A hierarchical information service method based on edge computing

By using a hierarchical information service system based on edge computing, the problem of efficient and accurate information distribution in unmanned aerial vehicle (UAV) swarms is solved. It enables information service quality assessment and adaptive hierarchical processing in a dynamic self-organizing network, meeting the diverse needs of UAV swarms.

CN115987980BActive Publication Date: 2025-11-07CHINA ACAD OF LAUNCH VEHICLE TECH
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202211442042.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-11-15
Publication Date
2025-11-07
Estimated Expiration
2042-11-15

AI Technical Summary

Technical Problem

In unmanned aerial vehicle (UAV) swarms, how can we provide hierarchical information services under dynamic self-organizing network conditions, combined with network transmission service quality, to solve the problem of efficient and accurate information sharing and distribution, especially by fully leveraging the advantages of edge computing in dedicated networks and dedicated protocols?

Method used

A hierarchical information service system based on edge computing is adopted, including an aircraft cluster edge computing platform, a sensor agent module, an actuator agent module, an information gateway module, a mission-level information product generation module, a service quality assessment model module, and an information distribution module. Through transfer learning of pre-trained models and fine-tuning of service quality assessment models, information can be processed locally and distributed accurately.

Benefits of technology

It enables efficient and accurate information distribution among unmanned aerial vehicle (UAV) swarms under resource-constrained conditions, enhances the dynamic adaptability of information service capabilities and the scalability of the system, and supports diverse mission requirements and application scenarios.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN115987980B_ABST
    Figure CN115987980B_ABST
Patent Text Reader

Abstract

The application discloses a hierarchical information service system based on edge computing, which comprises an aircraft cluster edge computing platform, a sensor agent module, an actuator agent module, an information gateway module, a task hierarchical information product generation module, a service quality evaluation model module and an information distribution module; the aircraft information is processed nearby by using the edge computing technology, different levels of information products are formed, a service quality evaluation model is constructed, and based on the pre-training model migration learning, according to the task requirements and the link service quality, the suitable information product and information distribution strategy are selected, the information sharing of the long-distance, large-scale and resource-limited aircraft cluster is realized, and the efficient and accurate distribution of information under the resource-limited condition is solved; the platform+plug-in architecture is adopted, the message processing flow is encapsulated by using the micro-service mode, and the diversified and expandable requirements of the unmanned aircraft cooperation are met.
Need to check novelty before this filing date? Find Prior Art

Description

TECHNICAL FIELD

[0001] The present application relates to a hierarchical information service method based on edge computing, belonging to the cross field of computer software and communication technology. BACKGROUND

[0002] Unmanned aerial vehicles do not need to consider human factors in design and use, and can complete many boring, dirty or dangerous tasks that manned aircraft cannot complete, and have wide application in military, civil and other fields. Unmanned aerial vehicles have wide operating areas, many types, diverse tasks and complex application environments. In particular, under the condition of confrontation, the self-organizing network links between unmanned aerial vehicles are unstable, interrupted and weakly connected, and different types of information such as instructions, coordination, intelligence (including image and simple intelligence), and state have different characteristics and transmission requirements. These diverse requirements bring great problems to the networking communication and information sharing between unmanned aerial vehicles.

[0003] Abroad, in the field of unmanned aerial vehicle networking communication, DARPA has successively sponsored PRNET, SURAN and GloMo projects to solve the problem of networking communication between unmanned aerial vehicles. The Packet Radio Network (PRNET) project improves the capacity of wireless networks through multi-hop relay technology and builds the prototype of a multi-hop wireless network. The Survivable Adaptive Network (SURAN) and Global Information Systems (GloMo) projects develop a wireless communication network that can be quickly established in special environments on the basis of PRNET, and improve the defects of PRNET in security, survivability, scalability, processing capacity and energy management. However, there is no specific technical approach related to the optimization of information sharing in the face of tasks combined with the status of wireless networking links.

