CN-121787915-B - Hoisting risk early warning method and system for hoisting unmanned aerial vehicle

CN121787915BCN 121787915 BCN121787915 BCN 121787915BCN-121787915-B

Abstract

The invention relates to the technical field of safety early warning, in particular to a lifting risk early warning method and system for a lifting unmanned aerial vehicle, comprising the steps of analyzing the real-time stress state and stability of a lifting object and evaluating the risk level of the lifting object; dynamically calculating weight coefficients of personnel risks, equipment risks and lifting article risks; weighting and fusing are carried out based on the determined dynamic weight coefficient, and a comprehensive risk index is generated; the invention introduces a dynamic weight calculation mechanism based on multidimensional real-time cost factors, carries out quantitative fusion on safety cost, article value cost, equipment loss cost, energy consumption cost and time cost, and adjusts the weight ratio of different risk categories in real time according to the quantitative fusion; according to the scheme, the risk perception center of gravity of the whole early warning system can be intelligently moved along with an operation scene, the risk perception center of gravity is more conserved under a severe environment to ensure safety, the integrity is more concerned when valuables are lifted, and intelligent self-adaption and flexible decision of a risk assessment strategy are realized.

Inventors

XIE QINGLIANG
LI ZHUO
SHEN JUNHAO
ZHANG CHENG
XU MAOCHUN
GONG JINJUN
ZHENG GUANGFU
ZHANG CHAOYUE
XU BOYANG

Assignees

中建三局第二建设(深圳)有限公司
中建三局集团有限公司
中建三局第二建设工程(广东)有限责任公司

Dates

Publication Date: 20260508
Application Date: 20260305

Claims (7)

