Microcosmic traffic simulation speed trajectory regeneration method responding to emission evaluation

Abstract

The invention belongs to the field of green traffic, and particularly relates to a microcosmic traffic simulation speed trajectory regeneration method responding to emission evaluation. Comprising the following steps: 1, selecting an emission factor model, and simulating and outputting second-by-second speed, lane and position data of a vehicle; step 2, constructing acceleration-jerk envelope line models of different vehicle types; 3, considering constraint conditions of kinematics consistency, dynamics controllability, jerk boundary, following safety and stroke boundary consistency, and constructing a vehicle speed trajectory control model; 4, performing linearization processing on nonlinear constraints in the vehicle speed trajectory control model by using an SOS2 technology; and 5, designing a'relaxation-strengthening 'solving algorithm, and optimizing output to obtain a speed trajectory sequence responding to emission evaluation and physical consistency. According to the method, a transferable technical path is provided for a robust coupling microscopic simulation and emission model, and the method can be used for hot spot analysis of a project scale, near-path air quality evaluation and design of a refined emission reduction strategy.

Description

A method for speed trajectory regeneration based on microscopic traffic simulation in response to emissions assessment Technical Field

[0001] This invention belongs to the field of green transportation, and specifically relates to a method for regenerating speed trajectories in response to emission assessments based on microscopic traffic simulation. Background Technology

[0002] Quantifying road vehicle emissions is a crucial step in identifying major sources of air pollution, assessing the environmental effects of traffic management measures, and supporting sustainable transportation development. Due to the high cost and organizational challenges of conducting large-scale real-vehicle emissions testing, current engineering practices primarily rely on models to estimate or predict vehicle emissions.

[0003] Vehicle emissions are influenced by a combination of factors, including vehicle type, engine operating status, environmental conditions, and driving behavior. To characterize engine load, the current field generally employs the Vehicle-Specific Power (VSP) model, which uniformly represents factors such as vehicle mass, rolling resistance coefficient, drag coefficient, operating speed, acceleration, and road longitudinal slope, and establishes a mapping relationship between VSP and fuel consumption or emission rate. Based on this principle, existing emission models (such as MOVES, CMEM, and FEC) can estimate vehicle emissions using the vehicle speed-acceleration trajectory every second.

[0004] On the other hand, microscopic traffic simulation tools (such as SUMO, TESSNG, VISSIM, CORSIM, PARAMICS, AIMSUN, etc.) can output second-by-second vehicle trajectory information and have been widely used for performance evaluation of traffic control schemes, road design, and management strategies. Relevant guidelines explicitly state that second-by-second vehicle trajectories from microscopic simulations can be used for project-level emissions assessments such as hotspot analysis. Therefore, in practical engineering, the output of microscopic simulations is often directly used as input to emissions models.

[0005] However, existing microscopic traffic simulation platforms are mainly used to reproduce traffic flow characteristics. Their built-in car-following models and driving behavior models typically use traffic performance indicators such as flow rate, speed, delay, and queue length as calibration targets. These models struggle to realistically simulate fine-grained vehicle dynamics (such as real-world speed-acceleration changes and instantaneous engine load), leading to significant deviations between simulated trajectories and real trajectories in key dynamic variables such as acceleration and jerk. In VSP-based emissions modeling, these deviations directly affect engine power estimation, causing problems such as incorrect emission factor selection and misjudgment of emission rates, ultimately resulting in systematic biases in emissions assessment results.

[0006] Current technology lacks a method to correct the simulation output trajectory without altering the original microscopic simulation framework, ensuring it simultaneously satisfies dynamic consistency constraints, Jerk boundary constraints, and car-following safety constraints, while guaranteeing that the operational characteristics are consistent with real-world vehicle behavior. If the microscopic simulation velocity trajectory could be regenerated to conform to vehicle physics and meet the input requirements of emission models, the accuracy and reliability of emission assessment results could be effectively improved, and the effective coupling between microscopic simulation and emission models could be promoted. Summary of the Invention

