Multi-terminal cooperation intelligent scheduling platform and method for flexible employment management

Abstract

The invention discloses a multi-terminal cooperation intelligent scheduling platform and method for flexible employment management, and particularly relates to the technical field of employment management, and the method comprises the following steps: presetting a scheduling rule algorithm, carrying out the work scheduling according to flexible employment demands, and generating a scheduling arrangement form of each time; in each scheduling period, the fairness and rationality of work arrangement are evaluated by extracting a feature set influencing scheduling, and an evaluation report is generated; and generating a feedback signal according to the evaluation report, and determining whether to optimize the scheduling rule algorithm. In the optimization process, an optimization scheme is designed through feedback signals, a scheduling algorithm is iteratively trained, so that the scheduling algorithm better adapts to the employment demand, it is ensured that work distribution is more fair and reasonable, and the optimized algorithm is applied to scheduling of the next period; according to the method, whether task allocation is fair and reasonable or not can be scientifically evaluated, targeted optimization training can be carried out on unreasonable parts in task allocation, and more scientific and efficient task allocation in a labor use environment is gradually realized.

Description

Technical Field

[0001] The present invention relates to the field of employment management technology, and more specifically, to a multi-terminal collaborative intelligent scheduling platform and method for flexible employment management. Background Art

[0002] With the rapid development of the economy and increasingly fierce market competition, enterprises' demand for labor is becoming more flexible and diverse. Flexible employment has gradually become an important human resource management method and is widely used in e-commerce, logistics, services and other industries. Traditional employment management models are often based on long-term labor relations and cannot effectively respond to temporary, phased or dynamic labor needs. This rigid employment management method is prone to the following problems: fixed employment models are prone to waste of labor resources when demand fluctuates greatly, and there may be a shortage of labor during peak periods, which makes it difficult to meet the flexible and efficient management needs of enterprises. Traditional scheduling methods are usually based on manual experience and lack scientific and intelligent support, which can easily lead to unfair or unreasonable task allocation, affecting employment efficiency and employee satisfaction. When enterprises face dynamic task requirements, the existing scheduling system lacks real-time adjustment capabilities, resulting in delayed task allocation and affecting the overall operation effect. Therefore, this paper proposes a multi-terminal collaborative intelligent scheduling platform and method for flexible employment management to solve the above problems. Summary of the invention

[0003] To achieve the above object, the present invention provides the following technical solutions:

[0004] The multi-terminal collaborative intelligent scheduling method for flexible employment management includes the following steps:

[0005] A scheduling rule model is preset, and the preset scheduling rule model is used to schedule work according to flexible employment needs, and obtain a work scheduling arrangement form for each time;

[0006] In each scheduling cycle, based on all the work scheduling forms in the scheduling cycle, feature extraction is performed to extract relevant features that affect the labor scheduling and obtain an influencing feature set;

[0007] According to the impact feature set, the fairness and rationality of each work scheduling result within the scheduling cycle are evaluated and an evaluation report is generated;

[0008] Generate different types of feedback signals according to the evaluation report, and decide whether to optimize the preset scheduling rule model according to the type of feedback signal;

[0009] During optimization training, an optimization training plan is generated according to the type of feedback signal. The preset scheduling rule model is optimized through iterative training to gradually adapt it to the scheduling needs in the employment environment. The optimized scheduling rule model is applied to the next cycle of employment scheduling to ensure that work distribution is more fair and reasonable.

[0010] In a preferred embodiment, extracting relevant features that affect employment scheduling refers to extracting a fairness feature group and a rationality feature group respectively, and the fairness feature group and the rationality feature group together constitute an influencing feature set.

[0011] In a preferred embodiment, the evaluation report includes a scheduling cycle fairness index obtained by performing a fairness evaluation and a scheduling cycle rationality index obtained by performing a rationality evaluation.

[0012] In a preferred implementation, the logic for obtaining the scheduling cycle fairness index is:

[0013] For each worker i, its workload and working hours They are defined as:

[0014] ; Indicates the preset difficulty coefficient corresponding to the workload in time period t;

[0015] ; represents the working time in time period t, Indicates whether worker i works in time period t. If he works in time period t, then The value is 1. If it does not work during time period t, then The value is 0, and T represents the total number of time periods;

[0016] Get the overall workload and work duration balance target:

[0017] First, calculate the overall workload and working time respectively. Suppose the total workload and total working time are and : ; ; n represents the total number of workers;

[0018] Then calculate the equilibrium target of each worker, namely the target workload and target working time:

[0019] Target workloads: ; represents the target workload of each worker;

[0020] Target working hours: ; represents the target working hours for each worker;

[0021] The workload deviation and working time deviation of each worker are:

[0022] ; represents the workload deviation of worker i;

[0023] ; represents the working time deviation of worker i;

[0024] Introducing differences in worker preferences for adjustment:

[0025] ; represents worker i’s preference or resistance to feedback in time period t. If worker i prefers to work in time period t, then The value is 1, if you resist working in time period t, then The value is 0. represents the preference difference value of worker i, represents the preset preference ratio coefficient of worker i in time period t;

[0026] The calculation formula of scheduling cycle fairness index is:

[0027] ; , are all preset non-zero trade-off coefficients, represents the total rest time of worker i, represents the minimum allowed rest time, F represents the scheduling cycle fairness index, and C is a preset non-zero constant.

