Intelligent factory personnel positioning and safety early warning method

Abstract

The invention discloses an intelligent factory personnel positioning and safety early warning method, which comprises the following steps of: accurately identifying a risk source, dividing direct and indirect risk induction areas, and marking a position and an influence range by a three-dimensional map to plan a personnel route, so that a high-risk area is effectively avoided, path planning is optimized to enable personnel to be far away from potential danger, and an exposure risk is reduced; a risk chain model is constructed to quantify a risk transmission path and an influence weight, a path deviation value is calculated in combination with real-time positioning, and a risk value can be dynamically and accurately calculated based on the constructed risk assessment model, so that the risk assessment accuracy is improved; when the risk value exceeds the threshold value, early warning is performed in time, the accident response time is shortened, the accident occurrence probability is greatly reduced, personnel and equipment safety is comprehensively guaranteed, and a firm defense line is built for factory safety production.

Description

Technical Field

[0001] The present invention relates to the field of positioning and early warning technology, and in particular to a method for positioning personnel and providing safety early warning for smart factories. Background Art

[0002] In traditional factory production environments, personnel safety management faces numerous challenges. First, the factory environment is complex, with numerous hazardous areas and potential risk sources. Manual inspections struggle to accurately and accurately determine personnel locations in real time, preventing them from entering high-risk areas. Second, risk warning mechanisms often lag behind, relying heavily on manual judgment and experience, lacking scientific evidence and dynamic monitoring. With the advancement of factory automation and intelligence, interactions between personnel and equipment are increasing, and the complexity of risks is increasing. Therefore, a method for personnel positioning and safety warning in smart factories is urgently needed to accurately locate personnel and promptly warn of risks, ensuring personnel safety and stable production.

[0003] The invention patent with patent application number 202010641443.4 discloses a factory personnel positioning and safety warning system, including a positioning information acquisition module, a three-dimensional management module, an information calculation module, a display module, an alarm management module and a history tracking module. A positioning information acquisition module is provided, and the location information of factory personnel can be collected in real time. Through the three-dimensional management module and the information calculation module, the update and action path are calculated to map the position to the coordinate system, improve the accuracy of positioning, and promptly feed back the calculation results of the factory personnel safety status to the display module, so that the system can grasp the safety status of factory personnel at the first time, make safety reminders, and effectively improve the effectiveness of safety management. When a safety problem occurs, an alarm can be issued to factory personnel in time to minimize manpower losses. At the same time, it can respond to sudden safety problems in a timely manner and improve the utilization rate of various resources in the factory.

[0004] However, the actual internal environment of a factory is complex, with various risk sources widely distributed and interconnected. However, the regional demarcation of risk impacts is unclear, making it difficult to intuitively present the scope of direct and indirect effects of risk sources. This can lead to personnel unknowingly entering high-risk areas during work. Furthermore, the lack of systematic research on the laws governing risk transmission makes it impossible to accurately grasp the transmission paths and impact levels of risks between different areas and links, making it difficult to construct an effective risk assessment system. Regarding personnel path management, simple planning is the only option, making it impossible to grasp personnel's actual movement trajectories in real time and unable to promptly detect personnel deviations from safe routes. Once a risk occurs, the lack of a scientific early warning mechanism makes it difficult to issue an alert in a timely manner, which prolongs the exposure of personnel and equipment to danger, increasing the probability of accidents and the extent of losses. Summary of the Invention

[0005] This application provides a method for positioning personnel and providing safety warnings in smart factories, accurately identifying risk areas, quantifying risk transmission, and providing real-time warnings.

[0006] This application provides a method for personnel positioning and safety early warning in a smart factory, including:

[0007] S1, identify the direct risk sensing areas and indirect risk sensing areas of risk sources within the factory, mark their locations and impact ranges on a 3D map, and plan personnel work routes based on the locations of each risk sensing area;

[0008] S2, define the risk transmission path and impact weights between risk sources to build a risk chain model;

[0009] S3, real-time positioning and recording of the personnel’s actual path to generate key points of the path, and calculating the distance deviation between the actual path and the work route;

[0010] S4: Based on the relationship between the deviation value and the risk chain, a risk assessment model is constructed to calculate the risk value of the current risk chain; based on the risk chain model, the risk transmission path between the risk sources is identified and the risk value of each path point on the transmission path is calculated; the transmission attenuation coefficient and the work fatigue coefficient on the transmission path are calculated to modify the risk value of the path point; the transmission attenuation coefficient and the work fatigue coefficient are used to calculate the risk delay coefficient of the modified path point risk value;

[0011] S5: When the risk value is greater than the safety threshold, a risk warning is issued.

