Passenger car carbon emission monitoring and regulation system and method based on multi-source spatio-temporal data

Abstract

The application relates to the technical field of intelligent transportation and carbon governance, in particular to a passenger car carbon emission monitoring and regulation system and method based on multi-source spatio-temporal data. The application solves the problem of scattered and inconsistent data by preprocessing and uniformly mapping multi-source spatio-temporal data. A three-dimensional carbon emission calculation model of vehicle-space unit-time period is constructed to realize accurate calculation of single-trip carbon emission of a single vehicle and positioning of high-carbon emission hotspots. A clustering algorithm is used to complete differentiated division of user groups. A dynamic early warning threshold is generated based on a warning threshold prediction model, regional carbon emission targets and historical data to reduce false positives and false negatives. A warning type-policy tool-adaptation object mapping rule library is established, and a regulation strategy is generated by combining early warning signals and real-time data to realize automatic and accurate generation of regulation strategies from problem identification. The model parameter, early warning threshold and mapping rule are optimized by using policy implementation effect data to form a monitoring, early warning, regulation, optimization closed-loop management and realize accurate regulation.

Description

Technical Field

[0001] This application relates to the fields of intelligent transportation and carbon governance technology, and in particular to a system and method for monitoring and controlling carbon emissions of passenger vehicles based on multi-source spatiotemporal data. Background Technology

[0002] With the continuous growth of urban motor vehicle ownership, carbon dioxide emissions from passenger vehicle operation have become a significant component of urban carbon emissions. Their high proportion and rapid growth not only exacerbate global climate change but also directly contribute to urban air pollution and public health crises. Therefore, monitoring and managing passenger vehicle carbon emissions is extremely important. However, current monitoring technologies suffer from problems such as fragmented and inconsistent data sources; low accuracy in carbon emission accounting, making it difficult to identify high-emission hotspots and differentiated groups; early warning mechanisms relying on static thresholds, resulting in high false alarm and false negative rates; and policy formulation largely based on experience, lacking dynamic linkage with real-time monitoring data, leading to "one-size-fits-all" approaches and delayed adjustments. Current traditional solutions largely rely on annual statistical data and static analysis models, failing to achieve real-time, accurate spatiotemporal perception and dynamic control. The data, analysis, early warning, and policy processes are disconnected, failing to form an effective closed loop and thus failing to achieve precise control. Summary of the Invention

[0003] To address the aforementioned technical problems, the purpose of this application is to provide a passenger vehicle carbon emission monitoring and control system and method based on multi-source spatiotemporal data, aiming to solve at least one of the technical problems mentioned above.

[0004] In a first aspect, embodiments of this application provide a passenger vehicle carbon emission monitoring and control system based on multi-source spatiotemporal data, the system comprising:

[0005] The data acquisition and fusion module is used to acquire multi-source spatiotemporal data, preprocess the multi-source spatiotemporal data, and uniformly map it to a standard spatial network and time slice, outputting fused data with unified spatiotemporal coding;

[0006] The carbon emission calculation and user analysis modules are used to calculate the carbon emission intensity of a single vehicle based on fused data using a pre-built three-dimensional carbon emission calculation model of vehicle-space unit-time period; and to segment user groups using clustering algorithms.

[0007] The early warning module is used to generate dynamic early warning thresholds based on a pre-built early warning threshold prediction model, combined with regional carbon emission targets and historical data. Based on early warning indicators and dynamic early warning thresholds, a two-condition linkage mechanism is used to trigger graded early warnings and generate early warning signals, whereby if the number of abnormal monitoring units within the region is greater than or equal to the first threshold, the range is determined to be within the standard; if the actual value of the early warning indicator deviates from the dynamic early warning threshold by a magnitude greater than or equal to the deviation from the baseline, the intensity is determined to be within the standard.

[0008] The regulation simulation and feedback module is used to dynamically calculate policy parameters based on a pre-built mapping rule base of early warning type, policy tool, and adaptable object, according to early warning signals, regional carbon emission gap rate, and low-carbon penetration gap rate; obtain policy implementation effect data after the policy parameters are implemented, and feed the policy implementation effect data back to the carbon emission calculation and user analysis module to drive iterative optimization of model parameters, dynamic early warning thresholds, and mapping rule base; wherein, the adaptable object is the regional level and user group whose policy effect is clearly defined by rules.

[0009] Furthermore, the calculation of single-vehicle carbon emission intensity based on fused data and a pre-built three-dimensional carbon emission calculation model of vehicle-space unit-time period includes:

[0010] Based on the vehicle identification number of the target vehicle, the vehicle model and current speed of the target vehicle are obtained from the fused data, and the baseline energy consumption of the target vehicle at the current speed is obtained based on the vehicle model and the current speed.

[0011] Calculate the vehicle power-to-weight ratio of the target vehicle;

[0012] The corresponding driving behavior correction factor is obtained by querying the vehicle power ratio.

[0013] If the target vehicle is an electric vehicle, then based on the power grid area and time, the dynamic power grid carbon emission intensity of that time period is queried, and the instantaneous carbon emission intensity of the target vehicle is calculated based on the benchmark energy consumption, the driving behavior correction factor, the dynamic power grid carbon emission intensity and the preset duration.

[0014] If the target vehicle is a gasoline vehicle, then the fixed carbon emission factor corresponding to the model of the target vehicle is obtained, and the instantaneous carbon emission intensity of the target vehicle is calculated based on the baseline energy consumption, the driving behavior correction factor, the fixed carbon emission factor and the preset duration.

[0015] Furthermore, the method of using clustering algorithms to segment user groups includes:

[0016] Acquire user characteristic data; wherein, the user characteristic data includes user generation information, city level of residence, and commuting distance;

[0017] The user feature data were clustered using the K-means++ clustering algorithm.