[0004] In the field of UAV networking communication, universities and research institutions such as Air Force Engineering University, National University of Defense Technology, and research institutes under China Electronics Technology Group have carried out related research on mobile ad hoc networks in combination with applications such as UAVs. Air Force Engineering University proposed the concept of "aviation cluster cloud network" to meet the needs of future aviation cluster operations, providing a theoretical reference for conducting specific communication technology research. In December 2017, National University of Defense Technology conducted a test of fixed-wing UAV cluster autonomous flight and exploration under complex conditions. In the test, dozens of fixed-wing UAVs could quickly form a formation in the air and fly to a designated area to perform reconnaissance tasks. China Electronics Technology Group, together with Tsinghua University and Beijing Poisson Technology Co., Ltd., overcame key technologies such as airborne small high-dynamic wireless ad hoc networks and centerless cluster control, and successfully conducted cluster flight tests of 67, 119, and 200 fixed-wing UAVs, respectively. However, none of them involved providing hierarchical information services under dynamic ad hoc networks for different task requirements in combination with network transmission quality of service.

[0005] In the field of mobile edge computing, both domestically and internationally, research has focused on how to correctly deploy services between resource-constrained edge servers. Authors Y. Zhai et al. considered the user's request mode and resource constraints and used reinforcement learning to better deploy services to edge servers. In addition, N. Fernando et al. proposed that groups of mobile terminals near resource-constrained devices can also serve as an effective edge cloud. The authors proposed a work-sharing model to balance the independent workloads between heterogeneous mobile nodes, allowing applicable computing nodes to randomly leave and join the system. For IoT scenarios, F. Wang et al. proposed a new type of wireless-powered mobile edge computing system. For scenarios with multiple IoT gateways and multiple users, T. X. Tran et al. considered maximizing the user's computing task offloading benefit. The authors studied the joint task offloading and resource allocation problem, measuring the offloading benefit of user computing tasks by the weighted sum of task completion time and energy consumption. In the context of network virtualization MEC, X. Chen et al. used deep reinforcement learning tools to design an innovative computing offloading strategy for network virtualization MEC systems, which adapts well to the time-varying network dynamic topology. However, most of these studies are based on 5G networks, while UAVs, especially UAV clusters, often require dedicated networks and protocols due to their unique task execution characteristics. How to fully leverage the advantages of edge computing in dedicated networks and protocols and achieve efficient information sharing under SWaP constraints remains a pressing problem. SUMMARY

[0006] The technical problem solved by this invention is to overcome the shortcomings of existing technologies, process the large amount of raw data generated by various unmanned aerial vehicles based on edge computing technology, form data products of different levels according to mission needs, adaptively select appropriate data products in combination with network service quality, provide accurate information distribution services to different data users, and fine-tune the service quality evaluation model based on pre-trained model transfer learning to solve the system initialization problem when samples are scarce, thereby improving the dynamic adaptability of information service capabilities to mission scenarios.

[0007] The technical solution of this invention is:

[0008] This invention discloses a hierarchical information service system based on edge computing, characterized by comprising: an aircraft cluster edge computing platform, a sensor agent module, an actuator agent module, an information gateway module, a task hierarchical information product generation module, a service quality assessment model module, and an information distribution module, wherein:

[0009] The aircraft cluster edge computing platform monitors network status and generates network status parameters; it combines and orchestrates several external service plugins; it receives sensor information from the sensor agent module; it sends actuator control commands to the actuator agent module; it receives hierarchical information products of the corresponding level from the task hierarchical information product generation module and sends sensor information to the task hierarchical information product generation module; it receives service quality assessment results from the service quality assessment model module and sends network status parameters to the service quality assessment model module; it receives business protocol data from the information gateway module and sends network transmission data to the information gateway module; it receives network transmission data from the information distribution module and sends hierarchical information products and business protocol data to the information distribution module.

[0010] Sensor agent module: Receives sensor data from external sensors, parses the sensor data, generates sensor information, and sends it to the edge computing platform of the aircraft cluster;

[0011] Execution agent module: Converts the execution control commands sent by the aircraft cluster edge computing platform into execution instruction codes according to the execution interface protocol, and sends them to the external execution agencies;

[0012] Information gateway module: Converts the service data sent by external communication devices into protocol data, generates service protocol data, and sends it to the aircraft cluster edge computing platform; converts the network data sent by the aircraft cluster edge computing platform into protocol data and sends it to the external communication devices.

