A Cross-Domain Database Collaborative Access Method Based on Multi-Scale Decoupling and Atomized Encapsulation in IVCPS

Abstract

This invention belongs to the field of IVCPS and database technology, and discloses an IVCPS cross-domain database collaborative access method based on multi-scale decoupling and atomic encapsulation. The method comprises seven steps: data access and cleaning, multi-scale decoupling, atomic encapsulation, layered database entry, task reconstruction, lock-free extraction and scheduling, and self-healing maintenance. It constructs a vehicle-road-network multi-scale orthogonal decoupling model, encapsulates state vectors into five-tuple atomic components and routes them to edge / regional / central cloud layered databases. Cross-domain requests are decomposed into lock-free snapshot extraction instructions through a reconstruction matrix, and data closure is achieved by combining load scheduling and lifecycle self-healing. This invention eliminates cross-layer coupling redundancy, solves memory backlog and access deadlock problems, reduces business response latency, achieves multi-tenant data isolation and database data consistency, adapts to distributed computing architectures, and significantly improves the robustness, throughput, and real-time performance of the IVCPS cloud control platform, supporting the commercialization of vehicle-road-cloud integration.

Description

Technical Field

[0001] This invention belongs to the field of intelligent vehicle cyber-physical systems and database technology, specifically relating to an IVCPS cross-domain database collaborative access method based on multi-scale decoupling and atomic encapsulation. Background Technology

[0002] With the deepening development of Intelligent Vehicle CyberPhysical Systems (IVCPS), a "vehicle-road-cloud integrated" system, the cloud control platform needs to simultaneously support the micro-control of autonomous vehicles, the meso-level coordinated management and control of intersection signals, and the macro-level scheduling of city-level road networks. In this system, massive amounts of multi-source heterogeneous sensing data flow frequently between edge clouds, regional clouds, and central clouds, creating extremely complex cross-domain, multi-tenant concurrent data access requirements. However, existing distributed relational or time-series databases, when supporting the high-dimensional, strongly coupled applications of IVCPS, reveal a fundamental flaw: a severe disconnect between the underlying data architecture and physical traffic characteristics. Specifically:

[0003] (1) The physical organization of data suffers from "scale collapse", making it difficult to support cross-level collaborative computing;

[0004] Real-world physical traffic systems possess strict, orthogonal spatial-temporal scale attributes (i.e., high-frequency microscopic vehicle scale, mid-frequency mesoscopic intersection scale, and low-frequency macroscopic road network scale). Existing database technologies typically employ "flat" data table structures to store heterogeneous data, or perform coarse-grained database construction based solely on data source type (such as radar flow tables or V2X signaling tables). This storage method, which ignores physical scale, directly leads to "scale collapse." When cross-domain services (such as "global green wave speed dynamic guidance") need to simultaneously access microscopic vehicle instantaneous speeds and macroscopic global congestion states, the underlying database engine must perform extremely complex cascading queries (Joins) across data tables with vastly different sampling frequencies. This can easily trigger system I / O bottlenecks and global deadlocks, completely failing to meet the requirements of IVCPS for hard real-time safety control.

[0005] (2) Simple "service decoupling" and "information decoupling" fall into a technical deadlock;

[0006] Currently, the IT industry commonly uses microservice architecture for conventional "service decoupling." However, in IVCPS, there is strong physical coupling between traffic elements. Simply splitting microservices at the software interface level still involves calling the highly coupled physical traffic data tables at the underlying level, making frequent API calls between microservices prone to distributed transaction deadlocks. Furthermore, attempting pure "information decoupling" at the bitstream level is extremely difficult due to the nonlinearity and strong randomness of the traffic environment, making it impossible to establish a unified model that adapts to the dynamic evolution of traffic business logic.

[0007] (3) Cross-domain concurrent task calls lack object-oriented database encapsulation and reconstruction mechanisms;

[0008] Different business domains (such as autonomous vehicles and traffic police control systems) have drastically different view requirements for data in the same physical space. Existing databases lack the "atomic" capability to decouple, encapsulate, and finally reconstruct complex traffic needs. This forces each business domain to operate independently, with "one set of data stored separately and repeated collection on demand," resulting in severe siloed data islands. Faced with high-concurrency cross-domain access, the lack of a mechanism to break down large-granularity business processes into fine-grained components leads to frequent resource contention and priority inversion.

