Edge cloud collaborative integrated management architecture

Abstract

The invention relates to the technical field of aerospace management, in particular to an edge-cloud collaborative integrated management architecture, which comprises edge computing node equipment, a cloud platform server cluster and a networking sensor network, through a side cloud cooperation mechanism and the unified management platform, centralized management and efficient scheduling of operation and maintenance of the low-altitude aircraft are realized, management links and operation processes are reduced, the management efficiency is improved, and the unified management system can dynamically allocate resources according to operation requirements and real-time states of the low-altitude aircraft, so that the management efficiency of the low-altitude aircraft is improved. The system has the advantages that high-efficiency utilization of resources is guaranteed, waste and bottleneck of the resources are avoided, the real-time monitoring and fault early warning system can find potential faults in advance, occurrence and expansion of the faults are reduced, the maintenance cost is reduced, the downtime of the aircraft is shortened, the flight safety and the operation reliability are improved through the autonomous 4D flight path generation technology and the real-time monitoring system, and the service life of the aircraft is prolonged. And a powerful guarantee is provided for wide application of low-altitude aircrafts.

Description

Technical Field

[0001] The present invention relates to the technical field of aerospace management, and specifically to an edge-cloud collaborative integrated management architecture. Background Art

[0002] The low-altitude traffic management system is a comprehensive system for comprehensively managing, monitoring, and commanding various types of aircraft in the low-altitude airspace, and its core lies in spatio-temporal resource management. This system deeply integrates cutting-edge technologies such as computer vision, radar detection, satellite positioning, big data analysis, artificial intelligence algorithms, and 5G communication to achieve collaborative control of low-altitude aircraft and ground traffic. The low-altitude traffic management system can real-time sense the traffic environment, accurately identify the state of aircraft, intelligently predict traffic trends, and provide key services such as flight path planning, traffic flow optimization, conflict warning and avoidance, and meteorological information docking. In modern air transportation, the low-altitude traffic management system plays a crucial role. By scientifically planning and efficiently utilizing airspace resources, it significantly improves air traffic capacity and operation efficiency, ensures flight safety and smoothness, reduces operating costs, enhances economic benefits, and lays a solid foundation for the booming development of general aviation and the low-altitude economy.

[0003] Currently, the low-altitude traffic management system faces many challenges. The traditional centralized architecture has the problem of functional fragmentation, and the integration of links such as airspace management, air traffic services, traffic flow management, and traffic control is not smooth, resulting in low management efficiency. Against the background of the increasing number of low-altitude aircraft, the complexity of its operation and maintenance is also becoming increasingly prominent. The traditional operation and maintenance mode is difficult to meet the requirements of real-time monitoring, rapid response, and efficient management, and there are problems such as unreasonable resource allocation and untimely fault handling. Summary of the Invention

[0004] The purpose of the present invention is to provide an edge-cloud collaborative integrated management architecture to solve the problems raised in the above background art.

[0005] To achieve the above purpose, the present invention provides the following technical solution: An edge-cloud collaborative integrated management architecture includes edge computing node devices, a cloud platform server cluster, and an Internet of Things sensor network. The edge computing node devices include a processor, a memory, and a communication module. The cloud platform server cluster adopts high-performance server hardware and a distributed storage system. The Internet of Things sensor network is composed of sensors installed on key components and systems of low-altitude aircraft. A distributed computing framework, a big data analysis platform, and unified information management system software are installed on the cloud platform server cluster. A lightweight operating system and data processing software are installed on the edge computing node devices.

[0006] Preferably, the sensors in the Internet of Things sensor network include temperature sensors, pressure sensors, vibration sensors, etc.