[0013] Task hierarchical information product generation module: according to the service quality evaluation result obtained from the aircraft cluster edge computing platform, the sensor information obtained from the aircraft cluster edge computing platform and the forward data sent by the external task system are processed to generate hierarchical information products of corresponding levels, which are sent to the aircraft cluster edge computing platform to adapt to the requirements of network transmission link;

[0014] Service quality evaluation model module: according to the quality characteristic evaluation method, the network service quality in a period of time is modeled to generate information distribution strategy matched with data producer and data consumer, and the information distribution strategy is sent to information distribution module; based on pre-training model transfer learning, the service quality evaluation model is fine-tuned to generate service quality evaluation result; the service quality evaluation result is sent to aircraft cluster edge computing platform; network state parameters sent by aircraft cluster edge computing platform are received;

[0015] Information distribution module: according to information service mode and information distribution strategy sent by service quality evaluation model module, hierarchical information products sent by aircraft cluster edge computing platform are received, priority of hierarchical information products is adjusted, network sending data is generated, and network sending data is sent to aircraft cluster edge computing platform; business protocol data sent by aircraft cluster edge computing platform is received and forwarded to external aircraft task system.

[0016] In the above service system, the information distribution strategy includes direct distribution, frequency reduction distribution, compression distribution and multi-node cooperative distribution.

[0017] In the above service system, the information service mode includes request and response, subscription notification and active push.

[0018] In the above service system, the network service quality includes distribution real-time, data validity, system resource overhead, data transmission control and data completeness.

[0019] In the above service system, the aircraft cluster edge computing platform includes resource management module, service orchestration module and basic algorithm library; wherein:

[0020] The resource management module receives sensor information sent by the external sensor agent module, receives service protocol data sent by the information gateway module, sends network sending data to the information gateway module, calls the basic algorithm library, sends the actuator control instruction to the actuator agent module, and realizes unified management of the sensor information, the service protocol data, the network sending data and the control instruction; the service arrangement module combines and arranges various external service plug-ins according to the service quality evaluation result sent by the service quality evaluation model module and the hierarchical information product of the task grading information product generation module; the basic algorithm library provides space-time benchmark alignment and intelligent algorithm engine functions; the resource management module monitors the network state, generates network state parameters, and sends the network state parameters to the service quality evaluation model module.

[0021] In the service system, the sensor agent module comprises a sensor data transceiver module and a sensor data analysis module; the sensor data transceiver module receives sensor data from external sensors; and the sensor data analysis module analyzes the sensor data according to a sensor interface protocol to generate sensor information.

[0022] In the service system, the actuator agent module comprises an actuator interface module and an actuator instruction conversion module; the actuator interface module is responsible for connection and interface interaction with external actuators; and the actuator instruction conversion module converts the actuator control instruction into an executable instruction code of the actuator according to an interface protocol of the actuator, and sends the executable instruction code to the external actuator.

[0023] In the service system, the information gateway module comprises a network information receiving module, a network protocol conversion module and a network information sending module; the network information receiving module receives service data from external communication devices; the network protocol conversion module analyzes and converts the format of the service data according to a protocol to generate service protocol data and send the service protocol data to the aircraft cluster edge computing platform; and the network information sending module reads network sending data from the aircraft cluster edge computing platform, performs protocol conversion and sends the network sending data to external communication devices.

[0024] In the service system, the sensor information and forward data sent by the external task system are processed according to the service quality evaluation result obtained from the aircraft cluster edge computing platform to generate hierarchical information products of corresponding levels, and the specific method is as follows:

[0025] The forward data of the external aircraft task system is decomposed and mapped into an atomic task;

[0026] According to the atomic task and the service quality evaluation result, the sensor information and the forward data are processed by frequency reduction, lossless compression or lossy compression to output hierarchical information products.

