Vehicle carbon emission safety calculation method and system based on Internet of Vehicles

Abstract

The invention discloses a vehicle carbon emission safety calculation method and system based on the Internet of Vehicles, and relates to the technical field of carbon emission monitoring. According to the invention, through the vehicle-mounted carbon metering terminal deployed on each vehicle, the carbon emission is calculated based on the embedded local carbon metering model, so that the real-time performance and high precision of carbon emission monitoring are ensured; through a traditional encryption algorithm and a homomorphic encryption technology, a strong security guarantee is provided in a data transmission process, and the privacy and security of carbon data are ensured. The carbon metering operation platform performs calculation and optimization of a global carbon emission model on the encrypted data uploaded by the plurality of vehicle-mounted terminals based on a federated learning technology, so that the efficiency and precision of large-scale data processing are effectively improved. Besides, the system supports various communication networks, can perform data interaction with an external system, provides support for generation of a carbon emission report and carbon capitalization application, and meets application requirements in different scenes.

Description

Technical Field

[0001] The present invention relates to the technical field of carbon emission monitoring, and in particular to a method and system for safely calculating vehicle carbon emissions based on an Internet of Vehicles. Background Art

[0002] At present, the vehicle carbon emission monitoring systems on the market mainly rely on the extraction and analysis of post-data, and have problems such as poor real-time performance and low accuracy. Especially for new energy vehicles, the calculation of carbon emission reduction is not clear due to unclear baselines and accounting methods, resulting in poor operability of carbon emission reduction assetization. Most existing monitoring systems rely on vehicle charging data for statistics, with large granularity, unable to achieve accurate segmented metering and real-time monitoring, and difficult to adapt to the dynamic carbon emission monitoring needs under different working conditions.

[0003] With the rapid development of Internet of Vehicles (IoV) technology, it has become possible to collect and analyze vehicle operation data in real time using on-board equipment. IoV technology obtains the vehicle's instantaneous operation data in real time through the vehicle's CAN bus, providing a basis for the accurate calculation of the vehicle's carbon emissions. However, the carbon emission calculation models of traditional systems mostly rely on centralized processing in the cloud. This centralized computing model is prone to system performance bottlenecks when faced with a large amount of vehicle data, and it is difficult to achieve high concurrency and real-time processing in multiple scenarios. Secondly, there are also great risks in data security, especially the vehicle's energy consumption data is susceptible to theft or tampering during transmission, which seriously affects the credibility and reliability of the data.

[0004] In addition, existing carbon measurement methods are difficult to support a variety of different types of vehicles, including traditional fuel vehicles, plug-in hybrid vehicles, pure electric vehicles, etc. In order to meet the carbon emission monitoring needs of different types of vehicles, a system that can dynamically adjust the calculation model is urgently needed to ensure accurate and safe calculation of vehicle operation data under various working conditions. Summary of the invention

[0005] In view of the problems existing in the current vehicle carbon emissions calculation, such as unclear accounting baseline, unclear accounting method, poor operability, and insufficient security in the data collection and transmission process, the present invention proposes a vehicle carbon emissions safety calculation method and system based on the Internet of Vehicles. Starting from the actual energy consumption of the vehicle, the mileage traveled per unit time is calculated based on the energy consumption, accurate to the individual vehicle, and the carbon emission reduction is accurately calculated by benchmarking the standard baseline, thereby promoting the assetization of carbon emission reduction in new energy vehicle travel.

[0006] In order to achieve the above object, the present invention is implemented by the following technical solutions:

[0007] A method for safely calculating vehicle carbon emissions based on an Internet of Vehicles, the method comprising:

[0008] Deploy an on-board carbon metering terminal on each vehicle and connect it to the carbon metering operation platform in the cloud through the IoV network;

[0009] The vehicle-mounted carbon metering terminal is embedded with a local carbon metering model, which obtains the real-time operation data of the vehicle from the vehicle CAN bus, and calculates the real-time operation data through the local carbon metering model to obtain the carbon data of the vehicle;

[0010] The real-time operation data is encrypted using a traditional encryption algorithm, the carbon data is encrypted using a homomorphic encryption technology, and the encrypted real-time operation data and carbon data are transmitted to a carbon metering operation platform;

[0011] The carbon metering operation platform obtains the vehicle's CAN message information from the vehicle manufacturer, combines the encrypted real-time operation data and carbon data uploaded by multiple on-board carbon metering terminals, calculates and optimizes the global carbon emission model based on federated learning technology, and generates global calculation results of carbon emissions.