[0007] To address the problems of unrealistic speed fluctuations, unreasonable acceleration and jerk values, and difficulty in meeting the input requirements of emission models in existing microscopic traffic simulations for emission assessment, this invention proposes a microscopic traffic simulation speed trajectory regeneration method that responds to emission assessment. By performing physical consistency correction on the second-by-second speed sequence output by the simulation, a vehicle trajectory usable for emission modeling is obtained. Based on measured data from floating cars, this invention constructs acceleration-jerk envelope models for different vehicle types; retains key characteristics of simulated vehicle operation, and considers kinematic consistency, dynamic controllability, jerk boundary, car-following safety, and travel boundary consistency constraints to construct a vehicle speed trajectory control model; utilizes SOS2 technology and a piecewise linear interpolation structure to linearize the nonlinear constraints in the vehicle speed trajectory control model; constructs a multi-objective function containing both speed tracking error and acceleration smoothing terms, designs a "relaxation-reinforcement" solution algorithm, and optimizes the output to obtain a speed trajectory sequence that responds to emission assessment and achieves physical consistency.

[0008] The technical solution of this invention is as follows: a microscopic traffic simulation speed trajectory regeneration method for response emission assessment, comprising the following steps: Step 1: Selecting an emission factor model; building a simulation environment using a microscopic traffic simulation platform, constructing a basic road network based on actual road structure, setting traffic demand and signal timing schemes, and outputting vehicle speed, lane, and position data per second; Step 2: Constructing acceleration-jerk envelope models for different vehicle types based on floating car data; Step 3: Retaining key characteristics of simulated vehicle operation, considering kinematic consistency, dynamic controllability, jerk boundary, car-following safety, and travel boundary consistency constraints, constructing a vehicle speed trajectory control model; Step 4: Using SOS2 (Special Ordered Set of Type 2) technology, employing a piecewise linear interpolation structure, linearizing the nonlinear constraints in the vehicle speed trajectory control model; Step 5: Constructing a multi-objective function containing two types of terms: speed tracking error and acceleration smoothing term, designing a "relaxation-reinforcement" solution algorithm, and optimizing the output to obtain a speed trajectory sequence that responds to emission assessment and has physical consistency.

[0009] Preferably, step 1 includes the following steps: Step 11: Selecting an emission factor model; selecting an instantaneous emission factor model that includes multiple vehicle types and multiple pollutants, wherein it includes at least three vehicle types: passenger cars, vans, and multi-purpose vehicles (MPVs), and at least includes PM2.5. 2.5 Two types of pollutants, NOx and H2O; Step 12: Use a microscopic traffic simulation platform to build a simulation operating environment, construct a basic road network based on the actual road structure, set traffic demand and signal timing schemes, and output vehicle speed, lane and location data per second.

[0010] Preferably, step 2 includes the following steps: Step 21: Floating car data preprocessing; acquiring vehicle trajectory data collected by the floating car, including timestamps, location coordinates, and speed information; for abrupt speed changes (such as points where the instantaneous speed increase exceeds 100 km / h), using cubic spline interpolation to fill in missing speed points, calculating the vehicle's instantaneous acceleration per second based on the difference method; for floating car data with inconsistent sampling time intervals (sampling intervals are usually between 1 and 5 seconds), using cubic splines to resample the speed sequence at a frequency of 1 Hz; Step 22: Jerk sequence calculation and anomaly handling; calculating the jerk value per second based on the acceleration sequence using the difference method, removing peak values ​​caused by noise during acceleration and jerk calculation, including filtering out abnormal jerk values ​​exceeding 99% and below 1% quantiles, and classifying and storing the jerk-acceleration data according to vehicle type; Step 23: Constructing a jerk-acceleration envelope fitting model for each vehicle type; calculating the jerk value at 5% of the acceleration dimension for each vehicle type's dataset. Using the 95th conditional quantile, the acceleration is discretized according to a preset interval, and the upper and lower quantile boundaries of jerk are calculated at each discrete acceleration point to form the minimum value. With the maximum value Two sets of points are fitted using piecewise linear, spline, or polynomial methods respectively, ensuring continuity at the segment points while obtaining a continuous jerk envelope fitting function; the upper and lower limit envelope fitting functions of jerk for passenger cars, MPVs, and vans are as follows: (1)-(6) (1) (2) (3) (4) (5) (6) , The distribution represents the fitting function of the upper and lower bounds of the jerk envelope of passenger vehicles. , The distribution represents the fitting function of the upper and lower bounds of the MPV jerk envelope. , The distribution represents the fitting function of the upper and lower bounds of the jerk envelope of the van.