[0027] In the service system, the service quality evaluation model module comprises a service quality collection module, a producer QoS policy generation module, a consumer QoS policy generation module, and a pre-trained model transfer learning module; the service quality collection module periodically reads network state parameters from the aircraft cluster edge computing platform; the producer QoS policy generation module is responsible for generating a QoS policy of a data producer; the consumer QoS policy generation module is responsible for generating a matching consumer QoS policy by coordinating with the producer QoS policy generation module, to ensure that the policies of the information sending and receiving parties are matched; and the pre-trained model transfer learning module fine-tunes the service quality evaluation model to generate a service quality evaluation result.

[0028] In the service system, the information distribution module comprises a distribution information reading module, an information distribution mode selection module, an information distribution priority adjustment module, and a distribution information sending module; the distribution information reading module reads hierarchical information products from the aircraft cluster edge computing platform; the information distribution priority adjustment module adjusts the priority of the hierarchical information products according to the information distribution strategy sent by the service quality evaluation model module, to generate priority hierarchical information products; the information distribution mode selection module selects a corresponding information distribution mode according to the information service mode; the distribution information sending module generates network sending data according to the information distribution mode and the priority hierarchical information products, sends the network sending data to the aircraft cluster edge computing platform, receives service protocol data sent by the aircraft cluster edge computing platform, and forwards the service protocol data to an external aircraft task system.

[0029] The present application has the following advantages over the prior art:

[0030] (1) The present application realizes an information service method for a long-distance, large-scale, and resource-limited aircraft cluster, adopts edge computing technology to process aircraft information nearby, forms different levels of information products, selects appropriate information products and information distribution strategies according to task requirements and link service quality, and solves the problem of efficient and accurate distribution of information under resource-limited conditions.

[0031] (2) The present application realizes a service quality evaluation model for hierarchical information services, constructs a service quality evaluation model from four aspects of data effectiveness, data real-time performance, system resources, and transmission control, automatically matches the strategies of data producers and data consumers, and solves the problem of adaptation of information service capability and task demand.

[0032] (3) The present application is based on pre-trained model transfer learning, fine-tunes the service quality evaluation model, solves the problem of system initialization when the sample is scarce, and improves the dynamic adaptability of information service capability to task scenarios.

[0033] (4) The present application adopts a platform+plug-in architecture, only the architecture and interface of hierarchical information service are specified, but the specific implementation is not limited. Different equipment or system can re-implement the interface according to different needs to meet the specific business requirements. Open architecture, easy to extend.

[0034] (5) The present application adopts micro-service mode to encapsulate message processing flow, with the characteristics of reusable function, flexible service combination, etc. According to the function requirements of equipment or system, different micro-service sets are pre-set to meet the diversified needs of unmanned aerial vehicle cooperation.

[0035] (6) The present application adopts three information service modes of request and response, subscription notification and active push, supports different application requirements such as point-to-point, single point to multi-point, multi-point to multi-point, can be compatible with existing applications, and can realize flexible information sharing and system function expansion through information service mode. BRIEF DESCRIPTION OF DRAWINGS

[0036] Figure 1 is a system structure diagram of the present application;

[0037] Figure 2 is a UAV cluster edge computing platform module structure diagram of the present application;

[0038] Figure 3 is a sensor agent module structure diagram of the present application;

[0039] Figure 4 is an actuator agent module structure diagram of the present application;

[0040] Figure 5 is an information gateway module structure diagram of the present application;

[0041] Figure 6 is a task hierarchical information product generation module structure diagram of the present application;

[0042] Figure 7 is a service quality evaluation model module structure diagram of the present application;

[0043] Figure 8 is an information distribution module structure diagram of the present application. DETAILED DESCRIPTION

[0044] The present application patent will be further described in detail below in combination with the drawings and specific embodiments.

[0045] As shown in Figure 1 , the present application discloses a hierarchical information service system based on edge computing, comprising: a UAV cluster edge computing platform, a sensor agent module, an actuator agent module, an information gateway module, a task hierarchical information product generation module, a service quality evaluation model module and an information distribution module, wherein:

[0046] As shown in Figure 2 The aircraft cluster edge computing platform: monitors network status, generates network status parameters; combines and arranges a plurality of external service plug-ins; receives sensor information sent by a sensor agent module; sends actuator control instructions to an actuator agent module; receives corresponding level of hierarchical information products sent by a task hierarchical information product generation module, and sends sensor information to the task hierarchical information product generation module; receives service quality evaluation results sent by a service quality evaluation model module, and sends network status parameters to the service quality evaluation model module; receives business protocol data sent by an information gateway module, and sends network sending data to the information gateway module; receives network sending data sent by an information distribution module, and sends hierarchical information products and business protocol data to the information distribution module; the aircraft cluster edge computing platform is a basic base, and is a basic support platform for realizing task-oriented hierarchical information services.

