Healthcare staff scheduling system based on healthcare staff skills

Abstract

The application relates to the technical field of human resource management, in particular to a medical staff scheduling system based on medical staff skills, which comprises a standard skill input and classification device, which is used for inputting the working skills of each department, dividing the same working skills of each department into general skills, and dividing the special working skills of each department into special skills; a department patient input device, which obtains the treatment information of patients in each department ward, extracts the predicted workload of each patient in the next scheduling cycle, and counts the predicted workload of each department; the predicted workload comprises the execution times of each general skill and the execution times of each special skill; the system realizes cross-department dynamic scheduling by constructing a hospital-level skill resource pool and a demand prediction mechanism, standardizes the general skills into shared resources, and enables medical staff to flexibly support departments with high demand peaks.

Description

Technical Field

[0001] This application relates to the field of human resource management technology, and more specifically, to a medical staff scheduling system based on the skills of medical staff. Background Technology

[0002] The content in this section provides only background information related to this application and may not constitute prior art.

[0003] Currently, medical institutions generally face an uneven distribution of medical staff resources. Frontline clinical departments, especially high-pressure departments such as emergency, general surgery, urology, radiotherapy and chemotherapy, hematology, and gastroenterology, chronically suffer from staff shortages. This structural imbalance leads to medical staff in high-burden departments being under constant overwork, exacerbating the risk of burnout and directly extending patient waiting times due to limited capacity, thus reducing the accessibility and responsiveness of medical services.

[0004] The existing scheduling model relies heavily on departmental autonomy, creating strict "skill silos." Because different departments have varying levels of professional qualifications (such as aseptic operating room procedures, specialized surgical nursing, and dialysis equipment operation) and domain knowledge (such as specialist medication guidelines and emergency procedures), medical staff are strictly confined to their own departmental scheduling system. This departmental-boundary-based scheduling mechanism essentially blocks the possibility of cross-departmental human resource collaboration; even if a neighboring department experiences a temporary staff shortage, it is impossible to utilize available personnel from other departments to fill the gap.

[0005] The rigid framework of departmental self-management scheduling amplifies the negative impact of resource misallocation. The workload of key departments often fluctuates seasonally (e.g., a surge in respiratory patients in winter, and a high incidence of trauma and gastrointestinal cases in summer), but staffing adjustments are constrained by skill silos, preventing dynamic matching. Consequently, overloaded departments are forced to work overtime during peak hours, leading to increased medical error rates and staff turnover; while idle staff in departments during low-demand periods result in resource waste. Furthermore, the hospital's internal staffing freeze mechanism makes it difficult to alleviate pressure through temporary staffing increases. Thus, a vicious cycle of "fixed busy and idle periods" easily forms within the hospital, systematically reducing the overall efficiency of medical resource utilization.

[0006] Meanwhile, existing hospital scheduling systems generally arrange staff numbers according to the workload requirements of each department, which is a linear scheduling system. When dealing with cross-departmental staff scheduling, the data processing model of the current scheduling system cannot consider staff arrangements from a global perspective, which can easily lead to the staffing of medical staff getting stuck in a local optimal solution. Summary of the Invention

[0007] The summary section of this application is intended to provide a brief overview of the concepts, which will be described in detail in the detailed description section below. This summary section is not intended to identify key or essential features of the claimed technical solutions, nor is it intended to limit the scope of the claimed technical solutions.

[0008] Some embodiments of this application propose a healthcare staff scheduling system based on healthcare staff skills to address the technical problems mentioned in the background section above.

[0009] As a first aspect of this application, some embodiments of this application provide a healthcare worker scheduling system based on healthcare worker skills, including:

[0010] The standard skills input and classification device is used to input the work skills of each department, classify the same work skills in each department into general skills, and classify the special work skills of each department into specialized skills.

[0011] General skills are the job skills that all medical staff in all departments need to master;

[0012] Specialized technical skills are the work skills required for special operations in specific departments;

[0013] The medical skills input device records and updates the general skills and specialized skills mastered by each medical staff member in real time, and simultaneously updates the number of general skills and specialized skills that each medical staff member can perform during the shift cycle.

[0014] The departmental patient entry device acquires treatment information of patients in each ward, extracts the estimated workload for each patient in the next scheduling cycle, and calculates the estimated workload for each department. The estimated workload includes the number of times each general skill is executed and the number of times each specialized skill is executed.

[0015] The staff scheduling device uses the constraint that the total workload of the medical staff assigned to each department during the scheduling cycle is greater than the expected workload of the corresponding department, and aims to minimize the total number of medical staff assigned, to allocate matching medical staff to all departments.

[0016] This system breaks down departmental barriers and enables dynamic cross-departmental scheduling by constructing a hospital-wide skills resource pool and demand forecasting mechanism. It standardizes common skills into hospital-wide shared resources, allowing qualified medical staff to flexibly support departments with peak demand. Simultaneously, it accurately quantifies the skill requirements of each department for future scheduling cycles (such as the number of intravenous punctures, urinary catheterizations, and ventilator operation times) based on patient treatment data and automatically matches them with the medical staff skills database and available working hours. Without increasing staffing levels, it significantly improves human resource utilization efficiency, effectively solves structural resource mismatch and uneven workloads, and achieves systematic optimization of hospital-wide human resources. Furthermore, the personnel scheduling device in this solution is equipped with constraints and optimization objectives. When setting scheduling schemes, it does not simply use linear scheduling, but rather, under the constraints, strives to find a scheduling scheme that best achieves the optimization objective, thus avoiding the scheduling scheme from getting trapped in local optima.

[0017] Furthermore, the medical skills input device includes:

[0018] The medical staff information entry module assigns a unique ID number to each medical staff member in each department. The ID number includes the department code and the personnel code.

[0019] The skill information extraction module extracts the execution time spent by medical staff when performing various general skills, as well as the execution time spent by medical staff when performing various specialized skills.