[0011] Preferably, step 3 includes the following steps: Step 31: Kinematic consistency constraint; to ensure position ,speed With acceleration The inherent consistency of the three within the time discrete framework is expressed by the Euler update relation, as shown in formulas (7)-(8): (7) (8) Among them, Indicates the end time of the trip. The sampling step size, express Optimization speed at any time express Optimization acceleration at all times Indicates vehicle Optimization position at time; Step 32: Dynamic controllability constraints; Introduce jerk (denoted as ) as the control variable, the acceleration ( The evolution of ) and jerk ( Directly related, as shown in formula (9), global upper and lower bounds are set for speed and acceleration to meet road speed limits and vehicle capacity limitations, as shown in formulas (10)-(11). To avoid comfort and mechanical constraint problems caused by excessive jerk, a global bound is set for jerk, as shown in formula (12). (9) (10) (11) (12) , These represent the minimum and maximum values ​​of jerk, respectively. , Let represent the minimum and maximum values ​​of acceleration, respectively; Step 33: State-dependent jerk envelope constraint; Considering the significant asymmetry and vehicle type differences in the reachable range of jerk under different acceleration states, a local boundary that varies with acceleration is introduced based on the global boundary of formula (12), as shown in formula (13): (13) Step 34: Follow-up safety constraints; optimize the speed trajectory to meet the longitudinal safe distance throughout the entire time domain, using a unified constraint of "minimum safe distance" + "reaction time", as shown in formulas (14)-(15): (14) (15) Indicates the position of the vehicle in front. Indicates the position of the following vehicle. Indicates the minimum safe following distance. Indicates the driver's reaction time. Indicates the simulated speed of the vehicle. The simulated speed of the rear vehicle is represented; Step 35: Consistency constraints of the travel boundary; The travel start-end points and the reference trajectory must be consistent in speed and position. The Big-M method and 0–1 variables are used to construct open / closeable boundary equations, as shown in the formula: (16) (17) (18) (19) Among them, express Simulate speed in real time. express Simulate position at all times. Represents a very large positive integer. and It is a 0-1 parameter, where, express The optimized position of the vehicle at any given time is equal to the simulated position. express The optimized position of a vehicle at any given time is not the same as its simulated position. express The optimized vehicle speed at any given time equals the simulated speed. express The optimized speed of the vehicle at any given time is not equal to the simulated speed. (or When ), formulas (16) and (17) degenerate into (or );when (or When the constraint is "unlocked" by big-M, no forced alignment is applied at that moment.

[0012] In one embodiment of the present invention, .

[0013] Preferably, step 4 includes the following steps: Step 41: Define nodes; define acceleration intervals Discretize into K nodes: And pre-define the upper and lower limits of jerk on the nodes: and Step 42: Weight Combination; Introducing Weight Variables It satisfies formula (20); and conforms to The constraint states that at most two adjacent values ​​exist at any given time t. Both can be positive at the same time; , (20) Step 43: Perform linear interpolation using weights, as shown in formulas (21)-(23): (twenty one) (twenty two) (twenty three).

[0014] Preferably, step 5 includes the following steps: Step 51: Divide the simulated output speed trajectory into multiple continuous running strokes, using the zero speed point as the boundary; specifically, any trajectory stroke must ensure that the starting speed is at least 0 or the ending speed is 0, the speed interval has no zero point, and the running stroke is the smallest calculation unit for trajectory optimization. To ensure consistency with the simulated reference speed sequence while obtaining a smoother speed trajectory, the sampling step size is set to... Discrete time And constructed two objective functions including velocity tracking error and acceleration smoothing term, as shown in equation (24): (twenty four) Represents the simulation output The velocity at any given moment is used as the reference trajectory velocity. express Optimization speed at any time express Optimization acceleration at any moment; The optimization weights represent acceleration. The optimization weights representing speed, Indicates the end time of the trip. and The weights of speed tracking and smoothing penalty are respectively represented; Step 52: The "relaxation-reinforcement" solution algorithm is adopted: First, relaxation calculation is performed without applying the jerk envelope, that is, the constraint formula (13) is not considered, and a feasible and relatively smooth trajectory is obtained; Then, the result is used as a reference input and filled into the optimization model containing the jerk envelope constraint formulas (20)-(23); Step 53: The solver is called and the optimized output is obtained to obtain a speed trajectory sequence that conforms to physical consistency; Based on the emission factor model, the speed trajectory regeneration in response to emission assessment is realized by matching the corresponding emission factor according to the instantaneous speed-acceleration.

