Motor vehicle emission supervision method and system based on remote measurement and OBD fusion

Abstract

The invention relates to the technical field of motor vehicle exhaust pollutant analysis and monitoring, and particularly discloses a motor vehicle emission supervision method and system based on remote measurement and OBD fusion, and more accurate motor vehicle emission supervision is realized by combining a road remote measurement technology and OBD data. The method comprises the following steps: firstly, acquiring vehicle passing characteristics, actually measured emission concentration and OBD real-time parameters, and generating a group reference interval to perform credibility evaluation; generating a dynamic self-inspection instruction packet for the suspicious vehicle, triggering the vehicle-mounted terminal to execute a diagnosis excitation script at the target remote measurement point, and collecting excitation response data; generating a deep diagnosis conclusion by analyzing the dynamic response consistency of the tail gas concentration in the excitation period and the excitation response data, updating the credibility score of the vehicle data according to the deep diagnosis conclusion, and generating an emission supervision instruction; according to the invention, data authenticity and law enforcement credibility are improved, and optimal configuration of supervision resources is realized.

Description

Technical Field

[0001] This invention belongs to the field of motor vehicle exhaust pollutant analysis and monitoring technology, and relates to a motor vehicle emission monitoring method and system based on the integration of telemetry and OBD. Background Technology

[0002] Motor vehicle emission regulation is a crucial aspect of air pollution control. Its core lies in using specific technologies to measure and analyze pollutants emitted into the atmosphere by in-use vehicles to ensure compliance with environmental regulations. The main regulatory technologies in this area include road telemetry and On-Board Diagnostics (OBD) systems. Road telemetry uses optical and other equipment deployed along the roadside to non-contactly measure and analyze the composition and concentration of exhaust gases emitted by vehicles in motion. On-Board Diagnostics, on the other hand, is an electronic system integrated into the vehicle that monitors engine operating status and emission control system performance parameters in real time. Combining the data obtained from these two technologies to comprehensively assess the actual emissions of vehicles represents a significant development direction in the fields of data analysis and environmental monitoring.

[0003] In existing technologies, to improve regulatory efficiency and coverage, a common approach is to simply aggregate and compare road telemetry data and OBD data. One common practice is to retrieve the vehicle's OBD data when a telemetry device detects excessive emissions, checking for related fault codes. Alternatively, vehicle OBD data is periodically reported to a regulatory platform and loosely correlated with road telemetry screening results, with vehicles showing anomalies simultaneously receiving focused attention. These methods attempt to identify high-emission vehicles by cross-referencing data from two different sources.

[0004] However, the aforementioned existing technical solutions have inherent technical flaws. First, this simple data comparison lacks a deep fusion mechanism. A significant information gap exists between the instantaneous objective data from telemetry and the continuous subjective data from OBD. When contradictions arise—for example, telemetry readings exceeding limits but no fault codes appearing in the OBD—the system struggles to determine whether the issue stems from instantaneous operating conditions or from OBD system tampering or diagnostic blind spots. Second, OBD data itself is generated internally by the vehicle, and its authenticity cannot be effectively guaranteed. It is at risk of being maliciously blocked or modified, rendering judgments based on this data unreliable. Therefore, existing technical combinations often remain at the level of superficial data splicing, failing to create an effective synergy to overcome their respective limitations. Summary of the Invention

[0005] In view of this, in order to solve the problems mentioned in the background technology, a method and system for monitoring motor vehicle emissions based on the integration of telemetry and OBD is proposed.

[0006] The objective of this invention can be achieved through the following technical solution: The first aspect of this invention provides a motor vehicle emission monitoring method based on the fusion of telemetry and OBD, including: S1, acquiring the vehicle traffic characteristics when the target vehicle passes through the telemetry point, the measured emission concentration collected by the telemetry point, and the real-time OBD parameter set synchronously reported by its on-board terminal.

[0007] S2. Based on vehicle traffic characteristics, select multiple mirror vehicles from the database storing historical regulatory data, and process the historical data of the multiple mirror vehicles to generate a group baseline interval.

[0008] S3. Compare and analyze the real-time OBD parameter set with the measured emission concentration joint group benchmark interval to generate the initial credibility assessment results.

[0009] S4. When the initial credibility assessment result indicates that the target vehicle is a suspicious vehicle, generate a dynamic self-test instruction package and determine the next target telemetry point of the target vehicle to bind the task with the dynamic self-test instruction package.

[0010] S5. When the target vehicle arrives at the target telemetry point, a dynamic self-test command package is sent to its on-board terminal to trigger the on-board terminal to execute the diagnostic activation script and collect activation response data.

[0011] S6. Receive the exhaust gas concentration during the excitation period synchronously collected by the target telemetry point and the excitation response data reported by the vehicle terminal, and perform dynamic response consistency analysis on the exhaust gas concentration during the excitation period and the excitation response data to generate in-depth diagnostic conclusions.

[0012] S7. Update the data credibility score of the target vehicle based on the in-depth diagnostic results, and generate the final emission regulatory directive based on the data credibility score.

[0013] The second aspect of the present invention provides a motor vehicle emission monitoring system based on the fusion of telemetry and OBD, comprising: a target vehicle data acquisition module, which acquires the vehicle traffic characteristics when the target vehicle passes through the telemetry point, the measured emission concentration collected by the telemetry point, and the real-time OBD parameter set synchronously reported by its on-board terminal.

[0014] The group baseline interval generation module selects multiple mirror vehicles from the database storing historical regulatory data based on vehicle traffic characteristics, and processes the historical data of the multiple mirror vehicles to generate the group baseline interval.

[0015] The credibility assessment result generation module compares and analyzes the real-time OBD parameter set with the measured emission concentration joint group benchmark interval to generate the initial credibility assessment result.

[0016] The dynamic self-test instruction package generation module generates a dynamic self-test instruction package when the initial credibility assessment result indicates that the target vehicle is a suspicious vehicle, and determines the next target telemetry point of the target vehicle to bind the dynamic self-test instruction package to the task.

[0017] The excitation response data acquisition module sends a dynamic self-test command package to the vehicle terminal when the target vehicle arrives at the target telemetry point, so as to trigger the vehicle terminal to execute the diagnostic excitation script and collect excitation response data.