[0020] The skills information recording module records the average execution time of each medical staff member for performing various general skills and the average execution time for performing various specialized skills, and generates a personnel skills table.

[0021] The skills information update module records the execution time of each medical staff member for each general skill and the execution time for each specialized skill, and updates the average execution time of each general skill and the average execution time of each specialized skill.

[0022] This device continuously tracks and dynamically updates the actual time each medical staff member spends performing various general and specialized skills, building a precise personal skills time database. This real-time updated average execution time data provides an objective and quantifiable core basis for subsequent staff scheduling and task allocation, effectively improving the accuracy of person-job matching, ensuring a balanced workload, and thus optimizing human resource allocation.

[0023] Furthermore, the departmental patient entry device includes:

[0024] The patient entry module extracts the inpatient information of all patients in the hospital and generates a patient ID for each patient, which includes the department code and the patient code.

[0025] The department workload entry module retrieves the medical record information of all patients, extracts the various operations that each patient needs to perform in the next scheduling cycle from the medical record information, so as to obtain the estimated workload of each patient, and makes a table of the estimated workload of patients in the same department to obtain the estimated workload table of each department.

[0026] The workload update module monitors patient admission information in real time, obtains the number of patients entering each department in the next scheduling cycle based on the newly entered admission information, and updates the estimated workload table for each department.

[0027] This device integrates hospital-wide patient inpatient information and case data to accurately predict and dynamically update the skill operation requirements (i.e., estimated workload) for each department in the next scheduling cycle. It transforms patient treatment plans into quantifiable departmental-level skill execution lists (such as the number of specific operations) and updates them in real-time in response to newly admitted patient information. This provides an accurate and timely demand-side data foundation for subsequent skills-based human resource matching.

[0028] Furthermore, the workload update module includes:

[0029] The outpatient information monitor is used to monitor outpatient information that will be available for the next scheduling cycle in real time.

[0030] The emergency information monitor is used to predict which emergency cases will need to enter the department in the next scheduling cycle based on historical emergency data.

[0031] The information updater updates the projected workload table based on outpatient and emergency information.

[0032] This module enables rolling and precise updates to departmental workload for the next scheduling cycle by tracking outpatient admission dynamics in real time and predicting emergency room inflows based on historical data. This dual-channel (outpatient + emergency) real-time demand perception mechanism significantly improves the scheduling system's responsiveness to sudden patient inflows and regular demand fluctuations, ensuring that departmental projected workload tables always reflect the latest and most realistic future demands, providing solid and timely data support for dynamic human resource allocation.

[0033] Furthermore, the personnel scheduling device includes:

[0034] The personnel information acquisition module is connected to the skills information recording module by signal. It extracts the medical staff who can be on duty in the next shift cycle from the personnel skills table and generates a candidate personnel skills table.

[0035] The department demand acquisition module is connected to the department workload entry module to obtain the estimated workload table for each department in the next scheduling cycle;

[0036] The scheduling design module establishes a scheduling allocation model based on the expected workload of each department in the next scheduling cycle as a constraint. It extracts the minimum number of medical staff from the candidate personnel skill list as the fitness function of the scheduling allocation model. The scheduling allocation model is solved to generate the scheduling plan for the next scheduling cycle.

[0037] The task assignment module assigns corresponding work tasks to medical staff based on their work skills when they execute the shift schedule.

[0038] Furthermore, the shift allocation model is as follows:

[0039] Department code i, i∈{1, 2, 3, ... I}, i represents the department index, and I represents the total number of departments;

[0040] The job skill is encoded as k, k∈{1, 2, 3, ... K}, where k represents the job skill index and K represents the total number of job skills.

[0041] The medical staff code j, j∈{1, 2, 3, ... J}, j represents the medical staff index, and J represents the total number of medical staff;

[0042] The job skill code of medical staff member J , The label represents the skill level of the j-th medical worker. =0 indicates that medical staff j has not mastered the k-th job skill. =1 indicates that medical staff j has mastered the kth job skill;

[0043] Workload sequence D of department i i , ; This represents the quantity of job skill k required by department i. It is a non-negative integer;

[0044] Define the allocation status X of medical staff; , This indicates the allocation status of the j-th medical staff member;

[0045] ,when When = 0, it indicates that the j-th medical staff member has not been assigned a shift. When =i, it indicates that the j-th medical staff member will be assigned to the i-th department for duty in the next shift cycle;

[0046] Constraint 1: ;

[0047] Constraint 2: ; Indicates the length of the scheduling cycle. This represents the total average time taken for all medical staff in department i to perform work skill k during the current scheduling cycle;

[0048] The main objective function is: .

[0049] It is an indicator function; its value is 1 when the condition is true, and 0 otherwise.

[0050] Where J represents the total number of medical staff, and j represents the index of medical staff. The label represents the skill level of the j-th medical worker. This represents the number of job skills k required by department i, where k represents the job skill index, K represents the total number of job skills, and i represents the department index. This represents the number of medical staff assigned to department i who possess job skill k. This represents the total number of medical staff assigned to department i.

[0051] In this scheme, the main objective function minimizes personnel input by optimizing human resource allocation. The particle swarm optimization algorithm drives the scheduling scheme to continuously reduce unnecessary personnel arrangements, accurately control human resource costs to reduce operating expenses, significantly improve personnel utilization efficiency to avoid resource idleness, realize resource intensification to eliminate redundant scheduling, and continuously guide particles to converge toward the unassigned state through the algorithm's inherent guidance mechanism, ultimately achieving the goal of meeting the department's needs with the fewest medical staff.

[0052] Furthermore, the department codes i are arranged in the order of adjacent physical locations of the departments;

[0053] The scheduling model also includes a first sub-objective function L1:

[0054] ;

[0055] .

