Intelligent peasant worker salary accounting and payment system based on digital twinning

Abstract

The invention discloses a peasant worker salary intelligent accounting and payment system based on digital twinning, which relates to the technical field of peasant worker salary intelligent accounting and payment and comprises an initial acquisition setting module, a data acquisition window module, a data analysis and feature extraction module, a task change classification module and a dynamic acquisition regulation and control module. The initial collection setting module sets an initial collection interval duration to collect data of worker work tasks according to the work arrangement of the construction site and the initial complexity of the worker tasks. According to the invention, the acquisition interval is intelligently and dynamically regulated and controlled, and the data acquisition mode is optimized according to the complexity and the change mode of the work task of a worker. For complex and changeable tasks, the system shortens the acquisition interval, accurately records high-intensity and low-intensity task switching, and ensures the fairness of salary accounting. For a single task, the collection interval is prolonged, redundant data is reduced, the resource utilization rate is improved, and the accuracy of salary accounting and the system operation efficiency are effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent calculation and payment of migrant workers' wages, and specifically relates to an intelligent calculation and payment system for migrant workers' wages based on digital twins. Background Art

[0002] The intelligent calculation and payment system for migrant workers' wages based on digital twins is an integrated solution that combines the Internet of Things, artificial intelligence, etc. This system creates a digital twin model of the construction site and collects the work data of migrant workers in real time, including attendance time, workload, and type of work information. Through AI algorithms for data analysis and wage calculation, it ensures the accuracy and fairness of wage calculation, realizes transparent and traceable wage payment, prevents wage arrears and disputes, and protects the legitimate rights and interests of migrant workers. This intelligent system improves the efficiency of construction site management and at the same time enhances the transparency and timeliness of migrant workers' wage payment.

[0003] The digital twin account system tool refers to relying on digital banking technology to establish a virtual model for contract deposit, process management, fund account management, fund transactions, fund settlement, etc. for all parties in the real - world scenario, which is mapped one - by - one with real transactions and records the deposit. After the calculation of migrant workers' wages is completed, wage payment will be carried out subsequently. The payment of migrant workers' wages mainly depends on the "special account for migrant workers' wages" (enterprise - level general account) opened by the enterprise and the project twin account. The enterprise first opens an enterprise - level general account in a bank with digital twin account technology, and then creates a project twin account for each engineering project on the platform. When the wages are allocated, the construction unit allocates the wages to the project twin account. When paying wages, after the general contractor for construction checks the wage list, it completes the wage payment to the workers through the platform. The wages enter the individual twin account of the workers from the project twin account, and the workers can withdraw the money to the physical bank card through the digital twin personal mini - program terminal. The platform ensures the accuracy of payment through three - factor authentication (worker's name, ID number, bank card number), and at the same time, the system supports a retry mechanism for failed withdrawals to ensure the safety of funds and the smooth completion of payment. Relying on the transaction fund supervision system of the head office, it provides comprehensive financial services such as online authentication of member status, fund sharing management, payment settlement, interest management, end - of - day reconciliation, and online printing of electronic vouchers for eligible enterprises and individual users of merchants.

[0004] In terms of virtual account management, it can strictly follow the bank's risk management requirements, execute the verification of the four - element business information of the enterprise and the online verification of personal public security information, and perform authentication verification on the settlement account bound to the virtual account. In addition, set the single - transaction / daily / annual cumulative limit for virtual account cash - out transactions, and add a multi - dimensional monitoring and early - warning mechanism at the system level to further prevent transaction risks. Adopt a leading microservices architecture design with dynamic load balancing to meet the high - frequency transaction processing needs of merchants and ensure the stable operation of the system.

[0005] When performing intelligent calculation of migrant workers' wages based on digital twins, the existing technology usually identifies the work tasks of migrant workers during the work process at a fixed collection interval, aiming to record the work conditions of workers in real time and accurately, including key data such as the types of tasks they perform, work intensity, and working hours. This method ensures that the construction site management system can automatically collect and update the labor data of workers, reduce errors and omissions in manual records, and provide an accurate data basis for wage calculation. Through this real-time data collection, the system can effectively track the work progress and efficiency of workers, adjust work arrangements in a timely manner, optimize resource allocation, and fairly reflect the actual labor achievements and contributions of workers during wage settlement.

[0006] The existing technology has the following deficiencies:

[0007] If the work tasks of workers are relatively complex and the set collection interval is large, the system may confuse or combine the high-intensity tasks completed by workers with low-intensity tasks, and calculate wages according to the low-intensity task standard. This will result in the underestimation of the time of high-intensity tasks completed by workers, thus making the actual wages lower than the due level. If the work tasks of workers are relatively single and the set collection interval is small, frequent data collection will result in the generation of a large amount of redundant data. These data do not add effective information to the record of single tasks, but instead increase the burden of data storage and processing.