[0018] Within a preset range of cluster numbers, different K values ​​are traversed, and clustering iterations are performed for each K value until a preset condition is met; wherein, the preset condition is that the offset of the cluster center is less than a preset offset threshold or the maximum number of iterations is reached;

[0019] For each K value, calculate its silhouette coefficient and select the K value that maximizes the silhouette coefficient as the optimal number of clusters.

[0020] Based on the clustering results of the optimal number of clusters, the clustering label for each user and the feature center of each cluster are output;

[0021] Based on the feature centers of each cluster and combined with business semantic rules, each cluster is mapped to its corresponding group name;

[0022] The clustering labels are associated with and stored as group names.

[0023] Furthermore, the dynamic early warning threshold is generated based on a pre-built early warning threshold prediction model, combined with regional carbon emission targets and historical data; the tiered early warning is triggered through a dual-condition linkage mechanism where both the scope and intensity meet the standards, based on early warning indicators and dynamic early warning thresholds, including:

[0024] Construct a predictive model for early warning thresholds;

[0025] Based on historical data, the early warning indicator prediction values ​​for the next M periods are predicted using the aforementioned early warning threshold prediction model; wherein, the early warning indicators include regional carbon emission intensity and low carbon penetration rate.

[0026] The predicted values ​​of the early warning indicators are compared with the decomposed values ​​of the regional carbon emission targets, and combined with the historical fluctuation standard deviation, to generate a dynamic early warning threshold; wherein, the historical data includes regional carbon emission intensity and low carbon penetration rate, or the historical data includes regional carbon emission intensity and low carbon penetration rate, and includes one or more of traffic index, meteorological data and holiday indicators;

[0027] Obtain the actual value of the early warning indicator, and calculate the deviation of the actual value of the early warning indicator from the dynamic early warning threshold based on the actual value of the early warning indicator and the dynamic early warning threshold;

[0028] Determine whether the range and intensity standards are met simultaneously. If they are met, issue a graded warning based on the magnitude of the deviation.

[0029] Furthermore, the warning threshold prediction model is a GRU neural network-based warning threshold prediction model, which is a time-series prediction model. The GRU neural network-based warning threshold prediction model includes a GRU network, an Attention layer, and a fully connected layer, which are connected sequentially.

[0030] Furthermore, the first threshold is determined based on the total number of monitoring units.

[0031] Furthermore, the deviation range is calculated according to the following formula:

[0032] Deviation range = (actual value of warning indicator / dynamic warning threshold - 1) * 100%.

[0033] Secondly, embodiments of this application provide a method for monitoring and controlling carbon emissions of passenger vehicles based on multi-source spatiotemporal data, the method comprising:

[0034] S1. Collect multi-source spatiotemporal data, preprocess the multi-source spatiotemporal data, and uniformly map it to a standard spatial network and time slice, outputting fused data with unified spatiotemporal coding;

[0035] S2. Based on fused data, calculate the carbon emission intensity of a single vehicle using a pre-built three-dimensional carbon emission calculation model of vehicle-spatial unit-time period; and use clustering algorithms to segment user groups;

[0036] S3. Based on the pre-built early warning threshold prediction model, combined with regional carbon emission targets and historical data, a dynamic early warning threshold is generated; based on the early warning indicators and the dynamic early warning threshold, a graded early warning is triggered through a dual-condition linkage mechanism where both the scope and intensity meet the standards, and an early warning signal is generated; if the number of abnormal monitoring units within the region is greater than or equal to the first threshold, the scope is determined to meet the standard; if the actual value of the early warning indicator deviates from the dynamic early warning threshold by a magnitude greater than or equal to the deviation from the baseline, the intensity is determined to meet the standard.

[0037] S4. Based on the pre-built mapping rule base of early warning type-policy tool-adaptive object, dynamically calculate policy parameters according to early warning signal, regional carbon emission gap rate and low carbon penetration gap rate; obtain policy implementation effect data after the policy parameters are implemented, and feed the policy implementation effect data back to step S2 to drive iterative optimization of model parameters, dynamic early warning threshold and mapping rule base; wherein, the adapted object is the regional level and user group where the rules clearly define the policy effect.

[0038] This application addresses the issues of fragmented and inconsistent data sources in the field of low-carbon governance for passenger vehicles by preprocessing multi-source spatiotemporal data and uniformly mapping it to a standard spatial network and time slices, outputting fused data with unified spatiotemporal coding. This provides high-quality data support for end-to-end analysis. By constructing a three-dimensional carbon emission calculation model of vehicle-spatial unit-time period, it achieves accurate calculation of carbon emissions for a single vehicle trip and the location of high-carbon emission hotspots, improving the accuracy of traditional macro-level accounting to the vehicle level. The use of clustering algorithms to segment user groups solves the problem of ambiguous objects, achieving differentiated group segmentation. Based on a pre-built early warning threshold prediction model, combined with regional carbon emission targets and historical data, dynamic early warning thresholds are generated. Compared to static thresholds, the generated dynamic thresholds better reflect actual trend changes, effectively reducing false alarms or missed alarms caused by rigid thresholds. The combination of dual-condition triggering logic for both scope and intensity compliance significantly reduces the probability of false and missed warnings. By pre-constructing a mapping rule base of early warning types, policy tools, and applicable objects, and dynamically calculating policy parameters based on early warning signals, the mapping rule base, and real-time monitoring data, the system achieves automatic and precise generation from problem identification to structured control strategies. This solves problems such as policy formulation relying heavily on experience, lacking dynamic linkage with real-time monitoring data, resulting in "one-size-fits-all" approaches and delayed adjustments. By acquiring policy implementation effect data after the policy parameters are implemented, this data is fed back to carbon emission calculation and user analysis modules to drive iterative optimization of model parameters, dynamic early warning thresholds, and the mapping rule base, forming a closed-loop management system of monitoring, early warning, control, and optimization, thus achieving precise control. Attached Figure Description