[0056] Indicates the parameter factor, Indicates an intermediate function. This represents the sum of human resources allocated to department i. This represents the sum of human resources allocated to department i+1. This represents the sum of human resources allocated to department i-1. This represents the sum of the workload that department i-1 needs to perform in the next scheduling cycle. This represents the sum of the workload that department i needs to perform in the next scheduling cycle. This represents the sum of the workload that department i+1 needs to perform in the next scheduling cycle.

[0057] The first objective function constructs a collaborative working mechanism between adjacent departments and dynamically adjusts the workload allocation in particle swarm optimization to ensure that at least one department maintains a significant workload surplus. This enables medical staff to be quickly transferred between physically adjacent departments during the actual scheduling cycle, greatly improving the efficiency of responding to sudden demands. At the same time, the algorithm stability is maintained through fixed parameters, ultimately forming a flexible human resource buffer pool to cope with the uncertainties and fluctuations in the medical scenario.

[0058] Furthermore, the scheduling model also includes a second sub-objective function L2:

[0059] ;

[0060] ;

[0061] ;

[0062] , These represent the first and second intermediate parameters, respectively. This indicates the number of work skills k that need to be performed in department i during the previous scheduling cycle. This represents the quantity of job skill k required by department i. This represents the sum of the number of all job skills k in department i. This represents the sum of the number of work skills k that need to be performed in department i during the previous scheduling cycle. This represents the sum of human resources allocated to department i. This represents the sum of the workload that department i needs to perform in the next scheduling cycle.

[0063] The second objective function establishes a prediction-driven adaptive adjustment mechanism in particle swarm optimization by dynamically calibrating the matching degree between prediction errors and actual resource allocation: when the demand prediction is too high, the workload buffer is automatically increased; when the prediction is too low, the resource input is actively reduced. The nonlinear coupling of historical data and current scheduling is used to continuously optimize the decision, so that the scheduling scheme has the ability to adapt to prediction fluctuations, and ultimately achieves an intelligent balance between the uncertainty of medical resource supply and demand.

[0064] Furthermore, the solution process for the shift allocation model includes the following steps:

[0065] Step 1: Randomly generate N particles that simultaneously satisfy both constraint 1 and constraint 2. Each particle corresponds to a medical staff assignment state X. Use these N particles as the initial population.

[0066] Step 2: Calculate the fitness function value F(X) for each particle;

[0067] F(X) = λ0L0 + λ1L1 + λ2L2;

[0068] L0 represents the primary objective function, L1 represents the first secondary objective function, L2 represents the second secondary objective function, and λ0, λ1, and λ2 represent the first weighting factor, the second weighting factor, and the third weighting factor, respectively.

[0069] Step 3: Sort all particles in descending order according to their fitness function values ​​to obtain a particle count.

[0070] Step 4: Set the fitness function value dividing line F0, and use particles with fitness function values ​​less than F0 as the diffusion population, and use particles with fitness function values ​​not less than F0 as the update population;

[0071] Step 5: For all particles in the updated population, randomly diffuse with the center point of all particles in the diffusion population as the target. For all particles in the diffusion population, randomly diffuse with the global optimal position as the direction.

[0072] Step 6: Repeat steps 3 to 5 until the termination condition is met. The termination condition is reaching the maximum number of iterations or the gradient of the fitness function value of the global optimal solution is less than a preset threshold.

[0073] This application achieves a dynamic balance between global optimization and local development through an innovative dual-population co-evolution mechanism: First, the population intelligence is divided into an update population and a diffusion population using a fitness dividing line, allowing high-fitness particles to diffuse towards the center of the low-fitness population to break through the local optimum trap, while guiding low-fitness particles to converge towards the global optimum to accelerate the search for optimization; This bidirectional diffusion strategy continuously stimulates the synergistic effect of exploration and development during the iteration process, maintaining population diversity to avoid premature convergence, and achieving precise termination through the fitness gradient threshold, thereby efficiently generating the scheduling scheme with the lowest human cost, optimal departmental collaboration, and the strongest predictive adaptability under complex medical constraints.

[0074] Furthermore, ;

[0075] Where t represents the index of the iteration number. Indicates the maximum number of iterations. Let represent the minimum fitness function value among all particles at the t-th iteration. This represents the maximum fitness function value among all particles at the t-th iteration.

[0076] The dynamic fitness split line F0 achieves an intelligent balance between the algorithm's exploration and development capabilities through an iterative adaptive threshold adjustment mechanism. As the number of iterations increases, F0 gradually increases from its initial value close to the minimum fitness value towards the maximum fitness value, causing the size of the updating population (high-fitness particles) to continuously decrease while the size of the diffusion population (low-fitness particles) increases accordingly. This gradual adjustment strengthens the central diffusion of high-fitness particles in the early stages of the algorithm to maintain population diversity and avoid premature convergence, while in the later stages, it increases the convergence strength of low-fitness particles towards the global optimum to accelerate the optimization efficiency. Ultimately, it forms an intelligent optimization rhythm that automatically adjusts with the iteration process, effectively solving the exploration-development imbalance problem caused by the fixed threshold in the traditional particle swarm optimization algorithm. Attached Figure Description

[0077] Figure 1 This is a schematic diagram of a healthcare worker scheduling system based on healthcare worker skills. Detailed Implementation

[0078] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below in conjunction with specific embodiments. The same reference numerals in the accompanying drawings represent the same components. It should be noted that the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the described embodiments of this application without creative effort are within the scope of protection of this application.

[0079] Compared to the embodiments shown in the accompanying drawings, feasible embodiments within the scope of this application may have fewer components, other components not shown in the drawings, different components, differently arranged components, or components with different connections, etc. Furthermore, two or more components in the drawings may be implemented in a single component, or a single component shown in the drawings may be implemented as multiple separate components.

[0080] Unless otherwise defined, the technical or scientific terms used herein shall have the ordinary meaning as understood by one of ordinary skill in the art to which this application pertains. The terms “first,” “second,” and similar terms used in this specification and claims do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Similarly, the terms “an” or “a” and similar terms do not necessarily indicate a quantity limitation. Terms such as “upper” and “lower” are used only to indicate relative positional relationships, and these relative positional relationships may change accordingly when the absolute position of the described object changes.