[0009] Existing technologies cannot solve the aforementioned pain points of cross-domain database collaborative access in IVCPS. There is an urgent need to design new underlying database processing methods adapted to the characteristics of IVCPS from the perspective of physical transportation, so as to achieve efficient collaborative access across multiple business domains. Summary of the Invention

[0010] In view of this, the purpose of this invention is to provide an IVCPS cross-domain database collaborative access method based on multi-scale decoupling and atomic encapsulation. This method aims to solve problems such as data storage scale collapse, deadlock in cross-domain access services and information decoupling technology, and lack of atomic encapsulation and reconstruction mechanisms for high-concurrency cross-domain tasks caused by the disconnect between the underlying data architecture and the spatiotemporal scale characteristics of physical traffic when supporting vehicle-road-cloud integrated IVCPS applications. It aims to overcome I / O bottlenecks and global deadlock problems in cross-level collaborative computing, eliminate resource contention, priority inversion, and data silos in multi-tenant cross-domain access, and achieve efficient cross-domain concurrent access and collaborative processing of multi-source heterogeneous perception data in IVCPS. This meets the hard real-time safety control requirements of autonomous driving micro-control, intersection meso-level management, and road network macro-scheduling, adapts to the distributed hierarchical computing architecture of vehicle-road-cloud integration, and improves the robustness, throughput, and response efficiency of the IVCPS cloud control platform.

[0011] To achieve the above objectives, the present invention adopts the following technical solution:

[0012] A cross-domain database collaborative access method based on multi-scale decoupling and atomic encapsulation in IVCPS includes the following steps:

[0013] S1. Heterogeneous data concurrent access and cleaning: The message gateway of the cloud control base concurrently receives heterogeneous bit streams uploaded by connected vehicle OBU and roadside infrastructure RSU, and performs spatial timestamp alignment and noise reduction processing on the heterogeneous bit streams.

[0014] S2. Multi-scale physical state decoupling: Construct a multi-scale orthogonal projection operator S to decouple the mixed data stream processed in step S1 into microscopic vehicle kinematics state vector, mesoscopic intersection traffic flow and signal control state vector, and macroscopic road network topology state vector according to the physical range and time refresh frequency.

[0015] S3. Database Atomized Component Encapsulation: Inject the three types of state vectors obtained from decoupling in step S2 into a standardized quintuple template to generate an atomic data component (ADC). The structure of a single ADC is as follows: ;

[0016] Wherein, UID is a globally unique hash primary key with a spatial geocoding prefix; Enumerate labels for scale; The core state data payload; This is the permission mask for role-based access control. The dynamic lifetime decay factor represents the remaining lifetime value of the atomic data component ADC at time t.

[0017] S4. Hierarchical Database Classification and Input: The database routing engine parses the scale labels of the ADCs and distributes the micro, meso, and macro-scale ADCs to the physically isolated edge cloud vehicle scale database (DB) respectively. V Regional Cloud Intersection Scale Database DB I Central Cloud Road Network Scale Database DB N Storage is completed in the middle;

[0018] S5. Task Decoupling and Concurrency Mapping: Intercepting Cross-Domain Business Requests Initiated by Multiple Business Domains Construct a decoupling task reconstruction matrix The cross-domain service request Dimensionality reduction and decomposition for DB V DB I DB N List of concurrent ADC extraction instructions;

[0019] S6. Lock-free snapshot extraction and resource scheduling: Based on load-sensitive dynamic resource quota scheduling function. Resource channels are allocated for each extraction instruction, lock-free snapshot reading based on multi-version concurrency control is performed in databases at all levels, the extracted ADCs are aggregated and assembled into a real-time service view in application layer memory, and control instructions are issued based on the real-time service view.

[0020] S7. Lifecycle self-healing and physical consistency maintenance: Continuously monitor the ADC's lifecycle decay factor. For ADCs whose lifecycle approaches zero, memory reclamation and archiving operations are automatically triggered. When the data in the edge cloud and the central cloud database are physically offset, the divergence self-healing mechanism is triggered to complete the precise synchronization of data. The feedback data from the physical world re-enters step S1, forming a data processing closed loop.

[0021] Furthermore, step S2 includes the following sub-steps:

[0022] S2.1 Constructing the multi-scale orthogonal projection operator S;

[0023]

[0024] In the formula, This represents the scale category label that the i-th data record is ultimately assigned, with a value of [value]. One of them corresponds to the micro-scale of vehicles, the meso-scale of intersections, and the macro-scale of road networks, respectively. The preset scale feature mapping coefficients are calibrated offline based on the communication bandwidth of the IVCPS system and the physical sensor deployment density. The spatial coverage feature of the i-th data point at scale S; Let S be the time update frequency of the i-th data point at scale S; Let S be the sampling precision of the i-th data point at scale S;

[0025] S2.2 Based on the multi-scale orthogonal projection operator S calculated in step S2.1, the mixed data stream processed in step S1 is decoupled into microscopic vehicle kinematics state vector, mesoscopic intersection traffic flow and signal control state vector and macroscopic road network topology state vector according to the physical range of action and time refresh frequency.

