Toll station anti-collision early warning method and system based on motion trail prediction

Abstract

The invention discloses a toll station anti-collision early warning method and system based on motion trail prediction, and belongs to the technical field of traffic management, and the method comprises the steps: obtaining the information of a moving vehicle through a collection unit disposed in a square; based on the moving vehicle information, analyzing to obtain a motion trail prediction value; and executing corresponding early warning measures based on the motion trail predicted value and the square grid. The mobile vehicle information is acquired through the acquisition unit arranged in the square, the square can provide an appropriate and existing data acquisition space, and enough mobile vehicle information can be acquired for subsequent judgment; based on the moving vehicle information, analyzing to obtain a motion trail prediction value, and accurately describing a potential risk source; and corresponding early warning measures are executed based on the motion trail predicted value and the square grid, so that targeted measures can be executed, and the defense efficiency is improved.

Description

Technical Field

[0001] This invention relates to the field of traffic management technology, and in particular to a method and system for collision avoidance warning at toll stations based on motion trajectory prediction. Background Technology

[0002] Toll booths, such as those on highways, are located at key road junctions and traffic points. Because the toll collection process takes a while, congestion can occur in the relevant areas. Furthermore, if an accident occurs, it can lead to further traffic congestion and potentially trigger a chain reaction of accidents.

[0003] Among the types of accidents, the most common are collisions between vehicles and various facilities and equipment at toll stations, as well as collisions with other vehicles. Existing toll stations mostly adopt passive protection measures, such as various buffers and fixed objects. These measures can only be controlled after the fact, so the defense efficiency is not high. Summary of the Invention

[0004] This invention proposes a collision avoidance early warning method and system for toll stations based on motion trajectory prediction, in order to solve the problem of low efficiency of existing defense methods.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0006] A collision avoidance warning method for toll stations based on motion trajectory prediction includes: acquiring moving vehicle information through a data acquisition unit set up in the plaza; parsing the moving vehicle information to obtain a motion trajectory prediction value; and executing corresponding warning measures based on the motion trajectory prediction value and the plaza grid.

[0007] Furthermore, the mobile vehicle information includes vehicle model, size, license plate, weight, movement data, and personnel information; correspondingly, the acquisition unit includes a camera, a weighbridge, an RFID tag, and a speed measuring radar.

[0008] Furthermore, the step of parsing the motion trajectory prediction value based on the moving vehicle information includes: determining the distance the vehicle moves by setting up reference points in the square, and calculating the motion data by combining the image acquisition time; acquiring the motion data through the speed measuring radar; identifying the individual information of the object through the radio frequency tag and matching it with the motion data, wherein the individual information includes the vehicle model, the license plate, and the personnel information; obtaining the vehicle's weight information through the weighbridge; processing the motion data to obtain an initial motion trajectory prediction value; and modifying the initial motion trajectory prediction value based on the individual information and / or the weight information to obtain the final motion trajectory prediction value.

[0009] Furthermore, the step of processing the moving data to obtain the initial motion trajectory prediction value includes: processing the moving data based on a b-spline curve fitting algorithm to obtain the initial motion trajectory prediction value.

[0010] Furthermore, the B-spline curve fitting algorithm for processing the moving data includes: data preprocessing: using Kalman filtering or moving average trajectory to process the original trajectory point sequence of the moving data to obtain a smoothed trajectory point sequence; parameterization and node vector generation: combining the B-spline order and the trajectory point sequence to output the corresponding node vector; B-spline fitting: selecting the order, calculating the node vector, and solving for the control points using the least squares method; trajectory prediction: extending the prediction from the current parameter value, extending the node vector, extrapolating the B-spline curve, calculating future time points, and solving for the control points.

[0011] Furthermore, the step of modifying the initial motion trajectory prediction value based on the individual information and / or the weight information to obtain the motion trajectory prediction value includes: obtaining the corresponding historical driving record based on the individual information; analyzing the historical driving record, and if there is an abnormal driving record, increasing the initial motion trajectory prediction value, and if there is no abnormal driving record, maintaining the initial motion trajectory prediction value; and if the weight information is greater than a specified threshold, maintaining the initial motion trajectory prediction value.

[0012] Furthermore, the plaza grid obtained by gridding the toll station area is configured with corresponding risk levels; correspondingly, the early warning measures based on the predicted motion trajectory value and the plaza grid include: determining overlapping grids based on the predicted motion trajectory value, and executing corresponding early warning measures based on the risk level, overlap probability, and overlap time of the overlapping grids.

