A remote identification method and device for heavy-duty vehicles with excessive emissions that can replace PEMS testing

Abstract

The present invention relates to the field of transportation technology, and in particular to a method and device for remotely identifying heavy-duty vehicles with excessive emissions that can replace PEMS testing. The method comprises: obtaining test data of the heavy-duty vehicle and remote transmission data sent by a target platform; calculating a first average emission value of the heavy-duty vehicle based on the test data, and calculating a second average emission value of the heavy-duty vehicle using multiple evaluation methods and remote transmission data; selecting a target evaluation method from multiple evaluation methods based on the first and second average emission values, evaluating the remote transmission data using the target evaluation method, and generating a data set based on the evaluated data; training an evaluation model using the data set, and evaluating the emission status of the heavy-duty vehicle using the trained evaluation model. This solves the problem of large deviations in online determination of the emission status of heavy-duty vehicles caused by differences in window division basis and emission calculation theory among different evaluation methods in related technologies.

Description

Technical Field

[0001] The present invention relates to the technical field of transportation, and in particular to a remote identification method and device for over-emission heavy vehicles capable of replacing a PEMS test. Background Art

[0002] With the advancement of IoT cloud technology and the development of remote monitoring systems for heavy-duty vehicles, online assessment of heavy-duty vehicle emissions status based on remotely transmitted data has become possible. The remote monitoring system, a heavy-duty vehicle driving and emissions monitoring system, collects real-time on-road driving data and remotely transmitted data from connected heavy-duty vehicles. Using remotely transmitted data instead of PEMS (Portable Emissions Measurement System) test data allows for online assessment of heavy-duty vehicle emissions status, enabling online identification of vehicles exceeding emission standards. This provides targeted leads for real-world heavy-duty vehicle emissions regulation, thereby improving regulatory authorities' efficiency in identifying vehicles exceeding emission standards.

[0003] Due to the parameter limitations of telemetry data, particularly the lack of the critical parameter of heavy-duty vehicle cycle power, the power-based window method is not applicable to telemetry data. To address this deficiency, other evaluation methods can be introduced. These methods use specific strategies to divide continuous vehicle driving and emission data into distinct windows. These methods then perform more detailed analysis and processing of the emission data within each window to obtain a multi-dimensional and comprehensive picture of the true level of vehicle emissions, providing effective technical support for evaluating the quality of heavy-duty vehicle emissions.

[0004] In related technologies, due to differences in window division basis and emission calculation theory among different assessment methods, there are also large differences in the emission levels of heavy-duty vehicles obtained using different methods. This leads to large deviations in the online determination of the emission status of heavy-duty vehicles, and is unable to effectively provide targeted and highly accurate clues to illegal vehicles. This poses great challenges to the supervision and management of heavy-duty vehicle emissions. Summary of the Invention

[0005] The present invention provides a remote identification method, device, equipment, medium and program for heavy-duty vehicles with excessive emissions that can replace the PEMS test, so as to solve the problem that different evaluation methods in related technologies have different window division basis and emission calculation theory, resulting in large deviations in the online determination of the emission status of heavy-duty vehicles.

[0006] The first aspect of the present invention provides a remote identification method for heavy-duty vehicles with excessive emissions that can replace the PEMS test, including the following steps: obtaining test data of the heavy-duty vehicle and remote transmission data sent by the target platform; calculating a first emission average value of the heavy-duty vehicle based on the test data, and calculating a second emission average value of the heavy-duty vehicle using multiple evaluation methods and the remote transmission data; selecting a target evaluation method from multiple evaluation methods based on the first emission average value and the second emission average value, evaluating the remote transmission data using the target evaluation method, and generating a data set based on the evaluated data; training an evaluation model using the data set, and evaluating the emission status of the heavy-duty vehicle using the trained evaluation model.

[0007] Optionally, the use of multiple evaluation methods and the remote transmission data to calculate the second emission average value of heavy-duty vehicles includes: using multiple evaluation methods to divide the remote transmission data into windows to generate remote transmission data of multiple windows; calculating the corresponding NOx emission average value based on the remote transmission data of each window; and calculating the second emission average value of heavy-duty vehicles based on the cumulative value of the NOx emission average values ​​of all windows and the number of windows.

[0008] Optionally, based on the first emission average value and the second emission average value, a target assessment method is selected from a plurality of assessment methods, including: calculating an emission assessment equivalence coefficient based on the first emission average value and the second emission average value; and screening a target assessment method based on the emission assessment equivalence coefficient.

[0009] Optionally, the screening of the target evaluation method based on the emission evaluation equivalence coefficient includes: screening multiple evaluation methods corresponding to the emission evaluation equivalence coefficient approaching the target threshold; and selecting the evaluation method with the smallest absolute value of the emission evaluation equivalence coefficient from the multiple evaluation methods as the target evaluation method.

[0010] Optionally, the evaluation model includes an input layer, a long short-term memory network structure layer and an output layer, wherein the input layer inputs the heavy-duty vehicle window feature value generated according to the remote transmission data, each layer of the long short-term memory network structure includes a forget gate, an input gate and an output gate, and the output layer outputs the vehicle emission status, wherein the forget gate calculates the weight of the heavy-duty vehicle window feature value in the previous time step to obtain the first target information; the input gate calculates the weight of the heavy-duty vehicle window feature value in the current time step to obtain the second target information, and fuses the first target information and the second target information to obtain the target information of the current time step, and the output gate generates the vehicle emission status according to the target information of the current time step.

[0011] Optionally, after the heavy-duty vehicle emission data evaluation model outputs the emission status of the heavy-duty vehicle, it includes: if the emission status of the heavy-duty vehicle exceeds the standard, obtaining the actual position and identification code of the heavy-duty vehicle; determining the target heavy-duty vehicle based on the actual position and identification code of the heavy-duty vehicle; identifying the actual emission status of the target heavy-duty vehicle; and reversely updating the evaluation model based on the actual emission status.

[0012] The second aspect of the present invention provides a remote identification device for heavy-duty vehicles with excessive emissions that can replace the PEMS test, including: an acquisition module for acquiring test data of the heavy-duty vehicle and remote transmission data sent by the target platform; a calculation module for calculating a first emission average value of the heavy-duty vehicle based on the test data, and calculating a second emission average value of the heavy-duty vehicle using a plurality of evaluation methods and the remote transmission data; a selection module for selecting a target evaluation method from a plurality of evaluation methods based on the first emission average value and the second emission average value, evaluating the remote transmission data using the target evaluation method, and generating a data set based on the evaluated data; an evaluation module for training an evaluation model using the data set, and evaluating the emission status of the heavy-duty vehicle using the trained evaluation model.