[0012] As a preferred solution of the present invention, the real-time operation data includes instantaneous voltage, instantaneous current, mileage, speed, acceleration, vehicle VIN code and other related data; the vehicle carbon data includes carbon emissions, carbon emission reductions and vehicle energy consumption of each vehicle.

[0013] As a preferred solution of the present invention, the real-time operation data is calculated by the local carbon metering model, further comprising: reviewing the operation data and carbon data calculation results of the vehicle in the historical time through the local experience playback mechanism, and comparing them with the current operation state of the vehicle, so as to adjust the weights and parameters of the current local carbon metering model, the formula is:

[0014]

[0015] Where, L local-clone is the local replay loss function; q is the qth related event in the historical time; h r is the historical hidden state; μ local (qh r ) means that in the local experience playback process, given the historical hidden state h r The probability of event q occurring under Represents the current local carbon measurement model parameter α local Next, given the historical hidden state h r The probability distribution of event q occurring under the condition .

[0016] As a preferred solution of the present invention, the traditional encryption algorithms include symmetric encryption algorithm AES, triple data encryption algorithm 3DES, asymmetric encryption algorithm RSA, stream cipher algorithm ChaCha20 and elliptic curve encryption algorithm ECC.

[0017] As a preferred solution of the present invention, the carbon data is encrypted using homomorphic encryption technology, specifically including:

[0018] The weight matrix A of the local carbon accounting model j Encoding processing, the formula is:

[0019] Encode(A j )=Flatten(Pad(A j ,0,n));

[0020] In the formula, Encode(A j ) is the encoded weight matrix; Pad represents the weight matrix A j Fill, 0 represents the filling value, n is the size of the matrix after encoding; Flatten represents the flattening operation, which is used to convert a multi-dimensional matrix into a one-dimensional vector;

[0021] Using the public key p k The encoded weight matrix Encode(A j ) is encrypted, the formula is:

[0022]

[0023] In the formula, Represents the encrypted weight matrix, and Enc is the encryption function.

[0024] As a preferred solution of the present invention, the carbon metering operation platform calculates and optimizes the global carbon emission model based on the federated learning technology, and also includes: performing global experience playback based on the CAN message information of different vehicles, the formula is:

[0025]

[0026] Where, L global-clone is the global replay loss function; q is the qth related event in the historical time; h r is the historical hidden state; μ global (qh r ) means that in the global playback process, given the historical hidden state h r The probability of event q occurring under

[0027] Indicates the current global carbon emission model parameter α globalNext, given the historical hidden state h r The probability distribution of event q occurring under the condition .

[0028] As a preferred solution of the present invention, the carbon metering operation platform uses a weighted aggregation algorithm to optimize and update the global carbon emission model. The update formula is:

[0029]

[0030] In the formula, represents the updated global carbon emission model parameters; is the local carbon accounting model parameter of the nth vehicle, M is the number of vehicles participating in the joint learning; is the global carbon emission model parameter used in the current iteration of the server; τ represents the current iteration round of joint learning; Add(·) represents the addition operation of the matrix; Mul(·) represents the multiplication operation of the matrix.

[0031] As a preferred solution of the present invention, the method also includes: the carbon metering operation platform generates a carbon emission report based on the global calculation results of carbon emissions, and interacts with the external system for data. At the same time, the global carbon emission model optimized by federated learning is fed back to each vehicle-mounted carbon metering terminal to improve the local carbon metering model and realize dynamic adjustment and optimization.

[0032] A vehicle carbon emissions security calculation system based on the Internet of Vehicles, the system comprising an on-board carbon metering terminal deployed on each vehicle, and a carbon metering operation platform;

[0033] The vehicle-mounted carbon metering terminal comprises:

[0034] CAN communication module, used for data transmission between vehicle CAN bus and MCU module;

[0035] The MCU module, as the core processing unit of the vehicle-mounted carbon metering terminal, is embedded with a local carbon metering model, obtains the real-time operation data of the vehicle from the vehicle CAN bus through the CAN communication module, and calculates the real-time operation data through the local carbon metering model to obtain the carbon data of the vehicle;

[0036] A storage module for temporarily storing the vehicle's real-time operating data and carbon data when the network is unstable or disconnected;

[0037] A data encryption module, used to encrypt the real-time operation data using a traditional encryption algorithm and to encrypt the carbon data using a homomorphic encryption technique;

[0038] External communication module, used to transmit encrypted real-time operation data and carbon data to the carbon measurement operation platform via 4G CAT1, 5G or 6G;

[0039] The carbon metering operation platform obtains the vehicle's CAN message information from the vehicle manufacturer, combines the encrypted real-time operation data and carbon data uploaded by multiple on-board carbon metering terminals, calculates and optimizes the global carbon emission model based on federated learning technology, generates global calculation results of carbon emissions, and generates a carbon emission report based on the global calculation results, and interacts with external systems for data.