[0028] In a preferred embodiment, the logic for obtaining the scheduling cycle rationality index is:

[0029] The timeliness of task delivery is used to measure whether the task can be completed within the scheduled time. The calculation formula is as follows:

[0030] ; N represents the total number of work tasks, represents the deadline of the kth task, represents the actual completion time of the kth task, and TI represents the timeliness of task delivery;

[0031] The worker load balance is used to evaluate the reasonable distribution of each worker's working time and load. The calculation formula is as follows:

[0032] ; represents the actual working time of the ith worker, represents the total time required for the task assigned to the i-th worker, LB represents the balance of worker load, and C is a preset non-zero constant;

[0033] The reasonable matching degree of personnel scheduling is evaluated by the matching adaptability between task requirements and worker scheduling. The calculation formula is as follows:

[0034] ; represents the similarity between the skills of the ith worker and the skills required for each scheduling task, represents the experience value of the i-th worker, , They are all preset non-zero measurement coefficients, and PP represents matching suitability;

[0035] Get the distances from task delivery timeliness TI, worker load balance LB, and matching adaptability PP to the ideal solution and negative ideal solution. Let the distance to the ideal solution be JL1 and the distance to the negative ideal solution be JL2. The calculation formula for the scheduling cycle rationality index is:

[0036] ; Y represents the scheduling cycle rationality index.

[0037] In a preferred embodiment, the experience value of the i-th worker The acquisition logic is:

[0038] ; V represents the total number of historical scheduling times of the i-th worker, v represents the index of the historical scheduling times, Indicates the preset adjustment factor, represents the quality score of the vth scheduling of the ith worker, represents the historical average quality score of the ith worker.

[0039] In a preferred embodiment, generating different types of feedback signals according to the evaluation report refers to:

[0040] Using fuzzy reasoning, the scheduling cycle fairness index and the scheduling cycle rationality index are used as input data for fuzzy reasoning. The output result of fuzzy reasoning is the type of feedback signal. The types of feedback signals include class one, class two, class three, and class four. Only when the type of feedback signal is class one, the preset scheduling rule model is not optimized and trained.

[0041] In a preferred embodiment, during optimization training, generating an optimization training scheme according to the type of feedback signal refers to:

[0042] The number of training data of a type of feedback signal is selected as Z, and the training data of the type of feedback signal corresponding to the training number in the range of [2Z, 4Z] are selected to together constitute a training set for optimization training, and in the training set for optimization training, the number of all training data is not less than a preset minimum training quantity threshold.

[0043] In a preferred embodiment, a multi-terminal collaborative intelligent scheduling platform for flexible employment management includes:

[0044] The work scheduling module has a preset scheduling rule model, and the preset scheduling rule model is used to schedule work according to the flexible employment needs and generate a work scheduling form for each time;

[0045] The feature extraction module performs feature extraction in each scheduling cycle based on all work scheduling forms in the scheduling cycle, extracts relevant features that affect the labor scheduling, and obtains an influencing feature set;

[0046] The evaluation module evaluates the fairness and rationality of each work scheduling result within the scheduling cycle based on the impact feature set and generates an evaluation report;

[0047] A feedback module generates different types of feedback signals according to the evaluation report, and determines whether to optimize the preset scheduling rule model according to the type of feedback signal;

[0048] The optimization module generates an optimization training plan according to the type of feedback signal during optimization training, optimizes the preset scheduling rule model through iterative training, so that it gradually adapts to the scheduling needs under the employment environment, and transmits the optimized scheduling rule model to the work scheduling module. The optimized scheduling rule model is applied to the employment scheduling of the next cycle to ensure that the work distribution is more fair and reasonable.

[0049] Technical effects and advantages of the present invention:

[0050] Based on the convolutional neural network intelligent algorithm, the present invention can quickly analyze and match the input labor demand and worker characteristics, automatically generate task allocation plans, and significantly reduce the time cost of traditional manual scheduling. The scheduling platform supports real-time updating of scheduling plans, can quickly respond to sudden task demands, realize dynamic scheduling, and avoid reducing work efficiency due to delays or insufficient resource allocation.

[0051] Through the fairness index, the system can scientifically evaluate whether the task allocation is fair, avoid excessive concentration of tasks on a small number of people, reduce employee dissatisfaction, and the scheduling plan fully considers the workers' skill level, historical experience and actual needs, generates a rationality index for task allocation, and ensures that the allocation meets the actual needs of the position and the difficulty of the task. The system analyzes the fairness and rationality index through fuzzy reasoning, and can conduct targeted optimization training on the unreasonable parts of the task allocation, gradually realizing a more scientific and efficient task allocation in the employment environment.