[0015] Compared with existing technologies, the present invention offers the following advantages: Based on measured data, the present invention constructs an acceleration-jerk envelope and, combined with constraints such as kinematic consistency, dynamic controllability, and car-following safety, regenerates the simulated trajectory, ensuring that the generated velocity sequence is realistically achievable, smooth, and meets safety requirements. Through optimized reconstruction of the velocity-acceleration sequence, the second-by-second trajectory generated by the present invention accurately reflects the vehicle's dynamic state, solving the problem that existing simulated trajectories are difficult to use for emissions assessment and providing reliable input for emissions models. Attached Figure Description

[0016] Figure 1 is a general flowchart of the present invention; Figure 2 shows the calculation of NOx and PM using different methods in embodiments of the present invention. 2.5 Total emissions comparison chart; Figure 3 shows the calculation of NOx and PM using different methods in embodiments of the present invention. 2.5 Comparison chart of emission factors per vehicle kilometer. Detailed Implementation

[0017] The present invention provides a detailed description of a microscopic traffic simulation speed trajectory regeneration method for response emission assessment, in conjunction with the accompanying drawings and embodiments. The present invention is not limited to this single example; any changes, modifications, substitutions, combinations, or simplifications made that deviate from the spirit and principle of the present invention shall be considered equivalent substitutions and are included within the protection scope of the present invention.

[0018] Example 1: A microscopic traffic simulation speed trajectory regeneration method for response emission assessment, the specific steps are as follows: (as shown in Figure 1) Step 1: Select an emission factor model; build a simulation environment using a microscopic traffic simulation platform, construct a basic road network based on the actual road structure, set traffic demand and signal timing schemes, and output vehicle speed, lane, and position data per second; Step 2: Construct acceleration-jerk envelope models for different vehicle types based on floating car data; Step 3: Retain key characteristics of simulated vehicle operation, consider kinematic consistency, dynamic controllability, jerk boundary, car-following safety, and travel boundary consistency constraints, and construct a vehicle speed trajectory control model; Step 4: Use SOS2 technology and a piecewise linear interpolation structure to linearize the nonlinear constraints in the vehicle speed trajectory control model; Step 5: Construct a multi-objective function containing two types of terms: speed tracking error term and acceleration smoothing term, design a "relaxation-reinforcement" solution algorithm, and optimize the output to obtain a speed trajectory sequence that responds to emission assessment and has physical consistency.

[0019] Example 2: Based on Example 1, in step 1, an emission factor model is selected, and a simulation environment is built using a microscopic traffic simulation platform. Based on the actual road structure, a basic road network is constructed, traffic demand and signal timing schemes are set, and vehicle speed, lane, and location data are output per second. Specifically, the following steps are included: Step 11: Select an emission factor model; Select an instantaneous emission factor model that includes multiple vehicle types and multiple pollutants, including at least three vehicle types: passenger cars, vans, and multi-purpose vehicles (MPVs), and at least PM2.5 emissions. 2.5Two types of pollutants, NOx and H2O; Step 12: Use a microscopic traffic simulation platform to build a simulation operating environment, construct a basic road network based on the actual road structure, set traffic demand and signal timing schemes, and output vehicle speed, lane and location data per second.

[0020] The microscopic traffic simulation platform can be selected from SUMO, VISSIM, TESSNG, or AIMSUN.