[0012] Preferably, in said S4, the path point risk value is modified and the calculation includes: R corrected =R current ·λ(p)·(1+γ(t)), where λ(p) is the conduction attenuation coefficient, λ(p)=e -β·d(p，s) μ env , (p, s) is the Euclidean distance between the path point p and the risk source s, β is the medium attenuation coefficient; γ(t) is the working fatigue coefficient, γ(t) = 1 + δ ln(1 + t), t is the working time (hours), δ is the fatigue growth coefficient; R current is the original risk value.

[0013] Preferably, the original risk value includes: Among them, w k is the impact weight of the k-th risk chain path, P transfer，k is the risk transfer probability of the kth path, and m is the number of associated paths.

[0014] Preferably, the direct risk sensing area and the indirect risk sensing area specifically include: the direct risk sensing area is a spherical or cylindrical area with a radius r1 centered on the risk source, and people entering the area will directly face high risks; The indirect risk sensing area is a spherical or cylindrical area with a radius of r2, centered on the risk source, and r2>r1. (x0, y0, z0) are the coordinates of the risk source.

[0015] Preferably, the method further includes: using the conduction attenuation coefficient and the work fatigue coefficient to calculate the risk delay coefficient of the modified path point risk value to adjust the timeliness of the risk assessment, and the calculation formula is: ε=τ·λ(p)·γ(t), where τ is the time delay factor.

[0016] Preferably, the path points include: the path points are the risk values ​​of the starting point, intermediate nodes and end point in the risk transmission path; the risk transmission path is a risk chain relationship composed of each path from the starting point of the source risk source to the end point of the risk source propagation range.

[0017] Preferably, in S41, identifying the risk transmission path between risk sources further includes:

[0018] 1A, collect historical risk transmission paths and record the risk type of each path point, group them according to risk type, and establish a risk type database;

[0019] 1B, set a risk value threshold for each path point, and dynamically calculate the risk value of each path point in the group in real time;

[0020] 1C, when the risk value of a path point within a group exceeds the threshold, the risk value change trend of the path point is analyzed and the group adjustment mechanism is triggered;

[0021] 1D, regroup the waypoints whose risk values ​​exceed the threshold with the waypoints with the same risk value change trend.

[0022] Preferably, the establishment of a risk type database is specifically as follows: collecting risk transmission path data from historical records, grouping path points with the same risk type into a group, and assigning a unique identifier to each group; creating a database to store information on risk type groups, and the database should contain a group identifier, a risk type description, and a list of path points within the group.

[0023] Preferably, the triggering mechanism for formation adjustment includes: real-time monitoring of the risk value of each path point in the formation, comparing the real-time risk value of each path point with a pre-set risk value threshold, and when the real-time risk value of a path point is greater than the set threshold, marking the path point as an over-threshold path point; collecting risk value data of the over-threshold path point in the past window period to form a time series data set; establishing a linear regression model y=ax+b with time as the independent variable (x) and risk value as the dependent variable (y); calculating the regression coefficients a and b by the least squares method and other methods, setting the conditions for triggering formation adjustment according to the results of the risk value change trend analysis, and immediately starting the formation adjustment process when the triggering conditions are met.

[0024] Preferably, the regrouping specifically includes: screening out path points with the same risk value change trend among all path points whose risk values ​​exceed a threshold; removing the screened path points from the original grouping and regrouping into a new grouping; and updating relevant information in the risk type database.

[0025] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0026] By accurately identifying risk sources within the factory and dividing direct and indirect risk-sensing zones, visual annotation and path planning are implemented on a three-dimensional map, effectively avoiding high-risk areas and optimizing personnel work routes. A risk chain model is also constructed to quantify risk transmission paths and impact weights. Real-time positioning technology is used to calculate personnel path deviations, which in turn creates a risk assessment model to dynamically calculate current risk values ​​and provide timely warnings when risk thresholds are exceeded. This solution comprehensively enhances the intelligence and refinement of factory safety management, effectively ensuring the safety of personnel and equipment, and reducing accident risks.

