Intelligent personnel scheduling management method and system based on Internet of Things

Abstract

The invention relates to the technical field of the Internet of Things, in particular to an intelligent personnel scheduling management method based on the Internet of Things, which comprises the following steps of: acquiring multi-source data such as positions, physiology and behaviors of personnel, and calculating individual risk scores by fusing surrounding personnel and historical relationships; generating an aggregation risk topological graph from the risk scores through a preset kernel function; and analyzing a risk peak value in the graph, matching an optimal executor through a multi-objective optimization algorithm, and generating a scheduling instruction. According to the method, discrete individual risks are converted into a visual global risk view, automation and optimization of scheduling decisions are realized based on dynamic analysis of the risk situation, and the risk early warning accuracy and the emergency response efficiency are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of Internet of Things, in particular to a personnel intelligent scheduling management method and system based on Internet of Things. BACKGROUND

[0002] The operation and management of specific places are extremely complex, and the dynamic scheduling and risk control of human resources are rigorous. Introducing advanced data processing methods to improve management efficiency and decision-making quality is the core demand.

[0003] The existing personnel management system mostly uses scheduling methods based on static rules or fixed cycles, and the resource allocation is rigid, which is difficult to cope with sudden changes in human resource demand and complex task assignment. The decision-making process relies on the personal experience of managers, and lacks data-driven risk assessment and dynamic optimization capabilities, resulting in low operational efficiency and potential management risks. Therefore, the existing management method needs to solve the problem of how to integrate multi-dimensional operation data to realize dynamic, intelligent and risk-aware resource scheduling from static scheduling.

[0004] Therefore, a personnel intelligent scheduling management method and system based on Internet of Things are proposed. SUMMARY

[0005] The purpose of the present application is to provide a personnel intelligent scheduling management method and system based on Internet of Things, which builds a multi-layer intelligent perception and decision-making framework integrating individual physiological / behavior, social context and spatial dimension, aiming to solve the problems of risk identification lag, situation awareness fragmentation and low scheduling response efficiency in the management of specific places.

[0006] To achieve the above purpose, the present application provides the following technical scheme:

[0007] A personnel intelligent scheduling management method based on Internet of Things, comprising:

[0008] Real-time acquisition of multi-source data of each personnel in the monitoring area, the multi-source data including real-time position data, physiological data and behavior data of each personnel; based on a situation-related risk assessment model, dynamically calculating the current individual risk score of each personnel; the calculation of the individual risk score integrates the multi-source data collected by the personnel itself, the real-time multi-source data of other personnel in the surrounding area, and a dynamically updated personnel relationship knowledge graph;

[0009] Mapping the physical space of the monitoring area to a coordinate system, taking the real-time position of each personnel as the coordinate center, and taking the individual risk score as the weight, generating an aggregated risk topology graph that can represent the continuous distribution of risks in the entire monitoring area through a preset kernel function;

[0010] Real-time analysis of the aggregation risk topology, identify risk peak area with risk value higher than the preset threshold, determine the risk cause type and its dynamic evolution trend of the risk peak area; and based on the risk peak area, risk level, risk type and dynamic trend, through multi-objective optimization scheduling algorithm, match the optimal performer and generate scheduling instructions containing recommended response level.

[0011] Preferably, the context-related risk assessment model comprises a data input and preprocessing unit, an individual state assessment unit, a social context analysis unit and a risk fusion decision unit; the data input and preprocessing unit is used to receive the multi-source data, and to clean, normalize and feature extract the multi-source data, to generate a standardized input data stream; the individual state assessment unit calculates the basic risk level of the target person in the isolated state based on the standardized input data stream of the target person itself; the social context analysis unit is used to query the nodes and relationships of the target person and the surrounding personnel in the personnel relationship knowledge graph according to the real-time location of the target person, analyze and quantify the potential risk influence of the current social environment through graph calculation and reasoning, and generate a context risk correction factor; the risk fusion decision unit is used to fuse the basic risk level output by the individual state assessment unit and the context risk correction factor output by the social context analysis unit, and calculate through a weighted algorithm to output the individual risk score of the target person.

[0012] Preferably, the generation step of the aggregation risk topology comprises: discretizing the physical space of the monitoring area into a two-dimensional and / or three-dimensional digital grid composed of multiple grid points, and zeroing the initial risk values of all grid points; for each person in the monitoring area, taking his / her current real-time location as the center and his / her individual risk score as the amplitude, applying a preset kernel function to construct a risk distribution function representing the influence range and intensity of the individual risk of the person in space; traverse each grid point in the digital grid, calculate and accumulate the function values of the risk distribution functions of all persons on the grid point to obtain the accumulated aggregation risk value on the grid point; render all grid points and their corresponding aggregation risk values to generate the aggregation risk topology, and different aggregation risk values are visualized and presented through different color gradients.