[0013] A third aspect of the present invention provides an electronic device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor. The processor executes the program to execute the remote identification method for over-emission heavy vehicles that can replace the PEMS test as described in the above embodiment.

[0014] A fourth embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, which is executed by a processor to perform the remote identification method for heavy-duty vehicles with excessive emissions that can replace the PEMS test as described in the above embodiment.

[0015] A fifth aspect of the present invention provides a computer program product, comprising a computer program or instructions, characterized in that when the computer program or instructions are executed, a remote identification method for heavy-duty vehicles exceeding emission limits that can replace the PEMS test as described in the above embodiment is implemented.

[0016] Therefore, the present invention has at least the following beneficial effects:

[0017] The embodiment of the present invention can calculate the first emission average value of heavy-duty vehicles based on test data, and calculate the second emission average value of heavy-duty vehicles using multiple evaluation methods and remote data, and screen out a target evaluation method that is equivalent to the power-based window method, which solves the problem that remote data cannot be directly applied to regulatory standards due to the lack of cycle power parameters. The target evaluation method is used to evaluate the remote data, and the remote data that meets the conditions is screened according to the evaluation results to generate a training set to train the evaluation model, thereby improving the accuracy of online remote identification of the emission status of heavy-duty vehicles. The trained evaluation model is used to evaluate the emission status of heavy-duty vehicles, accurately capturing the dynamic emission laws of vehicles, and solving the defect that traditional static methods cannot associate multi-window time series information, and has high applicability.

[0018] Additional aspects and advantages of the present invention will be set forth in part in the description which follows and, in part, will be obvious from the description which follows, or may be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0020] Figure 1 A flowchart of a method for remotely identifying a heavy-duty vehicle with excessive emissions that can replace a PEMS test according to an embodiment of the present invention;

[0021] Figure 2 A technical roadmap for a method and device for remotely identifying heavy-duty vehicles with excessive emissions that can replace PEMS testing according to an embodiment of the present invention;

[0022] Figure 3 A schematic diagram of a work-base window method according to an embodiment of the present invention;

[0023] Figure 4 A schematic diagram of emission level calculation values ​​and emission assessment equivalent coefficients according to different assessment methods provided in an embodiment of the present invention;

[0024] Figure 5 This is an example diagram of a remote identification device for heavy-duty vehicles with excessive emissions that can replace PEMS testing according to an embodiment of the present invention;

[0025] Figure 6 A schematic structural diagram of an electronic device provided according to an embodiment of the present invention. DETAILED DESCRIPTION

[0026] The following describes embodiments of the present invention in detail, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, and are not to be construed as limiting the present invention.

[0027] As heavy-duty vehicle emissions increasingly impact the environment, it's crucial for regulators to promptly and effectively identify vehicles with excessive emissions. Currently, the primary method for determining whether heavy-duty vehicle emissions exceed standards is based on emissions regulations. The specific process involves conducting a PEMS test on a target vehicle, processing the test data using the power-based window method to determine the target vehicle's emission level, and comparing this data to the regulatory limits to determine whether the target vehicle's emissions exceed standards. However, the high cost of PEMS testing makes it difficult to conduct PEMS testing on all heavy-duty vehicles, making it difficult to promptly and effectively identify vehicles with excessive emissions.

[0028] Therefore, the present invention proposes a remote identification method for heavy-duty vehicles with excessive emissions that can replace the PEMS test, obtains an evaluation method equivalent to the power-based window method, and introduces a machine learning method to establish a heavy-duty vehicle emission status identification model, thereby realizing accurate online judgment of the emission status of heavy-duty vehicles.

[0029] The following describes, with reference to the accompanying drawings, a method, device, electronic device, storage medium, and program for remotely identifying heavy-duty vehicles with excessive emissions that can replace PEMS testing according to embodiments of the present invention.

[0030] Specifically, Figure 1 The present invention provides a flow chart of a remote identification method for heavy-duty vehicles with excessive emissions that can replace the PEMS test.

[0031] like Figure 1 As shown, the remote identification method for heavy-duty vehicles with excessive emissions that can replace the PEMS test includes the following steps:

[0032] In step S101, the test data of the heavy-duty vehicle and the remote transmission data sent by the target platform are obtained.

[0033] It is understandable that the embodiment of the present invention can obtain the test data of the heavy-duty vehicle and the telemetry data sent by the target platform, so as to facilitate the subsequent selection of the target evaluation method and the training data set.

[0034] It should be noted that the present invention selects several test heavy-duty vehicles to carry out PEMS testing, and determines the emission status of the test heavy-duty vehicles according to emission regulations, for example: normal emissions or emissions exceeding the standard; and marks the emission status, and simultaneously downloads the remote transmission data of the test heavy-duty vehicles for subsequent optimal method selection and model training.

[0035] The heavy-duty vehicle remote monitoring system is used to remotely download the telemetry data of the test heavy-duty vehicle. The data includes parameters such as time, vehicle speed, acceleration, engine torque, engine speed, fuel flow, NOx emission rate, atmospheric pressure, ambient temperature, altitude, etc. The telemetry data sampling frequency is 1 Hz. The heavy-duty vehicle remote monitoring system here is a heavy-duty vehicle driving and emission data monitoring system that can collect the actual road driving data of networked heavy-duty vehicles in real time, that is, telemetry data.

[0036] In step S102 , a first average emission value of the heavy-duty vehicle is calculated based on the test data, and a second average emission value of the heavy-duty vehicle is calculated using multiple evaluation methods and remote data.

[0037] It is understandable that the embodiment of the present invention can calculate the first emission average value of heavy-duty vehicles based on test data, and calculate the second emission average value of heavy-duty vehicles using multiple evaluation methods and remote transmission data, so as to facilitate the subsequent selection of a target evaluation method from multiple evaluation methods based on the first emission average value and the second emission average value.

[0038] It should be noted that the first emission average value is calculated based on the power-based window method to process PEMS test data, and the second emission average value is calculated based on multiple evaluation methods to process remote transmission data. The evaluation method of the present invention is divided into two steps: window division and emission data evaluation.

[0039] Specifically, the power-based window method is used to process the PEMS test data. It is determined whether the heavy-duty vehicle under test exceeds the emission limit according to the emission regulations. The emission status of the heavy-duty vehicle under test is marked as a heavy-duty vehicle with excessive emissions or a heavy-duty vehicle with normal emissions. The emission status marking results of the heavy-duty vehicle under test are obtained. Specifically:

[0040] First, the PEMS test data is processed using the work-based window method, and the ratio of the total pollutant emissions to the total engine work in each window is calculated, which is recorded as specific emissions. e NOx Secondly, it will be more e NOx By comparing with the limit standards in the emission regulations, it is determined whether the test heavy-duty vehicle exceeds the emission limit, and the emission status of the test heavy-duty vehicle is marked as a heavy-duty vehicle with excessive emissions and a heavy-duty vehicle with normal emissions, thereby obtaining the emission status marking results of the test heavy-duty vehicle.