[0027] The risk chain model identifies risk transmission paths and quantifies the risk values ​​at each point along the path. The transmission attenuation coefficient is calculated by combining the medium attenuation coefficient and the environmental correction factor, reflecting the natural attenuation of risk along the transmission path. The fatigue growth coefficient and working hours are also incorporated to calculate the work fatigue coefficient, quantifying the impact of human fatigue on risk perception. Finally, the risk delay coefficient is calculated using the time delay factor, path length, and movement speed to dynamically adjust the timeliness of risk assessment. This solution integrates physical attenuation, human factors, and dynamic time delay to achieve accurate quantification of risk transmission and real-time early warning, enhancing the scientific nature and effectiveness of safety management.

[0028] By grouping path points with similar risk characteristics or correlations, centralized risk management and precise monitoring are achieved, helping to optimize resource allocation and improve resource utilization efficiency. Furthermore, risk trend forecasting and early warning based on groupings enable the development of targeted strategies and coordinated prevention and control measures, effectively improving risk management accuracy, enhancing risk response capabilities, and reducing risk-related losses. This systematized management of risk information enables accurate identification and dynamic risk monitoring, allowing for timely adjustments to groupings to adapt to changing risks. This effectively improves the efficiency and accuracy of risk response, reduces risk losses, and ensures stable plant operations.

[0029] By deeply mining historical risk data and accurately simulating risk transmission, we can predict the direction of risk at specific points along the path, providing early warning of risks. Compared to traditional methods, this approach moves away from passive risk response and enables the preemptive deployment of pre-emptive adjustments, significantly improving the timeliness and proactivity of risk response. It also precisely locates risk transmission paths and key nodes, enabling the rational allocation of resources, optimizing risk management strategies, and effectively mitigating the damage caused by risks. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 The figure is a flow chart of a method for personnel positioning and safety warning in a smart factory according to an embodiment of the present invention. DETAILED DESCRIPTION

[0031] To facilitate understanding of the present invention, the present application will be described more comprehensively below with reference to the relevant drawings; the drawings show preferred embodiments of the present invention, but the present invention can be implemented in many different forms and is not limited to the embodiments described herein; on the contrary, the purpose of providing these embodiments is to enable a more thorough and comprehensive understanding of the disclosed content of the present invention.

[0032] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this invention pertains; the terms used herein in the specification of the present invention are for the purpose of describing specific embodiments only and are not intended to limit the present invention; the term "and / or" used herein includes any and all combinations of one or more of the associated listed items.

[0033] Example 1: Figure 1 This is a flow chart of a method for personnel positioning and safety warning in a smart factory according to an embodiment of the present invention.

[0034] like Figure 1 As shown, a method for personnel positioning and safety warning in a smart factory includes the following steps:

[0035] S1, identify the direct risk sensing areas and indirect risk sensing areas of risk sources within the factory, mark their locations and impact ranges on a three-dimensional map, and plan personnel work routes based on the locations of each risk sensing area.

[0036] The direct risk sensing area is a spherical or cylindrical area with a radius of r1 (e.g., 1-3 meters) centered on the risk source. People entering this area will directly face high risks. The indirect risk sensing area is a spherical or cylindrical area with a radius of r2 (e.g. 5-10 meters) centered on the risk source, and r2>r1. Although people do not directly enter, they may face potential risks due to proximity. (x0, y0, z0) are the coordinates of the risk source. The specific area range is further limited according to the actual application scenario.

[0037] Specifically, the team traverses the factory area to identify all potential risk sources, such as high-temperature equipment, high-voltage areas, and toxic chemical storage areas. Each risk source's type, location, and risk level are recorded, and criteria for direct / indirect risk sensing zones are determined. Using 3D modeling software or GIS tools, the risk sources and their sensing zone boundaries are marked on a 3D map to create a visual risk map.

[0038] Based on a three-dimensional risk map, a path planning algorithm is used to plan personnel work routes, ensuring that they avoid direct risk-sensing areas and minimize time spent in indirect risk-sensing areas. Taking personnel task requirements (such as material handling and equipment maintenance) into consideration, routes are optimized to balance efficiency and safety, generating the optimal work route. Using the A* algorithm as an example, the shortest path from starting point S to end point G is found, avoiding direct risk-sensing areas. The heuristic function is defined as f(n) = g(n) + h(n), where g(n) is the actual cost from the starting point to node n, and h(n) is the estimated cost (e.g., Euclidean distance) from node n to the end point. The constraint is that no node on the path must be located in a direct risk-sensing area. The path length L and the exposure time T to the indirect risk-sensing area are minimized. The formula is min(w1·L+w2·T), where w1 and w2 are weight coefficients, satisfying w1+w2=1.