[0052] The scheduling platform in the present invention collaborates with multiple terminals such as the cloud, mobile terminals, and PC terminals, so that enterprise managers and scheduling personnel can better view and receive the arrangement forms of scheduling operations, greatly improving the flexibility and convenience of management. The task scheduling table and execution status can be synchronized to the worker terminal in real time. Workers can receive task instructions through the mobile terminal, thereby improving the transparency of employment and employee experience. The platform supports efficient collaboration between multiple roles such as enterprise managers and employees, and promotes the smooth connection of task allocation and execution.

[0053] The present invention generates first, second, third, and fourth type feedback signals based on the fairness and rationality evaluation results of the scheduling cycle, and dynamically adjusts the optimization direction. For example, for feedback signals with insufficient fairness, the system focuses on optimizing the balance of task allocation. Through historical data and periodic feedback, the system can optimize the scheduling rule model when necessary after each scheduling cycle, so that it can gradually adapt to the complex and changeable employment environment and achieve self-optimization.

[0054] The flexible employment management model in the present invention can dynamically adjust the employment scale according to actual needs, avoid the redundant costs of fixed labor relations, reduce the waste of human resources, and through scientific and reasonable task allocation, workers can complete their work more efficiently and reduce the additional expenses caused by task delays or failures. The scheduling platform helps enterprises optimize their employment strategies, further reduce management costs and improve efficiency, and greatly shortens the response time of task allocation through intelligent scheduling, enabling enterprises to respond to market changes and sudden demands more quickly. The intelligent, data-driven scheduling method makes enterprise employment management more refined and improves overall operational efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] In order to facilitate understanding by those skilled in the art, the present invention is further described below in conjunction with the accompanying drawings;

[0056] Figure 1 This is a schematic diagram of the multi-terminal collaborative intelligent scheduling method for flexible employment management in the present invention.

[0057] Figure 2 This is a schematic diagram of the multi-terminal collaborative intelligent scheduling platform for flexible employment management in the present invention. DETAILED DESCRIPTION

[0058] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. 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 creative work are within the scope of protection of the present invention.

[0059] Reference Figure 1 - Figure 2 The following embodiments are obtained:

[0060] Example 1: The flexible employment management in the present invention refers to a non-traditional labor management method, in which enterprises flexibly deploy human resources and assign tasks according to actual job requirements and business changes. It has the characteristics of flexible employment methods, diverse management methods, and efficient cost control. Job demand orientation includes: Dynamic job demand: The enterprise determines job demand based on business changes (such as project launch, emergency tasks, etc.) and does not fix labor relations. Task-oriented employment: Personnel are arranged in units of tasks, and positions are closely linked to task requirements. Tasks are assigned to the most suitable personnel through preset rules or intelligent algorithms.

[0061] Multi-terminal collaboration refers to the integration of cloud, mobile and PC collaboration to achieve intelligent scheduling and efficient management of flexible employment. For example, the multi-terminal collaborative intelligent scheduling system for flexible employment management of the present invention is deployed in the cloud. The mobile terminal is a port for collecting personal information of workers, and is used for sending work scheduling forms on the cloud to target workers. The PC terminal is a port for work task requirements released by the enterprise, and is used to display work scheduling forms generated in the cloud to the enterprise. The main functions of the system deployed in the cloud include matching job requirements with human resources, task allocation, scheduling feedback in each scheduling cycle, and deciding whether to optimize and train the preset scheduling rule model.

[0062] More specifically, the present invention proposes an intelligent scheduling method for multi-terminal collaboration for flexible employment management, comprising the following steps:

[0063] A scheduling rule model is preset, and the preset scheduling rule model is used to schedule work according to flexible employment needs, and obtain each work scheduling arrangement form; this is the basic setting of the system, which means that the scheduling process is based on pre-defined rules. These rules take into account the flexibility of employment needs (such as skill requirements, etc.), and generate a work schedule on this basis. Scheduling rules may include working time restrictions, etc., in order to achieve a preliminary work arrangement. Ensure that the system can generate a scheduling plan that meets actual needs based on established rules. The scheduling rule model can be a model trained based on a convolutional neural network (CNN) and deployed in the cloud.

[0064] In each scheduling cycle, based on all the work scheduling forms in the scheduling cycle, feature extraction is performed to extract the relevant features that affect the labor scheduling and obtain the influencing feature set; this step analyzes all the work scheduling forms in each scheduling cycle to identify those factors (i.e., features) that have a significant impact on the work scheduling results. These features may include the efficiency of work scheduling, the workload of employees, the completion of tasks, the flexibility of scheduling, etc. By extracting and summarizing these key features, the system can provide the necessary data support for subsequent evaluation and optimization.

