A method for evaluating actual pollution reduction benefits of plug-in hybrid electric vehicles considering user behavior

Abstract

The present application belongs to the field of green traffic, and particularly relates to a method for evaluating actual pollution reduction benefits of plug-in hybrid vehicles considering user behavior. The method comprises the following steps: step 1: collecting actual road emission data of the plug-in hybrid vehicle; obtaining user trip data of the plug-in hybrid vehicle through field research; step 2: constructing an emission factor matrix of the plug-in hybrid vehicle responding to SOC from two dimensions of speed-acceleration joint distribution and average vehicle speed; step 3: characterizing user charging characteristics and constructing a multi-level layered trip intensity matrix responding to user behavior; step 4: jointly calculating the trip emission amount of the plug-in hybrid vehicle considering user behavior by using the emission factor matrix and the daily trip intensity matrix; and step 5: calculating the total exhaust pollutant emission amount of the plug-in hybrid vehicle. The present application comprehensively considers the diversity of energy management strategies of the plug-in hybrid vehicle and the diversification of user behavior, and realizes quantitative calculation of exhaust emission amount of the plug-in hybrid vehicle in the actual use process.

Description

Technical Field

[0001] This invention belongs to the field of green transportation, and specifically relates to a method for evaluating the actual pollution reduction and emission reduction benefits of plug-in hybrid vehicles that takes into account user behavior. Background Technology

[0002] With increasingly cleaner electricity sources, promoting plug-in hybrid electric vehicles (PHEVs) has become an important option for mitigating greenhouse gas and air pollution from the global road transport sector. Globally, governments have widely implemented policies to promote the rapid growth of PHEV ownership. However, the development of charging infrastructure in many countries lags behind, making it difficult to meet the increasing charging demand in the short term. Furthermore, many consumers purchase PHEVs primarily to take advantage of tax breaks and road toll incentives, even without regular charging access. This leads to the underutilization of the electric drive ratio in PHEVs and its associated environmental benefits. Therefore, research on the actual environmental benefits of PHEVs considering user behavior (such as charging frequency, driving range, and driving time) is of great significance.

[0003] Currently, the emission reduction benefits of plug-in hybrid vehicles (PHEVs) are mostly based on laboratory tests, without considering actual charging frequency and driving intensity, and verification in real-world driving is limited. There are two main reasons for this: first, the complexity of PHEV energy management strategies; and second, the frequent changes in power distribution between the engine and electric motor, frequent engine restarts, and fluctuations in catalytic converter performance, all of which complicate the exhaust emission mechanisms of PHEVs. First, based on the battery's state of charge (SOC), PHEVs are designed to operate in two driving modes: Discharge Mode (CD) and Charge Maintenance Mode (CS). In CD mode, the electric motor primarily provides power to the vehicle, and the engine remains off during low power demands (such as idling or light acceleration), only starting when higher power is required. Once the SOC reaches a predetermined threshold, the vehicle switches to CS mode to maintain the SOC and prevent further discharge. Second, CO, HC, and NOx emissions in the exhaust are primarily affected by the operating status of the internal combustion engine, rather than total fuel consumption. Both CO and HC originate from incomplete combustion. CO production primarily depends on oxygen availability, while HC emissions often increase under partial load conditions due to low combustion efficiency. Frequent engine restarts and vehicle speed fluctuations, especially in urban driving, further exacerbate CO and HC emissions. NOx emissions are generated under high temperature and pressure, typically occurring during cold starts and sustained high-temperature phases. Although catalytic converters play a role in emission reduction, their efficiency is unstable, and a significant portion of NOx may still be released into the atmosphere. Furthermore, CO2 comes from complete combustion, and emissions are directly proportional to fuel consumption. In CS mode, the continuous operation of the internal combustion engine not only meets the increased power demand but also maintains a high-temperature and high-pressure environment, contributing to complete fuel combustion and leading to increased CO2 and NOx emissions. Numerous studies have shown that under certain conditions, such as high-speed driving or rapid acceleration, the emissions of plug-in hybrid vehicles in CS mode may exceed those in CD mode, and may even exceed the emissions of comparable gasoline vehicles. Therefore, considering the diverse driving modes and varied user behaviors of plug-in hybrid vehicles, it is urgent to establish a method for assessing the actual environmental benefits of plug-in hybrid vehicles that takes into account user behavior, in order to verify the pollution reduction and emission reduction benefits of plug-in hybrid vehicles. Summary of the Invention

[0004] To address the problems existing in the prior art, this invention proposes to consider the energy management strategy characteristics of plug-in hybrid vehicles. Based on actual road driving emission data, an emission factor matrix for plug-in hybrid vehicles responding to State of Charge (SOC) is established from two dimensions: the joint distribution of speed and acceleration and the average vehicle speed. By utilizing user trip data of plug-in hybrid vehicles obtained through field surveys, a multi-level hierarchical trip intensity matrix responding to user behavior is constructed. By combining the emission factor matrix and the trip intensity matrix, the trip emissions of plug-in hybrid vehicles considering user behavior are calculated.

[0005] Technical solution of the present invention:

[0006] A method for evaluating the actual pollution reduction and emission reduction benefits of plug-in hybrid vehicles that takes user behavior into account, with the following specific steps:

[0007] Step 1: Collect actual road emissions data of plug-in hybrid vehicles; obtain user trip data of plug-in hybrid vehicles through field surveys;

[0008] The actual road emissions data for the plug-in hybrid vehicle specifically includes: vehicle speed, acceleration, latitude and longitude, altitude, and the emissions of carbon dioxide (CO2), carbon monoxide (CO), hydrocarbons (HC), and nitrogen oxides (NO) in the exhaust. x Emission coefficients, engine speed, and vehicle SOC value were collected with a data acquisition accuracy of 1Hz.

[0009] The trip data of plug-in hybrid vehicle users obtained through field research includes: trip timestamps, SOC value, cumulative mileage, vehicle operating status (pure electric, fuel, hybrid) and latitude and longitude, and the data is collected at intervals of no less than 30 seconds.

[0010] Step 2: Construct the emission factor matrix of the plug-in hybrid vehicle in response to SOC from two dimensions: the joint distribution of velocity-acceleration and the average vehicle speed.

