Traffic control model training method and system based on block chain and federated learning

Abstract

The invention discloses a traffic control model training method and system based on a block chain and federated learning, and the method comprises the steps: enabling each target intelligent vehicle to train a local traffic control model according to the data of a corresponding to-be-trained vehicle-mounted sensor, and obtaining a local model training result, each local model training result is sent to the block chain network; evaluating each local model training result through an intelligent contract in the block chain network to determine a qualified model training result; all the qualified model training results are aggregated through the block chain network to obtain a global model, the global model is distributed to all the target intelligent vehicles, and all the target intelligent vehicles update local traffic control models according to the global model; and the local traffic control model is used for realizing intelligent management and control of traffic scenes. The method can avoid data privacy leakage and single-point fault risk, improves the model training quality and training efficiency, improves the safety and reliability of an Internet of Vehicles system, and can be widely applied to the technical field of computers.

Description

Technical Field

[0001] This application relates to the field of computer technology, and in particular to a traffic control model training method and system based on blockchain and federated learning. Background Art

[0002] With the development of intelligent transportation systems, vehicle networking technology has become the key to improving road efficiency and traffic safety. Intelligent vehicles collect a large amount of data through on-vehicle sensors, and this data can be used to train traffic control models to achieve more intelligent traffic management. However, traditional model training methods usually require centralized data processing on a central server, which not only increases the risk of data leakage, but also may cause problems such as single-point failures and communication bottlenecks.

[0003] In summary, the technical problems existing in the related technologies need to be improved. Summary of the Invention

[0004] The embodiments of this application aim to at least solve one of the technical problems in the related technologies to some extent. For this reason, the main purpose of the embodiments of this application is to propose a traffic control model training method and system based on blockchain and federated learning, which can avoid the risk of data privacy leakage and single-point failures, improve the quality and efficiency of model training, and enhance the security and reliability of the vehicle networking system.

[0005] To achieve the above object, on the one hand, the embodiments of this application propose a traffic control model training method based on blockchain and federated learning, and the method includes the following steps:

[0006] Obtain the to-be-trained on-vehicle sensor data corresponding to a number of intelligent vehicles in the vehicle networking system;

[0007] Authenticate the identities of each of the intelligent vehicles through the blockchain network to determine a number of target intelligent vehicles participating in federated learning;

[0008] Each of the target intelligent vehicles trains a local traffic control model according to the corresponding to-be-trained on-vehicle sensor data, obtains a number of local model training results, and sends each of the local model training results to the blockchain network;

[0009] Evaluate each of the local model training results through a smart contract in the blockchain network to determine a number of qualified model training results;

[0010] Aggregate each of the qualified model training results through the blockchain network to obtain a global model, and distribute the global model to each of the target intelligent vehicles, so that each of the target intelligent vehicles updates the local traffic control model according to the global model; the local traffic control model is used to achieve intelligent management and control of traffic scenarios.

[0011] In some embodiments, before obtaining the to-be-trained vehicle-mounted sensor data corresponding to a plurality of intelligent vehicles in the vehicle networking system, the method further includes:

[0012] Obtain the local vehicle-mounted sensor data corresponding to a plurality of the intelligent vehicles in the vehicle networking system;

[0013] Perform encryption processing on each of the local vehicle-mounted sensor data by using a target encryption algorithm, and perform preprocessing on each of the encrypted local vehicle-mounted sensor data to obtain a plurality of the to-be-trained vehicle-mounted sensor data.

[0014] In some embodiments, performing encryption processing on each of the local vehicle-mounted sensor data by using a target encryption algorithm, and performing preprocessing on each of the encrypted local vehicle-mounted sensor data to obtain a plurality of the to-be-trained vehicle-mounted sensor data includes:

[0015] Perform encryption processing on each of the local vehicle-mounted sensor data by using the target encryption algorithm, and perform denoising processing on each of the encrypted local vehicle-mounted sensor data to obtain a plurality of local clean sensor data;

[0016] Perform feature extraction processing on each of the local clean sensor data to obtain a plurality of local initial traffic feature data;

[0017] Perform data fusion processing on each of the local initial traffic feature data to obtain a plurality of local multi-source traffic feature data;

[0018] Perform deep learning on each of the local multi-source traffic feature data to extract a plurality of local multi-source key traffic feature data;

[0019] Perform selection and dimensionality reduction processing on each of the local multi-source key traffic feature data to obtain a plurality of the to-be-trained vehicle-mounted sensor data.

[0020] In some embodiments, the method further includes:

[0021] Perform encryption processing on the intermediate parameters generated during the training process of the local traffic control model by using a target encryption algorithm.

[0022] In some embodiments, evaluating each of the local model training results through the smart contract in the blockchain network to determine a plurality of qualified model training results includes:

[0023] Evaluate the training quality of each of the local model training results through the smart contract in the blockchain network to determine whether each of the local model training results meets a preset training quality standard;

[0024] Perform compliance checks on each of the local model training results through the smart contract to determine whether each of the local model training results meets the preset compliance standard;

[0025] Use the local model training results that meet the preset training quality standard and the preset compliance standard as the qualified model training results.

[0026] In some embodiments, the method further includes:

[0027] Give incentives to the target intelligent vehicle corresponding to the local model training result that meets the preset training quality standard;

[0028] Give penalties to the target intelligent vehicle corresponding to the local model training result that does not meet the preset training quality standard.

[0029] In some embodiments, the aggregating, through the blockchain network, each of the qualified model training results to obtain a global model, and distributing the global model to each of the target intelligent vehicles, so that each of the target intelligent vehicles updates the local traffic control model according to the global model, includes:

[0030] Perform an aggregation process on each of the qualified model training results through the blockchain network using the decentralized feature to obtain the global model;

[0031] Distribute the global model to each of the target intelligent vehicles through the blockchain network, so that each of the target intelligent vehicles trains the corresponding local traffic control model according to the global model parameters in the global model to obtain the updated local traffic control model.

[0032] To achieve the above object, on the other hand, an embodiment of the present application proposes a traffic control model training system based on blockchain and federated learning, and the system includes the following modules:

[0033] A to-be-trained vehicle-mounted sensor data acquisition module, configured to acquire to-be-trained vehicle-mounted sensor data corresponding to a plurality of intelligent vehicles in a vehicle networking system;

[0034] An intelligent vehicle identity authentication module, configured to authenticate the identity of each of the intelligent vehicles through a blockchain network to determine a plurality of target intelligent vehicles participating in federated learning;

[0035] A local traffic control model training module, which is used for each of the target intelligent vehicles to train a local traffic control model according to the corresponding to-be-trained vehicle-mounted sensor data, obtain several local model training results, and send each of the local model training results to the blockchain network;

[0036] A local model training result evaluation module, which is used to evaluate each of the local model training results through a smart contract in the blockchain network to determine several qualified model training results;

[0037] A global model aggregation module, which is used to aggregate each of the qualified model training results through the blockchain network to obtain a global model, and distribute the global model to each of the target intelligent vehicles, so that each of the target intelligent vehicles updates the local traffic control model according to the global model; the local traffic control model is used to realize intelligent management and control of traffic scenarios.

[0038] To achieve the above object, another aspect of the embodiments of the present application proposes an electronic device, which includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, the method described above is implemented.

[0039] To achieve the above object, another aspect of the embodiments of the present application proposes a computer-readable storage medium, which stores a computer program, and when the computer program is executed by a processor, the method described above is implemented.

[0040] The embodiments of the present application at least include the following beneficial effects: The present application provides a traffic control model training method and system based on blockchain and federated learning. The solution obtains the to-be-trained in-vehicle sensor data corresponding to several intelligent vehicles in the vehicle networking system; authenticates the identities of each intelligent vehicle through the blockchain network to determine several target intelligent vehicles participating in federated learning; each target intelligent vehicle trains the local traffic control model according to the corresponding to-be-trained in-vehicle sensor data, obtains several local model training results, and sends each local model training result to the blockchain network; evaluates each local model training result through the intelligent contract in the blockchain network to determine several qualified model training results; aggregates each qualified model training result through the blockchain network to obtain a global model, and distributes the global model to each target intelligent vehicle so that each target intelligent vehicle updates the local traffic control model according to the global model; the local traffic control model is used to realize the intelligent management and control of traffic scenarios. The embodiments of the present application adopt the federated learning method, enabling intelligent vehicles to train models locally without uploading the original data to the blockchain network, reducing the risk of user privacy leakage; adopting blockchain technology enables the vehicle networking system to eliminate the single-point failure risk brought by the traditional centralized solution, and the decentralized model aggregation method enables the training results of all participating nodes to be effectively verified and monitored, preventing malicious attacks and data tampering. At the same time, automated model quality evaluation is realized based on the blockchain intelligent contract to ensure that the aggregated global model has high quality, thereby improving the reliability of the traffic control system; by introducing the combination of blockchain technology and federated learning, the need for a large amount of data transmission is reduced, the communication cost is lowered, and the security and reliability of the federated learning process in the vehicle networking system are improved; at the same time, the local traffic control model can be updated in real time according to the global model. Through the continuous update and optimization of the local traffic control model, intelligent vehicles can better adapt to complex traffic scenarios and achieve more accurate traffic management and control. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 is a flowchart of the traffic control model training method based on blockchain and federated learning provided by the embodiments of the present application;

[0042] Figure 2 is a schematic structural diagram of the blockchain-enabled federated learning framework design provided by the embodiments of the present application;

[0043] Figure 3 is a schematic structural diagram of the traffic control model training system based on blockchain and federated learning provided by the embodiments of the present application;

[0044] Figure 4 is a schematic hardware structure diagram of the electronic device provided by the embodiments of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0045] In order to make the objectives, technical solutions and advantages of the present application more clear and understandable, the present application will be further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application. When the following description relates to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The implementation manners described in the following exemplary embodiments do not represent all implementation manners consistent with the embodiments of the present application. They are only examples of systems and methods consistent with some aspects of the embodiments of the present application detailed in the appended claims.

[0046] It can be understood that the terms "first", "second", etc. used in the present application may be used in this document to describe various concepts, but unless otherwise specified, these concepts are not limited by these terms. These terms are only used to distinguish one concept from another. For example, without departing from the scope of the embodiments of the present application, the first information may also be referred to as the second information. Similarly, the second information may also be referred to as the first information. Depending on the context, the words "if", "in case" used herein may be interpreted as "when" or "while" or "in response to determining".