[0040] As a preferred solution of the present invention, the MCU module calculates the real-time operation data through the local carbon metering model, and also includes: reviewing the operation data and carbon data calculation results of the vehicle in the historical time through the local experience playback mechanism, and comparing them with the current operation state of the vehicle, so as to adjust the weights and parameters of the current local carbon metering model, and the formula is:

[0041]

[0042] Where, L local-clone is the local replay loss function; q is the qth related event in the historical time; h r is the historical hidden state; μ local (qh r ) means that in the local experience playback process, given the historical hidden state h r The probability of event q occurring under Represents the current local carbon measurement model parameter α local Next, given the historical hidden state h r The probability distribution of event q under the condition of

[0043] The carbon metering operation platform calculates and optimizes the global carbon emission model based on the federated learning technology, and also includes: performing global experience playback based on the CAN message information of different vehicles, the formula is:

[0044]

[0045] Where, L global-clone is the global replay loss function; q is the qth related event in the historical time; h r is the historical hidden state; μ global (qh r ) means that in the global playback process, given the historical hidden state h r The probability of event q occurring under

[0046] Indicates the current global carbon emission model parameter α global Next, given the historical hidden state h r The probability distribution of event q occurring under the condition .

[0047] The beneficial effects of the present invention are as follows: by deploying the on-board carbon metering terminal on each vehicle, it is possible to obtain operating data from the vehicle CAN bus in real time, and accurately calculate the carbon emissions through the embedded local carbon metering model, thereby ensuring the real-time and high-precision carbon emission monitoring. At the same time, the system has the function of storing data when the network is unstable or disconnected, thereby ensuring the integrity of the data. Through traditional encryption algorithms and homomorphic encryption technology, the system provides strong security protection during data transmission to ensure the privacy and security of carbon data. Based on federated learning technology, the carbon metering operation platform calculates and optimizes the global carbon emission model for encrypted data uploaded by multiple on-board terminals, effectively improving the efficiency and accuracy of large-scale data processing. In addition, the system supports a variety of communication networks and can interact with external systems for data, providing support for the generation of carbon emission reports and carbon asset applications, meeting application requirements in different scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative labor.

[0049] in:

[0050] Figure 1 is a flow chart of the method of the present invention;

[0051] Figure 2 It is a system modular structure diagram of the present invention. DETAILED DESCRIPTION

[0052] In order to make the purpose, technical solution and advantages of the embodiment of the present invention clearer, the technical solution of the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings of the embodiment of the present invention. Obviously, the described embodiment is a part of the embodiment of the present invention, not all of the embodiments. Based on the described embodiment of the present invention, all other embodiments obtained by ordinary technicians in this field belong to the scope of protection of the present invention.

[0053] like Figure 1 As shown, an embodiment of the present invention provides a safe calculation method for vehicle carbon emissions based on the Internet of Vehicles. According to the reference vehicle line selection method based on the vehicle curb mass (i.e. the weight of the vehicle when unloaded) in the group standard "Technical Specifications for Assessment of Greenhouse Gas Emission Reductions for Alternative Travel by New Energy Vehicles TCAS 536-2021", a standardized carbon emission reduction assessment calculation is performed to ensure the consistency and accuracy of the calculation under different vehicle types.

[0054] The method comprises the following steps:

[0055] S1: Deploy an on-board carbon metering terminal on each vehicle to form a distributed node, which is connected to the carbon metering operation platform in the cloud through the IoV (Internet of Vehicles) network to achieve independent and distributed collection of carbon data;

[0056] Different from the traditional method of periodic statistics of carbon emission data and the method of collecting data through discrete and shared links, the present invention adopts an independent and secure data collection link to establish the real-time collection capability of carbon data;

[0057] S2: The on-board carbon metering terminal is embedded with a local carbon metering model, which obtains the real-time operation data of the vehicle from the vehicle CAN bus, calculates the real-time operation data through the local carbon metering model, and obtains the carbon data of the vehicle;

[0058] Specifically, real-time operating data may include instantaneous voltage, instantaneous current, mileage (single trip and total mileage), speed, acceleration, vehicle VIN code (vehicle identification code) and other related data (such as battery status, vehicle load), etc.; after collection, preliminary preprocessing is performed, including outlier processing, noise filtering and data standardization to ensure the accuracy of subsequent calculations.