1. A hoisting risk early warning method for hoisting an unmanned aerial vehicle is characterized by comprising the following steps: s1, monitoring stress data of a plurality of groups of hoisting ropes in a hoisting system and attitude data of hoisted articles in real time; S2, analyzing the real-time stress state and stability of the hoisted article based on the obtained stress data and attitude data, and evaluating the risk level of the hoisted article; S3, dynamically calculating weight coefficients of personnel risks, equipment risks and lifting article risks according to the real-time operation environment and task parameters; S4, combining a lifting article risk assessment result, comprehensively assessing personnel risks and equipment risks, and carrying out weighted fusion based on the determined dynamic weight coefficient to generate a comprehensive risk index; s5, constructing a risk causal graph by adopting a causal discovery algorithm based on historical monitoring data, a control instruction sequence and a risk event record, and predicting the root cause of the current risk state; S6, deducing a preset control program and a control instruction input in real time by using the obtained risk and cause graph, and predicting potential risks in a future time period; S7, triggering a risk early warning signal of a corresponding grade according to the comprehensive risk index and the future risk prediction result, and outputting a control adjustment suggestion according to the early warning grade; wherein, when evaluating hoist and mount article risk level, specifically include: s21, calculating resultant force, resultant moment and stress unbalance parameters of the hoisted objects based on real-time stress data of each hoisting cable; s22, analyzing vibration characteristics, attitude stability and torsion angle parameters of the hoisted article based on the acquired attitude data of the hoisted article; s23, carrying out coupling analysis on the stress unbalance parameter and the attitude stability parameter, and identifying a instability mode of the hoisted article; S24, calculating a risk factor of the hoisted article by adopting a risk quantization model based on the instability mode; s25, dynamically adjusting a risk level threshold interval according to the material property, the structural property and the current environmental condition of the hoisted article, judging risk factors of the hoisted article according to the risk level threshold interval, and judging a specific risk level; wherein, when the stress unbalance parameter and the gesture stability parameter are subjected to coupling analysis, when the unstability mode of the hoisted article is identified, the method specifically comprises the following steps: S231, combining the stress unbalance parameter, the reciprocal of the attitude stability parameter, the dangerous vibration energy duty ratio parameter and the torsion angle parameter to construct a coupling analysis feature vector; S232, respectively carrying out matching degree calculation on the coupling analysis feature vector and a group of preset destabilization mode feature vectors; S233, determining a destabilization mode with highest matching degree with the current state according to the calculated matching degree, and identifying the destabilization mode as a current main destabilization mode; when dynamically calculating weight coefficients of personnel risks, equipment risks and lifting article risks according to real-time operation environment and task parameters, the method specifically comprises the following steps: S31, acquiring environmental parameters of current operation, task attribute parameters of a hoisting task, state parameters of unmanned aerial vehicle equipment and task constraint parameters from an external system in real time; S32, respectively calculating a safety cost factor, a hoisting object cost factor, an equipment loss cost factor, an energy consumption cost factor and a time cost factor based on various parameters; S33, converting the safety cost factor, the hoisting article cost factor, the equipment loss cost factor, the energy consumption cost factor and the time cost factor into initial values of a personnel risk weight coefficient, an equipment risk weight coefficient and a hoisting article risk weight coefficient through a preset mapping rule; and S34, carrying out normalization processing on initial values of the personnel risk weight coefficient, the equipment risk weight coefficient and the lifting article risk weight coefficient, and outputting the initial values as final dynamic weight coefficients.
2. The hoisting risk early warning method for hoisting unmanned aerial vehicle according to claim 1, wherein the real-time monitoring of stress data of a plurality of groups of hoisting cables and gesture data of hoisting objects in a hoisting system comprises the following specific steps: s11, laying a force sensor on each hoisting cable, and synchronously collecting tension data of each cable at a sampling frequency of 100 Hz; S12, installing an inertial measurement unit on the hoisted article, and collecting acceleration, angular velocity and attitude angle data of the hoisted article in real time; s13, monitoring the relative position and distance between the hoisted article and the unmanned aerial vehicle body through a ranging module arranged on the unmanned aerial vehicle.
3. The hoisting risk early warning method for hoisting the unmanned aerial vehicle according to claim 2, wherein the predetermined mapping rule is: The method comprises the steps of determining a safety cost factor, determining an initial value of a personnel risk weight coefficient, determining an initial value of a lifting article cost factor, determining an initial value of a lifting article risk weight coefficient, determining an initial value of an equipment risk weight coefficient by using an equipment loss cost factor and an energy consumption cost factor, wherein the initial value of the equipment risk weight coefficient is a weighted sum result of two cost factors, and performing global adjustment on the initial values of the personnel risk weight coefficient, the equipment risk weight coefficient and the lifting article risk weight coefficient by using a time cost factor.
4. The hoisting risk early warning method for hoisting the unmanned aerial vehicle according to claim 3, wherein the method is characterized in that when a causal map is constructed by adopting a causal discovery algorithm based on historical monitoring data, a control instruction sequence and risk event records, and a root cause of a current risk state is presumed, the method specifically comprises the following steps: s51, collecting multi-source time sequence data in a historical time window; s52, analyzing and determining the causal relationship among the data variables by adopting a causal discovery algorithm based on the multi-source time sequence data; S53, constructing a directed graph model as a risk causal graph according to the determined causal relationship; S54, based on the constructed risk and cause graph, the root cause node and the key cause and effect path which cause the current risk state are presumed by combining the current real-time monitoring data and the state.
5. The hoisting risk early warning method for hoisting unmanned aerial vehicle according to claim 4, wherein when predicting root cause nodes and key cause and effect paths which cause the current risk state by combining current real-time monitoring data and states based on the constructed risk and effect graph, the method specifically comprises the following steps: S541, mapping current real-time monitoring data and system states into evidences of corresponding observation variable nodes and control variable nodes in a risk causal graph; s542, in the risk causal graph, taking the triggered risk variable nodes as starting points, tracing back along the directed edges, and identifying all possible precursor nodes as candidate reason nodes; s543, calculating the contribution degree of each candidate cause node to the current risk state based on the causal strength and the degree of the real-time data deviating from the normal threshold value; And S544, according to the contribution degree ranking, estimating a plurality of candidate reason nodes ranked at the top and paths connecting the candidate reason nodes to the risk nodes as root causes and key cause and effect paths for causing the current risk.
6. The hoisting risk early warning method for hoisting an unmanned aerial vehicle according to claim 5, wherein when a preset control program and a control instruction input in real time are deduced by using the obtained risk-cause graph, the method specifically comprises the following steps: s61, analyzing a control program to be executed currently and to be executed in the future, capturing a control instruction input in real time, and converting the control instruction into an input sequence of a corresponding control variable node in the risk-cause-effect diagram; S62, based on the risk and causal graph, combining the current system state, and deducing the influence which is caused by the control input sequence in the future time period along the direction pointed by the causal relation in the causal graph; And S63, predicting the possibility of occurrence of a specific risk event in a future time period based on the deduction result, and outputting a risk prediction result.
7. A hoisting risk early warning system for hoisting an unmanned aerial vehicle, configured to implement the hoisting risk early warning method for hoisting an unmanned aerial vehicle according to any one of claims 1 to 6, comprising: The data acquisition module is used for monitoring stress data of a plurality of groups of hoisting ropes in the hoisting system and attitude data of the hoisted articles in real time; The lifting article risk analysis module is used for analyzing the real-time stress state and stability of the lifting article based on the obtained stress data and posture data and evaluating the risk level of the lifting article; The dynamic weight calculation module is used for dynamically calculating weight coefficients of personnel risks, equipment risks and lifting article risks according to the real-time operation environment and task parameters; the comprehensive risk assessment module is used for comprehensively and independently assessing personnel risks and equipment risks, and carrying out weighted fusion based on the dynamic weight coefficients determined by the dynamic weight calculation module to generate a comprehensive risk index; the causal analysis and root cause speculation module is used for constructing a causal map of the risk by adopting a causal discovery algorithm based on historical monitoring data, control instruction sequences and risk event records, and speculating the root cause of the current risk state; The risk prediction module is used for deducting a preset control program or a control instruction input in real time by utilizing a risk and cause graph, and predicting potential risks in a future time period; and the early warning and response module is used for triggering a risk early warning signal of a corresponding grade according to the comprehensive risk index and the future risk prediction result and outputting a control adjustment suggestion according to the early warning grade.