[0047] The terms "at least one", "a plurality of", "each", "any one", etc. used in the present application, at least one includes one, two or more than two, a plurality of includes two or more than two, each refers to each of the corresponding plurality, and any one refers to any one of the plurality.

[0048] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the technical field to which this application belongs. The terms used herein are only for the purpose of describing the embodiments of this application and are not intended to limit this application.

[0049] With the development of intelligent transportation systems, vehicle networking technology has become the key to improving road efficiency and traffic safety. Intelligent vehicles collect a large amount of data through on-vehicle sensors, and this data can be used to train traffic control models to achieve more intelligent traffic management. However, traditional model training methods usually require aggregating data to a central server for processing, which not only increases the risk of data leakage, but may also cause problems such as single-point failures and communication bottlenecks.

[0050] For example, at present, there have been some studies that combine artificial intelligence with blockchain for related applications of intelligent Internet of Vehicles to improve the technical problems of blockchain. By combining blockchain with artificial intelligence, while retaining the security and credibility of blockchain, the Internet of Vehicles and vehicle-mounted systems are endowed with the optimization decision-making capabilities of artificial intelligence. However, there are also some problems with the combination of the two: First, traditional machine learning algorithms rely on a centralized data set to train models. Generally, the larger the data set, the more accurate the trained model. To this end, the server needs to collect a lot of data from the user side, which increases the risk of user data privacy leakage. Federated learning is an emerging machine learning solution. Unlike traditional centralized machine learning, federated learning can expand the training data set as a whole by delegating training tasks to the user side for training and introducing more user participation, thereby improving the quality of the overall model. However, federated learning also faces some challenges: cross-device federated learning executes model aggregation algorithms through parameter servers, but centralized parameter servers may be subject to malicious attacks and intercepted or even tampered with the intermediate parameters of the model aggregation process. In addition, centralized parameter servers enable a large amount of remote data communication between participating nodes and parameter servers, which also leads to data tampering and privacy leakage risks.

[0051] In view of this, the embodiments of the present application provide a traffic control model training method and system based on blockchain and federated learning. The solution obtains the to-be-trained in-vehicle sensor data corresponding to a number of intelligent vehicles in the vehicle networking system; authenticates the identities of each intelligent vehicle through the blockchain network to determine a number of target intelligent vehicles participating in federated learning; each target intelligent vehicle trains the local traffic control model according to the corresponding to-be-trained in-vehicle sensor data to obtain a number of local model training results, and sends each local model training result to the blockchain network; evaluates each local model training result through the intelligent contract in the blockchain network to determine a number of qualified model training results; aggregates each qualified model training result through the blockchain network to obtain a global model, and distributes the global model to each target intelligent vehicle so that each target intelligent vehicle updates the local traffic control model according to the global model; the local traffic control model is used to realize intelligent management and control of traffic scenarios. The embodiments of the present application adopt the federated learning method, enabling intelligent vehicles to train models locally without uploading raw data to the blockchain network, reducing the risk of user privacy leakage; adopting blockchain technology enables the vehicle networking system to eliminate the single-point failure risk brought by traditional centralized solutions, and the decentralized model aggregation method enables the training results of all participating nodes to be effectively verified and monitored, preventing malicious attacks and data tampering. At the same time, automated model quality evaluation is realized based on blockchain intelligent contracts to ensure that the aggregated global model has high quality, thereby improving the reliability of the traffic control system; by introducing the combination of blockchain technology and federated learning, the need for a large amount of data transmission is reduced, the communication cost is lowered, and the security and reliability of the federated learning process in the vehicle networking system are improved; at the same time, the local traffic control model can be updated in real time according to the global model. Through the continuous update and optimization of the local traffic control model, intelligent vehicles can better adapt to complex traffic scenarios and achieve more accurate traffic management and control.

[0052] The traffic control model training method based on blockchain and federated learning provided by the embodiments of the present application relates to the field of computer technology. The traffic control model training method based on blockchain and federated learning provided by the embodiments of the present application can be applied to a terminal, or to a server, or can be software running on a terminal or a server. In some embodiments, the terminal can be a smart phone, a tablet computer, a laptop computer, a desktop computer, a smart speaker, a smart watch, a vehicle-mounted terminal, etc., but is not limited thereto; the server side can be configured as an independent physical server, or can be configured as a server cluster or a distributed system composed of multiple physical servers, or can be configured as a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN (Content Delivery Network), and big data and artificial intelligence platforms. The server can also be a node server in a blockchain network; the software can be an application that implements the traffic control model training method based on blockchain and federated learning, etc., but is not limited to the above forms.

[0053] The present application can be used in many general-purpose or special-purpose computer system environments or configurations. For example: personal computers, server computers, handheld or portable devices, tablet-type devices, multi-processor systems, microprocessor-based systems, set-top boxes, programmable consumer electronic devices, network PCs (Personal Computers), minicomputers, mainframe computers, distributed computing environments including any of the above systems or devices, and so on. The present application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. The present application can also be practiced in a distributed computing environment, where tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules can be located in local and remote computer storage media including storage devices.

[0054] It should be noted that in each specific embodiment of the present application, when it comes to relevant processing based on data related to the user's identity or characteristics, such as user information, user behavior data, user historical data, and user location information, the user's permission or consent will be obtained first. Moreover, the collection, use, and processing of these data will comply with relevant laws, regulations, and standards. In addition, when the embodiments of the present application need to obtain the user's sensitive personal information, the user's separate permission or separate consent will be obtained through methods such as pop-up windows or jumping to a confirmation page. After clearly obtaining the user's separate permission or separate consent, the necessary user-related data for the normal operation of the embodiments of the present application will be obtained.

[0055] Please refer to Figure 1 , Figure 1 which is an optional flowchart of the traffic control model training method based on blockchain and federated learning provided by the embodiments of the present application. Figure 1 The method in

[0056] Step S101: Obtain the to-be-trained in-vehicle sensor data corresponding to several intelligent vehicles in the vehicle networking system.

[0057] In some embodiments, before step S101, it may further include: obtaining the local in-vehicle sensor data corresponding to several intelligent vehicles in the vehicle networking system; encrypting each local in-vehicle sensor data using a target encryption algorithm, and preprocessing the encrypted local in-vehicle sensor data to obtain several to-be-trained in-vehicle sensor data.

[0058] In some specific embodiments, the step of encrypting each local in-vehicle sensor data using a target encryption algorithm and preprocessing the encrypted local in-vehicle sensor data to obtain several to-be-trained in-vehicle sensor data may include: encrypting each local in-vehicle sensor data using a target encryption algorithm, and respectively performing denoising processing on the encrypted local in-vehicle sensor data to obtain several local clean sensor data; respectively performing feature extraction processing on each local clean sensor data to obtain several local initial traffic feature data; respectively performing data fusion processing on each local initial traffic feature data to obtain several local multi-source traffic feature data; performing deep learning on each local multi-source traffic feature data to extract several local multi-source key traffic feature data; respectively performing selection and dimensionality reduction processing on each local multi-source key traffic feature data to obtain several to-be-trained in-vehicle sensor data.

[0059] Among them, the vehicle networking system refers to a system that realizes intelligent information exchange and sharing between vehicles, between vehicles and infrastructure, between vehicles and pedestrians, and between vehicles and the network through information and communication technologies. The vehicle networking system may include components such as in-vehicle sensors, on-vehicle units, roadside units, data centers, and communication networks. The main purpose of vehicle networking technology is to improve road safety, traffic efficiency, and driving comfort.

[0060] For an intelligent vehicle, which is a component in the vehicle networking system, an intelligent vehicle refers to a vehicle equipped with advanced in-vehicle sensors, controllers, actuators, computing platforms, and communication devices, and is capable of perceiving, understanding, making decisions, and controlling the vehicle state, surrounding environment, and driving behavior. Among them, an intelligent vehicle can be regarded as a node, and the target intelligent vehicle below can be regarded as a participating node allowed to join federated learning.

[0061] Among them, local in-vehicle sensor data refers to the data collected by various sensors mounted on an intelligent vehicle, which is local data without any processing. The local in-vehicle sensor data is used for model training on the local device of the intelligent vehicle. The sensors mounted on the intelligent vehicle may include, but are not limited to: GPS (Global Positioning System) sensors, cameras, lidars, ultrasonic sensors, etc. The local in-vehicle sensor data may include, but is not limited to: position data, speed data, direction data, and sensor feedback information (such as lidar, camera images, etc.). Among them, the position data is the real-time geographical location of the vehicle, the speed data is the driving speed of the vehicle, the direction data is the driving direction or heading angle of the vehicle, and the sensor feedback information includes object recognition, obstacle detection, road conditions, etc.

[0062] For the target encryption algorithm, it can be selected according to the actual situation, and the embodiments of the present application do not limit this. Exemplarily, the target encryption algorithm may be a homomorphic encryption algorithm or a differential privacy algorithm.

[0063] In the embodiments of the present application, the training data (local data) will not be directly sent to the blockchain. The core is to ensure that user privacy is not leaked. Specifically, after the local data is obtained, the local data is first encrypted on the local device, then the encrypted local data is preprocessed, and then the preprocessed encrypted data is used for model training locally. Finally, the model parameters generated during the training process are sent to the blockchain. In this way, even if the training data is used and processed during the training process, the original personal data will not be leaked.

[0064] Among them, the preprocessing may include denoising processing, preliminary feature extraction processing, data fusion processing, deep learning feature extraction processing, and feature selection and dimensionality reduction processing. The specific preprocessing steps are as follows:

[0065] (1) Since vehicle network data is highly dynamic and heterogeneous, before feature extraction, it is first necessary to perform denoising processing on the original data. The methods of denoising processing can include, but are not limited to: smoothing algorithms, time window processing methods, and outlier detection methods. Among them, the smoothing algorithm refers to using a low-pass filter to eliminate random fluctuations in the data; the time window processing method refers to based on time series data, using a sliding window to smooth data such as speed and position; the outlier detection method refers to removing abnormal data caused by sensor failures or data transmission errors. That is, corresponding to the step of separately performing denoising processing on each encrypted local vehicle sensor data to obtain a number of local clean sensor data.