[0039] For example, consider identifying high-temperature equipment (such as a furnace) at coordinates (10,20,5), where the risk level is high. The radius of the direct risk sensing zone, r1, is 3 meters, and the radius of the indirect risk sensing zone, r2, is 8 meters. Starting from S (0,0,0) and ending at G (20,30,0), the A* algorithm is used to plan a path that avoids the direct risk sensing zone. This generates the following path node sequence: (0,0,0)→(5,10,0)→(15,25,0)→(20,30,0). Verify that the path node is within the direct risk sensing zone: (Safety). Calculate the path length L = 15 + 10 + 5 = 30 meters and the exposure time T of the indirect risk sensing zone (assuming a speed of 1 meter per second): The closest distance between the path from node (5, 10, 0) to (15, 25, 0) and the boundary of the indirect risk sensing zone is 8 - 7.07 ≈ 0.93 meters (exposure time T = 10 seconds). Optimize the objective function min(0.6 30 + 0.4 10) = 22.

[0040] S2, define the risk transmission path and impact weights between risk sources to build a risk chain model.

[0041] Specifically, based on plant layout, process flow, and historical accident data, we analyze the potential correlations between risk sources (such as physical proximity, upstream and downstream processes, energy / material transfer, etc.) and identify triggering conditions that may cause risk transfer (such as high-temperature equipment failure causing a fire in an adjacent area, or a chemical leak spreading to downstream areas). Based on the correlation between risk sources, we define risk transfer paths, that is, the possible paths by which risks are transferred from one risk source to another. The risk transfer paths are marked on the three-dimensional risk map, including the path's starting point, end point, intermediate nodes (such as equipment, pipelines, vents, etc.), and transfer direction.

[0042] Based on factors such as the intensity, frequency, and duration of the risk transmission path, impact weights are assigned to each path. The impact weights are quantified using the Analytic Hierarchy Process (AHP) or historical data analysis. Weight values ​​typically range from 0 to 1, with larger values ​​indicating a greater likelihood of risk transmission or a greater degree of impact. Risk sources, their risk transmission paths, and impact weights are integrated into a risk chain model. The model is represented as a directed graph, with nodes representing risk sources, edges representing risk transmission paths, and edge weights representing impact weights. The integrity and rationality of the risk chain model are verified to ensure that all key risk transmission paths are covered.

[0043] S3, real-time positioning and recording of the actual path of the personnel to generate key points of the path, and calculate the distance deviation value between the actual path and the work route.

[0044] Specifically, high-precision positioning equipment is installed in the factory to ensure coverage of all work areas, and positioning tags are equipped for personnel to upload location data to the positioning server in real time. The positioning server records the personnel location coordinates (x t ,y t , z t ), forming time series data, filtering the location data to eliminate noise and outliers and improve path accuracy.

[0045] Based on the path change characteristics (direction mutation, speed change, stop point), the path key points are extracted and the key point sequence P is generated. acti ={p1, p2, ..., p n}, where p i =(x i ,y i , z i ). Load the pre-planned work route from the 3D risk map, represented as a line segment sequence L work ={l1, l2, ..., l m}, where l j A line segment is defined by a start point and an end point.

[0046] For each actual path key point p i , calculate its route to the work route L work The shortest distance d(p i , L work The distance deviation value can be the average distance deviation (the average value of the shortest distances from all key points to the work route) and the maximum distance deviation (the maximum value of the shortest distances from all key points to the work route). According to the safety management requirements, the appropriate distance deviation value is selected as the evaluation indicator.

[0047] S4, based on the relationship between the deviation value and the risk chain, build a risk assessment model to calculate the risk value of the current risk chain.

[0048] Specifically, based on historical data, the relationship between the path deviation value and the risk transfer probability in the risk chain is established, the risk level (low risk, medium risk, high risk) corresponding to the deviation value range is determined, and mapped to the risk transfer path in the risk chain. The path deviation value is used as input, combined with the impact weight in the risk chain model, to construct a risk assessment model. The model can use a weighted method to calculate the current risk value R current , considering the impact weights and risk transmission probabilities of all associated risk chain paths. The calculation formula is: Among them, w k is the impact weight of the k-th risk chain path, P transfer，k is the risk transfer probability of the kth path (with the deviation value D k related), m is the number of related paths.

