CN117473741A - A full-sample high-resolution vehicle trajectory robust reconstruction method, equipment, and medium - Google Patents

Abstract

The invention relates to a full-sample high-resolution vehicle track robust reconstruction method, device and medium, wherein the reconstruction method comprises the following steps: acquiring coil detector data and network connection data, and estimating a space-time velocity matrix by a self-adaptive smoothing method based on traffic jam flow and free flow characteristic improvement; selecting a non-network vehicle to be reconstructed, and generating candidate tracks through an IDM model according to the upstream and downstream known tracks; calculating the weight of the candidate track by taking the generated space-time velocity matrix as constraint on the non-network vehicle connected with the generated candidate track, fusing the candidate track by using a weighting method, and reconstructing the high-resolution vehicle track of the non-network vehicle connected with the vehicle; and sequentially repeating the candidate track generation and fusion process for each non-networked vehicle. Compared with the prior art, the invention improves the reconstruction precision and robustness; the track information of the vehicle in different scenes of the blocking flow and the free flow is accurately estimated, and the optimal reconstruction result can be approached through simple parameter adjustment.

Description

A full-sample high-resolution vehicle trajectory robust reconstruction method, equipment, and medium

Technical field

The invention relates to the field of intelligent transportation technology, and in particular to a robust reconstruction method, equipment, and medium of full-sample high-resolution vehicle trajectories with scene flexibility.

Background technique

High-resolution vehicle trajectories can provide a large amount of traffic spatio-temporal information, which can not only extract vehicle group characteristics such as flow, speed, density, etc., but also be used to analyze the microscopic behavior of interactions between individual vehicles, providing information for traffic state estimation, traffic flow modeling, and signaling. It provides a good foundation for applications such as control optimization and traffic emission measurement. However, the deployment rate of traditional fixed detectors is not high, and the penetration rate of mobile detectors is low, so they cannot be used to directly obtain high-resolution vehicle trajectories; the cost of obtaining high-resolution trajectories directly through drone photography and other means is relatively high. In the context of difficult trajectory acquisition, many vehicle trajectory reconstruction methods have emerged, but the following problems still exist: model assumptions are inconsistent with reality, high data requirements, and single design scenarios hinder the practical application of these methods. Therefore, it is necessary to fully integrate and utilize the data detected by various sensors to construct a vehicle trajectory reconstruction model with reasonable assumptions, adaptable to various data conditions, and flexible response to traffic scenarios, and reconstruct complete high-resolution vehicle trajectory data from sparse data observations. , which is of great significance to the field of intelligent transportation.

Early vehicle trajectory reconstruction methods were limited by detector technology. They mainly used fixed detector data (coils, radars, license plate recognition, etc.) to reconstruct trajectories through variation theory, traffic flow basic diagrams and other methods. The reconstruction results were mainly used for vehicles. Travel time estimates, etc. This type of method largely ignores complex behaviors such as acceleration, deceleration and car-following of vehicles, and does not have the ability to characterize the microscopic behavior between vehicles, and cannot be used for more microscopic applications (traffic safety, traffic shock research, etc.). With the development of motion detectors, some methods are based on connected and autonomous vehicle data and use vehicle following models to reconstruct trajectories, usually assuming that connected and autonomous vehicles have a high penetration rate. However, in actual scenarios, the penetration rate of connected and autonomous vehicles is basically below 10%, and it is difficult to increase significantly in a short period of time.

There are the following defects in the existing technology:

(1) Most existing methods fail to solve the problem of vehicle microscopic behavior characterization and compatibility with sparse data scenarios at the same time. The method based on fixed point detection data has low data requirements, but cannot accurately represent vehicle behaviors such as acceleration and deceleration; the method based on mobile detector data better depicts the microscopic behavior of the vehicle, but has high data quality requirements and requires high upload frequency and high The premise of penetration rate is inconsistent with the real scenario.

(2) Multi-source data fusion is insufficient in existing methods. Although a framework for constraining micro-trajectory reconstruction based on macro-traffic status has been proposed, the interaction between macro-micro modules is insufficient, the degree of data fusion and data usage efficiency are not high, and the accuracy of vehicle trajectory reconstruction needs to be further improved.