[0066] (2) Preliminary feature extraction processing: After denoising processing, according to the specific requirements of the traffic scenario, extract features related to vehicle driving safety, path planning, obstacle recognition, etc. The preliminary features can include, but are not limited to: location-related features, dynamic state features, traffic flow features, road condition features, and spatio-temporal features. Among them, the location-related features can include the longitude and latitude information of the vehicle and the relative distance (for example, the distance between the vehicle and the intersection, the distance between the vehicle and other vehicles, etc.); the dynamic state features can include the speed, acceleration, acceleration change rate, etc. of the vehicle, which can help understand the driving state of the vehicle; the traffic flow feature is the current traffic flow situation estimated by collecting the driving speed, position, traffic density, etc. of surrounding vehicles; the road condition feature refers to identifying traffic signs, traffic lights, roadblocks, pedestrians, etc. by combining road information provided by sensors such as cameras or lidar; the spatio-temporal feature can include time series data of vehicle speed and position, which can help analyze the comparison between the current moment and past historical moments and predict future traffic conditions. That is, corresponding to the step of separately performing feature extraction processing on each local clean sensor data to obtain a number of local initial traffic feature data.

[0067] (3) Data fusion processing: Since there are often multiple sensor data sources in the vehicle network and these data are often heterogeneous, data fusion is required to combine data from multiple sensors to extract more accurate features. For example, using algorithms such as Kalman filtering to fuse GPS data and speed sensor data to optimize vehicle position and speed estimation. That is, corresponding to the step of separately performing data fusion processing on each local initial traffic feature data to obtain a number of local multi-source traffic feature data.

[0068] (4) Deep learning feature extraction processing: In some cases, the original sensor data (such as images, video streams) may need to further extract features through deep learning methods. Exemplarily, a Convolutional Neural Network (CNN) and a Recurrent Neural Network (RNN) can be used for feature extraction. Specifically, for camera image data, a convolutional neural network can be used to extract features such as object recognition, road sign recognition, and obstacle detection; for time series data (e.g., vehicle position, speed, etc.), methods such as RNN or LSTM (Long Short-Term Memory) can be used to extract temporal features. That is, corresponding to the above-mentioned step of performing deep learning on each local multi-source traffic feature data to extract several local multi-source key traffic feature data.

[0069] (5) Feature selection and dimensionality reduction processing: To improve the model training efficiency and reduce the computational complexity, the extracted features are selected and dimensionally reduced to remove redundant or irrelevant features. Exemplarily, the Principal Component Analysis (PCA) method can be used to transform multiple related variables into several important principal components through linear transformation; feature importance evaluation can be performed through methods such as information gain and correlation analysis to select the most representative features. That is, corresponding to the above-mentioned step of separately performing selection and dimensionality reduction processing on each local multi-source key traffic feature data to obtain several vehicle-mounted sensor data to be trained.

[0070] The features (vehicle-mounted sensor data to be trained) extracted through the above preprocessing methods will provide the necessary input data for the subsequent federated learning model training, while ensuring data privacy and security. In the blockchain environment, these features will be encrypted and recorded through a distributed ledger to ensure transparency and traceability. In an intelligent vehicle networking system, the process of extracting traffic scenario-related features is crucial. Traffic scenario-related features can help vehicles understand the surrounding environment and provide valuable information for subsequent deep learning model training.

[0071] Step S102, authenticate the identities of each of the intelligent vehicles through the blockchain network to determine several target intelligent vehicles participating in the federated learning;

[0072] Among them, the target intelligent vehicle refers to an intelligent device that can pass the identity authentication of the blockchain and has legality, and can be used as a vehicle participating in the federated learning.

[0073] In the specific implementation, vehicles or smart terminal devices are authenticated through the blockchain system to ensure that only trusted devices can participate in the federated learning process. Through the consensus mechanism of the blockchain (such as PoW (Proof of Work), PoS (Proof of Stake), DPoS (Delegated Proof of Stake)), the legitimacy and credibility of each participant in the model training process are guaranteed. Specifically, in the specific implementation process of participant authentication, the following steps are taken:

[0074] (1) Device registration and identity authentication;

[0075] Device registration: Each participating smart vehicle or terminal device needs to register in the blockchain system before participating in federated learning. The device registration information includes the vehicle's unique identifier (such as VIN (Vehicle Identification Number) code or device serial number), device type, hardware configuration, location (optional) and other basic information.

[0076] Public key generation and registration: Each device participates in the blockchain network by generating a public-private key pair, with the public key serving as the unique identity of the device. The private key is used for identity authentication and encrypted communication, while the public key serves as the public identity stored on the blockchain. The device's public key and related information will be registered in the blockchain as the device's "ID card".

[0077] (2) Blockchain identity verification;

[0078] Authentication request: Before a device participates in a federated learning task, it initiates an authentication request to the blockchain network, submitting its registered public key and necessary authentication information (such as device model, geographic location, etc.);

[0079] Consensus mechanism authentication: The device registration information is verified through the consensus mechanism of the blockchain (such as PoW, PoS or delegated proof mechanism, etc.). If the device identity information is verified by consensus (for example, some verification nodes confirm the identity of the device based on the device's history or reputation rating), the device is considered trustworthy and allowed to participate in model training.

[0080] (3) Permission control and access management;

[0081] Smart contract permission check: Deploy a smart contract in the blockchain to control the access rights of devices. The smart contract will automatically determine whether to allow a device to participate in training based on the device's public key and identity information. If the device identity fails verification or the device's reputation score is too low (e.g., due to past malicious behavior), the smart contract will automatically reject the device's request and prohibit it from conducting model training for federated learning.

[0082] Device reputation management: The smart contract can also dynamically evaluate the reputation of a device based on its historical records of participating in training, the quality of the uploaded model, and the device's behavior. The device's reputation score will affect its priority and rewards for participating in future training. Devices with low reputation may be restricted from participating or removed from the blockchain network.

[0083] (4) Privacy protection for authentication;

[0084] Zero-Knowledge Proof (ZKP): To enhance privacy protection, zero-knowledge proof technology can be introduced during the authentication process. A device can prove its identity without revealing any personal information. For example, a device can prove that it meets certain preset conditions (such as whether it is an authenticated vehicle) without providing specific personal or vehicle information to the blockchain.

[0085] Anonymous authentication: To further protect user privacy, an anonymous authentication mechanism, such as ring signatures or homomorphic encryption, can be used to ensure that no sensitive data is leaked during the device identity authentication process.

[0086] (5) Blockchain logs and audits;

[0087] Authentication log recording: The process of each device identity authentication will be recorded by the blockchain and an immutable log will be generated. These logs include the device's registration information, verification requests, consensus process, and the device's verification results. Each round of verification can be traced and provides evidence for subsequent system security audits.

[0088] Dynamic updates and audits: The device identity and reputation information will be continuously updated over time. If a device exhibits anomalies during model training (such as uploading false model parameters, participating in attacks, etc.), the smart contract can record it as a non-compliant node and deduct its reputation. All updates are recorded on the blockchain to ensure the transparency and credibility of the system.

[0089] Among them, the selection of the consensus mechanism can include PoW, PoS, and DPoS. Specifically: When using PoW, through the "proof of work" mechanism, the device needs to perform certain computing tasks to verify its identity, ensuring that only verified devices can participate in the model training process and preventing malicious devices from launching attacks; When using PoS, for certain specific devices, the "proof of stake" mechanism can be used. Among them, the device needs to pledge a certain amount of resources (such as tokens, computing power, etc.) to prove its legitimacy and credibility in participating in federated learning; When using DPoS, through the "delegated proof of stake" mechanism, trusted verification nodes can be selected to authenticate the device identity, reducing the computing consumption during the authentication process and improving efficiency.

[0090] Step S103, each of the target intelligent vehicles trains the local traffic control model according to the corresponding to-be-trained vehicle-mounted sensor data, obtains a number of local model training results, and sends each of the local model training results to the blockchain network;

[0091] Among them, the local traffic control model is a deep learning model. Each participant (i.e., the target intelligent vehicle) trains the deep learning model on the local device according to its own corresponding local data (the to-be-trained vehicle-mounted sensor data after preprocessing). Among them, each device will only send the local model training results (i.e., model parameters) rather than the original data (local data) to the blockchain network.

[0092] Among them, the deep learning model (local traffic control model) refers to a machine learning model used in applications such as intelligent transportation systems, autonomous driving, and cooperative driving in the Internet of Vehicles (IoV). Specifically, the purpose of the deep learning model is to be trained on local data of in-vehicle devices (such as intelligent vehicles) to achieve automated recognition, prediction, and decision-making for traffic scenarios. The uses of the local traffic control model can include, but are not limited to, traffic scene recognition, autonomous driving decision-making, cooperative autonomous driving, and traffic prediction. Among them, traffic scene recognition refers to being able to identify important information such as traffic signals, road obstacles, traffic signs, and pedestrians through the local traffic control model based on data collected by vehicle sensors (such as lidar, cameras, GPS, etc.), so as to provide environmental perception capabilities for the autonomous driving system; autonomous driving decision-making means that by training the local traffic control model, the vehicle can make driving decisions (such as steering, accelerating, braking, etc.) based on real-time data (such as speed, position, traffic status, etc.); cooperative autonomous driving refers to sharing information among multiple intelligent vehicles using the local traffic control model to conduct cooperative driving and improve road safety and traffic efficiency. For example, communication between vehicles can help achieve automated platoon driving, intelligent traffic signal control, etc.; traffic prediction means that the local traffic control model can also be used to predict traffic flow, congestion conditions, road conditions, etc., and provide navigation suggestions for drivers or autonomous driving systems.

[0093] For the initial construction steps of the deep learning model (local traffic control model), specifically: (1) Data collection and preprocessing: Each in-vehicle device obtains data from its local sensors (such as cameras, radars, GPS, etc.). This data usually contains information such as the vehicle's speed, position, direction, and surrounding environment. Due to the high dynamics and heterogeneity of IoV data, the data needs to be preprocessed, including operations such as noise removal, feature extraction, and data standardization; (2) Model selection and construction: The deep learning model uses multi-layer neural networks (such as convolutional neural network CNN, long short-term memory network LSTM, etc.). Exemplarily, in environmental perception tasks, a convolutional neural network (CNN) may be used to extract image features, or an LSTM may be used to process the vehicle's time-series data (such as speed changes, position changes, etc.). These models are trained on local vehicles to capture patterns and regularities in the data.