[0013] Preferably, the generating step of the scheduling instruction comprises: feature extraction on the identified risk peak area to obtain its position, risk level, risk type and dynamic evolution trend as input parameters of scheduling decision; for each performer, based on the input parameters, comprehensive evaluation is carried out on the performer by the multi-objective optimization scheduling algorithm to calculate sub-item indicators; the sub-item indicators of each performer are weighted to determine the optimal performer under the current situation; and a scheduling instruction is generated for the optimal performer, which contains intervention position, risk type summary and recommended response level.

[0014] Preferably, the sub-item indicators include intervention time indicator, skill matching degree indicator and task interruption cost indicator; the intervention time indicator is determined based on the shortest reachable path predicted travel time calculated by a path planning algorithm based on the current real-time position of the performer and the position of the risk peak area; the skill matching degree indicator is determined by comparing and quantitatively scoring the preset required skill corresponding to the risk type of the risk peak area with the skill items and proficiency levels in the skill profile of the performer; and the task interruption cost indicator is a cost quantification value comprehensively evaluated according to the preset priority of the task currently being performed by the performer and the attribute of whether the task can be safely interrupted.

[0015] Preferably, the personnel relationship knowledge graph is generated by a social network analysis method, specifically comprising: based on the real-time position data, continuously recording the close contact events of n personnel within a preset distance threshold in the monitoring area, and storing the contact duration and identity information; when a risk peak area appears on the aggregated risk topology graph and / or a security event report is received from the outside, the close contact events related to the event before and after the event occurrence time point are associated analyzed; according to the results of the association analysis, the relationship attributes between the personnel involved in the close contact events are quantitatively updated; the updated relationship attributes are stored in a relationship database for calling by the situational correlation risk assessment model when calculating individual risk scores.

[0016] A personnel intelligent scheduling management system based on Internet of Things, comprising:

[0017] The data acquisition module acquires multi-source data from each person within the monitoring area in real time, including their real-time location, physiological data, and behavioral data. The risk assessment module dynamically calculates the current individual risk score for each person based on a context-related risk assessment model. This calculation integrates the individual's own collected multi-source data, real-time multi-source data from other individuals in the surrounding area, and a dynamically updated knowledge graph of interpersonal relationships. The topology generation module maps the physical space of the monitoring area into a coordinate system, using each person's real-time location as the coordinate center and their individual risk score as weights. It then generates an aggregated risk topology map representing the continuous risk distribution across the entire monitoring area using a preset kernel function. The scheduling decision module analyzes the aggregated risk topology map in real time, identifies risk peak areas with risk values ​​exceeding a preset threshold, determines the risk causes and dynamic evolution trends of these peak areas, and, based on the risk peak areas, risk levels, risk types, and dynamic trends, uses a multi-objective optimization scheduling algorithm to match the optimal executor and generate scheduling instructions with suggested response levels.

[0018] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0019] 1. This invention constructs a context-related risk assessment model, which not only analyzes individual states but also innovatively builds and utilizes a knowledge graph of interpersonal relationships to transform discrete, unstructured historical interaction data into a structured relationship network. This allows risk assessment to move beyond simple historical event queries, enabling deep insights into potential conflicts arising from complex interpersonal networks through graph reasoning and analysis. It achieves a shift from superficial monitoring to deep knowledge-driven risk prediction, significantly improving the reliability and foresight of early warnings. This method can identify hidden risks that are difficult to detect using traditional methods earlier and more accurately, significantly enhancing the reliability and foresight of early warnings.

[0020] 2. The aggregated risk topology map proposed in this invention integrates and elevates massive, discrete personnel risk data into a continuous and intuitive global risk distribution map. Managers no longer need to monitor individual alerts one by one; instead, they can observe the risk like weather forecasts. Figure One In this way, the distribution, form, and evolution of the risk "high-pressure zones" and "low-pressure zones" in the entire monitoring area can be clearly understood at a glance. This greatly simplifies the information processing burden and realizes a fundamental transformation from fragmented point-based monitoring to integrated, global situational awareness.

[0021] 3, The application can automatically match the optimal intervention performer and give a recommended response level according to multiple dimensions such as response time, skill matching degree, task interruption cost, etc., based on the comprehensive judgment of the causes, grades and dynamic trends of risk peaks in the risk topology graph, and the multi-objective optimization scheduling algorithm. This completely changes the mode of traditional scheduling relying on artificial experience and response lag, ensures that each intervention is an optimal resource, appropriate strategy and precise response, and realizes the dual improvement of management resource utilization efficiency and emergency response effectiveness. BRIEF DESCRIPTION OF DRAWINGS

[0022] Figure 1 A method flowchart of a personnel intelligent scheduling management method based on the Internet of Things is provided for the embodiments of the application.