[0011] Step 3: Characterize user charging characteristics and construct a multi-level hierarchical trip intensity matrix to respond to user behavior;

[0012] Step 4: Combine the emission factor matrix and the daily travel intensity matrix to calculate the travel emissions of plug-in hybrid vehicles that take into account user behavior;

[0013] Step 5: Using "maximizing new energy" and "de-emphasis on new energy" as two baseline scenarios and actual user driving of plug-in hybrid vehicles as the real scenario, calculate the total exhaust pollutant emissions of plug-in hybrid vehicles under the three scenarios, and quantitatively evaluate the environmental benefits of actual pollution reduction and emission reduction of plug-in hybrid vehicles.

[0014] Preferably, step 1 includes the following steps:

[0015] Step 11: Select exhaust emission data monitoring equipment for plug-in hybrid vehicles, including: Portable Emissions Measurement System (PEMS), On-Board Diagnostics (OBD), environmental sensors, BeiDou positioning satellite system, and action camera;

[0016] PEMS is responsible for recording timestamps, greenhouse gases and air pollutants (carbon dioxide CO2, carbon monoxide CO, hydrocarbons HC, nitrogen oxides NO) in exhaust gases.x The system collects data on the volume concentration and exhaust gas volume flow rate of the vehicle. The OBD system is responsible for collecting data on the engine speed and vehicle speed of the plug-in hybrid vehicle. The Beidou positioning satellite system is responsible for collecting information on vehicle speed, vehicle location (latitude and longitude), and altitude. The environmental sensor records the temperature and dew point in the atmospheric environment. The action camera is responsible for capturing information on the vehicle's dashboard during the experiment, including SOC value, timestamp, and vehicle speed.

[0017] Step 12: Design a real-world emission monitoring test scheme for plug-in hybrid vehicles; the number of real-world emission monitoring tests for plug-in hybrid vehicles shall not be less than 4 times per vehicle, with each real-world driving time not less than 90 minutes, and the cumulative driving time not less than 10 hours per vehicle. Among these tests, the cumulative driving time on highways shall not be less than 3 hours and the cumulative mileage shall not be less than 120 kilometers, and the cumulative driving time on non-highway roads shall not be less than 6 hours and the cumulative mileage shall not be less than 150 kilometers. In the at least 4 real-world emission monitoring tests, the initial SOC of the vehicle shall not be less than 60% in at least two of them.

[0018] Step 13: Preprocessing and outlier removal of experimental data for real-world emissions from plug-in hybrid vehicles; Based on statistical principles, the interquartile range (interquartile range) is used to locate outliers in the experimental data, and Nan values ​​are used to replace them. Considering the data characteristics of the experimental parameters, the upper or lower bounds of the interquartile range are modified; Appropriate smoothing parameters are selected, and smoothing curves are used to analyze CO2, CO, HC, and NO in the exhaust gas. x The emission volume concentration, exhaust gas volume flow rate, BeiDou vehicle speed, and OBD vehicle speed are smoothed to improve data quality; according to the national standard "Limits and Measurement Methods for Pollutant Emissions from Light-Duty Vehicles (China VI)" (GB18352.6), altitude is corrected, and the concentrations of CO2, CO, HC, and NO in the exhaust gas are calculated. x The emission factor is expressed in g / s; the Paddle deep learning framework is used to identify information such as vehicle speed, SOC value, and time in the captured video.

[0019] Step 14: Alignment processing of experimental data for actual road emissions of plug-in hybrid vehicles; Considering that PMES, OBD, and action camera in the actual road emissions monitoring equipment for plug-in hybrid vehicles are independent of each other, and that the environmental sensor and Beidou positioning satellite system rely on PEMS for power supply and communication, time alignment is performed according to "PMES time - dashboard time", and vehicle speed alignment is performed according to "Beidou vehicle speed - OBD vehicle speed - dashboard vehicle speed" to ensure data alignment between different devices;

[0020] Step 15: Map matching based on Hidden Markov Model (HMM) and Viterbi algorithm; determine the geographic coordinate system according to the actual road emission monitoring test location of the plug-in hybrid vehicle, and import the road data and map files (such as shapefile files) of the city area where the test is located; encode the map data based on a binary tree structure to reduce the search space of the Markov model; map each trajectory point to a potential location on the road centerline based on the principle of minimizing the vertical projection distance; set the matching parameters of the Hidden Markov Model, call the Viterbi algorithm to perform map matching, and output the road name and road type corresponding to each trajectory point of the vehicle;

[0021] Step 16: Sub-trip segmentation; Based on the output of Step 15, the road types mentioned above are divided into highways and non-highway roads; Using timestamps as features and vehicle speed as the target variable, a decision tree regression model is constructed, and the predicted speed change locations (or timestamps) are calculated; By setting thresholds, change points are filtered and merged. If the difference between the mean speeds before and after a change point exceeds the threshold, the change point is marked. If the mean speeds before and after a change point are both below 3 or the distance between them is less than 400, the change point is marked as invalid; For each experimental data, the vehicle speed column data is traversed to identify segments with speeds below 5, and the start and end timestamps of these low-speed segments are recorded; Change points of the trip and road type are checked, new change points are added, and finally, a trip is divided into multiple sub-trips, distinguishing between high-speed trips and non-high-speed trips.

[0022] Preferably, step 2 constructs the emission factor matrix of the plug-in hybrid vehicle in response to SOC from two dimensions: the joint distribution of velocity-acceleration and the average vehicle speed, and includes the following steps:

[0023] Step 21: Construct the plug-in hybrid vehicle emission factor matrix E based on the velocity-acceleration joint distribution and the state of charge (SOC). p (v i ,a j ,m k ),as follows:

[0024] The actual road data of all plug-in hybrid vehicles processed in step 1 were divided into speed intervals of 1 km / h and acceleration intervals of 0.1 m / s², and a speed v was established. i In the range of 0–100 km / h, the acceleration a j The combined velocity-acceleration distribution in the range of -4 to 4 m / s²;

[0025] Based on the vehicle's State of Charge (SOC) value, the driving modes of plug-in hybrid vehicles are divided into two types: energy consumption mode (CD) and energy sustaining mode (CS), and different driving modes are constructed m kThe combined velocity-acceleration distribution was analyzed to calculate the CO2, CO, HC, and NO content in the exhaust gas of each velocity-acceleration group under both CD and CS modes. x The average emission factor, in g / s;

[0026] The emission factor matrix of the plug-in hybrid vehicle based on the joint velocity-acceleration distribution of the state of charge (SOC) is constructed as follows:

[0027]

[0028] v = (0, ..., 100) T

[0029] a = (-4, ..., 4) T

[0030]

[0031] Where speed bin represents the velocity range, acc bin represents the acceleration range, ceil(v) and ceil(10*a) represent rounding up to multiples of 10 for velocity and acceleration, respectively, and m represents the drive mode, m∈{CD,CS}. Indicates velocity v i acceleration a j and drive mode m k The emission factor of pollutant p, expressed in g / s, E p (v i ,a j ,m k ) represents the emission factor matrix of pollutant p.