[0059] The carbon data of a vehicle includes carbon emissions, carbon emission reductions, vehicle energy consumption, etc.

[0060] Carbon emissions calculations can be converted into carbon emissions by combining the vehicle's electrical energy consumption (power obtained by multiplying the instantaneous voltage and current) with the vehicle's mileage and battery efficiency; baseline carbon emissions can be estimated based on traditional vehicle fuel consumption and mileage data, and compared with the current vehicle's carbon emissions to obtain a carbon reduction value; vehicle energy consumption can be calculated by the vehicle's power consumption (the product of the instantaneous voltage and current) combined with the distance traveled and load conditions.

[0061] Starting from the actual energy consumption of the vehicle, the mileage traveled per unit time is calculated, accurately to the individual vehicle, and the carbon emission reduction is accurately calculated by benchmarking the standard baseline, thereby promoting the assetization of carbon emission reduction from new energy vehicle travel.

[0062] The calculated carbon data is temporarily stored in the storage module; if the network connection is unstable or interrupted, the data will be temporarily stored and uploaded after the network is restored.

[0063] Furthermore, in one embodiment, the calculation of real-time operation data by the local carbon metering model also includes: reviewing historical data through the local experience playback mechanism to obtain carbon emission calculation experience of early driving behavior, so as to prevent the model from forgetting important historical information. Specifically, the vehicle operation data and carbon data calculation results stored in the local historical time are reviewed, and compared with the current operation status of the vehicle, so as to adjust the weights and parameters of the current local carbon metering model. The formula is:

[0064]

[0065] Where, L local-clone is the local replay loss function, which is used to measure the difference between the distribution of historical data and the output distribution of the current model during the local experience replay process; q is the qth related event in the historical time, representing a possible state or output in the historical data; h r It is the historical hidden state, which indicates the state of the system in the past period of time. It is used to represent the historical behavior or characteristics of the system and to help the model learn from past experience during local playback. local represents the conditional probability distribution of historical data, μ local (qh r ) Specifically, in the local experience playback process, given the historical hidden state h r The probability of event q occurring under the same conditions. This probability represents the distribution of historical data and reflects the frequency or possibility of an event occurring under the same conditions in the past. represents the conditional probability distribution of the current local carbon accounting model output, Specifically, in the current local model parameter α local Next, given the historical hidden state h r The probability distribution of event q under the condition of , which represents the predicted probability of the model under the current conditions. The difference between the historical data distribution and the current model output distribution is measured by logarithmic operation. The larger the difference, the farther the output of the current model deviates from the historical experience. By minimizing this difference, the model can better adapt to historical experience and prevent forgetting important information from the past.

[0066] S3: Use traditional encryption algorithms to encrypt real-time operation data, use homomorphic encryption technology to encrypt carbon data, and transmit the encrypted real-time operation data and carbon data to the carbon measurement operation platform;

[0067] Before the data is ready to be uploaded, it is encrypted by the encryption module to ensure that it will not be leaked or tampered with during the data transmission process.

[0068] Specifically, traditional encryption algorithms include symmetric encryption algorithm AES (Advanced Encryption Standard), triple data encryption algorithm 3DES (Triple Data Encryption Standard), asymmetric encryption algorithm RSA

[0069] (Rivest-Shamir-Adleman), stream cipher algorithm ChaCha20, elliptic curve encryption algorithm ECC (Elliptic Curve Cryptography), etc. Since the amount of real-time running data is large and requires efficient encryption, AES is the best choice, especially AES-256, which can provide higher security and encryption speed. If the system requires higher computing resources and requires lower power consumption, ChaCha20 is also a good choice.