[0008] The above information disclosed in the background art section is only used to strengthen the understanding of the background of the present disclosure. Therefore, it may include information that does not constitute the prior art known to those of ordinary skill in the art. Summary of the Invention

[0009] The object of the present invention is to provide an intelligent calculation and payment system for migrant workers' wages based on digital twins. By intelligently regulating the collection interval, it effectively solves the problems of data confusion and redundancy caused by improper collection methods. For complex and changeable work tasks, the system will shorten the collection interval to accurately capture task switching situations, prevent the underestimation of high-intensity task time, and ensure the fairness and accuracy of wage calculation. For single tasks, the system will extend the collection interval, reduce unnecessary data collection, reduce the burden of data storage and processing, and improve resource utilization rate. By dynamically adjusting the collection strategy, the system can adapt to different task modes, improve the accuracy of wage calculation and the operation efficiency of the system, so as to solve the problems in the above background art.

[0010] To achieve the above object, the present invention provides the following technical solution: An intelligent calculation and payment system for migrant workers' wages based on digital twins, including an initial collection setting module, a data collection window module, a data analysis and feature extraction module, a task change classification module, and a dynamic collection regulation module;

[0011] An initial acquisition setting module sets an initial acquisition interval duration to collect data on workers' work tasks according to the work arrangements at the construction site and the preliminary complexity of the workers' tasks.

[0012] A data acquisition window module creates an acquisition window of a fixed duration when collecting initial data. Within this acquisition window, it continuously collects and records data on workers' work tasks.

[0013] A data analysis and feature extraction module analyzes the work task data collected from workers after the initial data acquisition is completed under the acquisition window, extracts key features representing the changes in workers' work tasks, and uses a pre-trained machine learning model to evaluate the extracted features to identify the change patterns of workers' work tasks.

[0014] A task change classification module classifies the task change situations of workers under the acquisition window into three types according to the evaluation results of the machine learning model, including complex and variable type, normal change type, and simple change type.

[0015] A dynamic acquisition regulation module dynamically regulates the initial acquisition interval duration based on the evaluation results of the machine learning model for the complex and variable type and simple change type of work task changes, and regulates the actual acquisition interval duration for identifying migrant workers' work tasks to adapt to different task changes.

[0016] Preferably, the initial acquisition interval duration refers to the time interval initially set for collecting data on workers' work tasks when identifying workers' work tasks based on digital twin technology. Secondly, the initial acquisition interval duration needs to satisfy the requirement of effectively monitoring workers' work status without excessively consuming system resources.

[0017] Preferably, the key features representing the changes in workers' work tasks extracted include the frequency of workers switching from one task to another and the differences in the time spent on completing various tasks. After analyzing the frequency of workers switching from one task to another and the differences in the time spent on completing various tasks, a task switching frequency index and a task duration variation index are respectively generated. The task switching frequency index and the task duration variation index are input into a pre-trained machine learning model to generate a task change coefficient, and the change pattern of workers' work tasks is intelligently evaluated through the task change coefficient.

[0018] Preferably, the task change coefficient generated after evaluation using the machine learning model is compared and analyzed with a pre-set reference threshold range of the task change coefficient to divide the task change situations of workers under the acquisition window. The specific division steps are as follows:

[0019] If the task change coefficient is within the reference threshold range of the task change coefficient, the task change situation of the worker under the collection window is classified as a normal change type;

[0020] If the task change coefficient is less than the minimum value of the reference threshold range of the task change coefficient, the task change situation of the worker under the collection window is classified as a simple change type;

[0021] If the task change coefficient is greater than the maximum value of the reference threshold range of the task change coefficient, the task change situation of the worker under the collection window is classified as a complex change type.

[0022] Preferably, after analyzing the frequency of the worker switching from one task to another under the collection window, the specific steps for generating the task switching frequency index are as follows:

[0023] Within the collection window, record all the task sequences executed by the worker. Represent the task sequences with T. The task sequence represents the tasks recorded in chronological order within the collection window. Then: , where, where represents the i-th task;

[0024] By analyzing the task sequence T, count the number of different task types. The counting expression is: , in the formula, is the task type count, representing the set of different task types that appear in the task sequence, represents the operation of removing duplicates from the tasks in the task sequence T to obtain a set of non-repeated tasks. Unique means removing duplicate items and only retaining different and unique elements;

[0025] By calculating the difference degree between adjacent tasks of the worker within the collection window, measure the breadth of task switching. The calculation expression is: , represents the difference degree between adjacent tasks, measuring the breadth of the worker switching from one task to another, represents the task and the task between the difference degree, , where, and respectively represent the tasks and the task number, represents the absolute difference between task numbers, represents the maximum difference degree among all task pairs;

[0026] According to the task type count and the difference degree between adjacent tasks Calculate the task switching frequency index, and the calculation expression is: , where represents the task switching frequency index, which measures the diversity and dispersion of workers in task switching, is the logarithmic transformation of the task type count, representing the logarithmic scaling of task diversity, and n represents the total length of the task sequence, that is, the number of tasks within the acquisition window.