[0007] The method for building a management architecture includes the following steps: Step S1: Deploy edge computing nodes on low-altitude aircraft. The edge computing nodes are directly connected to devices such as sensors and actuators of the aircraft, and local data collection and processing are achieved through high-speed data interfaces. The edge computing nodes perform encryption processing on the transmitted data. Step S2: Establish a cloud platform adopting a distributed computing architecture. The cloud platform is connected to the edge computing nodes through a high-speed and stable communication network. Step S3: Formulate a workflow and protocol for cloud-edge collaboration, clarify the responsibilities and division of labor of the edge computing nodes and the cloud platform in aspects such as data processing and decision-making, and utilize powerful computing resources for complex data analysis and model training. Step S4: Develop a unified information management system based on the SWIM concept, integrate various data sources related to the operation and maintenance of low-altitude aircraft, adopt standardized data interfaces and protocols to achieve data interaction and sharing between different data sources, introduce data fusion algorithms, and perform fusion processing on data from different data sources.

[0008] Preferably, in step S1, the edge computing nodes have a lightweight operating system and computing resources, and adopt a redundant design.

[0009] Preferably, in step S2, the cloud platform consists of multiple computing nodes and storage nodes, which are interconnected through a high-speed network to form a highly available and highly scalable computing and storage resource pool.

[0010] Preferably, in step S2, the cloud platform adopts load balancing and failover technologies.

[0011] Preferably, in step S3, after the division of labor, the edge computing nodes are responsible for real-time processing of local data and performing preliminary data cleaning and feature extraction operations, and the cloud platform is responsible for global optimization and long-term trend analysis.

[0012] Preferably, in step S3, to further optimize the collaboration efficiency, a collaborative decision-making algorithm can be introduced, and its formula is expressed as: D = argmax dinD (alpha * L(d) + beta * G(d) + gamma * C(d)); In the formula: D represents the decision set, d represents a specific decision option, L(d) represents the local evaluation value of the decision d by the edge computing node, G(d) represents the global evaluation value of the decision d by the cloud platform, C(d) represents the collaborative degree evaluation value of the decision d, and alpha, beta, and gamma respectively represent the weight coefficients of the local evaluation value, global evaluation value, and collaborative degree evaluation value.

[0013] Preferably, in the step S4, various data sources include aircraft status data, airspace resource information, and meteorological data.

[0014] Preferably, in the step S4, the data fusion algorithm formula is expressed as: F = sum i=1 w i cdot(D i +epsilon i ); In the formula, F represents the fused data, Di represents the data from the i-th data source, wi represents the weight coefficient of the i-th data source, epsiloni represents the data noise of the i-th data source, and n represents the number of data sources.

[0015] Compared with the prior art, the beneficial effects of the present invention are as follows: Through the edge-cloud collaboration mechanism and the unified management platform, the present invention realizes the centralized management and efficient scheduling of the operation and maintenance of low-altitude aircraft, reduces the management links and operation processes, improves the management efficiency, and the unified management system can dynamically allocate resources according to the operation requirements and real-time status of low-altitude aircraft to ensure the efficient utilization of resources, avoid the waste and bottleneck of resources, and the real-time monitoring and fault warning system can detect potential faults in advance, reduce the occurrence and expansion of faults, reduce the maintenance cost and the downtime of the aircraft, and the autonomous 4D flight path generation technology and the real-time monitoring system improve the flight safety and operation reliability, providing a strong guarantee for the wide application of low-altitude aircraft. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 is the architecture diagram of the present invention; Figure 2 is the schematic diagram of the step flow when building the management architecture of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0017] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0018] The edge-cloud collaborative integrated management architecture includes edge computing node devices, cloud platform server clusters, and an Internet of Things (IoT) sensor network. The edge computing node devices include processors, memories, and communication modules, ensuring correct connections between the edge computing nodes and devices such as sensors and actuators of the aircraft. The cloud platform server cluster uses high-performance server hardware and a distributed storage system to ensure that the cloud platform has powerful data processing and storage capabilities. The IoT sensor network is composed of sensors installed on key components and systems of the low-altitude aircraft, such as temperature sensors, pressure sensors, vibration sensors, etc., to achieve real-time monitoring of the aircraft's status. A lightweight operating system and data processing software are installed on the edge computing node devices, and autonomous 4D flight path generation algorithms and real-time data processing algorithms are configured. A distributed computing framework, a big data analysis platform, and unified information management system software are installed on the cloud platform server cluster, and resource scheduling algorithms, fault prediction models, and maintenance management function modules are configured. A unified management interface and tools are developed, using Web technology or mobile application technology to achieve cross-platform management and operation. The edge computing nodes, cloud platform, and IoT sensor network are system-integrated to ensure normal data interaction and collaborative work between components. System testing is carried out, including functional testing, performance testing, and reliability testing, to verify whether the functions of the edge-cloud collaborative integrated management architecture are normal, whether the performance meets the requirements, and whether the system is stable and reliable.