[0094] In a specific implementation, each vehicle trains a local traffic control model using its local processing capabilities (such as in-vehicle computing devices). The model training process is carried out using sensor data collected by in-vehicle devices. This training process may include steps such as forward propagation, loss calculation, and backpropagation. The steps of these training processes can refer to the technical content of relevant model training for reference, and are not elaborated in the embodiments of this application. Among them, the goal of training is to optimize the model parameters so that it can accurately perform traffic scene recognition and decision-making tasks.

[0095] In some embodiments, it may further include: encrypting the intermediate parameters generated during the training process of the local traffic control model using a target encryption algorithm.

[0096] Among them, the intermediate parameters refer to the parameters and data generated at each stage during the training process, and these parameters and data are the intermediate results of model training. For example, during the deep learning training process, the model parameters are continuously updated, and the loss function values during training will also change accordingly. These intermediate parameters do not represent the final model parameters, but refer to the parameters and related information updated each time during the training process. The model parameters are the parameters finally optimized after multiple rounds of training and are the core information for inference and prediction. The local model training results not only include the model parameters but also the loss function. The loss function is used to measure the error of the model after each round of training, and the loss function is closely related to the intermediate parameters because the performance of the model is evaluated through the loss function after each round of training and the model parameters are further adjusted. Therefore, the intermediate parameters include but are not equal to the model parameters and the loss function. The intermediate parameters are the dynamic data during the training process and are finally aggregated into the final model parameters.

[0097] For the encryption method of intermediate parameters, its method flow for encrypting local vehicle sensor data is the same. The difference between the two is the encryption stage. Specifically, the encryption of local vehicle sensor data first occurs in the local data collection and processing stage. That is, when each vehicle device processes sensor data locally, the original data is encrypted. Then, these encrypted data are used for local training, and only the encrypted training results (such as model parameters) are sent to the blockchain network. It can be understood that the time point of local data encryption is that once the original data is collected, it will be encrypted locally without leaking any privacy information; while the time point of intermediate parameter encryption is during the training process, and the parameters (intermediate parameters) updated in each round of model training will also be protected by an encryption algorithm to avoid being leaked or tampered with during transmission. The encryption of intermediate parameters and data encryption are two parallel processes, both aiming to ensure the security of privacy throughout the training process. The encryption of training data and intermediate parameters will be carried out at different stages, but both are completed on local devices, and the encryption methods and purposes of the two are the same, both to ensure the privacy and security of data. During the transmission stage, only the encrypted parameters will be transmitted to ensure that privacy is not leaked during both data transmission and storage.

[0098] In the embodiments of the present application, different from the traditional centralized training method, each vehicle device only uploads the trained model parameters (instead of the original data) to the blockchain network, protecting data privacy, and each device can still use the dispersed data sets for global model training. In the subsequent global model aggregation process, through the federated learning mechanism, the model parameters from multiple intelligent vehicles are aggregated in the blockchain network to form a global model. Finally, each vehicle will perform the next round of local training based on the global model to further improve the accuracy and reliability of the model.

[0099] In specific implementation, to protect user privacy during the training process, an encryption algorithm (such as homomorphic encryption or differential privacy technology) is used to encrypt the training data and intermediate parameters to prevent data leakage during transmission. Therefore, since the training data always remains on the local device (such as the data of vehicle sensors), even if the trained model parameters are transmitted through the blockchain, it does not involve the upload and sharing of the original data; in addition, even if the intermediate parameters (such as the intermediate results of model training) are recorded by the blockchain, the intermediate parameters are also in an encrypted version. Encrypted by an encryption algorithm (such as homomorphic encryption or differential privacy), even if these encrypted data are transmitted to the blockchain network, they cannot be decrypted, effectively preventing the risk of data leakage.

[0100] Step S104, evaluate each of the local model training results through a smart contract in the blockchain network to determine a number of qualified model training results;

[0101] In some embodiments, step S104 may include: evaluating the training quality of each local model training result through a smart contract in the blockchain network to determine whether each local model training result meets a preset training quality standard; performing a compliance check on each local model training result through the smart contract to determine whether each local model training result meets a preset compliance standard; and using the local model training results that meet the preset training quality standard and the preset compliance standard as qualified model training results.

[0102] In some embodiments, it may further include: giving an incentive to the target intelligent vehicle corresponding to the local model training result that meets the preset training quality standard; and giving a penalty to the target intelligent vehicle corresponding to the local model training result that does not meet the preset training quality standard.

[0103] In a specific implementation, each participating node (target intelligent vehicle) uploads its training result (model parameters and loss function) to the blockchain network, and the blockchain network records it as an immutable transaction record. The blockchain network, through the method of a smart contract, enables the blockchain to verify whether the training results submitted by each participating node are compliant and records the contributions of the nodes. At the same time, each block in the blockchain network stores the intermediate parameters and verification information of the training, and these intermediate parameters are used for the aggregation and verification of the global model. Through the decentralized blockchain ledger, the single-point failure and malicious attacks that a traditional centralized server may encounter are eliminated.

[0104] Among them, whether each local model training result meets the preset training quality standard is mainly judged based on the contributions of each participating node. The contributions of the participating nodes are mainly evaluated by the performance of the participating nodes in local training and the quality of the model parameters provided. The role of the smart contract is to automatically verify the contributions of each participating node on the blockchain and ensure that these contributions are compliant, thereby ensuring the quality and transparency of the global model training.

[0105] To ensure that the quality of the model parameters submitted by each node meets the standard, it is necessary to evaluate the model results of each node during each round of training. The specific evaluation process includes:

[0106] (1) Evaluation standard setting: Certain standards can be set: indicators such as loss function, accuracy, and F1-score are used to measure the training effect of the model. For example, when using a classification task, the accuracy of the model parameters can be judged by evaluating the classification accuracy. If it is a regression task, it can be evaluated by the mean squared error (MSE), etc.

[0107] (2) Loss function comparison: After multiple nodes complete local training, the values of their respective loss functions will be calculated. If the model loss function of a certain node is significantly higher than that of other nodes, it indicates that the quality of its model is poor, and there may be problems such as overfitting, underfitting, or data noise.

[0108] (3) Cross-validation: The model updates submitted by each node can be cross-validated on part of the data. That is, the data is divided into multiple subsets, and the performance of the model on different subsets is verified to ensure its good generalization ability.

[0109] (4) Consistency test: By performing a consistency test with the parameters submitted by other nodes, that is, detecting whether the model parameters of each node are within a certain standard deviation range. If the model parameters of a certain node differ greatly from those of other nodes, it means that the model submitted by this node may have deviations or problems.

[0110] Specifically, the contribution of participating nodes refers to the effective contributions made by participating nodes during the model training process, mainly including training quality, participation degree, computing resources, and model update contributions. The specific contents are as follows:

[0111] (1) Training quality: The quality and effectiveness of the model parameters submitted by participating nodes during local training. For example, whether the loss function during training decreases, and the performance of the model on the local dataset. If the model provided by a participating node can effectively improve the accuracy of the global model, the contribution of this participating node will be considered relatively large;

[0112] (2) Participation degree: The frequency of a node's participation in training and the amount of data involved in training are also part of the contribution. For example, a node may participate in multiple rounds of training, and the training results provided each time meet the requirements, which will increase the weight of its contribution;

[0113] (3) Computing resources: In some implementation processes, nodes may also calculate their total contributions based on their computing resource contributions. For example, nodes with stronger computing capabilities may show higher efficiency when processing more complex models, so their contributions can be weighted according to the computing resources provided;

[0114] (4) Model update contribution: The model parameters improved by a node through local training will be aggregated into the global model. If the model parameters provided by this node make a large contribution to the global aggregation (for example, significantly improving the accuracy of the global model), then its contribution value will also be higher.

[0115] For smart contracts, smart contracts in a blockchain network are programs used for automatic execution and enforcement of conditional agreements. Smart contracts are mainly used in the following four steps:

[0116] (1) Submission of training results: After each node (such as an intelligent vehicle) completes training locally, it uploads the parameters of the model (rather than the original data) to the blockchain network. The uploaded data includes information such as the parameters of the trained model, the value of the loss function, the number of training batches, and the algorithms used;

[0117] (2) Verification of training results: The model parameters submitted by each node will be verified through a smart contract. The smart contract will verify whether the training results submitted by each node meet certain quality standards according to predetermined criteria and verification logic. For example, the smart contract can check: 1) Whether the model parameters conform to the predetermined format and data range; 2) Whether the value of the loss function is within an appropriate range and conforms to the expected training effect; 3) Whether there is abnormal data (such as unstable training results, etc.);

[0118] (3) Hash verification: The smart contract can verify the hash value of the model parameters to ensure that they have not been tampered with during transmission.

[0119] (4) Compliance check: If the training results of a certain node are verified by the smart contract as non-compliant (such as poor model performance or abnormal submitted data), then the node will be punished or recorded as non-compliant. Conversely, compliant nodes will be incentivized and may receive rewards such as increased participation rights and enhanced reputation. Among them, the submission of each node will affect its reputation score. For example, if the training results of a node fail to meet the expected standards twice in a row, the reputation score of the node will decrease and it may be restricted from participating in subsequent training tasks.

[0120] Exemplarily, assume that there are 3 intelligent vehicles (nodes A, B, and C) participating in the training of a global traffic scenario model. During the model training process, the process of each node is as follows: (1) Node A collects local traffic data (such as vehicle speed, location, etc.) and trains a model locally. Node A calculates the loss function and uploads the training results (such as model parameters, loss function values, etc.) to the blockchain. (2) The smart contract automatically verifies the training results submitted by node A on the blockchain. The smart contract checks whether the model of node A is compliant and meets the quality standards. If it is qualified, then the training results of node A will be recorded as valid and its contribution to the global model will be evaluated. Among them, nodes B and C perform the same process and upload their respective training results. The training results of each node will be verified through the smart contract.

