CN-122024466-A - Vehicle traffic state assessment method and system based on millimeter wave radar

CN122024466ACN 122024466 ACN122024466 ACN 122024466ACN-122024466-A

Abstract

The invention discloses a vehicle traffic state assessment method and system based on millimeter wave radar, wherein the vehicle traffic state assessment method comprises the steps of obtaining a starting state and remaining green time of each lane, correcting the lane to be a green light starting passing state if a lane state correction condition is met in the green light starting passing state, judging the lane state step by step under the green light starting passing state and correspondingly correcting the starting state of the lane, traversing all the lanes to complete the correction of the starting state and the remaining green time of the lane, and assessing the starting state of a queuing vehicle based on the starting state of the nearest preceding vehicle of the queuing vehicle and the starting state of the corresponding lane. According to the method, the vehicle starting state is evaluated by combining the lane starting state and the vehicle queuing situation in the lane, when a certain lane is in a green light starting passing state, the queued vehicles can automatically start to travel after the starting point cloud is not matched for a long time, and the situation that the queued vehicles always wait in situ and are collided, rolled and deleted by other vehicles is avoided.

Inventors

ZHOU JUNJIE
MA ZHENGYANG
QI ZHAOGE
YUAN XUEFENG
JU BOWEN

Assignees

浙江中控信息产业股份有限公司

Dates

Publication Date: 20260512
Application Date: 20241112

Claims (10)

1. A millimeter wave radar-based vehicle traffic state evaluation method, characterized by comprising: Acquiring starting states and remaining green light time of each lane; if the lane is not in the green light starting traffic state, the lane is corrected to the green light starting traffic state, and the lane is in the green light starting traffic state, the step-by-step condition judgment of the lane state is carried out and the starting state of the lane is correspondingly corrected; and estimating the starting state of the queuing vehicle based on the starting state of the nearest preceding vehicle of the queuing vehicle and the starting state of the corresponding lane.
2. The millimeter wave radar-based vehicle traffic state evaluation method according to claim 1, wherein the stepwise condition determination comprises: Firstly, judging the state of a first-stage lane, if yes, updating and prolonging the remaining green light time, otherwise, not correcting; a second level lane state determination is then made, If the traffic signal is not met, the green light threshold value is compared, and if the remaining green light time is greater than the green light threshold value, the green light is not corrected, otherwise, the traffic signal is corrected to be not in the green light starting traffic state.
3. The millimeter wave radar-based vehicle traffic state evaluation method according to claim 1 or 2, wherein the lane state correction condition includes: The head car of the lane is in a starting state and is in a queuing state at the last moment, or The other queuing vehicles except the head vehicle of the lane are matched with the starting point cloud; And if the lane state correction condition is met, correcting to be a green light starting traffic state, and updating the remaining green light time.
4. The millimeter wave radar-based vehicle traffic state evaluation method according to claim 2, wherein the first-stage lane state determination comprises: The head car of the lane has traveled above a first distance threshold from its own queuing position, or The speed of the lead vehicle of the lane in the direction of extension of the road has currently reached above the first speed threshold.
5. The vehicle traffic state evaluation method based on millimeter wave radar according to claim 2 or 4, wherein the second-stage lane state determination includes: The head car of the lane has traveled above a first distance threshold from its own queuing position, or At least one of the other vehicles of the lane except the head vehicle is in a starting state.
6. The millimeter wave radar-based vehicle traffic state evaluation method according to claim 5, wherein the third-stage lane state determination comprises: The first vehicle distance of the lane is within a second distance threshold, the absolute speed of the first vehicle is smaller than the second speed threshold, the first vehicle is lost for more than the time point cloud, and the first vehicle never enters the queuing state, or The distance of the first vehicle from the queuing and stopping line of the lane is within a second distance threshold, the absolute speed of the first vehicle is smaller than the second speed threshold, the first vehicle is lost for more than the time point cloud, and the first vehicle enters a queuing state.
7. A vehicle traffic state evaluation method based on millimeter wave radar according to claim 1 or 2 or 4 or 6, wherein the evaluation of the starting state of the queuing vehicle comprises: for the head vehicle of the lane, if the head vehicle is matched with the starting point cloud, judging the head vehicle as a starting state, otherwise, keeping a queuing state; and for other queuing vehicles except the head vehicle, estimating and obtaining the starting state of the queuing vehicle based on the starting state of the nearest preceding vehicle of the queuing vehicle and the starting state of the corresponding lane.
8. The vehicle traffic state evaluation method based on millimeter wave radar according to claim 7, wherein the queuing state is maintained when the lane corresponding to the queuing vehicle is not in a green light starting traffic state; when the corresponding lane is in a green light starting and passing state, If the vehicle does not enter the starting state recently, keeping a queuing state; if the nearest preceding vehicle enters a starting state and the queuing vehicle is matched with the starting point cloud, entering the starting state; if the nearest preceding vehicle enters a starting state but the queuing vehicle does not match the starting point cloud, entering the starting state under the condition that the preset condition is met after the nearest preceding vehicle starts.
9. A millimeter wave radar-based vehicle traffic state evaluation system adapted to the vehicle traffic state evaluation method according to any one of claims 1 to 8, comprising: the lane state acquisition module acquires the starting state and the remaining green light time of each lane; the lane state correction module corrects the starting state and the remaining green time of the lane acquired by the lane state acquisition module; The vehicle state evaluation module is used for evaluating the starting state of the vehicle by combining the starting state of the nearest preceding vehicle based on the corrected starting state of the lane and the residual green time.
10. The millimeter wave radar-based vehicle traffic state assessment system according to claim 9, wherein said lane state correction module comprises: The state correction judging unit is used for judging and correcting the lanes which are not in the green light starting traffic state; the first-level condition judging unit judges and corrects the lane in the green light starting traffic state; a second-level condition determination unit that determines a lane passing through the first-level condition determination unit and then enters the lane; a third-level condition determination unit that determines whether or not the start state of the lane needs to be corrected according to a third-level lane state determination condition; And the green light threshold judging unit judges whether the starting state of the lane needs to be corrected according to the remaining green light time.