[0070] In one embodiment, carbon data is encrypted using homomorphic encryption technology, specifically including:

[0071] The weight matrix A of the local carbon accounting model j Encoding processing, the formula is:

[0072] Encode(A j )=Flatten(Pad(A j ,0,n));

[0073] In the formula, Encode(A j ) is the encoded weight matrix; Pad represents the weight matrix A j Fill, 0 represents the filling value, and n is the size of the matrix after encoding; the purpose of this operation is to ensure that the size of the matrix is ​​suitable for subsequent encryption processing; Flatten represents a flattening operation, which is used to convert a multi-dimensional matrix into a one-dimensional vector;

[0074] Using the public key p k The encoded weight matrix Encode(A j ) is encrypted, the formula is:

[0075]

[0076] In the formula, is the encrypted weight matrix, and Enc is the encryption function.

[0077] By using the public key p k , ensuring that after data is encrypted, homomorphic computing can still be performed without decryption, and the data is safe even if it is intercepted during transmission.

[0078] S4: The carbon metering operation platform obtains the CAN message information of the vehicle from the car company, combines the encrypted real-time operation data and carbon data uploaded by multiple on-board carbon metering terminals, calculates and optimizes the global carbon emission model based on federated learning technology, and generates the global calculation results of carbon emissions;

[0079] CAN message information includes the vehicle's historical operating data and system performance parameters, which are used to supplement and correct the real-time operating data uploaded by the on-board carbon metering terminal;

[0080] The traditional method is usually to aggregate the data of multiple vehicles to the cloud, and then perform centralized calculations, but this centralized calculation method is prone to privacy risks. Therefore, the present invention introduces federated learning technology, so that each vehicle can calculate local carbon emission data locally. Each on-board carbon metering terminal can learn the vehicle's operating mode and optimize the carbon emission model locally through its own computing resources (such as MCU module), and upload the encrypted model parameters of each vehicle to the carbon metering operation platform through federated learning. The carbon metering operation platform performs weighted aggregation on the model parameters of different vehicles, generates a global optimization model, and then distributes the model back to the on-board carbon metering terminal, thereby improving the accuracy of the carbon metering model and avoiding the impact of centralized data processing on vehicle privacy.

[0081] This distributed computing model not only improves data privacy protection, but also enables personalized carbon emission modeling based on different types of vehicles and different driving environments.

[0082] Furthermore, the carbon metering operation platform calculates and optimizes the global carbon emission model based on federated learning technology, and also includes: performing global experience playback based on CAN message information of different vehicles, the formula is:

[0083]

[0084] Where, L global-clone is the global replay loss function; q is the qth related event in the historical time; h r is the historical hidden state; μ global (qh r ) means that in the global playback process, given the historical hidden state h r The probability of event q occurring under

[0085] Indicates the current global carbon emission model parameter α global Next, given the historical hidden state h r The probability distribution of event q occurring under the condition .

[0086] The carbon metering operation platform supports dynamic learning and continuous optimization of vehicle carbon data, and adopts experience playback technology to ensure that the platform can maintain the long-term stability and accuracy of the carbon emission model according to different driving scenarios and long-term usage. It can also be set to automatically adjust the carbon emission calculation weights according to changes in different driving environments (such as rainy days, mountainous areas, etc.) to adapt to the carbon emission calculation needs in complex scenarios.

[0087] The carbon measurement operation platform aggregates the local model parameters of different vehicles through a joint learning algorithm to form a global carbon emission model. The local experience playback data of each vehicle will be uploaded to the platform through joint learning, and the carbon measurement operation platform will form a more accurate global model by aggregating the encrypted local model parameters.

[0088] In one specific embodiment, the carbon metering operation platform uses a weighted aggregation algorithm to optimize and update the global carbon emission model. The update formula is:

[0089]

[0090] In the formula, Represents the updated global carbon emission model parameters, which are the global model weights obtained after each round of training in the joint learning process and are used for the next round of model training and optimization; are the local carbon measurement model parameters of the nth vehicle. These vehicles train the model locally based on their own data, and the obtained parameters are encrypted to protect privacy; M is the number of vehicles participating in the joint learning; is the global carbon emission model parameter used in the current iteration of the server; τ represents the current iteration round of joint learning; Add(·) represents the addition operation of the matrix; Mul(·) represents the multiplication operation of the matrix.

[0091] S5: The carbon metering operation platform generates a carbon emission report based on the global calculation results of carbon emissions, and interacts with external systems (such as carbon trading platforms) for data. At the same time, the global carbon emission model optimized by federated learning is fed back to each on-board carbon metering terminal to improve the local carbon metering model and achieve dynamic adjustment and optimization.

[0092] The carbon emission report contains accurate carbon emission reductions calculated based on local and global playback mechanisms, and blockchain technology can be used to ensure data transparency and immutability.