[0018] The deep diagnostic conclusion generation module receives the exhaust gas concentration during the excitation period synchronously collected by the target telemetry point and the excitation response data reported by the vehicle terminal, and performs dynamic response consistency analysis on the exhaust gas concentration during the excitation period and the excitation response data to generate a deep diagnostic conclusion.

[0019] The emission regulatory instruction generation module updates the data credibility score of the target vehicle based on the in-depth diagnostic results, and generates the final emission regulatory instruction based on the data credibility score.

[0020] Compared with the prior art, the embodiments of the present invention have at least the following advantages or beneficial effects: (1) The present invention achieves an objective assessment of the credibility of vehicle OBD data by dynamically constructing a group benchmark interval and comparing it with individual data, overcomes the regulatory distortion defect caused by the tamperability of a single OBD data source, improves the data authenticity and law enforcement credibility of the entire emission supervision system, and ensures the reliability of the analysis data.

[0021] This invention designs a spatiotemporally triggered dynamic self-testing mechanism, upgrading traditional passive data acquisition to proactive, on-the-go system function testing. By analyzing the vehicle's internal and external dynamic responses to minute disturbances, it can accurately identify hidden faults, early performance degradation, and targeted cheating behaviors that are difficult to detect during steady-state operation, improving the depth and accuracy of fault diagnosis and making the tracing of the causes of excessive emissions more precise.

[0022] This invention employs a strategy that combines broad-area initial screening with precise puncture to conduct low-cost and high-efficiency preliminary screening of a massive number of vehicles, while initiating in-depth diagnostic processes only for a small number of suspicious vehicles. This achieves optimized allocation and efficient utilization of regulatory resources, ensuring regulatory coverage while reducing the cost of monitoring per vehicle and improving overall regulatory effectiveness.

[0023] This invention constructs a closed-loop feedback system from identification and diagnosis to learning. By feeding confirmed new anomaly case data back into the analysis model, the system can continuously learn and evolve, automatically adapting to and identifying new failure modes and cheating methods, thereby enabling the regulatory system to possess sustainable and dynamic anti-interference capabilities. Attached Figure Description

[0024] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 This is a schematic diagram of the method steps of the present invention.

[0026] Figure 2 This is a schematic diagram of the system structure connection of the present invention. Detailed Implementation

[0027] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0028] Please see Figure 1 The first aspect of the present invention provides a motor vehicle emission monitoring method based on the fusion of telemetry and OBD, comprising: S1, acquiring the vehicle traffic characteristics when the target vehicle passes through the telemetry point, the measured emission concentration collected by the telemetry point, and the real-time OBD parameter set synchronously reported by its on-board terminal.

[0029] In a specific embodiment of the present invention, obtaining the vehicle passage characteristics when the target vehicle passes through the telemetry point includes: capturing vehicle images by an image recognition unit deployed at the telemetry point, and parsing out the identity information of the target vehicle from them.

[0030] It should be noted that the purpose of this step is to capture a precise and multi-dimensional on-site context for each telemetry event, generating structured vehicle traffic characteristics. These characteristics are not only unique identifiers associated with specific vehicles, but also the foundational data input for subsequent homogeneous group comparisons and operational condition analyses, ensuring the accuracy and traceability of all subsequent analyses. The process begins when the target vehicle physically enters the effective monitoring area of ​​a pre-defined telemetry point. The image recognition unit deployed at the telemetry point, typically a system integrating a high-definition industrial camera and dedicated supplementary lighting, automatically triggers and captures images of the vehicle's rear. This image is immediately fed into an optical character recognition algorithm module running on an embedded or edge computing node. This module, specifically trained for license plate characters, can parse the license plate number and color from the image within milliseconds, forming the target vehicle's identification information.

[0031] The instantaneous speed and instantaneous acceleration of the target vehicle are measured by auxiliary sensing units deployed at telemetry points.

[0032] It should be noted that, simultaneously with image capture, an auxiliary sensing unit working in conjunction with the camera is also activated. This unit is typically implemented using a Doppler radar or lidar sensor. The auxiliary sensing unit emits electromagnetic waves or laser beams along the vehicle's path and, by analyzing the frequency or time-of-flight changes of the echo signal, accurately measures the instantaneous speed of the vehicle as it passes a specific measurement line. To obtain this instantaneous acceleration, the auxiliary sensing unit performs two consecutive speed measurements within a very short time interval, such as 100 milliseconds. The vehicle's acceleration value is then obtained by calculating the rate of change of speed, which is calculated by dividing the speed difference between the two measurements by the time difference.

[0033] The vehicle's traffic characteristics are formed by combining identification information, instantaneous speed, and instantaneous acceleration.

[0034] It should be noted that, ultimately, the data aggregation and preprocessing module deployed at the telemetry points will structurally combine the discrete data generated by the two parallel processes mentioned above. It will encapsulate the parsed identification information, instantaneous velocity and instantaneous acceleration values ​​synchronized with precise timestamps, into a unified data record, namely, vehicle passage characteristics. These vehicle passage characteristics, as an indivisible data unit, are uploaded to the cloud-based integrated monitoring platform, providing accurate and reliable input for subsequent retrieval of vehicle files based on vehicle identity and for filtering mirror vehicles with similar operating conditions based on their dynamic operating parameters.

[0035] S2. Based on vehicle traffic characteristics, select multiple mirror vehicles from the database storing historical regulatory data, and process the historical data of the multiple mirror vehicles to generate a group baseline interval.

[0036] In a specific embodiment of the present invention, the specific steps of processing historical data of multiple mirror vehicles to generate a group baseline interval include: extracting a subset of historical OBD parameters and historical telemetry emission concentration values ​​corresponding to multiple mirror vehicles from a database storing historical regulatory data.

[0037] A joint statistical analysis was conducted on a subset of historical OBD parameters and historical telemetry emission concentration values ​​to establish a joint distribution model that can characterize the normal operating state of vehicles.

