CN-119742064-B - AI (advanced technology attachment) -recognition-based worker health condition assessment method

CN119742064BCN 119742064 BCN119742064 BCN 119742064BCN-119742064-B

Abstract

The invention relates to the technical field of engineering construction and provides a worker health condition assessment method based on AI identification, which comprises the steps of acquiring worker health data in engineering construction and building a health file; the health record is imported into a preconfigured exposure factor identification network, exposure factors in engineering construction are identified, attribution risk functions of different exposure factors are built, attribution risk values of different exposure factors relative to workers are determined according to the attribution risk functions, and a worker health condition assessment report is generated. The invention not only provides powerful support for engineering safety management, but also shows remarkable effect in solving practical problems. According to the method, sources, influence degrees and occurrence probability of various potential risk factors in the engineering construction process are deeply analyzed, so that accurate quantitative assessment of the risk degree is realized. And corresponding risk prevention and control measures are formulated according to the evaluation result, so that the possibility of accident occurrence is effectively reduced, and the safe and stable performance of engineering construction is ensured.

Inventors

ZHANG YUNFENG
WANG YI
Zong Qiangwen
XU YUNSHENG
PANG CHENG
LIAO YANG
GU BIN
WANG YUNFENG
LUO XIAO
ZHAO KUN

Assignees

三峡高科信息技术有限责任公司

Dates

Publication Date: 20260508
Application Date: 20241211

Claims (8)

1. An AI-recognition-based worker health condition assessment method, comprising: Acquiring worker health data in engineering construction, and building a health file; Importing the health file into a preconfigured exposure factor identification network, identifying exposure factors in engineering construction, and building attribution risk functions of different exposure factors; determining the attributive risk values of different exposure factors relative to workers according to the attributive risk functions, and generating a worker health condition assessment report; the constructing attribution risk functions of different exposure factors comprises the following steps: According to exposure factors, pre-defining the risk degrees of different exposure factors, and collecting exposure behaviors of corresponding exposure factors; determining the statistical risk rate of different exposure behaviors according to the exposure behaviors, wherein the statistical risk rate is the incidence rate of the different exposure behaviors in workers; calculating the attribution risk percentage of the crowd according to the statistical risk rate; calculating the relative risk of different exposure behaviors according to the crowd attribution percentage; constructing attribution risk functions of different exposure factors according to the relative risk; The specific formula of the attribution risk function is as follows: CRF=C*RT*SI*IP+C*MD*P(CI0-CI1) Wherein: c=risk coefficient, value range [0,6] Rt=risk transfer rate, value range [0,1] dimensionless Md=rate of change, value range [ -1,1] dimensionless P=potential risk possibility, value range [0,1] dimensionless CI0=safety index before change, and value range [0,1] is dimensionless CI1=safety index after change, and the value range [0,1] is dimensionless Si=safety index, value range [0,1] dimensionless IP = implementation plan, value range [0,1] dimensionless Wherein c=cy+ck+ cj+Ct+Cs+Cp Cy is a noise pollution risk coefficient, ck is an air pollution risk coefficient, cj is a building risk coefficient, ct is an equipment risk coefficient, cs is a water resource risk coefficient, and Cp is a water resource risk coefficient.
2. The AI-identification-based assessment of health of a worker of claim 1, wherein the health profile includes a personal profile and an engineering profile, wherein: the personal archive is used for acquiring physiological and biochemical data, life style data, personal health history data and family health history data of workers; the project file is used for acquiring exposure environment data in project construction, wherein the exposure environment data comprises noise data, air pollution data, construction safety Quannan Fengxian data, equipment safety risk data, water resource pollution data and body temperature state data.
3. The AI-recognition-based worker health assessment method of claim 2, wherein the exposure environment data further comprises, during the acquiring: acquiring exposure environment data in historical engineering construction, and determining environmental ecological factor characteristics; determining generation time points of environmental ecological factors in different environmental ecologies according to the environmental ecological factor characteristics; Building an environmental ecology statistical matrix according to the generation time points of the environmental ecology factors; determining attribute intervals of different environmental ecological factors according to the environmental ecological statistical matrix, wherein the attribute intervals comprise time intervals and intensity intervals; and determining distribution rules of different environmental ecological factors according to the attribute interval, and setting distribution identifiers of the different environmental ecological factors, wherein the distribution rules comprise area distribution, intensity distribution and time distribution.
4. The AI-identification-based assessment method of claim 2, wherein the health profile is further configured with a class fill mechanism, wherein: a data attribute similar configuration mode for configuring a plurality of health data, wherein: In a similar configuration mode, the health file is provided with a plurality of groups of filling positions, and the data attribute of each group of filling positions in the plurality of groups of filling positions is consistent with the attribute of the target phrase; when the worker health data responds to any target phrase attribute, the worker health data and the filling position of the corresponding phrase are subjected to classified traversal, wherein, And executing the health data filling when the filling rule of any filling position is met after the classification traversal.
5. The AI-identification-based assessment method of health status of a worker of claim 1, wherein importing the health profile into a pre-configured exposure identification network identifies exposure in the engineering construction, comprising: Identifying the environmental ecology data in the health file one by one according to the exposure factor identification network, and determining environmental ecology characteristics; determining strong association relations between ecological features of different environments and health conditions of workers, and constructing a strong association network, wherein: The ecological characteristics of different environments and the health conditions of workers in the strongly-correlated network are provided with unique connected components, Determining the positive influence and the negative influence of different environmental ecological characteristics relative to the health state of workers according to the strong correlation network; and determining the corresponding ecological characteristics of the target environment according to the negative influence, and marking the ecological characteristics as exposure factors.
6. The AI-recognition-based assessment method of health of a worker of claim 1, wherein determining the attribution risk value for different exposure factors relative to the worker based on an attribution risk function comprises: Sequentially carrying out first simulation evolution on different exposure factors according to the attribution risk function, wherein the simulation calculation comprises initial exposure time and exposure ending time of the different exposure factors; according to simulation calculation, original calculation time of a plurality of different exposure factors is sequentially calculated, and adjustment time of the original calculation time is determined, wherein the adjacent adjustment time has a proportionality coefficient; Performing second simulation evolution on different exposure factors according to the adjustment time, and comparing the first simulation evolution result with the second simulation evolution result; When the second simulation evolution result is larger than the first simulation evolution result, the target of the second simulation evolution result is attributed to the risk value; when the second simulation evolution result is larger than the first simulation evolution result, taking the first simulation evolution result or the first simulation evolution result as a target attribution risk value; And when the second simulation evolution result is smaller than the first simulation evolution result, taking the first simulation evolution result as a target attribution risk value.
7. The AI-recognition-based worker health assessment method of claim 1, wherein generating a worker health assessment report includes: Configuring an evaluation map of different exposure factors; Dividing the health condition of a worker into a plurality of evaluation display pages through different exposure factors based on the evaluation map, wherein each evaluation display page corresponds to the display page of the different exposure factors; based on the display page, determining at least one health risk type corresponding to different exposure factors, and docking different evaluation contents of the display page by utilizing the health risk type to generate a visual evaluation report.
8. The AI-identification-based worker health assessment method of claim 1, wherein generating a worker health assessment report further comprises: displaying at least one visual area in the visual evaluation report, and setting a trigger instruction, wherein the trigger instruction is used for triggering a corresponding evaluation flow; when the evaluation flow is triggered, configuring a corresponding attribution mapping map according to physical risk data of workers; And determining the exposure distribution state of the user in the engineering construction according to the attribution mapping map, and generating an exposure factor arrangement table of the engineering construction according to the exposure distribution state.