[0019] The method for building its management architecture includes the following steps: Step S1: Deploy edge computing nodes on the low-altitude aircraft. These nodes not only have a lightweight operating system and computing resources, can run specific data processing and analysis algorithms, but also adopt a redundant design to ensure that the system can still operate normally when some nodes fail. The edge computing nodes are directly connected to devices such as sensors and actuators of the aircraft, and local data collection and processing are realized through high-speed data interfaces, providing a real-time and accurate data basis for subsequent cloud collaboration processing. To enhance data security, the edge computing nodes encrypt the transmitted data. Step S2: Establish a cloud platform, adopting a distributed computing architecture, with powerful data processing, storage, and analysis capabilities. The cloud platform consists of multiple computing nodes and storage nodes, interconnected through a high-speed network to form a highly available and highly scalable computing and storage resource pool. The cloud platform is connected to the edge computing nodes through a high-speed and stable communication network to ensure real-time data transmission and interaction, forming an "end-edge-cloud" collaborative data processing link. To further improve the fault tolerance of the cloud platform, load balancing and fault transfer technologies are adopted to ensure that the service can still be continuously provided when some nodes fail. Step S3: Develop the workflow and protocol for cloud-edge collaboration, and clarify the responsibilities and division of labor between edge computing nodes and cloud platforms in aspects such as data processing and decision-making. For example, edge computing nodes are responsible for real-time processing of local data, performing preliminary data cleaning, feature extraction, etc., while the cloud platform is responsible for global optimization and long-term trend analysis, using powerful computing resources for complex data analysis and model training. To further optimize the collaboration efficiency, a collaborative decision-making algorithm can be introduced, and its formula is expressed as: D = argmax dinD (alpha * L(d) + beta * G(d) + gamma * C(d)); In the formula: D represents the decision set, d represents a specific decision option, L(d) represents the local evaluation value of the edge computing node for decision d, G(d) represents the global evaluation value of the cloud platform for decision d, C(d) represents the collaboration degree evaluation value of decision d, and alpha, beta, and gamma respectively represent the weight coefficients of the local evaluation value, global evaluation value, and collaboration degree evaluation value. Through this algorithm, collaborative decision-making between edge computing nodes and cloud platforms can be achieved, improving the overall performance of the system; Step S4: Develop a unified information management system based on the SWIM concept, integrate various data sources related to the operation and maintenance of low-altitude aircraft, including aircraft status data, airspace resource information, meteorological data, etc., and adopt standardized data interfaces and protocols to achieve data interaction and sharing between different data sources. To improve the quality and accuracy of data, a data fusion algorithm is introduced to perform fusion processing on data from different data sources. The formula of the data fusion algorithm is expressed as: F = sum i=1 w i cdot (D i + epsilon i ); In the formula, F represents the fused data, Di represents the data from the i-th data source, wi represents the weight coefficient of the i-th data source, epsilon_i represents the data noise of the i-th data source, and n represents the number of data sources. Through this algorithm, denoising and fusion processing of data from different data sources can be performed, improving the quality and accuracy of data, and providing more reliable data support for subsequent decision-making. At the same time, the unified information management system also provides rich visualization interfaces and tools to facilitate managers to monitor and schedule the operation status of low-altitude aircraft in real time.