[0065] According to the impact feature set, the fairness and rationality of each work scheduling result within the scheduling cycle are evaluated and an evaluation report is generated; in this step, the system analyzes each work scheduling arrangement based on the impact features extracted in the previous step to evaluate whether they are fair and reasonable. The fairness evaluation may consider whether each employee's workload, working time allocation, etc. are treated reasonably; while the rationality evaluation may include whether resources are effectively utilized during the execution of the task. Ensure that the work scheduling arrangement achieves the expected goals in terms of fairness and rationality, and identify potential problems or unreasonableness in the scheduling.

[0066] Different types of feedback signals are generated based on the evaluation report, and the type of feedback signal determines whether to optimize the preset scheduling rule model. Different types of feedback signals are generated in the evaluation report based on different fairness and rationality results. For example, the feedback signal may be of type 2: "Increase fairness", type 3: "Increase rationality", type 4: "Increase rationality and fairness" or type 1: "Fairness and rationality meet the requirements"; based on these feedback signals, the system decides whether the scheduling rules need to be optimized. Ensure timely feedback on the scheduling algorithm based on the actual evaluation results to avoid continued scheduling according to inappropriate rules, thereby promoting system optimization.

[0067] During the optimization training, an optimization training plan is generated according to the type of feedback signal. The preset scheduling rule model is optimized through iterative training to gradually adapt it to the scheduling needs under the employment environment, and the optimized scheduling rule model is applied to the next cycle of employment scheduling to ensure that the work distribution is more fair and reasonable. This stage is the self-improvement process of the entire system. According to the feedback signal obtained in the previous step, the system will generate an optimization training plan and train the scheduling rule model through multiple iterations to make it better adapt to the actual employment needs. Through continuous training and optimization, the scheduling rules are made more accurate and flexible, and can play a better role in the next cycle of employment scheduling. The ultimate goal is to ensure that the work distribution in the scheduling process is fair and reasonable, that is, on the basis of meeting the employment needs, the workload of each employee is balanced as much as possible to avoid overwork or unfair work distribution. At the same time, the efficiency and resource utilization of the scheduling should also be maximized. Improve the sustainability of the system and employee satisfaction, and improve the overall work efficiency and resource utilization through fair and reasonable scheduling arrangements.

[0068] Extracting relevant features that affect labor scheduling refers to extracting fairness feature groups and rationality feature groups respectively, and the fairness feature groups and rationality feature groups together constitute the influencing feature set. The evaluation report includes the scheduling cycle fairness index obtained by the fairness evaluation and the scheduling cycle rationality index obtained by the rationality evaluation.

[0069] The extraction of fairness feature groups is to ensure that the job opportunities, fair treatment and task allocation obtained by each employee or worker in the process of labor scheduling are fair, avoiding favoritism, discrimination or unequal treatment. Usually, fairness assessment will focus on the following aspects: Fairness of workload: whether the workload of employees is evenly distributed to avoid some employees taking on too much or too little work. Fairness of scheduling time: whether the arrangements of workers' shifts, working hours, etc. are balanced. Working time distribution: whether there are unfair phenomena such as long-term work and excessive overtime.

[0070] The rationality feature group focuses on evaluating the actual feasibility and efficiency of the scheduling plan to ensure that the labor scheduling can run efficiently and economically while meeting the needs. The aspects that the rationality feature group focuses on usually include: Reasonable allocation of work tasks: whether the task allocation meets the requirements of workers' skill level, experience, etc., and avoids overly complex or easy task arrangements. Matching of production / service needs: whether the scheduling can meet the actual requirements such as production plans and customer needs, and avoid excessive or insufficient labor input. Optimization of resource utilization: whether existing resources are effectively utilized to avoid waste or shortages.

[0071] Define a fairness index F, which reflects the fairness of work allocation when scheduling work among multiple workers. Generally, the higher the fairness, the closer the index is to 1; when the fairness is poor, the index is lower. The logic for obtaining the fairness index of this scheduling cycle is:

[0072] For each worker i, its workload and working hours They are defined as:

[0073] ; Indicates the preset difficulty coefficient corresponding to the workload in time period t;

[0074] ; represents the working time in time period t, Indicates whether worker i works in time period t. If he works in time period t, then The value is 1. If it does not work during time period t, then The value is 0, and T represents the total number of time periods;

[0075] Get the overall workload and work duration balance target:

[0076] First, calculate the overall workload and working time respectively. Suppose the total workload and total working time are and : ; ; n represents the total number of workers;

[0077] Then calculate the equilibrium target of each worker, namely the target workload and target working time:

[0078] Target workloads: ; represents the target workload of each worker;

[0079] Target working hours: ; represents the target working hours for each worker;

[0080] The workload deviation and working time deviation of each worker are:

[0081] ; represents the workload deviation of worker i;

[0082] ; represents the working time deviation of worker i;

[0083] Introducing worker preference differences for adjustment, assuming that worker i prefers the morning shift and resists the night shift, then his preference adjustment factor can be obtained by calculating the difference between the actual working hours and the preference:

[0084] ; represents worker i’s preference or resistance to feedback in time period t. If worker i prefers to work in time period t, then The value is 1, if you resist working in time period t, then The value is 0. represents the preference difference value of worker i, represents the preset preference ratio coefficient of worker i in time period t;

[0085] If some workers do not get enough rest time, their fairness index needs to be penalized and adjusted. If the number of scheduling cycles is too small, the adjustment factor increases. Therefore, the scheduling cycle fairness index calculation formula is:

[0086] ; , are preset non-zero trade-off coefficients that adjust the impact of rest time and worker preferences on the fairness index. represents the total rest time of worker i, Indicates the minimum permissible rest time, It is used to measure the adequacy of workers' rest time. F represents the fairness index of the scheduling cycle. The closer the value of F is to 1, the fairer the scheduling is, and the closer it is to 0, the unfair the scheduling is.

[0087] The logic for obtaining the scheduling cycle rationality index is:

[0088] The timeliness of task delivery is used to measure whether the task can be completed within the scheduled time. The calculation formula is as follows:

[0089] ; N represents the total number of work tasks, represents the deadline of the kth task, represents the actual completion time of the kth task, and TI represents the timeliness of task delivery; the timeliness of task delivery measures the closeness between the actual completion time of each task and the deadline. The closer the value is to 1, the higher the proportion of tasks completed on time and the more timely they are.

[0090] The worker load balance is used to evaluate the reasonable distribution of each worker's working time and load. The calculation formula is as follows:

[0091] ; represents the actual working time of the ith worker, It represents the total time required for the task assigned to the i-th worker, and LB represents the balance of worker load; it measures whether the load distribution among workers is balanced. The closer the value is to 1, the smaller the difference between the worker's actual working time and the estimated time of the assigned task, and the more balanced the load.

[0092] The reasonable matching degree of personnel scheduling is evaluated by the matching adaptability between task requirements and worker scheduling. The calculation formula is as follows:

[0093] ; Indicates the similarity between the skills of the i-th worker and the skills required for each scheduling task, such as cosine similarity, represents the experience value of the i-th worker, , They are all preset non-zero measurement coefficients. PP represents matching adaptability, which measures the degree of matching between workers' abilities and task requirements. The higher the value, the more the workers' abilities can meet the task requirements.

[0094] Get the distances from task delivery timeliness TI, worker load balance LB, and matching adaptability PP to the ideal solution and negative ideal solution. Let the distance to the ideal solution be JL1 and the distance to the negative ideal solution be JL2. The calculation formula for the scheduling cycle rationality index is:

[0095] ; Y represents the rationality index of the scheduling cycle. The rationality index of the scheduling cycle measures the degree of proximity to the ideal goal. The closer the value is to 1, the more reasonable and fair the task allocation in the scheduling cycle is. The larger the rationality index Y of the scheduling cycle is, the closer the current scheduling rules are to the ideal goal, and the more fair and reasonable the task allocation is. When the index is close to 1, it means that the load balance of workers, the timeliness of task delivery, and the degree of ability matching have reached a high level. When the rationality index is large, the scheduling can fully meet the actual needs of the employment environment and improve the overall management efficiency and resource utilization.

[0096] Ideal solution: refers to the comprehensive state when all evaluation indicators (TI, LB, PP) reach the optimal value. The ideal solution represents the performance of the system under the most ideal conditions. Calculation method: Assume that the ideal value of each indicator is: Ideal value of task delivery timeliness (TI): maximum value (usually 1, indicating that all tasks are completed on time). Ideal value of worker load balance (LB): maximum value (usually 1, indicating that all workers are completely balanced). Ideal value of matching suitability (PP): maximum value (usually 1, indicating that tasks and workers are completely matched). According to the ideal value of the indicator, the Euclidean distance between the current state (TI, LB, PP) and the ideal solution is calculated as JL1.

[0097] Negative ideal solution: refers to the comprehensive state when all evaluation indicators (TI, LB, PP) reach the worst value. The negative ideal solution represents the performance of the system under the most undesirable conditions. Calculation method: Assume that the negative ideal value of each indicator is: Task delivery timeliness (TI) negative ideal value: minimum value (usually 0, indicating that all tasks are not completed on time). Worker load balance (LB) negative ideal value: minimum value (usually 0, indicating that the load of all workers is extremely unbalanced). Matching suitability (PP) negative ideal value: minimum value (usually 0, indicating that the task and the worker are completely mismatched). According to the negative ideal value of the indicator, the Euclidean distance between the current state (TI, LB, PP) and the negative ideal solution is calculated as JL2.