[0081] refer to Figure 1Example 1: A medical staff scheduling system based on medical staff skills, including a standard skills input and classification device, a medical skills input device, a departmental patient input device, and a staff scheduling device. The standard skills input and classification device, the medical skills input device, and the departmental patient input device are respectively connected to the staff scheduling device.

[0082] The standard skills input and classification device is used to input the work skills of each department, classify the same work skills in each department into general skills, and classify the special work skills of each department into specialized skills; general skills are the work skills that all medical staff in all departments need to master; specialized skills are the work skills that need to be mastered for special operations in a specific department.

[0083] Specifically, general skills refer to the basic and common medical and nursing operational abilities that all medical and nursing staff in all departments must master. Their characteristic is cross-departmental applicability, forming the basis for the sharing and flexible allocation of human resources throughout the hospital. Specialized skills, on the other hand, refer to highly specialized operational abilities required only in specific departments or for specific diseases or equipment.

[0084] For example, procedures such as venipuncture, preparation of intravenous medications, and administration of intravenous drugs are considered general skills.

[0085] Enterostomy enema and enterostomy care are specialized skills of the general surgery department.

[0086] For example, the procedure for transostomy enema in general surgery is as follows: The patient is positioned comfortably, the stoma is exposed, the enema tube tip is lubricated, and it is gently inserted into the stoma about 10-15 cm. The enema bag is connected, and warm enema fluid (usually 500-1000 ml) is slowly injected. The flow rate is controlled, and the patient's response and stoma drainage are observed. After enema, the tube is clamped, and after a short wait, it is slowly withdrawn. The patient is assisted with toileting or the ostomy bag is used to collect excrement. The entire process requires precise control of the flow rate, pressure, and insertion depth to prevent backflow, perforation, or retention of the irrigation fluid; therefore, it is a specialized skill in general surgery.

[0087] Bladder irrigation and bladder instillation are specialized skills in urology.

[0088] For example, the procedure for bladder irrigation is as follows: Connect the irrigation port of a three-lumen catheter (or a double-lumen catheter plus a separate irrigation tube) to a sterile irrigation solution bag / bottle, and connect the drainage port to a urine bag. Clamp the drainage tube, open the irrigation tube, and inject the irrigation solution at an appropriate rate (e.g., 80-100 drops / min). Observe the patient's response and the characteristics of the outflow. After injecting an appropriate amount (usually 100-200ml), clamp the irrigation tube and open the drainage tube to empty the bladder. This can be repeated until the outflow is clear. Throughout the process, the irrigation solution level (pressure) and flow rate must be strictly controlled to prevent bladder overdistension, excessive pressure leading to bladder damage, urine reflux, or even pyelonephritis. Therefore, it is a specialized skill in urology.

[0089] The safe preparation and administration of chemotherapy drugs, as well as the management of chemotherapy drug extravasation, are specialized skills of the radiotherapy and chemotherapy department.

[0090] For example, although the preparation and administration of chemical drugs are the same as those of ordinary drugs—either orally or intravenously—the preparation must be carried out in a dedicated biosafety cabinet. The person preparing the drug must wear double gloves, protective clothing, and goggles / face shields, and strictly verify the medication. Simultaneously, it is necessary to prevent splashing and aerosol generation during the dissolution, extraction, and dilution of the drug. The entire process is very complex, thus requiring specialized skills from the radiotherapy and chemotherapy department.

[0091] Special blood collection techniques for patients with blood diseases and the management of component blood transfusions are specialized skills of the hematology department.

[0092] For example, the procedure for special blood collection techniques for patients with hematological diseases involves: using specialized transfusion sets (e.g., special filters are needed for platelets). Administering anti-allergy medication before transfusion. Initiating a slow infusion (closely observing for the first 15 minutes), adjusting to the required rate after no reaction (e.g., rapid infusion for platelets). Throughout the process, it is crucial to quickly and accurately identify various transfusion reactions (especially acute hemolytic reactions, allergic reactions, TRALI, TACO, etc.) and immediately initiate emergency treatment procedures. The timeliness of reaction management is extremely important. Therefore, this is a specialized skill for hematology. In addition, skills such as infusion port administration, infusion port maintenance, PICC line placement, arterial blood collection, the use of the GI-1 type low-frequency therapy device, and the use of enteral nutrition infusion pumps, which are required by nurses in certain departments, can also be designated as specialized skills. These skills, not mastered by all nurses, can also be considered specialized skills.

[0093] In essence, general skills are common skills shared by all departments, and these skills constitute the main tasks of each department. However, each department, due to different treatment methods, has specific operational tasks, which fall under the category of specialized skills. Therefore, when nurses from different departments are seconded, it's primarily because they cannot perform the specialized skills required by those departments.

[0094] The medical skills input device records and updates the general skills and specialized skills mastered by each medical staff member in real time, and simultaneously updates the number of general skills and specialized skills that each medical staff member can perform during the shift cycle.

[0095] The medical and nursing skills input device includes: a medical and nursing personnel information input module, a skills information extraction module, a skills information recording module, and a skills information update module. The medical and nursing personnel information input module, skills information extraction module, and skills information recording module are connected in sequence, and the skills information update module is connected to the skills information recording module.

[0096] The medical staff information entry module assigns a unique ID number to each medical staff member in each department. The ID number includes the department code and the personnel code.

[0097] For example, suppose the department code for the Department of Gastroenterology is GI, and the 15th medical staff member in this department is assigned a unique ID number in the system: GI-015. Thus, the department of a medical staff member can be accurately determined using this ID number.

[0098] The skill information extraction module extracts the execution time spent by medical staff when performing various general skills, as well as the execution time spent by medical staff when performing various specialized skills.