[0027] Preferably, after analyzing the differences in the time spent on completing various tasks under the acquisition window, the specific steps for generating the task duration variation index are as follows:

[0028] Within the acquisition window, collect the time data of all tasks completed by the worker, including the start time, end time, and task type of each task. Among them, represents the time spent by the worker on completing the kth task, and N represents the total number of tasks completed within the acquisition window;

[0029] Calculate the average value and standard deviation of the time spent by the worker on completing all tasks, and the calculation expression is: , , where represents the average task time, indicating the average time spent by the worker on each task, represents the task time standard deviation, reflecting the dispersion degree of task time, that is, the difference in the time spent by the worker on different tasks;

[0030] Calculate the task time coefficient of variation, and the calculation expression is: , where represents the task time coefficient of variation,

[0031] Calculate the task duration variation index based on the task time coefficient of variation and the task time standard deviation, which is used to describe the variation degree of task time. The calculation expression is: , where represents the task duration variation index, reflecting the complexity and variability of the worker's task time allocation, Adjust the task time coefficient of variation using the natural logarithm function, Combine the task time standard deviation with the total number of tasks to further quantify the impact of the dispersion degree of task time on the task duration variation index.

[0032] Preferably, for the complex and variable and simple variable work task changes, dynamically adjust the initial acquisition interval duration based on the evaluation results of the machine learning model. The specific steps are as follows:

[0033] Calculate the deviation between the task change coefficient and the reference threshold range of the task change coefficient. The calculation expression is: , where represents the task change coefficient, reflecting the degree of change in the current work task, and respectively represent the minimum and maximum values of the reference threshold range of the task change coefficient, represents the deviation of the task change coefficient;

[0034] Based on the deviation of the task change coefficient, an initial regulation factor is calculated to initially adjust the initial acquisition interval duration. The calculation expression is: , where represents the initial regulation factor, which is used to adjust the initial degree of the initial acquisition interval duration, represents the exponential decay term to avoid over-regulation, is used to evaluate the non-linear deviation of the task change coefficient relative to the reference threshold range of the task change coefficient, providing a more sensitive regulation response;

[0035] According to the initial regulation factor, a regulation factor adjustment term is calculated to meet the specific requirements of different task change patterns. The calculation expression is: , where represents the regulation factor adjustment term, which is based on the initial regulation factor and introduces a sine function to increase the sensitivity to the task change pattern, Introducing periodic changes through the sine function further enhances the flexibility of the initial regulation factor, enabling it to better adapt to various change patterns.

[0036] Preferably, according to the regulation factor adjustment term, the adjustment ratio of the initial acquisition interval duration is calculated. The calculation expression is: , where represents the adjustment ratio of the initial acquisition interval duration, reflecting the specific amplitude of the adjustment of the initial acquisition interval duration, Non-linearly adjust the initial acquisition interval duration according to the ratio of the regulation factor adjustment term and the task change coefficient;

[0037] According to the initially set interval duration and the actual interval duration adjustment ratio, the actual acquisition interval duration is updated to adapt to the current task change situation of the worker. The expression for updating the actual acquisition interval duration is: , where represents the updated actual acquisition interval duration, reflecting the optimal acquisition frequency based on the task change pattern, represents the initial acquisition interval duration.

[0038] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0039] The present invention intelligently and dynamically adjusts the acquisition interval duration based on the complexity and change pattern of workers' tasks, effectively solving the problems of data confusion and redundancy caused by inappropriate acquisition frequencies. When the workers' tasks are complex and variable, the system can shorten the acquisition interval duration to accurately capture the switching between high- and low-intensity tasks, avoiding underestimating the high-intensity task time, thereby ensuring the fairness and accuracy of wage calculation; when the workers' tasks are relatively simple, the system can extend the acquisition interval duration, reduce unnecessary data acquisition, lower the data storage and processing burden, and optimize the system resource utilization rate. Through this intelligent adjustment mechanism, the system can dynamically adjust the acquisition strategy according to task changes, adapt to different work task modes, and improve the accuracy of wage calculation and the operation efficiency of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] To more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required in the embodiments. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those of ordinary skill in the art, other drawings can also be obtained based on these drawings.

[0041] Figure 1 It is a schematic diagram of the modules of the intelligent wage calculation and payment system for migrant workers based on digital twins according to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0042] Now, the exemplary embodiments will be described more fully with reference to the accompanying drawings. However, the exemplary embodiments can be implemented in various forms and should not be construed as limited to the examples set forth herein; rather, these exemplary embodiments are provided so that this disclosure will be more complete and comprehensive, and will fully convey the concept of the exemplary embodiments to those skilled in the art.

[0043] The present invention provides an intelligent wage calculation and payment system for migrant workers based on digital twins as Figure 1 shown, including an initial acquisition setting module, a data acquisition window module, a data analysis and feature extraction module, a task change classification module, and a dynamic acquisition regulation module;

[0044] The initial acquisition setting module sets an initial acquisition interval duration to collect data on workers' tasks according to the work arrangements at the construction site and the preliminary complexity of the workers' tasks;

[0045] At this stage, the collected data mainly includes the location of the workers, task types, task durations, work intensities, etc. These data lay the foundation for subsequent analysis and evaluation.