Contents of the invention

The purpose of this invention is to overcome the shortcomings of the above-mentioned existing technologies and provide a full-sample high-resolution vehicle trajectory robust reconstruction method with scene flexibility. In terms of data fusion, the error impedance model (ER) is introduced to increase the macroscopic The interaction between speed information and micro-trajectory information makes full use of multi-source data and greatly improves the reconstruction accuracy and robustness; on the other hand, the intelligent driver model (IDM) is used to accurately estimate the vehicle's flow in blocked flow and Free flow trajectory information in different scenarios, and the optimal reconstruction results can be approached through simple parameter adjustment.

The object of the present invention can be achieved through the following technical solutions:

A first aspect of the present invention provides a full-sample high-resolution vehicle trajectory robust reconstruction method, which includes the following steps:

S1: Obtain coil detector data and connected vehicle data, and estimate the space-time velocity matrix through an improved adaptive smoothing method based on traffic jam flow and free flow characteristics;

S2: Select the non-connected vehicle that needs to reconstruct the trajectory, determine the upstream and downstream reference trajectories of the vehicle based on the error impedance model (ER), and generate the candidate trajectory of the vehicle through the IDM model based on the reference trajectory;

S3: For the non-connected vehicle that has generated candidate trajectories in step S2, use the space-time velocity matrix generated in step S1 as a constraint, calculate the weight of the two candidate trajectories, and use the weighted method to fuse the two candidate trajectories to reconstruct the non-networked vehicle For the trajectories of connected vehicles, return to S2 until all trajectories of non-connected vehicles are reconstructed in sequence.

High-resolution vehicle trajectories refer to data describing vehicle position and movement with fine granularity in time and space.

Further, in S1, the coil detector data includes: detection record ID, vehicle instantaneous speed, and vehicle elapsed time;

The connected vehicle data includes: detected vehicle ID, timestamp, vehicle coordinates, and vehicle instantaneous speed.

Further, the acquisition process of the space-time velocity matrix includes:

S1-1: For the space-time domain on a single lane, divide the space-time velocity matrix with 3 meters of space and 3 seconds of time as the minimum aggregation unit, and use the coil detector data and connected vehicle data to know the x position and time t The traffic speed v initializes the space-time velocity matrix;

S1-2: Calculate the smoothing kernel φ(·) and normalization factor for each unknown matrix element (x, t)

S1-3: Considering the traffic jam flow and free flow characteristics at the same time, adjust the smoothing kernel V _free (x, t) and V _cong (x, t) respectively;

S1·4: Calculate the weight w(x, t) to weigh the free flow and blocked flow characteristics;

S1-5: Estimate the unknown traffic speed in space-time (x, t), and complete the space-time speed matrix:

V _refer (x, t) = w (x, t) V _cong (x, t) + [1-w (x, t)] V _free (x, t).

Further, in S1-2, the smoothing kernel φ(·) and the normalization factor The method of obtaining is:

Among them, x _i , ti _{, vi} (i=1,..., n) are the known position, time and traffic speed in the corresponding time and space respectively. The smoothing width σ in the space coordinate is 80m, and the smoothing width σ in the time coordinate is 80m. The smoothing width τ is taken as 6.5s;

In S1-3, adjust the smoothing kernel V _free (x, t) and V _cong (x, t) as:

Among them, the propagation speed of traffic disturbance in free flow c _free is taken as 70km/h, and the propagation speed of traffic disturbance in blocked flow c _cong is taken as -15km/h;

In S1-4, the calculation process of weight w(x, t) is:

The threshold V _thr between free flow and blocked flow is 60km/h, and the transition width ΔV between free flow and blocked flow is 20km/h.