[0099] Example of skill execution time extraction (taking intravenous infusion as an example):

[0100] Healthcare professionals use a handheld device (such as a PDA) to scan the patient's wristband barcode (to verify patient identity) and the infusion bag / infusion device barcode (to confirm the recipient). The system uses this dual scanning action as the start signal for the intravenous infusion procedure and automatically records the precise timestamp. After the tubing is secured, healthcare professionals again use the terminal to scan the patient's wristband barcode and the same infusion device barcode (or scan a specific barcode representing "complete"). The system uses this scanning action as the end signal for the procedure and automatically records the precise timestamp as the end time of the intravenous infusion procedure.

[0101] The skill information extraction module automatically calculates the difference between the end timestamp and the start timestamp, and binds this actual execution time to the operation (intravenous infusion) and the performing medical staff. In this way, the time when the medical staff performed the intravenous operation can be obtained.

[0102] In practice, because it is necessary to identify the responsible personnel for each operational task to avoid administering the wrong medication, current hospital management systems require medical staff to upload node information at each process node. This node information facilitates the retrieval of time spent on skills.

[0103] The skills information recording module records the average execution time of each medical staff member for performing various general skills and the average execution time for performing various specialized skills, and generates a personnel skills table.

[0104] The skills information update module records the execution time of each medical staff member for each general skill and the execution time for each specialized skill, and updates the average execution time of each general skill and the average execution time of each specialized skill.

[0105] Each healthcare worker has varying abilities, and each patient's condition is different. For example, intravenous puncture is more difficult and time-consuming for both elderly and young patients. Therefore, using average execution time allows for accurate measurement of the actual execution time for each task.

[0106] In this way, the medical skills recording device continuously tracks and dynamically updates the actual time each medical staff member spends performing various general skills and specialized skills, thus building an accurate personal skills time database.

[0107] The departmental patient entry device acquires treatment information of patients in each ward, extracts the estimated workload for each patient in the next scheduling cycle, and calculates the estimated workload for each department. The estimated workload includes the number of times each general skill is executed and the number of times each specialized skill is executed.

[0108] The departmental patient entry device includes: a patient entry module, a departmental workload entry module, and a workload update module. The workload update module and the patient entry module are respectively connected to the departmental workload entry module.

[0109] The patient entry module extracts the hospitalization information of all patients in the hospital and generates a patient ID for each patient, which includes the department code and the patient code.

[0110] For example, a patient in the gastroenterology department might have the number RE-103, where RE represents the gastroenterology department code and 103 represents the patient's patient code.

[0111] The department workload entry module retrieves the medical record information of all patients, extracts the various operations that each patient needs to perform in the next scheduling cycle from the medical record information, so as to obtain the estimated workload of each patient, and creates a table of the estimated workload of patients in the same department to obtain the estimated workload table of each department.

[0112] For example, the departmental workload entry module automatically parses the electronic medical record information (including doctor's orders, nursing plans, treatment plans, scheduled examination / surgery records, etc.) of all inpatients in the next scheduling cycle (e.g., the next 8 hours), identifies and extracts all medical and nursing procedures planned for each patient in that cycle (e.g., Patient A needs 2 intravenous infusions, 1 hourly monitoring of vital signs, and 1 postoperative dressing change; Patient B needs 1 daily hemodialysis session and 1 specialized wound care session). The specific identification and extraction methods are existing technologies and will not be elaborated here. In practice, cases can also be manually identified and relevant information entered and extracted.

[0113] The department workload entry module converts each operation into an estimated time consumption based on the standard operation time in the "Medical and Nursing Skills Library" (e.g., intravenous infusion = 15 minutes / time, hemodialysis = 240 minutes / time), and sums them up to obtain the individual estimated total workload for each patient. Finally, the department workload entry module automatically categorizes patients by their department, summarizes and calculates the total number of operations and total time consumption for all patients in the same department, and generates a structured table—the Department Estimated Workload Table (containing core fields such as department name, scheduling cycle, operation name, estimated total number of operations, and estimated total time consumption).

[0114] .

[0115] Table 1: Projected Workload of a Department;

[0116] The workload update module monitors patient admission information in real time, obtains the number of patients entering each department in the next scheduling cycle based on the newly entered admission information, and updates the estimated workload table for each department.

[0117] The workload update module tracks outpatient admission dynamics in real time and intelligently predicts emergency admission trends, dynamically updating the expected number of patients and corresponding workload for each department in the next scheduling cycle, ensuring that departmental workload predictions are always based on the latest and most comprehensive admission information.

[0118] In practice, outpatients generally need to make an appointment for admission when applying for hospitalization, so the time outpatients enter each department is pre-entered. The time emergency patients are transferred to departments cannot be fixed; they are all transferred temporarily. Therefore, emergency patient information can only be predicted. Specifically:

[0119] The workload update module includes:

[0120] The outpatient information monitor is used to monitor outpatient information that will be available for the next scheduling cycle in real time.

[0121] The emergency information monitor is used to predict which emergency cases will need to enter the department in the next scheduling cycle based on historical emergency data.

[0122] The information updater updates the projected workload table based on outpatient and emergency information.

[0123] For example, one hour before the start of the scheduling cycle, the outpatient information monitor in the workload update module scans the system in real time and finds that three new patients have been added to the orthopedics department who have already made appointments. Simultaneously, the emergency information monitor, based on historical data analysis, predicts that approximately two emergency patients may be transferred to the orthopedics department in the next scheduling cycle due to accidental injuries. The information updater immediately acquires this new information (a total of five people), combines it with the average standard workload per orthopedic patient (e.g., X operations per person per day), automatically calculates the increased workload, adds it to the existing "estimated workload table" for the orthopedics department, and generates an updated total workload forecast for the orthopedics department for the following week in real time.

[0124] In practice, based on historical emergency room data, the expected number of emergency patients entering the department in the next scheduling cycle can be directly calculated using the average number of patients transferred to each department within each scheduling cycle (implying how many specialized and general skills each patient needs to perform). Alternatively, other methods can be used for prediction, such as neural networks and big data analysis.