[0093] like Figure 2 As shown, another embodiment of the present invention provides a vehicle carbon emission security calculation system based on the Internet of Vehicles, including a vehicle-mounted carbon metering terminal and a carbon metering operation platform;

[0094] The vehicle-mounted carbon metering terminal includes:

[0095] CAN communication module, used for data transmission between vehicle CAN bus and MCU module;

[0096] The MCU module, as the core processing unit of the vehicle-mounted carbon metering terminal, is embedded with a local carbon metering model. It obtains the real-time operation data of the vehicle from the vehicle CAN bus through the CAN communication module, and calculates the real-time operation data through the local carbon metering model to obtain the carbon data of the vehicle;

[0097] A storage module for temporarily storing the vehicle's real-time operating data and carbon data when the network is unstable or disconnected;

[0098] Data encryption module, used to encrypt real-time operation data using traditional encryption algorithms and encrypt carbon data using homomorphic encryption technology;

[0099] External communication module, used to transmit encrypted real-time operation data and carbon data to the carbon measurement operation platform via 4G CAT1, 5G or 6G;

[0100] The carbon metering operation platform obtains the vehicle's CAN message information from the vehicle manufacturer, combines the encrypted real-time operation data and carbon data uploaded by multiple on-board carbon metering terminals, calculates and optimizes the global carbon emission model based on federated learning technology, generates global calculation results of carbon emissions, and generates carbon emission reports based on the global calculation results, and interacts with external systems for data.

[0101] 4G CAT1 is a specific 4G standard that focuses on low power consumption, low cost, and moderate data transmission performance while maintaining basic 4G network characteristics. 4G CAT1 modules are designed with energy efficiency in mind, and consume less power in standby and transmission states than traditional 4G modules. It can provide relatively low network latency, which is beneficial for applications with high real-time requirements.

[0102] Use low-cost MCU modules combined with 4G CAT1 communication modules and CAN bus to access the vehicle, obtain real-time operation data from the vehicle CAN communication protocol, ensure the independence and authority of data collection, and be able to adapt to different types of new energy vehicles (plug-in hybrid, fuel vehicles, etc.).

[0103] Furthermore, the MCU module calculates the real-time operation data through the local carbon metering model, and also includes: reviewing the vehicle's operation data and carbon data calculation results in the historical time through the local experience playback mechanism, and comparing them with the vehicle's current operation status, so as to adjust the weights and parameters of the current local carbon metering model. The formula is:

[0104]

[0105] Where, L local-cloneis the local replay loss function; q is the qth related event in the historical time; h r is the historical hidden state; μ local (qh r ) means that in the local experience playback process, given the historical hidden state h r The probability of event q occurring under Represents the current local carbon measurement model parameter α local Next, given the historical hidden state h r The probability distribution of event q under the condition of

[0106] The carbon metering operation platform calculates and optimizes the global carbon emission model based on federated learning technology, and also includes: performing global experience playback based on CAN message information of different vehicles. The formula is:

[0107]

[0108] Where, L global-clone is the global replay loss function; q is the qth related event in the historical time; h r is the historical hidden state; μ global (qh r ) means that in the global playback process, given the historical hidden state h r The probability of event q occurring under

[0109] Indicates the current global carbon emission model parameter α global Next, given the historical hidden state h r The probability distribution of event q occurring under the condition .

[0110] In summary, the present invention can obtain the vehicle's operating data from the vehicle's CAN bus in real time through the on-board carbon metering terminal deployed on each vehicle, and calculate these data in real time through the embedded local carbon metering model to generate the vehicle's carbon emissions. This design ensures the immediacy and high accuracy of carbon emission data, and is particularly suitable for the dynamic monitoring needs of different types of vehicles. Compared with the traditional method that relies on post-data extraction, this system significantly improves the real-time calculation capability of carbon emissions and ensures the accuracy and timeliness of the data.

[0111] Equipped with a storage module, it can temporarily store the vehicle's real-time operating data and carbon data when the network is unstable or disconnected. This function effectively ensures the integrity of the data, and key data will not be lost even when the transmission is interrupted. After the network is restored, the system can automatically retransmit this data to the carbon metering operation platform, thereby ensuring the continuity and reliability of the entire data flow. This redundant design greatly improves the stability of the system under harsh network conditions.