[0041] The calculation process of the work-based window method is as follows: Figure 3 As shown, the calculation formula of the power-based window method is as follows (1) and (2):

[0042] (1)

[0043] (2)

[0044] Where, Indicates the i The end time of the window; Indicates the i The start time of the window; Indicates the i A moment within a window; W ( t ) indicates from 0 to t The accumulated instantaneous engine output power at the time, W ref The work required to complete a WHTC (World Harmonized Transient Cycle) test condition for a heavy-duty vehicle.

[0045] The work-based window method uses the WHTC cycle work of heavy vehicles ( W ref ) as a benchmark to establish a moving window for the work-based window method. WHTC cycle work refers to the work required for a heavy-duty vehicle to complete one WHTC test condition. The work-based window method uses engine torque and engine speed from PEMS test data to calculate the instantaneous engine output power. The instantaneous engine output power is then accumulated second by second until it is greater than or equal to the corresponding WHTC cycle work, which is recorded as a window. The window is moved until all PEMS test data for the heavy-duty vehicle is fully covered. The total amount of work performed by the vehicle within each window is similar across all acquired windows.

[0046] In an embodiment of the present invention, a second emission average value of heavy-duty vehicles is calculated using multiple evaluation methods and remote data, including: using multiple evaluation methods to divide the remote data into windows to generate remote data of multiple windows; calculating the corresponding NOx emission average value based on the remote data of each window; and calculating the second emission average value of heavy-duty vehicles based on the cumulative value of the NOx emission average values ​​of all windows and the number of windows.

[0047] It is understandable that the embodiments of the present invention can use multiple evaluation methods to divide the remote transmission data into windows to generate remote transmission data for multiple windows; calculate the corresponding NOx emission average value based on the remote transmission data of each window; calculate the second emission average value of the heavy-duty vehicle based on the cumulative value of the NOx emission average value of all windows and the number of windows, so as to comprehensively evaluate the emission level of the heavy-duty vehicle, but it is necessary to rely on EEEC (Emissions Evaluation Equivalent Coefficient) to screen the optimal method to ensure that the results are close to regulatory standards.

[0048] It should be noted that the evaluation method refers to dividing the time-continuous vehicle emission data into different windows through a specific strategy, and conducting more detailed analysis and processing of the emission data in each window, so as to obtain the multi-dimensional and comprehensive true level of vehicle emissions, and provide effective technical support for evaluating the quality of heavy-duty vehicle emissions.

[0049] Specifically, the evaluation method of the present invention is described using several commonly used methods, such as the moving window average method, the three-zone moving average window method, the VSP-Bin-based window division method, and the data reconstruction PEMS window division method. The calculation process for using the above evaluation methods to process the heavy-duty vehicle remote transmission data to obtain the window and evaluate the heavy-duty vehicle emission level is as follows:

[0050] (1) The moving window averaging method divides the windows based on the time length, which is usually set to 30 seconds. The moving window averaging method can divide the telemetry data of heavy-duty vehicles into several windows with a time length of 30 seconds. Based on this, the ratio of the cumulative sum of the NOx emission rates at all times in each window to the window time length is calculated as the average NOx emission rate of the corresponding window. The average NOx emission rates of all windows are added and divided by the number of windows to obtain the heavy-duty vehicle emission level assessment result of the moving window averaging method.

[0051] (2) The three-zone moving average window method uses 300 seconds of time and the load ratio of heavy vehicles as the standard to divide the heavy vehicle remote transmission data.

[0052] First, the telemetry data of the heavy-duty vehicle is divided into several segments with a time length of 300 s, and the load ratio of each segment of the heavy-duty vehicle is calculated using equations (3) and (4). At the same time, the ratio of the cumulative sum of the NOx emission rates at all moments in the segment to the segment time length is calculated as the average NOx emission rate of the corresponding segment.

[0053] Then, according to the load ratio of heavy vehicles, all segments are divided into three intervals: idling area, low load area and medium and high load area; The segments with a load ratio less than 0.06 are classified into the idle zone, the segments with a load ratio greater than 0.2 are classified into the medium and high load zone, and the remaining segments are classified into the low load zone.

[0054] Then, the ratio of the cumulative sum of the average NOx emission rates of all segments in each interval to the total number of segments in the interval is calculated using equation (5) as the average NOx emission rate of the interval.

[0055] Finally, the heavy-duty vehicle emission level assessment results of the three-zone moving average window method are calculated using equation (6).

[0056] ;(3)

[0057] ;(4)

[0058] ;(5)

[0059] ;(6)

[0060] Where: CO2 test results for this vehicle model based on target requirements; is the rated power of the vehicle's engine; 300 s; 1 s; Fuel flow for heavy vehicles; is the density of diesel; 44 is the relative molecular mass of CO2; is the specific emission of the zone; a is the emission, NOx; b is the type of zone, which can be the idle zone, low load zone or medium-high load zone; is the number of NOx fragments contained in area b; is the instantaneous mass emission of NOx; is the total number of fragments in the three regions; is the number of fragments in region b, The heavy-duty vehicle emission level assessment results were obtained using the three-zone moving average window method; k The first speed in a certain range (idle zone, low load zone or medium-high load zone) k fragments.

[0061] (3) The window partitioning method based on VSP-Bin uses vehicle specific power as the criterion to divide the telematic data of heavy-duty vehicles. VSP refers to vehicle specific power. The calculation formula is as shown in Equation (7). Based on VSP, the interval division can obtain several vehicle specific power intervals. The interval here is usually referred to as Bin, so the vehicle specific power interval is VSP-Bin.

[0062] First, the vehicle specific power of the heavy-duty vehicle at each moment is calculated using Equation (7). Then, 10 vehicle specific power intervals are constructed with a vehicle specific power interval of 2 kW / ton, as shown in Table 1 below, and the telematic data of the heavy-duty vehicle is divided into the corresponding specific power intervals.

[0063] Subsequently, the ratio of the cumulative sum of the NOx emission rates at all times within each vehicle specific power interval to the data volume of the corresponding vehicle specific power interval is calculated as the average NOx emission rate of the corresponding vehicle specific power interval. Finally, the heavy-duty vehicle emission level assessment result based on the VSP-Bin window partitioning method is calculated using Equation (8).