[0098] The experience value of the i-th worker , is a comprehensive evaluation indicator to measure its performance and ability in historical scheduling tasks. The acquisition logic is:

[0099] ; V represents the total number of historical scheduling of the i-th worker, which refers to the number of schedulings that the worker has participated in. It is used to calculate the overall experience value of the worker and is part of the denominator, used to find the average value. v represents the index of the number of historical schedulings, ranging from 1 to V, and is used to traverse all historical scheduling tasks in which the worker has participated. Indicates the preset adjustment factor, control The strength of the influence determines the sensitivity of the score gap to the experience value calculation. The larger the value, the stronger the response of the experience value to the score deviation. represents the quality score of the vth dispatch of the ith worker, obtained by scoring the historical task completion performance. represents the historical average quality score of the ith worker. It is an S-shaped function. When the value approaches 1, it means that the worker's performance in this dispatch is better than the average level. When the value approaches 0.5, it means that the worker's performance is at the average level. When the value approaches 0, it means that the worker's performance in this dispatch is below the average level. It will be dynamically adjusted as the number of dispatches increases and workers' performance changes, and can reflect the changing trend of workers' abilities.

[0100] Generating different types of feedback signals based on the evaluation report refers to:

[0101] Using fuzzy reasoning, the scheduling cycle fairness index and the scheduling cycle rationality index are used as input data for fuzzy reasoning. The output result of fuzzy reasoning is the type of feedback signal. The types of feedback signals include class one, class two, class three, and class four. Only when the type of feedback signal is class one, the preset scheduling rule model is not optimized and trained.

[0102] In the scheduling cycle, the input variables include the fairness index and rationality index of the scheduling cycle. The first step of fuzzy reasoning is to convert these exact values ​​into fuzzy values ​​in fuzzy sets. Fuzzy sets of fairness index: Low: indicates poor fairness and significant imbalance in task allocation. Medium: indicates that fairness is acceptable, but there is room for improvement. High: indicates that task allocation is very fair and meets expectations.

[0103] Fuzzy set of rationality index: Low: indicates that the task allocation lacks rationality and the overall scheduling effect is poor. Medium: indicates that the rationality is insufficient but acceptable. High: indicates that the task allocation is relatively rational and meets the requirements.

[0104] The specific values ​​of the fairness and rationality index are attributed to the fuzzy set through the preset membership function. For example, a trapezoidal or triangular membership function can be used to map the input value to the fuzzy membership of "low", "medium" or "high".

[0105] Fuzzy rules are rules based on experience and logic, describing the relationship between input variables and output results. The rules are related to the combination of fairness and rationality indexes, mainly including the following:

[0106] If fairness is high and rationality is high, the feedback signal is "Category 1: Both fairness and rationality meet the requirements". If fairness is medium and rationality is high, the feedback signal is "Category 2: Increase fairness". If fairness is high and rationality is medium, the feedback signal is "Category 3: Increase rationality". If fairness is low or rationality is low, the feedback signal is "Category 4: Increase rationality and fairness".

[0107] The fuzzy reasoning engine derives the fuzzy output based on the fuzzy rules and the fuzzy membership of the input variables. Common reasoning methods include the maximum membership method and the weighted average method. For all fuzzy combinations of input variables, the output of the corresponding rule is determined. Combined with the membership of the input variables, the contribution of each rule to the output is calculated. For example, if fairness is medium (membership is 0.6) and rationality is high (membership is 0.8), the contribution corresponding to the rule "medium fairness and high rationality" is 0.6.

[0108] Fuzzy reasoning generates fuzzy output, which needs to be converted into a specific feedback signal type. Common defuzzification methods include the centroid method and the maximum membership method. Centroid method: Calculate the centroid of the output fuzzy set and use it as the specific output value. For example, the fuzzy set of signal type may be mapped to the interval from 0 to 1, corresponding to different feedback signal types. Maximum membership method: Select the result corresponding to the output fuzzy set with the largest membership as the final output. In the present invention, the defuzzification result directly corresponds to the type of feedback signal, including: Category 1: Both fairness and rationality meet the requirements and no optimization is required. Category 2: Optimization is required to increase fairness. Category 3: Optimization is required to increase rationality. Category 4: Simultaneous optimization is required to increase fairness and rationality.

[0109] According to the defuzzification results, a specific feedback signal type is generated, and it is used to decide whether to optimize the dispatch rule model: when the feedback signal is of type one, the dispatch rule does not need to be optimized, indicating that the existing rules have met the requirements of fairness and rationality. When the feedback signal is of type two, three, or four, the system determines the optimization direction (improving fairness, rationality, or both) based on the feedback type and triggers the optimization training process.

[0110] During optimization training, generating an optimization training plan based on the type of feedback signal refers to:

[0111] The number of training data of a type of feedback signal is selected as Z, and the training data of the type of feedback signal corresponding to the training number in the range of [2Z, 4Z] are selected to together constitute a training set for optimization training, and in the training set for optimization training, the number of all training data is not less than a preset minimum training quantity threshold.