[0032] Step 22: Construct the emission factor matrix of plug-in hybrid vehicles based on the State of Charge (SOC) of average vehicle speed;

[0033] Based on vehicle speed, the range is divided into two groups: 0-2.5 km / h and 2.5 km / h-102.5 km / h, with intervals of 5 km / h, namely 2.5-7.5 km / h, 7.5-12.5 km / h, 12.5-17.5 km / h, ..., 97.5-102.5 km / h.

[0034] Multiply the plug-in hybrid vehicle emission factor matrix based on the joint speed-acceleration distribution output in step 21 with the speed-acceleration distribution matrix within each vehicle speed group to obtain the plug-in hybrid vehicle emission factor matrix based on the average vehicle speed, as follows:

[0035]

[0036] Where avg speed bin represents the average vehicle speed range. Indicates average vehicle speed. Indicates average vehicle speed velocity v within the interval i acceleration a j frequency, Indicates based on average vehicle speed Velocity-acceleration distribution matrix

[0037]

[0038] Where ° represents matrix dot product, This represents the p-emission factor matrix of pollutants from plug-in hybrid vehicles based on the State of Charge (SOC) at average vehicle speed. They represent average vehicle speeds. Plug-in hybrid vehicle drive mode m k Emission factor of pollutant p p (v i ,a j ,CD) and E p (v i ,a j ,CS) represent E p (v i ,a j ,m k In the matrix m k A submatrix whose columns are equal to CD or CS.

[0039] Step 3: Characterize user charging characteristics and construct a multi-level hierarchical trip intensity matrix to respond to user behavior;

[0040] The user behaviors include: the user's daily charging frequency and daily driving mileage.

[0041] Step 3 specifically includes the following steps:

[0042] Step 31: Collect and improve user trip data for each hour: Obtain trip data of plug-in hybrid vehicle users through on-site surveys, and supplement the data with the cumulative mileage and SOC value of users for each 24 hours of the day through interpolation; provide the improved user trip data to Step 32.

[0043] Step 32: Based on the improved user trip data, calculate the user's daily mileage and average speed using timestamps and cumulative mileage data; calculate the user's average daily mileage for a single user, thereby determining the user's average daily mileage group label;

[0044] Preferably, the average daily mileage is divided into mileage ranges of 0-30km, 30-90km, 90-210km, and greater than 210km at equal intervals of 3km, 6km, 12km, and 24km, respectively, and the average driving speed is grouped according to step 22, i.e., avgspeed bin.

[0045] Furthermore, plug-in hybrid vehicles with a battery range of 80km or more are classified as long-range plug-in hybrid vehicles, while the rest are classified as short-range plug-in hybrid vehicles. User trip data is grouped, and the average daily mileage group label is calculated for each group.

[0046] Step 33: Determine whether the user is charging the vehicle by using the vehicle's location latitude and longitude information and SOC value changes; improve the user's daily charging information and characterize the user's average daily charging frequency characteristics by grouping them into long cruising and short range categories.

[0047] Specifically, if the change in the vehicle's longitude and latitude is less than or equal to one millionth of a degree for more than 15 consecutive minutes and the SOC value gradually increases, the vehicle is considered to be charging; otherwise, it is not charging.

[0048] Step 34: Establish a SOC distribution matrix based on average daily mileage; based on the improved user trip data, group and label the hybrid vehicle drive mode status by average daily mileage and distinguish it by hour, i.e. CD mode or CS mode.

[0049] Step 35: Group by range and obtain the mileage-speed distribution matrix in response to user behavior.

[0050] Preferably, step 4, which combines the emission factor matrix and the daily travel intensity matrix to calculate the travel emissions of the plug-in hybrid vehicle considering user behavior, includes the following steps:

[0051] Step 41: Calculate the emissions of the plug-in hybrid vehicle considering user behavior; based on the plug-in hybrid vehicle emission factor matrix with response SOC based on average vehicle speed output in Step 2, multiply it simultaneously with the vehicle speed distribution matrix in response to user behavior in Step 3 to calculate CO2, CO, HC, and NO in the exhaust gas of the plug-in hybrid vehicle during the user's daily journey. x The emissions from the journey are as follows:

[0052] The vehicle speed distribution matrix S in response to user behavior, and the hourly average vehicle speed matrix S for plug-in hybrid vehicles i grouped by range n on driving day d. n，i，d For example, S (h) n，i，d (h) is a submatrix of S, as follows:

[0053]

[0054] Where i represents a plug-in hybrid vehicle; n represents the range group number of the plug-in hybrid vehicle, n∈{N1, N2}, where N1 and N2 represent the short-range and long-range plug-in hybrid vehicle groups, respectively; d represents the driving day; m h Indicates the drive mode for hour h, m h ∈{CD, CS}; S n，i，d (h) represents the hourly average vehicle speed matrix of plug-in hybrid vehicle i with range group n on driving day d; h represents the hour (0 to 23), v h V represents the vehicle speed in h hours. h The speed group label for h hours corresponds to avg speedbin in step 22;

[0055]

[0056] in, Indicates based on shared columns (specifically S) n，i，d V in matrix (h) h and m h ,and In the matrix and m k This involves merging the vehicle speed grouping and driving mode (represented by the matrix S). The specific steps are as follows: First, for matrix S... n，i，d Each row (h, v) in (h) h V h m h Search Does a matching V exist in? h and m h Secondly, if a matching V exists h and m h Then the corresponding emission factors are combined. If there is no matching V h and m h If the value is 0, then fill the row with 0; U n，i，d，p This represents the pollutant emission matrix for user i of plug-in hybrid vehicle with range group n during driving day d.