[0026] I. Microscopic vehicle kinematic state vector:

[0027]

[0028] In the formula, Let be the microscopic vehicle kinematic state vector at time t; Let be the vehicle's real-time lateral position coordinates at time t; Here are the real-time longitudinal coordinates of the vehicle at time t; Let be the vehicle's real-time speed at time t; Let be the real-time acceleration of the vehicle at time t; Let be the vehicle's real-time heading angle at time t; This refers to the real-time operating mode of the vehicle at time t.

[0029] II. Traffic Flow and Signal Control State Vectors at Intersections (Medium Level):

[0030]

[0031] In the formula, Let be the mesoscopic traffic flow and signal control state vector of the j-th intersection at time t; Let be the cross-sectional flow rate at the j-th intersection; Let the spatial density be the value at the j-th intersection. Let J be the average vehicle speed at the j-th intersection; The multi-phase timing state matrix for the j-th intersection; Let be the queue length at the j-th intersection;

[0032] III. Macro-level road network topology state vector:

[0033]

[0034] In the formula, Let be the macroscopic road network topology state vector at time t; Represents graph structure functions; The road network segment connection weight matrix; This represents the distribution vector of the road network congestion index. A macro-level travel demand matrix for the road network;

[0035] Core state data payload State vector after multi-scale decoupling , , The direct carrier.

[0036] Furthermore, in step S3, the dynamic lifetime decay factor The decay function is:

[0037]

[0038] In the formula, This represents the initial lifetime value of the ADC. For ADC generation time; The decay constant is a scale-dependent property.

[0039] Furthermore, in step S4:

[0040] The vehicle scale database DB V It is a memory-based key-value architecture used to store microscale ADCs that are generated and destroyed at high frequencies;

[0041] The intersection-scale database DB I It is a time-series database used to store mesoscale ADCs to support traffic control services in intersection areas;

[0042] The road network scale database DB NIt serves as a graph database for storing macro-scale ADCs to support global topology calculations for city-level road networks.

[0043] Furthermore, in step S5, cross-domain service requests The reconstruction mapping relationship is as follows:

[0044]

[0045] In the formula, These are the survival ADC vectors at the micro, meso, and macro scales, respectively. The activation matrix is ​​extracted as a Boolean value, which is the decoupling task reconstruction matrix; This represents the k-th surviving ADC component, from... Any scale set in the middle; ,matrix Chinese correspondence Element; This indicates that the current business request needs to read this ADC component. This triggers a lock-free snapshot read; This indicates that the current business request does not require the ADC component and should skip reading it.

[0046] Furthermore, in step S6, the load-sensitive dynamic scheduling resource quota function The calculation method is as follows:

[0047]

[0048] In the formula, Priority coefficients based on business security attributes; This refers to the actual number of ADCs at each scale invoked during task reconstruction. The physical data scale weights for each scale of the ADC; K represents the total I / O and computing resources of the system; K represents the total number of business domains that initiated cross-domain business requests.

[0049] Beneficial effects:

[0050] 1. Eradicate the "scale collapse" problem in data storage and significantly reduce system overhead;

[0051] This invention pioneers a multi-scale orthogonal decoupling model for "vehicle-intersection-road network" based on physical spatiotemporal characteristics. This model aligns the database data organization skeleton with the topological logic of the real physical traffic space at the pixel level, abandoning the traditional, crude large-table stacking model and eliminating cross-layer coupling redundancy at its source. Cross-layer collaborative computation eliminates the need for complex multi-table cascading queries, reducing the redundancy filtering overhead of cross-layer queries by nearly an order of magnitude. This effectively solves system I / O bottlenecks and global deadlock problems, meeting the core requirements of IVCPS for hard real-time safety control.