Further, in S2, the specific process is as follows:

S2-1: There are I connected vehicle trajectories in the connected vehicle trajectory set Y, and the reconstruction interval between the adjacent upstream connected vehicle trajectory Y ⁱ and the downstream connected vehicle trajectory Y ⁱ⁺¹ is obtained. Among them, N represents the number of non-connected vehicle trajectories that need to be reconstructed in this interval. In the initial state, i = 1;

S2-2: pair interval For the nth non-connected vehicle in , determine its upstream reference trajectory X _upREF and downstream reference trajectory X _downREF . In the initial state, n=1;

S2-3: Based on the IDM model, generate the candidate trajectory of the nth non-connected vehicle based on the upstream reference trajectory X _upREF

S2-4: Based on the IDM model, generate the candidate trajectory of the nth non-connected vehicle based on the downstream reference trajectory X _downREF

Further, in S2-2, the determination process of the upstream reference trajectory X _upREF and the downstream reference trajectory X _downREF includes:

a. If n=1, X _upREF =Y ⁱ ,

b. If n=N, X _downREF =Y ⁱ⁺¹ ;

c. If n≠1 and n≠N,

in is the reconstructed trajectory of the n-1th non-connected vehicle;/> It is the candidate trajectory of the n+2 non-connected vehicle downstream of the n+1 non-connected vehicle/> Candidate trajectories generated for reference,/> It is a candidate trajectory generated based on the downstream connected vehicle trajectory Y ⁱ⁺¹ ;

In S2-3, the candidate trajectory of the nth non-connected vehicle is generated based on the upstream reference trajectory X _upREF . The process is:

a. by Calculate/>

b. by Calculate/>

min location error＝|x′ _upREF (t-2)-x _upREF (t-2)|

Among them, the initial Provided by coil detection data,/> and/> is the estimated acceleration, speed and position of non-connected vehicle n with reference to the vehicle in front of it at time t, v _upREF (t) and x _upREF (t) are the speed and position of the vehicle in front of non-connected vehicle n at time t, a′ , v′ and x′ are estimated values. The maximum acceleration a is taken as 2.75m/s ² , the most comfortable deceleration b is taken as 2.25m/s ² , the free flow speed v ₀ is taken as 32m/s, and the safe distance between vehicles at s ₀ is taken as 8m. s ^* is the distance between vehicles, and the reaction time T is 1.1s.

Further, in S2-4, the candidate trajectory of the nth non-connected vehicle is generated according to the downstream reference trajectory X _downREF. The process is:

a. by Calculate/>

minlocation error＝|x′ _downREF (t+2)-x _downREF (t+2)|

b. by Calculate/>

Among them, the initial Provided by coil detection data,/> and/> is the estimated acceleration, speed and position of the vehicle behind the non-connected vehicle n at time t, v _downREF (t) and x _downREF (t) are the speed and position of the vehicle behind the non-connected vehicle n at time t.

Further, in S3, the specific process is:

S3-1: Use the estimated space-time velocity matrix V _refer in S1 as a constraint to solve the candidate trajectory of non-connected vehicle n in S2 and/> weight/> and/>

S3-2: Calculate the high-resolution trajectory of non-connected vehicle n according to the weighted method

S3-3: Set n=n+1, return to S2 and continue to reconstruct the trajectory of the next non-connected vehicle until the reconstruction interval The N non-connected vehicles in are all reconstructed, that is, n=N.

S3-4: Let i=i+1, n=1, return to S2 and continue to reconstruct the trajectories of N non-connected vehicles in the next interval until all non-connected vehicles in the interval are reconstructed, that is, i= I-1.

A second aspect of the present invention provides an electronic device, including a memory and a processor. The processor is configured to execute a program in the memory, thereby realizing the above-mentioned robust reconstruction method of full-sample high-resolution vehicle trajectory.

A third aspect of the present invention provides a storage medium containing computer-executable instructions. When executed by a computer processor, the storage medium of computer-executable instructions is used to perform the above-mentioned full-sample high-resolution vehicle trajectory robust reconstruction. construction method.

Compared with the existing technology, the present invention has the following technical advantages:

(1) Although existing methods also use macroscopic traffic information to constrain trajectory reconstruction, they do not fully utilize the macroscopic information, causing errors to accumulate as the number of non-connected vehicles to be reconstructed increases, and the method lacks robustness. . The proposed method enables a more complete interaction between macro and micro traffic information and increases the robustness of high-resolution trajectory reconstruction.

(2) Although the existing methods also combine traffic macro and micro models, the Newell model of the micro module in the existing methods assumes that the trajectories of the front and rear vehicles are consistent and lacks an accurate description of the interaction between vehicles. Structure trajectory distortion. The proposed method is compatible with vehicle trajectory reconstruction in blocked flow and free flow scenarios, and can characterize microscopic vehicle behavior more reasonably and accurately.