[0047] The sensor agent module: receives sensor data of an external sensor, analyzes the sensor data, generates sensor information, and sends the sensor information to the aircraft cluster edge computing platform;

[0048] The actuator agent module: converts actuator control instructions sent by the aircraft cluster edge computing platform into actuator instruction codes according to an actuator interface protocol, and sends the actuator instruction codes to an external actuator;

[0049] The information gateway module: generates business protocol data by performing protocol conversion on business data sent by an external communication device, and sends the business protocol data to the aircraft cluster edge computing platform; and sends network sending data sent by the aircraft cluster edge computing platform to the external communication device after performing protocol conversion on the network sending data;

[0050] The task hierarchical information product generation module: processes sensor information obtained from the aircraft cluster edge computing platform and forward data sent by an external task system according to service quality evaluation results obtained from the aircraft cluster edge computing platform, generates hierarchical information products of corresponding levels, and sends the hierarchical information products to the aircraft cluster edge computing platform to adapt to requirements of a network transmission link;

[0051] The service quality evaluation model module: according to the quality characteristic evaluation method, the network service quality in a period of time is modeled, the information distribution strategy of data producer and data consumer matching is generated, the information distribution strategy is sent to the information distribution module; based on the pre-training model transfer learning, the service quality evaluation model is fine-tuned and trained, and the service quality evaluation result is generated; the service quality evaluation result is sent to the aircraft cluster edge computing platform; the network state parameters sent by the aircraft cluster edge computing platform are received; the information distribution strategy includes direct distribution, frequency reduction distribution, compression distribution and multi-node cooperative distribution. The network service quality includes distribution real-time performance, data validity, system resource overhead, data transmission control and data completeness;

[0052] The information distribution module: according to the information service mode and the information distribution strategy sent by the service quality evaluation model module, the hierarchical information product sent by the aircraft cluster edge computing platform is received, the priority of the hierarchical information product is adjusted, the network sending data is generated, and the network sending data is sent to the aircraft cluster edge computing platform; the service protocol data sent by the aircraft cluster edge computing platform is received and forwarded to the external aircraft task system.

[0053] As shown in Figure 2 The aircraft cluster edge computing platform includes a resource management module, a service arrangement module and a basic algorithm library; mainly provides basic data processing algorithms and system integration platforms for sensor agent modules, actuator agent modules, information gateway modules, task hierarchical information product generation modules, service quality evaluation model modules and information distribution modules, wherein: the resource management module receives sensor information sent by external sensor agent modules and service protocol data sent by information gateway modules; network sending data is sent to the information gateway module; the basic algorithm library is called, and actuator control instructions are sent to the actuator agent module to realize unified management of sensor information, service protocol data, network sending data and control instructions; the service arrangement module combines and arranges various external service plug-ins according to the service quality evaluation result sent by the service quality evaluation model module and the hierarchical information product of the task hierarchical information product generation module; the basic algorithm library provides space-time benchmark alignment and intelligent algorithm engine functions; the resource management module monitors the network state and generates network state parameters, which are sent to the service quality evaluation model module. The service plug-in includes coordinate conversion plug-in, data registration plug-in and data filtering plug-in.

[0054] As shown in Figure 3As shown, the sensor agent module includes a sensor data transceiver module and a sensor data analysis module, and mainly implements receiving input from external sensors, performing analysis, and generating semantic expression of sensor data in a unified format. The sensor data transceiver module receives sensor data from external sensors; the sensor data analysis module analyzes sensor data according to a sensor interface protocol, and forms a unified representation of sensor data in a semantic manner to generate sensor information.

