Express industry personnel management and occupational safety dynamic association analysis system and method

Abstract

The invention relates to the technical field of intelligent task scheduling and employee health management, and discloses an express industry personnel management and occupational safety dynamic association analysis system and method, which comprises the steps of collecting multi-dimensional data of employees in real time through various sensors, the multi-dimensional data comprising health data, environment data, psychological state data and work task data; performing standardization processing on the collected multi-dimensional data and generating a standardized data set; performing association analysis on the standardized data by using a graph neural network to generate a safety risk score of the employee; performing dynamic scheduling optimization on tasks of the employees through a reinforcement learning algorithm based on the multi-dimensional data; task scheduling and security strategies are dynamically adjusted through a feedback mechanism, and self-adaptive optimization and continuous learning of the system are ensured. Task scheduling and employee psychological state analysis are carried out by adopting a graph neural network and a multi-objective optimization algorithm, and the technical effect of dynamically optimizing task allocation is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent task scheduling and employee health management, and specifically to a system method for dynamic correlation analysis of personnel management and occupational safety in the express delivery industry. Background Art

[0002] Traditional task scheduling systems typically rely on fixed rules or manual experience to assign tasks. These methods are unable to adjust task allocation strategies in real time and, in particular, ignore changes in employee mental health and work environments. As workload and psychological stress increase, employee productivity can be impacted, even leading to burnout or mental health issues. However, existing technologies often fail to timely capture changes in employee mental states, resulting in irrational task allocation and an inability to effectively improve employee productivity and well-being.

[0003] Furthermore, existing task scheduling methods are often static and fail to fully account for individual employee differences. While some scheduling algorithms can allocate tasks based on employee capabilities, these often overlook employees' mental states, mood swings, and current health conditions. Without systematic feedback and dynamic adjustment mechanisms, high workloads can easily lead to long-term negative impacts on employee health, potentially causing employee turnover or sick leave, and increasing pressure on corporate management and human resources.

[0004] Furthermore, existing employee health management systems are mostly independent and lack effective integration with task scheduling systems. This information silo prevents synergy between task allocation and employee mental health management. For example, when an employee experiences depression due to excessive stress, the system cannot automatically adjust their workload or provide effective mental health intervention during work. Consequently, traditional technologies cannot form effective dynamic feedback and real-time intervention mechanisms in practical applications, which can lead to employees' physical and mental health issues not being addressed promptly.

[0005] In summary, existing technologies cannot accurately match task assignments with employee health status, lack real-time adjustment capabilities, and overlook the close relationship between employee mental health and task execution efficiency. These shortcomings severely restrict the effectiveness of task scheduling systems in complex work environments and increase the difficulty of enterprise management. Therefore, an intelligent system that comprehensively considers tasks, employee mental health, and the work environment is needed to address these issues. Summary of the Invention

[0006] In response to the shortcomings of the existing technology, the present invention provides a system method for dynamic correlation analysis of personnel management and occupational safety in the express delivery industry, which solves the problems in the existing technology such as lack of dynamic correlation analysis between employee health and work tasks, lack of real-time feedback mechanism and dynamic adjustment capability, failure to comprehensively consider multi-dimensional factors in employee management, and lack of long-term optimization and personalized management.

[0007] To achieve the above objectives, the present invention is implemented through the following technical solutions: a system method for dynamic correlation analysis of personnel management and occupational safety in the express delivery industry, comprising the following steps:

[0008] S1. Collecting multi-dimensional data of employees in real time through various sensors, including health data, environmental data, psychological state data, and work task data;

[0009] S2, standardize the collected multidimensional data and generate a standardized data set;

[0010] S3. Use graph neural networks to perform correlation analysis on standardized data to generate employee safety risk scores;

[0011] S4. Dynamically schedule and optimize employee tasks using reinforcement learning algorithms based on multidimensional data;

[0012] S5. Provide security warnings based on employees’ security risk scores, and issue warning signals when the scores fall below the preset threshold;

[0013] S6. Dynamically adjust task scheduling and security strategies through feedback mechanisms to ensure adaptive optimization and continuous learning of the system.

[0014] Preferably, the multiple sensors in step S1 include a heart rate monitor, a body temperature sensor, a blood oxygen sensor, an environmental sensor, an emotion monitoring device and a GPS positioning module.