[0046] The initial acquisition interval duration refers to the time interval initially set by the system for collecting workers' work task data when identifying workers' work tasks based on digital twin technology. This time interval is used to define the frequency of data collection by the system during the work process. For example, data is collected every 15 minutes or every 30 minutes, aiming to obtain detailed information about the workers' current work status. The setting of this initial time interval is usually determined based on the understanding of the construction site work arrangements and the preliminary complexity assessment of the workers' tasks.

[0047] The initial acquisition interval duration should be frequent enough to capture possible task changes of workers during the work process, ensuring that it can accurately reflect the actual situation of work tasks. Secondly, the time interval should not be too short to avoid generating too much redundant data and increasing the burden of data storage and processing. Therefore, the selection of the initial acquisition interval duration needs to balance the accuracy of data capture and the effective utilization of system resources, ensuring that it can effectively monitor the workers' work status without over-consuming system resources.

[0048] The data acquisition window module, when collecting initial data, establishes an acquisition window with a fixed duration (such as 1 hour). During this acquisition window, it continuously collects and records the workers' work task data;

[0049] The acquisition window with a fixed duration can help the system accumulate enough data volume in the initial stage, providing a complete information perspective for subsequent analysis.

[0050] Establishing an acquisition window with a fixed duration (such as 1 hour) when collecting initial data means that the system will continuously collect and record the workers' work task data within a preset time period. In this way, the system can obtain enough data samples in a short time to capture various work tasks and their changes of workers during this time period. The acquisition window with a fixed duration helps to initially establish a basic understanding of the workers' work patterns, facilitating the analysis of characteristics such as the time distribution, intensity changes, and task switching frequency of different tasks. The continuous acquisition of this window can also ensure the integrity and continuity of data, reducing information loss and errors caused by intermittent acquisition, enabling the system to have a comprehensive and accurate record of the workers' work conditions in the early stage, and laying a foundation for subsequent feature extraction and model evaluation.

[0051] The data analysis and feature extraction module, under the acquisition window, after the initial data collection is completed, analyzes the collected work task data of workers, extracts the key features representing the changes in workers' work tasks, and uses a pre-trained machine learning model to evaluate the extracted features to identify the change patterns of workers' work tasks;

[0052] The key features extracted to represent the changes in workers' tasks include the frequency of task switching from one task to another by workers and the differences in the time spent on completing various tasks. After analyzing the frequency of task switching from one task to another by workers and the differences in the time spent on completing various tasks, a task switching frequency index and a task duration variation index are generated respectively. The task switching frequency index and the task duration variation index are input into a pre-trained machine learning model to generate a task change coefficient, and the change pattern of workers' tasks is intelligently evaluated through the task change coefficient.

[0053] The pre-trained machine learning model refers to an algorithm model that has been trained and optimized with a large amount of historical data. These data include various change patterns of workers' tasks and their related features, such as the differences in task switching frequency and task duration. The training process usually involves inputting these feature data and using methods such as supervised learning, unsupervised learning, or reinforcement learning to enable the model to learn how to accurately predict or classify the change patterns of tasks based on these features. In the present invention, the pre-trained model has learned how to map the input data of the task switching frequency index and the task duration variation index into a task change coefficient, so as to effectively evaluate the complexity and variability of workers' tasks. Because the model has been verified and optimized through historical data, it can predict the newly input data with high accuracy and efficiency, providing a basis for the intelligent evaluation of the changes in workers' tasks.

[0054] Under the collection window, a relatively high frequency of task switching from one task to another by workers usually indicates a more complex change pattern of workers' tasks. This high-frequency task switching means that workers need to perform multiple different types of tasks within a short period of time, which not only requires workers to have a wide range of skills and adaptability, but may also involve different work intensities and operation processes. Frequent task switching increases the dynamics and uncertainty of the work, making task management more complex. This situation often requires more detailed monitoring and analysis to accurately reflect the workers' workload and labor contribution, and ensure the fairness and accuracy of wage calculation. A high frequency of task switching is an important sign of the complex and variable nature of workers' tasks, indicating that the working environment has a high degree of variability and challenge.

[0055] Under the collection window, after analyzing the frequency of task switching from one task to another by workers, the specific steps for generating the task switching frequency index are as follows:

[0056] Within the collection window, record all the task sequences executed by the workers. Represent the task sequences with T. The task sequences represent the tasks recorded in chronological order within the collection window. Then: , where, where represents the i-th task;

[0057] Each task (such as handling, installation, inspection, etc.) is recorded in chronological order, which can reflect the actual task switching situation of the workers.

[0058] By analyzing the task sequence T, the number of different task types is counted. The statistical expression is: , where is the task type count, representing the set of different task types that appear in the task sequence. represents the operation of removing duplicates from the tasks in the task sequence T to obtain a set of non-repeated tasks. Unique means removing duplicate items and only retaining different and unique elements.

[0059] This step helps to understand the task diversity involved by the workers within the collection window, that is, how many different tasks there are.