[0057] Preferably, step 5 includes the following steps:

[0058] Step 51: Under the "maximizing new energy" baseline scenario (S1, scenario 1), the plug-in hybrid vehicle maintains CD mode throughout the entire journey; under the "de-energization" baseline scenario (S2, scenario 2), the plug-in hybrid vehicle maintains CS mode throughout the entire journey. Taking the actual driving of the plug-in hybrid vehicle as scenario 3 (S3), calculate the CO2, CO, HC, and NO emissions from the exhaust gas of the plug-in hybrid vehicle under the three scenarios. x Total emissions, to assess the actual environmental benefits of pollution reduction and emission reduction of plug-in hybrid vehicles, are as follows:

[0059] The baseline scenario for "maximizing new energy" (Scenario 1, S1) is the user's hourly average vehicle speed matrix S. n，i，d (h) in m h Set all to CD mode, that is In the baseline scenario of "de-energization" (Scenario 2, S2), the user's hourly average vehicle speed matrix S n，i，d (h) in m h Set all to CS mode, that is

[0060] Therefore, the user trip emissions are calculated as follows under scenarios S1 and S2:

[0061]

[0062] Among them, U′ n，i，d，p and U″ n，i，d，p The table shows the pollutant p emission matrix for plug-in hybrid vehicle i with range group n under scenarios S1 and S2, respectively, during the trip on driving day d.

[0063] The total emissions calculations for S1, S2, and the actual scenario for plug-in hybrid vehicles are as follows:

[0064] total S1，p =∑ n ∑ i ∑ d U′ n，i，d，p

[0065] total S2，p =∑ n ∑ i ∑ d U″ n，i，d，p

[0066] total S3，p =∑ n ∑ i ∑ d U n，i，d，p

[0067] Among them, total S1，p total S2,p and total actual，p These represent the total pollutant emissions p from plug-in hybrid vehicles under scenarios S1, S2, and S3, respectively.

[0068] Compared with the prior art, the present invention has the following advantages:

[0069] This invention utilizes real-world road emission data from plug-in hybrid vehicles (PHEVs) to establish a multi-dimensional emission factor matrix for PHEVs that responds to State of Charge (SOC). It also utilizes user trip data from field surveys to establish a multi-level, hierarchical trip intensity matrix that responds to user behavior. By combining the emission factor matrix and the daily trip intensity matrix, the invention calculates the PHEV trip emissions considering user behavior. Furthermore, through baseline scenarios of "maximizing new energy" and "de-reliance on new energy," the invention quantitatively evaluates the actual pollution reduction and emission reduction benefits of PHEVs. This invention integrates user behavior, vehicle characteristics, and emission generation mechanisms from multiple dimensions, overcoming the limitations of traditional laboratory testing. It provides a highly adaptable solution for calculating the trip emissions of PHEVs, thereby quantitatively evaluating the environmental benefits of PHEVs in daily use. This provides important technical support for environmental management policy formulation, the promotion of green travel, and the optimized design of PHEVs. Attached Figure Description

[0070] Figure 1 This is the overall flowchart of the present invention. Detailed Implementation

[0071] The accompanying drawings and embodiments provide a detailed description of the method for calculating the range emissions of plug-in hybrid vehicles with real-world SOC. This invention is not limited to this single instance; any changes, modifications, substitutions, combinations, or simplifications made that depart from the spirit and principle of this invention should be considered equivalent replacements and are included within the scope of protection of this invention.

[0072] Example 1:

[0073] A method for evaluating the actual pollution reduction and emission reduction benefits of plug-in hybrid vehicles that takes into account user behavior includes the following steps:

[0074] Step 1: Design an experimental plan, select experimental equipment, and collect actual road emission data of plug-in hybrid vehicles; obtain user trip data of plug-in hybrid vehicles through field surveys;

[0075] Step 2: Construct the emission factor matrix of the plug-in hybrid vehicle in response to SOC from two dimensions: the joint distribution of velocity-acceleration and the average vehicle speed.

[0076] Step 3: Characterize user charging characteristics and construct a multi-level hierarchical trip intensity matrix to respond to user behavior;

[0077] Step 4: Combine the emission factor matrix from Step 2 and the daily travel intensity matrix from Step 3 to calculate the travel emissions of plug-in hybrid vehicles considering user behavior;

[0078] Step 5: Set up "maximizing new energy" and "de-emphasis on new energy" as two baseline scenarios respectively. Using actual user driving of plug-in hybrid vehicles as the real scenario, calculate the total exhaust pollutant emissions of plug-in hybrid vehicles under the three scenarios, and quantitatively evaluate the environmental benefits of actual pollution reduction and emission reduction of plug-in hybrid vehicles.

[0079] Example 2:

[0080] Based on Example 1, step 1 involves designing an experimental scheme, selecting experimental equipment, and collecting actual road emission data from the plug-in hybrid vehicle. Specifically, this includes: vehicle speed, acceleration, latitude and longitude, altitude, and emissions of carbon dioxide (CO2), carbon monoxide (CO), hydrocarbons (HC), and nitrogen oxides (NOx) in the exhaust gas. x Emission coefficients, engine speed, and vehicle SOC values ​​are collected with a data acquisition accuracy of 1Hz. Trip data from plug-in hybrid vehicle users is obtained through field surveys, including trip timestamps, SOC values, cumulative mileage, and vehicle operating status (pure electric, gasoline, hybrid). Data collection is performed at intervals of at least 30 seconds, and includes the following steps:

[0081] Step 11: Select exhaust emission data monitoring equipment for plug-in hybrid vehicles, including Portable Emissions Measurement System (PEMS), On-Board Diagnostics (OBD), environmental sensors, BeiDou positioning satellite system, and action camera; among them, PEMS is responsible for recording timestamps, greenhouse gases and air pollutants (carbon dioxide CO2, carbon monoxide CO, hydrocarbons HC, nitrogen oxides NO) in the exhaust gas. x The system collects data on the volume concentration and exhaust gas volume flow rate of the vehicle. The OBD system is responsible for collecting data on the engine speed and vehicle speed of the plug-in hybrid vehicle. The Beidou positioning satellite system is responsible for collecting information on vehicle speed, vehicle location (latitude and longitude), and altitude. The environmental sensor records the temperature and dew point in the atmospheric environment. The action camera is responsible for capturing information on the vehicle's dashboard during the experiment, including SOC value, timestamp, and vehicle speed.