[0125] The staff scheduling device uses the constraint that the total workload of the medical staff assigned to each department during the scheduling cycle is greater than the expected workload of the corresponding department, and aims to minimize the total number of medical staff assigned, to allocate matching medical staff to all departments.

[0126] The personnel scheduling device includes: a personnel information acquisition module,

[0127] The personnel information acquisition module is connected to the skills information recording module by signal. It extracts the medical staff who can be on duty in the next shift cycle from the personnel skills table and generates a candidate personnel skills table.

[0128] The skills list for candidates essentially identifies those who will be able to work in the next scheduling cycle. In practice, it's impossible to assign all medical staff to work in the department; instead, it's necessary to assign as few medical staff as possible to allow for sufficient rotation and rest.

[0129] The departmental demand acquisition module is connected to the departmental workload entry module to obtain the estimated workload table for each department in the next scheduling cycle.

[0130] The scheduling design module establishes a scheduling allocation model based on the expected workload of each department in the next scheduling cycle as a constraint. It extracts the minimum number of medical staff from the candidate personnel skill list as the fitness function of the scheduling allocation model. The scheduling allocation model is solved to generate the scheduling plan for the next scheduling cycle.

[0131] After the shift schedule for the next shift cycle is generated, a notification will be sent to the corresponding medical staff through the hospital's management system. These medical staff will then arrive at their respective departments to work at the designated shift change time according to the notification.

[0132] The task assignment module assigns corresponding work tasks to medical staff based on their work skills when they execute the shift schedule.

[0133] In this plan, each medical staff member will move between different departments. Therefore, when assigning work tasks, the medical staff's skills need to be taken into account.

[0134] During the scheduling phase, it has been ensured that the human resources allocated to each department can meet the department's expected workload. Therefore, within each department, tasks can be assigned and arranged sequentially based on the availability and skill level of medical staff. For example, in the urology department, if catheter care is currently required, this task can be assigned to medical staff capable of performing catheter care. Tasks related to general skills, such as intravenous puncture and medication preparation, can be prioritized for medical staff without urological expertise. The actual tasks performed and the order of task allocation within the department are determined by the medical staff based on the situation. The scheduling plan provided in this application ensures a human resource surplus, so internal staff only need to adapt their work accordingly.

[0135] The key to this application's ability to precisely assign tasks to each healthcare worker lies in the hospital's current digital management system, which requires each process to have a designated staff member and maintain records. For example, after medications are transferred from the central pharmacy to the department, healthcare workers must scan the corresponding barcode with a tablet to identify the dispenser, the administerer, and the administration time. Thus, simply issuing electronic tasks to each healthcare worker's tablet in the backend allows for efficient task allocation. While task allocation requires configuration based on actual circumstances, further optimization is not feasible. Assigning tasks to tablets is a current technological practice and will not be elaborated upon here.

[0136] The key to this application lies in constructing a shift allocation model and solving the shift allocation model. Based on this, this application provides Embodiment 2:

[0137] Example 2: Example 2 provides a method for constructing a shift allocation model based on Example 1.

[0138] The scheduling model can be implemented using integer programming, heuristic algorithms, and constrained programming. The core reason for adopting the particle swarm optimization algorithm in this scheme is that its swarm intelligence optimization characteristics can efficiently handle high-dimensional, multi-constraint combinatorial optimization problems in medical scheduling. In the dynamically adjusted search process, it quickly approaches the global optimal solution of "minimum manpower configuration" while taking into account the complex constraints of matching medical staff skills and balancing departmental workload.

[0139] The scheduling model is as follows:

[0140] Department code i, i∈{1, 2, 3, ... I}, i represents the department index, and I represents the total number of departments;

[0141] The job skill is encoded as k, k∈{1, 2, 3, ... K}, where k represents the job skill index and K represents the total number of job skills.

[0142] The medical staff code j, j∈{1, 2, 3, ... J}, j represents the medical staff index, and J represents the total number of medical staff;

[0143] Job skill coding for medical staff member J , , The label represents the skill level of the j-th medical worker. =0 indicates that medical staff j has not mastered the k-th job skill. =1 indicates that medical staff j has mastered the kth job skill;

[0144] Workload sequence D of department i i , ; This represents the quantity of job skill k required by department i. It is a non-negative integer;

[0145] Define the allocation status X of medical staff; , This indicates the allocation status of the j-th medical staff member;

[0146] ,when When = 0, it indicates that the j-th medical staff member has not been assigned a shift. When =i, it indicates that the j-th medical staff member will be assigned to the i-th department for duty in the next shift cycle;

[0147] Constraint 1: ;

[0148] Constraint 2: ; Indicates the length of the scheduling cycle. This represents the total average time taken for all medical staff in department i to perform work skill k within the current scheduling cycle.

[0149] The meaning of this number of terms is that all items are extracted. The medical staff j (the medical staff assigned to department i) are then assessed, and their mastery of job skill k is obtained. ,in this way, Ultimately, this represents the number of medical staff assigned to department i who possess job skill k. This represents the average number of times the medical staff assigned to department i perform work skill k in the next scheduling cycle. and Multiplication yields the total number of times medical staff assigned to department i perform work skill k.

[0150] The meaning of this number of terms is that all items are extracted. Medical staff j (medical staff assigned to department i). This equals the total number of medical staff assigned to department i. This represents the total working hours of the medical staff assigned to department i;

[0151] This indicates the time required for nurses assigned to department i to perform job skill k. This represents the total time required for nurses in department i to perform all their work skills k.

[0152] For healthcare workers who are unable to perform job skill k, when calculating the average time to perform job skill k, a multiplier gt is set, where gt = 3~10. The multiplier gt is multiplied by the highest time to perform job skill k, which is the value introduced by the average time when the healthcare worker who is unable to perform job skill k performs job skill k.