[0049] According to the deviation value of the actual path, query the associated risk chain path and its impact weight, calculate the current risk value, and consider the amplifying effect of the deviation value on the risk transmission probability.

[0050] Among them, the path deviation value D is related to the risk transfer probability P in the risk chain. transfer The association relationship is:

[0051]

[0052] It should be noted that the threshold and probability value can be adjusted according to actual needs.

[0053] For example, assuming that the deviation value of the actual path is D = 1.2 meters, according to the mapping relationship, the risk transfer probability P is obtained. transfer =0.3, assuming that two risk chain paths are associated, the impact weights are w1 = 0.6 and w2 = 0.4, and the risk transmission probability is P transfer，1 =P transfer，2 =0.3, calculate the current risk value R current =0.6×0.3+0.4×0.3=0.18+0.12=0.3

[0054] S5: When the risk value is greater than the safety threshold, a risk warning is issued.

[0055] The technical solutions in the above embodiments of the present application have at least the following technical effects or advantages:

[0056] By accurately identifying risk sources within the factory and dividing direct and indirect risk-sensing zones, visual annotation and path planning are implemented on a three-dimensional map, effectively avoiding high-risk areas and optimizing personnel work routes. A risk chain model is also constructed to quantify risk transmission paths and impact weights. Real-time positioning technology is used to calculate personnel path deviations, which in turn creates a risk assessment model to dynamically calculate current risk values ​​and provide timely warnings when risk thresholds are exceeded. This solution comprehensively enhances the intelligence and refinement of factory safety management, effectively ensuring the safety of personnel and equipment, and reducing accident risks.

[0057] Example 2: In Example 1, the risk value calculation of the risk transmission path relies only on the initial risk superposition of a single path point, and does not fully consider the influence of medium attenuation, dynamic changes in the environment, and personnel operating status on risk perception during the transmission process. However, the attenuation characteristics of different transmission media (such as air and liquid) are significantly different, and real-time fluctuations in environmental conditions (such as ventilation efficiency and temperature) will cause changes in the intensity of risk transmission; at the same time, the fatigue state of personnel after long-term work will reduce their ability to respond to risks, resulting in a lag in risk perception. If a unified initial risk value calculation model is still used, it will inevitably fail to accurately reflect the dynamic characteristics of risk transmission, resulting in deviations in risk assessment and early warning delays. In order to more accurately quantify the attenuation effect and the influence of human factors in the risk transmission process, it is necessary to simultaneously introduce the medium attenuation coefficient, environmental correction factor, and work fatigue coefficient to dynamically correct the path point risk value, thereby improving the real-time and accuracy of risk assessment.

[0058] In some embodiments, the current risk value is calculated, and step S4 further includes:

[0059] S41, based on the risk chain model, identifies the risk transmission path between risk sources and calculates the risk value of each path point on the transmission path.

[0060] Among them, the risk transmission path also includes each path consisting of a starting point (source risk source), an end point (end point of the risk source transmission range) and an intermediate node (equipment, pipelines, vents, etc. involved in the middle); the path point is the risk value of the starting point position, intermediate node and end point position in the risk transmission path.

[0061] S42, calculating the conduction attenuation coefficient and the working fatigue coefficient on the conduction path, and correcting the risk value of the path point.

[0062] The conduction attenuation coefficient represents the degree of risk attenuation along the conduction path, and the calculation formula is: λ(p) = e -β·d(p，s) .μ env , (p, s) is the Euclidean distance between the path point p and the risk source s, β is the medium attenuation coefficient (preset according to the type of conductive medium, such as air = 0.1, liquid = 0.05), μ env is the environmental correction factor (dynamically adjusted according to ventilation, temperature, etc., ranging from 0.8-1.2).

[0063] The work fatigue coefficient indicates the degree of decline in a person's risk perception and response ability after working for a long time. The calculation formula is: γ(t) = 1 + δ·ln(1+t), where t is the working time (hours) and δ is the fatigue growth coefficient (preset according to work intensity, high intensity = 0.3, medium intensity = 0.2, and low intensity = 0.1).

[0064] The formula for modifying the path point risk value is: R corrected =R current ·λ(p)·(1+γ(t)).

[0065] S43, calculating a risk delay coefficient for correcting the risk value of the path point using the conduction attenuation coefficient and the work fatigue coefficient.