[0015] Preferably, the step of standardizing and generating a standardized data set includes:

[0016] Normalize employee health data, environmental data, psychological status data, and work task data, converting each type of data into a unified standard scale;

[0017] Calculate the mean and standard deviation for each type of data and set the standardization formula:

[0018]

[0019] Where X is the data to be standardized; μ is the mean of the data; σ is the standard deviation of the data; X norm is the standardized data;

[0020] Merge various standardized data into a unified data set to provide standardized input for subsequent data analysis.

[0021] Preferably, the step S3 specifically includes:

[0022] Represent the standardized multidimensional data as a graph structure, where each node represents the data feature of an employee and each edge represents the relationship between employee data;

[0023] Use graph convolutional networks to process graph structure data. Node features are propagated through the features of adjacent nodes to update the security risk features of each node.

[0024] The security risk characteristics of each node are weighted and summed to generate the employee's comprehensive security risk score. The calculation formula is as follows:

[0025]

[0026] in, is the updated node feature of the l+1th layer; is the set of neighbor nodes of node v; c v u is the normalization coefficient between nodes v and u; W (l) is the weight matrix of the lth layer; b (l) is the bias term; σ is the activation function; is the updated feature vector of node u at layer l; u is a node adjacent to node v;

[0027] This method is used to iteratively calculate the features of each node until convergence, and finally output the safety risk score of each employee.

[0028] Preferably, the specific update rules of the reinforcement learning algorithm are as follows:

[0029]

[0030] Among them, Q(S t ,A t ) is state S t and action A t Q value; α is the learning rate; γ is the discount factor; R t is the reward value of the current action; S t+1 For the next state, A t+1 The action chosen for the next step;

[0031] By continuously iteratively updating the Q value, the system selects the optimal task scheduling plan based on the learned Q value, thereby maximizing overall employee safety and work efficiency.

[0032] Preferably, the step S5 specifically includes:

[0033] Generate a risk score for each employee based on the employee's safety score calculation formula. When the risk score falls below the set threshold, the system automatically triggers a safety warning.

[0034] Security warnings are displayed through the visual interface of the management platform, and administrators are notified through the alarm system to take appropriate intervention measures.

[0035] Preferably, the feedback mechanism is specifically manifested in that the system automatically adjusts task allocation and workload and optimizes safety strategies according to changes in health data, environmental data, psychological state data and work task data.

[0036] Preferably, the automatic adjustment of task allocation and workload is specifically performed by analyzing the current workload and health data, calculating the employee's workload index, and adjusting the workload using the following formula:

[0037] L v =f(h v ,t v );

[0038] Among them, L v is the workload of employee v; h v is the employee's health status characteristic; v is the current task load; f is the load adjustment function.

[0039] Preferably, the optimized security policy is specifically manifested in that when the security risk score of certain employees changes by more than a certain threshold, the system automatically adjusts the security policy. The adjustment of the security policy is achieved through the following rules:

[0040] S new =S current +ΔS;

[0041] Among them, S new is the adjusted security policy; S current is the current safety policy; ΔS is the change in the safety policy, which is dynamically calculated based on the employee's health, task situation, and environmental factors.

[0042] The express industry personnel management and occupational safety dynamic correlation analysis system preferably includes:

[0043] Data collection module: used to collect employees' health data, environmental data, psychological status data and work task data in real time. The data collection module includes multiple sensors and GPS positioning equipment;

[0044] Data processing module: used to standardize the collected data and transmit the standardized data to the data analysis module;

[0045] Data analysis module: used to perform correlation analysis on standardized data using graph neural networks to generate a safety risk score for each employee;

[0046] Task Scheduling Module: Through reinforcement learning algorithms, it dynamically schedules and optimizes employee tasks based on collected real-time data, optimizing task allocation to ensure employee health and safety;

[0047] Safety early warning module: Provides real-time safety warnings based on employees' safety risk scores, and automatically issues warning signals when the scores fall below the preset threshold;

[0048] Feedback and optimization module: Dynamically adjust task scheduling and security strategies through feedback mechanisms to achieve adaptive optimization and continuous learning of the system.

[0049] The present invention provides a system method for dynamic correlation analysis of personnel management and occupational safety in the express delivery industry. It has the following beneficial effects:

[0050] 1. This invention utilizes graph neural networks and multi-objective optimization algorithms for task scheduling and employee psychological state analysis, achieving the technical effect of dynamically optimizing task allocation. Compared to traditional task allocation methods in the prior art, this invention can analyze employees' psychological burden and task execution status in real time, avoiding the problem of over-assignment of tasks under high-pressure situations and reducing the impact of excessive work on employee health.