[0020] Example: The edge computing nodes deployed on low-altitude aircraft adopt a modular architecture design, equipped with a deeply optimized real-time operating system (such as ROS2 or xWorks), integrated with multi-core heterogeneous processors and FPGA acceleration units to form a dynamically scalable computing resource pool. Through an adaptive task scheduling algorithm, computing power can be flexibly allocated according to flight mission requirements. A dual-redundant CAN bus and a time-sensitive network (TSN) architecture are built inside the nodes to achieve nanosecond-level synchronous communication between the sensor array and the actuator. At the same time, a distributed storage system based on blockchain technology is used for local data caching to ensure a millisecond-level processing delay for flight attitude data, environmental perception information, and navigation signals. These nodes are equipped with a lightweight AI inference engine, which can run point cloud segmentation, SLAM mapping, and anomaly detection algorithms in real time, and form an anti-jamming Mesh network with spaceborne edge nodes through a quantum encryption channel. During flight, the encryption strategy can be dynamically adjusted (such as a hybrid encryption of national cipher SM4 / AES-256), and an end-to-end security protection system is built in combination with the TEE trusted execution environment. The self-healing fault tolerance mechanism built into the nodes realizes hardware anomaly prediction through a digital twin model. Combined with the Docker containerized service hot migration technology, the computing load can be seamlessly transferred to adjacent nodes when a single node fails, and the dynamic offloading of computing tasks is achieved through an edge-fog-cloud three-level collaborative architecture, significantly improving the system robustness and task continuity in a complex electromagnetic environment, and providing a highly reliable edge intelligence infrastructure for the collaborative perception and autonomous decision-making of multi-aircraft formations.

[0021] The constructed cloud platform is based on a hybrid cloud-native architecture, adopting a distributed resource scheduling framework that combines Kubernetes federated clusters and Apache-Mesos. At the bottom layer, a hyper-converged storage system with horizontal scalability is deployed (integrating Ceph object storage and Alluxio memory acceleration layer), and trillions of spatio-temporal data are mapped to heterogeneous storage nodes through intelligent data sharding technology. The core of the platform is equipped with a stream-batch integrated processing engine that mixes Spark-Streaming and Flink, supporting real-time feature extraction and federated learning modeling of multi-modal data (lidar point cloud, millimeter-wave radar trajectory, electro-optical pod video stream) uploaded by aircraft clusters. At the same time, it integrates the OpenStack and Kubernetes virtualization layers to achieve hybrid orchestration of bare-metal servers and container instances, and relies on a 400Gbps RoCEv3 RDMA network to build a low-latency communication plane for in-memory computing. The cloud-edge collaboration channel adopts multi-link aggregation technology (integrating 5G sliced private network, low-earth orbit satellite Q / V band, and terahertz directional backhaul), realizes data priority transmission from edge nodes to the cloud through a dynamic bandwidth allocation algorithm based on deep reinforcement learning, and deploys trusted computing nodes based on SGX-enclave to perform homomorphic encryption and multi-party secure computing on sensitive data. The platform has a three-layer fault tolerance system: at the hardware layer, it uses N+M hot standby servers and erasure code distributed storage; at the service layer, it realizes canary release and automatic fusing across availability zones through the Istio service mesh; at the data layer, it uses blockchain evidence storage technology to ensure the non-tampering of data traceability. The intelligent operation and maintenance system integrates digital twins to map the physical resource status in real time, combines the Prometheus time series database and the LSTM prediction model to achieve sub-healthy warning of hardware failures. At the same time, through an elastic scaling strategy driven by reinforcement learning, it can dynamically reorganize the GPU computing power pool and the hierarchical storage architecture of hot and cold data according to the edge-end task load, forming a full-stack intelligent processing closed loop from edge perception to cloud cognition, meeting the intelligent operation requirements of high concurrency, strong real-time performance, and anti-damage for tens of billions of IoT terminals in the low-altitude economy field.