[0038] It should be noted that the purpose of this step is to provide a dynamic and highly correlated group reference system, i.e., a group baseline interval, for the emission and operating condition data of individual vehicles. This process replaces the traditional static and universal threshold by jointly modeling historical data of a large number of homogeneous vehicles under specific driving conditions, thereby achieving a precise and context-aware assessment of the reliability of the target vehicle's data. The system first performs a multi-dimensional matching query in the historical regulatory database based on the vehicle traffic characteristics obtained from the target vehicle to extract data from multiple mirror vehicles. The matching conditions for this query include vehicle model, engine displacement, fuel type, geographical identifier of the driving segment, and speed and acceleration range when passing telemetry points, for example, speed between 55-65 km / h and acceleration between -0.2 and +0.2 m / s². 2 Within this range, a dataset containing, for example, 50 to 500 mirror vehicles is selected. Then, for each selected mirror vehicle, the system extracts a subset of its historical OBD parameters when passing the same preset telemetry point and historical telemetry emission concentration values ​​synchronized with its precise timestamp. The subset of historical OBD parameters includes key operating condition parameters such as engine load percentage, engine speed, and air-fuel ratio feedback values, while the historical telemetry emission concentration values ​​are the measured concentrations of carbon monoxide, hydrocarbons, or nitrogen oxides. The system treats each set of "historical OBD parameter subset - historical telemetry emission concentration values" as a high-dimensional data point and applies joint statistical analysis to this high-dimensional data point set formed by all mirror vehicles. In engineering, a Gaussian mixture model can be used to fit this joint distribution; this model can identify the dominant data clusters representing normally operating vehicles in the dataset.

[0039] It should also be noted that, to achieve joint statistical analysis of historical data, this invention specifically employs a Gaussian mixture model to fit the probability density of this high-dimensional data point set in engineering practice. First, a subset of historical OBD parameters – historical telemetry emission concentration values ​​– for each mirror vehicle is considered as a data point in a high-dimensional space, and the data from all mirror vehicles are aggregated to form a training set. The Gaussian mixture model assumes that this training set is composed of multiple potential Gaussian distributions (i.e., data clusters), each representing a specific vehicle operation and emission coupling state. The fitting process uses iterative algorithms such as expectation maximization, continuously calculating the posterior probability of each data point generated by each Gaussian distribution, and updating the mean, covariance, and mixing weights of each Gaussian distribution based on this probability until the model parameters converge. After fitting, the system analyzes the mixing weights of each Gaussian distribution component. The Gaussian distribution component with the largest weight, representing the densest and most mainstream sample group in the dataset, is identified as the dominant data cluster characterizing normally operating vehicles. The mean vector and covariance matrix of this dominant data cluster mathematically define the intrinsic correlation and fluctuation range between OBD parameters and emission concentrations under normal operating conditions, providing a solid model foundation for the subsequent generation of population baseline intervals.

[0040] Based on the joint distribution model, a population baseline interval is derived, which includes the normal range of OBD parameters and the expected range of emission concentrations that are dynamically correlated with the normal range of OBD parameters.

[0041] It should be noted that, based on the fitted Gaussian mixture model, the system ultimately generates a population baseline interval. This baseline interval consists of two interrelated parts. The first part is the normal OBD parameter interval, defined as the high-probability projection region of the dominant Gaussian component in the Gaussian mixture model onto the OBD parameter dimension; for example, a confidence ellipsoid covering 95% of the data points for that component. The second part is the expected emission concentration interval associated with the normal OBD parameter interval. This interval is not a fixed range but is dynamically generated. Its calculation process can be expressed as follows: , ,in, A vector representing the real-time OBD parameter set of the target vehicle, which is obtained through the vehicle terminal; The Gaussian mixture model, which has been trained, is stored on a cloud-based fusion monitoring platform. This is a function that calculates the conditional expectation value based on a Gaussian mixture model, used to output the conditional expectation value given an OBD parameter vector. Below, the expected average emission concentration ; This represents the standard deviation of the expected emission concentration output by the Gaussian mixture model under the same conditions. This is a confidence coefficient, typically with a value of 2, representing a statistical confidence level. This is the final calculated expected range of emission concentrations. This process ensures that the emission concentration assessment criteria can be dynamically adjusted according to the real-time operating conditions of the vehicle, thus providing an accurate and practically meaningful reference for subsequent joint comparisons.

[0042] S3. Compare and analyze the real-time OBD parameter set with the measured emission concentration joint group benchmark interval to generate the initial credibility assessment results.

[0043] In a specific embodiment of the present invention, the specific steps for generating the initial credibility assessment result include: determining whether the real-time OBD parameter set falls within the normal range of OBD parameters in the group baseline range.

[0044] Based on the real-time parameter set of OBD, the corresponding expected emission concentration range is obtained from the group baseline range, and it is determined whether the measured emission concentration falls within the expected emission concentration range.

[0045] Based on the combination of the above two judgment results, an initial credibility assessment result containing a clear abnormal pattern identifier is generated.

[0046] It should be noted that the purpose of this step is to perform a two-dimensional cross-validation based on population data to conduct a preliminary and rapid health status screening and credibility classification of a single passage event of the target vehicle. The core of this process is to avoid evaluating OBD or telemetry data in isolation, but rather to accurately classify the vehicle status into different abnormal patterns through a pre-set logical matrix, thereby providing clear guidance for subsequent differentiated diagnosis.

[0047] The system first performs a first-dimensional judgment, comparing the real-time OBD parameter set uploaded by the vehicle terminal with the normal range of OBD parameters in the group benchmark interval to determine whether the OBD data is "normal within the range". Next, the system performs a second-dimensional judgment, comparing the measured emission concentration obtained by the telemetry device with the expected emission concentration range dynamically generated based on the current OBD parameters to determine whether the emission result is "normal within expectations". Finally, the system combines the results of the above two dimensions to generate a structured initial credibility assessment result. This result is not just a simple "pass / fail" conclusion, but includes a clear anomaly pattern identifier, specifically categorized as follows: Normal Mode: When the real-time OBD parameter set falls within the normal range of OBD parameters, and the measured emission concentration also falls within the expected emission concentration range. In this case, the vehicle is judged as normal, requiring no further in-depth diagnosis. Anomaly Mode A: Data-Performance Contradiction: When the real-time OBD parameter set falls within the normal range of OBD parameters, but the measured emission concentration does not fall within the expected emission concentration range, the system generates this assessment result. This mode presents a clear logical contradiction: the vehicle's internal system claims to be "normal," but its external performance is "abnormal." This strongly suggests that the OBD system has diagnostic blind spots, performance degradation, or that its data has been maliciously tampered with or blocked. Therefore, vehicles in this mode are marked as having suspicious OBD data and are considered the highest priority suspicious vehicles. Abnormal Mode B: Joint Indicator Abnormality: This assessment result is generated when the real-time OBD parameter set does not fall within the normal range for OBD parameters, and the measured emission concentration also does not fall within the expected emission concentration range. In this mode, both internal data and external performance point to a problem with the vehicle, constituting a statistically significant event indicating a high probability of physical fault. Abnormal Mode C: Internal Warning - External Normal: This assessment result is generated when the real-time OBD parameter set does not fall within the normal range for OBD parameters, but the measured emission concentration still falls within the expected emission concentration range. This mode may indicate that the OBD system is overly sensitive, has false alarms, or that the emission control system still has redundancy. Vehicles in this mode have lower priority but can serve as samples for long-term monitoring and model optimization.