[0052] 2. Achieve precise "metabolism" of high-concurrency memory resources to improve system robustness;

[0053] This invention designs a database atomic component (ADC) encapsulation technology with lifecycle and access permission constraints, embedding a dynamic lifecycle decay factor into the component layer, enabling the underlying database engine to have data "shelf life" awareness for the first time. Extremely high-frequency micro-traffic data is naturally degraded and destroyed throughout its lifecycle, completely solving the persistent problem of memory backlog in high-concurrency environments. This ensures high throughput and strong robustness of the cloud control platform under extreme concurrency scenarios such as morning and evening rush hours, avoiding resource contention and priority inversion issues.

[0054] 3. Break the deadlock of cross-domain concurrent access and minimize business response latency;

[0055] This invention constructs a decoupled task reconstruction matrix, dynamically reducing complex cross-domain high-order business requirements into lock-free snapshot extraction instructions for atomic components of different scales. It transforms the global lock contention in multi-table joins, which is prone to deadlocks in traditional relational databases, into high-speed, stateless I / O reads. This significantly improves the efficiency of cross-domain concurrent access by compressing response latency from hundreds of milliseconds to microseconds when handling complex cross-domain businesses such as "multi-vehicle cluster collaborative passage" and "emergency vehicle green wave passage."

[0056] 4. Break down data silos and industry integration barriers to empower the commercialization of vehicle-road-cloud integration;

[0057] The atomic component encapsulation mechanism of this invention endows data with independent physical meaning and permission isolation capabilities, supporting logical data isolation for multiple tenants (police, car manufacturers, traffic management departments, etc.), and solving the siloed data island problem of "one data, one storage, and repeated collection on demand" in the traditional model. At the same time, the layered database encapsulation system is naturally adapted to the distributed hierarchical computing power architecture of cloud-network convergence, allowing the business logic of macro-level road network scheduling, meso-level intersection control, and micro-level vehicle control to coexist harmoniously in the same underlying data pool through the free assembly of components of different scales. This breaks down the data barriers in the intelligent transportation industry, provides underlying data support for the complex multi-objective collaboration of vehicle-road-cloud integration, and promotes its commercialization.

[0058] 5. Achieve intelligent management and physical consistency maintenance throughout the entire data lifecycle to ensure the closed-loop reliability of the system;

[0059] This invention achieves automatic recycling and archiving of atomic components through a lifecycle decay mechanism, and combines it with a divergence self-healing mechanism to achieve precise data synchronization between edge cloud and central cloud databases, constructing a fully closed-loop processing chain of "data access-decoupling-encapsulation-storage-calling-self-healing". This not only realizes refined management of data resources, but also ensures the physical data consistency of the distributed database cluster, allowing the IVCPS cloud control platform to maintain a stable and reliable operating state even in high-frequency data flow scenarios.

[0060] 6. It has good scalability and adaptability, and is compatible with multiple business scenarios of intelligent transportation;

[0061] The multi-scale decoupling model, atomized component definition, and task reconfiguration mechanism of this invention all adopt standardized and modular designs. Core parameters such as scale feature mapping coefficients, lifecycle decay constants, and resource scheduling weights can be flexibly calibrated according to the communication bandwidth, sensor deployment density, and business security priority of the IVCPS system. This invention's method is not only applicable to core scenarios such as autonomous vehicle control, intersection signal coordination, and urban road network scheduling, but can also be extended to all intelligent transportation business scenarios such as traffic emergency response, road network traffic optimization, and multi-vehicle cooperative passage, possessing broad engineering application value.

[0062] Other advantages, objectives, and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination, or may be learned from practice of the invention. The objectives and other advantages of the invention can be realized and obtained through the following description. Attached Figure Description

[0063] Figure 1 This is a flowchart of an IVCPS cross-domain database collaborative access method based on multi-scale decoupling and atomic encapsulation according to the present invention. Detailed Implementation

[0064] To make the technical solutions, advantages, and objectives of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the described embodiments of the present invention without creative effort are within the protection scope of this application.

[0065] like Figure 1 As shown, this invention provides a cross-domain database collaborative access method for IVCPS based on multi-scale decoupling and atomic encapsulation, specifically including the following steps:

[0066] S1. Heterogeneous data concurrent access and cleaning: The message gateway of the cloud control base concurrently receives heterogeneous bit streams uploaded by connected vehicle OBU and roadside infrastructure RSU, and performs spatial timestamp alignment and noise reduction processing on the heterogeneous bit streams.

[0067] S2. Multi-scale physical state decoupling: Construct a multi-scale orthogonal projection operator S to decouple the mixed data stream processed in step S1 into microscopic vehicle kinematics state vector, mesoscopic intersection traffic flow and signal control state vector, and macroscopic road network topology state vector according to the physical range and time refresh frequency.