[0039] To more clearly illustrate the technical solution of this application, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0040] Figure 1 This is a schematic diagram of the structure of the passenger vehicle carbon emission monitoring and control system based on multi-source spatiotemporal data provided in the embodiments of this application;

[0041] Figure 2 This is a flowchart illustrating the method for monitoring and controlling carbon emissions of passenger vehicles based on multi-source spatiotemporal data provided in this application embodiment. Detailed Implementation

[0042] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0043] Those skilled in the art will understand that, unless explicitly stated otherwise, the singular forms “a,” “an,” “the,” and “the” used herein may also include the plural forms. It should be further understood that the term “comprising” as used in the specification of this application means the presence of features, integers, steps, operations, elements, modules, and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, modules, components, and / or groups thereof. It should be understood that when we say an element is “connected” or “coupled” to another element, it can be directly connected or coupled to the other element, or there may be intermediate elements. Furthermore, “connected” or “coupled” as used herein can include wireless connections or wireless coupling. The term “and / or” as used herein includes all or any modules and all combinations of one or more associated listed items.

[0044] Those skilled in the art will understand that, unless otherwise defined, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains. It should also be understood that terms such as those defined in general dictionaries should be understood to have the same meaning as in the context of the prior art, and should not be interpreted in an idealized or overly formal sense unless specifically defined as herein.

[0045] Please see Figure 1 This application provides a passenger vehicle carbon emission monitoring and control system based on multi-source spatiotemporal data, the system comprising:

[0046] The data acquisition and fusion module 1 is used to acquire multi-source spatiotemporal data, preprocess the multi-source spatiotemporal data, and uniformly map it to a standard spatial network and time slice, outputting fused data with unified spatiotemporal coding.

[0047] The carbon emission calculation and user analysis module 2 is used to calculate the carbon emission intensity of a single vehicle based on fused data using a pre-built three-dimensional carbon emission calculation model of vehicle-space unit-time period; and to segment user groups using clustering algorithms.

[0048] Early warning module 3 is used to generate dynamic early warning thresholds based on a pre-built early warning threshold prediction model, combined with regional carbon emission targets and historical data; based on early warning indicators and dynamic early warning thresholds, it triggers graded early warnings through a dual-condition linkage mechanism where both range and intensity meet the standards, generating early warning signals; wherein, if the number of abnormal monitoring units within the region is greater than or equal to the first threshold, the range is determined to meet the standard; if the actual value of the early warning indicator deviates from the dynamic early warning threshold by a magnitude greater than or equal to the deviation from the baseline, the intensity is determined to meet the standard.

[0049] The regulation simulation and feedback module 4 is used to dynamically calculate policy parameters based on a pre-built mapping rule base of early warning type, policy tool, and adaptable object, according to early warning signals, regional carbon emission gap rate, and low carbon penetration gap rate; obtain policy implementation effect data after the policy parameters are implemented, and feed the policy implementation effect data back to the carbon emission calculation and user analysis module to drive iterative optimization of model parameters, dynamic early warning thresholds, and mapping rule base; wherein, the adaptable object is the regional level and user group whose policy role is clearly defined by rules.

[0050] In this embodiment of the application, the data acquisition and fusion module 1, the carbon emission calculation and user analysis module 2, the early warning module 3, and the control simulation and feedback module 4 are connected sequentially. In the data acquisition and fusion module, the multi-source spatiotemporal data includes real-time data from vehicle terminals, roadside unit data, consumption statistics data, and macro-target data. The real-time data from vehicle terminals includes one or more of instantaneous fuel / electricity consumption, GPS (Global Positioning System) coordinates, VIN (Vehicle Identification Number) code, vehicle speed, and acceleration. The roadside unit data includes one or more of traffic flow, vehicle type identification, and average speed. The consumption statistics data includes user age and region, and the macro-target data includes regional carbon emission targets.

[0051] For vehicle terminal data, it can be collected via the OBD-II interface using the CAN bus protocol at a set frequency. For example, through the OBD-II interface using the CAN bus protocol, data such as vehicle VIN code, instantaneous fuel / electricity consumption, GPS coordinates, and vehicle speed can be collected at a frequency of 30 seconds per acquisition.

[0052] For roadside unit data, traffic flow, vehicle type recognition, and average speed can be obtained in real time through roadside radar and cameras.

[0053] For consumption statistics, vehicle registration information from vehicle management agencies can be periodically extracted via an application programming interface or direct database connection based on the HTTPS (Hypertext Transfer Protocol Secure) encrypted transmission protocol. This registration information includes the user's age and region.

[0054] For macro-level target data, regional carbon emission targets can be collected periodically via application programming interfaces based on the HTTPS encrypted transmission protocol or direct database connection.

[0055] The data mentioned above uses a unified transmission protocol. In one embodiment, the MQTT 3.1.1 protocol is used to set different QoS (Quality of Service) levels for different data sources. For example, the data from the vehicle terminal is set to QoS level 2 to ensure that critical data is not lost.

[0056] The process involves preprocessing the multi-source spatiotemporal data and mapping it uniformly to a standard spatial network and time slices, outputting a fused data stream with unified spatiotemporal coding. Specifically, the multi-source spatiotemporal data is cleaned, coordinate transformed, and time synchronized, uniformly mapped to a standard spatial grid and time slices, and outputting fused data with unified spatiotemporal coding.

[0057] For data cleaning, in one embodiment, distributed stream processing is implemented within the Spark framework. Through spatial filtering, records with GPS positioning drift exceeding 100 meters are discarded. Based on different vehicle models, records with instantaneous fuel / electricity consumption deviating from the historical average by ±30% are removed and configured according to the 3σ principle. Finally, null values ​​or extreme values ​​are marked or removed.