[0121] Among them, if the contribution of a certain node is non-compliant, the smart contract will trigger a punishment mechanism, and the punishment may be manifested as reducing its future right to participate in training, reducing rewards, or recording the low-quality performance of the node. If a node submits valid training results and makes effective contributions, it will receive corresponding rewards (such as increasing reputation points or obtaining certain material rewards).

[0122] Among them, for the nodes that submit high-quality model parameters, the smart contract will give rewards. The forms of rewards can be: (1) Reputation score increase: Each time a node submits model parameters that meet the standards, the node's reputation score will increase, and it can obtain more tasks in the future and even higher weights in future training. (2) Token incentives: If the blockchain system adopts a token incentive mechanism (such as based on cryptocurrencies or points), excellent nodes can receive token rewards, which can be used to participate in more tasks in the future or exchange for value.

[0123] For the nodes that submit unqualified model parameters, the following penalty measures will be taken: (1) Loss of participation right: If the model quality submitted by a certain node continuously fails to meet the standards, the smart contract will automatically list it as a "penalty node" and restrict its continued participation in the subsequent training process; (2) Reputation decline: Each unqualified submission will cause the node's reputation score to decline. Nodes with low reputation scores may lose the priority to participate in subsequent training or even be removed from the system; (3) Weight reduction: The system can adjust the weight of a node in model aggregation according to its reputation score. If a node has a low reputation score, its model contribution will be reduced in the update of the global model.

[0124] Exemplarily, assume that after a certain round of training, node A submits its model parameters to the blockchain. After evaluation by the smart contract, it is found that the model accuracy of node A is 70%, while the accuracies of other nodes are generally above 85%. At this time, according to the preset standards, the smart contract believes that the model quality of node A does not meet the standards, so node A will be punished, its reputation score will be reduced, and it may lose the qualification to participate in the next two rounds of training; in addition, during the update of the global model, the model of node A is removed from the aggregation of the global model, and the contribution model parameters of other nodes are weighted and integrated. At this time, the updated global model will continue to be propagated and updated among the valid nodes.

[0125] In specific implementation, for each round of training, the quality and accuracy of the model parameters will be verified according to the contributions provided by each node. If the model parameters submitted by a certain node do not meet the standards, it will be punished and may lose the participation right or receive a reputation penalty. During the subsequent aggregation process, the smart contract evaluates the training quality of each node to ensure that only qualified training results can enter the aggregation of the global model.

[0126] Step S105, aggregate the qualified model training results through the blockchain network to obtain a global model, and distribute the global model to each of the target intelligent vehicles, so that each of the target intelligent vehicles updates the local traffic control model according to the global model; the local traffic control model is used to implement intelligent management and control of traffic scenarios.

[0127] In some embodiments, step S105 may include: using the decentralized feature of the blockchain network to aggregate the training results of each qualified model to obtain a global model; distributing the global model to each target intelligent vehicle through the blockchain network, so that each target intelligent vehicle trains the corresponding local traffic control model according to the global model parameters in the global model to obtain an updated local traffic control model.

[0128] Wherein, when a certain node is punished, the model update includes the following steps:

[0129] (1) Removing unqualified models: When a certain node is punished and removed, the model parameters submitted by it will be excluded from the global model. If the parameters of this node have participated in the aggregation and affected the global model, a "rollback" mechanism needs to be used for correction to ensure that the accuracy of the global model is not affected.

[0130] (2) Dynamic update: The system will re-perform global aggregation based on the valid models submitted by other nodes. At this time, the model parameters originally contributed by the unqualified node will be discarded, and the model aggregation will continue to be updated by the valid parameters of the remaining nodes.

[0131] (3) Calibration and optimization: After a certain node is punished, the system will re-calibrate the training load of the remaining participating nodes to ensure that the model can continue to be optimized through high-quality nodes.

[0132] Among them, the distributed global model that has been aggregated and updated contains the training information and optimization results from all participating nodes (in-vehicle devices). This model has been aggregated and its correctness and integrity are guaranteed by the blockchain. The role of the global model is to be used by intelligent vehicles (or other participating nodes) for the next round of local training and optimization. In the application scenario of the vehicle networking, vehicles will perform local training according to the global model to improve tasks such as driving behavior prediction, traffic signal optimization, and autonomous driving algorithms, so as to provide more efficient and accurate decision-making support in subsequent applications.

[0133] In specific implementation, using the decentralized feature of the blockchain, the training results are aggregated on the basis of all participating parties to obtain a global model. The aggregated global model is updated and distributed to all participating nodes through the blockchain network, and each participating node performs the next round of local training according to this global model. The blockchain ensures the transparency and security of model updates. All participants can view and verify the process and results of model updates in real time. Records of all training processes are saved as immutable blockchain transaction logs, and each node can query the blockchain to trace and audit the training process to ensure that there is no external malicious attack. Among them, the aggregation algorithm can adopt weighted average or other adaptive aggregation methods, and the embodiments of the present application do not limit this.

[0134] Specifically, after all nodes submit their training results to the blockchain, the blockchain aggregates the contributions of each node through a smart contract and performs decentralized model aggregation. This process includes calculating the contribution values of each node and weighted updating of the global model according to the contribution degree. Finally, the aggregated global model is updated and distributed to all nodes through the smart contract to ensure that each participant can perform the next round of training based on the new global model, continuously updating and optimizing the local model and the global model.

[0135] In the embodiments of the present application, by dynamically adjusting and verifying model parameters, quality evaluation, automated verification of smart contracts, and introduction of reward and punishment mechanisms, the quality of model training and the collaboration effect of participants can be ensured. By using decentralized blockchain technology, transparency and efficiency are achieved, while interference from malicious nodes is prevented, effectively guaranteeing the security and fairness of the system.

[0136] In specific implementation, the embodiments of the present application integrate federated learning and blockchain, potentially changing the intelligent vehicle networking with decentralized and insecure properties. Specifically, using a decentralized blockchain can eliminate the need for a central server in the federated learning training of traffic scenarios. Instead, a shared immutable ledger is used to aggregate the global model and distribute the global updates to the learning clients for direct calculation on the devices. The decentralization of model aggregation not only reduces the risk of single point of failure, thereby improving training reliability, but also reduces the burden on the central server in global model aggregation, especially when there are multiple clients in the intelligent vehicle networking. The learning updates are appended to the immutable blocks for information exchange between clients during training, thus ensuring high security of the training against external attacks. Copying the blocks to the entire network also allows all clients to verify and track the training progress to ensure a high level of trust and transparency in the new blockchain-enabled framework. Eliminating the central server for traffic scenario model aggregation may reduce communication costs and attract more vehicle networking users to participate in data training based on its decentralized network topology to improve the scalability of the intelligent vehicle network. In addition, the embodiments of the present application focus on the research of a new blockchain-enabled federated learning framework. During the process of model training based on federated learning, the information necessary for training the model is transmitted between the participating parties through the blockchain, but the data cannot be transmitted to effectively perform federated learning-based calculations while protecting user privacy.

[0137] Steps S101 to S105 illustrated in the embodiments of the present application include obtaining the to-be-trained in-vehicle sensor data corresponding to a number of intelligent vehicles in the vehicle networking system; authenticating the identities of each intelligent vehicle through the blockchain network to determine a number of target intelligent vehicles participating in federated learning; each target intelligent vehicle training the local traffic control model according to the corresponding to-be-trained in-vehicle sensor data to obtain a number of local model training results, and sending each local model training result to the blockchain network; evaluating each local model training result through the smart contract in the blockchain network to determine a number of qualified model training results; aggregating each qualified model training result through the blockchain network to obtain a global model, and distributing the global model to each target intelligent vehicle so that each target intelligent vehicle updates the local traffic control model according to the global model; the local traffic control model is used to implement intelligent management and control of traffic scenarios. The embodiments of the present application adopt the federated learning method, enabling intelligent vehicles to train models locally without uploading raw data to the blockchain network, reducing the risk of user privacy leakage; adopting blockchain technology enables the vehicle networking system to eliminate the single-point failure risk brought by traditional centralized solutions, and the decentralized model aggregation method enables the training results of all participating nodes to be effectively verified and monitored, preventing malicious attacks and data tampering. At the same time, automated model quality evaluation is realized based on the blockchain smart contract to ensure that the aggregated global model has high quality, thereby improving the reliability of the traffic control system; by introducing the combination of blockchain technology and federated learning, the need for a large amount of data transmission is reduced, the communication cost is lowered, and the security and reliability of the federated learning process in the vehicle networking system are improved; at the same time, the local traffic control model can be updated in real time according to the global model. Through the continuous update and optimization of the local traffic control model, intelligent vehicles can better adapt to complex traffic scenarios and achieve more accurate traffic management and control.

[0138] To explain the principle of the technical solution of the present invention in detail, the overall process of the present invention will be described below in conjunction with some specific embodiments. It is easy to understand that the following is an explanation of the technical principle of the present invention and should not be regarded as a limitation of the present invention.

[0139] Please refer to Figure 2 , Figure 2 which is a schematic structural diagram of the blockchain-enabled federated learning framework design provided by the embodiments of the present application. Figure 2 It is a federated learning framework enabled by blockchain. The traffic control model training method based on blockchain and federated learning provided in the embodiments of the present application works based on this federated learning framework. As Figure 2 shown, the specific work content and working principle of each data processing layer in the blockchain-enabled federated learning framework are as follows:

[0140] (1) Data access layer;

[0141] 1) Model Invocation: At this level, the model invocation operation is triggered by the blockchain network, and the model is invoked through the execution of smart contracts. When the model is shared among nodes, it is accessed through the API (Application Programming Interface), which conforms to the transparency and decentralization characteristics of blockchain technology.

[0142] 2) Model Transaction: Model transaction refers to the parameter exchange during the model update process, and the transaction is recorded by the blockchain to ensure the integrity of data and model updates.

[0143] 3) Parameter Configuration: Each participating party adjusts and configures the parameters of the model according to the training needs. The immutability of the blockchain ensures the transparency and security of these configuration processes.

[0144] 4) API Interface: Provides a communication interface for model access and configuration to ensure the secure transmission of information among nodes.