[0023] Figure 2 A flowchart of a situational correlation risk assessment model is provided for the embodiments of the application.

[0024] Figure 3 A system structure diagram of a personnel intelligent scheduling management system based on the Internet of Things is provided for the embodiments of the application. DETAILED DESCRIPTION

[0025] The technical solutions in the embodiments of the application will be described clearly and completely below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, not all the embodiments. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the application.

[0026] Embodiment one

[0027] Please refer to Figures 1-2 The application provides a personnel intelligent scheduling management method based on the Internet of Things, and the technical solutions are as follows:

[0028] A personnel intelligent scheduling management method based on the Internet of Things, as shown in Figure 1 , includes:

[0029] Real-time acquisition of multi-source data of each personnel in the monitored area, the multi-source data including real-time position data, physiological data and behavior data of each personnel; based on a situational correlation risk assessment model, dynamically calculating the current individual risk score of each personnel; the calculation of the individual risk score fuses the multi-source data collected by the personnel itself, the real-time multi-source data of other personnel in the surrounding area, and a dynamically updated personnel relationship knowledge graph;

[0030] mapping the physical space of the monitoring area into a coordinate system, taking the real-time position of each person as the coordinate center and taking the individual risk score as the weight, generating an aggregated risk topology graph capable of representing the continuous distribution of risks in the entire monitoring area through a preset kernel function;

[0031] analyzing the aggregated risk topology graph in real time, identifying a risk peak area with a risk value higher than a preset threshold, determining the risk cause type and dynamic evolution trend of the risk peak area, and based on the risk peak area, risk level, risk type and dynamic trend, matching an optimal executor and generating a scheduling instruction containing a recommended response level through a multi-objective optimization scheduling algorithm.

[0032] Further, the physiological data is heart rate data and skin conductance response data collected by a wearable bracelet, and the behavior data is personnel abnormal running, falling and arm attack action data analyzed by a camera at the top of the monitoring area using a gesture recognition algorithm.

[0033] Further, as shown in the figure, Figure 2 The context-related risk assessment model includes a data input and preprocessing unit, an individual state assessment unit, a social context analysis unit and a risk fusion decision unit. The data input and preprocessing unit is used to receive the multi-source data, and clean, normalize and feature extract the multi-source data to generate a standardized input data stream. The individual state assessment unit calculates the basic risk level of the target person in the isolated state based on the standardized input data stream of the target person itself. The social context analysis unit is used to query the nodes and relationships of the target person and surrounding persons in the personnel relationship knowledge graph according to the real-time position of the target person, analyze and quantify the potential risk influence of the current social environment through graph calculation and reasoning, and generate a context risk correction factor. The risk fusion decision unit is used to fuse the basic risk level output by the individual state assessment unit and the context risk correction factor output by the social context analysis unit, and calculate through a weighting algorithm to output the individual risk score of the target person.

[0034] Specifically, the calculation method of the basic risk level, which first identifies the key indicators strongly related to the risk state from the physiological data and behavior data of the personnel. Then, one or more risk thresholds are set for each key indicator. When the real-time value of the indicator exceeds the preset threshold, a predefined risk score is triggered. The basic risk level is finally obtained by weighted summation of the risk scores corresponding to all triggered key indicators, so as to integrate the discrete physiological and behavior signals into a quantitative baseline score representing the danger level of the individual in the isolated state.

[0035] Specifically, when any indicator is triggered, the corresponding risk score is accumulated, and the final basic risk level is obtained. For example, a person's heart rate is 140 bpm (+10 points), and there is a swinging attack action (+40 points), so the basic risk level is 50 points. The calculation example of the basic risk level can refer to Table 1 below.

[0036] Table 1: Reference example of basic risk level

[0037] Key Indicators Risk Threshold Risk Score Static Heart Rate >120 bpm or <50 bpm 10 Behavior Detection of "Swing Attack" Posture 40 Behavior Detection of "Fall Down" for more than 10 seconds 30 Location Entry into an unauthorized area 20

[0038] The quantitative method of the situational risk correction factor is to build and apply a dynamic personnel relationship knowledge graph. The graph defines each person in the monitoring area as an entity node with attributes such as identity, skill, and real-time risk level. The relationship between nodes is used to represent the interaction history, such as the "close contact" relationship based on real-time location data, and the "conflict" relationship determined by risk event analysis.

[0039] The personnel relationship knowledge graph has the ability to dynamically learn and update. When the system identifies new interactions or risk events, it will update the relationship edge weights between the involved personnel nodes in real time according to the nature of the event. When performing risk assessment, the social situation analysis unit queries the adjacent nodes of the target personnel in the graph and their relationship weights, and through graph calculation and reasoning, accurately quantifies the potential risks brought by the surrounding social environment. This method converts the ambiguous "historical relationship" into a structured and quantifiable social situation, ultimately generating a situational risk correction factor that accurately reflects the current situation.