[0058] The coordinate transformation and time synchronization described above are achieved using a spatiotemporal alignment algorithm. For time alignment, UTC timestamps are used as the reference, and all data is divided and aggregated into windows according to preset time slices, such as 1-hour time segments. A sliding window algorithm is used to correct minor time discrepancies between devices. For spatial alignment, a standard spatial grid system (e.g., a 500m × 500m spatial grid) is established. After all spatial data is transformed from WGS84 to the local projected coordinate system, it is matched to a unique grid ID.

[0059] The output is fused data with unified spatiotemporal coding. Specifically, it outputs standardized data records with (timestamp, network ID, and vehicle ID) as the primary key for use by downstream modules.

[0060] This acquisition and fusion module preprocesses multi-source spatiotemporal data and maps it uniformly to a standard spatial network and time slices, outputting fused data with unified spatiotemporal coding. This solves the core problem of fragmented and inconsistent data sources in the field of low-carbon governance of passenger vehicles, providing high-quality data support for full-process analysis.

[0061] In the carbon emission calculation and user analysis modules, the aforementioned vehicle-spatial unit-time period three-dimensional carbon emission calculation model is a micro-level carbon emission calculation model indexed by vehicle ID, spatial ID, and time period ID. This model enables accurate calculation of carbon emissions for a single vehicle trip and the location of high-carbon emission hotspots, elevating the accuracy of traditional macro-level accounting to the vehicle level. By employing clustering algorithms to segment user groups, the problem of ambiguous objects is solved, achieving differentiated group segmentation.

[0062] Compared to static thresholds, the dynamic thresholds generated by this early warning module are more in line with actual trend changes, effectively reducing false alarms or missed alarms caused by rigid thresholds.

[0063] In the regulation simulation and feedback module, the early warning types include transition lag early warning and high-carbon scenario early warning. The policy parameters include subsidy amount, carbon tax rate, and restricted driving periods. The core mapping relationships in the mapping rule base include: transition lag early warning -> triggering special subsidies and simplified pure electric operation subsidies for the elderly; high-carbon scenario early warning -> triggering differentiated carbon taxes and low-carbon vehicle traffic incentives. The regional levels include core areas, suburbs, and remote suburbs, and the user groups include Generation Z, middle-aged groups, and the elderly. This regulation simulation and feedback module establishes a quantitative mapping relationship between early warning signals and regulation policies, achieving automatic connection from "problem identification" to "strategy generation." The policy implementation effect is used as a feedback signal to continuously optimize model parameters, dynamic early warning thresholds, and the mapping rule base (optimizing the strategy rules in the mapping rule base), enabling the system to have autonomous learning and evolution capabilities. The policy parameters are dynamically calculated based on the early warning signal, regional carbon emission gap rate, and low-carbon penetration gap rate. Taking pure electric vehicle subsidies as an example: the subsidy amount is jointly determined by the regional carbon emission gap rate and the low-carbon penetration gap rate. It should be noted that when calculating policy parameters such as subsidy amounts, carbon tax rates, and restricted driving periods based on the aforementioned regional carbon emission gap rate and low-carbon penetration gap rate, it is necessary to ensure that the policy intensity matches the severity of the problem. The regional carbon emission gap rate and low-carbon penetration gap rate are data obtained through real-time calculation. In embodiments where the early warning types include transition lag early warning and high-carbon scenario early warning, the low-carbon penetration rate corresponds to the early warning indicator for transition lag early warning; the regional carbon emission intensity corresponds to the early warning indicator for high-carbon scenario early warning. The regional carbon emission gap rate represents the gap between the overall carbon emissions of the region and the target. The low-carbon penetration gap rate represents the gap between the current low-carbon penetration rate and the target. The aforementioned low-carbon penetration rate specifically refers to the penetration rate of low-carbon vehicle models. This application embodiment constructs a rule base that maps early warning types, policy tools, and applicable objects to each other, and establishes a dynamic policy parameter calculation model with the regional carbon emission gap rate and low-carbon penetration gap rate as the core, realizing the automatic and accurate generation from problem identification to structured control strategies. The process of feeding back policy implementation effect data to the carbon emission calculation and user analysis modules drives iterative optimization of model parameters, dynamic early warning thresholds, and mapping rule bases. Specifically, the policy implementation effect data includes subsidy distribution data, low-carbon vehicle usage data, and regional carbon emission monitoring data after policy implementation. These data are fed back to the carbon emission calculation and user analysis modules through a standardized interface.

[0064] In summary, this application provides a quantifiable, executable, and optimizable systematic technical solution for carbon emission reduction in the transportation sector.

[0065] This application addresses the issues of fragmented and inconsistent data sources in the field of low-carbon governance for passenger vehicles by preprocessing multi-source spatiotemporal data and uniformly mapping it to a standard spatial network and time slices, outputting fused data with unified spatiotemporal coding. This provides high-quality data support for end-to-end analysis. By constructing a three-dimensional carbon emission calculation model of vehicle-spatial unit-time period, it achieves accurate calculation of carbon emissions for a single vehicle trip and the location of high-carbon emission hotspots, improving the accuracy of traditional macro-level accounting to the vehicle level. The use of clustering algorithms to segment user groups solves the problem of ambiguous objects, achieving differentiated group segmentation. Based on a pre-built early warning threshold prediction model, combined with regional carbon emission targets and historical data, dynamic early warning thresholds are generated. Compared to static thresholds, the generated dynamic thresholds better reflect actual trend changes, effectively reducing false alarms or missed alarms caused by rigid thresholds. The combination of dual-condition triggering logic for both scope and intensity compliance significantly reduces the probability of false and missed warnings. By pre-constructing a mapping rule base of early warning types, policy tools, and applicable objects, and dynamically calculating policy parameters based on early warning signals, the mapping rule base, regional carbon emission deficit rates, and low-carbon penetration deficit rates, the system achieves automatic and precise generation of structured control strategies from problem identification. This solves problems such as policy formulation relying heavily on experience, lacking dynamic linkage with real-time monitoring data, resulting in "one-size-fits-all" approaches and delayed adjustments. By acquiring policy implementation effect data after the policy parameters are implemented, this data is fed back to the carbon emission calculation and user analysis modules to drive iterative optimization of model parameters, dynamic early warning thresholds, and the mapping rule base, forming a closed-loop management system of monitoring, early warning, control, and optimization, thus achieving precise control.