[0112] By using traditional encryption algorithms (such as AES, RSA, etc.) to encrypt the vehicle's real-time operating data and using homomorphic encryption technology for carbon data, the system provides strong security protection during data transmission. The use of homomorphic encryption allows encrypted data to be calculated and processed without decryption, ensuring the security and privacy of carbon data during transmission and processing in the cloud. This design not only effectively prevents the risk of data tampering and theft during transmission, but also improves the protection capabilities of the entire system in terms of data security, meeting the current high standards for data privacy protection.

[0113] The carbon metering operation platform can obtain encrypted real-time operation data and carbon data from multiple on-board carbon metering terminals, and calculate and optimize the global carbon emission model based on federated learning technology. Federated learning allows multiple data sources to collaboratively train the global model without sharing the original data. This distributed learning method protects data privacy while ensuring that the system can process carbon emission data from a large number of vehicles and generate a more accurate global carbon emission model. Compared with the traditional centralized data processing model, this system greatly improves the efficiency and accuracy of large-scale carbon data processing, and can achieve dynamic optimization of carbon emission models between different vehicles.

[0114] Through the external communication module, it supports high-speed data transmission of 4G CAT1, 5G or 6G networks, can quickly upload encrypted data to the carbon measurement operation platform, and interact with external systems. This flexible communication method ensures the system's application expansion capabilities under different network conditions in the future, while supporting data sharing with various external systems, such as the generation of carbon emission reports and carbon asset management. This provides a technical basis for cross-industry and cross-regional data sharing and certification, especially in the field of international carbon emission reduction data mutual recognition, which has important application value.

[0115] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of various changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A vehicle carbon emissions safety calculation method based on vehicle networking, characterized in that: The method comprises: Deploy an on-board carbon metering terminal on each vehicle and connect it to the carbon metering operation platform in the cloud through the IoV network; The vehicle-mounted carbon metering terminal is embedded with a local carbon metering model, which obtains the real-time operation data of the vehicle from the vehicle CAN bus, and calculates the real-time operation data through the local carbon metering model to obtain the carbon data of the vehicle; The real-time operation data is encrypted using a traditional encryption algorithm, the carbon data is encrypted using a homomorphic encryption technology, and the encrypted real-time operation data and carbon data are transmitted to a carbon metering operation platform; The carbon metering operation platform obtains the vehicle's CAN message information from the vehicle manufacturer, combines the encrypted real-time operation data and carbon data uploaded by multiple on-board carbon metering terminals, calculates and optimizes the global carbon emission model based on federated learning technology, and generates global calculation results of carbon emissions.

2. The method for safely calculating vehicle carbon emissions based on the Internet of Vehicles according to claim 1, characterized in that: The real-time operation data includes instantaneous voltage, instantaneous current, mileage, speed, acceleration, vehicle VIN code and other relevant data; the vehicle carbon data includes carbon emissions, carbon reduction and vehicle energy consumption of each vehicle.

3. The method for safely calculating vehicle carbon emissions based on the Internet of Vehicles according to claim 1, characterized in that: Calculating the real-time operation data through the local carbon metering model also includes: reviewing the operation data and carbon data calculation results of the vehicle in the historical time through the local experience playback mechanism, and comparing them with the current operation status of the vehicle, so as to adjust the weights and parameters of the current local carbon metering model. The formula is: Where, L local-clone is the local replay loss function; q is the qth related event in the historical time; h r is the historical hidden state; μ local (q|h r ) means that in the local experience playback process, given the historical hidden state h r The probability of event q occurring under Represents the current local carbon measurement model parameter α local Next, given the historical hidden state h r The probability distribution of event q occurring under the condition .

4. The method for safely calculating vehicle carbon emissions based on the Internet of Vehicles according to claim 1, characterized in that: The traditional encryption algorithms include symmetric encryption algorithm AES, triple data encryption algorithm 3DES, asymmetric encryption algorithm RSA, stream cipher algorithm ChaCha20 and elliptic curve encryption algorithm ECC.

5. The method for safely calculating vehicle carbon emissions based on the Internet of Vehicles according to claim 1, characterized in that: The method of encrypting the carbon data using homomorphic encryption technology specifically includes: The weight matrix A of the local carbon accounting model j Encoding processing, the formula is: Encode(A j )=Flatten(Pad(A j ,0,n)); In the formula, Encode(A j ) is the encoded weight matrix; Pad represents the weight matrix A j Fill, 0 represents the filling value, n is the size of the matrix after encoding; Flatten represents the flattening operation, which is used to convert a multi-dimensional matrix into a one-dimensional vector; Using the public key p k The encoded weight matrix Encode(A j ) is encrypted, the formula is: In the formula, Represents the encrypted weight matrix, and Enc is the encryption function.