[0051] 2. This invention improves the personalization and accuracy of task scheduling through continuous learning and optimization of deep learning models. Compared with traditional static scheduling solutions, the system can continuously adjust task allocation based on employee health status, workload, and environmental changes, greatly improving work efficiency and employee satisfaction.

[0052] 3. This invention utilizes real-time monitoring and feedback mechanisms to intervene and manage employee mental health, achieving the technical benefits of preventing excessive psychological stress and improving employee work efficiency. Compared to existing single mental health management methods, this invention can flexibly adjust based on real-time data, promptly alleviating employee stress, thereby improving work efficiency and preventing psychological problems.

[0053] 4. This invention uses multiple sensors for data collection, combined with standardized processing and in-depth analysis, to achieve comprehensive optimization of task scheduling and employee health. Unlike traditional methods that rely on manual adjustment of task scheduling, this invention, through comprehensive data integration and scientific analysis, addresses the shortcomings of traditional task scheduling systems that fail to fully consider employee mental health and environmental factors, significantly improving the scientific and rational nature of task allocation. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 A diagram showing the steps of the method of the present invention;

[0055] Figure 2 It is a system module diagram of the present invention. DETAILED DESCRIPTION

[0056] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0057] like Figure 1 As shown, the express industry personnel management and occupational safety dynamic correlation analysis system method may include the following steps:

[0058] S1. Collecting multi-dimensional data of employees in real time through various sensors, including health data, environmental data, psychological state data, and work task data;

[0059] S2, standardize the collected multidimensional data and generate a standardized data set;

[0060] S3. Use graph neural networks to perform correlation analysis on standardized data to generate employee safety risk scores;

[0061] S4. Dynamically schedule and optimize employee tasks using reinforcement learning algorithms based on multidimensional data;

[0062] S5. Provide security warnings based on employees’ security risk scores, and issue warning signals when the scores fall below the preset threshold;

[0063] S6. Dynamically adjust task scheduling and security strategies through feedback mechanisms to ensure adaptive optimization and continuous learning of the system.

[0064] The following is a detailed description of each step in the method of the present invention, which comprehensively explains the specific implementation principles, technical details and processes of each step.

[0065] In step S1, employee health data is collected using wearable devices (such as smartwatches and heart rate monitors). These devices can monitor the employee's physiological status in real time, including key health indicators such as heart rate, blood oxygen saturation, and body temperature. The collected health data is sent to the data processing system via a wireless communication module (such as Bluetooth or Wi-Fi), ensuring immediate data transmission and processing.

[0066] Second, environmental data is collected by environmental sensors (such as temperature and humidity sensors, air quality sensors, and noise monitoring equipment). These sensors monitor temperature, humidity, air quality, noise, and other indicators in the employee's work environment, thereby assessing the impact of the environment on employee health. All environmental data is transmitted to a central data storage system via a sensor network.

[0067] At the same time, psychological state data is collected through emotion analysis and psychological assessment tools. For example, wearable devices combined with biofeedback technology can be used to monitor employees' stress levels, mood swings, fatigue, and other psychological states in real time. In addition to biofeedback data, psychological health assessment tools can also be used to further collect employee psychological state data. Specific implementation methods include:

[0068] Emotional self-assessment forms (e.g., self-assessment mood questionnaires): Employees regularly complete these self-assessment forms, recording their mood swings, feelings of mental fatigue, and perceived stress. These questionnaires include multiple emotional state assessment questions, such as "Do you feel anxious today?" and "Do you feel tired?" These self-assessment forms are completed via a mobile app or work platform, and the results are uploaded to a data system.

[0069] Mental Health Assessment System: The system regularly pushes mental health assessment forms (such as common depression, anxiety, and stress test scales) to help employees report their mental health status through self-assessment. This data can reflect employees' mood swings, anxiety levels, stress index, and other mental health indicators. Psychological data is automatically converted into digital information and uploaded to the system.

[0070] Furthermore, work task data is monitored in real time through integrated GPS positioning devices and a task management system. Data such as each employee's task assignment, task start and end times, task duration, and task intensity are recorded in real time. This information is combined with the employee's GPS location data through the task scheduling system to ensure the integrity and accuracy of work task data.

[0071] All of this data is consolidated and standardized through a unified data processing platform to ensure consistency and comparability, providing a foundation for subsequent analysis and decision-making. The data processing platform utilizes a highly integrated data management system to monitor, record, and analyze all types of data in real time. This platform integrates data from various devices into a unified database and performs data cleaning, noise removal, and preprocessing.