Description of the drawings

Figure 1 shows the effect of the IDM model before and after use under 5% connected vehicle penetration rate;

Figure 2 shows the effects of the ER model before and after use under 5% connected vehicle penetration;

Figure 3 shows the error of the method proposed in this solution under different parameter settings;

Figure 4 is a schematic diagram of the overall process in this solution.

Detailed ways

The present invention provides a full-sample high-resolution trajectory robust reconstruction method with scene flexibility: using sparse connected vehicle data and easily accessible coil detector data, considering the error impedance model (ER) at the macro and micro level Under the module interaction framework, the macroscopic space-time velocity matrix is first estimated, and then the microscopic vehicle trajectory is generated using the intelligent driver model (IDM) with the macroscopic speed information as a constraint. On the one hand, the introduction of the ER model in data fusion increases the interaction between macroscopic speed information and microscopic trajectory information, greatly improving the reconstruction accuracy and robustness; on the other hand, using the IDM model to accurately estimate the vehicle's flow in blocked flow and free flow Stream trajectory information in different scenarios, and get close to the optimal reconstruction results through simple parameter adjustment.

The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. Features such as component models, material names, connection structures, control methods, algorithms, etc. that are not explicitly stated in this technical solution are regarded as common technical features disclosed in the prior art.

Example 1

In this embodiment, the full-sample high-resolution vehicle trajectory robust reconstruction method with scene flexibility includes the following steps, see Figure 4:

Step 1: Input the coil detector data and connected vehicle data, and estimate the space-time velocity matrix through the adaptive smoothing method (ASM) improved based on traffic jam flow and free flow characteristics, that is, the traffic macro model.

Step 2: Select the non-connected vehicle whose trajectory needs to be reconstructed, determine the upstream and downstream reference trajectories of the vehicle based on the error impedance model (ER), and generate the candidate trajectory of the vehicle through the IDM model (traffic micro model) based on the reference trajectory.

Step 3: For the non-connected vehicle that has generated candidate trajectories in Step 2, use the space-time velocity matrix generated in Step 1 as a constraint, calculate the weight of the two candidate trajectories, and use the weighted method to fuse the two candidate trajectories to reconstruct the non-connected vehicle. Trajectories of connected vehicles; return to step 2 until all non-connected vehicle trajectories are reconstructed in sequence.

Among them, IDM (Intelligent Driver Model) intelligent driver model has a small number of parameters and clear meaning, and can use a unified model to describe different states from free flow to completely congested flow.

The specific process of step 1 is as follows:

(1) Enter coil detector data, including detection record ID, vehicle instantaneous speed, and vehicle elapsed time; input connected vehicle data, including detection vehicle ID, timestamp, vehicle coordinates, and vehicle instantaneous speed.

(2) For the space-time domain on a single lane, use 3 meters of space and 3 seconds of time as the minimum aggregation unit, divide the space-time speed matrix, and use the traffic speed v at position x and time t that has been recorded in the coil and connected vehicle data Initialize matrix.

(3) Calculate the smoothing kernel φ(·) and normalization factor for each unknown matrix element (x, t)

Among them, x _i , ti _{, vi} (i=1,..., n) are the known position, time and traffic speed in the corresponding time and space respectively. The smoothing width σ in the space coordinate is 80m, and the smoothing width σ in the time coordinate is 80m. The smoothing width τ is taken as 6.5s.

(4) Considering the characteristics of traffic jam flow and free flow at the same time, the smoothing kernel is adjusted to:

Among them, the propagation speed of traffic disturbance in free flow c _free is taken as 70km/h, and the propagation speed of traffic disturbance in blocked flow c _cong is taken as -15km/h.

(5) Calculate the weight w(x, t) to weigh the free flow and blocked flow characteristics:

The threshold V _thr between free flow and blocked flow is 60km/h, and the transition width ΔV between free flow and blocked flow is 20km/h.