[0021] Example 3: Based on Example 2, step 2 constructs acceleration-jerk envelope models for different vehicle types based on floating car data. Specifically, this includes the following steps: Step 21: Floating car data preprocessing; acquiring vehicle trajectory data collected by the floating car, including timestamps, location coordinates, and speed information; for abrupt speed changes (e.g., points with instantaneous speed increases exceeding 100 km / h), using cubic spline interpolation to fill in missing speed points; calculating the vehicle's instantaneous acceleration per second using the difference method; for floating car data with inconsistent sampling time intervals (typically between 1 and 5 seconds), resampling the speed sequence at a frequency of 1 Hz using cubic splines; Step 22: Jerk sequence calculation and anomaly handling; calculating the jerk value per second based on the acceleration sequence using the difference method; removing peak values ​​caused by noise during acceleration and jerk calculations, including anomalies in the 99th percentile and below the 1% quantile range. The values ​​are filtered, and the jerk-acceleration data are classified and stored according to vehicle type; Step 23: Construct a jerk-acceleration envelope fitting model for each vehicle type; Calculate the 5% and 95% conditional quantiles of jerk in the acceleration dimension for each vehicle type dataset, discretize the acceleration according to a preset interval, and find the upper and lower quantile boundaries of jerk at each discrete acceleration point to form the minimum value. With the maximum value Two sets of points are fitted using piecewise linear, spline, or polynomial methods respectively, ensuring continuity at the segment points while obtaining a continuous jerk envelope fitting function; the upper and lower limit envelope fitting functions of jerk for passenger cars, MPVs, and vans are as follows: (1)-(6) (1) (2) (3) (4) (5) (6) , The distribution represents the fitting function of the upper and lower bounds of the jerk envelope of passenger vehicles. , The distribution represents the fitting function of the upper and lower bounds of the MPV jerk envelope. , The distribution represents the fitting function of the upper and lower bounds of the jerk envelope of the van. Acceleration For values ​​greater than 4, take 4; for values ​​less than -4, take -4.

[0022] Example 4: Based on Example 3, step 3 retains the key characteristics of the simulated vehicle operation, considering kinematic consistency, dynamic controllability, jerk boundary, car-following safety, and travel boundary consistency constraints, to construct a vehicle speed trajectory control model, specifically including the following steps: Step 31: Kinematic consistency constraints; to ensure position ,speed With acceleration The inherent consistency of the three within the time discrete framework is achieved by adopting the most basic Euler update relation, as shown in formulas (7)-(8): (7) (8) express Optimization speed at any time express Optimization acceleration at all times Indicates vehicle Optimization position at time; Step 32: Dynamic controllability constraints; Explicitly control the rate of change of acceleration in the model, introducing jerk (denoted as...). ) as the control variable, the acceleration ( The evolution of ) and jerk ( Directly related, as shown in formula (9), global upper and lower bounds are set for speed and acceleration to meet road speed limits and vehicle capacity restrictions, as shown in formulas (10)-(11). In order to avoid comfort and mechanical constraint problems caused by excessive jerk, a global bound is also given for jerk, as shown in formula (12): (9) (10) (11) (12) , These represent the minimum and maximum values ​​of jerk, respectively. , Let represent the minimum and maximum values ​​of acceleration, respectively; Step 33: State-dependent jerk envelope constraint; Considering the significant asymmetry and vehicle type differences in the reachable range of jerk under different acceleration states, based on the global bound of formula (12), a local boundary that varies with the state is further introduced, as shown in formula (13): (13) Step 34: Follow-up safety constraints; optimize the speed trajectory to meet the longitudinal safe distance throughout the entire time domain, and adopt the unified form of minimum safe distance + reaction time distance constraint, as shown in formulas (14)-(15): (14) (15) Indicates the position of the vehicle in front. Indicates the position of the following vehicle. Indicates the minimum safe following distance. Indicates the driver's reaction time. Indicates the simulated speed of the rear vehicle; Step 35: Consistency constraints of the travel boundary; The travel start-end points and the reference trajectory must be consistent in speed and position. An open / closed boundary equation is constructed using big-M and 0–1 indicator variables, as shown in the formula: (16) (17) (18) (19) and It is a 0-1 parameter, where, express The optimized position of the vehicle at any given time is strictly equal to the simulated position. express The optimized position of the vehicle at any given time may not be the same as the simulated position. express The optimized vehicle speed at any given time is strictly equal to the simulated speed. express The optimized speed of the vehicle at any given time may not be equal to the simulated speed. (or When ), formulas (16) and (17) degenerate into (or );when (or When this happens, the constraints are "unbound" by big-M, and no forced alignment is applied at that moment. (Selection) .

[0023] Example 5: Based on Example 4, step 4 utilizes SOS2 technology and a piecewise linear interpolation structure to linearize the nonlinear constraints in the vehicle speed trajectory control model. Specifically, it includes the following steps: Step 41: Define nodes; define the acceleration interval... Discretize into K nodes: And pre-define the upper and lower limits of jerk on the nodes: and Step 42: Weight Combination; Introducing Weight Variables , satisfying the formula (20); The constraint states that at most two adjacent values ​​exist at any given time t. Both can be positive at the same time; , (20) Step 43: Perform linear interpolation using weights, as shown in formulas (21)-(23): (twenty one) (twenty two) (twenty three).