[0064] (7)

[0065] (8)

[0066] Where VSP refers to vehicle specific power, v is vehicle speed, a is acceleration; is the data volume of the i-th power ratio interval; is the total data volume; is the average value of NOx emission rate in the i-th specific power interval, The heavy-duty vehicle emission level assessment results are calculated using the window partitioning method based on VSP-Bin.

[0067] Table 1 Vehicle specific power division interval table

[0068]

[0069] (4) Data reconstruction PEMS window division method reconstructs the telematic data of heavy-duty vehicles into actual road conditions in several PEMS tests, and evaluates the emission level of heavy-duty vehicles based on this.

[0070] First, the telematics data of heavy-duty vehicles is cut into several short-trip segments, where the vehicle speed at the start and end of each segment is 0, and the vehicle speed within each segment is greater than 0.

[0071] Then, the average speed of all short trip segments is calculated, and based on the average speed, all short trip segments are divided into urban areas (speed <50 km / h), suburban areas (speed between 50 and 75 km / h), and high-speed areas (speed greater than 75 km / h).

[0072] Subsequently, based on the characteristics of the PEMS test conditions in the standard, corresponding short-trip segments were selected from urban, suburban, and highway sections to reconstruct the actual road conditions. The selection and combination rules for short-trip segments should meet the following requirements: reasonable boundary conditions between adjacent segments (the average speed difference between the previous segment and the next segment should be less than 10 km / h); the mileage of the combined urban segment segments should account for 15% to 20% of the total mileage, the mileage of the combined suburban segment segments should account for 20% to 25% of the total mileage, and the mileage of the combined highway segment segments should account for 50% to 55% of the total mileage; the average speed of the combined urban segment segments should be between 15 and 30 km / h, and the average speed of the combined suburban segment segments should be between 45 and 70 km / h.

[0073] Finally, the ratio of the cumulative sum of the NOx emission rates at all moments in the reconstructed actual road condition to the length of the reconstructed condition is calculated as the heavy-duty vehicle emission level evaluation result of the data reconstruction PEMS window division method.

[0074] In step S103, a target evaluation method is selected from a plurality of evaluation methods based on the first average emission value and the second average emission value, the remote transmission data is evaluated using the target evaluation method, and a data set is generated based on the evaluated data.

[0075] It can be understood that the embodiment of the present invention selects a target evaluation method from a plurality of evaluation methods based on the first emission average value and the second emission average value, uses the target evaluation method to evaluate the remote data, and generates a data set based on the evaluated data, thereby improving the accuracy of online remote identification of the emission status of heavy-duty vehicles.

[0076] It should be noted that the target evaluation method can be understood as the one that most closely matches the calculation results of the power-based window method among multiple evaluation methods. Using this optimal evaluation method, the telemetry data of the test heavy-duty vehicle is processed, divided into several windows, and the characteristic values ​​of each window parameter are calculated. These characteristic values ​​include average vehicle speed, average acceleration, average engine torque, average fuel flow, average NOx emission rate, average atmospheric pressure, average ambient temperature, and average altitude. The window characteristics of the test heavy-duty vehicle are combined with the emission status labeling results of the test heavy-duty vehicle to establish the model training dataset.

[0077] In an embodiment of the present invention, a target evaluation method is selected from a plurality of evaluation methods based on the first emission average value and the second emission average value, including: calculating an emission evaluation equivalence coefficient based on the first emission average value and the second emission average value; and screening the target evaluation method based on the emission evaluation equivalence coefficient.

[0078] It can be understood that the embodiment of the present invention can calculate the emission evaluation equivalence coefficient based on the first emission average value and the second emission average value; screen the target evaluation method based on the emission evaluation equivalence coefficient, and thus select the evaluation method that is closest to the calculation result of the power basis window method from multiple evaluation methods as the target evaluation method to improve the accuracy of online remote identification of the emission status of heavy-duty vehicles.

[0079] It should be noted that the calculation method of the emission assessment equivalence coefficient is:

[0080] (9)

[0081] Where, E j Indicates the heavy-duty vehicle emission level assessment result obtained by processing remote transmission data using a certain assessment method; It represents the average value of heavy-duty vehicle emission levels obtained by processing the PEMS test data of heavy-duty vehicles using the power-based window method.

[0082] To facilitate evaluation, the calculation results of the power-based window method have been converted to units. The specific conversion method is as follows: the heavy-duty vehicle PEMS test data is processed using the power-based window method to obtain several windows. On this basis, the ratio of the cumulative sum of the NOx emission rates at all times in each window to the window time length is calculated as the average NOx emission rate of the corresponding window; the average NOx emission rates of all windows are added together and divided by the number of windows to obtain the heavy-duty vehicle emission level evaluation results of the power-based window method.

[0083] In an embodiment of the present invention, the target evaluation method is screened according to the emission evaluation equivalence coefficient, including: screening multiple evaluation methods whose emission evaluation equivalence coefficients are close to the target threshold value; and selecting the evaluation method with the smallest absolute value of the emission evaluation equivalence coefficient from the multiple evaluation methods as the target evaluation method.

[0084] The target threshold can be 0, without any specific limitation.

[0085] It can be understood that the embodiment of the present invention can select the evaluation method whose emission evaluation equivalent coefficient is close to the target threshold as the target evaluation method, thereby selecting the evaluation method that is closest to the calculation result of the power basis window method from multiple evaluation methods as the target evaluation method, thereby improving the consistency of remote monitoring results with regulatory standards, and thus improving the accuracy of online remote identification of the emission status of heavy-duty vehicles.

[0086] It should be noted that the physical meaning of the emission assessment equivalence coefficient EEEC can be interpreted as:

[0087] When EEEC=0, it means that the vehicle emission level evaluated by the current method is consistent with the vehicle emission level evaluated by the power-based window method, and there is no difference in effect.

[0088] When EEEC>0, it means that the vehicle emission level assessed using the current method is generally higher than the vehicle emission level assessed using the power-based window method, and there is a certain degree of overestimation.

[0089] When EEEC<0, it means that the vehicle emission level assessed using the current method is generally lower than the vehicle emission level assessed using the power-based window method, and there is a certain degree of underestimation.

[0090] When EEEC is closer to 0, it means that the vehicle emission level evaluated by the current method is closer to the evaluation level of the power-based window method. It also shows that the current method is better and closer to the evaluation and judgment effect using the power-based window method in emission regulations.

[0091] In step S104, the evaluation model is trained using the data set, and the trained evaluation model is used to evaluate the emission status of heavy-duty vehicles.