[0060] By calculating the difference degree between adjacent tasks of the worker within the collection window, the extent of task switching is measured. The calculation expression is: , represents the difference degree between adjacent tasks, measuring the extent of the worker's task switching from one task to another. represents task and task The difference degree between them. , where and respectively represent the numbers of task and task . represents the absolute difference between task numbers. represents the maximum difference degree among all task pairs.

[0061] plays a role in normalization, making the task switching breadth vary between 0 and 1. A larger difference degree indicates a more extensive and diverse task switching, reflecting the complexity of the tasks.

[0062] According to the task type count and the difference degree between adjacent tasks The task switching frequency index is calculated. The calculation expression is: , where represents the task switching frequency index, measuring the diversity and dispersion of the worker in task switching. is the logarithmic transformation of the task type count, representing the logarithmic scaling of task diversity. n represents the total length of the task sequence, that is, the number of tasks within the collection window.

[0063] It can be seen from the task switching frequency index that under the acquisition window, the larger the performance value of the task switching frequency index generated after analyzing the frequency of workers switching from one task to another, the more frequently workers change different tasks in a short period of time. This indicates that the pattern of workers' task changes is relatively complex, involving multiple task types and frequent task switches. Such high-frequency task switches may require workers to possess various skills and rapid adaptability. Therefore, it can be considered that the working environment and task arrangements for workers are relatively dynamic and diverse. Conversely, if the task switching frequency index is small, it indicates that the number of task types for workers within a certain period of time is small and stable, with infrequent changes. This means that the pattern of workers' task changes is relatively simple and consistent, and the task types are relatively single.

[0064] A large difference in the time spent on completing various tasks usually indicates a relatively complex pattern of workers' task changes. When workers need to complete multiple types of tasks within a short period of time, and the time required for each task varies greatly, this reflects the diversity and imbalance of tasks, indicating that workers face challenges of frequent task switches and adapting to different task requirements in their work. Such a large time difference often means that workers must continuously adjust their work rhythms and methods to meet the needs of various tasks, indicating the complexity and variability of work tasks. Therefore, a large task time difference is an important indicator of the complexity and variability of workers' work tasks.

[0065] Under the acquisition window, the specific steps for generating the task duration variation index after analyzing the difference in the time spent on completing various tasks are as follows:

[0066] Within the acquisition window, collect the time data of all tasks completed by workers, including the start time, end time, and task type of each task. Among them, use to represent the time (in minutes) spent by the worker on completing the kth task, and use N to represent the total number of tasks completed within the acquisition window;

[0067] Calculate the average value and standard deviation of the time spent by the worker on completing all tasks. The calculation expressions are: , , where represents the average task time, indicating the average time spent by the worker on each task, represents the task time standard deviation, reflecting the degree of dispersion of task times, that is, the difference in the time spent by the worker on different tasks;

[0068] Calculate the coefficient of variation of task time, which measures the relative degree of change in task time. Different from directly using the standard deviation of task time, the coefficient of variation of task time takes into account the influence of the average task time, enabling the variation in task time to be evaluated relative to the average value, and more accurately reflecting the relative degree of change in task time. The calculation formula for the coefficient of variation of task time is as follows: , where represents the coefficient of variation of task time, which is used to represent the relative variability of task time and standardizes the fluctuation of task time;

[0069] Based on the coefficient of variation of task time and the standard deviation of task time calculate the task duration variation index. The task duration variation index is a quantitative indicator used to describe the degree of variation in task time. The calculation formula is as follows: , where represents the task duration variation index, which reflects the complexity and variability of the worker's task time allocation. Adjust the coefficient of variation of task time using the natural logarithm function, so that the growth of the task duration variation index slows down as the coefficient of variation of task time increases, preventing the exponential growth from being too fast. Combine the standard deviation of task time with the total number of tasks to further quantify the impact of the dispersion degree of task time on the task duration variation index.

[0070] From the task duration variation index, it can be seen that under the acquisition window, the larger the value of the task duration variation index generated after analyzing the differences in the time spent on completing various tasks, usually indicates that the change pattern of the worker's tasks is more complex. This is because a higher task duration variation index reflects a greater difference in the time spent by the worker between different tasks, frequent task switching, and uneven changes in task intensity and duration, showing the complex and variable nature of the work tasks. On the contrary, a lower task duration variation index indicates that the difference in the time spent by the worker on various tasks is smaller, the task switching is less, or the task duration is relatively consistent, which means that the change pattern of the worker's tasks is relatively simple and stable. Therefore, the task duration variation index can be used as an effective indicator to evaluate the variability and complexity of the worker's tasks.

[0071] The machine learning model is not specifically limited here. Any machine learning model that can achieve comprehensive analysis of the task switching frequency index and the task duration variation index to generate the task change coefficient is acceptable. To implement the technical solution of the present invention, the present invention provides a specific implementation method;

[0072] Task change coefficient The generated calculation formula is as follows: , where and are the preset proportionality coefficients of the task switching frequency index and the task duration variation index respectively, and and are both greater than 0.