[0048] S4. When the initial credibility assessment result indicates that the target vehicle is a suspicious vehicle, generate a dynamic self-test instruction package and determine the next target telemetry point of the target vehicle to bind the task with the dynamic self-test instruction package.

[0049] In a specific embodiment of the present invention, the specific steps of generating a dynamic self-test instruction package and determining the next target telemetry point of the target vehicle to bind the dynamic self-test instruction package to a task include: selecting a matching script template from the diagnostic script template library to generate a dynamic self-test instruction package based on the abnormal pattern indicated in the initial confidence assessment result.

[0050] It should be noted that the purpose of this step is to transform the assessment conclusions generated in the previous steps, which contain clear anomaly patterns, into a precise follow-up regulatory action that matches the risk level. This process reflects the optimized allocation of regulatory resources, taking intervention measures of varying strengths for different types of potential problems. Based on the anomaly pattern identifiers clearly identified in the initial credibility assessment results, the cloud-based integrated regulatory platform executes the following differentiated responses: For "Anomaly Pattern A: Data-Performance Contradiction" or "Anomaly Pattern B: Joint Indicator Anomaly": These two patterns represent a high probability of data fraud and a high probability of physical failure, respectively, both posing a direct or potential serious threat to the environment, and are therefore listed as the highest priority. The system will immediately initiate a deep diagnostic process, selecting a targeted script template from a pre-set diagnostic script template library. For example, for "Anomaly Pattern A," the system will select a script template designed to "expose the deception." This template might include making minor disturbances to the air-fuel ratio to test the true response speed of the oxygen sensor, or briefly changing the operating conditions to observe the true rate of change in catalytic converter temperature, thereby verifying the actual performance of these key components. In "Abnormal Mode B," the system selects a script template designed for "fault reproduction and confirmation." This template may guide the vehicle into specific operating conditions (such as high load or a specific RPM range) to observe whether the fault parameters reported by the OBD stably appear and verify their correlation with external emissions. After selecting a suitable script template, the system generates a customized dynamic self-test instruction package and plans the optimal execution time and space node (i.e., the target telemetry point), achieving a leap from passive analysis to proactive activation. For "Abnormal Mode C: Internal Warning - External Normal": In this mode, the vehicle is not currently producing excessive emissions, and the direct risk is low. Therefore, to conserve regulatory resources, the system will not immediately trigger the costly dynamic self-test instruction package issuance process. Instead, the system will implement an "observation and learning" strategy: 1) Marking and recording: The system marks the vehicle as a "potential risk observation object" and records the time, operating conditions, and specific OBD abnormal parameters of the event in detail in its digital file; 2) Increasing monitoring level: The system can appropriately increase the frequency of subsequent routine data reporting or analysis priority for the vehicle to more closely track its status evolution. If the vehicle frequently experiences such situations in a short period of time, or subsequently transforms into abnormal mode A or B, its risk level will automatically increase; 3) Data feedback to the model: More importantly, this type of "internal warning but external normal performance" data, as an important sample for calibrating the model, will be automatically input into the learning and optimization module of the group baseline interval. By analyzing a large number of such samples, the system can learn and calibrate the diagnostic threshold sensitivity of the OBD system for specific vehicle models, thereby improving the overall regulatory system's accuracy in identifying real fault warnings and reducing misjudgments of occasional, non-critical OBD fluctuations.

[0051] The system acquires the real-time location and driving status data of the target vehicle and combines this data with road network information to predict its future driving trajectory.

[0052] It should be noted that, next, the system must determine a physical location and time for the issuance of this dynamic self-check command package. This process is accomplished by predicting the target vehicle's trajectory. Using real-time GPS coordinates, direction of travel, and speed reported by the vehicle's onboard terminal, combined with the integrated digital road network map, the system predicts the most likely path the vehicle will take in the next time period, such as the next 5 to 15 minutes. The system will identify all candidate telemetry points along this path where roadside units are deployed. To select the optimal target telemetry point from these candidate points, the system uses a scoring function for decision-making, which can be expressed as: ,in, It is the first The scores of each candidate telemetry point; The estimated arrival time is calculated using a path planning algorithm. The time of each candidate telemetry point; It is a time window function; It is the confidence level of the path prediction; and These are preset weighting coefficients used to balance time suitability and prediction reliability. The system will select the candidate telemetry point with the highest score as the final target telemetry point.

[0053] It should also be noted that, firstly, regarding the estimated arrival time... The calculation is not a simple division of distance by speed limit, but rather a precise calculation using an advanced path planning algorithm that integrates real-time traffic data. This algorithm models the urban road network as a graph, where intersections are nodes and roads are edges. The weight (i.e., toll cost) of each edge is dynamically obtained from a real-time traffic flow API provided by a third-party map service provider, reflecting current traffic conditions such as congestion and accidents. The algorithm starts with the vehicle's current GPS location and uses candidate telemetry points... Using the destination as the starting point, calculate the path with the shortest total travel time. The total time for this path is... .

[0054] Secondly, the time window function It is designed as a normalized Gaussian function to quantify the suitability of the arrival time. Its specific formula is: ,in, Indicates the first The estimated arrival times of the candidate telemetry points Indicates the ideal central time. This represents the standard deviation of the control time window width. This represents the natural constant. In this embodiment, the ideal center time... It can be set to 5.5 minutes, which represents the ideal time from the issuance of the command to the vehicle stably entering the test condition and reaching the telemetry point; standard deviation It can be set to 1.5 minutes, which controls the width of the time window. This design makes it possible to... At exactly 5.5 minutes, The maximum value is 1; when Within a 3 to 8 minute window, The value is relatively high; while when When the time is much less than 3 minutes or much greater than 8 minutes, The value will smoothly and rapidly approach 0, thus effectively penalizing candidate points that are not time-appropriate.