[0082] In this embodiment, step 11 selects the equipment for the actual road driving emission monitoring experiment of the vehicle. The specific equipment selection and data acquisition in this embodiment are shown in Table 1.

[0083] Table 1 is a summary table of experimental equipment and data content.

[0084]

[0085]

[0086] Step 12: Design a real-world emission monitoring test scheme for plug-in hybrid vehicles; the number of real-world emission monitoring tests for plug-in hybrid vehicles shall not be less than 4 times per vehicle, with each real-world driving time not less than 90 minutes, and the cumulative driving time not less than 10 hours per vehicle. Among these tests, the cumulative driving time on highways shall not be less than 3 hours and the cumulative mileage shall not be less than 120 kilometers, and the cumulative driving time on non-highway roads shall not be less than 6 hours and the cumulative mileage shall not be less than 150 kilometers. In the at least 4 real-world emission monitoring tests, the initial SOC of the vehicle shall not be less than 60% in at least two of them.

[0087] Step 13: Preprocessing and outlier removal of experimental data for real-world emissions from plug-in hybrid vehicles; Based on statistical principles, the interquartile range (interquartile range) is used to locate outliers in the experimental data, and Nan values ​​are used to replace them. Considering the data characteristics of the experimental parameters, the upper or lower bounds of the interquartile range are modified; Appropriate smoothing parameters are selected, and smoothing curves are used to analyze CO2, CO, HC, and NO in the exhaust gas. x The emission volume concentration, exhaust gas volume flow rate, BeiDou vehicle speed, and OBD vehicle speed are smoothed to improve data quality; according to the national standard "Limits and Measurement Methods for Pollutant Emissions from Light-Duty Vehicles (China VI)" (GB18352.6), altitude is corrected, and the concentrations of CO2, CO, HC, and NO in the exhaust gas are calculated. x The emission factor is expressed in g / s; the Paddle deep learning framework is used to identify information such as vehicle speed, SOC value, and time in the captured video.

[0088] Step 14: Alignment processing of experimental data for actual road emissions of plug-in hybrid vehicles; Considering that PMES, OBD, and action camera in the actual road emissions monitoring equipment for plug-in hybrid vehicles are independent of each other, and that the environmental sensor and Beidou positioning satellite system rely on PEMS for power supply and communication, time alignment is performed according to "PMES time - dashboard time", and vehicle speed alignment is performed according to "Beidou vehicle speed - OBD vehicle speed - dashboard vehicle speed" to ensure data alignment between different devices;

[0089] Step 15: Map matching based on Hidden Markov Model (HMM) and Viterbi algorithm; determine the geographic coordinate system according to the actual road emission monitoring test location of the plug-in hybrid vehicle, and import the road data and map files (such as shapefile files) of the city area where the test is located; encode the map data based on a binary tree structure to reduce the search space of the Markov model; map each trajectory point to a potential location on the road centerline based on the principle of minimizing the vertical projection distance; set the matching parameters of the Hidden Markov Model, call the Viterbi algorithm to perform map matching, and output the road name and road type corresponding to each trajectory point of the vehicle;

[0090] Step 16: Sub-trip segmentation; Based on the output of Step 15, the road types mentioned above are divided into highways and non-highway roads; Using timestamps as features and vehicle speed as the target variable, a decision tree regression model is constructed, and the predicted speed change locations (or timestamps) are calculated; By setting thresholds, change points are filtered and merged. If the difference between the mean speeds before and after a change point exceeds the threshold, the change point is marked. If the mean speeds before and after a change point are both below 3 or the distance between them is less than 400, the change point is marked as invalid; For each experimental data, the vehicle speed column data is traversed to identify segments with speeds below 5, and the start and end timestamps of these low-speed segments are recorded; Change points of the trip and road type are checked, new change points are added, and finally, a trip is divided into multiple sub-trips, distinguishing between high-speed trips and non-high-speed trips.

[0091] Example 3:

[0092] Based on Example 2, step 2 constructs the emission factor matrix of the plug-in hybrid vehicle in response to SOC from multiple dimensions, including the joint distribution of speed and acceleration and the average vehicle speed. Specifically, the steps are as follows:

[0093] Step 21: Construct the emission factor matrix of plug-in hybrid vehicles with joint speed-acceleration distribution of response SOC;

[0094] The actual road data of all plug-in hybrid vehicles processed in step 1 were divided into speed intervals of 1 km / h and acceleration intervals of 0.1 m / s², and a speed v was established. i In the 0-100km / h range, acceleration a j The combined velocity-acceleration distribution in the range of -4 to 4 m / s²;

[0095] Based on the vehicle's State of Charge (SOC) value, the driving modes of plug-in hybrid vehicles are divided into two types: energy consumption mode (CD) and energy sustaining mode (CS), and different driving modes are constructed m k The combined velocity-acceleration distribution was analyzed to calculate the CO2, CO, HC, and NO content in the exhaust gas of each velocity-acceleration group under both CD and CS modes. x The average emission factor, in g / s;

[0096] Construct the emission factor matrix E of plug-in hybrid vehicles with joint speed-acceleration distribution of response SOC. p (v i a j m k ),as follows:

[0097]

[0098] v = (0, ..., 100) T

[0099] a = (-4, ..., 4) T

[0100]

[0101] Where speed bin represents the velocity range, aCc bin represents the acceleration range, ceil(v) and ceil(10*a) represent rounding up to the nearest 10 for velocity and acceleration, respectively, and m represents the drive mode, m∈{CD, CS}. Indicates velocity v i acceleration a j and drive mode m k The emission factor of pollutant p, expressed in g / s, E p (v i a j m k ) represents the emission factor matrix of pollutant p.

[0102] Step 22: Construct the emission factor matrix for plug-in hybrid vehicles based on average vehicle speed;

[0103] Based on vehicle speed, the range is divided into two groups: 0-2.5 km / h and 2.5 km / h-102.5 km / h, with intervals of 5 km / h, namely 2.5-7.5 km / h, 7.5-12.5 km / h, 12.5-17.5 km / h, ..., 97.5-102.5 km / h.