(6) Estimate the unknown traffic speed in space-time (x, t) and complete the space-time speed matrix:

V _refer (x, t) = w (x, t) V _cong (x, t) + [1-w (x, t)] V _free (x, t)

The specific process of step 2 is as follows:

(1) There are I connected vehicle trajectories in the connected vehicle trajectory set Y, and there is a reconstruction interval between the adjacent upstream connected vehicle trajectory Y ⁱ and the downstream connected vehicle trajectory Y ⁱ⁺¹ N represents the number of non-connected vehicle trajectories that need to be reconstructed in this interval, and i=1 under the initial conditions.

(2) For the interval For the nth non-connected vehicle in , determine its upstream reference trajectory X _upREF and downstream reference trajectory X _downREF . Under the initial condition, n=1:

a. If n=1, X _upREF =Y ⁱ ,

b. If n=N, X _downREF =Y ⁱ⁺¹ ;

c. If n≠1 and n≠N,

in is the reconstructed trajectory of the n-1th non-connected vehicle. By using the information in the reconstructed trajectory, the occurrence of errors can be further reduced, which is called the error impedance model (ER);/> It is the candidate trajectory of the n+2 non-connected vehicle downstream of the n+1 non-connected vehicle/> (n+2<N) generates candidate trajectories for reference,/> The candidate trajectory is generated based on the downstream connected vehicle trajectory Y ⁱ⁺¹ .

(3) Based on the IDM model, generate the candidate trajectory of the nth non-connected vehicle based on the upstream reference trajectory X _upREF

a. by Calculate/>

b. by Calculate/>

min location error＝|x′ _upREF (t-2)-x _upREF (t-2)|

where the initial Provided by coil detection data,/> and/> are the acceleration, speed and position estimated by the reference vehicle in front of non-connected vehicle n at time t, v _upREF (t) and x _upREF (t) are the acceleration, speed and position of the vehicle in front of non-connected vehicle n (i.e. the upstream reference vehicle) at time t Velocity and position, a′, v′ and x′ are estimates. The maximum acceleration a is taken as 2.75m/s ² , the most comfortable deceleration b is taken as 2.25m/s ² , the free flow speed v ₀ is taken as 32m/s, the safe vehicle distance of s ₀ is taken as 8m, s ^* is the vehicle distance, and the reaction time T is taken as 1.1s.

(4) Based on the IDM model, the candidate trajectory of the nth non-connected vehicle is generated based on the downstream reference trajectory X _downREF .

a. by Calculate/>

min location error＝|x′ _downREF (t+2)-x _downREF (t+2)|

b. by Calculate/>

where the initial Provided by coil detection data,/> and/> is the estimated acceleration, speed and position of the reference vehicle behind the non-connected vehicle n at time t, v _downREF (t) and x _downREF (t) are the speed of the vehicle behind the non-connected vehicle n (that is, the downstream reference vehicle) at time t and location.

The specific process of step 3 is as follows:

(1) Use the space-time velocity matrix V _refet estimated in step 1 as a constraint to solve the candidate trajectory of non-connected vehicle n in step 2 and/> weight/> and/>

(2) Calculate the high-resolution trajectory of non-connected vehicle n according to the weighted method

(3) Set n=n+1 and return to S2 to continue calculating the candidate trajectory of the next non-connected vehicle until the reconstruction interval The N non-connected vehicles in are all reconstructed, that is, n=N.

(4) Let i=i+1, n=1, return to S2 and continue to calculate the candidate trajectories of N non-connected vehicles in the next reconstruction interval until all non-connected vehicles in the interval are reconstructed, that is, i =I-1.

Verification example 1

The present invention conducts case analysis on the above method, and the trajectory reconstruction effect is good. The case data comes from the NGSIM data set collected by the U.S. Federal Highway Administration. Analyze vehicle trajectory data captured by drones from 7:50 to 7:55 a.m. in the leftmost lane of Highway 101 in Los Angeles, California, traveling from north to south.

In order to verify the compatibility of this method for low penetration scenarios, sparse data scenarios with 85%, 90%, and 95% missing connected vehicle data were constructed to test the performance. The root mean square error (RMSE), the average absolute error (MAE), and the average absolute percentage were measured. Error MAPE is used as an indicator to measure the accuracy of data recovery. The calculation method is as follows:

in is the observation value trajectory point,/> is the reconstruction result trajectory point. MAE represents the average position error, RMSE is sensitive to outliers and extreme values, and MAPE is the size of the error relative to the length of the reconstructed road segment.