[0055] As shown in Figure 4 The actuator agent module includes an actuator interface module and an actuator instruction conversion module, and mainly implements converting actuator control instructions generated by the platform task system or received through the communication device into instruction codes executable by the actuator according to the actuator interface protocol. The actuator interface module is responsible for connection and interface interaction with external actuators; the actuator instruction conversion module converts actuator control instructions into instruction codes executable by the actuator according to the actuator interface protocol, and sends them to external actuators.

[0056] As shown in Figure 5 The information gateway module includes a network information receiving module, a network protocol conversion module, and a network information sending module, and mainly functions to receive service data sent by external platforms through the communication device, perform protocol conversion, and send them to the aircraft cluster edge computing platform for use by other modules. The network information receiving module receives service data from external communication devices; the network protocol conversion module completes analysis and protocol conversion of the service data format according to the protocol, generates service protocol data, and sends it to the aircraft cluster edge computing platform; the network information sending module reads network sending data from the aircraft cluster edge computing platform, sends it to external communication devices after protocol conversion. The network information sending module sends information sent by the information distribution module according to the priority information carried by the information.

[0057] As shown in Figure 6 In the task hierarchical information product generation module, the sensor information obtained from the aircraft cluster edge computing platform and the forward data sent by the external task system are processed according to the service quality evaluation results obtained from the aircraft cluster edge computing platform to generate hierarchical information products of corresponding levels. The specific method is to decompose and map the forward data of the external aircraft task system into atomic tasks; according to the atomic tasks and the service quality evaluation results, the sensor information and the forward data are processed for frequency reduction, lossless compression, or lossy compression to output hierarchical information products.

[0058] As shown in Figure 6As shown, the task hierarchical information product generation module includes four parts: task requirement decomposition module, data frequency reduction processing module, data compression processing module and hierarchical product output module. The main function is to generate hierarchical information products of corresponding levels according to the network link conditions provided by the service quality evaluation model module, so as to adapt to the requirements of network transmission link, for the sensor data generated by the sensor agent module and the forward data generated by the external task system. The task requirement decomposition module decomposes and maps the task requirements of the aircraft task system into atomic tasks; the data frequency reduction processing module performs frequency reduction processing on the sensor data, task system forward data and the like, so as to reduce the amount of network transmission information; the data compression processing module performs lossless compression and lossy compression processing on the sensor data, task system forward data and the like, so as to adapt to the network transmission link; and the hierarchical product output module outputs appropriate hierarchical information products according to the task requirements and service quality requirements.

[0059] As shown in Figure 7 The service quality evaluation model module includes a service quality collection module, a producer QoS policy generation module, a consumer QoS policy generation module and a pre-trained model transfer learning module. The service quality collection module periodically reads network state parameters from the aircraft cluster edge computing platform; the producer QoS policy generation module is responsible for generating the QoS policy of the data producer; the consumer QoS policy generation module is responsible for generating the matching consumer QoS policy through coordination with the producer QoS policy generation module, to ensure the strategy matching of the information sending and receiving parties; and the pre-trained model transfer learning module fine-tunes the service quality evaluation model to generate the service quality evaluation result, solves the system initialization problem in the case of sample shortage, and improves the dynamic adaptability of information service capability to task scenarios.

[0060] As shown in Figure 8The information distribution module includes a distribution information reading module, an information distribution mode selection module, an information distribution priority adjustment module and a distribution information sending module, and mainly functions to increase multi-level and multi-priority distribution mechanisms based on three information service modes of request response, subscription notification and active push, automatically select a suitable data product level in combination with an information distribution strategy output by a service quality evaluation model module, adjust a priority of to-be-sent data, and provide efficient and accurate hierarchical information service for a task. The distribution information reading module reads hierarchical information products from a spacecraft cluster edge computing platform. The information distribution priority adjustment module adjusts a priority of the hierarchical information products according to an information distribution strategy sent by the service quality evaluation model module, and generates priority hierarchical information products. The information distribution mode selection module selects a corresponding information distribution mode from the three information service modes of request response, subscription notification and active push according to a task system demand. The information distribution priority adjustment module automatically adjusts a priority of information according to a link state provided by the service quality evaluation model and an importance of information to a task. The distribution information sending module generates network sending data according to the information distribution mode and the priority hierarchical information products, sends the network sending data to the spacecraft cluster edge computing platform, receives service protocol data sent by the spacecraft cluster edge computing platform and forwards the service protocol data to an external spacecraft task system.