[0055] Secondly, path prediction confidence. The calculation combines the inherent complexity of the path with historical behavior data. The specific calculation formula is as follows: ,in, For historical paths to follow probability, This is an adjustment factor, for example, 0.1. From the vehicle's current position to the... The number of decision points that need to be turned on the optimal path of each candidate point. This refers to the distance from the vehicle's current location to the candidate point. The optimal path is determined by the number of intersections where the driving direction needs to be changed. This design makes the simpler and closer to a straight line the path, the closer its attenuation factor is to 1; while the more winding and turning the path is, the greater the possibility that the driver will deviate from the intended route, and the smaller the attenuation factor will be. This is derived by analyzing the historical trajectory data of the vehicle model or the driver, representing the statistical probability that they will follow navigation or conventional routes when choosing similar paths. If historical data shows that there are multiple popular routes after the fork in the road, The value will decrease accordingly. Therefore, The higher the value, the greater the system's confidence that the vehicle will actually pass through the candidate point.

[0056] Finally, regarding the preset weighting coefficients and These are used to balance the relative importance of time and certainty in decision-making, and are typically set to a sum of 1. In a typical regulatory application scenario, The value should be higher than ,For example, 0.7 is acceptable. A value of 0.3 is acceptable. This value is based on the premise that ensuring successful vehicle interception and testing is the primary objective of the mission, taking precedence over testing at a "perfect" time. Choosing a point with high certainty but slight timing deviation is far better than choosing a point with perfect timing where the vehicle is unlikely to pass through at all. Of course, these two weighting coefficients are configurable parameters of the system, allowing regulatory agencies to flexibly adjust them according to specific enforcement strategies or urban traffic characteristics.

[0057] Based on the future driving trajectory, the optimal telemetry point is selected from multiple candidate telemetry points as the target telemetry point, and the dynamic self-test command package is bound to the identifier and trigger time window of the target telemetry point.

[0058] It should be noted that, finally, the system binds the generated dynamic self-check instruction package, the unique identifier of the selected target telemetry point, and a preset trigger time window calculated based on the expected arrival time into a data structure. This bound task record is stored in the pending execution queue of the cloud-based integrated monitoring platform, waiting to be triggered and issued when the target vehicle physically enters the communication range of the target telemetry point and meets the time window conditions.

[0059] It should also be noted that, finally, the system binds the generated dynamic self-test instruction package, the unique identifier of the selected target telemetry point, and a preset trigger time window calculated based on the estimated arrival time into a data structure. The calculation method for this trigger time window can be simply understood as setting a "smart alarm clock" for the roadside detection equipment, ensuring it "wakes up" and "monitors" the target vehicle within the appropriate time period. The specific calculation process consists of three steps: 1) Determining a center point: First, the system calculates the vehicle's "estimated arrival time" (e.g., 3:10 PM) based on map navigation and uses this time as the center reference point for the entire time window. 2) Setting the start time (lead time): To avoid missing potentially early-arriving vehicles, the system sets an advance start time. This lead time is dynamically calculated: on the one hand, it ensures the equipment has at least a fixed "warm-up preparation time" of several seconds; on the other hand, it adds a dynamic buffer based on the "uncertainty" of this journey. For example, if the vehicle's route is simple, almost a straight line, then the uncertainty is low and the dynamic buffer is short; if the route is complex, requiring multiple turns, then the uncertainty is high and the dynamic buffer will be longer. The system will take the longer of the "warm-up time" and the "dynamic buffer" as the final lead time. 3) Setting the end time (delay): Considering real-world factors such as traffic jams and traffic lights, the probability of a vehicle arriving late is much greater than arriving early. Therefore, the system will extend a "tolerance time" backward from the center point. This tolerance time is also linked to the "uncertainty" of the journey; the longer and more complex the route, the longer the tolerance waiting time given by the system, usually much larger than the lead time. In this way, the system calculates the trigger time window: for simple, predictable journeys, the window is narrow, saving equipment energy; while for complex, variable journeys, the window will be widened accordingly to ensure that the vehicle can be successfully captured regardless of whether it arrives early or late. This ensures both a very high monitoring success rate and the most efficient use of system resources.

[0060] S5. When the target vehicle arrives at the target telemetry point, a dynamic self-test command package is sent to its on-board terminal to trigger the on-board terminal to execute the diagnostic activation script and collect activation response data.

[0061] In a specific embodiment of the present invention, the specific steps of triggering the vehicle terminal to execute the diagnostic excitation script and collect excitation response data include: the vehicle terminal parses the dynamic self-test instruction package and sends a parameter disturbance instruction to the vehicle engine control unit.

[0062] It's important to note that the purpose of this step is to transform the abstract test commands issued from the cloud into a series of precise, controlled physical actions and high-frequency data acquisition behaviors within the vehicle. When the onboard terminal receives the verified dynamic self-test command package, it first parses the command content and enters execution mode. Based on the preset diagnostic trigger script included in the command, the onboard terminal sends a preset parameter perturbation command to the engine control system through its interface with the vehicle's CAN bus. This command does not directly modify the engine's core operating pattern, but rather fine-tunes a control parameter within a very short time (e.g., 200-500 milliseconds) by sending simulated sensor signals or diagnostic commands. For example, the command might require temporarily shifting the target air-fuel ratio by 2% from its current position, or increasing or decreasing the fuel quantity by 5% in a single injection pulse. This level of manipulation is carefully designed so that its impact on engine output torque is far below the driver's perception threshold, ensuring the concealment and safety of the operation.

[0063] While sending parameter perturbation commands, the system records time-series waveforms of multiple deep response parameters in the vehicle emission control system at a higher acquisition frequency than the conventional polling rate.

[0064] It should be noted that at the precise moment the preset parameter disturbance command is sent, the time synchronization module inside the vehicle terminal triggers a preset high-speed data recording task. This task is no longer the conventional low-frequency OBD data polling (usually 1-10Hz), but rather a targeted acquisition of multiple preset deep response parameters in the OBD system at a higher data acquisition rate (e.g., 50-100Hz). These parameters are key characteristics of the dynamic behavior of the emission system, directly reflecting the health and response speed of the control loop. For example, the vehicle terminal will read the voltage waveform of the oxygen sensor signal at high speed to capture its response delay and jump rate under changes in air-fuel ratio; simultaneously, it will record the instantaneous readings of the temperature sensors before and after the catalytic converter to assess the rate of chemical reaction heat change of the catalytic converter in response to changes in the input exhaust gas composition; in addition, it will monitor the value of the misfire detection counter to confirm whether the disturbance has unexpectedly caused combustion instability.