[0145] (2) Smart Contract Layer;

[0146] 1) Automated Model Aggregation: Within the federated learning framework, the model aggregation process is automated, and all models from different nodes are aggregated through blockchain technology without the intervention of a central server.

[0147] 2) Automated Parameter Verification: The smart contract is used to verify the validity of the model parameters submitted by each node to ensure that only qualified parameters can enter the global model aggregation process.

[0148] 3) Automated Reputation Scoring: To encourage high-quality model contributions, the smart contract also performs reputation scoring based on the quality of node contributions. The incentive mechanism promotes participants to comply with the norms and prevents malicious behavior.

[0149] (3) Incentive Mechanism Layer;

[0150] 1) Minimization of Resource Overhead: This layer optimizes the use of computing resources through smart contracts, reduces unnecessary resource waste, and lowers the operating cost, especially in a decentralized environment.

[0151] 2) Maximization of Node Utility: By rewarding and incentivizing participants, it ensures the maximum utility output of nodes, that is, nodes provide high-quality model updates as much as possible.

[0152] 3) Maximization of Overall Revenue: The incentive mechanism ensures the maximum revenue of the entire federated learning process through the incentive and penalty mechanisms for nodes, that is, the improvement of model training quality and efficiency.

[0153] (4) Consensus Protocol Layer;

[0154] 1) Proof of Node Contribution: The consensus mechanism of the blockchain (such as PoW or PoS) ensures that each node proves its validity according to its contribution, ensuring the honesty of participants.

[0155] 2) Auditor of Collaborator Parameters: The blockchain provides an auditing function for the parameter upload and update of all nodes to ensure that collaborators do not upload false parameters.

[0156] 3) Competition for Block Bookkeeping Rights: Nodes obtain the opportunity to record blocks according to their contributions, and ensure the decentralization and security of the blockchain through the right of bookkeeping.

[0157] (5) Network Communication Layer;

[0158] 1) P2P (Peer-to-Peer) Communication Network: All nodes communicate directly through the P2P network, without relying on a central server, which conforms to the distributed characteristics of blockchain technology, enhancing decentralization and data privacy.

[0159] 2) Gossip Protocol: This protocol is used to quickly spread information and data, ensuring that nodes in the network can quickly obtain global information.

[0160] 3) Auxiliary Communication Facilities: Ensure communication and synchronization between nodes, and can enhance the stability and reliability of information transmission.

[0161] (6) Data Storage Layer;

[0162] 1) Transaction Data Blocks: Each upload and update of model parameters are recorded in the blockchain in the form of transactions, ensuring the immutability of information.

[0163] 2) Hash Chain Structure: The hash value of each data block is linked to the previous block, forming a chain structure, enhancing the security of data.

[0164] 3) Merkle Data Verification: The Merkle tree used for data verification helps ensure that the transmitted data is complete and accurate, and can efficiently verify the data in the blockchain.

[0165] (7) Federated Learning Layer;

[0166] 1) Distributed Model Training: The training process of the model is distributed. Each node uses local data for local training and shares model parameters to prevent data leakage.

[0167] 2) Adaptive Model Aggregation: According to the contributions and training quality of different nodes, use an adaptive method for model aggregation to optimize the final global model.

[0168] 3) Privacy requirement perception: When conducting model training, fully consider privacy protection requirements and ensure data privacy is not leaked through encryption technologies such as homomorphic encryption or differential privacy technologies.

[0169] 4) Node training incentive: Each node obtains rewards by providing high-quality training results, motivating the nodes to conduct efficient training.

[0170] 5) Joint training group construction: Each node is formed into a training group, and more effective training is carried out through in-group collaboration to improve the model quality.

[0171] 6) Model parameter auditing: Audit the uploaded model parameters to ensure the quality and credibility of the parameters for further optimizing the global model.

[0172] Through the mutual collaboration of each level of the above-mentioned federated learning framework, the security, transparency, and efficiency of the entire training process are ensured. At the same time, the addition of smart contracts, decentralized protocols, and privacy protection mechanisms can effectively reduce communication costs and improve training quality.

[0173] In the embodiment of this application, the overall implementation process of the traffic control model training method based on blockchain and federated learning is as follows:

[0174] Step 1, data preprocessing and participant authentication;

[0175] Obtain the local data of intelligent vehicles. By preprocessing the local data of intelligent vehicles, the noise in the data is removed, and traffic scene-related features are extracted. At the same time, use the blockchain system to authenticate the identity of each participating node to ensure that only legitimate devices participate in federated learning. Due to the immutability and consensus mechanism of the blockchain, the legitimacy of each participant is ensured, thus avoiding the addition of malicious nodes and enhancing the credibility of the overall system.

[0176] Step 2, local model training;

[0177] Each participating node conducts deep learning model training locally according to its own sensor data and uploads the trained model parameters to the blockchain network. By only transmitting the trained model parameters rather than the original data to the blockchain network, the protection of data privacy is achieved. Since the data does not leave the local device, the privacy information of users is protected, and at the same time, the security risks that may be brought by data concentration are avoided.

[0178] Step 3, blockchain recording and intermediate parameter storage;

[0179] The training results (model parameters and loss function) are recorded on the blockchain. The decentralized nature of the blockchain eliminates the dependence on traditional centralized servers. The intermediate parameters of each training process are stored in the immutable blocks of the blockchain, effectively avoiding the risks of data tampering and privacy leakage. At the same time, the contributions and training quality of each node are also recorded to ensure the transparency and traceability of the model.

[0180] Step 4: Model aggregation and update;

[0181] Use the decentralized nature of the blockchain for model aggregation. Automatically evaluate the training quality of each node through smart contracts to ensure that the global model updated in each round is based on qualified model parameters. This decentralized aggregation method improves the security and transparency of the training process. Among them, the existence of malicious nodes is identified by smart contracts and punished, ensuring the fairness and quality of the system; for nodes that submit high-quality model parameters, smart contracts will give rewards.

[0182] Step 5: Model update and distribution;

[0183] The aggregated global model is distributed to all participating nodes through the blockchain network. Each participating node conducts the next round of local training based on this global model, ensuring the transparency and security of the update process. All training records and logs of model updates are stored in the blockchain, ensuring a high degree of credibility of the training process. The immutability of the blockchain enables every detail in the training process to be traced and verified, preventing external malicious attacks.

[0184] In the specific implementation, the specific implementation process of the traffic control model training method based on blockchain and federated learning is as follows:

[0185] (1) Data collection and training; Each participant (i.e., intelligent vehicle) conducts training based on its local sensor data (such as speed, position, direction, etc. information). Each participant (i.e., intelligent vehicle) can obtain its corresponding training parameters (intermediate parameters) by training the model locally. These intermediate parameters usually include network weights, bias values, and loss function values, etc. In this process, the training data itself does not leave the vehicle, but encrypts the training results through encryption methods (such as homomorphic encryption or differential privacy) to ensure the protection of user data privacy.

[0186] (2) Generate a block and upload intermediate parameters; After each participant (i.e., intelligent vehicle) completes local training, it will generate a training result (such as model parameters, loss function values, etc.) and package it into a "training result block", which will be uploaded to the network for storage through the blockchain network. Each block will contain the following content:

[0187] 1) Training results: such as the parameters of the model (e.g., weights and biases) and the calculated loss function;

[0188] 2) Verification information: including the ID (Identification) of the participant to ensure legality, signature, timestamp, etc., which are used to verify the source of the submitted results;

[0189] 3) Smart contract verification: Each uploaded block will trigger a smart contract on the blockchain. The smart contract will perform a preliminary verification on the uploaded training results to check whether they meet specific format and legality requirements.

[0190] Exemplarily, assume that vehicle A uploads its training result block, which includes the weights, biases, and loss value of the trained model, along with the digital signature of the in-vehicle device. The smart contract will first verify whether these data are valid. After passing the verification, it will generate a "legal" tag and store it on the blockchain.

[0191] (3) Aggregate intermediate parameters: Once multiple vehicles have submitted their corresponding training results, the system will aggregate these training results for the model. The aggregation is carried out based on the decentralized characteristics of the blockchain, rather than through a centralized parameter server.

[0192] Decentralized aggregation: The model parameters (e.g., weights, biases) included in each uploaded block will be verified by other nodes (such as vehicle B, vehicle C) and weighted averaged or other aggregation methods will be used. The specific aggregation method can be defined through a smart contract:

[0193] 1) Weighted average: Different weights are given for aggregation according to the quality of training of each node (such as the low or high of the loss function, the good or bad of the training effect).

[0194] 2) Other aggregation methods: such as adaptive aggregation algorithms, weighted aggregation based on node reputation, etc.

[0195] Exemplarily, assume that the weights uploaded by vehicles A, B, and C are W A 、W B 、W C , and after verifying their validity through a smart contract, the aggregated weight is W agg =α*W A +β*W B +γ*W C , where α, β, and γ are weight coefficients based on the contribution of each vehicle model.

[0196] (4) Verification and checking:

[0197] 1) Verify intermediate parameters: To ensure that the uploaded intermediate parameters (such as model weights) have not been tampered with, the immutability feature of the blockchain enables each submitted block and training result to have a digital signature, and the integrity of the block data is ensured through a hash chain. Each submitted training result needs to be verified by multiple nodes in the blockchain network to ensure that the submitted result has not been maliciously tampered with;

[0198] 2) Verification mechanism: The verification process is automatically executed by a smart contract. The specific operations include:

[0199] Verify parameters: Verify whether the model parameters in the training result meet the predetermined standards (for example, whether the loss function is low enough and whether the model parameters are within a certain range);

[0200] Verify the legality of the submitted training result: Check whether the participant's identity is legal (for example, verify using the vehicle's public key) to prevent malicious nodes from submitting false results;

[0201] Verify integrity: Through the blockchain's hash link, ensure that the submitted block has not been tampered with. If the data of a certain block changes, the hash value of the subsequent block will change, causing the entire chain to become invalid, thus enabling illegal operations to be detected in a timely manner.

[0202] Exemplarily, assume that the training parameters submitted by vehicle B fail the verification process of the smart contract (such as too high a loss function or an unqualified model). The system will reject the addition of this block, and the penalty mechanism will handle vehicle B accordingly (such as reducing the reputation score, suspending participation in training, etc.).

