CN-121997140-A - Space structure health monitoring method and system based on sky-ground cooperation

CN121997140ACN 121997140 ACN121997140 ACN 121997140ACN-121997140-A

Abstract

The invention discloses a space structure health monitoring method and system based on sky-ground cooperation, and relates to the technical field of space structure health monitoring. The method comprises the specific steps of respectively monitoring a space structure in three dimensions of the sky, the air and the ground to obtain multi-dimensional monitoring information, carrying out data space-time synchronization on the multi-dimensional monitoring information to obtain panoramic data, carrying out comprehensive evaluation and life evolution trend prediction on static and dynamic performance of the space structure based on ground sensing data, carrying out multi-source data feature extraction and fusion on the panoramic data to construct a comprehensive risk index evaluation model, carrying out three-layer hierarchical early warning, introducing an LLM model after early warning triggering, and automatically generating a treatment scheme according to an early warning structure. The invention deeply fuses three monitoring means of satellite remote sensing, unmanned aerial vehicle inspection and ground sensing network, and realizes omnibearing, multi-scale and multi-dimensional sensing and monitoring of a space structure in the whole construction and operation period.

Inventors

YANG LIBIN
ZHOU JUNJIE
GONG BAOJUN
XU YIBIN
ZHOU XINYI

Assignees

浙江江南工程管理股份有限公司
杭州市工程咨询中心有限公司

Dates

Publication Date: 20260508
Application Date: 20260129

Claims (10)

1. The space structure health monitoring method based on sky-ground cooperation is characterized by comprising the following specific steps: respectively monitoring the space structure in three dimensions of the sky, the air and the ground to obtain multi-dimensional monitoring information; Carrying out comprehensive evaluation and life evolution trend prediction on static and dynamic performances of the space structure based on the ground-based layer perception data; carrying out multi-source data feature extraction and fusion on the panoramic data, constructing a comprehensive risk index assessment model, and carrying out three-level grading early warning; and after the early warning is triggered, introducing an LLM model, and automatically generating a treatment scheme according to the early warning structure.
2. The method for monitoring the health of a spatial structure based on sky-ground coordination according to claim 1, wherein the specific steps of monitoring the spatial structure at a sky-base layer are as follows: Acquiring a multi-phase high-resolution image through Beidou satellite positioning and synthetic aperture radar interferometry, and finishing image preprocessing to obtain an input image; Performing building identification and boundary frame extraction on the input image by using a deep learning target detection algorithm, and performing differential analysis on detection results of the same region and different time phases to obtain a potential fluctuation range of a building; further identifying the potential variation range of the building through a pixel-level Softmax classification layer, and outputting a variation probability map; and carrying out morphological post-processing and noise suppression operation on the change probability map to obtain a change mask, and superposing the change mask on a building vector boundary around the space structure to form a dynamically updated environment evolution database.
3. The space structure health monitoring method based on sky-ground coordination according to claim 1, wherein the space structure is monitored at a sky-ground level, the method further comprises the steps of obtaining regional atmospheric environment monitoring data including temperature, humidity, air pressure, wind speed, wind direction and rainfall, sensing extreme weather events by combining the regional atmospheric environment monitoring data with a disaster prediction model, and automatically triggering high-frequency sampling of a stratum sensor and encryption inspection of an empty layer unmanned aerial vehicle when a prediction index exceeds a set threshold value to realize dynamic adjustment of a monitoring strategy.
4. The space structure health monitoring method based on sky-ground coordination according to claim 1, wherein the space structure is monitored on an empty substrate through unmanned aerial vehicle-mounted multi-sensing equipment, and the space structure health monitoring method comprises apparent disease identification and overall deformation monitoring.
5. The spatial structure health monitoring method based on sky-ground coordination is characterized by comprising the specific modes of acquiring high-density three-dimensional point cloud data by carrying laser radars on an unmanned aerial vehicle, generating a high-precision three-dimensional geometric model of a spatial structure through point cloud filtering, registering and modeling algorithms, calculating geometric offset of any time t and a reference state t 0 based on point cloud differential analysis, so that the identification of the integral camber change of a roof and the displacement of a key member is realized, and simultaneously, establishing a dynamic database of the structure form evolving along with time through multi-time phase point cloud registering.
6. The spatial structure health monitoring method based on sky-ground cooperation according to claim 4, wherein the specific way of identifying the apparent disease on the empty base layer is as follows: analyzing the high-definition image acquired by the unmanned aerial vehicle by utilizing a YOLO series target detection network, and outputting a category label of apparent diseases and a rectangular boundary box thereof in a forward propagation process, so as to realize the primary screening of the diseases in the large-area image; Inputting the candidate region output by the YOLO as a region of interest into a pixel level dividing network, and carrying out fine division on the disease boundary to generate a pixel level mask; And quantifying various apparent diseases based on the pixel-level mask.
7. The space structure health monitoring method based on sky-ground coordination according to claim 1, wherein a multi-dimensional sensing system which is jointly supported by DIC displacement monitoring, MEMS multi-parameter environment and power sensing and independent mechanical monitoring is formed on a ground layer, and unified access is realized through a wireless network.
8. The spatial structure health monitoring method based on sky-ground collaboration according to claim 1, wherein the step of constructing the comprehensive risk index evaluation model and performing three-level hierarchical early warning is as follows: unified conversion of indexes of three dimensions of the environmental threat degree E, the overall health degree H and the local safety degree S into standardized features, and construction of the comprehensive risk index assessment model: ; wherein, alpha, beta and gamma are weighting coefficients, which are determined by expert weighting or analytic hierarchy process, satisfying alpha+beta+gamma=1; The method comprises the steps of acquiring weather and remote sensing image data based on the basic level monitoring, extracting external load parameters, calculating an environmental risk index by adopting a normalization method to obtain the environmental threat degree E, forming a global health degree index by fuzzy comprehensive evaluation based on point cloud data acquired by the basic level monitoring to obtain the overall health degree H; And comparing the comprehensive risk index R with the grading threshold value to form a multi-level early warning result.
9. A space structure health monitoring system based on sky-ground cooperation is characterized by comprising a multi-dimensional monitoring module, a data processing module, a performance evaluation module, a risk early warning module and an intelligent decision module, wherein, The multi-dimensional monitoring module is used for respectively monitoring the space structure in three dimensions of the sky, the air and the ground to obtain multi-dimensional monitoring information; The data processing module is used for carrying out data space-time synchronization on the multi-dimensional monitoring information to obtain panoramic data; The performance evaluation module is used for comprehensively evaluating the static and dynamic performance of the space structure and predicting the life evolution trend based on the ground-based layer perception data structure; the risk early warning module is used for carrying out multi-source data feature extraction and fusion on the panoramic data, constructing a comprehensive risk index assessment model and carrying out three-level hierarchical early warning; The intelligent decision module is used for introducing the LLM model after the early warning is triggered, and automatically generating a treatment scheme according to the early warning structure.
10. The spatial structure health monitoring system based on sky-ground collaboration according to claim 9, wherein the multi-dimensional monitoring module, the data processing module, the performance evaluation module, the risk early warning module and the intelligent decision module are functionally integrated through a lightweight integrated platform and are connected with a multi-terminal collaborative management module.