6. The method for safely calculating vehicle carbon emissions based on the Internet of Vehicles according to claim 1, characterized in that: The carbon metering operation platform calculates and optimizes the global carbon emission model based on the federated learning technology, and also includes: performing global experience playback based on the CAN message information of different vehicles, the formula is: Where, L global-clone is the global replay loss function; q is the qth related event in the historical time; h r is the historical hidden state; μ global (q|h r ) means that in the global playback process, given the historical hidden state h r The probability of event q occurring under Indicates the current global carbon emission model parameter α global Next, given the historical hidden state h r The probability distribution of event q occurring under the condition .

7. The method for safely calculating vehicle carbon emissions based on the Internet of Vehicles according to claim 1, characterized in that: The carbon measurement operation platform uses a weighted aggregation algorithm to optimize and update the global carbon emission model. The update formula is: In the formula, represents the updated global carbon emission model parameters; is the local carbon accounting model parameter of the nth vehicle, M is the number of vehicles participating in the joint learning; is the global carbon emission model parameter used in the current iteration of the server; τ represents the current iteration round of joint learning; Add(·) represents the addition operation of the matrix; Mul(·) represents the multiplication operation of the matrix.

8. The method for safely calculating vehicle carbon emissions based on the Internet of Vehicles according to claim 1, characterized in that: The method also includes: the carbon metering operation platform generates a carbon emission report according to the global calculation results of carbon emissions, and interacts with the external system for data. At the same time, the global carbon emission model optimized by federated learning is fed back to each vehicle-mounted carbon metering terminal to improve the local carbon metering model and realize dynamic adjustment and optimization.

9. A vehicle carbon emissions security calculation system based on a vehicle networking-based carbon emissions security calculation method according to any one of claims 1 to 8, characterized in that: The system includes an on-board carbon metering terminal deployed on each vehicle, and a carbon metering operation platform; The vehicle-mounted carbon metering terminal comprises: CAN communication module, used for data transmission between vehicle CAN bus and MCU module; The MCU module, as the core processing unit of the vehicle-mounted carbon metering terminal, is embedded with a local carbon metering model, obtains the real-time operation data of the vehicle from the vehicle CAN bus through the CAN communication module, and calculates the real-time operation data through the local carbon metering model to obtain the carbon data of the vehicle; A storage module for temporarily storing the vehicle's real-time operating data and carbon data when the network is unstable or disconnected; A data encryption module, used to encrypt the real-time operation data using a traditional encryption algorithm and to encrypt the carbon data using a homomorphic encryption technique; External communication module, used to transmit encrypted real-time operation data and carbon data to the carbon measurement operation platform via 4G CAT1, 5G or 6G; The carbon metering operation platform obtains the vehicle's CAN message information from the vehicle manufacturer, combines the encrypted real-time operation data and carbon data uploaded by multiple on-board carbon metering terminals, calculates and optimizes the global carbon emission model based on federated learning technology, generates global calculation results of carbon emissions, and generates a carbon emission report based on the global calculation results, and interacts with external systems for data.

10. The vehicle carbon emissions safety calculation system according to claim 9, characterized in that: The MCU module calculates the real-time operation data through the local carbon metering model, and also includes: reviewing the operation data and carbon data calculation results of the vehicle in the historical time through the local experience playback mechanism, and comparing them with the current operation status of the vehicle, so as to adjust the weights and parameters of the current local carbon metering model. The formula is: Where, L local-clone is the local replay loss function; q is the qth related event in the historical time; h r is the historical hidden state; μ local (q|h r ) means that in the local experience playback process, given the historical hidden state h r The probability of event q occurring under Represents the current local carbon measurement model parameter α local Next, given the historical hidden state h r The probability distribution of event q under the condition of The carbon metering operation platform calculates and optimizes the global carbon emission model based on the federated learning technology, and also includes: performing global experience playback based on the CAN message information of different vehicles, the formula is: Where, L global-clone is the global replay loss function; q is the qth related event in the historical time; h r is the historical hidden state; μ global (q|h r ) means that in the global playback process, given the historical hidden state h r The probability of event q occurring under Indicates the current global carbon emission model parameter α global Next, given the historical hidden state h r The probability distribution of event q occurring under the condition .

Images