[0065] The recorded time-series waveforms are encapsulated as excitation response data, in which deep response parameters include oxygen sensor response signals, catalyst temperature change rate, and engine misfire count values.

[0066] It should be noted that the high-speed data recording process lasts for a preset short window, such as 1-2 seconds before and after the command is sent, to fully capture the entire process from the application of the disturbance to the system reaching a new transient equilibrium. During this period, all collected high-frequency data points are cached in the memory of the vehicle terminal. After the recording window ends, the vehicle terminal encapsulates the waveform changes of all deep response parameters in the cache into a data packet, namely the excitation response data, according to a preset data format. This data packet is timestamped and assigned a task ID, ready to be reported to the cloud-based fusion monitoring platform via the wireless communication module, thus completing the closed loop from command execution to data generation.

[0067] S6. Receive the exhaust gas concentration during the excitation period synchronously collected by the target telemetry point and the excitation response data reported by the vehicle terminal, and perform dynamic response consistency analysis on the exhaust gas concentration during the excitation period and the excitation response data to generate in-depth diagnostic conclusions.

[0068] In a specific embodiment of the present invention, the specific steps of performing dynamic response consistency analysis on the exhaust gas concentration and excitation response data during the excitation period to generate in-depth diagnostic conclusions include: extracting a set of internal response features from the excitation response data and extracting a set of external emission response features from the exhaust gas concentration during the excitation period.

[0069] It should be noted that the purpose of this step is to perform deep fusion and causal analysis on the vehicle's internal response data and external emission data collected synchronously in the previous steps, in order to reveal the true dynamic performance of the emission control system under active excitation. The core of this process is to establish and evaluate a quantitative dynamic transfer characteristic model, transforming fuzzy, qualitative data anomalies into quantitative diagnostic conclusions about the system's health status, thereby identifying hidden faults or tampering. The system first extracts features from the two sets of received data. For the excitation response data reported from the onboard terminal, the system extracts key quantitative indicators characterizing the system's internal dynamic response, i.e., internal response features. For example, for the voltage waveform of the oxygen sensor signal, the system acquires its response delay time, rise and fall times, and steady-state switching frequency; for catalytic converter temperature changes, it acquires its temperature rise slope. These features quantify the response speed and magnitude of key components in the control loop to disturbances. Simultaneously, the system processes the exhaust gas concentration data during the excitation period, synchronously collected from the target telemetry point, to extract external emission response features. This includes the timing, height, and width of peak concentrations of specific pollutants in the exhaust gas, which quantify the macroscopic results ultimately output by the entire system.

[0070] Establish a dynamic transfer characteristic model to describe the causal relationship between internal response characteristics and external emission response characteristics.

[0071] The dynamic transmission characteristic model is compared with the benchmark dynamic transmission characteristic model representing a healthy vehicle to calculate the deviation index, and a deep diagnostic conclusion is generated based on the deviation index.

[0072] It should be noted that, next, the system will perform time-series correlation and causal relationship modeling between the extracted internal response features and external emission response features. Since the application time of the internal disturbance command is precisely known, the system can use this as a time reference to analyze the time lag relationship and amplitude correlation between various response features. In engineering, this can be implemented to construct a simplified system identification model to describe the dynamic transfer characteristics from internal disturbance to external emission output. This model can be a transfer function or a state-space model, whose parameters are fitted by the extracted response features. For example, the delay time from the disturbance command to the occurrence of the peak concentration of a specific pollutant, and the proportional relationship between the amplitude of the peak concentration of a specific pollutant and the catalytic converter temperature rise slope, can be used as the core parameters of this dynamic transfer characteristic. Finally, the system will quantitatively compare the dynamic transfer characteristics constructed for the target vehicle with a preset benchmark of dynamic transfer characteristics for healthy vehicles. This benchmark dynamic transfer characteristic model is established in advance by executing the same micro-diagnostic excitation script on a large number of healthy vehicles of the same model and analyzing their response data; it represents the dynamic response paradigm that a normal system should have. The comparison process calculates the deviation index between the two. To accomplish this, it can be represented as: ,in, The first is the dynamic transmission characteristic of the target vehicle. Parameter values, such as response latency; These are the corresponding parameter values ​​in the baseline dynamic transmission characteristic model of a healthy vehicle; It is the first The standard deviation of each parameter in a healthy vehicle population is used for normalization. It is the first The system calculates the degree of deviation by assigning importance weights to each parameter in the diagnosis. Compare with a preset threshold. If If the threshold is exceeded, the system will determine that there is a significant abnormality in the vehicle's emission control system and generate a clear, in-depth diagnostic conclusion indicating a hidden fault or data tampering.

[0073] It should also be noted that if this difference measure value If the emissions exceed a preset threshold, the system determines that the vehicle's emissions control system has a significant anomaly. A typical value for this preset threshold is 50%. This value is chosen because it is significantly higher than the normal combined measurement error of telemetry and on-board sensors, effectively avoiding false alarms caused by normal equipment fluctuations; at the same time, this threshold is sensitive enough to identify significant deviations caused by severe aftertreatment system failure or data fraud, as such anomalies typically result in differences far exceeding 50%. Furthermore, this setting also references the regulatory concept of allowing reasonable fluctuations in real-world emissions in mainstream emissions regulations, ensuring the scientific rigor and impartiality of the diagnostic conclusions.

[0074] S7. Update the data credibility score of the target vehicle based on the in-depth diagnostic results, and generate the final emission regulatory directive based on the data credibility score.

[0075] In a specific embodiment of the present invention, the specific steps of updating the data credibility score of the target vehicle based on the in-depth diagnostic conclusion and generating the final emission regulatory instruction based on the data credibility score include: if the in-depth diagnostic conclusion confirms a real physical fault, then the OBD data credibility weight of the target vehicle is lowered and its telemetry data credibility weight is raised accordingly.

[0076] If the in-depth diagnostic conclusion confirms that the OBD data has been tampered with, the OBD data credibility weight of the target vehicle will be set to the lowest level.

[0077] The updated OBD data confidence weights and telemetry data confidence weights will be applied to all future data analyses of the target vehicle.