[0112] A specific example is as follows: During each optimization training, a type of feedback signal (fairness and rationality meet the requirements) with a historical training data quantity of Z is preferentially selected as the basic data. At the same time, according to the type of feedback signal in the current scheduling cycle (such as type 2, type 3, type 4), the corresponding feedback signal training data is selected, the quantity is 2Z, to highlight the pertinence and focus on optimizing the deficiencies in the current scheduling cycle. In this way, by balancing the historical high-quality scheduling data (as a basic reference) and the current deficiencies data (as the focus of optimization), the scheduling rule model can be more effectively improved in its ability to adapt to the deficiencies.

[0113] Select the number of data Z of a type of feedback signal: A type of feedback signal data represents an ideal case in which fairness and rationality meet the requirements in the scheduling cycle. It is used as the basic data for model learning to maintain the overall balance of the scheduling algorithm. The number of data is fixed at Z and is extracted from the historical scheduling cycle.

[0114] Selection rules for the current feedback signal type: If the feedback signal type is type II (increasing fairness), select type II feedback signal data related to "increasing fairness". If the feedback signal type is type III (increasing rationality), select type III feedback signal data related to "increasing rationality". If the feedback signal type is type IV (increasing fairness and rationality), select type IV feedback signal data related to both aspects.

[0115] The amount of data corresponding to the feedback signal of the current cycle is 2Z, which is twice the data of the first-class feedback signal. It is used to strengthen the model's learning ability for the current deficiencies. The data Z of the first-class feedback signal and the data 2Z of the current feedback signal are combined to form an optimized training set. In the training set: the data of the first-class feedback signal accounts for one-third (maintaining the basic reference), and the data of the current feedback signal accounts for two-thirds (optimization focus). In the optimization training, the data of the first-class feedback signal is learned first to maintain the overall fairness and rationality of the scheduling algorithm, and to strengthen the data learning of the current feedback signal type. The current feedback signal data can be used when the problem is serious. The proportion is increased when the problem is serious. The scheduling quality is high.

[0116] Embodiment 2: A multi-terminal collaborative intelligent scheduling platform for flexible employment management, deployed in the cloud, includes:

[0117] The work scheduling module has a preset scheduling rule model, and the preset scheduling rule model is used to schedule work according to the flexible employment needs and generate a work scheduling form for each time;

[0118] The feature extraction module performs feature extraction in each scheduling cycle based on all work scheduling forms in the scheduling cycle, extracts relevant features that affect the labor scheduling, and obtains an influencing feature set;

[0119] The evaluation module evaluates the fairness and rationality of each work scheduling result within the scheduling cycle based on the impact feature set and generates an evaluation report;

[0120] A feedback module generates different types of feedback signals according to the evaluation report, and determines whether to optimize the preset scheduling rule model according to the type of feedback signal;

[0121] The optimization module generates an optimization training plan according to the type of feedback signal during optimization training, optimizes the preset scheduling rule model through iterative training, so that it gradually adapts to the scheduling needs under the employment environment, and transmits the optimized scheduling rule model to the work scheduling module. The optimized scheduling rule model is applied to the employment scheduling of the next cycle to ensure that the work distribution is more fair and reasonable.

[0122] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0123] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0124] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0125] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0126] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A multi-terminal collaborative intelligent scheduling method for flexible employment management, characterized in that: The following steps are involved: A scheduling rule model is preset, and the preset scheduling rule model is used to schedule work according to flexible employment needs, and obtain a work scheduling arrangement form for each time; In each scheduling cycle, based on all the work scheduling forms in the scheduling cycle, feature extraction is performed to extract relevant features that affect the labor scheduling and obtain an influencing feature set; According to the impact feature set, the fairness and rationality of each work scheduling result within the scheduling cycle are evaluated and an evaluation report is generated; Generate different types of feedback signals according to the evaluation report, and decide whether to optimize the preset scheduling rule model according to the type of feedback signal; During optimization training, an optimization training plan is generated according to the type of feedback signal. The preset scheduling rule model is optimized through iterative training to gradually adapt it to the scheduling needs in the employment environment. The optimized scheduling rule model is applied to the next cycle of employment scheduling to ensure that work distribution is more fair and reasonable.

2. The multi-terminal collaborative intelligent scheduling method for flexible employment management according to claim 1 is characterized in that: Extracting relevant features that affect labor scheduling refers to extracting a fairness feature group and a rationality feature group respectively. The fairness feature group and the rationality feature group together constitute an influencing feature set.

3. The multi-terminal collaborative intelligent scheduling method for flexible employment management according to claim 2 is characterized in that: The evaluation report includes a scheduling cycle fairness index obtained by a fairness evaluation and a scheduling cycle rationality index obtained by a rationality evaluation.