[0073] From the task change coefficient, it can be seen that under the acquisition window, the larger the performance value of the task switching frequency index generated by analyzing the frequency of workers switching from one task to another, and the larger the performance value of the task duration variation index generated by analyzing the differences in the time spent on completing various tasks, that is, the larger the performance value of the task change coefficient generated by using the machine learning model for evaluation, the more complex the change pattern of the workers' work tasks; on the contrary, it indicates that the change pattern of the workers' work tasks is less complex.

[0074] The task change classification module classifies the task change situation of workers under the acquisition window into three types according to the evaluation results of the machine learning model, including complex and changeable type, normal change type, and simple change type;

[0075] Compare and analyze the task change coefficient generated by using the machine learning model with the preset reference threshold range of the task change coefficient to divide the task change situation of workers under the acquisition window. The specific division steps are as follows:

[0076] If the task change coefficient is within the reference threshold range of the task change coefficient, then the task change situation of workers under the acquisition window is classified as the normal change type;

[0077] If the task change coefficient is less than the minimum value of the reference threshold range of the task change coefficient, then the task change situation of workers under the acquisition window is classified as the simple change type;

[0078] If the task change coefficient is greater than the maximum value of the reference threshold range of the task change coefficient, then the task change situation of workers under the acquisition window is classified as the complex change type.

[0079] The complex and changeable type means that the workers' work tasks in the acquisition window change frequently, and there are significant differences in the duration, intensity, and type of each task. This type usually includes frequent task switching, diverse task types, and significant task intensity fluctuations, indicating that workers need to continuously adapt to new tasks and work environments. This high-dynamic and high-complexity task pattern may lead to a higher work load and requires workers to have strong multitasking capabilities and flexibility to cope with various challenges in the work.

[0080] The normal change type describes that the frequency and type of workers' work tasks within the collection window are relatively stable, and the switching between tasks is regular and predictable. This type of task change does not frequently disrupt the workers' work rhythm, and the task intensity and duration are roughly the same or have slight variations. This situation usually reflects a common work pattern, in which workers have time to adapt to task switching, and the tasks themselves do not require rapid adaptation to new work requirements. Workers can complete tasks methodically without feeling fatigued or overly stressed due to frequent changes.

[0081] The simple change type means that there are very few changes in the workers' work tasks within the collection window. Even during the entire window period, only one type of task is performed, and the task type, intensity, and duration are all very consistent. This type of task pattern reflects the simplicity and repetitiveness of the workers' work content, and task switching hardly exists or occurs extremely slowly. This work pattern usually does not require workers to have a high adaptability because the work tasks are fixed and regular, and workers can form work proficiency and efficiency through repeated operations. This type of task may be suitable for simple operations or work environments that do not require frequent adjustments.

[0082] The dynamic collection control module dynamically adjusts the initial collection interval duration based on the evaluation results of the machine learning model for the complex and changeable type and the simple change type of work task changes, and adjusts the actual collection interval duration for the identification of migrant workers' work tasks to adapt to different task changes;

[0083] For the complex and changeable type and the simple change type of work task changes, the initial collection interval duration is dynamically adjusted based on the evaluation results of the machine learning model. The specific steps are as follows:

[0084] Calculate the deviation between the task change coefficient and the reference threshold range of the task change coefficient. The calculation expression is: , where represents the task change coefficient, reflecting the degree of change of the current work task, and represent the minimum and maximum values of the reference threshold range of the task change coefficient respectively, represents the task change coefficient deviation;

[0085] This step helps to determine the deviation of the task change degree from the expected range, providing a basis for the subsequent adjustment of the collection interval duration.

[0086] Based on the task change coefficient deviation, calculate the initial adjustment factor for initially adjusting the initial collection interval duration to ensure that it can quickly respond to the needs of different task change types. The calculation expression is: , where Represents the initial regulation factor, which is used to adjust the initial degree of the initial acquisition interval duration. Represents the exponential decay term, which ensures that the initial regulation factor gradually levels off in extreme cases (such as when the task change coefficient is extremely large or extremely small), avoiding over-regulation. It is used to evaluate the non-linear deviation of the task change coefficient relative to the reference threshold range of the task change coefficient, providing a more sensitive regulation response.

[0087] The initial degree for adjusting the initial acquisition interval duration means that in the process of dynamically regulating the data acquisition interval duration of workers' work tasks, the system first needs to determine a preliminary adjustment amount or direction based on the preliminary assessment of task changes. This initial adjustment degree is determined based on the task change coefficient and the task change coefficient deviation of the current task change, aiming to quickly respond to the needs of task change types.

[0088] Specifically, when the task change pattern is evaluated as complex and variable or simple change type, the initial regulation factor will determine how the initial acquisition interval duration should be adjusted (increased or decreased). If the task change is evaluated as complex and variable, the initial regulation factor will tend to shorten the initial acquisition interval duration in order to collect data more frequently to capture the rapidly changing task situation. On the contrary, if the task change is evaluated as simple change type, the initial regulation factor will tend to extend it, which may tend to reduce the data acquisition frequency, reduce unnecessary data redundancy and system resource consumption.