Description

AI (advanced technology attachment) -recognition-based worker health condition assessment method Technical Field The invention relates to the field of risk calculation control of engineering construction, in particular to a worker health condition assessment method based on AI identification. Background Along with the development of science and technology and the increase of health care of people, in the engineering construction process, the health problem of constructors is focused on by the country, and the attribution risk calculation method is taken as the core field, so that great change and convenience are brought to the life of people. Since the constructors work on the construction site in engineering construction, they often get away from home and live in a crude accommodation environment. The working time is long, usually more than 12 hours, from 6 a.m. to 6 a.m. or 8 a.m. and the illegally long working time is normal because of the long working time, their body is easily affected. Because of the long time high load work, many workers suffer from various occupational diseases. Such as asthma, bronchitis, etc. Even some workers are weakened due to long-term work, and even life-threatening situations occur, only by some conventional means: such as 1. Physical examination, 2. Occupational disease detection, 3. Screening for common diseases, 4. Mental health assessment, 5. Nutritional examination, 6. Individual protective supplies examination, etc. However, the conventional means can lead to missing the optimal treatment time due to improper examination and improper detection, and the sudden death event caused by the basic disease of the constructors. Therefore, the existing detection means and early warning measures cannot comprehensively analyze the health condition of the participants. Disclosure of Invention The invention provides a worker health condition assessment method based on AI identification, which is used for solving the problem that the existing detection means and early warning measures cannot comprehensively analyze the health condition of the participants. The invention provides a worker health condition assessment method based on AI identification, which comprises the following steps: Acquiring worker health data in engineering construction, and building a health file; Importing the health file into a preconfigured exposure factor identification network, identifying exposure factors in engineering construction, and building attribution risk functions of different exposure factors; And determining the attributive risk values of different exposure factors relative to the workers according to the attributive risk functions, and generating a worker health condition assessment report. Further, the health profile includes a personal profile and an engineering profile, wherein: the personal archive is used for acquiring physiological and biochemical data, life style data, personal health history data and family health history data of workers; the project file is used for acquiring exposure environment data in project construction, wherein the exposure environment data comprises noise data, air pollution data, construction safety Quannan Fengxian data, equipment safety risk data, water resource pollution data and body temperature state data. Further, the exposing environmental data further includes, in the acquiring process: acquiring exposure environment data in historical engineering construction, and determining environmental ecological factor characteristics; determining generation time points of environmental ecological factors in different environmental ecologies according to the environmental ecological factor characteristics; Building an environmental ecology statistical matrix according to the generation time points of the environmental ecology factors; determining attribute intervals of different environmental ecological factors according to the environmental ecological statistical matrix, wherein the attribute intervals comprise time intervals and intensity intervals; and determining distribution rules of different environmental ecological factors according to the attribute interval, and setting distribution identifiers of the different environmental ecological factors, wherein the distribution rules comprise area distribution, intensity distribution and time distribution. Further, the health profile is further configured with a class fill mechanism, wherein: a data attribute similar configuration mode for configuring a plurality of health data, wherein: In a similar configuration mode, the health file is provided with a plurality of groups of filling positions, and the data attribute of each group of filling positions in the plurality of groups of filling positions is consistent with the attribute of the target phrase; when the worker health data responds to any target phrase attribute, the worker health data and the filling position of the corresponding phrase are subjected to