[0072] In step S2, the standardized data processing model must process multidimensional data including health data, environmental data, psychological status data, and work task data. Each type of data has different dimensions, units, and ranges in its original state, which can lead to errors and inconsistencies in subsequent analysis. Therefore, these data must be standardized to ensure that they are compared on the same scale.

[0073] In the data standardization process, the Z-score standardization method is used. The standardization formula is as follows:

[0074]

[0075] Where X is the data to be standardized; μ is the mean of the data; σ is the standard deviation of the data; X norm The data are standardized.

[0076] This formula converts each data item into a standard normal distribution (mean 0, standard deviation 1), eliminating dimensional differences between different data sources and making the data comparable. This standardized data can then be used in the subsequent graph neural network analysis module to facilitate correlation analysis.

[0077] Furthermore, the processing of each type of data is as follows:

[0078] Health data: This includes physiological indicators such as employee heart rate, body temperature, and blood oxygen levels. After collecting this data, its mean and standard deviation are calculated, and the data is converted using the standardized formula mentioned above to ensure that health data across different employees is comparable.

[0079] Environmental data: This includes environmental factors such as temperature, humidity, and air quality. Similar to health data, the mean and standard deviation of each environmental parameter are first calculated and then standardized.

[0080] Psychological state data: including stress index, mood swings and other indicators. Use the same standardized method to process employee psychological state data.

[0081] Work task data: including task start time, duration, work intensity, etc. Calculate the mean and standard deviation based on this task data to ensure the consistency of the task data.

[0082] After this standardization process, all data sets will be converted to the same standard scale, which is convenient for subsequent analysis model processing.

[0083] For step S3, the core of step S3 is the graph neural network (GNN) model, which establishes a graph structure to transmit and update information for different types of nodes. The following is the design and implementation process of the graph neural network model.

[0084] Node definition and graph structure construction

[0085] In a graph neural network, nodes represent various entities in the system, such as employees, work tasks, environmental factors, etc. Each node carries corresponding data features, such as the employee's mental health status, the difficulty of the work task, and environmental data.

[0086] Employee Node: Each employee node contains standardized mental health data, task load, emotional index, etc.

[0087] Task Node: Task nodes include the duration, intensity, and required concentration of the task;

[0088] Environment node: The environment node contains data such as temperature, humidity, and noise of the current working environment.

[0089] These nodes form the graph's vertices, connected by edges that form the graph's structure. Graph edges represent relationships between different nodes, such as the connection between employees and work tasks, or the interaction between employees and their environment. Edge weights represent the strength of the relationship between nodes. Weights can be calculated based on factors such as the fit between employees and tasks, their mental health, and the complexity of the tasks.

[0090] Data transmission and message update mechanism of graph neural network

[0091] The working principle of a graph neural network is to update the status of each node by passing information between nodes. In this implementation, the input data of the graph neural network model is standardized employee, task, and environment data, and the output is the prediction result after multiple rounds of information transmission.

[0092] In each layer, the feature vector of a node shares information with its neighboring nodes through a message passing mechanism. This process can be expressed by the following formula:

[0093]

[0094] in, is the updated node feature of the l+1th layer; is the set of neighbor nodes of node v; c v u is the normalization coefficient between nodes v and u; W (l) is the weight matrix of the lth layer; b (l) is the bias term; σ is the activation function; is the updated feature vector of node u in layer l; u is a node adjacent to node v.

[0095] Through multiple graph convolutions (the iterative process of graph neural networks), the feature vector of each node is gradually updated, and eventually the state of each node will reflect the information propagation results in the entire graph structure.

[0096] Weight calculation and optimization

[0097] The edge weights between nodes are a key factor affecting the propagation process of graph neural networks. In this implementation, edge weights are calculated based on the following aspects:

[0098] Match between task difficulty and employee v : By analyzing the employee's historical performance and task requirements, the matching degree between the task and the employee is calculated. The calculation rules are:

[0099]

[0100] Among them: TaskLoad v Task load, which indicates the effort and complexity required for the task; EmployeeSkill v is the employee's skill level, which indicates the employee's ability to complete the task, usually evaluated by historical performance; β1 and β2 are adjustment factors that control the relationship between task load and employee skills.

[0101] The relationship between employees' psychological state and task load v : Based on employees' mental health data, calculate the impact of workload on employees' mental health. The calculation rules are:

[0102]

[0103] Among them, HRV v Heart rate variability, reflecting employees' physiological stress response; WorkLoad v is the workload of the current task; Threshold v is the employee's psychological stress threshold, indicating the maximum workload that the employee can bear; β3 is the adjustment factor; exp(.) is the exponential function.