Description

Hoisting risk early warning method and system for hoisting unmanned aerial vehicle Technical Field The invention relates to the technical field of safety early warning, in particular to a hoisting risk early warning method and system for a hoisting unmanned aerial vehicle. Background In the unmanned aerial vehicle hoisting process, the steady state of the hoisted article directly relates to operation safety, and in case risk events such as uneven stress of a mooring rope, instability and collision of the hoisted article occur, the damage of the hoisted article and the faults of unmanned aerial vehicle equipment can be caused, and the safety of surrounding personnel can be threatened, so that the accurate and efficient hoisting risk early warning system is constructed to be one of core requirements of the development of unmanned aerial vehicle hoisting technology. The risk early warning technology in the current unmanned aerial vehicle hoisting field still has a plurality of shortages, is difficult to satisfy the safety guarantee demand of complicated operation scene. In the prior art, fixed weights are mostly adopted during risk fusion evaluation, and cannot be dynamically adjusted according to the working environment, equipment states and task demands, so that the matching degree of an evaluation result and an actual scene is not high, meanwhile, most of the prior art only can realize real-time risk early warning, lacks traceability of risk factors and prejudgment capability of future potential risks, operators only can passively cope with the occurring risks, and an effective risk prevention and control closed loop is difficult to form. Aiming at the defects of the prior art, the invention provides a hoisting risk early warning method and a hoisting risk early warning system for hoisting an unmanned aerial vehicle, wherein a multi-cost factor driven dynamic weight mechanism is constructed to improve the accuracy of risk assessment, a causal discovery algorithm and a dynamic Bayesian network are introduced to realize the root cause tracing and future prediction, and finally, a targeted control adjustment suggestion is output through a grading early warning mechanism combining a real-time risk index and a future prediction result to form a complete risk prevention and control closed loop, so that the safety and reliability of the unmanned aerial vehicle hoisting operation are effectively improved. Disclosure of Invention In order to overcome the problems in the background art, the invention provides a hoisting risk early warning method and a hoisting risk early warning system for hoisting an unmanned aerial vehicle. The technical scheme of the invention is that the hoisting risk early warning method for hoisting the unmanned aerial vehicle comprises the following steps: s1, monitoring stress data of a plurality of groups of hoisting ropes in a hoisting system and attitude data of hoisted articles in real time; S2, analyzing the real-time stress state and stability of the hoisted article based on the obtained stress data and attitude data, and evaluating the risk level of the hoisted article; S3, dynamically calculating weight coefficients of personnel risks, equipment risks and lifting article risks according to real-time operation environment and task parameters, wherein the calculation of the weight coefficients comprises safety cost, lifting article cost, equipment loss cost, energy consumption cost and time cost; S4, combining a lifting article risk assessment result, comprehensively assessing personnel risks and equipment risks, and carrying out weighted fusion based on the determined dynamic weight coefficient to generate a comprehensive risk index; s5, constructing a risk causal graph by adopting a causal discovery algorithm based on historical monitoring data, a control instruction sequence and a risk event record, and predicting the root cause of the current risk state; S6, deducing a preset control program and a control instruction input in real time by using the obtained risk and cause graph, and predicting potential risks in a future time period; And S7, triggering a risk early warning signal of a corresponding grade according to the comprehensive risk index and the future risk prediction result, and outputting a control adjustment suggestion according to the early warning grade. Preferably, the real-time monitoring of stress data of a plurality of groups of hoisting ropes and attitude data of hoisted articles in the hoisting system specifically comprises: s11, laying a force sensor on each hoisting cable, and synchronously collecting tension data of each cable at a sampling frequency of 100 Hz; S12, installing an inertial measurement unit on the hoisted article, and collecting acceleration, angular velocity and attitude angle data of the hoisted article in real time; s13, monitoring the relative position and distance between the hoisted article and the unmanned aerial vehicle body through a ran