[0078] It's important to note that the purpose of this step is to dynamically adjust the system's reliance on different data sources for a specific vehicle based on the final conclusions of the in-depth diagnostics. The core of this process is to transform the system health status or data authenticity revealed by the in-depth diagnostics into a quantifiable data credibility score that guides future regulatory strategies. By dynamically modifying the weights of OBD and telemetry data in subsequent analyses, the system achieves individualized and adaptive adjustments to the regulatory strategy for that vehicle. The system first obtains the in-depth diagnostic conclusions generated in the previous step and performs differentiated weight updates based on the nature of these conclusions. The system maintains a data credibility score file for each vehicle, containing the credibility weights for both OBD and telemetry data. These two weights may initially be set equally or based on a baseline value derived from the vehicle's historical performance. When the in-depth diagnostic conclusion confirms a genuine fault, such as "catalytic converter aging leading to insufficient conversion efficiency," this indicates that the telemetry data accurately reflects the actual excessive emissions, while the OBD system may have failed to report the fault in a timely manner due to its wide diagnostic threshold settings or blind spots. In this case, the system will perform weight adjustments. The OBD data reliability weight will be lowered, for example, from 0.5 to 0.3, while the telemetry data reliability weight will be correspondingly increased, from 0.5 to 0.7. Conversely, if the in-depth diagnostic results confirm clear signs of tampering with the OBD real-time parameter set, such as "oxygen sensor signals being spoofed by the simulator," it means that the vehicle's OBD data source is completely unreliable. In this case, the system will perform a more aggressive weight adjustment. The OBD data reliability weight for the target vehicle will be directly set to the lowest level, typically 0. In engineering terms, this means that in all future emissions analysis models for this vehicle, the weight coefficient of the OBD data input will be set to zero, and the system will completely ignore any OBD parameters uploaded by the vehicle. At this point, the vehicle's emissions status assessment will rely entirely on its historical and future telemetry data sequences for analysis, such as by performing time-series trend analysis on multiple telemetry results to determine its emissions deterioration. This updated data reliability score is not only an internal parameter but also serves as an important basis for generating subsequent emissions regulatory directives. For example, vehicles with extremely low OBD data reliability weighting will still be marked as high-risk monitoring targets by the system, even if their subsequent uploaded OBD data shows normal results. This may increase the frequency with which they are triggered to execute dynamic self-check commands, or directly generate on-site inspection work orders. In this way, the system ensures that the results of each in-depth diagnosis can effectively and sustainably influence the regulatory approach for that specific vehicle, achieving intelligent allocation of regulatory resources.

[0079] Reference Figure 2The second aspect of the present invention provides a motor vehicle emission monitoring system based on the fusion of telemetry and OBD, comprising: a target vehicle data acquisition module, a group benchmark interval generation module, a credibility assessment result generation module, a dynamic self-test instruction package generation module, an excitation response data acquisition module, a deep diagnostic conclusion generation module, and an emission monitoring instruction generation module.

[0080] The target vehicle data acquisition module is connected to the group baseline interval generation module. Both the target vehicle data acquisition module and the group baseline interval generation module are connected to the credibility assessment result generation module. The credibility assessment result generation module is connected to the dynamic self-test instruction package generation module. The dynamic self-test instruction package generation module is connected to the excitation response data acquisition module. The excitation response data acquisition module is connected to the deep diagnostic conclusion generation module. The deep diagnostic conclusion generation module is connected to the emission regulatory instruction generation module.

[0081] The target vehicle data acquisition module acquires the vehicle traffic characteristics when the target vehicle passes through the telemetry point, the measured emission concentration collected by the telemetry point, and the real-time OBD parameter set synchronously reported by its on-board terminal.

[0082] The group baseline interval generation module selects multiple mirror vehicles from a database storing historical monitoring data based on vehicle traffic characteristics, and processes the historical data of the multiple mirror vehicles to generate the group baseline interval.

[0083] The credibility assessment result generation module compares and analyzes the real-time OBD parameter set with the measured emission concentration joint group benchmark interval to generate an initial credibility assessment result.

[0084] When the initial credibility assessment result indicates that the target vehicle is a suspicious vehicle, the dynamic self-test instruction package generation module generates a dynamic self-test instruction package and determines the next target telemetry point of the target vehicle to bind the dynamic self-test instruction package to the task.

[0085] The excitation response data acquisition module sends a dynamic self-test command package to the vehicle terminal when the target vehicle arrives at the target telemetry point, so as to trigger the vehicle terminal to execute the diagnostic excitation script and collect excitation response data.

[0086] The deep diagnostic conclusion generation module receives the exhaust gas concentration during the excitation period synchronously collected by the target telemetry point and the excitation response data reported by the vehicle terminal, and performs dynamic response consistency analysis on the exhaust gas concentration during the excitation period and the excitation response data to generate a deep diagnostic conclusion.

[0087] The emission regulatory instruction generation module updates the data credibility score of the target vehicle based on the in-depth diagnostic conclusions, and generates the final emission regulatory instruction based on the data credibility score.

[0088] The above content is merely an example and illustration of the concept of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described, or use similar methods to replace them, as long as they do not deviate from the concept of the invention or exceed the scope defined by the present invention, and all such modifications and additions should fall within the protection scope of the present invention.

Claims

1. A method for monitoring motor vehicle emissions based on the fusion of telemetry and OBD, characterized in that, include: S1. Obtain the vehicle traffic characteristics when the target vehicle passes through the telemetry point, the measured emission concentration collected by the telemetry point, and the real-time OBD parameter set synchronously reported by its on-board terminal. S2. Based on vehicle traffic characteristics, select multiple mirror vehicles from the database storing historical monitoring data, and process the historical data of the multiple mirror vehicles to generate a group baseline interval. S3. Compare and analyze the real-time OBD parameter set with the measured emission concentration combined with the group benchmark interval to generate an initial credibility assessment result. S4. When the initial credibility assessment result indicates that the target vehicle is a suspicious vehicle, generate a dynamic self-test instruction package and determine the next target telemetry point of the target vehicle to bind the task with the dynamic self-test instruction package. S5. When the target vehicle arrives at the target telemetry point, a dynamic self-test command package is sent to its on-board terminal to trigger the on-board terminal to execute the diagnostic activation script and collect activation response data. S6. Receive the exhaust gas concentration during the excitation period synchronously collected by the target telemetry point and the excitation response data reported by the vehicle terminal, and perform dynamic response consistency analysis on the exhaust gas concentration during the excitation period and the excitation response data to generate in-depth diagnostic conclusions. S7. Update the data credibility score of the target vehicle based on the in-depth diagnostic results, and generate the final emission regulatory directive based on the data credibility score.