[0066] In one embodiment, the multi-source spatiotemporal data includes real-time data from vehicle terminals, roadside unit data, consumption statistics data, and macroscopic target data.

[0067] This application embodiment uses multi-source spatiotemporal data composed of real-time data from vehicle terminals, roadside unit data, consumption statistics data, and macro-target data. This enables comprehensive acquisition of passenger vehicle carbon emission information from multiple dimensions, providing multi-dimensional data support for subsequent three-dimensional carbon emission calculation, user group segmentation, dynamic early warning threshold generation, and policy regulation simulation. This ensures accurate carbon emission monitoring results, objective user analysis, timely early warning response, and scientific and effective regulation strategies.

[0068] In one embodiment, the calculation of single-vehicle carbon emission intensity based on fused data and using a pre-built three-dimensional carbon emission calculation model of vehicle-space unit-time period includes:

[0069] Based on the vehicle identification number of the target vehicle, the vehicle model and current speed of the target vehicle are obtained from the fused data, and the baseline energy consumption of the target vehicle at the current speed is obtained based on the vehicle model and the current speed.

[0070] Calculate the vehicle power-to-weight ratio of the target vehicle;

[0071] The corresponding driving behavior correction factor is obtained by querying the vehicle power ratio.

[0072] If the target vehicle is an electric vehicle, then based on the power grid area and time, the dynamic power grid carbon emission intensity of that time period is queried, and the instantaneous carbon emission intensity of the target vehicle is calculated based on the benchmark energy consumption, the driving behavior correction factor, the dynamic power grid carbon emission intensity and the preset duration.

[0073] If the target vehicle is a gasoline vehicle, then the fixed carbon emission factor corresponding to the model of the target vehicle is obtained, and the instantaneous carbon emission intensity of the target vehicle is calculated based on the baseline energy consumption, the driving behavior correction factor, the fixed carbon emission factor and the preset duration.

[0074] In this embodiment, a relationship table between vehicle model, vehicle speed, and baseline energy consumption is pre-constructed. Based on the target vehicle's model and current speed, the baseline energy consumption of the target vehicle at the current speed is obtained by querying the relationship table. Specifically, the target vehicle's model and current speed are obtained from the fused data based on its vehicle identification number (VIN). Then, the relationship table is queried to obtain the baseline energy consumption of that vehicle model at the current speed.

[0075] The specific power of a vehicle is calculated using the following formula:

[0076] ;

[0077] Where 1.1 is the rotational mass coefficient. 0.000302 is the rolling resistance term, and 0.000302 is the air resistance term; Current vehicle speed It is acceleration.

[0078] The step of obtaining the corresponding driving behavior correction factor based on the vehicle power ratio includes: pre-constructing a correlation table between the vehicle power ratio and the driving behavior correction factor, and then querying the corresponding driving behavior correction factor based on the vehicle power ratio through the correlation table between the vehicle power ratio and the driving behavior correction factor.

[0079] The instantaneous carbon emission intensity of the target vehicle is calculated based on the baseline energy consumption, the driving behavior correction factor, the dynamic power grid carbon emission intensity, and the preset duration. Specifically, it is calculated according to the following formula:

[0080] Instantaneous carbon emission intensity = vehicle baseline energy consumption × driving behavior correction factor × dynamic grid carbon emission intensity × duration.

[0081] The instantaneous carbon emission intensity of the target vehicle is calculated based on the baseline energy consumption, the driving behavior correction factor, the fixed carbon emission factor, and the preset duration. Specifically, it is calculated using the following formula:

[0082] Instantaneous carbon emission intensity = vehicle baseline energy consumption × driving behavior correction factor × fixed carbon emission factor × duration.

[0083] It should be noted that the vehicle identification code is the vehicle ID. The above-mentioned vehicle-spatial unit-time period three-dimensional carbon emission calculation model is a micro-carbon emission calculation model indexed by vehicle ID, spatial ID, and time period ID. It realizes the accurate calculation of carbon emissions of a single vehicle per trip and the location of high carbon emission hotspots, improving the accuracy of traditional macro-calculation to the vehicle level.

[0084] In one embodiment, the step of using a clustering algorithm to segment user groups includes:

[0085] Acquire user characteristic data; wherein, the user characteristic data includes user generation information, city level of residence, and commuting distance;

[0086] The user feature data were clustered using the K-means++ clustering algorithm.

[0087] Within a preset range of cluster numbers, different K values ​​are traversed, and clustering iterations are performed for each K value until a preset condition is met; wherein, the preset condition is that the offset of the cluster center is less than a preset offset threshold or the maximum number of iterations is reached;

[0088] For each K value, calculate its silhouette coefficient and select the K value that maximizes the silhouette coefficient as the optimal number of clusters.

[0089] Based on the clustering results of the optimal number of clusters, the clustering label for each user and the feature center of each cluster are output;

[0090] Based on the feature centers of each cluster and combined with business semantic rules, each cluster is mapped to its corresponding group name;

[0091] The clustering labels are associated with and stored as group names.