[0066] Specifically, the risk delay factor is used to adjust the timeliness of risk assessments. The calculation formula is: ε = τ·λ(p)·γ(t), where τ is the time delay factor (related to path length and occupant speed, measured in seconds / meter). For example, if occupant speed v = 1 meter / second and path length L = 20 meters, then τ = L / v = 20 seconds.

[0067] The technical solutions in the above embodiments of the present application have at least the following technical effects or advantages:

[0068] The risk chain model identifies risk transmission paths and quantifies the risk values ​​at each point along the path. The transmission attenuation coefficient is calculated by combining the medium attenuation coefficient and the environmental correction factor, reflecting the natural attenuation of risk along the transmission path. The fatigue growth coefficient and working hours are also incorporated to calculate the work fatigue coefficient, quantifying the impact of human fatigue on risk perception. Finally, the risk delay coefficient is calculated using the time delay factor, path length, and movement speed to dynamically adjust the timeliness of risk assessment. This solution integrates physical attenuation, human factors, and dynamic time delay to achieve accurate quantification of risk transmission and real-time early warning, enhancing the scientific nature and effectiveness of safety management.

[0069] Example 3: In Example 2, for the grouping and monitoring of path points for risk management, the initial grouping is mainly based on a fixed risk level classification standard, and a unified response strategy is adopted for all path points in the grouping when the risk occurs. However, in actual situations, there are significant differences in the risk transmission mechanism and impact range of different path points, and it is difficult for a fixed grouping method to accurately capture the dynamic changes in the risks of each path point. At the same time, when faced with complex and changing risk scenarios, because the key factor of the risk value change trend is not fully considered, the response mechanism is triggered only based on whether the risk value exceeds the threshold, which is prone to misjudgment or response lag, resulting in poor risk management effect. In order to more effectively respond to complex risk environments, achieve accurate dynamic management of risk path points, and improve the accuracy and timeliness of risk response, this example proposes a path point grouping and dynamic adjustment method based on risk type and change trend, namely Example 3.

[0070] In some embodiments, in step S41, identifying the risk transmission path between risk sources further includes:

[0071] 1A. Collect historical risk transmission paths and record the risk type of each path point, group them according to risk type, and establish a risk type database.

[0072] Specifically, data on risk transmission paths should be collected from historical records. This data should include relevant information for each path point, such as path point identification, occurrence time, and risk description. The collected risk descriptions for each path point should be analyzed, and risks should be classified into different types based on factors such as the nature of the risk, the source of the risk, and the scope of impact.

[0073] Group waypoints with the same risk type and assign a unique identifier to each group. Create a database to store information about risk type groups. The database should contain fields such as the group identifier, risk type description, and a list of waypoints within the group.

[0074] 1B, set a risk value threshold for each path point, and dynamically calculate the risk value of each path point in the formation in real time.

[0075] Among them, the setting can be made by referring to historical risk value data.

[0076] 1C, when the risk value of a path point within a formation exceeds the threshold, the risk value change trend of the path point is analyzed and the formation adjustment mechanism is triggered.

[0077] Specifically, the risk value of each path point in the formation is continuously monitored in real time, and the real-time risk value of each path point is compared with the pre-set risk value threshold. When the real-time risk value of a path point is greater than the set threshold, the path point is marked as an over-threshold path point.

[0078] Collect risk value data for the over-threshold path point over a period of time (e.g., 6 hours, 12 hours; the specific time length can be determined based on needs and risk characteristics) to form a time series dataset. For example, record the risk value of the path point once a day to obtain a data series containing risk values ​​at multiple time points. With time as the independent variable (x) and risk value as the dependent variable (y), establish a linear regression model y = ax + b. Calculate the regression coefficients a and b using methods such as the least squares method. If the regression coefficient a > 0, it indicates that the risk value is increasing over time; if a < 0, it indicates that the risk value is decreasing over time.

[0079] Based on the results of the risk value trend analysis, set the conditions that trigger group adjustments. For example, when the risk value shows a sustained upward trend (the regression coefficient a in the linear regression analysis is greater than 0 and reaches a certain significance level), the group adjustment mechanism is triggered. Once the trigger conditions are met, the group adjustment process is immediately initiated.

[0080] 1D, regroup the waypoints whose risk values ​​exceed the threshold with the waypoints with the same risk value change trend.