[0104] The impact of environmental factors on employee status v :The impact of environmental factors such as temperature, humidity, and noise on employee emotions and work efficiency is calculated as follows:

[0105] Env v =λ4·Noise v +λ5·Temperature v +λ6 Humidity v ;

[0106] Among them, Noisev is the ambient noise level; Temperature v is the ambient temperature; Humidity v is the ambient humidity; λ4, λ5, and λ6 are the weight coefficients of environmental factors, which are used to reflect the degree of influence of different environmental factors on the psychological state of employees.

[0107] Based on the above factors, the weighted summation method is used to calculate the weight of each edge, so the weight of each edge α vu It can be expressed as:

[0108] α vu =λ1·Match v +λ2·Stress v +λ3·Env v ;

[0109] Among them, Match v Based on the employee's historical task performance and task workload, the employee's past work efficiency, task completion quality and other indicators are used to evaluate the employee; Stress v is the employee's stress index, which indicates the employee's psychological adaptability to task load; Env v is the impact of environmental factors on the psychological state of employees, indicating the degree of influence of the environment on the emotional fluctuations of employees; λ1, λ2, and λ3 are the weight coefficients of each factor, which are used to reflect the importance of each factor.

[0110] Network output and decision support

[0111] The calculated results are provided as input to the subsequent task scheduling optimization module to ensure that employees can obtain tasks that are suitable for their current mental state and work ability, and achieve optimal work efficiency and mental health levels.

[0112] In step S4, the goal of task scheduling optimization is to build an optimization model and schedule tasks based on employee mental health, task load, and other relevant data (such as work environment). This model uses a multi-objective optimization algorithm to optimize task allocation by comprehensively considering factors such as employee ability, workload, task priority, and work environment.

[0113] The multi-objective optimization algorithm mainly considers the following objectives:

[0114] Minimize employees' psychological burden: Reduce employees' psychological pressure through reasonable task allocation.

[0115] Maximize task completion efficiency: Ensure tasks are completed on time and optimize the efficiency of task execution.

[0116] Minimize the adverse effects of environmental factors: Consider the impact of environmental factors on employees, such as temperature, humidity, noise, etc., and optimize the working environment configuration.

[0117] The objective function can be expressed as:

[0118] f(x)=α1·Stress(x)+α2·Efficiency(x)+α3·EnvironmentImpact(x);

[0119] Among them, Stress(x) is the psychological pressure of employees after task scheduling; Efficiency(x) is the efficiency of task completion; EnvironmentImpact(x) is the impact of the environment on employees; α1, α2, and α3 are the weight coefficients of each goal, indicating the importance of different goals during optimization.

[0120] In the task scheduling process, in addition to optimizing the objective function, the following constraints need to be considered:

[0121] Task duration constraint: The working time of each task cannot exceed the predetermined maximum duration;

[0122] Employee capability constraints: Employees’ work capabilities (e.g., physical strength, skills, etc.) should match task requirements;

[0123] Environmental adaptability constraints: Ensure that the impact of environmental factors on employees is minimized when tasks are assigned.

[0124] These constraints can be expressed by the following mathematical expressions:

[0125] T i ≤T max ;

[0126] C j ≥C min ;

[0127] E k ≤E max ;

[0128] Among them, T i The duration of task i is T max is the maximum duration limit of the task; C j is the ability of employee j; C min The minimum competence requirements for employees; k is the environmental adaptability level required for task k; E max To meet the maximum adaptability requirements of the environment.

[0129] By defining the objective function and constraints, an intelligent optimization algorithm such as genetic algorithm or particle swarm optimization (PSO) is used to solve the task scheduling optimization model. The specific process is as follows:

[0130] First, the task scheduling plan is initialized. Each task-to-employee matching plan is randomly generated, and the initial plan will be reasonably allocated based on task requirements and employee capabilities.

[0131] For each task scheduling solution, the defined objective function is used to evaluate it and calculate its fitness. The fitness value reflects the degree of optimization of the solution. The fitness calculation formula is as follows:

[0132] F(x)=f(x)+λ·∑ i Penalty i ;

[0133] Among them, F(x) is the fitness of the task scheduling scheme; f(x) is the objective function; Penalty i is the penalty term for violating the constraint conditions, which is used to ensure that the solution meets the constraints; λ is the penalty factor, which is used to balance the optimization objective and the penalty for constraint violation; i is different constraint conditions.

[0134] Based on fitness, suitable individuals are selected for crossover and mutation operations to generate new task scheduling solutions. The crossover operation swaps parts of the two parent task allocation solutions to generate a child solution; the mutation operation randomly adjusts the matching relationship between tasks and employees to increase diversity.

