CN-121979714-A - Intelligent Beidou equipment parallel detection and maintenance system

CN121979714ACN 121979714 ACN121979714 ACN 121979714ACN-121979714-A

Abstract

The invention relates to the technical field of equipment detection and maintenance, in particular to an intelligent Beidou equipment parallel detection and maintenance system, which aims to solve the problems of stiff protocol adaptation, unreasonable parallel scheduling, no self-evolution capability in fault diagnosis, insufficient data safety and the like in the prior art. The system comprises a three-proofing flat hardware carrier with a hardware compatibility adaptation layer and overhaul software integrating multiple creative modules, wherein the protocol self-adaptive learning engine is used for realizing the automatic adaptation of a new customized protocol, the dynamic core scheduling and the resource elastic allocation are adopted for guaranteeing the parallel stability of multiple machines, the fault diagnosis accuracy is improved based on an FT-BN self-evolution mechanism, the cloud edge collaborative backup and the off-line multi-mode interaction suite are combined for widening the application scene, and the maintenance flow is optimized through a multiple fault conflict resolution algorithm and a visual auxiliary verification module. The system supports parallel access of multiple Beidou equipment, realizes zero-base accurate maintenance, adapts to multiple types of multiprotocol, and has the advantages of high efficiency, reliability and universality in overhauling of military Beidou equipment.

Inventors

JIN CHENG
BAO PANPAN
TIAN YASU
CHEN LINGLING
MA CHEN
WANG JIPENG
WANG JIAHAO
Cao Yizhuo
LI KAILI
ZHANG WENJING

Assignees

西安航光卫星测控技术有限公司

Dates

Publication Date: 20260505
Application Date: 20260126

Claims (10)

1. An intelligent Beidou equipment parallel detection and maintenance system is characterized by comprising a three-prevention flat-plate hardware carrier and Beidou equipment intelligent maintenance software; The three-proofing flat board is provided with a domestic Galaxy kylin operating system and is provided with a hardware compatibility adapting layer, a multi-path communication interface and 3 paths of aviation interfaces, the RS232 aviation interfaces can be converted into 5 paths of serial ports, 5 north bucket equipment is supported to be accessed in parallel, and the baud rate of each path of serial port is 1200-115200bps; The overhaul software comprises a serial port automatic detection and communication module, an equipment fault detection module, a maintenance operation guiding module, a data storage module, a man-machine interaction module, a protocol self-adaptive learning engine and a visual auxiliary checking module, wherein task intelligent allocation is realized through a heterogeneous computing architecture of a multi-core CPU and a GPU, a dynamic core scheduling and resource elastic allocation algorithm, a fault tree-Bayesian network fusion engine and a self-evolution mechanism are adopted to position faults, operation guidance is realized through multi-mode interaction and an off-line suite, and data safety is ensured by combining cloud edge collaborative backup and hierarchical storage; The serial port automatic detection and communication module is internally provided with a standard protocol library and a protocol self-adaptive learning engine, the maintenance operation guidance module integrates a multi-fault conflict resolution algorithm, and the vision auxiliary verification module verifies the repair state of the structural member through image recognition.
2. The intelligent Beidou equipment parallel detection and maintenance system of claim 1 is characterized in that the protocol self-adaptive learning engine comprises a protocol feature extraction unit, a frame structure analysis unit and a learning model training unit, wherein an initiator, a check rule and data segment format features are extracted by capturing unknown protocol data frames, a analysis function is trained based on a CNN deep learning model, equipment identifiers are automatically bound and a protocol library is updated, and a new customized protocol is adapted without manual intervention.
3. The intelligent Beidou equipment parallel detection and maintenance system of claim 1 is characterized in that the dynamic core scheduling and resource elastic allocation algorithm dynamically adjusts the core allocation quantity based on a real-time task complexity evaluation value and CPU core occupancy rate, sets a core resource pool and a standby queue, and allocates cores by adopting a time slice rotation mechanism when equipment with the same priority requests resources.
4. The intelligent Beidou equipment parallel detection and maintenance system of claim 1 is characterized in that the fault tree-Bayesian network self-evolution mechanism comprises a fault case acquisition unit, a probability updating unit and a fault tree expansion unit, new fault cases are automatically acquired, a Bayesian network condition probability table is updated after manual auditing, and event nodes are newly added and fault tree logic relations are optimized.
5. The intelligent Beidou equipment parallel detection and maintenance system of claim 1 is characterized in that the data storage module adopts a cloud-edge collaborative backup and hierarchical storage architecture, an edge-end InfluxDB stores real-time detection logs, a dream DM8 stores recent maintenance data, a cloud synchronous backup full-scale data and a hot data caching and cold data archiving strategy.
6. The intelligent Beidou equipment parallel detection and maintenance system of claim 1 is characterized in that the man-machine interaction module integrates an off-line multi-mode interaction suite, pre-stores a three-dimensional model, images and texts, voice packets and maintenance videos, automatically switches off-line modes in a network-free environment, and maintains a rendering frame rate of more than 30 fps.
7. The intelligent Beidou equipment parallel detection and maintenance system of claim 1 is characterized in that the multi-fault conflict resolution algorithm builds a priority matrix based on fault severity, maintenance operation dependency and component relevance, and reorders maintenance steps of conflict faults.
8. The intelligent Beidou equipment parallel detection and maintenance system of claim 1 is characterized in that the hardware compatibility adaptation layer comprises interface conversion drivers, level adaptation units and performance adaptation modules, and is compatible with serial port chips of different brands/batches of three-prevention flat plates and CPU/GPU configuration.
9. The intelligent Beidou equipment parallel detection and maintenance system of claim 1 is characterized in that the vision auxiliary verification module adopts YOLOv target detection algorithm, acquires structural member repair images through a three-proofing flat-panel camera, and compares the structural member repair images with standard images to output objective verification results.
10. The intelligent Beidou equipment parallel detection and maintenance system of claim 1 is characterized in that the equipment fault detection module disassembles a detection flow by adopting a task slicing technology, processes multiple serial port data streams by an asynchronous non-blocking I/O model and a multithreaded polling mechanism, polls for 10ms at intervals, and adopts mutual exclusion lock among threads to avoid resource conflict.