[0068] S3. Database Atomized Component Encapsulation: Inject the three types of state vectors obtained from decoupling in step S2 into a standardized quintuple template to generate an atomic data component (ADC). The structure of a single ADC is as follows: ;

[0069] Wherein, UID is a globally unique hash primary key with a spatial geocoding prefix; Enumerate labels for scale; The core state data payload; This is the permission mask for role-based access control. The dynamic lifetime decay factor represents the remaining lifetime value of the atomic data component ADC at time t.

[0070] S4. Hierarchical Database Classification and Input: The database routing engine parses the scale labels of the ADCs and distributes the micro, meso, and macro-scale ADCs to the physically isolated edge cloud vehicle scale database (DB) respectively. V Regional Cloud Intersection Scale Database DB I Central Cloud Road Network Scale Database DB N Storage is completed in the middle;

[0071] S5. Task Decoupling and Concurrency Mapping: Intercepting Cross-Domain Business Requests Initiated by Multiple Business Domains Construct a decoupling task reconstruction matrix The cross-domain service request Dimensionality reduction and decomposition for DB V DB I DB N List of concurrent ADC extraction instructions;

[0072] S6. Lock-free snapshot extraction and resource scheduling: Based on load-sensitive dynamic resource quota scheduling function. Resource channels are allocated for each extraction instruction, lock-free snapshot reading based on multi-version concurrency control is performed in databases at all levels, the extracted ADCs are aggregated and assembled into a real-time service view in application layer memory, and control instructions are issued based on the real-time service view.

[0073] S7. Lifecycle self-healing and physical consistency maintenance: Continuously monitor the ADC's lifecycle decay factor. For ADCs whose lifecycle approaches zero, memory reclamation and archiving operations are automatically triggered. When the data in the edge cloud and the central cloud database are physically offset, the divergence self-healing mechanism is triggered to complete the precise synchronization of data. The feedback data from the physical world re-enters step S1, forming a data processing closed loop.

[0074] Scheme Principle

[0075] 1. Design of IVCPS Cross-Domain Database Collaborative Architecture Based on Multi-Scale Decoupling

[0076] The underlying architecture design of this invention strictly follows the systems engineering principle that "physical characteristics determine data structure," and the overall collaborative processing link is divided into four logical evolution layers:

[0077] (1) Multi-scale feature decoupling layer: At the data access gateway, the scale mapping operator is used to forcibly separate the mixed data stream uploaded by the vehicle end and the road end into a three-dimensional orthogonal dataset of micro (vehicle), meso (intersection) and macro (road network) according to the physical scope and time refresh frequency.

[0078] (2) Atomized Component Encapsulation Layer: The decoupled dataset is injected into a standardized quintuple template to generate mutually independent atomic data components (ADCs), providing the system with the smallest, standard concurrent read and write unit.

[0079] (3) Layered database encapsulation layer: Based on the component's scale label, the atomic components are routed to distributed database clusters (edge ​​memory database, regional time series database, central graph database) that match the physical location and computing power characteristics.

[0080] (4) Decoupling Task Reconstruction Layer: Serves as a gearbox for cross-domain business calls. Using the reconstruction matrix, complex cross-domain instructions passed from the outside are "broken down" and mapped into lightweight tasks that concurrently extract atomic components at various scales, which are then assembled in memory and output.

[0081] 2. Spatial-temporal multi-scale orthogonal decoupling mathematical modeling of transportation systems

[0082] To achieve a precise mapping from the physical world to database storage, this module establishes a rigorous mathematical model and decouples the entire traffic state space by scale orthogonalization.

[0083] Define the set of original discrete heterogeneous sensor data frames in the input database as follows: (in For data sampling accuracy, For spatial coverage characteristics, (Update frequency over time). Construct a multi-scale orthogonal projection operator. Calculate its classification scale level determination function:

[0084]

[0085] in, These are preset scale feature mapping coefficients, whose values ​​are determined offline by the communication bandwidth of the IVCPS system and the physical sensor deployment density. The spatial coverage feature of the i-th data point at scale S; Let S be the time update frequency of the i-th data point at scale S; Let be the sampling precision of the i-th data point at scale S.

[0086] Based on the judgment results, the highly coupled global traffic state subspace will be... Strictly orthogonal decoupling results in three independent low-dimensional manifold spaces:

[0087]

[0088] Where: vehicle-scale service state vector (Microscopic level / millisecond level): Focusing on the extreme kinematic characteristics of individual vehicles.