[0081] Specifically, among all path points with risk values ​​exceeding a threshold, path points with the same risk value change trend are screened out. These filtered path points are removed from the original grouping and reorganized into a new grouping. Simultaneously, the relevant information in the risk type database is updated.

[0082] The technical solutions in the above embodiments of the present application have at least the following technical effects or advantages:

[0083] By grouping path points with similar risk characteristics or correlations, centralized risk management and precise monitoring are achieved, helping to optimize resource allocation and improve resource utilization efficiency. Furthermore, risk trend forecasting and early warning based on groupings enable the development of targeted strategies and coordinated prevention and control measures, effectively improving risk management accuracy, enhancing risk response capabilities, and reducing risk-related losses. This systematized management of risk information enables accurate identification and dynamic risk monitoring, allowing for timely adjustments to groupings to adapt to changing risks. This effectively improves the efficiency and accuracy of risk response, reduces risk losses, and ensures stable plant operations.

[0084] Example 4: In Example 3, for path point risk monitoring and formation adjustment, the response mechanism is triggered mainly based on the comparison between the current risk value and the fixed threshold, and the formation adjustment is relatively delayed. However, actual risk propagation is a dynamic and complex process, and the speed, scope and impact of risk propagation between different path points vary significantly. It is difficult to accurately capture the early signs of risk propagation based solely on the current risk value and the fixed threshold, and it is impossible to predict the direction of risk in advance, resulting in untimely formation adjustments and passive risk response. At the same time, since the time characteristics of risk propagation are not fully considered, the delay effect of different path points in the risk propagation process is ignored, making the risk warning not accurate enough. In order to more effectively respond to complex and changing risk environments, predict risks in advance and optimize formation adjustment strategies, and improve the timeliness and accuracy of risk response, this embodiment further optimizes the risk value change trend of path points.

[0085] In some embodiments, in step 1C, analyzing the risk value change trend of the path points further includes:

[0086] Collect the risk value change trends of historical path points to build a risk propagation model. According to the risk propagation model, simulate and predict the risk value change trends of different path points to obtain the risk propagation time of each path point. According to the different risk propagation times, inject the corresponding risk delay coefficient for each path point, and predict the formation in advance to make preliminary adjustments to the formation.

[0087] Specifically, the risk value trends of historical path points are classified and cleaned, noise data and outliers are removed, and the risk value data is standardized. Based on the characteristics of risk propagation, a complex network model is used, treating path points as nodes in the network and the connections between path points as edges to construct a risk propagation network. By analyzing historical risk propagation events, the probability of risk propagation from one path point to another under different conditions is determined, as well as the degree of risk attenuation during the propagation process.

[0088] Set different initial risk scenarios, such as setting different initial risk values ​​at certain key path points, to simulate various possible risk outbreaks. Determine the time step and total duration of the simulation. The time step can be set based on the speed and accuracy requirements of risk propagation, such as hourly, daily, etc. Use the constructed risk propagation model to perform simulations, calculating the risk value trend and risk propagation process for each path point within each time step. Record the risk value of each path point at different time steps to obtain time series data for risk propagation. Based on the risk propagation time series data obtained from the simulation, calculate the risk propagation time for each path point under different initial risk scenarios and calculate the corresponding delay coefficient for each path point. Based on the delay coefficient, obtain the path point risk value in advance to provide early risk warnings.

[0089] The technical solutions in the above embodiments of the present application have at least the following technical effects or advantages:

[0090] By deeply mining historical risk data and accurately simulating risk transmission, we can predict the direction of risk at specific points along the path, providing early warning of risks. Compared to traditional methods, this approach moves away from passive risk response and enables the preemptive deployment of pre-emptive adjustments, significantly improving the timeliness and proactivity of risk response. It also precisely locates risk transmission paths and key nodes, enabling the rational allocation of resources, optimizing risk management strategies, and effectively mitigating the damage caused by risks.

[0091] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Various modifications and variations are readily apparent to those skilled in the art. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. A method for personnel positioning and safety early warning in a smart factory, characterized in that: include: S1, identify the direct risk sensing areas and indirect risk sensing areas of risk sources within the factory, mark their locations and impact ranges on a 3D map, and plan personnel work routes based on the locations of each risk sensing area; S2, define the risk transmission path and impact weights between risk sources to build a risk chain model; S3, real-time positioning and recording of the personnel’s actual path to generate key points of the path, and calculating the distance deviation between the actual path and the work route; S4, based on the relationship between the deviation value and the risk chain, build a risk assessment model to calculate the risk value of the current risk chain; Based on the risk chain model, identify the risk transmission path between risk sources and calculate the risk value of each path point on the transmission path; Calculate the conduction attenuation coefficient and working fatigue coefficient on the conduction path and correct the risk value of the path point; The risk delay coefficient of the modified path point risk value is calculated using the conduction attenuation coefficient and the work fatigue coefficient; S5: When the risk value is greater than the safety threshold, a risk warning is issued.