2. The method for monitoring motor vehicle emissions based on the fusion of telemetry and OBD as described in claim 1, characterized in that, The acquisition of vehicle traffic characteristics when the target vehicle passes through the telemetry point includes: Vehicle images are captured by image recognition units deployed at telemetry points, and the identification information of the target vehicle is parsed from them; The instantaneous speed and instantaneous acceleration of the target vehicle are measured by auxiliary sensing units deployed at telemetry points as the vehicle passes by. The vehicle's traffic characteristics are formed by combining identification information, instantaneous speed, and instantaneous acceleration.

3. The method for monitoring motor vehicle emissions based on the fusion of telemetry and OBD as described in claim 1, characterized in that, The specific steps for processing historical data from multiple mirrored vehicles to generate a group baseline interval include: Extract subsets of historical OBD parameters and historical telemetry emission concentration values ​​corresponding to multiple mirror vehicles from the database storing historical regulatory data; A joint statistical analysis was conducted on a subset of historical OBD parameters and historical telemetry emission concentration values ​​to establish a joint distribution model that can characterize the normal operating state of vehicles. Based on the joint distribution model, a population baseline interval is derived, which includes the normal range of OBD parameters and the expected range of emission concentrations that are dynamically correlated with the normal range of OBD parameters.

4. The method for monitoring motor vehicle emissions based on the fusion of telemetry and OBD as described in claim 3, characterized in that, The specific steps for generating the initial credibility assessment result include: Determine whether the real-time OBD parameter set falls within the normal range of the OBD parameters in the group's baseline range; Based on the real-time parameter set of OBD, the corresponding emission concentration expected range is obtained from the group benchmark range, and it is determined whether the measured emission concentration falls within the emission concentration expected range. Based on the combination of the above two judgment results, an initial credibility assessment result containing a clear abnormal pattern identifier is generated.

5. A method for monitoring motor vehicle emissions based on the fusion of telemetry and OBD as described in claim 4, characterized in that, The specific steps for generating the dynamic self-test command package and determining the next target telemetry point of the target vehicle to bind the dynamic self-test command package to the task include: Based on the abnormal patterns indicated in the initial confidence assessment results, a matching script template is selected from the diagnostic script template library to generate a dynamic self-test instruction package; Acquire the real-time location and driving status data of the target vehicle, and combine it with road network information to predict its future driving trajectory; Based on the future driving trajectory, the optimal telemetry point is selected from multiple candidate telemetry points as the target telemetry point, and the dynamic self-test command package is bound to the identifier and trigger time window of the target telemetry point.

6. The method for monitoring motor vehicle emissions based on the fusion of telemetry and OBD as described in claim 1, characterized in that, The specific steps for triggering the vehicle-mounted terminal to execute the diagnostic activation script and collect activation response data include: The on-board terminal parses the dynamic self-test command package and sends parameter disturbance commands to the vehicle engine control unit. While sending parameter perturbation commands, the time-series waveforms of multiple deep response parameters in the vehicle emission control system are recorded at a higher acquisition frequency than the conventional polling rate. The recorded time-series waveforms are encapsulated as excitation response data, in which deep response parameters include oxygen sensor response signals, catalyst temperature change rate, and engine misfire count values.

7. The method for monitoring motor vehicle emissions based on the fusion of telemetry and OBD as described in claim 1, characterized in that, The specific steps for performing dynamic response consistency analysis on the exhaust gas concentration and excitation response data during the excitation period to generate in-depth diagnostic conclusions include: A set of internal response features are extracted from the excitation response data, and a set of external emission response features are extracted from the exhaust gas concentration during the excitation period; Establish a dynamic transfer characteristic model to describe the causal relationship between internal response characteristics and external emission response characteristics; The dynamic transmission characteristic model is compared with the benchmark dynamic transmission characteristic model representing a healthy vehicle to calculate the deviation index, and a deep diagnostic conclusion is generated based on the deviation index.

8. The method for monitoring motor vehicle emissions based on the fusion of telemetry and OBD as described in claim 1, characterized in that, The specific steps for updating the data credibility score of the target vehicle based on the in-depth diagnostic conclusions and generating the final emissions regulatory directive based on the data credibility score include: If the in-depth diagnostic conclusion confirms a real physical fault, the reliability weight of the target vehicle's OBD data will be lowered, and the reliability weight of its telemetry data will be raised accordingly. If the in-depth diagnostic conclusion confirms that the OBD data has been tampered with, the credibility weight of the target vehicle's OBD data will be set to the lowest level. The updated OBD data confidence weights and telemetry data confidence weights will be applied to all future data analyses of the target vehicle.

9. A motor vehicle emission monitoring system based on the fusion of telemetry and OBD, characterized in that, include: The target vehicle data acquisition module acquires the vehicle traffic characteristics when the target vehicle passes through the telemetry point, the measured emission concentration collected by the telemetry point, and the real-time OBD parameter set synchronously reported by its on-board terminal. The group baseline interval generation module selects multiple mirror vehicles from the database storing historical regulatory data based on vehicle traffic characteristics, and processes the historical data of the multiple mirror vehicles to generate the group baseline interval. The credibility assessment result generation module compares and analyzes the real-time OBD parameter set with the measured emission concentration combined with the group benchmark interval to generate the initial credibility assessment result. The dynamic self-test instruction package generation module generates a dynamic self-test instruction package when the initial credibility assessment result indicates that the target vehicle is a suspicious vehicle, and determines the next target telemetry point of the target vehicle to bind the dynamic self-test instruction package to the task. The excitation response data acquisition module sends a dynamic self-test command package to the vehicle terminal when the target vehicle arrives at the target telemetry point, so as to trigger the vehicle terminal to execute the diagnostic excitation script and collect excitation response data. The deep diagnostic conclusion generation module receives the exhaust gas concentration during the excitation period synchronously collected by the target telemetry point and the excitation response data reported by the vehicle terminal, and performs dynamic response consistency analysis on the exhaust gas concentration during the excitation period and the excitation response data to generate a deep diagnostic conclusion. The emission regulatory instruction generation module updates the data credibility score of the target vehicle based on the in-depth diagnostic results, and generates the final emission regulatory instruction based on the data credibility score.

Images