The performance of the method combining macro and micro models is better than the traditional interpolation and variation methods. In order to illustrate the superiority of the proposed method, macro and micro modules based on different models were selected to test the performance. The test results are shown in Table 1. As the penetration rate increases from 5% to 15%, and especially from 5% to 10%, the errors represented by the three indicators decrease. It shows that permeability significantly affects the performance of various reconstruction methods, and high-resolution vehicle trajectory reconstruction under extremely low permeability is challenging. In these low penetration rate scenarios, the proposed method not only reconstructs more accurate vehicle trajectories (the reconstruction accuracy under each penetration rate is improved by more than 10%), but also has better performance in sparse data scenarios. Robustness (reconstruction accuracy increased by 28.9% at 5% penetration rate).

Table 1 Vehicle trajectory vertical component performance of different macro and micro fusion methods

The introduction of the IDM model makes the method flexible in scenarios. Newell is a widely used car-following model. Figure 1 compares the reconstruction effects of using the Newell model and the IDM model in the microscopic module. The method used in the present invention more accurately reconstructs the process of the vehicle decelerating to a slow speed when the traffic flow is blocked, and then accelerating back to the free flow speed after a period of time, while ensuring the reconstruction effect of the vehicle trajectory in the free flow state.

In the process of macro-micro information interaction, the error impedance model (ER) is introduced to reduce the errors generated in the trajectory reconstruction process. Figure 2 compares the errors of candidate trajectories generated in step S2 before and after using the ER model. As shown in Figure 2(a), the error will increase as the number of non-connected vehicles that need to be reconstructed between adjacent connected vehicle trajectories increases. That is, the farther away from the connected vehicles, the greater the reconstruction error. Large, and the ER model reduces such errors; Figure 2(b) and Figure 2(c) intuitively reflect the effect of the ER model. Regardless of the free flow or blocked flow state, the candidate trajectories generated by this model are closer to actual value. These improvements are due to the fact that when estimating candidate trajectories based on the ER model, the reconstructed vehicle trajectory of the preceding vehicle is used as the baseline, and the reconstructed trajectory of the preceding vehicle is generated under the constraints of macro traffic speed information. In this way, the traffic macro speed information is more fully utilized and the reconstruction error is further reduced.

In order to measure the impact of parameters on model performance, a parameter sensitivity test was conducted on the invented method, and the results are shown in Figure 3. When the free flow velocity v ₀ is set to 20-36m/s, the average absolute error is stable at about 8m; when the safety distance is set to 6-9m, the average absolute error is stable below 9m. Near the optimal parameter values of maximum acceleration a and comfortable deceleration b, the performance of the model is also relatively stable. It shows that the proposed method is not sensitive to parameters, can approach the optimal effect through simple parameter adjustment, and is competitive in practical applications.

This embodiment also proposes a full-sample high-resolution vehicle trajectory robust reconstruction device. The device includes a processor and a memory. The processor and the memory are coupled. The memory stores program instructions. When the program instructions stored in the memory are processed by the processor Implement the above task management method during execution. The processor can be a general-purpose processor, including a central processing unit (Central Processing Unit, referred to as CPU), a network processor (Network Processor, referred to as NP), etc.; it can also be a digital signal processor (Digital Signal Processing, referred to as DSP), application specific integrated circuit (Application Specific Integrated Circuit) , ASIC for short), Field-Programmable Gate Array (FPGA for short) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components; the memory may include Random Access Memory (Random Access Memory, RAM for short) ), and may also include non-volatile memory (Non-VolatileMemory), such as at least one disk memory. The memory may be an internal memory of random access memory (Random Access Memory, RAM) type, and the processor and memory may be integrated into one or more independent circuits or hardware, such as an application specific integrated circuit (Application Specific Integrated Circuit, ASIC). It should be noted that the computer program in the above-mentioned memory can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on this understanding, the technical solution of the present invention essentially contributes to the existing technology or the part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several Instructions are used to cause a computer device (which may be a personal computer, electronic device, or network device, etc.) to execute all or part of the steps of the methods of various embodiments of the present invention.