[0135] Through multiple iterations of optimization, the task scheduling scheme is continuously adjusted until the fitness value converges or reaches the predetermined stopping condition. After each iteration, the fitness evaluation is performed to select the optimal task scheduling scheme.

[0136] After optimization calculations, the optimal task scheduling plan is finally output. This plan includes: task allocation for each employee, execution time for each task, and environmental adaptation plan, etc.

[0137] Through the task scheduling system, the optimized scheduling plan is applied to actual work to achieve dynamic and personalized task allocation.

[0138] Regarding step S5, when the system detects an abnormality in the employee's mental state (e.g., a high stress index) or a delay in task execution, the system will automatically adjust the task. The basis for task allocation adjustment mainly includes the following aspects:

[0139] The relationship between employee psychological state and task load: If an employee's psychological burden is too heavy (such as a high stress index), the system will automatically reduce their task load or assign tasks to employees with less stress;

[0140] Task completion efficiency: If a task is lagging behind in completion, the system will consider adding more resources (such as assigning more employees or extending the task time) to ensure that the task can be completed on time.

[0141] The process of task adjustment can be expressed by the following formula:

[0142] T adjusted =T original +ΔT adjustment ;

[0143] Among them, T adjusted is the adjusted task time or load; T original is the original task time or load; ΔT adjustment It is the amount of task adjustment determined by task execution or employee psychological state.

[0144] In addition to adjusting task allocation, environmental factors also need to be adjusted. Environmental adjustments can include:

[0145] Temperature and humidity adjustment: Based on environmental data, it is recommended to adjust the air conditioning or air purification equipment to maintain appropriate operating temperature and humidity;

[0146] Noise Management: If the noise level is too high, it is recommended to use headphones, noise reduction equipment or adjust the work area.

[0147] Environmental adjustment recommendations are calculated using the following formula:

[0148] E adjusted =E original +ΔE adjustment ;

[0149] Among them, E adjusted is the environmental parameter after adjustment; E original is the original environmental parameter; ΔE adjustment It is the adjustment amount determined by the results of environmental data analysis.

[0150] Ultimately, the adjusted tasks and environment settings are fed back to employees through the task execution module and globally monitored and updated in real time by the central control system. As employees perform the adjusted tasks in their new work environment, the system continues to track their progress and their mental state to ensure the adjustments are effective.

[0151] Step S6 focuses on optimizing the task allocation system through continuous learning and data feedback mechanisms, and utilizing deep learning algorithms to continuously improve task scheduling strategies and employee management. This process relies not only on the accumulation of historical data but also includes long-term monitoring of employee work conditions, mental health, and environmental factors. The following is an implementation guide for step S6.

[0152] Building a deep learning model

[0153] By learning from historical task execution data, the task scheduling system is optimized. The model mainly includes the following parts:

[0154] Input data: Real-time monitoring data from step S5, including employee task execution data, mental health data, environmental data, etc. This data will serve as input to the deep learning model for multi-level feature extraction and analysis.

[0155] Output data: The output of the model is the optimized task scheduling strategy, including task allocation plan, employee mental health prediction value, environmental optimization suggestions, etc.

[0156] The deep learning model uses a multi-layer perceptron (MLP) structure. The input layer receives standardized task execution data, employee health data, and environmental data. Several hidden layers perform nonlinear transformations on this data, ultimately generating optimized task scheduling results through the output layer.

[0157] During the model training process, regression analysis and classification algorithms are used to simultaneously predict task completion time, employee mental health level, and task fit. The training data includes historical task data, employee mental state records, and environmental parameters.

[0158] During the training process, the mean square error (MSE) loss function is used to minimize the task scheduling error and mental health prediction error. The loss function is calculated as follows:

[0159]

[0160] Among them, y i Actual task completion time or employee mental health status; is the task completion time or mental health status predicted by the model; N is the number of training samples.

[0161] Through the back-propagation algorithm, the model parameters are gradually adjusted to optimize the accuracy of task scheduling prediction.

[0162] Data feedback and continuous learning mechanism

[0163] The data feedback mechanism ensures that the system can continuously adjust the task scheduling strategy based on newly acquired real-time data. This is done through the following methods:

[0164] Dynamically adjust the learning rate: The model dynamically adjusts the learning rate based on real-time task execution feedback to quickly adapt to environmental changes. For example, if an employee's mental health status changes significantly, the system will increase the learning rate to accelerate the adjustment.