[0022] To formulate a cloud-edge collaborative workflow, a five-layer protocol stack framework needs to be constructed: at the physical layer, a hybrid network of IEEE 802.11ax and TSN time-sensitive network is adopted to achieve microsecond-level delay communication between the aircraft cluster and the edge nodes; at the transport layer, an enhanced HTTP / 3 bidirectional data channel based on the QUIC protocol is deployed to support priority marking and dynamic bandwidth allocation of the aircraft status data stream; at the collaborative decision-making layer, a three-layer task grading mechanism is established - the emergency obstacle avoidance instruction is autonomously responded by the edge node through the DDS real-time publish-subscribe system accelerated by FPGA hardware, the complex path planning task is completed on the edge side through the LSTM prediction model for short-term decision-making, and at the same time, the multi-aircraft collaborative strategy generation and large-scale environment modeling tasks are offloaded to the cloud heterogeneous computing cluster; the data protocol adopts an adaptive Protobuf / Avro hybrid serialization format, and a cross-layer data bus is constructed through Apache Kafka to achieve standardized encapsulation and version control of sensor raw data, feature vectors, and model parameters; in terms of intelligent collaborative algorithms, a fusion architecture of federated learning and edge distillation is designed, the lightweight AI model is periodically sent from the cloud to the edge nodes, the edge side aggregates the local model updates through differential privacy technology, and the dynamic offloading strategy (DRL-DOS) driven by reinforcement learning is used to optimize the computing task allocation. At the same time, a resource portrait system based on digital twin is established, and the intelligent mapping and elastic scheduling of cloud-edge resources are realized through Prometheus monitoring metrics and Grafana visualization dashboards. A two-way authentication TLS-1.3 encryption channel and a blockchain evidence storage module are integrated into the protocol stack to ensure the full-link credible traceability from sensor data collection to cloud decision feedback. Finally, a cloud-edge collaborative API system including 18 types of standard interfaces such as data preprocessing, model incremental update, task dynamic orchestration, and security audit is formed to meet the dual requirements of millisecond-level response and EB-level data processing in the low-altitude economy scenario.

[0023] The development of a unified information management system based on the SWIM concept adopts an aviation-grade SOA architecture, integrates the ICAO-SWIM standard and the UAVSC (Unmanned Aircraft System Communication) data model, constructs a multi-layer data bus through Apache Kafka, and realizes the millisecond-level synchronous acquisition of aircraft OBD-II bus data, ADS-B signals, airspace dynamic grid coding (such as U-Space four-dimensional trajectories), Doppler weather radar, and GNSS differential correction data. The system is deployed with a hybrid transport layer based on the HTTP / 3 QUIC protocol stack and MQTT5.0, supports dual-mode data encapsulation of JSON-Schema and Avro, and introduces a multi-source heterogeneous data fusion engine driven by digital twins. An improved cross-modal alignment algorithm combining Kalman filtering and deep neural networks is adopted to eliminate the spatio-temporal reference differences of multi-sensors. The data quality enhancement module integrates a federated learning framework, realizes data traceability tracking through the blockchain-enabled Data-Provenance mechanism, and at the same time deploys an LSTM-based abnormal data detection model and a missing value filling strategy driven by Bayesian inference to construct a dynamic credibility evaluation system containing 87 indicators. The system security layer uses the national cryptographic SM9 algorithm to implement attribute-based encryption (ABE), constructs a hierarchical access control model in combination with zero-knowledge proof technology, and realizes hard real-time interoperability with the air traffic control automation system through the OPC-UA over TSN protocol. A lightweight AI proxy node is deployed on the edge side to execute the prediction of route conflict probability and dynamic airspace division, and finally forms an intelligent information hub covering the entire link of data collection, cleaning, fusion, simulation, and decision-making, meeting the requirements of the low-altitude economy field for the trusted fusion of EB-level data streams and microsecond-level collaborative decision-making.