[0061] Although the present application has been described in detail by the preferred embodiments above, it should be appreciated that the above description should not be considered as limiting the present application. Various modifications and substitutions to the present application will be apparent to those skilled in the art after having read and understood the above description. Therefore, the scope of the present application should be defined by the appended claims.

[0062] The contents not described in detail in the specification of the present application are known to those skilled in the art.

Claims

1. An edge-computing-based hierarchical information service system, characterized by comprising: Comprise: An aircraft cluster edge computing platform, a sensor agent module, an actuator agent module, an information gateway module, a task hierarchical information product generation module, a service quality evaluation model module, and an information distribution module, wherein: The aircraft cluster edge computing platform: monitors the network state to generate network state parameters; combines and arranges a plurality of external service plug-ins; receives sensor information sent by the sensor agent module; sends actuator control instructions to the actuator agent module; receives hierarchical information products of corresponding levels sent by the task hierarchical information product generation module, and sends sensor information to the task hierarchical information product generation module; receives service quality evaluation results sent by the service quality evaluation model module, and sends network state parameters to the service quality evaluation model module; receives business protocol data sent by the information gateway module, and sends network transmission data to the information gateway module; receives network transmission data sent by the information distribution module, and sends hierarchical information products and business protocol data to the information distribution module; The sensor agent module: receives sensor data of external sensors, analyzes the sensor data, generates sensor information, and sends it to the aircraft cluster edge computing platform; The actuator agent module: converts the actuator control instructions sent by the aircraft cluster edge computing platform into actuator instruction codes according to the actuator interface protocol, and sends them to external actuators; The information gateway module: converts business data sent by external communication devices into business protocol data and sends it to the aircraft cluster edge computing platform; converts network transmission data sent by the aircraft cluster edge computing platform into protocol and sends it to external communication devices; The task hierarchical information product generation module: processes sensor information obtained from the aircraft cluster edge computing platform and forward data sent by an external task system based on service quality evaluation results obtained from the aircraft cluster edge computing platform, generates hierarchical information products of corresponding levels, and sends them to the aircraft cluster edge computing platform to meet the requirements of network transmission links; The service quality evaluation model module: models network service quality over a period of time based on quality characteristic evaluation methods, generates information distribution strategies that match data producers and data consumers, sends the information distribution strategies to the information distribution module, fine-tunes the service quality evaluation model based on pre-trained model transfer learning, generates service quality evaluation results, sends the service quality evaluation results to the aircraft cluster edge computing platform, and receives network state parameters sent by the aircraft cluster edge computing platform; The information distribution module: receives hierarchical information products sent by the aircraft cluster edge computing platform based on information service modes and information distribution strategies sent by the service quality evaluation model module, adjusts the priority of the hierarchical information products, generates network transmission data, and sends the network transmission data to the aircraft cluster edge computing platform; receives business protocol data sent by the aircraft cluster edge computing platform and forwards it to an external aircraft task system; The task hierarchical information product generation module comprises a task demand decomposition module, a data frequency reduction processing module, a data compression processing module and a hierarchical product output module; the task demand decomposition module decomposes and maps the task demand of the spacecraft task system into atomic tasks; the data frequency reduction processing module performs frequency reduction processing on sensor information and forward data to reduce the amount of network transmission information; the data compression processing module performs lossless compression or lossy compression processing on the sensor information and the forward data; and the hierarchical product output module outputs corresponding hierarchical information products according to the task requirements and the quality of service requirements. The service quality evaluation model module comprises a service quality collection module, a producer QoS policy generation module, a consumer QoS policy generation module and a pre-trained model transfer learning module; the service quality collection module periodically reads network state parameters from the spacecraft cluster edge computing platform; the producer QoS policy generation module is responsible for generating the QoS policy of the data producer; the consumer QoS policy generation module is responsible for generating a matching consumer QoS policy through coordination with the producer QoS policy generation module, to ensure the strategy matching of the information sending and receiving parties; and the pre-trained model transfer learning module fine-tunes the service quality evaluation model to generate a service quality evaluation result. 2.The hierarchical information service system based on edge computing according to claim 1, characterized in that: The information distribution strategy comprises direct distribution, frequency reduction distribution, compression distribution and multi-node collaborative distribution. 3.The hierarchical information service system based on edge computing of claim 1, wherein: The information service mode comprises request and response, subscription notification and active push.

