CN-122024446-A - Intelligent life support and risk early warning method and system for closed circulation submersible

CN122024446ACN 122024446 ACN122024446 ACN 122024446ACN-122024446-A

Abstract

The application relates to a diving safety management and control technology, in particular to an intelligent life support and risk early warning method and system for a closed circulation submersible, comprising the steps of collecting multisource perception data of the closed circulation submersible in real time; the method comprises the steps of performing timestamp alignment, filtering and calibration on data, integrating the data into a unified state matrix, inputting the matrix into a risk prediction algorithm model, dynamically predicting the change trend of oxygen partial pressure and carbon dioxide partial pressure, evaluating the physiological load level, operating a multi-risk coupling analysis model based on the prediction trend, evaluating the probability and severity of risks such as hypoxia, high carbon dioxide poisoning, physiological overload and the like, executing a grading early warning strategy according to the risk probability, outputting early warning information to a diver, recording response behaviors and parameter changes of the diver, evaluating early warning effectiveness and adjusting model parameters. According to the application, through multi-mode sensing fusion and an intelligent algorithm, accurate life support and dynamic risk prevention and control are realized, and the safety and rescue efficiency of underwater operation are remarkably improved.

Inventors

FANG GUANGMING
HOU XUAN
CAI YOUQUAN
PENG ZHIBING
LI LONGFEI
WU JIAWEI

Assignees

中国人民解放军陆军特种作战学院

Dates

Publication Date: 20260512
Application Date: 20260210

Claims (10)