[0024] Although the embodiments of the present invention have been shown and described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention, and the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An edge-cloud collaborative integrated management architecture, including edge computing node devices, a cloud platform server cluster, and a networked sensor network, is characterized in that: The edge computing node device includes a processor, a memory, and a communication module. The cloud platform server cluster adopts high-performance server hardware and a distributed storage system. The Internet of Things sensor network is composed of sensors installed on key components and systems of low-altitude aircraft. A distributed computing framework, a big data analysis platform, and unified information management system software are installed on the cloud platform server cluster. A lightweight operating system and data processing software are installed on the edge computing node device.

2. The edge-cloud collaborative integrated management architecture according to claim 1, wherein: The sensors in the Internet of Things sensor network include temperature sensors, pressure sensors, vibration sensors, etc.

3. The edge-cloud collaborative integrated management architecture according to claim 1, wherein The method for building the management architecture includes the following steps: Step S1: Deploy edge computing nodes on the low-altitude aircraft. The edge computing nodes are directly connected to devices such as the sensors and actuators of the aircraft, and local data collection and processing are realized through high-speed data interfaces. The edge computing nodes perform encryption processing on the transmitted data. Step S2: Establish a cloud platform adopting a distributed computing architecture. The cloud platform is connected to the edge computing nodes through a high-speed and stable communication network. Step S3: Formulate the working processes and protocols for cloud-edge collaboration, clarify the responsibilities and divisions of labor of the edge computing nodes and the cloud platform in aspects such as data processing and decision-making, and use powerful computing resources for complex data analysis and model training. Step S4: Develop a unified information management system based on the SWIM concept, integrate various data sources related to the operation and maintenance of low-altitude aircraft, adopt standardized data interfaces and protocols to realize data interaction and sharing between different data sources, introduce a data fusion algorithm, and perform fusion processing on data from different data sources.

4. The edge-cloud collaborative integrated management architecture according to claim 3, wherein: In the step S1, the edge computing node has a lightweight operating system and computing resources, and adopts a redundant design.

5. The edge-cloud collaborative integrated management architecture according to claim 3, wherein: In the step S2, the cloud platform is composed of multiple computing nodes and storage nodes, interconnected through a high-speed network, forming a highly available and highly scalable computing and storage resource pool.

6. The edge-cloud collaborative integrated management architecture according to claim 3, wherein: In the step S2, the cloud platform adopts load balancing and failover technologies.

7. The edge-cloud collaborative integrated management architecture according to claim 3, wherein: In the step S3, after the division of labor, the edge computing node is responsible for real-time processing of local data and performing preliminary data cleaning and feature extraction operations, and the cloud platform is responsible for global optimization and long-term trend analysis.

8. The edge-cloud collaborative integrated management architecture according to claim 3, wherein: In the step S3, to further optimize the collaboration efficiency, a collaborative decision-making algorithm can be introduced, and its formula is expressed as: D = argmax dinD (α ⋅ L(d) + β ⋅ G(d) + γ ⋅ C(d)); In the formula: D represents the decision set, d represents a specific decision option, L(d) represents the local evaluation value of the edge computing node for the decision d, G(d) represents the global evaluation value of the cloud platform for the decision d, C(d) represents the collaborative degree evaluation value of the decision d, and alpha, beta, and gamma respectively represent the weight coefficients of the local evaluation value, the global evaluation value, and the collaborative degree evaluation value.

9. The edge-cloud collaborative integrated management architecture according to claim 3, wherein: In the step S4, various data sources include aircraft status data, airspace resource information, and meteorological data.

10. The edge-cloud collaborative integrated management architecture according to claim 3, characterized in that: In the step S4, the formula of the data fusion algorithm is expressed as: F = sum i=1 w i ·(D i + ε i ); In the formula, F represents the fused data, Di represents the data from the i-th data source, wi represents the weight coefficient of the i-th data source, epsiloni represents the data noise of the i-th data source, and n represents the number of data sources.

Images