[0104] Multiply the plug-in hybrid vehicle emission factor matrix based on the joint speed-acceleration distribution output in step 21 with the speed-acceleration distribution matrix within each vehicle speed group to obtain the plug-in hybrid vehicle emission factor matrix based on the average vehicle speed, as follows:

[0105]

[0106] Where avg speed bin represents the average vehicle speed range. Indicates average vehicle speed. Indicates average vehicle speed velocity v within the interval i acceleration a j frequency, Indicates based on average vehicle speed Velocity-acceleration distribution matrix

[0107]

[0108] in, Represents matrix dot product. This represents the p-emission factor matrix of pollutants from plug-in hybrid vehicles based on average vehicle speed under response-driven mode.

[0109] The emission factor matrices of the two plug-in hybrid vehicles based on average vehicle speed in this embodiment are shown in Tables 2 and 3.

[0110]

[0111]

[0112] Example 4:

[0113] Based on Example 3, step 3 specifically includes the following steps:

[0114] Step 31: Collect and improve user trip data for each hour: Obtain trip data of plug-in hybrid vehicle users through on-site surveys, and supplement the data with the cumulative mileage and SOC value of users for each 24 hours of the day through interpolation; provide the improved user trip data to Step 32.

[0115] Step 32: Based on the improved user trip data, calculate the user's daily mileage and average speed using timestamps and cumulative mileage data; calculate the user's average daily mileage for a single user, thereby determining the user's average daily mileage group label;

[0116] Preferably, the average daily mileage is divided into mileage ranges of 0-30km, 30-90km, 90-210km, and greater than 210km at equal intervals of 3km, 6km, 12km, and 24km, respectively, and the average driving speed is grouped according to step 22, i.e., avgspeed bin.

[0117] Furthermore, plug-in hybrid vehicles with a battery range of 80km or more are classified as long-range plug-in hybrid vehicles, while the rest are classified as short-range plug-in hybrid vehicles. User trip data is grouped, and the average daily mileage group label is calculated for each group.

[0118] The distribution characteristics of the average daily mileage of the plug-in hybrid vehicle in this embodiment are shown in Table 4:

[0119] Average daily mileage (km / day) Distribution of short-range group (%) Percentage of long-range vehicles (%) (0,6] 12.26 15.54 (6,12] 6.25 1.67 (12,27] 23.2 32.07 (27,42] 3.85 0.88 (42,57] 2.57 0.49 (57,72] 1.85 0.36 (72,87] 1.46 0.17 (87,102] 12.8 14.85 (102,142] 2.35 0.38 (142,182] 6.74 8.2 (182,222] 4.23 3.17 (222,262] 15.87 19.26 (262,302] 3.6 1.83 >302 2.97 1.13

[0120] Taking user A's average vehicle speed over 24 hours on a specific driving day as an example, the table is shown below (to correspond to S). n,i,d (h) matrix):

[0121]

[0122]

[0123] Step 33: Determine whether the user is charging the vehicle by using the vehicle's location latitude and longitude information and SOC value changes; improve the user's daily charging information and characterize the user's average daily charging frequency characteristics by grouping them into long cruising and short range categories.

[0124] Specifically, if the change in the vehicle's longitude and latitude is less than or equal to one millionth of a degree for more than 15 consecutive minutes and the SOC value gradually increases, the vehicle is considered to be charging; otherwise, it is not charging.

[0125] The charging frequency distribution characteristics of plug-in hybrid vehicle users in this embodiment are shown in Table 5:

[0126] Average daily charging frequency (times / day) Distribution of short-range group (%) Percentage of long-range vehicles (%) ＝0 24.33 4.00 (0,0.25) 26.24 9.14 [0.25,0.5) 8.75 21.71 [0.5,0.75) 12.55 24.57 [0.75,1) 13.69 18.29 [1,1.25) 6.84 7.43 [1.25,1.5) 4.56 6.29 [1.5,1.75) 1.14 4.00 [1.75,2) 0.38 2.86 ≥2 1.52 1.71

[0127] Step 34: Establish a SOC distribution matrix based on average daily mileage; based on the improved user trip data, group and label the hybrid vehicle drive mode status by average daily mileage and distinguish it by hour, i.e. CD mode or CS mode.

[0128] Step 35: Construct a multi-level hierarchical trip intensity matrix for responding to user behavior based on mileage, vehicle speed, and SOC driving mode; aggregate users with the same daily average mileage group label, and use the daily average mileage group as the factor base to statistically obtain the mileage-vehicle speed distribution matrix for responding to user behavior.

[0129] Example 5:

[0130] Based on Example 4, step 4 specifically includes the following steps:

[0131] Step 41: Calculate the mileage emissions of the plug-in hybrid vehicle considering user behavior; based on the mileage-speed distribution matrix of the plug-in hybrid vehicle output in Step 23, multiply it simultaneously with the mileage-speed distribution matrix in response to user behavior in Step 35 to calculate the CO2, CO, HC, and NO emissions from the plug-in hybrid vehicle exhaust during the user's daily mileage. x The emissions from the trip.

[0132] Example 6:

[0133] Based on Example 5, step 5 specifically includes the following steps:

[0134] Step 51: Set up the following 3 scenarios:

[0135] Scenario 1: Under the baseline scenario of "maximizing new energy," the plug-in hybrid vehicle maintains CD mode throughout the entire journey.

[0136] Scenario 2: Under the baseline scenario of "de-energization", the plug-in hybrid vehicle maintains CS mode throughout the entire journey.

[0137] Scenario 3: Actual driving scenario of plug-in hybrid vehicles

[0138] Calculate the CO2, CO, HC, and NO emissions from the exhaust of plug-in hybrid vehicles under three scenarios. x Total emissions, to assess the actual environmental benefits of plug-in hybrid vehicles in reducing pollution and emissions.