[0089] According to the initial regulation factor, calculate the regulation factor adjustment term to adapt to the specific needs of different task change patterns. The calculation expression is: , where Represents the regulation factor adjustment term, which is based on the initial regulation factor and introduces a sine function to increase the sensitivity to the task change pattern. Introduce periodic changes through the sine function to further enhance the flexibility of the initial regulation factor, enabling it to better adapt to various change patterns.

[0090] According to the regulation factor adjustment term, calculate the adjustment ratio of the initial acquisition interval duration. The calculation expression is: , where Represents the adjustment ratio of the initial acquisition interval duration, reflecting the specific amplitude of the adjustment of the initial acquisition interval duration. Non-linearly adjust the initial acquisition interval duration according to the ratio of the regulation factor adjustment term and the task change coefficient, ensuring to increase the frequency in high-complexity tasks and decrease the frequency in low-complexity tasks.

[0091] Adjust the actual acquisition interval duration according to the initially set interval duration and the adjustment ratio of the actual interval duration, and update the actual acquisition interval duration to adapt to the current task change situation of the worker. The expression for updating the actual acquisition interval duration is: , where represents the updated actual acquisition interval duration, reflecting the optimal acquisition frequency based on the task change pattern, represents the initial acquisition interval duration.

[0092] The present invention intelligently and dynamically adjusts the acquisition interval duration through the complexity and change pattern of the worker's work tasks, effectively solving the problems of data confusion and redundancy caused by inappropriate acquisition frequencies. When the worker's work tasks are complex and variable, the system can shorten the acquisition interval duration, accurately capture the switching of high- and low-intensity tasks, and avoid underestimating the high-intensity task time, thereby ensuring the fairness and accuracy of wage calculation; when the worker's work tasks are relatively simple, the system can extend the acquisition interval duration, reduce unnecessary data acquisition, reduce the data storage and processing burden, and optimize the system resource utilization rate. Through this intelligent adjustment mechanism, the system can dynamically adjust the acquisition strategy according to task changes, adapt to different work task modes, and improve the accuracy of wage calculation and the operation efficiency of the system.

[0093] The above formulas are all calculated by taking the numerical values without dimensions. The formula is obtained by software simulation of a large amount of collected data to get a formula closest to the actual situation. The preset parameters in the formula are set by those skilled in the art according to the actual situation.

[0094] The above is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily think of changes or substitutions, which should all be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

[0095] Only some exemplary embodiments of the present invention have been described above by way of illustration. Undoubtedly, for those of ordinary skill in the art, the described embodiments can be modified in various different ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the protection scope of the claims of the present invention.

Claims

1. The intelligent calculation and payment system for migrant workers’ wages based on digital twins is characterized by: It includes initial collection setting module, data collection window module, data analysis and feature extraction module, task change classification module and dynamic collection control module; The initial collection setting module sets an initial collection interval to collect data on workers' work tasks based on the work schedule of the construction site and the initial complexity of the workers' tasks; The data collection window module establishes a fixed-length collection window when collecting initial data. During this collection window, the worker's work task data is continuously collected and recorded. The data analysis and feature extraction module, in the collection window, analyzes the work task data collected by the workers after the initial data collection is completed, extracts the key features representing the changes in the workers' work tasks, and uses the pre-trained machine learning model to evaluate the extracted features and identify the change pattern of the workers' work tasks; The task change classification module classifies the worker's task changes within the collection window into three types, including complex and variable types, normal change types, and simple change types, based on the evaluation results of the machine learning model. The dynamic collection and control module dynamically adjusts the initial collection interval duration based on the evaluation results of the machine learning model for complex and changeable work tasks and simple and changeable work tasks, and adjusts the actual collection interval duration of migrant workers' work task identification to adapt to different task changes.

2. The intelligent calculation and payment system for migrant workers' wages based on digital twins according to claim 1 is characterized in that: The initial collection interval refers to the time interval set for collecting workers' work task data when identifying workers' work tasks based on digital twin technology. Secondly, the initial collection interval must meet the requirements of effectively monitoring the workers' work status without excessively consuming system resources.

3. The intelligent calculation and payment system for migrant workers' wages based on digital twins according to claim 1 is characterized in that: The key features extracted to represent the changes in workers' work tasks include the frequency of workers switching from one task to another and the differences in the time taken to complete various tasks. After analyzing the frequency of workers switching from one task to another and the differences in the time taken to complete various tasks, a task switching frequency index and a task duration variation index are generated respectively. The task switching frequency index and the task duration variation index are input into a pre-trained machine learning model to generate a task change coefficient. The task change coefficient is used to intelligently evaluate the change pattern of workers' work tasks.