[0092] In this embodiment, the user generational information refers to age range information. Based on the feature centers of each cluster and combined with business semantic rules, each cluster is mapped to a corresponding group name. Specifically, based on the feature centers of each cluster, each cluster is mapped to a corresponding interpretable group name, including Generation Z, middle-aged group, and elderly group. In one embodiment, K=3, and the offset threshold is 0.01.

[0093] The embodiments of this application implement differentiated group segmentation.

[0094] In one embodiment, the dynamic early warning threshold is generated based on a pre-built early warning threshold prediction model, combined with regional carbon emission targets and historical data; the tiered early warning is triggered based on early warning indicators and dynamic early warning thresholds through a dual-condition linkage mechanism where both scope and intensity meet the standards, including:

[0095] Construct a predictive model for early warning thresholds;

[0096] Based on historical data, the early warning indicator prediction values ​​for the next M periods are predicted using the aforementioned early warning threshold prediction model; wherein, the early warning indicators include regional carbon emission intensity and low carbon penetration rate.

[0097] The predicted values ​​of the early warning indicators are compared with the decomposed values ​​of the regional carbon emission targets, and combined with the historical fluctuation standard deviation, to generate a dynamic early warning threshold; wherein, the historical data includes regional carbon emission intensity and low carbon penetration rate, or the historical data includes regional carbon emission intensity and low carbon penetration rate, and includes one or more of traffic index, meteorological data and holiday indicators;

[0098] Obtain the actual value of the early warning indicator, and calculate the deviation of the actual value of the early warning indicator from the dynamic early warning threshold based on the actual value of the early warning indicator and the dynamic early warning threshold;

[0099] Determine whether the range and intensity standards are met simultaneously. If they are met, issue a graded warning based on the magnitude of the deviation.

[0100] In this embodiment, the early warning indicators include regional carbon emission intensity and low-carbon penetration rate. The historical data includes regional carbon emission intensity and low-carbon penetration rate, or the historical data includes regional carbon emission intensity and low-carbon penetration rate, and includes one or more of traffic index, meteorological data, and holiday indicators. In one embodiment, the input features of the early warning threshold prediction model are: regional carbon emission intensity, low-carbon penetration rate, traffic index, meteorological data, and holiday indicators over a past period (e.g., 12 months). The early warning threshold prediction model makes predictions based on the input features and outputs predicted early warning indicator values ​​for the next M periods. Finally, the predicted early warning indicator values ​​are compared with the carbon emission target decomposition values, and combined with the historical fluctuation standard deviation, to generate a dynamic early warning threshold. Specifically, the dynamic early warning threshold is generated according to the following formula:

[0101] ;

[0102] Wherein, C0 represents the carbon emission target decomposition value. C is the predicted value of the early warning indicator, λ is the adjustment coefficient, 0 < λ ≤ 1. warn σ represents the dynamic early warning threshold, and σ is the standard deviation of historical fluctuations. Compared to static thresholds, the dynamic thresholds generated in this application embodiment better reflect actual trend changes, effectively reducing false alarms or missed alarms caused by threshold rigidity.

[0103] In one embodiment, the warning threshold prediction model is a GRU neural network-based warning threshold prediction model, which is a time-series prediction model. The GRU neural network-based warning threshold prediction model includes a GRU network, an Attention layer, and a fully connected layer, which are connected sequentially.

[0104] In this embodiment, a GRU network is used as the time-series prediction model, which has higher training efficiency than LSTM. An Attention layer is then applied to focus on key time periods, and finally, a fully connected layer outputs the predicted values ​​of the warning indicators for the next M periods. It should be noted that the main advantages of GRU are its efficiency and the number of parameters. Since carbon emission time-series data typically have clear periodicity and trends, and the data is monthly / quarterly with moderate sequence lengths, the simplified structure of GRU is sufficient to capture key patterns. Furthermore, the system has high real-time requirements; therefore, efficiency is a significant consideration in model selection. This application selects the GRU network as the time-series prediction model, which maintains comparable prediction accuracy to LSTM while having higher training efficiency and lower computational resource consumption, making it more suitable for the real-time and scalability requirements of this system.

[0105] In addition, this application uses a gated recurrent unit (GRU) network as a time-series prediction model, combines regional carbon emission targets to generate dynamic early warning thresholds, and reduces the probability of false and missed early warnings through dual-condition triggering logic of "range compliance" and "intensity compliance".

[0106] In one embodiment, the first threshold is determined based on the total number of monitoring units.

[0107] In this embodiment, the monitoring unit refers to a lower-level area responsible for early warning indicators (such as carbon emission intensity and low carbon penetration rate). For example, if the system is applied to city-level carbon management, then the "monitoring unit" can be the various administrative districts under the city's jurisdiction. The first threshold is determined based on the total number of monitoring units. Specifically, if the total number of monitoring units is less than or equal to a preset second threshold, the first threshold is equal to a preset first value; if the total number of monitoring units is greater than the preset second threshold, the first threshold is equal to a second value, where the second value = total number of monitoring units × P%; and the second value is greater than the first value. In one example, when the total number of monitoring units is less than or equal to 50, the first threshold is equal to 2; when the total number of monitoring units is greater than 50, the first threshold = total number of monitoring units × 5%.

[0108] In one embodiment, the deviation magnitude is calculated according to the following formula:

[0109] Deviation range = (actual value of warning indicator / dynamic warning threshold - 1) * 100%.

[0110] In one example, the tiered warning based on the magnitude of the deviation specifically includes:

[0111] If the deviation is less than or equal to 10% and less than 20%, a Level 3 warning will be issued.

[0112] If the deviation is 20% ≤ and less than 30%, a second-level warning will be issued.