[0040] The weighting algorithm is a mathematical model that combines the basic risk level and the situational risk correction factor. The algorithm is as follows: individual risk score = w_1 x basic risk level + w_2 x situational risk correction factor, where w_1 and w_2 are preset weights representing the importance of individual factors and environmental factors. By using this algorithm, the system can generate a comprehensive and dynamic individual risk assessment result.

[0041] The situational correlation risk assessment model improves the risk assessment from isolated and superficial monitoring to correlated and in-depth insight by integrating individual status, social situation, and historical interaction relationships. This model overcomes the one-sidedness and lag caused by traditional methods that only focus on individuals, enabling earlier and more accurate identification and quantification of potential conflicts caused by interpersonal relationships, significantly improving the reliability and forward-looking nature of risk warning, and providing a solid basis for decision-making for proactive intervention.

[0042] Further, the generating step of the aggregated risk topological map comprises: discretizing the physical space of the monitoring area into a two-dimensional and / or three-dimensional digital grid composed of a plurality of grid points, and setting the initial risk value of all grid points to zero; for each person in the monitoring area, taking the current real-time position as the center and the individual risk score as the amplitude, applying a preset kernel function to construct a risk distribution function representing the influence range and intensity of the individual risk in space; traversing each grid point in the digital grid, calculating and accumulating the function values of the risk distribution functions of all persons on the grid point to obtain the aggregated risk value of the grid point; rendering all grid points and their corresponding aggregated risk values to generate the aggregated risk topological map, and different aggregated risk values are visualized by different color gradients.

[0043] The preset kernel function is a mathematical tool for expanding each individual's discrete risk point into a continuous distribution field with spatial influence. In the present application, a Gaussian kernel function is preferably used. This function can smoothly attenuate the influence of individual risk from near to far, which conforms to the intuitive understanding of risk diffusion in physical space.

[0044] In the kernel function, the bandwidth parameter directly determines the influence range and intensity of individual risk in space. In this embodiment, a fixed bandwidth is preferably used. The risk distribution of all persons adopts a unified fixed value which is preset according to the physical environment of the monitoring area, such as the typical size of the monitoring area, the width of the corridor, etc. Preferably, in a typical 3-meter-wide corridor or 4x5-meter monitoring area environment, the fixed bandwidth value is set to 1.5 meters. This value can ensure that the individual risk can effectively cover the near range, while avoiding unreasonable superposition effect with individuals far away.

[0045] The specific calculation steps of the aggregated risk value are as follows: first, for the first person in the monitoring area, the risk influence of the person on the specific position point is calculated. The calculation of this influence value integrates three factors: first, the individual risk score of the person; second, the straight-line physical distance between the person and the position point; third, the risk influence range set for the person, i.e. the bandwidth. In the calculation, the physical distance is first divided by the bandwidth to obtain a relative distance value, and then a preset kernel function is used to convert the relative distance value into an influence coefficient. Finally, the influence coefficient is multiplied by the individual risk score of the person to obtain the risk value contributed by the person to the specific position point. For the second, third and last person in the monitoring area, the same calculation is repeated to obtain the risk value contributed by each person to the same position point. Adding all the risk values contributed by the persons, the sum is the final aggregated risk value of the specific position point.

[0046] The generation method integrates massive and discrete individual risk data into a continuous and visual global risk topology map through spatial gridding and kernel function application. This overcomes the information fragmentation and cognitive overload problems caused by traditional point monitoring, enabling managers to intuitively understand the overall distribution, intensity, and spatial form of risks, providing intuitive and scientific decision support for forward-looking spatial control and resource allocation.

[0047] Further, the risk cause type determination method is based on a multi-dimensional feature matching engine. The engine analyzes the data feature combinations of all personnel in the risk peak area, compares them with pre-set risk event templates, and identifies the most likely cause type. These templates are based on historical data and expert experience. For example:

[0048] When the system detects multiple individuals in an area, their behavior data shows high-amplitude violent movements, their physiological data shows a sharp and synchronized increase in heart rate, and their historical interaction data shows conflict records between them, the system determines the cause type of the event as "physical conflict".

[0049] If the behavior data of only one individual in the area shows "falling" or "long-term stillness", and its physiological data presents extreme abnormalities, while the data of other surrounding personnel is basically normal, the cause type is determined as "individual health problem".

[0050] If multiple individuals are detected to gather in an unregulated area for a long time, but their physiological and behavior data do not show significant abnormalities at the beginning, the cause type is determined as "unregulated gathering", which is a potential risk that requires early intervention.

[0051] The determination method of the dynamic evolution trend is achieved by analyzing the change characteristics of the risk peak over time. The system continuously tracks key indicators of the risk peak, such as the height of the peak, the area covered, and the number of people involved, and calculates the change rate in unit time.