1. The intelligent life support and risk early warning method for the closed-cycle submersible is characterized by comprising the following steps of: Acquiring multisource sensing data of the closed cycle submersible in real time, wherein the multisource sensing data comprise gas component data in a breathing loop, loop pressure data, diver vital sign data and environment and equipment state data; performing time stamp alignment, filtering and calibration on the multi-source sensing data to obtain processed multi-source sensing data, and integrating the processed multi-source sensing data into a unified state matrix; Inputting the unified state matrix into a risk prediction algorithm model, dynamically predicting the change trend of oxygen partial pressure and carbon dioxide partial pressure in the short time in the future, and evaluating the physiological load level of the diver; Based on the predicted oxygen partial pressure trend, carbon dioxide partial pressure trend and physiological load level, running a multi-risk coupling analysis model to evaluate the probability and severity of at least one risk of hypoxia, high carbon dioxide poisoning and physiological overload; and executing a grading early warning strategy according to the risk probability, and outputting early warning information to the diver based on the grading early warning strategy.
2. The intelligent life support and risk early warning method for a closed-cycle submersible according to claim 1, wherein the performing timestamp alignment, filtering and calibration processing on the multi-source sensing data to obtain processed multi-source sensing data, and integrating the processed multi-source sensing data into a unified state matrix specifically comprises: Performing time stamp alignment on the acquired multisource sensing data, synchronizing all the data to the same time reference, and performing filtering processing on the aligned data to remove noise; And fusing the filtered multi-source data to generate a unified state matrix, wherein the state matrix comprises gas parameters, pressure parameters, physiological parameters and environmental parameters, and the unified state matrix is stored for subsequent risk prediction and analysis.
3. The intelligent life support and risk early warning method for a closed-cycle submersible according to claim 1, wherein the step of inputting the unified state matrix into a risk prediction algorithm model dynamically predicts the trend of the oxygen partial pressure and the carbon dioxide partial pressure in a short time in the future and evaluates the physiological load level of the submersible, specifically comprises the steps of: extracting real-time oxygen partial pressure, carbon dioxide partial pressure, respiratory rate and heart rate data from the unified state matrix as input of a aerodynamic model; Based on the law of conservation of mass and the respiratory metabolism rate, predicting the change track of the partial pressure of oxygen and the partial pressure of carbon dioxide in future time by using a aerodynamic model; Acquiring depth, water temperature and exercise intensity data from the unified state matrix, inputting a physiological load assessment model, and calculating a physiological stress index of a diver; and integrating the gas prediction track and the physiological stress index, and outputting a comprehensive risk trend report comprising future values and confidence intervals of the key parameters.
4. The intelligent life support and risk early warning method for a closed cycle submersible according to claim 3, wherein the running of the multi-risk coupled analysis model based on the predicted oxygen partial pressure trend, carbon dioxide partial pressure trend and physiological load level, the assessment of the probability and severity of at least one risk of hypoxia, high carbon dioxide poisoning, physiological overload, specifically comprises: Based on the comprehensive risk trend report, identifying at least one risk signal in the oxygen partial pressure decreasing trend, the carbon dioxide partial pressure increasing trend or the physiological stress index overrun; based on the historical data and the real-time state, calculating the occurrence probability and the coupling effect of each risk signal, wherein the coupling effect comprises that the high carbon dioxide level aggravates the risk of oxygen poisoning; And evaluating the overall risk level according to the probability and the coupling effect, quantifying the severity, and generating a risk evaluation result comprising a risk type list, a probability value and a recommended intervention time window.
5. The intelligent life support and risk early warning method for a closed cycle submersible according to claim 1, wherein the step of executing a hierarchical early warning strategy according to the risk probability and outputting early warning information to a diver based on the hierarchical early warning strategy specifically comprises: based on the risk probability and a preset safety threshold envelope curve, determining whether the risk exceeds the limit; If the probability is higher than the first threshold value or the single parameter approaches the safety threshold value, judging that the risk probability is higher than the safety threshold value, judging that the risk probability is lower than the first threshold value and the severity is lower than the severity, judging that the risk probability is higher than the safety threshold value, judging that the risk probability is lower than the third-level serious alarm; And generating specific early warning contents including risk descriptions, expected occurrence time and treatment suggestions according to the early warning level.
6. The intelligent life support and risk early warning method for a closed-cycle submersible according to claim 1, wherein after the step of executing a step early warning strategy according to the risk level, outputting early warning information to a diver based on the step early warning strategy, the intelligent life support and risk early warning method for a closed-cycle submersible further comprises: recording response behaviors of divers to early warning and parameter change data of the closed-cycle submersible after early warning, and evaluating early warning effectiveness based on the recorded response behaviors and parameter changes; And according to the evaluation result, adjusting parameters of the risk prediction algorithm model, and updating a threshold value of the early warning grading logic.
7. The intelligent life support and risk early warning method for a closed cycle submersible according to claim 6, wherein the generating specific early warning content including risk description, expected time of occurrence and treatment advice according to the early warning level specifically comprises: Analyzing the early warning level and content, and if the early warning level and content are primary prompts, driving a visual display unit to display green icons and trend curves on a helmet display; if the warning is a secondary warning, triggering a visual display yellow flashing icon, an audible prompt tone and voice broadcasting; if the alarm is three-level, a visual red alarm, a high-frequency audible alarm and a strong tactile vibration are started, and the non-key information display is interrupted preferentially.
8. An intelligent life support and risk early warning system for a closed-cycle submersible, the intelligent life support and risk early warning system for a closed-cycle submersible comprising: The data acquisition module is used for acquiring multisource sensing data of the closed-cycle submersible in real time, wherein the multisource sensing data comprise gas composition data in a breathing loop, loop pressure data, diver vital sign data and environment and equipment state data; The multi-source perception data processing module is used for carrying out time stamp alignment, filtering and calibration processing on the multi-source perception data to obtain processed multi-source perception data, and integrating the processed multi-source perception data into a unified state matrix; the risk prediction module is used for inputting the unified state matrix into a risk prediction algorithm model, dynamically predicting the change trend of oxygen partial pressure and carbon dioxide partial pressure in short time in the future, and evaluating the physiological load level of the diver; The risk level assessment module is used for running the multi-risk coupling analysis model based on the predicted oxygen partial pressure trend, the carbon dioxide partial pressure trend and the physiological load level, and assessing the probability and the severity of at least one risk of hypoxia, high carbon dioxide poisoning and physiological overload; And the hierarchical early warning strategy generation module is used for executing a hierarchical early warning strategy according to the risk probability and outputting early warning information to the diver based on the hierarchical early warning strategy.
9. An electronic device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that the processor, when executing the computer program, carries out the steps of a smart life support and risk early warning method for a closed cycle submersible according to any one of claims 1 to 7.
10. A computer readable storage medium storing a computer program, characterized in that the computer program when executed by a processor implements the steps of an intelligent life support and risk early warning method for a closed cycle submersible according to any one of claims 1 to 7.