4. The hierarchical information service system based on edge computing according to claim 1, characterized in that: The network service quality comprises distribution real-time performance, data validity, system resource overhead, data transmission control and data completeness.

5. The hierarchical information service system based on edge computing according to claim 1, characterized in that: The spacecraft cluster edge computing platform comprises a resource management module, a service orchestration module and a basic algorithm library; wherein: The resource management module receives sensor information sent by the external sensor agent module and receives service protocol data sent by the information gateway module; sends network sending data to the information gateway module; calls the basic algorithm library to send an actuator control instruction to the actuator agent module, to realize unified management of the sensor information, the service protocol data, the network sending data and the control instruction; the service orchestration module combines and arranges various external service plug-ins according to the service quality evaluation result sent by the service quality evaluation model module and the hierarchical information product of the task hierarchical information product generation module; the basic algorithm library provides a space-time reference alignment and an intelligent algorithm engine function; and the resource management module monitors the network state to generate network state parameters and sends the network state parameters to the service quality evaluation model module. 6.The hierarchical information service system based on edge computing of claim 1, wherein: The sensor agent module comprises a sensor data transceiver module and a sensor data analysis module; the sensor data transceiver module receives sensor data from an external sensor; and the sensor data analysis module analyzes the sensor data according to a sensor interface protocol to generate sensor information. 7.The hierarchical information service system based on edge computing of claim 1, wherein: The actuator agent module comprises an actuator interface module and an actuator instruction conversion module; the actuator interface module is responsible for connection and interface interaction with an external actuator; the actuator instruction conversion module converts actuator control instructions into executable instruction codes of the actuator according to an interface protocol of the actuator, and sends the executable instruction codes to the external actuator. 8.The hierarchical information service system based on edge computing of claim 1, wherein: The information gateway module comprises a network information receiving module, a network protocol conversion module and a network information sending module; the network information receiving module receives service data from an external communication device; the network protocol conversion module completes analysis and protocol conversion on the service data format according to a protocol, generates service protocol data, and sends the service protocol data to the aircraft cluster edge computing platform; the network information sending module reads network sending data from the aircraft cluster edge computing platform, sends the network sending data to the external communication device after protocol conversion. 9.The hierarchical information service system based on edge computing of claim 1, wherein: According to the service quality evaluation result obtained from the aircraft cluster edge computing platform, the sensor information obtained from the aircraft cluster edge computing platform and the forward data sent by the external task system are processed to generate corresponding hierarchical information products, and the specific method is as follows: The forward data of the external aircraft task system is decomposed and mapped into an atomic task; According to the atomic task and the service quality evaluation result, the sensor information and the forward data are processed by frequency reduction, lossless compression or lossy compression to output hierarchical information products. 10.The hierarchical information service system based on edge computing of claim 1, wherein: The information distribution module comprises a distribution information reading module, an information distribution mode selection module, an information distribution priority adjustment module and a distribution information sending module; wherein: The distribution information reading module reads hierarchical information products from the aircraft cluster edge computing platform; The information distribution priority adjustment module adjusts the priority of the hierarchical information products according to the information distribution strategy sent by the service quality evaluation model module to generate priority hierarchical information products; The information distribution mode selection module selects a corresponding information distribution mode according to the information service mode; The distribution information sending module generates network sending data according to the information distribution mode and the priority hierarchical information products; sends the network sending data to the aircraft cluster edge computing platform; receives the service protocol data sent by the aircraft cluster edge computing platform and forwards the service protocol data to the external aircraft task system.

Citation Information

Patent Citations

  • Unmanned aerial vehicle remote measuring and control system and method based on mobile communication network

    CN104950906A

  • Resource scheduling system and method based on edge computing in heterogeneous environment

    CN111858054A