Description

Intelligent Beidou equipment parallel detection and maintenance system Technical Field The invention relates to the technical field of equipment detection and maintenance, in particular to an intelligent Beidou equipment parallel detection and maintenance system. Background The Beidou equipment is used as core supporting equipment in military operations and national defense construction, and plays an irreplaceable role in key scenes such as positioning navigation, emergency communication, accurate time service and the like. Along with the continuous upgrading of the Beidou system, the model iteration speed of Beidou equipment is increased, a multi-generation equipment system covering Beidou I, II and III is formed at present, and meanwhile, various military manufacturers push out a large number of special equipment with customized protocols according to specific combat requirements, so that the technical state of the equipment is in a complex situation of multiple models, miscellaneous protocols, different components and low system installation. The existing Beidou equipment maintenance technology has a plurality of outstanding problems that firstly, the protocol adaptation capability is rigidified, the existing detection tool can only adapt to a few preset standard protocols, the customized protocols of new manufacturers cannot be identified at all, an analysis program is required to be written manually, the adaptation period is long, the cost is high, and the rapid formation fight of new listed equipment is seriously influenced; secondly, the parallel detection scheduling is unreasonable, the existing system adopts a fixed core allocation mode, the core allocation proportion of complex tasks and simple tasks is solidified, when a plurality of equipment simultaneously carry out complex detection tasks, CPU core resource gaps are extremely easy to appear, detection is blocked, and resource waste is caused when the simple tasks occupy cores, the batch detection efficiency is low, thirdly, the fault diagnosis lacks self-evolution capability, the existing fault diagnosis model relies on initial experiment statistical data, novel faults caused by equipment aging and environmental changes can not be identified, the diagnosis accuracy gradually decreases along with the service time, the fault tree structure is fixed and cannot be dynamically expanded, fourthly, the data storage and backup are provided with a short board, when the existing double-database architecture carries out full-load detection on 5 equipment, the write-in rate of the detection log is close to the upper limit of InfluxDB, write-in blocking is easy to appear after the maintenance operation log is overlapped, and only local storage is relied on, the non-local backup mechanism is not needed, once a three-protection flat plate is damaged, key maintenance data and fault cases are permanently lost, fifthly, the environment adaptability is poor, the existing system is rendered according to the network environment, the functions such as voice, the three-dimensional model of the existing system can not be expanded, the existing system can not be dynamically expanded, the existing system can not generate a plurality of fault scenes when the existing system is in a field maintenance sequence, and can not normally generate a plurality of fault conditions, and can not normally occur when the fault scene is generated in a plurality of network has a conflict in a network service sequence, the system has the advantages that the dependence relationship between faults and the relevance of components are not considered, the secondary damage is easily caused by the collision of maintenance steps, the hardware compatibility is insufficient, the existing system depends on three-proofing plates with specific configuration, after plates with different brands or batches are replaced, the serial port data conversion and DMA transmission stability is reduced, even communication cannot be established, the fault verification is unreveable, and for structural member faults such as antenna cover damage, silica gel key failure and the like, only maintenance personnel are relied on to subjectively judge the repair results, misjudgment is easily caused, and the knowledge base data distortion is caused. In addition, the existing maintenance tool still has the problems of high upper threshold and complex operation, most detection equipment needs long-term training of professional technicians to be used skillfully, and a basic maintenance unit generally lacks high-level professional talents, so that a large number of fault equipment cannot be repaired in time, and the equipment combat readiness rate is affected. Therefore, research and development of a Beidou equipment parallel detection and maintenance system with high compatibility, high intelligence, high reliability and strong environmental adaptability has become an urgent need for solving the current Beidou equipment maint