[0113] If the deviation is ≥30%, a Level 1 warning will be issued.

[0114] In this embodiment, the smaller the level number, the higher the level. A third-level warning can be represented by a yellow warning, a second-level warning by an orange warning, and a first-level warning by a red warning.

[0115] In one embodiment, policy implementation effect data is fed back to the carbon emission calculation and user analysis modules to drive iterative optimization of model parameters, dynamic early warning thresholds, and mapping rule bases. Specifically, the optimization methods include: Regular optimization: On a monthly basis, the gradient descent algorithm is used to update the spatial unit-time period three-dimensional carbon emission calculation model parameters and early warning thresholds. Emergency optimization: When the monthly change in carbon emission indicators after policy implementation is <-1% or the growth rate of low-carbon penetration rate continues to fail to meet the target, optimization is immediately initiated to shorten the adjustment cycle. Rule iteration: The early warning-policy matching rules are continuously optimized based on feedback data to achieve continuous optimization of the mapping rule base of early warning type-policy tool-adapted object, realizing the system's adaptive evolution.

[0116] like Figure 2 As shown in the embodiments of this application, a method for monitoring and controlling carbon emissions of passenger vehicles based on multi-source spatiotemporal data is also provided. The method includes:

[0117] S1. Collect multi-source spatiotemporal data, preprocess the multi-source spatiotemporal data, and uniformly map it to a standard spatial network and time slice, outputting fused data with unified spatiotemporal coding;

[0118] S2. Based on fused data, calculate the carbon emission intensity of a single vehicle using a pre-built three-dimensional carbon emission calculation model of vehicle-spatial unit-time period; and use clustering algorithms to segment user groups;

[0119] S3. Based on the pre-built early warning threshold prediction model, combined with regional carbon emission targets and historical data, a dynamic early warning threshold is generated; based on the early warning indicators and the dynamic early warning threshold, a graded early warning is triggered through a dual-condition linkage mechanism where both the scope and intensity meet the standards, and an early warning signal is generated; if the number of abnormal monitoring units within the region is greater than or equal to the first threshold, the scope is determined to meet the standard; if the actual value of the early warning indicator deviates from the dynamic early warning threshold by a magnitude greater than or equal to the deviation from the baseline, the intensity is determined to meet the standard.

[0120] S4. Based on the pre-built mapping rule base of early warning type-policy tool-adaptive object, dynamically calculate policy parameters according to early warning signal, regional carbon emission gap rate and low carbon penetration gap rate; obtain policy implementation effect data after the policy parameters are implemented, and feed the policy implementation effect data back to step S2 to drive iterative optimization of model parameters, dynamic early warning threshold and mapping rule base; wherein, the adapted object is the regional level and user group where the rules clearly define the policy effect.

[0121] It should be noted that the technical solutions in the embodiments of this specification, if involving the processing of personal information, will all be processed under the premise of having a legal basis (such as obtaining the consent of the personal information subject), and will only be processed within the scope stipulated or agreed. A user's refusal to process personal information beyond what is necessary for basic functions will not affect the user's use of basic functions. The collection, storage, use, processing, transmission, provision, and presentation of related information all comply with relevant laws and regulations, do not infringe upon the privacy of others, and do not violate public order and good morals.

[0122] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in this application and in the embodiments can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual-speed SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0123] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, apparatus, article, or method that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, apparatus, article, or method. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, apparatus, article, or method that includes that element.

[0124] The above description is only a preferred embodiment of this application and does not limit the scope of this application. Any equivalent structural or procedural transformations made based on the content of this application specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the protection scope of this application.