[0165] Error Correction in Task Execution: By accumulating long-term data, the model can identify error patterns in task scheduling and correct these errors through adaptive optimization algorithms. For example, the system can detect that certain task types are poorly matched with specific employees and automatically adjust the task allocation strategy.

[0166] This continuous learning mechanism can make the system more intelligent and personalized, and adapt to various complex changes in the working environment.

[0167] Optimization algorithms and decision support

[0168] Based on the output of the deep learning model, the optimization algorithm further optimizes the task scheduling plan to ensure a balance between task execution efficiency and employee mental health. The algorithm includes the following steps:

[0169] Task scheduling strategy optimization

[0170] Based on the optimization results output by the model, the task scheduling system will combine information such as employee workload, task complexity, and environmental factors to generate a more refined task allocation plan. Specifically, the task scheduling optimization algorithm includes:

[0171] Capacity-based task allocation: Optimize task allocation based on each employee's historical performance, skills, and current workload to ensure that employees are assigned tasks within a reasonable range;

[0172] Stress adaptability optimization: When an employee's stress index is too high, the system will reduce their task load or adjust their working environment to ensure that the employee does not become overly fatigued.

[0173] Mental health prediction and intervention

[0174] By using deep learning models to predict employees' mental health conditions (such as stress levels and fatigue), the system can identify high-risk employees in advance and proactively provide psychological intervention measures, such as:

[0175] Rest reminder: When the system predicts that an employee's stress level is too high, it recommends that the employee take a break or adjust their work tasks;

[0176] Psychological counseling suggestions: Based on employees' emotional fluctuations, the system can recommend suitable psychological counseling courses or leisure activities to relieve employees' stress.

[0177] Implementation and Feedback

[0178] Task scheduling optimization plans and mental health intervention plans are delivered to employees in real time through the employee management module and task execution system, with ongoing tracking during execution. Employees receive real-time task adjustments and mental health management advice, ensuring tasks are completed efficiently and that employees receive appropriate psychological support.

[0179] In general, the present invention achieves the effect of improving employees' work efficiency and ensuring their mental health by combining real-time data collection, deep learning models and dynamic task scheduling optimization methods.

[0180] The express industry personnel management and occupational safety dynamic correlation analysis system described below and the express industry personnel management and occupational safety dynamic correlation analysis method described above can be referenced to each other.

[0181] Please see the attached Figure 2 The present invention also provides a dynamic correlation analysis system for personnel management and occupational safety in the express delivery industry, including:

[0182] Data collection module: used to collect employees' health data, environmental data, psychological status data and work task data in real time. The data collection module includes multiple sensors and GPS positioning equipment;

[0183] Data processing module: used to standardize the collected data and transmit the standardized data to the data analysis module;

[0184] Data analysis module: used to perform correlation analysis on standardized data using graph neural networks to generate a safety risk score for each employee;

[0185] Task Scheduling Module: Through reinforcement learning algorithms, it dynamically schedules and optimizes employee tasks based on collected real-time data, optimizing task allocation to ensure employee health and safety;

[0186] Safety early warning module: Provides real-time safety warnings based on employees' safety risk scores, and automatically issues warning signals when the scores fall below the preset threshold;

[0187] Feedback and optimization module: Dynamically adjust task scheduling and security strategies through feedback mechanisms to achieve adaptive optimization and continuous learning of the system.

[0188] The system of this embodiment can be used to execute the above method embodiments, and its principles and technical effects are similar, so they will not be repeated here.

[0189] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A systematic method for dynamic correlation analysis between personnel management and occupational safety in the express delivery industry, characterized by: The following steps are involved: S1. Collecting multi-dimensional data of employees in real time through various sensors, including health data, environmental data, psychological state data, and work task data; S2, standardize the collected multidimensional data and generate a standardized data set; S3. Use graph neural networks to perform correlation analysis on standardized data to generate employee safety risk scores; S4. Dynamically schedule and optimize employee tasks using reinforcement learning algorithms based on multidimensional data; S5. Provide security warnings based on employees’ security risk scores, and issue warning signals when the scores fall below the preset threshold; S6. Dynamically adjust task scheduling and security strategies through feedback mechanisms to ensure adaptive optimization and continuous learning of the system.

2. The express delivery industry personnel management and occupational safety dynamic correlation analysis system method according to claim 1 is characterized by: The multiple sensors in step S1 include a heart rate monitor, a body temperature sensor, a blood oxygen sensor, an environmental sensor, an emotion monitoring device, and a GPS positioning module.