[0013] Furthermore, the early warning measures include: activating audible and visual alarm devices, activating barrier devices, storing data of moving targets, and interacting with targets via data.

[0014] Furthermore, the step of implementing corresponding early warning measures based on the risk level, overlap probability, and overlap time of overlapping grids includes: controlling the signal strength of the audible and visual alarm device, the start-up speed of the barrier device, prioritizing data storage and / or deletion, and prioritizing data interaction based on the risk level; determining the audible and visual alarm device and the barrier device that need to be controlled based on the overlap probability; and pre-controlling the audible and visual alarm device and the barrier device based on the overlap time.

[0015] A toll station collision avoidance warning system based on motion trajectory prediction includes: a first module for acquiring moving vehicle information through a collection unit set in the plaza; a second module for parsing the moving vehicle information to obtain motion trajectory prediction values; and a third module for executing corresponding warning measures based on the motion trajectory prediction values ​​and the plaza grid.

[0016] By adopting the above technical solution, the present invention has the following beneficial effects:

[0017] 1. This invention acquires mobile vehicle information by setting up a collection unit in a square. The square provides a suitable and existing data collection space, enabling the collection of sufficient mobile vehicle information for subsequent judgment. Based on the mobile vehicle information, a motion trajectory prediction value is obtained, which can accurately describe potential risk sources. Based on the motion trajectory prediction value and the square grid, corresponding early warning measures are executed, enabling targeted measures and improving the efficiency of defense. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of a toll station collision avoidance early warning method based on motion trajectory prediction proposed in this invention. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] like Figure 1 The method for collision avoidance warning at toll stations based on motion trajectory prediction includes: S1, acquiring moving vehicle information through a collection unit set up in the plaza; S2, parsing the moving vehicle information to obtain motion trajectory prediction values; S3, executing corresponding warning measures based on the motion trajectory prediction values ​​and the plaza grid.

[0021] Toll booths are typically designed in a diamond shape, with the main structure resembling a strip. Small plazas are located on the outer side of each booth to accommodate vehicles slowing down or stopping. In addition to the buildings, toll booths also house numerous pieces of equipment. These structures and equipment are easily damaged by vehicle impacts, disrupting the normal operation of the toll booth.

[0022] Because the square has a relatively large area, it is possible to install large and numerous devices. Generally speaking, larger devices have stronger performance, and more devices have a wider coverage. By setting up acquisition units in the square to acquire information about targets within the square, information about key targets, namely moving vehicles, can be obtained.

[0023] Moving vehicle information comprises information about the vehicle itself and from external observations. It describes or defines various attributes of the vehicle, which can be used to predict the risks it poses to toll booths. Specifically, it analyzes the vehicle's future trajectory, determines whether the trajectory will pose a threat to the toll booth, and the trajectory-related data constitutes the predicted trajectory value.

[0024] To more accurately describe the plaza's situation, a grid is created based on area and location to depict the plaza's area, coordinates, and its surrounding structures—this is the plaza grid. By analyzing the relationship between predicted motion trajectories and the plaza grid, existing or temporary early warning measures can be implemented, allowing for defensive actions to mitigate risks based on specific circumstances.

[0025] The mobile vehicle information includes vehicle model, size, license plate, weight, movement data, and personnel information; correspondingly, the acquisition unit includes a camera, a weighbridge, an RFID tag, and a speed measuring radar.

[0026] Vehicle model refers to the type of vehicle itself, usually provided by the manufacturer. It can be obtained through a trained appearance recognition model. The vehicle model provides a rough estimate of its specifications and motion parameters (i.e., the parameters described by the manufacturer). The greater the difference between the actual motion parameters and the specified motion parameters, the greater the risk. Volume can be obtained through direct observation or indirectly through the vehicle model. Larger vehicles pose a greater risk of colliding with other objects. License plates can be used to identify and locate vehicles and, under legal authorization, to obtain driver information to determine if the driver intends to cause a collision (a large number of traffic violations increases the risk). Motion data includes the vehicle's coordinates, speed, and direction. Personnel information refers to the actual driver or vehicle owner. The risk is greater if the person is frequently involved in traffic accidents or is a new driver.