[0203] (5) Aggregated global model: After aggregation, the global parameters of the model (i.e., the weighted average of the training results of all participants) will be generated and distributed to all participants through the blockchain network.

[0204] Transparency and auditing: Each node can query its own contribution, as well as the contributions of other participants and the parameter update log of the entire training process, through the blockchain. The transparency and immutability of the blockchain ensure the fairness and traceability of all model updates and training processes.

[0205] Exemplarily, the global model parameters will be distributed to all participating vehicles through the blockchain. After each node receives the global model, it will use it as the initialization parameter to start the next round of local training. The process of global model update, the contribution of each node, and the reason for model update will be stored in the blockchain log and can be audited at any time.

[0206] Through the above process, while ensuring data privacy, data integrity, and preventing tampering, the blockchain can decentralize the aggregation and verification of the intermediate parameters generated during the training process, ensuring the transparency, security, and reliability of traffic scenario model training in the intelligent vehicle network.

[0207] In the embodiments of this application, based on the blockchain-enabled federated learning framework, the centralized parameter server of federated learning is constructed into a decentralized parameter aggregation chain. The blockchain is used to record the intermediate parameters of the traffic scenario model training process as evidence, and to incentivize collaborative nodes to verify the model parameters, punishing the participating nodes that upload false parameters or low-quality models to constrain their self-interest. In addition, taking the model quality as the evaluation basis, the dynamic adjustment of intermediate parameter privacy noise and adaptive model aggregation are realized to protect user privacy and resist potential vehicle network attacks. By replacing the traditional centralized parameter server with a decentralized blockchain network, the risks of data privacy leakage and centralized systems in traditional model training are avoided, improving the security, reliability of the system, and the efficiency of model training.

[0208] It is worth noting that although the solution proposed in the embodiments of this application is a decentralized one, in practical applications, it is still possible to choose to centrally transmit data to a trusted central server for model aggregation and protect data privacy through encryption means. However, the solution of centrally transmitting data to a trusted central server for model aggregation is vulnerable to single-point failures and malicious attacks, and data privacy is difficult to be fully guaranteed. Additionally, other distributed storage technologies, such as IPFS (InterPlanetary File System), can be used to replace the blockchain for the storage and distribution of model parameters. However, the decentralization and immutability of the blockchain proposed in the embodiments of this application can better ensure the security of the system compared to these distributed storage technologies. Therefore, in the design of the federated learning framework provided in the embodiments of this application, the blockchain network, smart contract, and encryption algorithm are necessary structures, while the central server is an optional structure. Specifically, the blockchain network is used to store the intermediate parameters during the training process and provide decentralized support for model aggregation; the smart contract is used to evaluate the quality of the model parameters provided by the participating parties and reward or punish the nodes according to the quality; the encryption algorithm is used to protect data privacy and ensure the security of data during the training process. However, in some scenarios with low requirements, a traditional centralized server can be selected for model aggregation, but the security of the server needs to be ensured.

[0209] It should be noted that this embodiment only briefly and schematically illustrates the general process of the traffic control model training method based on blockchain and federated learning. The detailed description of each step can refer to the relevant content in the foregoing embodiments and will not be elaborated here. It can be understood that the present invention is not limited thereto.

[0210] In the embodiment of the present application, the to-be-trained in-vehicle sensor data corresponding to a number of intelligent vehicles in the vehicle networking system is obtained; the identity authentication of each intelligent vehicle is performed through the blockchain network to determine a number of target intelligent vehicles participating in federated learning; each target intelligent vehicle trains the local traffic control model according to the corresponding to-be-trained in-vehicle sensor data, obtains a number of local model training results, and sends each local model training result to the blockchain network; the intelligent contract in the blockchain network evaluates each local model training result to determine a number of qualified model training results; the blockchain network aggregates each qualified model training result to obtain a global model, and distributes the global model to each target intelligent vehicle, so that each target intelligent vehicle updates the local traffic control model according to the global model; the local traffic control model is used to realize the intelligent management and control of the traffic scenario. The embodiment of the present application adopts the federated learning method, enabling intelligent vehicles to train models locally without uploading the original data to the blockchain network, reducing the risk of user privacy leakage; adopting blockchain technology enables the vehicle networking system to eliminate the single-point failure risk brought by the traditional centralized solution, and the decentralized model aggregation method enables the training results of all participating nodes to be effectively verified and monitored, preventing malicious attacks and data tampering. At the same time, the automatic model quality evaluation is realized based on the blockchain intelligent contract to ensure that the aggregated global model has high quality, thereby improving the reliability of the traffic control system; by introducing the combination of blockchain technology and federated learning, the demand for a large amount of data transmission is reduced, the communication cost is lowered, and the security and reliability of the federated learning process in the vehicle networking system are improved; at the same time, the local traffic control model can be updated in real time according to the global model. Through the continuous update and optimization of the local traffic control model, intelligent vehicles can better adapt to complex traffic scenarios and achieve more accurate traffic management and control.

[0211] The key points of the embodiment of the present application are as follows:

[0212] (1) Decentralized model aggregation: By introducing blockchain technology, the traditional centralized parameter server is replaced by a decentralized blockchain ledger, eliminating the single-point failure risk and improving the security and reliability of the system.

[0213] (2) Intermediate parameter privacy protection: Encryption algorithms (such as homomorphic encryption, differential privacy, etc.) are used to protect the intermediate parameters generated during the training process to ensure that the data content cannot be leaked or tampered with even during the transmission process.

[0214] (3) Automatic evaluation mechanism of intelligent contract: Based on the blockchain intelligent contract, an automatic model quality evaluation and penalty mechanism are realized to ensure that each participating party provides a qualified model update, encouraging cooperation and punishing malicious nodes.

[0215] (4) Credibility and transparency: The immutability and transparency of the blockchain enable the real-time traceability and verification of the model training process, parameter updates, and node contributions, enhancing the trust and security of the system.

[0216] In summary, the technical effects of the embodiments of the present application include the following:

[0217] (1) Improving system security and reliability: After adopting blockchain technology, the system eliminates the single-point failure risk brought by traditional centralized solutions. Moreover, the immutability and transparency of the blockchain enable the traceability and verification of the contributions and training processes of each node. All updates during the training process are recorded in the blockchain in a transparent manner, enhancing the trust of the system. The decentralized model aggregation method enables the effective verification and monitoring of the training results of all nodes, preventing malicious attacks and data tampering; and each node and training process are recorded on the blockchain, ensuring the security of data and model updates.

[0218] (2) Enhancing data privacy protection: Protect the intermediate parameters generated during the training process through encryption technology and record these encrypted parameters on the blockchain, avoiding data leakage or tampering during transmission. Compared with traditional solutions, by only transmitting model parameters instead of data, the integrity of user privacy is ensured.

[0219] (3) Effectiveness of the incentive and punishment mechanism: The automated model quality assessment and incentive mechanism implemented through smart contracts ensure that each participating party provides high-quality model updates. Malicious nodes submitting low-quality models will be identified and punished, ensuring the fairness and effectiveness of the federated learning process.

[0220] (4) Reducing communication costs: Traditional centralized servers need to frequently transmit a large amount of raw data, while the embodiments of the present application reduce the need for large-scale data transmission through decentralized model training and parameter transmission. Each node only uploads model parameters, thus significantly reducing communication bandwidth and latency. Through this solution, the communication costs of the vehicle networking system can be reduced by approximately 30% - 50%.

[0221] (5) Improving model training efficiency: The decentralized model aggregation method not only reduces the dependence on the central server but also automatically evaluates and optimizes model updates through smart contracts, improving the efficiency of model training. Experiments show that the model training framework combining blockchain and federated learning can shorten the training time by 20% - 40% compared with traditional methods.

[0222] (5) Enhance the scalability of the system: Through blockchain technology and decentralized aggregation, each participating node in the vehicle network can independently train and participate in the update of the global model. The system can scale as the number of participating nodes increases, while also ensuring the efficient use of data privacy and computing resources of each node. As the number of nodes increases, the training efficiency and system scalability are gradually improved, and the theoretical scalability is stronger.

[0223] By combining blockchain technology with federated learning in the embodiments of this application, the technical problems in aspects such as data privacy, model training efficiency, security, and system scalability of traditional vehicle network systems are effectively solved. Through a decentralized design and encryption technology, the overall security, transparency, and trust of the system are improved. At the same time, the use of smart contracts and decentralized aggregation methods ensures the efficiency and fairness of the training process, and has significant technical advantages and practical value.

[0224] Please refer to Figure 3 , the embodiments of this application also provide a traffic control model training system 300 based on blockchain and federated learning, which can implement the above-mentioned traffic control model training method based on blockchain and federated learning. The system 300 includes the following modules:

[0225] The module 301 for obtaining in-vehicle sensor data to be trained is used to obtain the in-vehicle sensor data to be trained corresponding to several intelligent vehicles in the vehicle network system;

[0226] The intelligent vehicle identity authentication module 302 is used to authenticate the identity of each of the intelligent vehicles through the blockchain network to determine several target intelligent vehicles participating in federated learning;

[0227] The local traffic control model training module 303 is used for each of the target intelligent vehicles to train the local traffic control model according to the corresponding in-vehicle sensor data to be trained, obtain several local model training results, and send each of the local model training results to the blockchain network;

[0228] The local model training result evaluation module 304 is used to evaluate each of the local model training results through the smart contract in the blockchain network to determine several qualified model training results;

[0229] The global model aggregation module 305 is used to aggregate each of the qualified model training results through the blockchain network to obtain a global model, and distribute the global model to each of the target intelligent vehicles, so that each of the target intelligent vehicles updates the local traffic control model according to the global model; the local traffic control model is used to realize the intelligent management and control of traffic scenarios.

[0230] It can be understood that the content in the above method embodiments is applicable to the system embodiments of the present application. The functions specifically implemented by the system embodiments of the present application are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those of the above method embodiments.

[0231] An embodiment of the present application further provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, it implements the above traffic control model training method based on blockchain and federated learning. The electronic device can be any intelligent terminal including a tablet computer, an in-vehicle computer, etc.

[0232] It can be understood that the content in the above method embodiments is applicable to the device embodiments of the present application. The functions specifically implemented by the device embodiments of the present application are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those of the above method embodiments.