[0092] The evaluation model includes an input layer, a long short-term memory network structure layer and an output layer. The input layer inputs the heavy-duty vehicle window feature value generated according to the remote transmission data. Each layer of the long short-term memory network structure contains a forget gate, an input gate and an output gate. The output layer outputs the vehicle emission status. The forget gate calculates the weight of the heavy-duty vehicle window feature value in the previous time step to obtain the first target information; the input gate calculates the weight of the heavy-duty vehicle window feature value in the current time step to obtain the second target information, and fuses the first target information and the second target information to obtain the target information of the current time step. The output gate generates the vehicle emission status according to the target information of the current time step.

[0093] It can be understood that the embodiments of the present invention can use the trained evaluation model to evaluate the emission status of heavy-duty vehicles. Through time series modeling and dynamic optimization, high-precision, real-time emission monitoring is achieved, thereby improving the accuracy of online remote identification of the emission status of heavy-duty vehicles.

[0094] Specifically, based on the model training dataset, a heavy-duty vehicle emission status recognition model is constructed using the LSTM neural network method: the model takes the characteristic values ​​of the heavy-duty vehicle window as input; and the heavy-duty vehicle emission status labeling results as output.

[0095] All windows of the same vehicle in the model training dataset constitute a sample. Each sample is fed into the model sequentially for training. The model calculates all windows of each sample, frame by frame, in chronological order. Each frame generates a memory unit and a hidden state. The memory unit is a core long-term memory carrier that can transfer information between the previous and next frames across time steps. The hidden state contains the output of each frame and the final model output.

[0096] Model Structure: The model consists of a four-layer LSTM network structure, each of which includes a forget gate, an input gate, and an output gate. The forget gate determines which information in the memory cells of the previous frame should be retained or discarded, thereby obtaining valuable old information. Here, a sigmoid function is used to output weights ranging from 0 to 1 to process the memory cells of the previous frame, where 0 indicates complete discard and 1 indicates complete retention. The input gate calculates the current window data frame by frame to obtain new information, then integrates the valuable old information transmitted by the forget gate of the previous frame to obtain the memory cells of the current frame. The output gate generates the hidden state or final model output based on the memory cells of the current frame. A sigmoid activation function is used to generate output weights ranging from 0 to 1, which are then multiplied by the memory cells processed by the Tanh activation function to obtain the hidden state of the current frame and pass it to the next frame.

[0097] In summary, the LSTM neural network method successfully constructs a heavy-duty vehicle emission status recognition model, resolving the problem that traditional methods can only solve static predictions. This problem typically manifests itself as: when inputting the characteristic value data of a window, the model constructed by the traditional method calculates a corresponding output value. This calculation method fails to account for the mutual influence between multiple consecutive windows in the time series, as well as the changes in emission patterns during vehicle driving. In the process of building a heavy-duty vehicle emission status recognition model, the present invention aims to solve the problem of determining whether a vehicle has excessive emissions. This requires using the characteristic values ​​of multiple consecutive windows in the time series as model input, and obtaining a final output value through model calculation, thereby achieving the identification of whether the target vehicle has excessive emissions. LSTM neural networks can capture long-term dependencies in time series data and are suitable for processing remote data with strong temporal continuity. Furthermore, the present invention further optimizes the remote data to obtain multiple consecutive windows and their characteristic values ​​in the time series, which are used to more clearly represent the changes in vehicle emission patterns during driving, thereby providing strong support for the subsequent construction of heavy-duty vehicle emission status recognition models and the identification of whether vehicles have excessive emissions.

[0098] In an embodiment of the present invention, after the heavy-duty vehicle emission data evaluation model outputs the emission status of the heavy-duty vehicle, the following steps are included: if the emission status of the heavy-duty vehicle exceeds the standard, the actual position and identification code of the heavy-duty vehicle are obtained; the target heavy-duty vehicle is determined based on the actual position and identification code of the heavy-duty vehicle; the actual emission status of the target heavy-duty vehicle is identified; and the evaluation model is reversely updated based on the actual emission status.

[0099] It can be understood that the embodiments of the present invention can obtain the actual location and identification code of the heavy-duty vehicle when the emission status of the heavy-duty vehicle exceeds the standard; determine the target heavy-duty vehicle based on the actual location and identification code of the heavy-duty vehicle, so that inspectors can use portable emission detection equipment to conduct on-site inspections, identify the actual emission status of the target heavy-duty vehicle, and verify the accuracy of the model; and reversely update the evaluation model based on the actual emission status to improve the accuracy of the online remote identification of the emission status of the heavy-duty vehicle. Through the closed-loop feedback mechanism of "exceeding standard identification-positioning verification-model update", the precise and intelligent emission supervision of heavy-duty vehicles is achieved.

[0100] Specifically, the heavy-duty vehicle emission status recognition model will be integrated into the heavy-duty vehicle remote monitoring system to remotely identify heavy-duty vehicles exceeding emission standards online and issue early warnings. Based on on-site law enforcement feedback, the heavy-duty vehicle emission status recognition model will be updated online to adapt to changes in vehicle status and the environment.

[0101] (1) First, the heavy-duty vehicle emission status recognition model is connected to the heavy-duty vehicle remote monitoring system, relying on the real-time collection of telemetric data from networked heavy-duty vehicles. The optimal evaluation method is used to process the telemetric data of all networked heavy-duty vehicles to obtain the model prediction data set of networked heavy-duty vehicles. Based on this, the heavy-duty vehicle emission status recognition model is used to identify the emission status of networked heavy-duty vehicles, identify heavy-duty vehicles with excessive emissions, mark them as target heavy-duty vehicles, and issue early warnings. At the same time, the enforcement clues of target heavy-duty vehicles with excessive emissions are provided, including vehicle GPS positioning and VIN (Vehicle Identification Number) codes.

[0102] (2) Based on the enforcement clues of the target heavy-duty vehicle's excessive emissions, the target heavy-duty vehicle can be located and subjected to on-site enforcement, and feedback on the on-site enforcement results can be obtained, including whether the target heavy-duty vehicle's actual emissions exceed the standards.

[0103] (3) Relying on the on-site law enforcement feedback results, the model prediction data set of the target heavy-duty vehicle and the on-site law enforcement feedback results are integrated as new supplementary model training data to perform online update training on the heavy-duty vehicle emission status recognition model, thereby improving the adaptability of the heavy-duty vehicle emission status recognition model to vehicle status and environmental changes.