[0027] The step of parsing the moving vehicle information to obtain the predicted motion trajectory value includes: determining the distance the vehicle traveled by setting up reference points in the square, and calculating the motion data by combining the image acquisition time; acquiring the motion data by the speed measuring radar; identifying the individual information of the object by the radio frequency tag and matching it with the motion data, wherein the individual information includes the vehicle model, the license plate, and the person information; obtaining the vehicle's weight information by the weighbridge; processing the motion data to obtain an initial motion trajectory prediction value; and modifying the initial motion trajectory prediction value based on the individual information and / or the weight information to obtain the final motion trajectory prediction value.

[0028] The plaza can be equipped with bricks of different colors, markers, or architectural structures as reference points. Since the coordinates and distances of these reference points can be measured in advance, the movement data can be calculated by combining the image acquisition time and the distance between the target and the reference points. Based on this movement data, an initial predicted movement trajectory value is obtained. Simultaneously, based on the individual information and / or the weight information, the initial predicted movement trajectory value is modified to obtain the final predicted movement trajectory value. This further refines the predicted movement trajectory value to predict risk.

[0029] The step of processing the movement data to obtain the initial motion trajectory prediction value includes: processing the movement data based on a b-spline curve fitting algorithm to obtain the initial motion trajectory prediction value.

[0030] The B-spline curve fitting algorithm processes the moving data, including: data preprocessing: using Kalman filtering or moving average trajectory to process the original trajectory point sequence of the moving data to obtain a smoothed trajectory point sequence; parameterization and node vector generation: combining the B-spline order and the trajectory point sequence to output the corresponding node vector; B-spline fitting: selecting the order, calculating the node vector, and solving for the control points using the least squares method; trajectory prediction: extending the prediction from the current parameter value, extending the node vector, extrapolating the B-spline curve, calculating future time points, and solving for the control points; where the B-spline curve is defined as: P(u) = Σ_{i=0}^n N_{i,p}(u)Q_i; where: Q_i is the control point, N_{i,p}(u) is the p-th order B-spline basis function, and u is the parameter variable; the basis function recursive formula is: N_{i,0}(u) = {1, if u_i ≤ u < u_{i+1} ;

[0031] {0, otherwise

[0032] N_{i,p}(u) = (u-u_i) / (u_{i+p}-u_i) N_{i,p-1}(u)

[0033] + (u_{i+p+1}-u) / (u_{i+p+1}-u_{i+1}) N_{i+1,p-1}(u).

[0034] B-spline fitting:

[0035] In vehicle trajectory prediction, B-spline functions provide smooth trajectory fitting and exhibit good local control characteristics. This algorithm uses B-spline functions to fit historical trajectory data and then predicts future trajectories based on the fitting results.

[0036] B-spline curves are defined as: P(u) = Σ_{i=0}^n N_{i,p}(u)Q_i; where: Q_i are control points; N_{i,p}(u) are p-th degree B-spline basis functions; u is a parameter variable;

[0037] Basis function recurrence formula:

[0038] N_{i,0}(u) = {1, if u_i ≤ u < u_{i+1};

[0039] {0, otherwise

[0040] N_{i,p}(u) = (u-u_i) / (u_{i+p}-u_i) N_{i,p-1}(u)

[0041] + (u_{i+p+1}-u) / (u_{i+p+1}-u_{i+1}) N_{i+1,p-1}(u);

[0042] Algorithm steps:

[0043] 1. Data Preprocessing

[0044] Preprocess the trajectory data using Kalman filtering or moving average.

[0045] Input: Original trajectory point sequence [(x1,y1,t1), (x2,y2,t2), ...]

[0046] Functions: 1. Outlier removal. 2. Data smoothing (optional). 3. Timestamp normalization.

[0047] Output: Smoothed sequence of trajectory points

[0048] 2. Parameterization and Node Vector Generation

[0049] Generate a uniform node vector input: a sequence of trajectory points, and the order of the B-spline.

[0050] Output: Node vector

[0051] 3.B Spline Fitting

[0052] Choose the order (usually 3rd order).

[0053] Calculate the node vector UU (uniform or chord-parameterized).

[0054] The control points PiPi are solved using the least squares method.

[0055] 4. Trajectory Prediction

[0056] Extend the prediction starting from the current parameter value

[0057] Extend the node vectors and extrapolate the B-spline curve.

[0058] Calculate the position corresponding to the future time point tN+1,...,tN+TtN+1,...,tN+T t)+ui+p+1−ui+1ui+1−tNi+1,p−1(t);

[0059] Control point solution (least squares method):

[0060] P∑k=1N∥C(tk)−(xk,yk)∥2Pmink=1∑N∥C(tk)−(xk,yk)∥2

[0061] The algorithm has a typical error of 0.3-1.2 meters (depending on road curvature) within a 50-meter prediction range, and a computation time of <10ms.