Claims

1. A passenger vehicle carbon emission monitoring and control system based on multi-source spatiotemporal data, characterized in that, The system includes: The data acquisition and fusion module is used to acquire multi-source spatiotemporal data, preprocess the multi-source spatiotemporal data, and uniformly map it to a standard spatial network and time slice, outputting fused data with unified spatiotemporal coding; wherein, the multi-source spatiotemporal data includes real-time data from vehicle terminals, roadside unit data, consumption statistics data, and macroscopic target data; The carbon emission calculation and user analysis modules are used to calculate the carbon emission intensity of a single vehicle based on fused data using a pre-built three-dimensional carbon emission calculation model of vehicle-space unit-time period; and to segment user groups using clustering algorithms. The early warning module is used to generate dynamic early warning thresholds based on a pre-built early warning threshold prediction model, combined with regional carbon emission targets and historical data. Based on early warning indicators and dynamic early warning thresholds, a two-condition linkage mechanism is used to trigger graded early warnings and generate early warning signals, whereby if the number of abnormal monitoring units within the region is greater than or equal to the first threshold, the range is determined to be within the standard; if the actual value of the early warning indicator deviates from the dynamic early warning threshold by a magnitude greater than or equal to the deviation from the baseline, the intensity is determined to be within the standard. The regulation simulation and feedback module is used to dynamically calculate policy parameters based on a pre-built mapping rule base of early warning type, policy tool, and applicable object, according to early warning signals, regional carbon emission gap rate, and low-carbon penetration gap rate; it acquires policy implementation effect data after the policy parameters are implemented, and feeds the policy implementation effect data back to the carbon emission calculation and user analysis module to drive iterative optimization of model parameters, dynamic early warning thresholds, and mapping rule base; wherein, the applicable object is the regional level and user group whose policy effect is clearly defined by the rules; The calculation of single-vehicle carbon emission intensity based on fused data and using a pre-constructed three-dimensional carbon emission calculation model of vehicle-space unit-time period includes: Based on the vehicle identification number of the target vehicle, the vehicle model and current speed of the target vehicle are obtained from the fused data, and the baseline energy consumption of the target vehicle at the current speed is obtained based on the vehicle model and the current speed. Calculate the vehicle power-to-weight ratio of the target vehicle; The corresponding driving behavior correction factor is obtained by querying the vehicle power ratio. If the target vehicle is an electric vehicle, then based on the power grid area and time, the dynamic power grid carbon emission intensity of that time period is queried, and the instantaneous carbon emission intensity of the target vehicle is calculated based on the benchmark energy consumption, the driving behavior correction factor, the dynamic power grid carbon emission intensity and the preset duration. If the target vehicle is a gasoline vehicle, then the fixed carbon emission factor corresponding to the model of the target vehicle is obtained, and the instantaneous carbon emission intensity of the target vehicle is calculated based on the benchmark energy consumption, the driving behavior correction factor, the fixed carbon emission factor and the preset duration. The pre-built early warning threshold prediction model, combined with regional carbon emission targets and historical data, generates dynamic early warning thresholds. Based on early warning indicators and dynamic early warning thresholds, a tiered early warning system is triggered through a dual-condition linkage mechanism where both scope and intensity meet the standards, including: Construct a predictive model for early warning thresholds; Based on historical data, the early warning indicator prediction values ​​for the next M periods are predicted using the aforementioned early warning threshold prediction model; wherein, the early warning indicators include regional carbon emission intensity and low carbon penetration rate. The predicted values ​​of the early warning indicators are compared with the decomposed values ​​of the regional carbon emission targets, and combined with the historical fluctuation standard deviation, to generate a dynamic early warning threshold; wherein, the historical data includes regional carbon emission intensity and low carbon penetration rate, or the historical data includes regional carbon emission intensity and low carbon penetration rate, and includes one or more of traffic index, meteorological data and holiday indicators; Obtain the actual value of the early warning indicator, and calculate the deviation of the actual value of the early warning indicator from the dynamic early warning threshold based on the actual value of the early warning indicator and the dynamic early warning threshold; Determine whether both the range and intensity standards are met simultaneously. If they are met, issue a graded warning based on the magnitude of the deviation. The dynamic early warning threshold is generated according to the following formula: ； Wherein, C0 represents the carbon emission target decomposition value. C is the predicted value of the early warning indicator, λ is the adjustment coefficient, 0 < λ ≤ 1. warn σ represents the dynamic early warning threshold, and σ represents the historical standard deviation of fluctuations.

2. The passenger vehicle carbon emission monitoring and control system based on multi-source spatiotemporal data according to claim 1, characterized in that, The method of using clustering algorithms to segment user groups includes: Acquire user characteristic data; wherein, the user characteristic data includes user generation information, city level of residence, and commuting distance; The user feature data were clustered using the K-means++ clustering algorithm. Within a preset range of cluster numbers, different K values ​​are traversed, and clustering iterations are performed for each K value until a preset condition is met; wherein, the preset condition is that the offset of the cluster center is less than a preset offset threshold or the maximum number of iterations is reached; K represents the number of clusters; For each K value, calculate its silhouette coefficient and select the K value that maximizes the silhouette coefficient as the optimal number of clusters. Based on the clustering results of the optimal number of clusters, the clustering label for each user and the feature center of each cluster are output; Based on the feature centers of each cluster and combined with business semantic rules, each cluster is mapped to its corresponding group name; The clustering labels are associated with and stored as group names.

3. The passenger vehicle carbon emission monitoring and control system based on multi-source spatiotemporal data according to claim 1, characterized in that, The warning threshold prediction model is a GRU neural network-based warning threshold prediction model, which is a time-series prediction model. The GRU neural network-based warning threshold prediction model includes a GRU network, an Attention layer, and a fully connected layer, which are connected sequentially.

4. The passenger vehicle carbon emission monitoring and control system based on multi-source spatiotemporal data according to claim 1, characterized in that, The first threshold is determined based on the total number of monitoring units.

5. The passenger vehicle carbon emission monitoring and control system based on multi-source spatiotemporal data according to claim 1, characterized in that, The deviation range is calculated according to the following formula: Deviation range = (actual value of warning indicator / dynamic warning threshold - 1) * 100%.

6. A method for monitoring and controlling carbon emissions of passenger vehicles based on multi-source spatiotemporal data, characterized in that, The method is performed by the passenger vehicle carbon emission monitoring and control system based on multi-source spatiotemporal data as described in any one of claims 1-5, and the method includes: S1. Collect multi-source spatiotemporal data, preprocess the multi-source spatiotemporal data, and uniformly map it to a standard spatial network and time slice, outputting fused data with unified spatiotemporal coding; S2. Based on fused data, calculate the carbon emission intensity of a single vehicle using a pre-built three-dimensional carbon emission calculation model of vehicle-spatial unit-time period; and use clustering algorithms to segment user groups; S3. Based on the pre-built early warning threshold prediction model, combined with regional carbon emission targets and historical data, a dynamic early warning threshold is generated; based on the early warning indicators and the dynamic early warning threshold, a graded early warning is triggered through a dual-condition linkage mechanism where both the scope and intensity meet the standards, and an early warning signal is generated; if the number of abnormal monitoring units within the region is greater than or equal to the first threshold, the scope is determined to meet the standard; if the actual value of the early warning indicator deviates from the dynamic early warning threshold by a magnitude greater than or equal to the deviation from the baseline, the intensity is determined to meet the standard. S4. Based on the pre-built mapping rule base of early warning type-policy tool-adaptive object, dynamically calculate policy parameters according to early warning signal, regional carbon emission gap rate and low carbon penetration gap rate; obtain policy implementation effect data after the policy parameters are implemented, and feed the policy implementation effect data back to step S2 to drive iterative optimization of model parameters, dynamic early warning threshold and mapping rule base; wherein, the adapted object is the regional level and user group where the rules clearly define the policy effect.

Images