[0104] According to the embodiment of the present invention, a remote identification method for over-emission heavy-duty vehicles that can replace the PEMS test is proposed. The first emission average value of the heavy-duty vehicle is calculated based on the test data, and the second emission average value of the heavy-duty vehicle is calculated using multiple evaluation methods and remote transmission data to screen out a target evaluation method equivalent to the power-based window method. This solves the problem that the remote transmission data cannot be directly applied to regulatory standards due to the lack of cycle power parameters. The target evaluation method is used to evaluate the remote transmission data. The remote transmission data that meets the conditions is screened according to the evaluation results to generate a training set to train the evaluation model, thereby improving the accuracy of online remote identification of the emission status of heavy-duty vehicles. The trained evaluation model is used to evaluate the emission status of heavy-duty vehicles, accurately capturing the dynamic emission patterns of vehicles, and solving the defect that the traditional static method cannot associate multi-window time series information. The method has high applicability.

[0105] The following will describe in detail the method for remotely identifying heavy-duty vehicles with excessive emissions that can replace the PEMS test according to the present invention with reference to specific embodiments, as follows:

[0106] 1. We collected and organized PEMS test data from 200 test heavy-duty vehicles. We determined and marked the emission status of the test heavy-duty vehicles according to emission regulations. We also downloaded the remote transmission data of the test heavy-duty vehicles and performed data preprocessing for subsequent optimal method selection and model training.

[0107] (1.1) First, all 200 test heavy-duty vehicles collected and collated met emission regulations before leaving the factory. Furthermore, the PEMS test data for these 200 test heavy-duty vehicles all had a mileage exceeding 10,000 kilometers before the PEMS tests were conducted. This means that none of the 200 test heavy-duty vehicles collected and collated were new vehicles.

[0108] (1.2) Determine the emission status of the test heavy-duty vehicle according to emission regulations. The specific steps are as follows: PEMS test data is processed using the work-based window method. Then, the ratio of total pollutant emissions to total engine work in each window is calculated, denoted as specific NOx. Finally, the specific eNOx is compared with the regulatory limit to determine whether the test heavy-duty vehicle exceeds the emission limit. The emission status of the test heavy-duty vehicle is then labeled as either a vehicle with excessive emissions or a vehicle with normal emissions, thereby obtaining the emission status labeling results for the test heavy-duty vehicles. Ultimately, 54 vehicles were found to have excessive emissions, and 146 vehicles were found to have normal emissions.

[0109] (1.3) Use the heavy-duty vehicle remote monitoring system to remotely download the telemetry data of the test heavy-duty vehicle. The data includes parameters such as time, vehicle speed, acceleration, engine torque, engine speed, fuel flow, NOx emission rate, atmospheric pressure, ambient temperature, altitude, etc. The telemetry data sampling frequency is 1 Hz.

[0110] 2. Define the Emission Evaluation Equivalence Coefficient (EEEC) to quantify the effectiveness of the evaluation method for heavy-duty vehicle emission levels. Select the evaluation method with the best Emission Evaluation Equivalence Coefficient (EEEC) that is suitable for telemetry data and establish a model training dataset.

[0111] (2.1) First, the evaluation method involves partitioning continuous vehicle emissions data into distinct windows using a specific strategy. The data within each window is then analyzed and processed in detail to obtain a multi-dimensional, comprehensive picture of the true vehicle emissions level, providing effective technical support for evaluating the quality of heavy-duty vehicle emissions. Finally, the ratio of the cumulative sum of NOx emission rates at all times within the reconstructed actual road condition to the duration of the reconstructed condition is calculated as the heavy-duty vehicle emission level assessment result using the data reconstruction PEMS windowing method.

[0112] (2.2) Then, define the emission evaluation equivalence coefficient EEEC. The physical meaning of the emission evaluation equivalence coefficient EEEC can be explained as follows: when EEEC=0, it means that the vehicle emission level assessed using the current method is consistent with the vehicle emission level assessed using the power-based window method, with no difference in effect. When EEEC>0, it means that the vehicle emission level assessed using the current method is generally higher than the vehicle emission level assessed using the power-based window method, indicating a certain degree of overestimation. When EEEC<0, it means that the vehicle emission level assessed using the current method is generally lower than the vehicle emission level assessed using the power-based window method, indicating a certain degree of underestimation. The closer EEEC is to 0, the closer the vehicle emission level assessed by the current method is to the assessment level by the power-based window method, which also indicates that the current method is better and closer to the assessment and judgment effect of the power-based window method in emission regulations.

[0113] (2.3) Then, based on the PEMS test data and telemetry data of the test heavy-duty vehicle obtained in step 1, the power-based window method is used to process the PEMS test data of the test heavy-duty vehicle to obtain the emission level assessment results of the test heavy-duty vehicle, such as Figure 4 The first column on the left side is about 0.255 g / s. The remote transmission data of the heavy-duty vehicle under evaluation were processed by the evaluation methods respectively, and the emission level evaluation results of the heavy-duty vehicle under different evaluation methods were obtained, such as Figure 4 Columns 2 to 5 in the table and calculate the corresponding emission assessment equivalence coefficient EEEC, such as Figure 4 The middle broken line. Select the evaluation method with the best emission evaluation equivalent coefficient EEEC and suitable for remote data. Figure 4 The results show that the emission level calculated using the moving window averaging method is closest to the power-based window method, at 0.261 g / s. Furthermore, this method achieves an EEEC value of 0.02, the best among all evaluation methods. Therefore, the moving window averaging method was selected as the processing method for subsequent telemetry data.

[0114] (2.4) Finally, the optimal evaluation method (i.e., moving window averaging) was used to process the telemetry data from the test heavy-duty vehicle. This data was divided into several windows, and the characteristic values ​​of each window parameter were calculated. These characteristic values ​​included average vehicle speed, average acceleration, average engine torque, average fuel flow, average NOx emission rate, average atmospheric pressure, average ambient temperature, and average altitude. The window characteristics of the test heavy-duty vehicle were combined with the emission status labeling results of the test heavy-duty vehicle to establish the model training dataset.

[0115] 3. Based on the model training data set, a heavy-duty vehicle emission status recognition model is constructed using the LSTM neural network method.

[0116] Among them, the LSTM neural network method's processing capabilities on time series data are used to train a heavy-duty vehicle emission status recognition model. The model uses all windows of each vehicle in the model training dataset as frame-by-frame input, the feature values ​​of each window as input parameters, and the emission status labeling results (i.e., emissions exceeding the standard and normal emissions) as the final output of the model. The heavy-duty vehicle emission status recognition model is trained and constructed.

[0117] All windows of the same vehicle in the model training dataset constitute a sample. Each sample is fed into the model sequentially for training. The model calculates all windows of each sample, frame by frame, in chronological order. Each frame generates a memory unit and a hidden state. A memory unit is a core long-term memory carrier that transfers information between the previous and next frames across time steps. The hidden state contains the output of each frame and the final model output.