[0153] For example, in department A, there are 4 medical staff. The average time taken by the first 3 staff to perform job skill k is 6, 12, and 12 respectively. The last staff member is unable to perform job skill k, so we take g as 4. The average time taken by the last staff member to perform job skill k is then 48. Thus, the total average time taken by all medical staff in department A to perform job skill k is (6+12+12+48) / 4 = 19.5.

[0154] Constraint 1 is used to constrain the number of times that medical staff with job skill k can perform their duties within the scheduling cycle among the medical staff assigned to department i, which is greater than the number of times department i needs job skill k within the scheduling cycle.

[0155] Constraint 2 is used to constrain the total workload performed by medical staff assigned to department i to be greater than the sum of the workloads of all work skills k performed by medical staff assigned to department i.

[0156] In other words, constraint 1 ensures that the medical staff assigned to department i are capable of performing all the work skills within department i and completing the workload for each skill. Constraint 2, on the other hand, ensures that the total number of human resources assigned to department i is greater than the total number of human resources required by department i. Furthermore, when calculating the total workload required by department i, the average time taken by the medical staff assigned to department i for each work skill is used. This also helps to avoid assigning staff whose skills are not well-suited to department i to that department.

[0157] The main objective function is: .

[0158] It's an indicator function; its value is 1 when the condition is true, and 0 otherwise. That is, for a given allocation state, there are J x... j If x j If x ≠ 0, it means that medical staff j has been assigned to a department, and this is recorded as 1. All x j By processing the data using an indicator function and then summing the results, we can obtain the total number of staff members scheduled for the next work cycle.

[0159] Department codes i are arranged in adjacent order according to the physical location of the departments;

[0160] The scheduling model also includes the first sub-objective function L1:

[0161] ;

[0162] .

[0163] Indicates the parameter factor, Indicates an intermediate function. This represents the sum of human resources allocated to department i. This represents the sum of human resources allocated to department i+1. This represents the sum of human resources allocated to department i-1. This represents the sum of the workload that department i-1 needs to perform in the next scheduling cycle. This represents the sum of the workload that department i needs to perform in the next scheduling cycle. This represents the sum of the workload that department i+1 needs to perform in the next scheduling cycle.

[0164] The scheduling model also includes a second sub-objective function L2:

[0165] ;

[0166] ;

[0167] ;

[0168] , These represent the first and second intermediate parameters, respectively. This indicates the number of work skills k that need to be performed in department i during the previous scheduling cycle. This represents the quantity of job skill k required by department i. This represents the sum of the number of all job skills k in department i. This represents the sum of the number of work skills k that need to be performed in department i during the previous scheduling cycle. This represents the sum of human resources allocated to department i. This represents the sum of the workload that department i needs to perform in the next scheduling cycle.

[0169] The above is the constructed shift allocation model. The core objective of solving the shift allocation model is to find an optimal allocation state X for medical staff, which can achieve optimal performance on all three objective functions while satisfying two constraints.

[0170] Therefore, this application employs the particle swarm optimization algorithm for solving the problem. The solution process for the shift allocation model includes the following steps:

[0171] Step 1: Randomly generate N particles that simultaneously satisfy both constraint 1 and constraint 2. Each particle corresponds to a medical staff assignment state X. Use the N particles as the initial population.

[0172] , ,when When = 0, it indicates that the j-th medical staff member has not been assigned a shift. When =i, it indicates that the j-th medical staff member will be assigned to the i-th department for duty in the next scheduling cycle.

[0173] Thus, each particle X is an I×J dimension matrix, and the element in the j-th row and i-th column of the matrix represents the j-th medical staff member to be assigned to the (i-1)-th department (department 0 needs to be considered) in the next scheduling cycle.

[0174] There is no limit to the number of randomly generated particles. It is foreseeable that the higher the number of randomly generated particles, the greater the advantage of the final scheduling scheme.

[0175] Step 2: Calculate the fitness function value F(X) for each particle;

[0176] F(X) = λ0L0 + λ1L1 + λ2L2;

[0177] L0 represents the primary objective function, L1 represents the first secondary objective function, L2 represents the second secondary objective function, and λ0, λ1, and λ2 represent the first weighting factor, the second weighting factor, and the third weighting factor, respectively.

[0178] The fitness function F(X) in this scheme guides the diffusion process of the particle swarm optimization algorithm during iteration. The main objective function is used to reduce the total number of personnel required for scheduling; that is... To keep it as small as possible means that as particles diffuse, they will move in the direction that reduces the need for personnel to work.

[0179] Department codes 'i' are arranged in order of physical proximity; physical proximity refers to the distance between departments, with the shortest physical distance between two adjacent departments. In case of emergencies, medical staff can be borrowed from adjacent departments for support. Therefore, the total human resources of the three departments need to have a certain surplus.

[0180] ;

[0181] ;

[0182] The first objective function is used to control the level of affluence among personnel. This represents the sum of human resources arranged in department i, department i-1, and department i+1 (the product of the total number of personnel and the length of the scheduling cycle). This represents the sum of the workload (the product of the number of times a task skill is performed and the corresponding execution time) required by departments i, i-1, and i+1 in the next scheduling cycle. Therefore, the larger F1 is, the greater the surplus workload of these three departments. U1 is a constant; the larger F1 is, the greater the surplus workload. The smaller the value, the better. Thus, this first objective function ensures that at least one or two of the three adjacent departments have a significant workload surplus. Consequently, these three departments can second medical staff, increasing staff response efficiency. Used to guarantee The denominator is greater than 0.

[0183] The second objective function is used to measure the accuracy of workload prediction and its impact on task allocation.

[0184] This indicates the extra working hours of department i in the next scheduling cycle;

[0185] This represents the difference between the sum of the number of all work skills k in department i and the sum of the number of work skills k that need to be performed in department i during the previous scheduling cycle; This represents the quantity of job skill k required by department i. This represents the sum of the number of all job skills k in department i. This represents the sum of the number of work skills k that need to be performed in department i during the previous scheduling cycle. This represents the sum of human resources allocated to department i. This represents the sum of the workload that department i needs to perform in the next scheduling cycle.