This embodiment also provides a computer-readable storage medium that stores computer instructions, and the computer instructions are used to cause the computer to execute the above-mentioned full-sample high-resolution vehicle trajectory robust reconstruction method. Storage media may be electronic media, magnetic media, optical media, electromagnetic media, infrared media or semiconductor systems or propagation media. Storage media may also include semiconductor or solid-state memory, magnetic tape, removable computer disks, random access memory (RAM), read-only memory (ROM), hard disks, and optical disks. Optical disks may include compact disk-read-only memory (CD-ROM), compact disk-read/write (CD-RW), and DVD.

The above description of the embodiments is to facilitate those of ordinary skill in the technical field to understand and use the invention. It is obvious that those skilled in the art can easily make various modifications to these embodiments and apply the general principles described herein to other embodiments without inventive efforts. Therefore, the present invention is not limited to the above embodiments. Based on the disclosure of the present invention, improvements and modifications made by those skilled in the art without departing from the scope of the present invention should be within the protection scope of the present invention.

Claims (10)

1. The full-sample high-resolution vehicle track robust reconstruction method is characterized by comprising the following steps of:

s1: acquiring coil detector data and network connection data, and estimating a space-time velocity matrix by a self-adaptive smoothing method based on traffic jam flow and free flow characteristic improvement;

s2: selecting a non-network vehicle with a track to be reconstructed, determining an upstream reference track and a downstream reference track of the vehicle based on an error impedance model (ER), and generating a candidate track of the vehicle through an IDM model according to the reference track;

s3: and (3) for the non-network connected vehicle with the candidate track generated in the step (S2), calculating weights of the two candidate tracks by taking the space-time velocity matrix generated in the step (S1) as constraint, fusing the two candidate tracks by using a weighting method to reconstruct the track of the non-network connected vehicle, and returning to the step (S2) until all the non-network connected vehicle tracks are sequentially reconstructed.

2. The full-sample high-resolution vehicle trajectory robust reconstruction method of claim 1, wherein in S1, said coil detector data comprises: detecting record ID, vehicle instantaneous speed and vehicle elapsed time;

the internet protocol data comprises: vehicle ID, timestamp, vehicle coordinates, vehicle instantaneous speed are detected.

3. The full-sample high-resolution vehicle trajectory robust reconstruction method of claim 1, wherein the space-time velocity matrix acquisition process comprises:

s1-1: dividing a space-time velocity matrix by taking 3 m space and 3 s time as minimum set units on a time-space domain on a single lane, and initializing the space-time velocity matrix by using recorded traffic velocity v at the time of an x position t in coil detector data and network vehicle connection data;

s1-2: calculating a smoothing kernel phi (,) and a normalization factor for each matrix element (x, t) of unknown numerical value

S1-3: simultaneously, the traffic blocking flow and the free flow characteristics are considered to respectively adjust the smooth kernel V _free (x, t) and V _cong (x，t)；

S1-4: calculating weights w (x, t) to trade-off free flow and blocked flow characteristics;

s1-5: estimating an unknown traffic speed under space-time (x, t), and complementing the space-time speed matrix:

V _refer (x，t)＝w(x，t)V _cong (x，t)+[1-w(x，t)]V _free (x，t)。

4. a full-sample high-resolution vehicle trajectory robust reconstruction method as recited in claim 3, wherein in S1-2, said smoothing kernel Φ (·) and normalization factorThe acquisition mode of (a) is as follows:

wherein x is _i 、t _i 、v _i (i=1,..n.) is the known position, time and traffic speed in the corresponding space-time, respectively, the smoothed width σ in the spatial coordinates is 80m and the smoothed width τ in the temporal coordinates is 6.5s;

in S1-3, smooth kernel V is adjusted _free (x, t) and V _cong (x, t) is:

wherein the propagation speed c of the traffic disturbance in the free flow _free Taking the propagation speed c of traffic disturbance in the blocking flow at 70km/h _cong Taking-15 km/h;

in S1-4, the calculation process of the weight w (x, t) is as follows:

wherein the threshold value V between free flow and blocked flow _thr The transition width DeltaV between the free stream and the choked stream was taken at 60km/h and 20km/h.