[0089]

[0090] Intersection-scale service state vector (Mesoscale / second-level): By performing time-window integration and statistical dimensionality reduction on data within the microscale space, we extract the characteristics of traffic flow and traffic control.

[0091]

[0092] In the formula, Let be the mesoscopic traffic flow and signal control state vector of the j-th intersection at time t; Let be the cross-sectional flow rate at the j-th intersection; Let the spatial density be the value at the j-th intersection. Let J be the average vehicle speed at the j-th intersection; The multi-phase timing state matrix for the j-th intersection; Let be the queue length at the j-th intersection.

[0093] Network-scale service state vector (Macro-level / Minute-level): Primarily based on topological dynamics graph features.

[0094]

[0095] In the formula, Let be the macroscopic road network topology state vector at time t; Represents graph structure functions; The road network segment connection weight matrix; This represents the distribution vector of the road network congestion index. This is a macroscopic travel demand matrix for the road network.

[0096] 3. Definition and encapsulation generation of Database Atomic Components (ADCs)

[0097] If the decoupled state vector is directly stored in the database, it is still prone to dirty reads during concurrent calls. This invention encapsulates it as a collection of object-oriented atomic components at the storage engine layer. The data structure model for a single component is defined as follows:

[0098]

[0099] Among them, UID is a globally unique hash primary key with a geo-encoded (GeoHash) prefix, which supports ultra-fast spatial clustering retrieval; Enumerate labels for scale This guides the database query planner to push down the index; Core state data payload, core state data payload State vector after multi-scale decoupling , , The direct carrier; It provides permission masks based on role-based access control (ACL) to support logical isolation for multi-tenant applications (such as police and automotive companies). This is the dynamic lifetime decay factor, representing the remaining lifetime value of the atomic data component ADC at time t. The decay function is defined as follows:

[0100]

[0101] in, For scale-dependent decay constants (microscopic components) Minimal, macroscopic components (Larger). When a component's lifecycle approaches zero, the database automatically triggers garbage collection, downgrading, and archiving, achieving precise "metabolism" of memory.

[0102] 4. Dynamic Reconfiguration Model for Decoupling Cross-Domain Business Tasks

[0103] When different business domains initiate complex cross-domain concurrent requests to the cloud platform (such as "multiple emergency vehicles coordinating green wave passage"), this module uses the "decoupling task reconstruction matrix" to decompose the fuzzy semantics of high-order business into precise extraction operations of multi-scale atomic components.

[0104] Define complex cross-domain requests from external inputs as The internal restructuring logic mapping of the system is as follows:

[0105]

[0106] in, These represent the vectors of surviving atomic components belonging to the micro, meso, and macro scales, respectively, in the database object pool; Extract the activation matrix (or query the weight mask) as a Boolean value. Represents specific components Independent extraction instructions. When At that time, it means that the cross-domain business concurrently triggered the atomic component. Snapshot Read.

[0107] Through this matrix mapping, large multi-table join transactions that are prone to deadlocks are mathematically equivalently transformed into multiple parallel, independent, stateless database concurrent SELECT commands, which are then concatenated and sent out in the application layer memory, completely bypassing the global lock contention of traditional relational databases.

[0108] 5. Physical encapsulation and cross-domain collaborative scheduling of hierarchical scale databases

[0109] The logical decoupling of scales ultimately manifests in the layered encapsulation system of the physical database cluster:

[0110]

[0111] Vehicle Scale Database Deployed on edge computing nodes, it adopts a memory-based key-value architecture and is dedicated to storing frequently updated and deleted data. Components.

[0112] Intersection Scale Database Deployed in a regional cloud time-series database, storing... The components are designed to support regional information control.

[0113] Road network scale database Deployed in the central cloud map database, storing... The component supports global topology calculations.

[0114] To address high concurrency requests from multiple tenants, the system introduces a load-sensitive dynamic scheduling resource quota function based on physical scale awareness:

[0115]

[0116] In the formula, Priority coefficients are based on business security attributes; core load requirements are no longer abstract IT metrics, but rather determined by the actual number of micro, meso, and macro components invoked concurrently during task refactoring. and its physical data size weight Dynamically determined joint operators; This sets a minimum limit for dynamically allocating I / O and computing resources to this business domain. This ensures that, in the event of a sudden traffic disaster, the extraction instructions for high-frequency microscopic physical components enjoy absolute data transmission priority.