[0052] If the height or area of the peak shows a sustained and rapid increase within a short period of time (such as the past 30 seconds), and the change rate exceeds the pre-set "rapid growth" threshold, the trend is determined as "escalating trend".

[0053] Conversely, if the height and covered area of the peak show a significant and sustained decline, the trend is determined as "mitigation trend".

[0054] If the peak maintains a high level, but its value fluctuates within a certain range with a change rate below the "significant change" threshold, the trend is determined as "continuous state".

[0055] If the risk value jumps from a safe level to a high risk level in a very short time, showing a steep rising curve, the trend is determined as "sudden formation", which usually requires the highest priority response.

[0056] By the above method, the present application analyzes an abstract, high-dimensional risk data group into "cause types" (such as fighting, illness) with clear business meanings and clear "evolution trends" (such as escalation, mitigation), thereby providing accurate and executable decision inputs for subsequent intelligent scheduling.

[0057] Further, the generation of the scheduling instruction includes: feature extraction on the identified risk peak area to obtain its position, risk level, risk type and dynamic evolution trend as input parameters for scheduling decisions; for each performer, based on the input parameters, the multi-objective optimization scheduling algorithm is used for comprehensive evaluation, and a sub-index is calculated; the multiple sub-indices of each performer are weighted and calculated to determine the optimal performer under the current situation; and a scheduling instruction is generated for the optimal performer, which includes intervention position, risk type summary and recommended response level.

[0058] The sub-indexes include intervention time index, skill matching degree index and task interruption cost index.

[0059] The calculation method of the weighted calculation is: the sub-indexes are normalized to unify their dimensions to the interval of 0 to 1; after normalization, the system calculates a comprehensive score for each candidate performer. The weight is not fixed but is dynamically adjusted according to the cause type and dynamic evolution trend of the current risk event.

[0060] For events determined as physical conflict and with an escalation trend, the system gives the highest weight to the intervention time index to ensure the fastest intervention.

[0061] For individual health problem events, the system significantly increases the weight of the skill matching degree index to preferentially dispatch personnel with first aid skills.

[0062] If the risk level is not high, but all candidate performers are performing high-priority tasks, the system appropriately increases the weight of the task interruption cost index to avoid excessive impact on existing work.

[0063] Finally, the system selects the performer with the highest comprehensive score as the optimal dispatch object under the current situation.

[0064] The scheduling instruction generation step promotes the scheduling decision from simple rule driving to intelligent tactical decision based on multi-objective optimization. Through comprehensive optimization of the dynamic characteristics of risks and multi-dimensional indicators of the performer, the system can go beyond the limitation of "dispatching nearby" and generate an optimal solution that takes into account response speed, skill matching and task cost. This ensures the accuracy of intervention and the rationality of resource allocation, significantly improving the professional level and overall operational efficiency of emergency response.

[0065] Further, the sub-item indicators include an intervention time indicator, a skill matching degree indicator and a task interruption cost indicator; the intervention time indicator is determined based on the current real-time location of the performer and the location of the risk peak area, using the expected travel time of the shortest reachable path calculated by a path planning algorithm; the skill matching degree indicator is determined by comparing and quantitatively scoring the preset required skills corresponding to the risk type of the risk peak area with the skill items and proficiency levels in the skill profile of the performer; and the task interruption cost indicator is a cost quantification value comprehensively evaluated according to the preset priority of the task currently being performed by the performer and the attribute of whether the task can be safely interrupted.

[0066] By accurately quantifying the decision dimensions of time, skill and task cost, the optimal choice is transformed from a vague management concept to a calculable objective technical indicator. This makes the optimization decision no longer an abstract trade-off, but a precise simulation of real-world opportunity cost. Therefore, the system can make highly context-aware optimal decisions, ensuring that the scheduling solution achieves a scientific balance in multiple dimensions, significantly improving the rationality and refinement level of resource allocation.

[0067] Further, the personnel relationship knowledge graph is generated by a social network analysis method, specifically including: based on the real-time location data, continuously recording the close contact events of n personnel within a preset distance threshold in the monitoring area, and storing the contact duration and identity information; when a risk peak area appears on the aggregated risk topology graph and / or a security event report is received from the outside, performing correlation analysis on the close contact events related to the event before and after the event occurrence time point; according to the results of the correlation analysis, quantitatively updating the relationship attributes between the personnel involved in the close contact events; storing the updated relationship attributes in a relationship database for the context-related risk assessment model to call when calculating individual risk scores.

[0068] The social network analysis method endows the system with the ability to dynamically learn and remember interpersonal relationships. It associates objective location data with risk events, making relationship assessment no longer dependent on subjective judgment or static profiles, but based on continuously evolving data evidence. This enables the system to automatically discover potential "conflict pairs" and endows the risk model with "social memory", significantly improving the depth and prediction accuracy of situational risk assessment, making decisions more insightful.