[0024] Example 6: Based on Example 5, step 5 constructs a multi-objective function containing two types of terms: speed tracking error and acceleration smoothing. A "relaxation-reinforcement" solution algorithm is designed to optimize the output, obtaining a speed trajectory sequence that meets emission assessment and physical consistency requirements. Specifically, the steps are as follows: Step 51: Divide the simulated output speed trajectory into multiple continuous running segments, using the zero-speed point as the boundary. Specifically, any segment must ensure at least a starting speed of 0 or an ending speed of 0, with no zero points in the interval. The running segment is the smallest computational unit for trajectory optimization. To ensure consistency with the simulated reference speed sequence while obtaining a smoother speed trajectory, the sampling step size is set to... Discrete time And constructed two objective functions including velocity tracking error and acceleration smoothing term, as shown in equation (24): (twenty four) Represents the simulation output The velocity at any given moment is used as the reference trajectory velocity. express Optimization speed at any time express Optimization acceleration at any moment; The optimization weights represent acceleration. The optimization weights representing speed, Indicates the end time of the trip. and The weights of speed tracking and smoothing penalty are respectively represented; Step 52: The "relaxation-reinforcement" solution algorithm is adopted: First, relaxation calculation is performed without applying the jerk envelope, that is, the constraint formula (13) is not considered, and a feasible and relatively smooth trajectory is obtained; Then, the result is used as a reference input and filled into the optimization model containing the jerk envelope constraint formulas (20)-(23); Step 53: The solver is called and the optimized output is obtained to obtain a speed trajectory sequence that conforms to physical consistency; Based on the emission factor model, the speed trajectory regeneration in response to emission assessment is realized by matching the corresponding emission factor according to the instantaneous speed-acceleration.

[0025] Example 7: To verify the applicability of the trajectory correction method proposed in this invention in emissions estimation, the estimation results of the method of this invention and the SUMO default scheme were compared and analyzed based on the measured data of an Unmanned Aerial Vehicle (UAV), as shown in Figures 2 and 3. The SUMO default scheme is the Wiedemann model shown in the figures. R-VST (re-generating vehicle speed trajectories) and R-VST relaxed are both methods proposed in this invention. The R-VST method is the first step result of the "relaxation-reinforcement" algorithm, while the R-VST relaxed method is the final result of the "relaxation-reinforcement" algorithm. Both methods have reference value.

[0026] The analysis covered NOx and PM. 2.5 Two typical pollutants are analyzed, focusing on both regional total emissions and vehicle-kilometer emission factors. The former reflects the overall scale of pollutant emissions in the study area or road segment, while the latter emphasizes the average emission intensity of vehicles and the calibration accuracy of the model. Joint analysis of these two indicators allows for a comprehensive evaluation of the performance of different methods in emission estimation, thereby validating the effectiveness and reliability of velocity trajectory regeneration technology at different scales.

[0027] Figure 2 shows the NOx and PM levels under different methods. 2.5 Total emissions comparison. The results show that the SUMO default scheme significantly overestimates both types of pollutants, with NOx total emissions being approximately 21.8% higher than the UAV measured values, and PM2.5... 2.5 The error is approximately 32.8% higher. After introducing the R-VST and R-VST relaxed methods, the total emission estimation is significantly improved: the error of R-VST is reduced to 8.5% and 12.3%, respectively, while the error of R-VST relaxed is 10.8% and 13.7%. Although R-VST relaxed is slightly inferior in overall accuracy, it still shows a significant improvement compared to the default SUMO scheme. Overall, the velocity trajectory regeneration method proposed in this invention can effectively alleviate the bias of simulation overestimation of total emissions, making the overall emission level closer to the measured benchmark, and verifying its applicability in regional or road segment emission assessment.

[0028] Figure 3 shows the NOx and PM levels under different methods. 2.5 Emissions per vehicle kilometer. Unlike total emissions, this indicator focuses more on reflecting the average emission intensity of vehicles. It can be seen that the default SUMO scheme also overestimates NOx and PM2.5 emissions. 2.5The deviations were 14.6% and 24.6%, respectively. Then, the R-VST method reduced the errors to 1.3% and 4.8%, demonstrating better accuracy; R-VST relaxed was slightly higher (3.7% and 6.2%), but still significantly better than the default SUMO scheme. Therefore, the trajectory regeneration method not only improves the total emission estimation but also effectively corrects the deviation of individual vehicle emission factors, making it more suitable for scenarios such as vehicle emission inventory construction and model calibration.