4. The multi-terminal collaborative intelligent scheduling method for flexible employment management according to claim 3 is characterized in that: The logic for obtaining the scheduling cycle fairness index is: For each worker i, its workload and working hours They are defined as: ; Indicates the preset difficulty coefficient corresponding to the workload in time period t; ; represents the working time in time period t, Indicates whether worker i works in time period t. If he works in time period t, then The value is 1. If it does not work during time period t, then The value is 0, and T represents the total number of time periods; Get the overall workload and work duration balance target: First, calculate the overall workload and working time respectively. Suppose the total workload and total working time are and : ; ; n represents the total number of workers; Then calculate the equilibrium target of each worker, namely the target workload and target working time: Target workloads: ; represents the target workload of each worker; Target working hours: ; represents the target working hours for each worker; The workload deviation and working time deviation of each worker are: ; represents the workload deviation of worker i; ; represents the working time deviation of worker i; Introducing differences in worker preferences for adjustment: ; represents worker i’s preference or resistance to feedback in time period t. If worker i prefers to work in time period t, then The value is 1, if you resist working in time period t, then The value is 0. represents the preference difference value of worker i, represents the preset preference ratio coefficient of worker i in time period t; The calculation formula of scheduling cycle fairness index is: ; , are all preset non-zero trade-off coefficients, represents the total rest time of worker i, represents the minimum allowed rest time, F represents the scheduling cycle fairness index, and C is a preset non-zero constant.

5. The multi-terminal collaborative intelligent scheduling method for flexible employment management according to claim 4 is characterized in that: The logic for obtaining the scheduling cycle rationality index is: The timeliness of task delivery is used to measure whether the task can be completed within the scheduled time. The calculation formula is as follows: ; N represents the total number of work tasks, represents the deadline of the kth task, represents the actual completion time of the kth task, and TI represents the timeliness of task delivery; The worker load balance is used to evaluate the reasonable distribution of each worker's working time and load. The calculation formula is as follows: ; represents the actual working time of the ith worker, represents the total time required for the task assigned to the i-th worker, LB represents the balance of worker load, and C is a preset non-zero constant; The reasonable matching degree of personnel scheduling is evaluated by the matching adaptability between task requirements and worker scheduling. The calculation formula is as follows: ; represents the similarity between the skills of the ith worker and the skills required for each scheduling task, represents the experience value of the i-th worker, , They are all preset non-zero measurement coefficients, and PP represents matching suitability; Get the distances from task delivery timeliness TI, worker load balance LB, and matching adaptability PP to the ideal solution and negative ideal solution. Let the distance to the ideal solution be JL1 and the distance to the negative ideal solution be JL2. The calculation formula for the scheduling cycle rationality index is: ; Y represents the scheduling cycle rationality index.

6. The multi-terminal collaborative intelligent scheduling method for flexible employment management according to claim 5 is characterized in that: The experience value of the i-th worker The acquisition logic is: ; V represents the total number of historical scheduling times of the i-th worker, v represents the index of the historical scheduling times, Indicates the preset adjustment factor, represents the quality score of the vth scheduling of the ith worker, represents the historical average quality score of the ith worker.

7. The multi-terminal collaborative intelligent scheduling method for flexible employment management according to claim 6, characterized in that: Generating different types of feedback signals based on the evaluation report refers to: Using fuzzy reasoning, the scheduling cycle fairness index and the scheduling cycle rationality index are used as input data for fuzzy reasoning. The output result of fuzzy reasoning is the type of feedback signal. The types of feedback signals include class one, class two, class three, and class four. Only when the type of feedback signal is class one, the preset scheduling rule model is not optimized and trained.

8. The multi-terminal collaborative intelligent scheduling method for flexible employment management according to claim 7 is characterized in that: During optimization training, generating an optimization training plan based on the type of feedback signal refers to: The number of training data of a type of feedback signal is selected as Z, and the training data of the type of feedback signal corresponding to the training number in the range of [2Z, 4Z] are selected to together constitute a training set for optimization training, and in the training set for optimization training, the number of all training data is not less than a preset minimum training quantity threshold.

9. An intelligent scheduling platform for multi-terminal collaboration for flexible employment management, used to implement the intelligent scheduling method for multi-terminal collaboration for flexible employment management as described in any one of claims 1 to 8, characterized in that: include: The work scheduling module has a preset scheduling rule model, and the preset scheduling rule model is used to schedule work according to the flexible employment needs and generate a work scheduling form for each time; The feature extraction module performs feature extraction in each scheduling cycle based on all work scheduling forms in the scheduling cycle, extracts relevant features that affect the labor scheduling, and obtains an influencing feature set; The evaluation module evaluates the fairness and rationality of each work scheduling result within the scheduling cycle based on the impact feature set and generates an evaluation report; A feedback module generates different types of feedback signals according to the evaluation report, and determines whether to optimize the preset scheduling rule model according to the type of feedback signal; The optimization module generates an optimization training plan according to the type of feedback signal during optimization training, optimizes the preset scheduling rule model through iterative training, so that it gradually adapts to the scheduling needs under the employment environment, and transmits the optimized scheduling rule model to the work scheduling module. The optimized scheduling rule model is applied to the employment scheduling of the next cycle to ensure that the work distribution is more fair and reasonable.

Images