[0069] The present application provides an intelligent method for comprehensively improving the management of monitoring areas from post-treatment and passive response to pre-warning and active intervention. By constructing a risk assessment model that integrates individual status, surrounding social context, and historical relationships, it achieves deep insight and accurate prediction of potential conflict risks, overcoming the one-sidedness and lag of traditional methods that rely solely on individual appearance. The unique aggregated risk topology graph elevates discrete and independent risk information to a global and visual overall situation, enabling managers to intuitively grasp the distribution, intensity, and evolution of risks, and achieving a leap from point monitoring to surface perception. Multi-objective optimization scheduling based on risk dynamic characteristics ensures that each intervention decision can achieve an optimal balance between response time, personnel skills, and task cost, realizing precise allocation of scheduling resources and maximization of intervention effect. The present application systematically solves the core pain points of risk identification lag, fragmented situation awareness, and low scheduling efficiency in traditional management modes.

[0070] Embodiment Two

[0071] This embodiment is based on a specific application scenario and provides a detailed description of a personnel intelligent scheduling management system based on the Internet of Things. As shown in Figure 3 the system includes a data acquisition module, a risk assessment module, a topology graph generation module, and a scheduling decision module.

[0072] A personnel intelligent scheduling management system based on the Internet of Things, comprising:

[0073] The data acquisition module: real-time acquisition of multi-source data of each personnel in the monitoring area, the multi-source data including real-time position data, physiological data and behavior data of each personnel; the risk assessment module: based on the context-related risk assessment model, dynamically calculating the current individual risk score of each personnel; the calculation of the individual risk score integrates the multi-source data collected by the personnel itself, the real-time multi-source data of other personnel in the surrounding area and a dynamically updated personnel relationship knowledge graph; the topological graph generation module: mapping the physical space of the monitoring area into a coordinate system, taking the real-time position of each personnel as the coordinate center and taking the individual risk score as the weight, generating an aggregated risk topological graph that can represent the continuous distribution of the risk of the entire monitoring area through a preset kernel function; the scheduling decision module: real-time analysis of the aggregated risk topological graph, identifying the risk peak area with a risk value higher than a preset threshold, determining the risk cause type and dynamic evolution trend of the risk peak area; and based on the risk peak area, risk level, risk type and dynamic trend, matching the optimal performer and generating a scheduling instruction containing a recommended response level through a multi-objective optimization scheduling algorithm.

[0074] The scenario is set as follows: time: Wednesday 15:30, public activity time in the monitoring area. Place: multi-functional activity room and adjacent corridor in monitoring area A. Personnel: monitored personnel A: has a history of violent conflict, recent emotional instability. Monitored personnel B: has a history of conflict with A, marked as "high conflict risk relationship pair" by the system. Other monitored personnel: 15, normally active in the activity room. Security personnel A: patrolling the corridor in A, 20 meters away from the activity room, with basic first aid skills. Security personnel B: on duty in the office in B, about 100 meters away from the activity room, with senior conflict mediator qualification. System initial state: the aggregated risk topological graph in monitoring area A as a whole presents "green" representing low risk.

[0075] System operation process:

[0076] First step: data acquisition and risk assessment

[0077] The monitored person A and the monitored person B enter the multi-functional activity room. The data acquisition module obtains the position data of the two persons in real time through the positioning bracelet they wear. The risk assessment module starts the situational correlation risk assessment model. The social situation analysis unit identifies that A and B have entered the preset 3-meter social distance threshold. It immediately calls the personnel relationship knowledge graph and confirms that the two have a "high conflict risk" record. Then, the system generates a higher "situational risk correction factor" for the two. At this time, the physiological and behavioral data of the two persons are still within the normal range, and the "basic risk level" calculated by the individual state assessment unit is still low. The risk fusion decision unit calculates the individual risk score of A and B through a weighting algorithm (for example: individual risk score = 0.6 x basic risk level + 0.4 x situational risk correction factor), and the individual risk score of A and B slightly increases.

[0078] Second step: generation and evolution of aggregated risk topology graph

[0079] The topology graph generation module re-renders the aggregated risk topology graph using the Gaussian kernel function according to the increased individual risk scores of A and B. In the northeast corner of the activity room, a small range of "yellow" area representing "low risk" appears on the topology graph.

[0080] The data acquisition module detects that the heart rate of A increases rapidly from 85bpm to 130bpm, exceeding the static heart rate risk threshold of 120bpm. At the same time, the top camera captures the dangerous action of "arm attack" of A through the posture recognition algorithm.

[0081] The individual state assessment unit in the risk assessment module increases the "basic risk level" of A to 50 points according to the preset rules (heart rate exceeding 10 points, arm attack 40 points).