[0029] The above description is merely a description of preferred embodiments of this application and is not intended to limit the scope of this application in any way. Any changes or modifications made by those skilled in the art based on the above-disclosed technical content should be considered as equivalent and valid embodiments and fall within the scope of protection of the technical solution of this application.

Claims

1. A method for regenerating speed trajectories from microscopic traffic simulation in response to emissions assessment, characterized in that, Includes the following steps: Step 1: Select an emission factor model; Step 1: A simulation environment is built using a microscopic traffic simulation platform. Based on the actual road structure, a basic road network is constructed, traffic demand and signal timing schemes are set, and vehicle speed, lane, and position data are output per second. Step 2: Based on floating car data, acceleration-jerk envelope models for different vehicle types are constructed. Step 3: Key characteristics of simulated vehicle operation are preserved, and considering kinematic consistency, dynamic controllability, jerk boundary, car-following safety, and travel boundary consistency constraints, a vehicle speed trajectory control model is constructed. Step 4: Using SOS2 technology, a piecewise linear interpolation structure is used to linearize the nonlinear constraints in the vehicle speed trajectory control model. Step 5: A multi-objective function containing both speed tracking error and acceleration smoothing terms is constructed, and a "relaxation-reinforcement" solution algorithm is designed to optimize the output and obtain a speed trajectory sequence with response emission assessment and physical consistency.

2. The method for regenerating speed trajectories from microscopic traffic simulation in response to emission assessment as described in claim 1, characterized in that, Step 1 includes the following steps: Step 11: Select an emission factor model; Select an instantaneous emission factor model that includes multiple vehicle types and multiple pollutants, including at least three vehicle types: passenger cars, vans, and multi-purpose vehicles, and at least PM2.

5. 2.5 Two types of pollutants, NOx and H2O; Step 12: Use a microscopic traffic simulation platform to build a simulation operating environment, construct a basic road network based on the actual road structure, set traffic demand and signal timing schemes, and output vehicle speed, lane and location data per second.

3. The method for regenerating speed trajectories in microscopic traffic simulation based on emission assessment as described in claim 1, characterized in that, Step 2, The steps include: Step 21: Floating car data preprocessing; acquiring vehicle trajectory data collected by the floating car, including timestamps, location coordinates, and speed information; For speed mutation anomalies, cubic spline interpolation is used to fill in the missing speed points. The instantaneous acceleration of the vehicle is calculated second by second based on the difference method. For floating car data with inconsistent sampling time intervals, cubic splines are used to resample the speed sequence at a frequency of 1Hz. Step 22: Jerk sequence calculation and anomaly handling. The finite difference method is used to calculate the second-by-second Jerk value based on the acceleration sequence, eliminating peak values ​​caused by noise during the acceleration and Jerk calculation process, including filtering out abnormal Jerk values ​​above the 99th percentile and below the 1% percentile. The acceleration-jerk data is classified and stored according to vehicle type. Step 23: Construct a Jerk-acceleration envelope fitting model for each vehicle type; calculate the 5% and 95th conditional quantiles of Jerk in the acceleration dimension for each vehicle type dataset, discretize the acceleration according to a preset interval, and find the upper and lower quantile boundaries of Jerk at each discrete acceleration point to form the minimum value. With the maximum value Two sets of points are fitted using piecewise linear, spline, or polynomial methods respectively, ensuring continuity at the segment points while obtaining a continuous Jerk envelope fitting function; the upper and lower limit envelope fitting functions of acceleration for passenger cars, MPVs, and vans are as follows: (1)-(6): (1) (2) (3) (4) (5) (6) 、 The distribution represents the fitting function of the upper and lower bounds of the acceleration envelope of passenger vehicles. 、 The distribution represents the fitting function of the upper and lower bounds of the MPV acceleration envelope. 、 The distribution represents the fitting function of the upper and lower bounds of the acceleration envelope of the van.