[0117] Technical effect analysis

[0118] Compared to existing microservice architectures and traditional distributed traffic databases, this invention possesses extremely high technical barriers and disruptive engineering advantages:

[0119] (1) Eradicate “scale collapse” and significantly reduce system storage and cross-layer query I / O overhead;

[0120] The pioneering "vehicle-intersection-road network" multi-scale orthogonal decoupling mapping model enables pixel-level alignment between the database's organizational skeleton and the topological logic of the real physical traffic space. By abandoning the crude stacking of large tables, the underlying data structure exhibits extremely high purity, reducing redundant filtering overhead during cross-level queries by nearly an order of magnitude.

[0121] (2) A unique atomic component lifecycle mechanism enables precise “metabolism” of high-concurrency memory resources;

[0122] Innovatively, physical decay time Embedded within the database component layer, this enables the underlying database engine to perceive the "shelf life" of data for the first time. Extremely high-frequency micro-trajectory data disappears naturally and rapidly over time, greatly avoiding the accumulation of useless historical fragments in memory and ensuring the strong robustness and high throughput of the cloud control platform when encountering extreme concurrency impacts during peak hours.

[0123] (3) Decoupling task reconstruction matrix to overturn cross-domain concurrency deadlock;

[0124] To address the problem of strong coupling and deadlocks in complex IVCPS business logic when calling underlying data tables, this invention decouples the large and ambiguous business logic into smaller parts through a task reconstruction matrix, transforming it into concurrent lock-free snapshot extraction of discrete atomic component groups. This completely avoids the fatal global lock contention issues in relational databases, enabling the system to reduce response latency from hundreds of milliseconds to microseconds when handling complex scenarios such as "multi-vehicle cluster collaborative passage."

[0125] (4) Truly empower the commercialization of complex, multi-objective collaboration in "vehicle-road-cloud integration";

[0126] The "scale decoupling-encapsulation-reconstruction-categorized storage" link constructed in this invention is naturally adapted to the distributed hierarchical computing architecture of cloud-network convergence. It enables the macro-level control intentions of traffic management departments and the micro-level collision avoidance control logic of autonomous vehicles to coexist harmoniously in the same underlying data pool through the free assembly of components of different scales, breaking down data barriers and integration silos in the intelligent transportation industry.

[0127] It is hereby declared that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A cross-domain database collaborative access method based on multi-scale decoupling and atomic encapsulation in IVCPS, characterized in that, Includes the following steps: S1. Heterogeneous data concurrent access and cleaning: The message gateway of the cloud control base concurrently receives heterogeneous bit streams uploaded by connected vehicle OBU and roadside infrastructure RSU, and performs spatial timestamp alignment and noise reduction processing on the heterogeneous bit streams. S2. Multi-scale physical state decoupling: Construct a multi-scale orthogonal projection operator S to decouple the mixed data stream processed in step S1 into microscopic vehicle kinematics state vector, mesoscopic intersection traffic flow and signal control state vector, and macroscopic road network topology state vector according to the physical range and time refresh frequency. S3. Database Atomized Component Encapsulation: Inject the three types of state vectors obtained from decoupling in step S2 into a standardized quintuple template to generate an atomic data component (ADC). The structure of a single ADC is as follows: ; Wherein, UID is a globally unique hash primary key with a spatial geocoding prefix; Enumerate labels for scale; The core state data payload; This is the permission mask for role-based access control. The dynamic lifetime decay factor represents the remaining lifetime value of the atomic data component ADC at time t. S4. Hierarchical Database Classification and Input: The database routing engine parses the scale labels of the ADCs and distributes the micro, meso, and macro-scale ADCs to the physically isolated edge cloud vehicle scale database (DB) respectively. V Regional Cloud Intersection Scale Database DB I Central Cloud Road Network Scale Database DB N Storage is completed in the middle; S5. Task Decoupling and Concurrency Mapping: Intercepting Cross-Domain Business Requests Initiated by Multiple Business Domains Construct a decoupling task reconstruction matrix The cross-domain service request Dimensionality reduction and decomposition for DB V DB I DB N List of concurrent ADC extraction instructions; S6. Lock-free snapshot extraction and resource scheduling: Based on load-sensitive dynamic scheduling resource quota function. Resource channels are allocated for each extraction instruction, lock-free snapshot reading based on multi-version concurrency control is performed in databases at all levels, the extracted ADCs are aggregated and assembled into a real-time service view in application layer memory, and control instructions are issued based on the real-time service view. S7. Lifecycle self-healing and physical consistency maintenance: Continuously monitor the ADC's lifecycle decay factor. For ADCs whose lifecycle approaches zero, memory reclamation and archiving operations are automatically triggered. When the data in the edge cloud and the central cloud database are physically offset, the divergence self-healing mechanism is triggered to complete the precise synchronization of data. The feedback data from the physical world re-enters step S1, forming a data processing closed loop.

