preset threshold value, determining that the main vehicle has a lane change behavior at the automatic driving moment.
2. The lane change behavior recognition method according to claim 1, wherein the step of determining the first confidence level of the first lane corresponding to the automatic driving time according to the position of the host vehicle corresponding to the automatic driving time and the lane change state of the host vehicle at the automatic driving time includes: according to the position of the main vehicle corresponding to the automatic driving moment and the lane central line position of the first lane, calculating a first distance between the main vehicle and the first lane at the automatic driving moment, and determining a first fraction based on the first distance; If the lane change state of the host vehicle at the automatic driving moment is a request lane change state, determining a first state duration of the request lane change state up to the automatic driving moment, determining a second score based on the first state duration, and determining a first confidence corresponding to the first lane at the automatic driving moment according to the first score and the second score; And if the lane change state of the host vehicle at the automatic driving moment is not the request lane change state, determining a first confidence degree corresponding to the first lane at the automatic driving moment according to the first score.
3. The method of claim 2, wherein the step of determining a first score based on the first spacing comprises: calculating a first ratio between the first distance and a lane width of the first lane; Processing the first ratio to determine the first score based on the following expression: ln(1+(e-1)·a) Wherein e is a natural constant and a is the first ratio.
4. The lane change behavior recognition method according to claim 1, wherein the step of determining the second confidence level of the second lane corresponding to the automatic driving time according to the position of the host vehicle corresponding to the automatic driving time and the lane change state of the host vehicle at the automatic driving time comprises: Calculating a second distance between the main vehicle and the second vehicle lane at the automatic driving moment according to the main vehicle position corresponding to the automatic driving moment and the lane central line position of the second lane, and determining a third score based on the second distance; If the lane change state of the host vehicle at the automatic driving moment is an ending lane change state, determining a second state duration of the ending lane change state at the automatic driving moment, determining a fourth score based on the second state duration, and determining a second confidence level corresponding to the second lane at the automatic driving moment according to the third score and the fourth score; And if the lane change state of the host vehicle at the automatic driving moment is not the ending lane change state, determining a second degree of confidence corresponding to the second lane at the automatic driving moment according to the third score.
5. The method of claim 4, wherein the step of determining a third score based on the second pitch comprises: Calculating a second ratio between a second distance between the main vehicle and the second vehicle lane at the automatic driving moment and a lane width of the second lane, and determining a fifth fraction based on the second ratio; Judging whether the second distance between the main vehicle and the second lane is unchanged at a plurality of automatic driving moments according to the second distance between the main vehicle and the second lane at least one front automatic driving moment and the second distance between the main vehicle and the second lane at the automatic driving moment, wherein each front automatic driving moment is earlier than the automatic driving moment; If a second distance between the main vehicle and the second vehicle road is kept unchanged at a plurality of automatic driving moments, determining a distance duration time for which the second distance is kept unchanged up to the automatic driving moment, determining a sixth score based on the distance duration time, and determining the third score according to the fifth score and the sixth score; And if the second distance between the host vehicle and the second vehicle at the automatic driving moment is different from the second distance between the host vehicle and the second vehicle at the previous automatic driving moment, taking the fifth score as the third score.
6. The method of claim 5, wherein the sixth score is correlated positively with the distance duration, and wherein the step of determining the third score based on the fifth score and the sixth score comprises: and if the sixth score is greater than or equal to a preset score threshold value, taking the sum of the fifth score and the sixth score as the third score, and otherwise, taking the fifth score as the third score.
7. The lane change behavior recognition method according to claim 1, wherein the method further comprises: For each automatic driving moment with lane change behaviors, calculating the product between the first confidence level corresponding to the first lane at the automatic driving moment and the second confidence level corresponding to the second lane at the automatic driving moment; And taking the lane change behavior corresponding to the product with the largest numerical value in the products as a target lane change behavior, and respectively obtaining the lane change starting time and the lane change ending time corresponding to the target lane change behavior.
8. The lane change behavior recognition method according to claim 7, wherein the step of obtaining the lane change start time and the lane change end time corresponding to the target lane change behavior, respectively, comprises: Determining a target third lane and a target fourth lane corresponding to the target lane change behavior; In each target first confidence level corresponding to the target third lane, determining a first target first confidence level greater than or equal to the first preset confidence level threshold according to the time sequence of each automatic driving moment, and taking the automatic driving moment corresponding to the first target first confidence level as a lane change starting moment corresponding to the target lane change behavior; and determining the maximum second confidence level in each second confidence level corresponding to the target fourth lane, and taking the automatic driving time corresponding to the maximum second confidence level as the lane change ending time corresponding to the target lane change behavior.