[0082] The topology graph generation module responds in real time, and the risk of A's location rapidly accumulates. The previously "yellow" risk area rapidly expands and changes to a "red" peak area representing "high risk" on the topology graph.

[0083] Third step: dispatching decision and instruction generation

[0084] The dispatching decision module automatically identifies this newly formed "risk peak area" with a risk value higher than the preset threshold. The module begins to analyze the risk cause type and dynamic evolution trend:

[0085] Cause determination: combined with multi-dimensional data such as "close contact between two people", "historical conflict relationship", "heart rate surge" and "aggressive action", the system determines the risk type as "physical conflict".

[0086] Trend determination: Since the risk peak rapidly grows from "yellow" to "red" within 30 seconds, the system determines its dynamic evolution trend as "escalating trend".

[0087] The system starts the multi-objective optimization scheduling algorithm to evaluate the optimal performer:

[0088] Security Officer A: Intervention time indicator: Excellent. Path planning shows it can be reached within 20 seconds. Skill matching degree indicator: General. The event is "physical conflict", and the "basic first aid" skill matching degree is not high. Task interruption cost indicator: Low. Its current task is "routine patrol", which can be interrupted at any time.

[0089] Security Officer B: Intervention time indicator: Poor. Path planning shows it needs 90 seconds to reach; Skill matching degree indicator: Excellent. "Advanced conflict mediator" qualification is highly matched with "physical conflict" event. Task interruption cost indicator: Low. It is currently "office guard", which can be interrupted.

[0090] Optimal performer determination: Since the event is determined to be "escalating trend", the system dynamically adjusts the weights of the weighted algorithm, setting the weight of "intervention time indicator" to the highest. After comprehensive calculation, the score of Security Officer A is higher than that of Security Officer B, and it is determined as the optimal performer.

[0091] Generate scheduling instructions: The system immediately sends instructions to Security Officer A's mobile terminal: "[Emergency] A area activity room northeast corner physical conflict, risk escalation, please go to deal with immediately!" The instruction is accompanied by a snapshot of the risk topology map of the scene and the recommended "level one response" level. At the same time, the system sends a secondary alarm to the command center and Security Officer B, suggesting that Security Officer B go to support as backup force.

[0092] Step 4: Post-processing and system learning

[0093] Security Officer A quickly arrives at the scene and successfully separates the two people before the conflict fully erupts. With the separation of personnel, the risk source is removed, and the "red" peak area on the aggregated risk topology map gradually subsides and returns to "green".

[0094] The system automatically records the whole process data of this event. Based on this event, the conflict relationship weight between A and B in the personnel relationship knowledge graph is further strengthened, so that the system will be more sensitive to their close contact in the future, thus achieving earlier warning.

[0095] The system acquires multi-source data such as personnel position, physiology and behavior in real time, and dynamically calculates individual risk scores by fusing surrounding personnel and historical interaction relationships. Subsequently, it generates discrete risk points into continuous and visual global risk topology map by using preset kernel function. Finally, the system can automatically analyze the causes and trends of risk peaks in the map, match the optimal performer through multi-objective optimization algorithm and generate precise scheduling instructions, thereby significantly improving risk warning capability and emergency disposal efficiency.

[0096] While embodiments of the present application have been shown and described with reference to particular embodiments thereof, it will be understood by those skilled in the art that various changes in form and details can be made therein without departing from the spirit and scope of the application. The scope of the application is defined by the appended claims and their equivalents.