Description

Vehicle traffic state assessment method and system based on millimeter wave radar Technical Field The invention relates to the technical field of intelligent traffic, in particular to a vehicle traffic state evaluation method and system based on millimeter wave radar. Background The traffic millimeter wave radar back-end processing algorithm is mainly divided into a clustering algorithm and a tracking algorithm, wherein the clustering algorithm is responsible for identifying and clustering scattered radar point clouds into individual vehicle target clusters, and the tracking algorithm predicts and tracks the clustering points of the vehicle point clouds on the basis of a clustering result. In urban roads, vehicles stop on road sections for waiting and queuing due to the fact that red lights are turned on, the state of waiting and queuing is changed into the state of starting and passing due to the fact that green lights are turned on, the state of starting and passing is displayed on radar point clouds, namely the point clouds of the vehicles disappear on the road sections when the red lights are turned on, and the point clouds reappear on the road sections when the green lights are turned on. In order to maintain the consistency of the id of the vehicle, when the vehicle stops due to the lighting of a red light and the point cloud disappears, a queuing state is given to the current vehicle target, at the moment, the vehicle target can stop in place in the queuing state, wait for the occurrence of the starting point cloud of the vehicle target, then transition from the queuing state to the starting state along with the starting point cloud of the vehicle target, and advance until the vehicle enters an intersection and the point cloud disappears. However, in an actual starting scene, because the speed of the vehicle is lower when the vehicle starts, and the front and rear of the queuing vehicle are blocked, the point cloud of the starting vehicle detected by the radar is sparse or the point cloud of part of the starting vehicle cannot be detected at all. This results in some of the in-line vehicles always waiting in place for an unsuccessful launch, and then being bumped by other vehicles and then erased as they approach the in-line vehicle location. The vehicle motion state estimation method based on the traffic millimeter wave radar disclosed in the China patent literature is disclosed as CN115657002A, and the publication date is 2023.01.31, and comprises the steps of deploying the traffic millimeter wave radar on a road side to collect track data of a vehicle in different motion states, constructing a vehicle motion state data set, training a support vector machine model by using the vehicle motion state data set, classifying the vehicle motion states by using the trained support vector machine model, constructing a hidden Markov chain model according to a transition relation between the motion states, and estimating the vehicle motion states. The traffic millimeter wave radar used by the technology has low cost, is suitable for various bad weather and various traffic scenes, and can be used for monitoring and early warning the running behavior of the vehicle by constructing a large number of data sets to train a support vector machine model and greatly improving the accuracy of the classification and estimation of the motion state. However, the technology is suitable for evaluating the state of the vehicle under the condition of a certain speed, and for the judgment that the vehicle is in a queuing state in situ due to the stop of the red light or the vehicle is in a starting state, accurate evaluation and judgment still cannot be performed due to the lack of radar point cloud data. Disclosure of Invention The invention provides a vehicle traffic state evaluation method and system based on millimeter wave radar, which aims to solve the problems that in the prior art, because the speed of a vehicle is lower when the vehicle starts, and the front and rear parts of the queuing vehicles are shielded, the point cloud of the starting vehicle detected by a radar is sparse or the point cloud of part of the starting vehicles cannot be detected at all, so that part of the queuing vehicles cannot update the starting state and are crashed and deleted by the following vehicles. In order to achieve the above purpose, the present invention adopts the following technical scheme: A millimeter wave radar-based vehicle traffic state assessment method, comprising: Acquiring starting states and remaining green light time of each lane; if the lane is not in the green light starting traffic state, the lane is corrected to the green light starting traffic state, and the lane is in the green light starting traffic state, the step-by-step condition judgment of the lane state is carried out and the starting state of the lane is correspondingly corrected; and estimating the starting state of the queuing vehicle based on the startin