[0118] The model consists of a four-layer LSTM network structure, each of which includes a forget gate, an input gate, and an output gate. The forget gate determines which information in the memory cells of the previous frame should be retained or discarded, thereby obtaining valuable old information. Here, a sigmoid function is used to output weights ranging from 0 to 1 to process the memory cells of the previous frame, with 0 indicating complete discard and 1 indicating complete retention. The input gate calculates the current window data frame by frame to obtain new information, which is then integrated with the valuable old information transmitted by the forget gate of the previous frame to ultimately obtain the memory cells of the current frame. The output gate generates the hidden state or final model output based on the memory cells of the current frame. A sigmoid function is used to generate output weights ranging from 0 to 1, which are then multiplied by the memory cells processed by the Tanh function to obtain the hidden state of the current frame and pass it to the next frame.

[0119] 4. Integrate the heavy-duty vehicle emission status recognition model into the heavy-duty vehicle remote monitoring system to remotely identify heavy-duty vehicles exceeding emission standards online and issue early warnings. Based on on-site law enforcement feedback, the heavy-duty vehicle emission status recognition model will be updated online to adapt to changes in vehicle status and the environment.

[0120] (4.1) First, the heavy-duty vehicle emission status recognition model was connected to the heavy-duty vehicle remote monitoring system, relying on real-time telemetry data collected from networked heavy-duty vehicles. To verify the recognition effectiveness of the present invention's technology, 100 networked heavy-duty vehicles were randomly selected from the remote monitoring system, with each vehicle downloading one month's worth of telemetry data. The optimal evaluation method obtained in step (2.3) was used to process the telemetry data from all networked heavy-duty vehicles, obtaining a model prediction dataset for each networked heavy-duty vehicle. Based on this dataset, the heavy-duty vehicle emission status recognition model was used to identify the emission status of these networked heavy-duty vehicles. Eleven networked heavy-duty vehicles were identified as exceeding emission standards. These vehicles were marked as target vehicles and issued warnings. Law enforcement clues related to these vehicles' excessive emissions were also provided, including vehicle GPS locations and VIN numbers.

[0121] (4.2) Based on enforcement clues regarding excessive emissions from target heavy-duty vehicles, the target vehicles can be located and subjected to on-site enforcement, generating feedback on whether the vehicles actually exceeded emissions standards. This feedback revealed that 10 of the vehicles actually exceeded emissions standards, demonstrating an online recognition accuracy exceeding 90.9%.

[0122] (4.3) Based on the on-site enforcement feedback results, the model prediction dataset of the target heavy-duty vehicle and the on-site enforcement feedback results are integrated as new supplementary model training data to conduct online update training of the heavy-duty vehicle emission status recognition model, thereby improving the heavy-duty vehicle emission status recognition model's adaptability to vehicle status and environmental changes.

[0123] Next, a remote identification device for heavy-duty vehicles exceeding emission limits that can replace the PEMS test according to an embodiment of the present invention will be described with reference to the accompanying drawings.

[0124] Figure 5 4 is a block diagram of a remote identification device for heavy-duty vehicles with excessive emissions that can replace PEMS testing according to an embodiment of the present invention.

[0125] like Figure 5 As shown, the remote identification device 10 for heavy-duty vehicles with excessive emissions that can replace the PEMS test includes: an acquisition module 100 , a calculation module 200 , a selection module 300 and an evaluation module 400 .

[0126] Among them, the acquisition module 100 obtains the test data of the heavy-duty vehicle and the remote transmission data sent by the target platform; the calculation module 200 is used to calculate the first emission average value of the heavy-duty vehicle based on the test data, and calculate the second emission average value of the heavy-duty vehicle using multiple evaluation methods and remote transmission data; the selection module 300 is used to select a target evaluation method from multiple evaluation methods based on the first emission average value and the second emission average value, evaluate the remote transmission data using the target evaluation method, and generate a data set based on the evaluated data; the evaluation module 400 is used to train the evaluation model using the data set, and evaluate the emission status of the heavy-duty vehicle using the trained evaluation model.

[0127] It should be noted that the aforementioned explanation of the embodiment of the remote identification method for heavy-duty vehicles with excessive emissions that can replace the PEMS test is also applicable to the remote identification device for heavy-duty vehicles with excessive emissions that can replace the PEMS test in this embodiment, and will not be repeated here.

[0128] According to the embodiment of the present invention, a remote identification device for over-emission heavy-duty vehicles that can replace the PEMS test is proposed. The first emission average value of the heavy-duty vehicle is calculated based on the test data, and the second emission average value of the heavy-duty vehicle is calculated using multiple evaluation methods and remote transmission data to screen out a target evaluation method equivalent to the power-based window method. This solves the problem that the remote transmission data cannot be directly applied to regulatory standards due to the lack of cycle power parameters. The target evaluation method is used to evaluate the remote transmission data, and the remote transmission data that meets the conditions is screened according to the evaluation results to generate a training set to train the evaluation model, thereby improving the accuracy of online remote identification of the emission status of heavy-duty vehicles. The trained evaluation model is used to evaluate the emission status of heavy-duty vehicles, accurately capturing the dynamic emission patterns of vehicles, and solving the defect that the traditional static method cannot associate multi-window time series information. The applicability is high.

[0129] Figure 6 This is a schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. The electronic device may include:

[0130] A memory 601 , a processor 602 , and a computer program stored in the memory 601 and executable on the processor 602 .

[0131] When the processor 602 executes the program, the remote identification method for heavy-duty vehicles exceeding emission limits, which can replace the PEMS test, is implemented in the above-mentioned embodiment.

[0132] Furthermore, the vehicle further comprises:

[0133] The communication interface 603 is used for communication between the memory 601 and the processor 602 .

[0134] The memory 601 is used to store computer programs that can be run on the processor 602 .

[0135] The memory 601 may include a high-speed RAM memory, and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.

[0136] If the memory 601, processor 602, and communication interface 603 are implemented independently, the communication interface 603, memory 601, and processor 602 can be interconnected via a bus and communicate with each other. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus. Buses can be divided into address buses, data buses, control buses, etc. For ease of representation, Figure 6 Only one thick line is used in the diagram, but this does not mean that there is only one bus or one type of bus.

[0137] Optionally, in a specific implementation, if the memory 601, the processor 602 and the communication interface 603 are integrated on a chip, the memory 601, the processor 602 and the communication interface 603 can communicate with each other through an internal interface.

[0138] The processor 602 may be a central processing unit (CPU), an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present invention.

[0139] An embodiment of the present invention further provides a computer-readable storage medium having a computer program or instruction stored thereon. When the computer program or instruction is executed by a processor, the method for remotely identifying heavy-duty vehicles exceeding emission limits, which can replace the PEMS test, is implemented.