Claims

1. A personnel intelligent scheduling management method based on Internet of Things, characterized in that, The application relates to a method for generating an aggregated risk topology map of a monitoring area. The method comprises the following steps: Real-time acquisition of multi-source data of each person in the monitoring area, wherein the multi-source data comprises real-time position data, physiological data and behavior data of each person; Dynamic calculation of a current individual risk score of each person based on a context-related risk assessment model, wherein the calculation of the individual risk score fuses multi-source data collected by the person himself / herself, real-time multi-source data of other persons in the surrounding area and a dynamically updated personnel relationship knowledge graph; Mapping of the physical space of the monitoring area into a coordinate system, taking the real-time position of each person as the coordinate center and taking the individual risk score as the weight, and generation of an aggregated risk topology map capable of representing the continuous distribution of risks in the whole monitoring area through a preset kernel function; Real-time analysis of the aggregated risk topology map, identification of a risk peak area with a risk value higher than a preset threshold, and determination of the risk cause type and dynamic evolution trend of the risk peak area; 2. The personnel intelligent scheduling management method based on the Internet of Things according to claim 1, characterized in that: And based on the risk peak area, risk level, risk type and dynamic trend, matching of an optimal executor and generation of a scheduling instruction containing a recommended response level through a multi-objective optimization scheduling algorithm. The context-related risk assessment model comprises a data input and preprocessing unit, an individual state assessment unit, a social context analysis unit and a risk fusion decision unit; the data input and preprocessing unit is used for receiving the multi-source data and performing cleaning, normalization and feature extraction on the multi-source data to generate a standardized input data stream; The individual state assessment unit calculates a basic risk level of a target person in an isolated state based on the standardized input data stream of the target person himself / herself; The social context analysis unit is used for querying the nodes and relationships of the target person and surrounding persons in the personnel relationship knowledge graph according to the real-time position of the target person, analyzing and quantifying the potential risk influence of the current social environment through graph calculation and reasoning, and generating a context risk correction factor; 3.The personnel intelligent scheduling management method based on the Internet of Things according to claim 1, characterized in that, The risk fusion decision unit is used for fusing the basic risk level output by the individual state assessment unit and the context risk correction factor output by the social context analysis unit, calculating through a weighting algorithm and outputting the individual risk score of the target person. The generation step of the aggregated risk topology map comprises the following steps: discretizing the physical space of the monitoring area into a two-dimensional and / or three-dimensional digital grid composed of a plurality of grid points, and performing zero processing on the initial risk values of all grid points; for each person in the monitoring area, taking the current real-time position as the center and taking the individual risk score as the amplitude, applying a preset kernel function to construct a risk distribution function representing the influence range and intensity of the individual risk of the person in space; traversing each grid point in the digital grid, calculating and accumulating the function values of the risk distribution functions of all persons on the grid point to obtain the accumulated aggregated risk value on the grid point; and rendering all grid points and the corresponding aggregated risk values to generate the aggregated risk topology map, wherein different aggregated risk values are visually presented through different color gradients. 4.The personnel intelligent scheduling management method based on the Internet of Things of claim 1, wherein, The generation step of the scheduling instruction comprises: feature extraction on the identified risk peak area to obtain its position, risk level, risk type and dynamic evolution trend as input parameters of scheduling decision; for each performer, based on the input parameters, the multi-objective optimization scheduling algorithm is used for comprehensive evaluation, and the sub-item index is calculated; the multiple sub-item indexes of each performer are weighted and calculated to determine the optimal performer under the current situation; and the scheduling instruction is generated for the optimal performer, which contains the intervention position, risk type summary and recommended response level.

5. The personnel intelligent scheduling management method based on the Internet of Things according to claim 4, characterized in that: The sub-item index includes intervention time index, skill matching degree index and task interruption cost index; the intervention time index is determined based on the shortest reachable path predicted travel time calculated by the path planning algorithm based on the current real-time position of the performer and the position of the risk peak area; the skill matching degree index is determined by comparing and quantitatively scoring the preset required skill corresponding to the risk type of the risk peak area with the skill items and proficiency level in the skill profile of the performer; and the task interruption cost index is a cost quantitative value comprehensively evaluated according to the preset priority of the task currently being performed by the performer and the attribute of whether the task can be safely interrupted. 6.The personnel intelligent scheduling management method based on the Internet of Things of claim 1, wherein, The personnel relationship knowledge graph is generated by a social network analysis method, specifically comprising: based on the real-time position data, continuously recording the close contact events of n personnel within a preset distance threshold in the monitoring area, and storing the contact duration and identity information; when the risk peak area appears on the aggregated risk topology graph and / or the external input security event report is received, the close contact events related to the event before and after the event occurrence time point are associated analyzed; according to the results of the association analysis, the relationship attributes between the personnel involved in the close contact events are quantitatively updated; the updated relationship attributes are stored in the relationship database for calling by the situational correlation risk assessment model when calculating the individual risk score.

7. An Internet of Things-based personnel intelligent scheduling management system, characterized in that, Comprise: Data acquisition module: real-time acquisition of multi-source data of each personnel in the monitoring area, the multi-source data including real-time position data, physiological data and behavior data of each personnel; risk assessment module: based on the situational correlation risk assessment model, dynamically calculating the current individual risk score of each personnel; The calculation of the individual risk score integrates the multi-source data collected by the personnel itself, the real-time multi-source data of other personnel in the surrounding area and a dynamically updated personnel relationship knowledge graph; Topology graph generation module: mapping the physical space of the monitoring area into a coordinate system, taking the real-time position of each personnel as the coordinate center, and taking the individual risk score as the weight, generating an aggregated risk topology graph that can represent the continuous distribution of risks in the entire monitoring area through a preset kernel function; Scheduling decision module: real-time analysis of the aggregated risk topology graph, identifying the risk peak area with a risk value higher than a preset threshold, and determining the risk cause type and its dynamic evolution trend of the risk peak area; And based on the risk peak area, risk level, risk type and dynamic trend, through multi-objective optimization scheduling algorithm, the optimal performer is matched and the scheduling instruction containing the recommended response level is generated.

Images