5. A full-sample high-resolution vehicle track robust reconstruction method as recited in claim 3, wherein in S2, the specific process is as follows:

s2-1: i net train tracks are shared in the net train track set Y, and adjacent upstream net train tracks Y are obtained ⁱ And downstream net train track Y ⁱ⁺¹ Reconstruction interval therebetweenWherein N represents the number of non-networked vehicle tracks to be reconstructed in the interval, and i=1 under the initial condition;

s2-2: for the intervalThe nth non-network vehicle in the system determines the upstream reference track X of the vehicle _upREF And a downstream reference trajectory X _downREF N=1 under initial conditions;

s2-3: based on IDM model, according to upstream reference track X _upREF Generating candidate tracks of the nth non-networked vehicle

S2-4: based on IDM model, according to downstream reference track X _downREF Generating candidate tracks of the nth non-networked vehicle

6. The full-sample high-resolution vehicle track robust reconstruction method as recited in claim 5, wherein in S2-2, the upstream reference track X _upREF And a downstream reference trajectory X _downREF The determining process of (1) comprises:

a. if n=1, x _upREF ＝Y ⁱ ，

b. If n=n,X _downREF ＝Y ⁱ⁺¹ ；

c. if N is not equal to 1 and N is not equal to N,

wherein the method comprises the steps ofIs the reconstruction track of the n-1 non-network vehicle; />Is the candidate track +.>Candidate trajectories generated for reference, +.>Is a downstream network-connected track Y ⁱ⁺¹ Candidate trajectories generated for the reference;

s2-3 according to the upstream reference track X _upREF Generating candidate tracks of the nth non-networked vehicleThe process of (1) is as follows:

a. from the following componentsCalculation of/>

b. From the following componentsCalculate->

min location error＝|x′ _upREF (t-2)-x _upREF (t-2)|

Wherein, initiallyIs provided by coil detection data->And->Acceleration, speed and position at time t estimated by non-networked vehicle n with reference to preceding vehicle, v _upREF (t) and x _upREF (t) the speed and position of the preceding vehicle of the non-networked vehicle n at the moment t, a ', v ' and x ' are estimated values, and the maximum acceleration a takes 2.75m/s ² The most comfortable deceleration b takes 2.25m/s ² Free flow vehicle speed v ₀ Taking 32m/s, s ₀ The distance between safety vehicles is 8m, s ^* The vehicle distance is 1.1s for the reaction time T.

7. The full-sample high-resolution vehicle track robust reconstruction method as recited in claim 5, wherein in S2-4, the reference track X is based on the downstream reference track _downREF Generating candidate tracks of the nth non-networked vehicleThe process of (1) is as follows:

a. from the following componentsCalculate->

min location error＝|x′ _dounREF (t+2)-x _downREF (t+2)|

b. From the following componentsCalculate->

Wherein, initiallyIs provided by coil detection data->And->Acceleration, speed and position at time t, v, estimated by non-networked vehicle n with reference to rear vehicle _downREF (t) and x _downREF (t) speed and position of the following vehicle of the non-networked vehicle n at the time t.

8. The full-sample high-resolution vehicle track robust reconstruction method according to claim 1, wherein in S3, the specific process is:

s3-1: the space-time velocity matrix V estimated in S1 _refer As constraint, solving candidate track of non-internet-connected vehicle n in S2And->Weight of +.>And->

S3-2: high-resolution track of non-internet-connected vehicle n is calculated according to weighting method

S3-3: returning to S2 to reconstruct the track of the next non-networked vehicle until the reconstruction intervalThe N non-networked vehicles in (a) are all reconstructed, namely n=n;

s3-4: let i=i+1, n=1, return S2 to continue reconstructing the trajectories of N non-networked vehicles in the next section until all non-networked vehicles in the section are reconstructed, i.e. i=i-1.

9. An electronic device comprising a memory, a processor, wherein the processor is configured to execute a program in the memory, thereby implementing the full-sample high-resolution vehicle trajectory robust reconstruction method according to any one of claims 1 to 8.

10. A storage medium containing computer executable instructions which, when executed by a computer processor, are for performing the full-sample high-resolution vehicle trajectory robust reconstruction method of any one of claims 1 to 8.