Description

Intelligent life support and risk early warning method and system for closed circulation submersible Technical Field The invention relates to the technical field of diving safety control, in particular to an intelligent life support and risk early warning method and system for a closed circulation submersible. Background With the rapid growth of marine resource exploration and emergency rescue requirements, closed cycle submersible vehicles are increasingly used in the military, scientific and commercial fields. The submersible can obviously prolong the underwater operation time by recycling the breathing gas, but the sealing of the internal environment of the submersible also brings unique safety challenges. Currently, underwater operations are proceeding toward deep hydration and long-term development, and physiological risk factors faced by divers are increasingly complex, including multiple threats such as unbalanced gas components, abnormal pressure, physiological overload and the like. The traditional life support method mainly relies on diver experience judgment and regular equipment inspection, and has obvious defects. Gas monitoring has mostly employed discrete sensors, such as electrochemical oxygen sensors and infrared carbon dioxide sensors, which typically work independently, data presentation fragmentation, and lack system level integration. The physiological monitoring is often limited to basic heart rate detection, and the comprehensive physiological load state of a diver is difficult to reflect in real time, so that in conclusion, the existing closed-cycle submersible life support technology has inherent defects of monitoring on one side, early warning lag, poor adaptability and the like, and therefore, a certain improvement space exists. Disclosure of Invention In order to solve the technical problems, the application provides an intelligent life support and risk early warning method and system for a closed-cycle submersible. The first object of the present application is achieved by the following technical solutions: An intelligent life support and risk early warning method for a closed-cycle submersible comprises the following steps: Acquiring multisource sensing data of the closed cycle submersible in real time, wherein the multisource sensing data comprise gas component data in a breathing loop, loop pressure data, diver vital sign data and environment and equipment state data; performing time stamp alignment, filtering and calibration on the multi-source sensing data to obtain processed multi-source sensing data, and integrating the processed multi-source sensing data into a unified state matrix; Inputting the unified state matrix into a risk prediction algorithm model, dynamically predicting the change trend of oxygen partial pressure and carbon dioxide partial pressure in the short time in the future, and evaluating the physiological load level of the diver; Based on the predicted oxygen partial pressure trend, carbon dioxide partial pressure trend and physiological load level, running a multi-risk coupling analysis model to evaluate the probability and severity of at least one risk of hypoxia, high carbon dioxide poisoning and physiological overload; and executing a grading early warning strategy according to the risk probability, and outputting early warning information to the diver based on the grading early warning strategy. By adopting the technical scheme, the integrity of data coverage is ensured through real-time acquisition of multisource sensing data (such as oxygen partial pressure, heart rate, depth and the like), risk blind spots caused by missed detection of a single sensor are avoided, so that the reliability and the real-time performance of monitoring are improved, secondly, the asynchronism and noise of multisource data are eliminated through time stamp alignment, filtering and calibration in the data processing step, a unified state matrix is generated, clean and consistent input is provided for risk prediction, the false alarm rate of a model is reduced, then, a risk prediction algorithm dynamically predicts the trend of oxygen partial pressure and carbon dioxide partial pressure based on aerodynamic and physiological models and combines physiological load evaluation, the conversion from passive monitoring to active early warning is realized, the foresight and the accuracy of a system are enhanced, then, the multiple risk coupling analysis model evaluates the coupling effects of risks such as hypoxia and high carbon dioxide poisoning, the quantification probability and the severity are avoided, the limitation of single risk analysis is avoided, the comprehensiveness and scientificity of early warning are improved, finally, the multi-mode early warning strategy is output according to the grades, the response of multi-mode early warning information is ensured, and meanwhile, the response of a self-adaption is reduced through the self-adaption, the complexity of the