[0140] An embodiment of the present invention further provides a computer program product, including a computer program or instructions, characterized in that when the computer program or instructions are executed, the above-mentioned remote identification method for heavy-duty vehicles with excessive emissions that can replace the PEMS test is implemented.

[0141] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or N embodiments or examples in a suitable manner. In addition, those skilled in the art can combine and combine different embodiments or examples described in this specification and the features of different embodiments or examples without contradiction.

[0142] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the number of technical features indicated. Thus, a feature specified as "first" or "second" may explicitly or implicitly include at least one such feature. In the description of the present invention, "N" means at least two, such as two, three, etc., unless otherwise specifically defined.

[0143] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code comprising one or N executable instructions for implementing a custom logical function or step of a process, and the scope of the preferred embodiments of the present invention includes alternative implementations in which functions may be performed out of the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by those skilled in the art to which the embodiments of the present invention pertain.

[0144] It should be understood that various components of the present invention can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiment, the N steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, it can be implemented using any one or a combination of the following technologies known in the art: discrete logic circuits having logic gate circuits for implementing logical functions on data signals, application-specific integrated circuits having suitable combinational logic gate circuits, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0145] Those skilled in the art will appreciate that all or part of the steps in the method for implementing the above-mentioned embodiment can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium. When the program is executed, it includes one or a combination of the steps of the method embodiment.

Claims

1. A remote identification method for heavy-duty vehicles with excessive emissions that can replace PEMS testing, characterized in that: The following steps are involved: Obtain test data of heavy-duty vehicles and telemetry data sent by the target platform; Calculating a first average emission value of the heavy-duty vehicle based on the test data, and calculating a second average emission value of the heavy-duty vehicle using multiple evaluation methods and the telemetry data; According to the first emission average value and the second emission average value, a target evaluation method is selected from a plurality of evaluation methods, the remote transmission data is evaluated using the target evaluation method, and a data set is generated based on the evaluated data, wherein, according to the first emission average value and the second emission average value, a target evaluation method is selected from a plurality of evaluation methods, including: calculating an emission evaluation equivalence coefficient based on the first emission average value and the second emission average value; screening a target evaluation method based on the emission evaluation equivalence coefficient; wherein, screening a target evaluation method based on the emission evaluation equivalence coefficient includes: screening a plurality of evaluation methods corresponding to emission evaluation equivalence coefficients that are close to a target threshold value; selecting an evaluation method with the smallest absolute value of the emission evaluation equivalence coefficient from a plurality of evaluation methods as the target evaluation method; The evaluation model is trained using the data set, and the emission status of heavy-duty vehicles is evaluated using the trained evaluation model, wherein all windows of the same vehicle in the data set constitute a sample, each sample enters the evaluation model for training in sequence, and the evaluation model calculates all windows of each sample frame by frame in chronological order.

2. The remote identification method for heavy-duty vehicles with excessive exhaust emissions that can replace PEMS testing according to claim 1 is characterized in that: The method of calculating the second emission average value of the heavy-duty vehicle using multiple evaluation methods and the remote transmission data includes: Using multiple evaluation methods to divide the remote transmission data into windows to generate multiple windows of remote transmission data; Calculate the corresponding average NOx emission value based on the remote data of each window; The second emission average value of the heavy-duty vehicle is calculated based on the accumulated NOx emission average values ​​of all windows and the number of windows.

3. The remote identification method for heavy-duty vehicles with excessive exhaust emissions that can replace PEMS testing according to claim 1 is characterized in that: The evaluation model includes an input layer, a long short-term memory network structure layer and an output layer, wherein the input layer inputs the heavy-duty vehicle window feature value generated according to the remote transmission data, each layer of the long short-term memory network structure includes a forget gate, an input gate and an output gate, and the output layer outputs the vehicle emission status, wherein the forget gate calculates the weight of the heavy-duty vehicle window feature value in the previous time step to obtain the first target information; the input gate calculates the weight of the heavy-duty vehicle window feature value in the current time step to obtain the second target information, and fuses the first target information and the second target information to obtain the target information of the current time step, and the output gate generates the vehicle emission status according to the target information of the current time step.

4. The remote identification method for heavy-duty vehicles with excessive exhaust emissions that can replace PEMS testing according to claim 3 is characterized in that: After the evaluation model outputs the heavy-duty vehicle emission status, it includes: If the emission status of the heavy-duty vehicle exceeds the standard, obtaining the actual location and identification code of the heavy-duty vehicle; Determining a target heavy vehicle based on the actual location and identification code of the heavy vehicle; Identify the actual emission status of target heavy-duty vehicles; The estimation model is updated in reverse according to the actual emission status.

5. A remote identification device for heavy-duty vehicles with excessive emissions that can replace PEMS testing, characterized in that: include: Acquisition module, which obtains the test data of heavy-duty vehicles and the telemetry data sent by the target platform; a calculation module, configured to calculate a first average emission value of the heavy-duty vehicle based on the test data, and calculate a second average emission value of the heavy-duty vehicle using a plurality of evaluation methods and the telemetry data; A selection module is configured to select a target evaluation method from a plurality of evaluation methods based on the first emission average value and the second emission average value, evaluate the remote transmission data using the target evaluation method, and generate a data set based on the evaluated data, wherein selecting the target evaluation method from a plurality of evaluation methods based on the first emission average value and the second emission average value includes: calculating an emission evaluation equivalence coefficient based on the first emission average value and the second emission average value; screening a target evaluation method based on the emission evaluation equivalence coefficient; wherein screening the target evaluation method based on the emission evaluation equivalence coefficient includes: screening a plurality of evaluation methods corresponding to emission evaluation equivalence coefficients approaching a target threshold; and selecting an evaluation method having a minimum absolute value of an emission evaluation equivalence coefficient from a plurality of evaluation methods as the target evaluation method; An evaluation module is used to train an evaluation model using the data set and evaluate the emission status of heavy-duty vehicles using the trained evaluation model, wherein all windows of the same vehicle in the data set constitute a sample, each sample enters the evaluation model for training in sequence, and the evaluation model calculates all windows of each sample frame by frame in chronological order.

6. An electronic device, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the remote identification method for over-emission heavy vehicles that can replace the PEMS test as described in any one of claims 1 to 4.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that: The program is executed by a processor to implement the remote identification method for heavy-duty vehicles with excessive emissions that can replace the PEMS test as described in any one of claims 1 to 4.

8. A computer program product, characterized in that The invention comprises a computer program, which, when executed by a processor, is used to implement the remote identification method for heavy-duty vehicles with excessive emission levels that can replace the PEMS test as described in any one of claims 1 to 4.

Images