[0062] The step of modifying the initial motion trajectory prediction value based on the individual information and / or the weight information to obtain the motion trajectory prediction value includes: obtaining the corresponding historical driving record based on the individual information; analyzing the historical driving record, and if there is an abnormal driving record, increasing the initial motion trajectory prediction value, and if there is no abnormal driving record, maintaining the initial motion trajectory prediction value; and if the weight information is greater than a specified threshold, maintaining the initial motion trajectory prediction value.

[0063] Specifically, different individual information / weight information can be assigned corresponding modification parameters, including modifying the direction and magnitude of the predicted motion trajectory, which results in higher prediction accuracy.

[0064] In its simplest case, historical driving records primarily involve obtaining traffic violations or accident records when legally authorized / permitted. Clearly, a high number of violations or accidents suggests unstable driving, a high probability of altered driving trajectories, and significant changes in trajectory, thus requiring greater parameter modifications. For example, if driver A tends to veer to the left or accelerates suddenly, and there's a tollbooth on the left side of the vehicle, driver A is considered to pose a higher risk to the tollbooth. Without authorization, records of driving on highways can be used to determine driving patterns. For instance, consistently staying close to the lane line and maintaining a safe distance requires less parameter modification, while swerving left and right and frequently changing lanes necessitates greater parameter modifications.

[0065] Similarly, for weight information, the heavier the weight, the smaller the range of motion changes, and therefore the smaller the parameter modifications.

[0066] The plaza grid is obtained by gridding the toll station area, and each grid is set with a corresponding risk level. Correspondingly, the early warning measures are implemented based on the predicted motion trajectory value and the plaza grid, including: determining overlapping grids according to the predicted motion trajectory value, and implementing corresponding early warning measures according to the risk level, overlap probability and overlap time of the overlapping grids.

[0067] The plaza is considered a two-dimensional plane, and then divided into grids according to certain specifications (such as size, location, and attachments), forming a plaza grid. Attachments include buildings, equipment, structures, and personnel. Clearly, grids with attachments have a better ability to withstand risk than those without, resulting in a lower risk level. Conversely, grids with higher-value attachments have a weaker ability to withstand risk, resulting in a higher risk level.

[0068] The higher the risk level, the greater the probability of overlap, and the closer the overlap time, the more necessary it is to take preventative measures, or to implement higher-level preventative measures. The purpose is to determine whether to take preventative measures or strengthen preventative measures based on the likelihood of damage and the value of the loss.

[0069] The early warning measures include: activating audible and visual alarm devices, activating barrier devices, storing data on moving targets, and interacting with targets via data.

[0070] Audible and visual alarm devices, including horns and lights, can alert drivers or plaza staff. Barrier devices can be used to intercept or buffer vehicles. The various devices in the plaza continuously generate data; without data deletion and backup, data overflow will occur. By highlighting specific data, important data (i.e., accident data) can be saved for subsequent liability determination. Data interaction with targets—that is, direct data exchange between the plaza and vehicles—can be implemented, such as stopping vehicles in advance or obtaining data from accident-involved vehicles for subsequent accident analysis.

[0071] The step of implementing corresponding early warning measures based on the risk level, overlap probability, and overlap time of overlapping grids includes: controlling the signal strength of the audible and visual alarm device, the start-up speed of the barrier device, prioritizing data storage and / or deletion, and prioritizing data interaction based on the risk level; determining the audible and visual alarm device and the barrier device that need to be controlled based on the overlap probability; and pre-controlling the audible and visual alarm device and the barrier device based on the overlap time.

[0072] By implementing different early warning measures, defenses can be tailored to specific situations, reducing the need for such measures to minimize damage, lower operating costs, and improve early warning efficiency.

[0073] A toll station collision avoidance warning system based on motion trajectory prediction includes: a first module for acquiring moving vehicle information through a collection unit set in the plaza; a second module for parsing the moving vehicle information to obtain motion trajectory prediction values; and a third module for executing corresponding warning measures based on the motion trajectory prediction values ​​and the plaza grid.

[0074] The above description is a detailed description of the preferred embodiments of the present invention. However, the embodiments are not intended to limit the scope of the patent application of the present invention. All equivalent changes or modifications made under the technical spirit of the present invention should fall within the patent scope covered by the present invention.