[0139] Table 6 shows a comparison of the total exhaust emissions of plug-in hybrid vehicles under three different scenarios in this embodiment:

[0140] scene <![CDATA[CO2 (tons)]]> CO (tons) HC (tons) NOx (tons) Scenario 1 124.607 0.015 0.046 0.161 Scenario 2 295.252 0.070 0.379 2.989 Scenario 3 220.229 0.048 0.234 1.750

[0141] 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 evaluating the actual pollution reduction and emission reduction benefits of plug-in hybrid vehicles considering user behavior, characterized in that, Includes the following steps: Step 1: Collect real-world emissions data for plug-in hybrid vehicles; Data on the travel itineraries of plug-in hybrid vehicle users was obtained through on-site surveys; The actual road emissions data for the plug-in hybrid vehicle specifically includes: vehicle speed, acceleration, latitude and longitude, altitude, and the emissions of carbon dioxide (CO2), carbon monoxide (CO), hydrocarbons (HC), and nitrogen oxides (NO) in the exhaust. x Emission coefficients, engine speed, and vehicle SOC value were collected with a data acquisition accuracy of 1Hz. The trip data of plug-in hybrid vehicle users obtained through field surveys includes: trip timestamps, SOC values, cumulative mileage, vehicle operating status, and latitude and longitude. The data is collected at intervals of no less than 30 seconds. The vehicle operating status includes pure electric, fuel, and hybrid modes. Step 2: Construct the emission factor matrix of the plug-in hybrid vehicle in response to SOC from two dimensions: the joint distribution of velocity-acceleration and the average vehicle speed. Step 3: Characterize user charging characteristics and construct a multi-level hierarchical trip intensity matrix to respond to user behavior; Step 4: Combine the emission factor matrix and the daily travel intensity matrix to calculate the travel emissions of plug-in hybrid vehicles that take into account user behavior; Step 5: Using "maximizing new energy" and "de-emphasis on new energy" as two baseline scenarios and the actual driving of plug-in hybrid vehicles by users as the real scenario, calculate the total exhaust pollutant emissions of plug-in hybrid vehicles under the three scenarios, and quantitatively evaluate the environmental benefits of actual pollution reduction and emission reduction of plug-in hybrid vehicles.

2. The method for evaluating the actual pollution reduction and emission reduction benefits of plug-in hybrid vehicles considering user behavior as described in claim 1, characterized in that, Step 1 includes the following steps: Step 11: Select the following equipment for monitoring exhaust emission data in plug-in hybrid vehicles: Portable Emission Measurement System (PEMS), On-Board Diagnostic (OBD) system, environmental sensors, BeiDou satellite positioning system, and action camera; among them, PEMS is responsible for recording timestamps, greenhouse gases and air pollutants in the exhaust gas including carbon dioxide (CO2), carbon monoxide (CO), hydrocarbons (HC), and nitrogen oxides (NO). x The data includes emission volume concentration and exhaust gas volume flow rate. The OBD system is responsible for collecting engine speed and vehicle speed data of the plug-in hybrid vehicle. The Beidou positioning satellite system is responsible for collecting vehicle speed, vehicle location latitude and longitude, and altitude information. The environmental sensor records the temperature and dew point in the atmospheric environment. The action camera is responsible for capturing information on the vehicle's dashboard during the experiment, including SOC value, timestamp, and vehicle speed. Step 12: Design a real-world emission monitoring test scheme for plug-in hybrid vehicles; the number of real-world emission monitoring tests for plug-in hybrid vehicles shall not be less than 4 times per vehicle, with each test lasting no less than 90 minutes and the cumulative road driving time no less than 10 hours per vehicle. Specifically, the cumulative driving time on highways shall not be less than 3 hours and the cumulative mileage no less than 120 kilometers, and the cumulative driving time on non-highway roads shall not be less than 6 hours and the cumulative mileage no less than 150 kilometers. In at least two of the 4 real-world emission monitoring tests, the initial SOC of the vehicle shall not be less than 60%. Step 13: Preprocessing and outlier removal of experimental data for real-world emissions from plug-in hybrid vehicles; Based on statistical principles, the interquartile range (interquartile range) is used to locate outliers in the experimental data, and Nan values ​​are used to replace them. Simultaneously, considering the data characteristics of the experimental parameters, the upper or lower bound parameters in the interquartile range are modified; Appropriate smoothing parameters are selected, and smoothing curves are used to analyze CO2, CO, HC, and NO in the exhaust gas. x The emission volume concentration, exhaust gas volume flow rate, BeiDou vehicle speed, and OBD vehicle speed are smoothed to improve data quality; according to the national standard "Limits and Measurement Methods for Pollutant Emissions from Light-Duty Vehicles (China VI)" (GB18352.6), altitude is corrected, and the concentrations of CO2, CO, HC, and NO in the exhaust gas are calculated. x The emission factor is expressed in g / s; the Paddle deep learning framework is used to identify vehicle speed, SOC value, and time information in the captured video. Step 14: Alignment processing of experimental data for actual road emissions of plug-in hybrid vehicles; Considering that PMES, OBD, and action camera in the actual road emissions monitoring equipment for plug-in hybrid vehicles are independent of each other, and that the environmental sensor and Beidou positioning satellite system rely on PEMS for power supply and communication, time alignment is performed according to "PMES time - dashboard time", and vehicle speed alignment is performed according to "Beidou vehicle speed - OBD vehicle speed - dashboard vehicle speed" to ensure data alignment between different devices; Step 15: Map matching based on Hidden Markov Model and Viterbi algorithm; determine the geographic coordinate system according to the actual road emission monitoring test location of the plug-in hybrid vehicle, and import the road data and map files of the city area where the test is located; encode the map data based on a binary tree structure to reduce the search space of the Markov model; map each trajectory point to a potential location on the road centerline based on the principle of minimizing the vertical projection distance; set the matching parameters of the Hidden Markov Model, call the Viterbi algorithm to perform map matching, and output the road name and road type corresponding to each trajectory point of the vehicle; Step 16: Sub-trip segmentation; Based on the output of Step 15, the road types mentioned above are divided into highways and non-highway roads; Using timestamps as features and vehicle speed as the target variable, a decision tree regression model is constructed, and the predicted speed change location or timestamp is calculated; By setting thresholds, change points are filtered and merged. If the difference between the mean speeds before and after a change point exceeds the threshold, the change point is marked. If the mean speeds before and after a change point are both below 3 or the distance between them is less than 400, the change point is marked as invalid; For each experimental data, the vehicle speed column data is traversed to identify segments with speeds below 5, and the start and end timestamps of these low-speed segments are recorded; Change points of the trip and road type are checked, new change points are added, and finally, a trip is divided into multiple sub-trips, distinguishing between high-speed trips and non-high-speed trips.