2. The smart factory personnel positioning and safety warning method according to claim 1, characterized in that: In said S4, the path point risk value is modified, and the calculation includes: R corrected =R current ·λ(p)·(1+γ(t)), where λ(p) is the conduction attenuation coefficient, λ(p)=e -β·d(p，s) μ env , (p, s) is the Euclidean distance between the path point p and the risk source s, β is the medium attenuation coefficient; γ(t) is the working fatigue coefficient, γ(t) = 1 + δ ln(1 + t), t is the working time (hours), δ is the fatigue growth coefficient; R current is the original risk value.

3. The smart factory personnel positioning and safety warning method according to claim 2, characterized in that: The original risk value includes: Among them, w k is the impact weight of the k-th risk chain path, P transfer，k is the risk transfer probability of the kth path, and m is the number of associated paths.

4. The smart factory personnel positioning and safety warning method according to claim 1, characterized in that: The direct risk sensing area and the indirect risk sensing area specifically include: the direct risk sensing area is a spherical or cylindrical area with a radius r1 centered on the risk source. People entering this area will directly face high risks. The indirect risk sensing area is a spherical or cylindrical area with a radius of r2, centered on the risk source, and r2>r1. (x0, y0, z0) are the coordinates of the risk source.

5. The smart factory personnel positioning and safety warning method according to claim 2, characterized in that: The method further includes: using the conduction attenuation coefficient and the work fatigue coefficient to calculate the risk delay coefficient of the modified path point risk value to adjust the timeliness of the risk assessment, and the calculation formula is: ε=τ·λ(p)·γ(t), where τ is the time delay factor.

6. The smart factory personnel positioning and safety warning method according to claim 1, characterized in that: The path points include: the path points are the risk values ​​of the starting point, intermediate nodes and end points in the risk transmission path; the risk transmission path is a risk chain relationship composed of each path from the source risk source starting point to the end point of the risk source propagation range.

7. The smart factory personnel positioning and safety warning method according to claim 1, characterized in that: In S41, identifying the risk transmission path between risk sources also includes: 1A, collect historical risk transmission paths and record the risk type of each path point, group them according to risk type, and establish a risk type database; 1B, set a risk value threshold for each path point, and dynamically calculate the risk value of each path point in the group in real time; 1C, when the risk value of a path point within a group exceeds the threshold, the risk value change trend of the path point is analyzed and the group adjustment mechanism is triggered; 1D, regroup the waypoints whose risk values ​​exceed the threshold with the waypoints with the same risk value change trend.

8. The method for personnel positioning and safety early warning in a smart factory according to claim 7, characterized in that: The establishment of the risk type database is specifically as follows: risk transmission path data is collected from historical records, path points with the same risk type are grouped together, and a unique identifier is assigned to each group; a database is created to store information on risk type groups, and the database should contain a group identifier, a risk type description, and a list of path points within the group.

9. The method for personnel positioning and safety early warning in a smart factory according to claim 7, characterized in that: The triggering mechanism for group adjustment includes: real-time monitoring of the risk value of each path point in the group, comparing the real-time risk value of each path point with a pre-set risk value threshold, and when the real-time risk value of a path point is greater than the set threshold, marking the path point as an over-threshold path point; collecting risk value data of the over-threshold path point over a past window period to form a time series data set; establishing a linear regression model y=ax+b with time as the independent variable (x) and risk value as the dependent variable (y); calculating the regression coefficients a and b by the least squares method and other methods, setting the conditions for triggering group adjustment based on the results of the risk value change trend analysis, and immediately starting the group adjustment process when the triggering conditions are met.

10. The smart factory personnel positioning and safety warning method according to claim 7, characterized in that: The regrouping specifically includes: screening out path points with the same risk value change trend among all path points whose risk values ​​exceed the threshold; removing the screened path points from the original grouping and regrouping into a new grouping; updating the relevant information in the risk type database.

Images