Description

Space structure health monitoring method and system based on sky-ground cooperation Technical Field The invention relates to the technical field of space structure health monitoring, in particular to a space structure health monitoring method and system based on sky-ground cooperation. Background In recent years, along with the continuous expansion of urban construction scale, large-scale space structures such as large-span stadiums, airport terminal buildings, exhibition centers and the like are widely applied at home and abroad, and the problems of safe operation and full life cycle management are increasingly focused on various social and academia. However, such structures generally have the characteristics of large span, complex component system, special stress mechanism and susceptibility to remarkable influence from external environment, so that two prominent technical short plates still exist in the existing monitoring mode: Firstly, the existing monitoring multi-dependency sensor is used for collecting local response of structural key components, and the overall safety state of the structure is difficult to comprehensively and truly reflect. Although related researches try to infer the overall performance of a structure through local response data, in practical engineering application, the superposition of multiple factors such as temperature fluctuation, load random change, environmental noise interference and the like is extremely easy to cause the accuracy deviation of performance inference, and further the distortion of the structure safety evaluation result is caused, so that reliable basis cannot be provided for operation and maintenance decision. Secondly, most of the existing monitoring systems stay in the primary stage of data perception and acquisition, lack of intelligent analysis, structural risk prediction and operation and maintenance decision support functions for monitoring data, and do not form a complete intelligent management closed loop of perception-analysis-prediction-decision. When potential safety hazards appear in the structure, the system is difficult to early warn in advance, and the system can only passively feed back after faults occur, so that the requirements of safety control of the whole life cycle of the large-scale space structure cannot be met. It can be seen that, for those skilled in the art, there is a need in the current field of space structure health monitoring to establish a multisource sensing system with space-air-ground cooperation, and to implement accurate assessment of structural state, early risk warning and full chain management closed loop through an intelligent technology. Disclosure of Invention The invention aims to provide a space structure health monitoring method and system based on sky-ground coordination, which are used for solving the problems in the background technology, and realizing the omnibearing, multi-scale and multi-dimensional sensing and monitoring of a space structure in the whole construction and operation period by means of deep fusion of three monitoring means of satellite remote sensing, unmanned aerial vehicle inspection and a ground sensing network. In order to achieve the purpose, the invention provides the following scheme, namely, on one hand, a space structure health monitoring method based on sky-ground cooperation is provided, and the method comprises the following specific steps: respectively monitoring the space structure in three dimensions of the sky, the air and the ground to obtain multi-dimensional monitoring information; Carrying out comprehensive evaluation and life evolution trend prediction on static and dynamic performances of the space structure based on the ground-based layer perception data; carrying out multi-source data feature extraction and fusion on the panoramic data, constructing a comprehensive risk index assessment model, and carrying out three-level grading early warning; and after the early warning is triggered, introducing an LLM model, and automatically generating a treatment scheme according to the early warning structure. Preferably, the specific steps for monitoring the space structure at the sky base layer are as follows: Acquiring a multi-phase high-resolution image through Beidou satellite positioning and synthetic aperture radar interferometry, and finishing image preprocessing to obtain an input image; Performing building identification and boundary frame extraction on the input image by using a deep learning target detection algorithm, and performing differential analysis on detection results of the same region and different time phases to obtain a potential fluctuation range of a building; further identifying the potential variation range of the building through a pixel-level Softmax classification layer, and outputting a variation probability map; and carrying out morphological post-processing and noise suppression operation on the change probability map to obtain a change mask