3. The method for evaluating the actual pollution reduction and emission reduction benefits of plug-in hybrid vehicles considering user behavior as described in claim 1, characterized in that... Step 2 includes the following steps: Step 21: Construct the emission factor matrix of plug-in hybrid vehicles based on the response SOC of the joint velocity-acceleration distribution. ,as follows: The actual road data of all plug-in hybrid vehicles processed in step 1 were divided into speed intervals of 1 km / h and acceleration intervals of 0.1 m / s², and a speed... In the 0 – 100 km / h range, acceleration The combined velocity-acceleration distribution in the range of -4 to -4 m / s²; Based on the vehicle's State of Charge (SOC) value, the driving modes of plug-in hybrid vehicles are divided into two types: energy consumption mode (CD) and energy sustaining mode (CS), and different driving modes are constructed. The combined velocity-acceleration distribution under different conditions was used to calculate the CO2, CO, HC, and NO content in the exhaust gas of each velocity-acceleration group in both CD and CS modes. x The average emission factor, in g / s; The emission factor matrix of the plug-in hybrid vehicle based on the joint velocity-acceleration distribution of the state of charge (SOC) is constructed as follows: in Indicates the speed range. Indicates the acceleration range. and These represent rounding up to the nearest multiple of 10 for velocity and acceleration, respectively. Indicates the driving mode. , Indicates speed acceleration and drive mode pollutants below The emission factor, expressed in g / s, Indicates pollutants The emission factor matrix; Step 22: Construct the emission factor matrix of plug-in hybrid vehicles based on the State of Charge (SOC) of average vehicle speed; Based on vehicle speed, the range is divided into two groups: 0-2.5 km / h and 2.5 km / h-102.5 km / h, with intervals of 5 km / h, namely 2.5-7.5 km / h, 7.5-12.5 km / h, 12.5-17.5 km / h, ..., 97.5-102.5 km / h. Multiply the plug-in hybrid vehicle emission factor matrix based on the joint speed-acceleration distribution output in step 21 with the speed-acceleration distribution matrix within each vehicle speed group to obtain the plug-in hybrid vehicle emission factor matrix based on the average vehicle speed, as follows: in, Indicates the average vehicle speed range. Indicates average vehicle speed. , Indicates average vehicle speed speed within the interval acceleration frequency, Indicates based on average vehicle speed Velocity-acceleration distribution matrix in, Represents matrix dot product. The pollutant emissions from plug-in hybrid vehicles are represented by the State of Charge (SOC) based on average vehicle speed. Emission factor matrix, They represent average vehicle speeds. Plug-in hybrid vehicle drive mode pollutants Emission factors, and They represent In the matrix A submatrix whose columns are equal to CD or CS.

4. The method for evaluating the actual pollution reduction and emission reduction benefits of plug-in hybrid vehicles considering user behavior as described in claim 1, characterized in that, Step 3 includes the following steps: Step 31: Collect and improve user trip data for each hour: Obtain trip data of plug-in hybrid vehicle users through on-site surveys, and supplement the data with the cumulative mileage and SOC value of users for each 24 hours of the day through interpolation; provide the improved user trip data to Step 32. Step 32: Based on the improved user trip data, calculate the user's daily mileage and average speed using timestamps and cumulative mileage data; calculate the user's average daily mileage for a single user, thereby determining the user's average daily mileage group label; Step 33: Determine whether the user is charging the vehicle by using the vehicle's location latitude and longitude information and SOC value changes; improve the user's daily charging information and characterize the user's average daily charging frequency characteristics by grouping them into long cruising and short range categories. Step 34: Establish a SOC distribution matrix based on average daily mileage; based on the improved user trip data, group and label the hybrid vehicle drive mode status by average daily mileage and distinguish it by hour, i.e. CD mode or CS mode. Step 35: Group by range and obtain the mileage-speed distribution matrix in response to user behavior.

5. The method for evaluating the actual pollution reduction and emission reduction benefits of plug-in hybrid vehicles considering user behavior as described in claim 4, characterized in that, In step 32, the average daily mileage is divided into mileage ranges of 0-30km, 30-90km, 90-210km, and greater than 210km at equal intervals of 3km, 6km, 12km, and 24km, respectively. The average driving speed is grouped in the same way as in step 22.

6. The method for evaluating the actual pollution reduction and emission reduction benefits of plug-in hybrid vehicles considering user behavior as described in claim 4, characterized in that, If the change in the vehicle's longitude and latitude is less than or equal to one millionth of a degree for more than 15 consecutive minutes, and the SOC value gradually increases, then the vehicle is considered to be charging; otherwise, it is not charging.

7. The method for evaluating the actual pollution reduction and emission reduction benefits of plug-in hybrid vehicles considering user behavior as described in claim 1, characterized in that, Step 4 is as follows: Step 41: Calculate the emissions of the plug-in hybrid vehicle considering user behavior; based on the plug-in hybrid vehicle emission factor matrix with response SOC based on average vehicle speed output in Step 2, multiply it simultaneously with the vehicle speed distribution matrix in response to user behavior in Step 3 to calculate CO2, CO, HC, and NO in the exhaust gas of the plug-in hybrid vehicle during the user's daily journey. x The emissions from the trip.

8. The method for evaluating the actual pollution reduction and emission reduction benefits of plug-in hybrid vehicles considering user behavior as described in claim 1, characterized in that, Step 5 specifically involves: Step 51: Set up the following 3 scenarios: Scenario 1: Under the baseline scenario of "maximizing new energy," the plug-in hybrid vehicle maintains CD mode throughout the entire journey. Scenario 2: Under the baseline scenario of "de-emphasis on new energy vehicles", the plug-in hybrid vehicle maintains CS mode throughout the entire journey. Scenario 3: Actual driving scenario of a plug-in hybrid vehicle; Calculate the CO2, CO, HC, and NO emissions from the exhaust of plug-in hybrid vehicles under three scenarios. x Total emissions, to assess the actual environmental benefits of pollution reduction and emission reduction of plug-in hybrid vehicles, are as follows: The baseline scenario for "maximizing new energy" (Scenario 1, S1) is the user's hourly average vehicle speed matrix. middle Set all to CD mode, that is In the baseline scenario of "de-energization" (Scenario 2, S2), the user's hourly average vehicle speed matrix middle Set all to CS mode, that is Therefore, the user trip emissions are calculated as follows under scenarios S1 and S2: in, and The table shows the battery life groupings under scenarios S1 and S2, respectively. plug-in hybrid vehicles User on driving day Pollutants during the journey Emissions matrix; The total emissions calculations for S1, S2, and the actual scenario for plug-in hybrid vehicles are as follows: in, , and These represent the pollutants from plug-in hybrid vehicles under scenarios S1, S2, and S3, respectively. Total emissions.

Images