4. The intelligent calculation and payment system for migrant workers' wages based on digital twins according to claim 3 is characterized in that: The task change coefficient generated after evaluation using the machine learning model is compared and analyzed with the preset reference threshold range of the task change coefficient, and the task change situations of the workers under the collection window are divided. The specific division steps are as follows: If the task change coefficient is within the task change coefficient reference threshold range, the worker's task change situation in the collection window is classified as a normal change type; If the task change coefficient is less than the minimum value of the task change coefficient reference threshold range, the worker's task change situation under the collection window is classified as a simple change type; If the task change coefficient is greater than the maximum value of the task change coefficient reference threshold range, the worker's task change situation within the collection window is classified as a complex change type.

5. The intelligent calculation and payment system for migrant workers' wages based on digital twins according to claim 3 is characterized in that: After analyzing the frequency of workers switching from one task to another in the collection window, the specific steps to generate the task switching frequency index are as follows: In the collection window, all task sequences performed by the workers are recorded. The task sequence is represented by T. The task sequence represents the tasks recorded in chronological order in the collection window. Then: , among which represents the i-th task; By analyzing the task sequence T, the number of different task types is counted. The statistical expression is: , where is the task type count, which represents the set of different task types that appear in the task sequence. It means to perform a deduplication operation on the tasks in the task sequence T to obtain a set of non-duplicate tasks. Unique means to remove duplicate items and only keep different and unique elements. The degree of task switching is measured by calculating the difference between adjacent tasks of workers within the acquisition window. The calculation expression is: , represents the difference between adjacent tasks, measuring the extent to which workers switch from one task to another. Indicates the task With the task The difference between ,in, and Respectively represent tasks and tasks The number of represents the absolute difference between task numbers, represents the maximum difference between all task pairs; Count by task type The difference between adjacent tasks Calculate the task switching frequency index, the calculation expression is: , where represents the task switching frequency index, which measures the diversity and dispersion of workers in task switching. The logarithmic transformation of the task type count represents the logarithmic scaling of task diversity, and n represents the total length of the task sequence, that is, the number of tasks within the acquisition window.

6. The intelligent calculation and payment system for migrant workers' wages based on digital twins according to claim 3 is characterized in that: In the collection window, after analyzing the differences in the time taken to complete various tasks, the specific steps to generate the task duration variation index are as follows: In the collection window, the time data of all tasks completed by the workers are collected, including the start time, end time and task type of each task. represents the time it takes a worker to complete the kth task, and N represents the total number of tasks completed within the acquisition window; Calculate the mean and standard deviation of the time it takes workers to complete all tasks. The expression for calculation is: , , where represents the average task time, which means the average time workers spend on each task. It represents the standard deviation of task time, which reflects the discreteness of task time, that is, the difference in the time workers spend on different tasks; Calculate the task time difference coefficient, the calculation expression is: , where represents the task time variance coefficient, The task duration variation index is calculated based on the task time difference coefficient and the task time standard deviation. It is used to describe the degree of variation of the task time. The calculation expression is: , where It represents the task duration variation index, reflecting the complexity and variability of workers’ task time allocation. The natural logarithmic function is used to adjust the task time difference coefficient. The standard deviation of task time is combined with the total number of tasks to further quantify the impact of the dispersion of task time on the task duration variation index.

7. The intelligent calculation and payment system for migrant workers' wages based on digital twins according to claim 4 is characterized in that: In response to the changes in complex and variable work tasks and simple and variable work tasks, the initial collection interval is dynamically adjusted based on the evaluation results of the machine learning model. The specific steps are as follows: Calculate the deviation between the task variation coefficient and the reference threshold range of the task variation coefficient. The calculation expression is: , where It represents the task change coefficient, which reflects the degree of change of the current work task. and Respectively represent the minimum and maximum values ​​of the reference threshold range of the task variation coefficient, represents the task variation coefficient deviation; Based on the task variation coefficient deviation, the initial control factor is calculated to preliminarily adjust the initial collection interval. The calculation expression is: , where Represents the initial control factor, which is used to adjust the initial degree of the initial collection interval. represents an exponential decay term to avoid over-regulation, It is used to evaluate the nonlinear deviation of the task variation coefficient relative to the reference threshold range of the task variation coefficient, and provide a more sensitive regulatory response; According to the initial control factor, the control factor adjustment item is calculated to adapt to the specific needs of different task change modes. The calculation expression is: , where represents the control factor adjustment term, which is based on the initial control factor and introduces a sine function to increase the sensitivity to the task change pattern. The flexibility of the initial control factors is further enhanced by introducing periodic changes through the sine function.

8. The intelligent calculation and payment system for migrant workers' wages based on digital twins according to claim 7 is characterized in that: According to the control factor adjustment item, the adjustment ratio of the initial collection interval is calculated. The calculation expression is: , where Indicates the adjustment ratio of the initial collection interval, reflecting the specific adjustment range of the initial collection interval. According to the ratio of the control factor adjustment item and the task change coefficient, the initial collection interval duration is adjusted nonlinearly; According to the ratio of the initial interval and the actual interval, the actual collection interval is updated to adapt to the current task changes of the workers. The expression for updating the actual collection interval is: , where Indicates the actual updated collection interval duration, reflecting the optimal collection frequency based on the task change pattern. Indicates the initial collection interval duration.

Images