[0233] Please refer to Figure 4 , Figure 4 which schematically shows the hardware structure of an electronic device according to another embodiment. The electronic device includes:

[0234] A processor 401, which can be implemented in the form of a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (ASIC for short), or one or more integrated circuits, etc., and is used to execute relevant programs to implement the technical solutions provided by the embodiments of the present application;

[0235] A memory 402, which can be implemented in the form of a read-only memory (ROM for short), a static storage device, a dynamic storage device, or a random access memory (RAM for short). The memory 402 can store an operating system and other application programs. When implementing the technical solutions provided by the embodiments of this specification through software or firmware, the relevant program codes are stored in the memory 402, and the processor 401 is used to call and execute the traffic control model training method based on blockchain and federated learning of the embodiments of the present application;

[0236] An input / output interface 403, which is used to implement information input and output;

[0237] A communication interface 404, which is used to implement communication interaction between this device and other devices, and can implement communication through a wired method (such as USB, network cable, etc.) or through a wireless method (such as mobile network, WIFI, Bluetooth, etc.);

[0238] A bus 405 that transmits information among various components of the device, such as a processor 401, a memory 402, an input / output interface 403, and a communication interface 404;

[0239] Among them, the processor 401, the memory 402, the input / output interface 403, and the communication interface 404 are communicatively connected to each other inside the device through the bus 405.

[0240] The embodiment of the present application also provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the above-mentioned traffic control model training method based on blockchain and federated learning is implemented.

[0241] It can be understood that the content in the above method embodiments is applicable to the present storage medium embodiment, and the functions specifically implemented by the present storage medium embodiment are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those of the above method embodiments.

[0242] As a non-transitory computer-readable storage medium, the memory can be used to store non-transitory software programs and non-transitory computer-executable programs. In addition, the memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one magnetic disk storage device, a flash memory device, or other non-transitory solid-state storage devices. In some embodiments, the memory optionally includes a memory remotely disposed relative to the processor, and these remote memories can be connected to the processor through a network. Examples of the above networks include, but are not limited to, the Internet, an enterprise intranet, a local area network, a mobile communication network, and combinations thereof.

[0243] The traffic control model training method and system based on blockchain and federated learning provided by the embodiments of the present application obtain the to-be-trained in-vehicle sensor data corresponding to a number of intelligent vehicles in the vehicle networking system; authenticate the identities of each intelligent vehicle through the blockchain network to determine a number of target intelligent vehicles participating in federated learning; each target intelligent vehicle trains the local traffic control model according to the corresponding to-be-trained in-vehicle sensor data to obtain a number of local model training results, and sends each local model training result to the blockchain network; evaluate each local model training result through the intelligent contract in the blockchain network to determine a number of qualified model training results; aggregate each qualified model training result through the blockchain network to obtain a global model, and distribute the global model to each target intelligent vehicle, so that each target intelligent vehicle updates the local traffic control model according to the global model; the local traffic control model is used to implement intelligent management and control of traffic scenarios. The embodiments of the present application adopt the federated learning method, enabling intelligent vehicles to train models locally without uploading raw data to the blockchain network, reducing the risk of user privacy leakage; adopting blockchain technology enables the vehicle networking system to eliminate the single point of failure risk brought by traditional centralized solutions. The decentralized model aggregation method enables the training results of all participating nodes to be effectively verified and monitored, preventing malicious attacks and data tampering. At the same time, automated model quality evaluation is realized based on the blockchain intelligent contract to ensure that the aggregated global model has high quality, thereby improving the reliability of the traffic control system; by introducing the combination of blockchain technology and federated learning, the need for a large amount of data transmission is reduced, the communication cost is lowered, and the security and reliability of the federated learning process in the vehicle networking system are improved; at the same time, the local traffic control model can be updated in real time according to the global model. Through the continuous update and optimization of the local traffic control model, intelligent vehicles can better adapt to complex traffic scenarios and achieve more accurate traffic management and control.

[0244] The embodiments described in the embodiments of the present application are for more clearly explaining the technical solutions of the embodiments of the present application, and do not constitute a limitation on the technical solutions provided by the embodiments of the present application. Those skilled in the art know that with the evolution of technology and the emergence of new application scenarios, the technical solutions provided by the embodiments of the present application are equally applicable to similar technical problems.

[0245] Those skilled in the art can understand that the technical solutions shown in the figures do not constitute a limitation on the embodiments of the present application, and may include more or fewer steps than those shown in the figures, or combine certain steps, or different steps.

[0246] The system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0247] Those of ordinary skill in the art can understand that all or some of the steps in the methods disclosed above, and the functional modules / units in the systems and devices, can be implemented as software, firmware, hardware, and their appropriate combinations.

[0248] In several embodiments provided in this application, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the system embodiments described above are merely illustrative. For example, the division of the above units is only a logical function division, and there may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces, and the indirect couplings or communication connections of the systems or units can be electrical, mechanical, or other forms.

[0249] The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0250] The preferred embodiments of the embodiments of this application have been described above with reference to the drawings, but this does not limit the scope of the rights of the embodiments of this application. Any modifications, equivalent replacements, and improvements made by those skilled in the art without departing from the scope and essence of the embodiments of this application shall be within the scope of the rights of the embodiments of this application.

Claims

1. A method for training a traffic control model based on blockchain and federated learning, characterized in that The method includes the following steps: Obtain the to-be-trained vehicle-mounted sensor data corresponding to a number of intelligent vehicles in the vehicle networking system; Authenticate the identity of each of the intelligent vehicles through the blockchain network to determine a number of target intelligent vehicles participating in federated learning; Each of the target intelligent vehicles trains a local traffic control model according to the corresponding to-be-trained vehicle-mounted sensor data, obtains a number of local model training results, and sends each of the local model training results to the blockchain network; Evaluate each of the local model training results through a smart contract in the blockchain network to determine a number of qualified model training results; Aggregate each of the qualified model training results through the blockchain network to obtain a global model, and distribute the global model to each of the target intelligent vehicles, so that each of the target intelligent vehicles updates the local traffic control model according to the global model; the local traffic control model is used to implement intelligent management and control of traffic scenarios.

2. The method according to claim 1, wherein Before obtaining the to-be-trained vehicle-mounted sensor data corresponding to a number of intelligent vehicles in the vehicle networking system, the method further includes: Obtain the local vehicle-mounted sensor data corresponding to a number of the intelligent vehicles in the vehicle networking system; Perform encryption processing on each of the local vehicle-mounted sensor data by using a target encryption algorithm, and perform preprocessing on each of the encrypted local vehicle-mounted sensor data to obtain a number of the to-be-trained vehicle-mounted sensor data.

3. The method according to claim 2, characterized in that, The performing encryption processing on each of the local vehicle-mounted sensor data by using a target encryption algorithm, and performing preprocessing on each of the encrypted local vehicle-mounted sensor data to obtain a number of the to-be-trained vehicle-mounted sensor data includes: Perform encryption processing on each of the local vehicle-mounted sensor data by using the target encryption algorithm, and perform denoising processing on each of the encrypted local vehicle-mounted sensor data respectively to obtain a number of local clean sensor data; Perform feature extraction processing on each of the local clean sensor data respectively to obtain a number of local initial traffic feature data; Perform data fusion processing on each of the local initial traffic feature data respectively to obtain a number of local multi-source traffic feature data; Perform deep learning on each of the local multi-source traffic feature data to extract a number of local multi-source key traffic feature data; Perform selection and dimensionality reduction processing on each of the local multi-source key traffic feature data respectively to obtain a number of the to-be-trained vehicle-mounted sensor data.

4. The method according to claim 1, wherein The method further includes: Perform encryption processing on the intermediate parameters generated during the training of the local traffic control model by using a target encryption algorithm.

5. The method according to claim 1, wherein The evaluating each of the local model training results through a smart contract in the blockchain network to determine a number of qualified model training results includes: Evaluate the training quality of each of the local model training results through the smart contract in the blockchain network to determine whether each of the local model training results meets a preset training quality standard; Perform compliance checks on each of the local model training results through the smart contract to determine whether each of the local model training results meets the preset compliance standards; Use the local model training results that meet the preset training quality standards and the preset compliance standards as the qualified model training results.

6. The method according to claim 5, characterized in that, The method further includes: Provide incentives to the target intelligent vehicles corresponding to the local model training results that meet the preset training quality standards; Impose penalties on the target intelligent vehicles corresponding to the local model training results that do not meet the preset training quality standards.

7. The method according to claim 1, wherein The aggregating, through the blockchain network, each of the qualified model training results to obtain a global model, and distributing the global model to each of the target intelligent vehicles, so that each of the target intelligent vehicles updates the local traffic control model according to the global model, includes: Perform an aggregation process on each of the qualified model training results through the blockchain network using the decentralized feature to obtain the global model; Distribute the global model to each of the target intelligent vehicles through the blockchain network, so that each of the target intelligent vehicles trains the corresponding local traffic control model according to the global model parameters in the global model to obtain the updated local traffic control model.

8. A traffic control model training system based on blockchain and federated learning, characterized in that, The system includes the following modules: A module for obtaining in-vehicle sensor data to be trained, configured to obtain the in-vehicle sensor data to be trained corresponding to a plurality of intelligent vehicles in the vehicle networking system; An intelligent vehicle identity authentication module, configured to authenticate the identity of each of the intelligent vehicles through the blockchain network to determine a plurality of target intelligent vehicles participating in federated learning; A local traffic control model training module, configured to each of the target intelligent vehicles train a local traffic control model according to the corresponding in-vehicle sensor data to be trained, obtain a plurality of local model training results, and send each of the local model training results to the blockchain network; A local model training result evaluation module, configured to evaluate each of the local model training results through a smart contract in the blockchain network to determine a plurality of qualified model training results; A global model aggregation module, configured to aggregate each of the qualified model training results through the blockchain network to obtain a global model, and distribute the global model to each of the target intelligent vehicles, so that each of the target intelligent vehicles updates the local traffic control model according to the global model; the local traffic control model is used to implement intelligent management and control of traffic scenarios.

9. An electronic device, characterized in that, The electronic device includes a memory and a processor, the memory stores a computer program, and when the processor executes the computer program, the method according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, the method according to any one of claims 1 to 7 is implemented.

Images