4. The method for regenerating speed trajectories in microscopic traffic simulation based on emission assessment as described in claim 1, characterized in that, Step 3 includes the following steps: Step 31: Kinematic consistency constraints; to ensure position ,speed With acceleration The inherent consistency of the three within the time discrete framework is expressed by the Euler update relation, as shown in formulas (7)-(8): (7) (8) Indicates vehicle Optimization speed at any time Indicates vehicle Optimization acceleration at all times Indicates vehicle Optimization position at time; Step 32: Dynamic controllability constraints; Introduce jerk notation. As a control variable, acceleration The evolution of jerk Directly related, as shown in formula (9), global upper and lower bounds are set for speed and acceleration to meet road speed limits and vehicle capacity restrictions, as shown in formulas (10)-(11). To avoid comfort and mechanical constraint problems caused by excessive jerk, a global bound is set for jerk, as shown in formula (12): (9) (10) (11) (12) 、 These represent the minimum and maximum values ​​of jerk, respectively. 、 Let represent the minimum and maximum values ​​of acceleration, respectively; Step 33: State-dependent jerk envelope constraint; Considering the significant asymmetry and vehicle type differences in the reachable range of jerk under different acceleration states, a local boundary that varies with acceleration is introduced based on the global boundary of formula (12), as shown in formula (13): (13) Step 34: Follow-up safety constraints; optimize the speed trajectory to meet the longitudinal safe distance throughout the entire time domain, using a unified constraint of "minimum safe distance" + "reaction time", as shown in formulas (14)-(15): (14) (15) Indicates the vehicle in front Time and location Indicates the following vehicle Time and location Indicates the minimum safe following distance. Indicates the driver's reaction time. Indicates the following vehicle Simulate speed at all times; Step 35: Consistency constraints of travel boundaries; The travel start-end points and the reference trajectory are kept consistent in speed and position. The big-M method and 0-1 variables are used to construct open / close boundary equations, as shown in formulas (16)-(19): (16) (17) (18) (19) Indicates vehicle Simulate speed in real time. Indicates vehicle Simulate position at all times. Represents a very large positive integer. and It is a 0-1 parameter. express The optimized position of the vehicle at any given time is equal to the simulated position. express The optimized position of a vehicle at any given time is not the same as its simulated position. express The optimized vehicle speed at any given time equals the simulated speed. express The optimized speed of the vehicle at any given time is not equal to the simulated speed. or At that time, formulas (16) and (17) degenerate into or ;when or At that time, the constraint is "released" by big-M, and no forced alignment is imposed at that moment.

5. The method for regenerating speed trajectories in microscopic traffic simulation for response emission assessment as described in claim 4, characterized in that, Step 4 includes the following steps: Step 41: Define nodes; define acceleration intervals Discretize into K nodes: And pre-define the upper and lower limits of jerk on the nodes: and Step 42: Weight Combination; Introducing Weight Variables It satisfies formula (20) and conforms to The constraint states that at most two adjacent values ​​exist at any given time t. Both can be positive simultaneously; , (20) Step 43: Perform linear interpolation using weights, as shown in formulas (21)-(23): (21) (22) (23)。 6. The method for regenerating speed trajectories in microscopic traffic simulation based on emission assessment as described in claim 5, characterized in that, Step 5 The steps include: Step 51: Divide the simulated output speed trajectory into multiple continuous running strokes, using the zero speed point as the boundary; specifically, any trajectory stroke must ensure that the starting speed is at least 0 or the ending speed is 0, the speed interval has no zero point, and the running stroke is the smallest calculation unit for trajectory optimization. To ensure consistency with the simulation reference speed sequence while obtaining a smoother speed trajectory, the sampling step size is set to... Discrete time And construct an objective function that includes two types of terms: velocity tracking error and acceleration smoothing term, as shown in equation (24): (24) Represents the simulation output The velocity at any given moment is used as the reference trajectory velocity. express Optimization speed at any time express Optimization acceleration at any moment; The optimization weights represent acceleration. The optimization weights representing speed, Indicates the end time of the trip. and The weights of speed tracking and smoothing penalty are respectively represented; Step 52: The "relaxation-reinforcement" solution algorithm is adopted: First, relaxation calculation is performed without applying the jerk envelope, that is, the constraint formula (13) is not considered, and a feasible and relatively smooth trajectory is obtained; Then, the result is used as a reference input and filled into the optimization model containing the jerk envelope constraint formulas (20)-(23); Step 53: The solver is called and the optimized output is obtained to obtain a speed trajectory sequence that conforms to physical consistency; Based on the emission factor model, the corresponding emission factor is matched according to the instantaneous speed-acceleration to realize the speed trajectory regeneration in response to emission assessment.

Images