9. A lane change behavior recognition apparatus, the apparatus comprising: the position acquisition module is used for acquiring the position of the main vehicle corresponding to each automatic driving moment and a first lane to which the main vehicle belongs; The first confidence acquiring module is used for determining, for each automatic driving moment, a first confidence corresponding to the first lane at the automatic driving moment according to the position of the host vehicle corresponding to the automatic driving moment and the lane changing state of the host vehicle at the automatic driving moment, wherein the first confidence is used for reflecting the possibility that the host vehicle changes lanes from the first lane to other lanes except the first lane; The second confidence acquiring module is configured to determine, for each of the autopilot moments, a second confidence corresponding to the second lane at the autopilot moment according to the position of the host vehicle corresponding to the autopilot moment and a lane change state of the host vehicle at the autopilot moment if at least a portion of the host vehicle runs in the second lane at the autopilot moment, where the second confidence is configured to reflect a possibility that the host vehicle changes lanes from lanes other than the second lane to the second lane; The identification module is used for determining that the main vehicle has a lane change behavior at the automatic driving moment if the first confidence level of the first lane at the automatic driving moment is greater than or equal to a first preset confidence threshold value and the second confidence level of the second lane adjacent to the first lane at the automatic driving moment is greater than or equal to a second preset threshold value.
10. A storage medium having stored therein computer readable instructions which, when executed by one or more processors, cause the one or more processors to perform the steps of the lane change behavior identification method of any one of claims 1 to 8.
11. A computer device includes one or more processors and a memory; stored in the memory are computer readable instructions which, when executed by the one or more processors, perform the steps of the lane change behavior identification method of any one of claims 1 to 8.

Description

Method and device for identifying lane change behavior, storage medium and computer equipment Technical Field The present application relates to the field of autopilot, and in particular, to a method and apparatus for identifying lane change behavior, a storage medium, and a computer device. Background In the automatic driving process, lane change behavior is an important link for evaluating the planning rationality of an automatic driving algorithm, so that when the automatic driving algorithm is subjected to simulation test, the lane change behavior appearing during the simulation test needs to be identified, and the identified lane change behavior is scored to determine the rationality and safety of the lane change behavior. Currently, the prior art only judges whether the main vehicle has a lane change behavior based on the information generated by the planning control algorithm module of the main vehicle. However, during traveling, the host vehicle may go away from the lane center line to detour or overrun the obstacle, or travel to the road bifurcation or merge, which may be erroneously identified as a lane change. Therefore, if the data in the planning control algorithm module is excessively relied on to identify the lane change behavior, erroneous judgment is likely to occur, and the problem of low identification accuracy exists. Disclosure of Invention The object of the present application is to solve at least one of the above-mentioned technical drawbacks, in particular the technical drawbacks of the prior art with low recognition accuracy. In a first aspect, an embodiment of the present application provides a method for identifying a lane change behavior, where the method includes: acquiring a main vehicle position corresponding to each automatic driving moment and a first lane to which the main vehicle belongs; for each automatic driving moment, determining a first confidence level corresponding to the first lane at the automatic driving moment according to the position of the host vehicle corresponding to the automatic driving moment and the lane changing state of the host vehicle at the automatic driving moment, wherein the first confidence level is used for reflecting the possibility that the host vehicle changes lanes from the first lane to other lanes except the first lane; For each automatic driving moment, if at least part of the host vehicle runs in a second lane at the automatic driving moment, determining a second confidence level corresponding to the second lane at the automatic driving moment according to the host vehicle position corresponding to the automatic driving moment and the lane changing state of the host vehicle at the automatic driving moment, wherein the second confidence level is used for reflecting the possibility that the host vehicle changes lanes from other lanes except the second lane to the second lane; and identifying lane change behaviors of the host vehicle in the automatic driving process according to the first confidence level and the second confidence level corresponding to each lane at each automatic driving time. In one embodiment, the step of determining the first confidence level of the first lane corresponding to the automatic driving time according to the position of the host vehicle corresponding to the automatic driving time and the lane changing state of the host vehicle at the automatic driving time includes: according to the position of the main vehicle corresponding to the automatic driving moment and the lane central line position of the first lane, calculating a first distance between the main vehicle and the first lane at the automatic driving moment, and determining a first fraction based on the first distance; If the lane change state of the host vehicle at the automatic driving moment is a request lane change state, determining a first state duration of the request lane change state up to the automatic driving moment, determining a second score based on the first state duration, and determining a first confidence corresponding to the first lane at the automatic driving moment according to the first score and the second score; And if the lane change state of the host vehicle at the automatic driving moment is not the request lane change state, determining a first confidence degree corresponding to the first lane at the automatic driving moment according to the first score. In one embodiment, the step of determining a first score based on the first pitch includes: calculating a first ratio between the first distance and a lane width of the first lane; Processing the first ratio to determine the first score based on the following expression: ln(1+(e-1)·a) Wherein e is a natural constant and a is the first ratio. In one embodiment, the step of determining the second confidence level of the second lane corresponding to the automatic driving time according to the position of the host vehicle corresponding to the automatic driving time and the lan