Claims

1. A collision avoidance early warning method for toll stations based on motion trajectory prediction, characterized in that, include: Information on moving vehicles is obtained through data collection units set up in the square; Based on the mobile vehicle information, the predicted motion trajectory value is obtained through analysis. Based on the predicted motion trajectory and the square grid, corresponding early warning measures are implemented.

2. The toll station collision avoidance early warning method based on motion trajectory prediction according to claim 1, characterized in that, The mobile vehicle information includes vehicle model, size, license plate, weight, movement data, and personnel information; Correspondingly, the acquisition unit includes a camera, a weighbridge, an RFID tag, and a speed measuring radar.

3. The toll station collision avoidance early warning method based on motion trajectory prediction according to claim 2, characterized in that, The step of parsing the motion trajectory prediction value based on the moving vehicle information includes: The distance the vehicle traveled was determined by setting up reference points in the square, and the movement data was calculated by combining this with the time of image acquisition. The movement data is acquired through the speed-measuring radar; The radio frequency tag is used to identify the individual information of the object and match it with the mobile data. The individual information includes the vehicle model, the license plate, and the person information. The vehicle's weight information is obtained through the weighbridge. Based on the movement data, an initial motion trajectory prediction value is obtained. Based on the individual information and / or the weight information, the initial motion trajectory prediction value is modified to obtain the final motion trajectory prediction value.

4. The toll station collision avoidance early warning method based on motion trajectory prediction according to claim 3, characterized in that, The process of obtaining an initial motion trajectory prediction value based on the motion data includes: The motion data is processed based on the b-spline curve fitting algorithm to obtain the predicted value of the initial motion trajectory.

5. The toll station collision avoidance early warning method based on motion trajectory prediction according to claim 4, characterized in that, The b-spline curve fitting algorithm processes the moving data, including: Data preprocessing: The original trajectory point sequence of the moving data is processed using Kalman filtering or moving average trajectory to obtain a smoothed trajectory point sequence; Parameterization and node vector generation: Combine the B-spline order and trajectory point sequence to output the corresponding node vector; B-spline fitting: Select the order, calculate the node vectors, and solve for the control points using the least squares method; Trajectory prediction: Starting from the current parameter values, the prediction is extended by extending the node vector, extrapolating the B-spline curve, calculating future time points, and solving for control points.

6. The toll station collision avoidance early warning method based on motion trajectory prediction according to claim 5, characterized in that, The step of modifying the initial motion trajectory prediction value based on the individual information and / or the weight information to obtain the motion trajectory prediction value includes: Based on the individual information, obtain the corresponding historical driving records; If the historical driving records are analyzed, and abnormal driving records are found, the initial motion trajectory prediction value is increased; otherwise, the initial motion trajectory prediction value is maintained. If the weight information is greater than the specified threshold, the initial motion trajectory prediction value is maintained.

7. The toll station collision avoidance early warning method based on motion trajectory prediction according to claim 6, characterized in that, The plaza grid is obtained by dividing the toll station area into grids, and each grid is set with a corresponding risk level. Correspondingly, based on the predicted motion trajectory value and the square grid, the corresponding early warning measures are executed, including: Based on the predicted motion trajectory values, overlapping grids are determined, and corresponding early warning measures are implemented according to the risk level, overlap probability, and overlap time of the overlapping grids.

8. The toll station collision avoidance early warning method based on motion trajectory prediction according to claim 7, characterized in that, The aforementioned early warning measures include: Activate the audible and visual alarm devices, start the barrier devices, focus on storing data of moving targets, and interact with the targets via data exchange.

9. The toll station collision avoidance early warning method based on motion trajectory prediction according to claim 8, characterized in that, The step of implementing corresponding early warning measures based on the risk level, overlap probability, and overlap time of overlapping grids includes: Based on the risk level, control the signal strength of the audible and visual alarm device, the start-up speed of the barrier device, prioritize data storage and / or deletion, and prioritize data interaction; Based on the overlap probability, determine the audible and visual alarm device and the railing device that need to be controlled; Based on the overlap time, the audible and visual alarm device and the railing device are controlled in advance.

10. A toll station collision avoidance warning system based on motion trajectory prediction, characterized in that, include: The first module is used to acquire information about moving vehicles through a collection unit set up in the square; The second module is used to parse and obtain the predicted motion trajectory value based on the mobile vehicle information; The third module is used to execute corresponding early warning measures based on the predicted motion trajectory value and the square grid.

Images