3. The express delivery industry personnel management and occupational safety dynamic correlation analysis system method according to claim 1 is characterized in that: The steps of standardizing and generating a standardized data set include: Normalize employee health data, environmental data, psychological status data, and work task data, converting each type of data into a unified standard scale; Calculate the mean and standard deviation for each type of data and set the standardization formula: Where X is the data to be standardized; μ is the mean of the data; σ is the standard deviation of the data; X norm is the standardized data; Merge various standardized data into a unified data set to provide standardized input for subsequent data analysis.

4. The express delivery industry personnel management and occupational safety dynamic correlation analysis system method according to claim 1 is characterized in that: The step S3 specifically includes: Represent the standardized multidimensional data as a graph structure, where each node represents the data feature of an employee and each edge represents the relationship between employee data; Use graph convolutional networks to process graph structure data. Node features are propagated through the features of adjacent nodes to update the security risk features of each node. The security risk characteristics of each node are weighted and summed to generate the employee's comprehensive security risk score. The calculation formula is as follows: in, is the updated node feature of the l+1th layer; is the set of neighbor nodes of node v; c v u is the normalization coefficient between nodes v and u; W (l) is the weight matrix of the lth layer; b (l) is the bias term; σ is the activation function; is the updated feature vector of node u at layer l; u is a node adjacent to node v; This method is used to iteratively calculate the features of each node until convergence, and finally output the safety risk score of each employee.

5. The express delivery industry personnel management and occupational safety dynamic correlation analysis system method according to claim 1 is characterized in that: The specific update rules of the reinforcement learning algorithm are as follows: Among them, Q(S t ,A t ) is state S t and action A t Q value; α is the learning rate; γ is the discount factor; R t is the reward value of the current action; S t+1 For the next state, A t+1 The action chosen for the next step; By continuously iteratively updating the Q value, the system selects the optimal task scheduling plan based on the learned Q value, thereby maximizing overall employee safety and work efficiency.

6. The express delivery industry personnel management and occupational safety dynamic correlation analysis system method according to claim 1 is characterized in that: The step S5 specifically includes: Generate a risk score for each employee based on the employee's safety score calculation formula. When the risk score falls below the set threshold, the system automatically triggers a safety warning. Security warnings are displayed through the visual interface of the management platform, and administrators are notified through the alarm system to take appropriate intervention measures.

7. The express delivery industry personnel management and occupational safety dynamic correlation analysis system method according to claim 1 is characterized by: The feedback mechanism is specifically manifested in that the system automatically adjusts task allocation and workload and optimizes safety strategies according to changes in health data, environmental data, psychological state data and work task data.

8. The express delivery industry personnel management and occupational safety dynamic correlation analysis system method according to claim 7 is characterized in that: The automatic adjustment of task allocation and workload is specifically performed by analyzing the current workload and health data, calculating the employee's workload index, and adjusting the workload using the following formula: L v =f(h v ,t v ); Among them, L v is the workload of employee v; h v is the employee's health status characteristic; v is the current task load; f is the load adjustment function.

9. The express delivery industry personnel management and occupational safety dynamic correlation analysis system method according to claim 7 is characterized in that: The optimized security policy is specifically manifested in that when the security risk score of certain employees changes by more than a certain threshold, the system automatically adjusts the security policy. The security policy adjustment is achieved through the following rules: S new =S current +ΔS; Among them, S new is the adjusted security policy; S current is the current safety policy; ΔS is the change in the safety policy, which is dynamically calculated based on the employee's health, task situation, and environmental factors.

10. A system for analyzing the dynamic correlation between personnel management and occupational safety in the express delivery industry, configured to execute the method for analyzing the dynamic correlation between personnel management and occupational safety in the express delivery industry according to any one of claims 1 to 9, characterized in that: include: Data collection module: used to collect employees' health data, environmental data, psychological status data and work task data in real time. The data collection module includes multiple sensors and GPS positioning equipment; Data processing module: used to standardize the collected data and transmit the standardized data to the data analysis module; Data analysis module: used to perform correlation analysis on standardized data using graph neural networks to generate a safety risk score for each employee; Task Scheduling Module: Through reinforcement learning algorithms, it dynamically schedules and optimizes employee tasks based on collected real-time data, optimizing task allocation to ensure employee health and safety; Safety early warning module: Provides real-time safety warnings based on employees' safety risk scores, and automatically issues warning signals when the scores fall below the preset threshold; Feedback and optimization module: Dynamically adjust task scheduling and security strategies through feedback mechanisms to achieve adaptive optimization and continuous learning of the system.

Images