2. The IVCPS cross-domain database collaborative access method based on multi-scale decoupling and atomic encapsulation according to claim 1, characterized in that, Step S2 includes the following sub-steps: S2.1 Constructing the multi-scale orthogonal projection operator S; In the formula, This represents the scale category label that the i-th data record is ultimately assigned, with a value of [value]. One of them corresponds to the micro-scale of vehicles, the meso-scale of intersections, and the macro-scale of road networks, respectively. The preset scale feature mapping coefficients are calibrated offline based on the communication bandwidth of the IVCPS system and the physical sensor deployment density. The spatial coverage feature of the i-th data point at scale S; Let S be the time update frequency of the i-th data point at scale S; Let S be the sampling precision of the i-th data point at scale S; S2.2 Based on the multi-scale orthogonal projection operator S calculated in step S2.1, the mixed data stream processed in step S1 is decoupled into microscopic vehicle kinematics state vector, mesoscopic intersection traffic flow and signal control state vector and macroscopic road network topology state vector according to the physical range of action and time refresh frequency. I. Microscopic vehicle kinematic state vector: In the formula, Let be the microscopic vehicle kinematic state vector at time t; Let be the vehicle's real-time lateral position coordinates at time t; Here are the real-time longitudinal coordinates of the vehicle at time t; Let be the vehicle's real-time speed at time t; Let be the real-time acceleration of the vehicle at time t; Let be the vehicle's real-time heading angle at time t; This refers to the real-time operating mode of the vehicle at time t. II. Traffic Flow and Signal Control State Vectors at Intersections (Medium Level): In the formula, Let be the mesoscopic traffic flow and signal control state vector of the j-th intersection at time t; Let be the cross-sectional flow rate at the j-th intersection; Let the spatial density be the value at the j-th intersection. Let J be the average vehicle speed at the j-th intersection; The multi-phase timing state matrix for the j-th intersection; Let be the queue length at the j-th intersection; III. Macro-level road network topology state vector: In the formula, Let be the macroscopic road network topology state vector at time t; Represents graph structure functions; The road network segment connection weight matrix; This represents the distribution vector of the road network congestion index. A macro-level travel demand matrix for the road network; Core state data payload State vector after multi-scale decoupling , , The direct carrier.

3. The IVCPS cross-domain database collaborative access method based on multi-scale decoupling and atomic encapsulation according to claim 2, characterized in that, In step S3, the dynamic lifetime decay factor The decay function is: In the formula, This represents the initial lifetime value of the ADC. For ADC generation time; The decay constant is a scale-dependent property.

4. The IVCPS cross-domain database collaborative access method based on multi-scale decoupling and atomic encapsulation according to claim 3, characterized in that, In step S4: The vehicle scale database DB V It is a memory-based key-value architecture used to store microscale ADCs that are generated and destroyed at high frequencies; The intersection-scale database DB I It is a time-series database used to store mesoscale ADCs to support traffic control services in intersection areas; The road network scale database DB N It serves as a graph database for storing macro-scale ADCs to support global topology calculations for city-level road networks.

5. The IVCPS cross-domain database collaborative access method based on multi-scale decoupling and atomic encapsulation according to claim 4, characterized in that, In step S5, cross-domain service requests The reconstruction mapping relationship is as follows: In the formula, These are the survival ADC vectors at the micro, meso, and macro scales, respectively. The activation matrix is ​​extracted as a Boolean value, which is the decoupling task reconstruction matrix. This represents the k-th surviving ADC component, from... Any scale set in the middle; ,matrix Chinese correspondence Element; This indicates that the current business request needs to read this ADC component. This triggers a lock-free snapshot read; This indicates that the current business request does not require the ADC component and should skip reading it.

6. The IVCPS cross-domain database collaborative access method based on multi-scale decoupling and atomic encapsulation according to claim 5, characterized in that, In step S6, the load-sensitive dynamic scheduling resource quota function The calculation method is as follows: In the formula, Priority coefficients based on business security attributes; This refers to the actual number of ADCs at each scale invoked during task reconstruction. The physical data scale weights for each scale of the ADC; This refers to the total I / O and computing resources of the system. K represents the total number of business domains that initiated cross-domain business requests.

Images