[0186] Thus, g i The larger the value, the higher the predicted result is compared to the actual result, in which case it is necessary to... As large as possible, corresponding The smaller the value, the greater the potential gain. Conversely, the larger the potential gain, the smaller the potential gain. i The smaller the value, the smaller the predicted result is compared to the actual result, in which case it is necessary to... As small as possible, corresponding The smaller it is; if g i Larger, s i Smaller, or g i Smaller, s i Larger sizes will lead to It gets bigger.

[0187] In this way, the scheme can influence the objective function based on the accuracy of the prediction results, thereby feeding back into the particle training process.

[0188] Step 3: Sort all particles in descending order according to their fitness function values ​​to obtain a particle count.

[0189] The smaller the fitness function value of a particle, the more optimal its position.

[0190] Step 4: Set the fitness function value dividing line F0, and use particles with fitness function values ​​less than F0 as the diffusion population, and use particles with fitness function values ​​not less than F0 as the update population;

[0191] ;

[0192] Where t represents the index of the iteration number. Indicates the maximum number of iterations. Let represent the minimum fitness function value among all particles at the t-th iteration. This represents the maximum fitness function value among all particles at the t-th iteration.

[0193] Thus, as iterations proceed, F0 gradually increases, and the dividing line dynamically changes. Furthermore, as iterations continue, the number of particles in the update population decreases, while the number of particles in the diffusion population increases. In other words, in the early stages of iteration, most particles belong to the update population, and their diffusion direction is towards the center of the diffusion population. This allows for rapid diffusion in the direction of reducing particle fitness values ​​without getting trapped in local optima. In the later stages of iteration, most particles follow the traditional iterative method, allowing the model to converge quickly.

[0194] Step 5: For all particles in the updated population, randomly diffuse with the center point of all particles in the diffusion population as the target. For all particles in the diffusion population, randomly diffuse with the global optimal position as the direction.

[0195] The particle update formula for each iteration is:

[0196] ;

[0197] in, Let represent the allocation state of medical staff for particle q in the t-th iteration. This represents the allocation state of medical staff for particle q in the (t+1)th iteration (the state after iteration). This represents the velocity of the q-th particle in the t-th iteration;

[0198] For updating particle q in the population, its velocity at the t-th iteration is: ;

[0199] ;

[0200] Where w represents the inertia weight, This represents the velocity of the q-th particle in the (t-1)-th iteration. Indicates the first learning factor. This represents the first random number (a value randomly selected between 0 and 1). The value represents the center point of all particles in the diffusion population at the t-th iteration. In reality, it represents the position of the particle whose fitness function value is closest to the median of the fitness function value range of the entire diffusion population.

[0201] For example, if a diffusing population has 5 particles with fitness function values ​​of 1, 1, 3, 6, and 5, and the median of the fitness function value range is 2.5, and the number closest to 2.5 is 3, then the center point of the diffusing population is the 3rd particle.

[0202] For particle q in the diffusing population, its velocity at the t-th iteration is: `;

[0203] ;

[0204] Where w represents the inertia weight, This represents the velocity of particle q in the (t-1)th iteration. Indicates the first learning factor. This represents the first random number (a value randomly selected between 0 and 1). This represents the optimal position among all particles in this iteration. Indicates the globally optimal position. r1 represents the second learning factor, and r2 represents the second random number.

[0205] Step 6: Repeat steps 3 to 5 until the termination condition is met. The termination condition is reaching the maximum number of iterations or the gradient of the fitness function value of the global optimal solution is less than a preset threshold.

[0206] The reason for configuring the particle swarm optimization (PSO) algorithm in this application is that it has high requirements for the convergence efficiency of the algorithm. That is, after iterations, most particles need to have good fitness function values, focusing on the average fitness function value of the population. While it's impossible to find the theoretically optimal solution, ensuring that the found solution is at least a good one is sufficient.

[0207] The above are merely preferred embodiments of this application and are not intended to limit this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A medical staff scheduling system based on medical staff skills, characterized in that, include: The standard skills input and classification device is used to input the work skills of each department, classify the same work skills in each department into general skills, and classify the special work skills of each department into specialized skills. General skills are the job skills that all medical staff in all departments need to master; Specialized technical skills are the work skills required for special operations in specific departments; The medical skills input device records and updates the general skills and specialized skills mastered by each medical staff member in real time, and also updates the number of general skills and specialized skills that each medical staff member can perform during the shift cycle. The departmental patient entry device acquires treatment information of patients in each ward, extracts the estimated workload for each patient in the next scheduling cycle, and calculates the estimated workload for each department; the estimated workload includes the number of times each general skill is executed and the number of times each specialized skill is executed. The staff scheduling device uses the constraint that the total workload of the medical staff assigned to each department during the scheduling cycle is greater than the expected workload of the corresponding department, and aims to minimize the total number of medical staff assigned, to allocate matching medical staff to all departments. The medical skills input device includes: The medical staff information entry module assigns a unique ID number to each medical staff member in each department. The ID number includes the department code and the personnel code. The skill information extraction module extracts the execution time spent by medical staff when performing various general skills, as well as the execution time spent by medical staff when performing various specialized skills. The skills information recording module records the average execution time of each medical staff member for performing various general skills and the average execution time for performing various specialized skills, and generates a personnel skills table. The department's patient entry device includes: The patient entry module extracts the inpatient information of all patients in the hospital and generates a patient ID for each patient, which includes the department code and the patient code. The department workload entry module retrieves the medical record information of all patients, extracts the various operations that each patient needs to perform in the next scheduling cycle from the medical record information, so as to obtain the estimated workload of each patient, and makes a table of the estimated workload of patients in the same department to obtain the estimated workload table of each department. Personnel scheduling devices include: The personnel information acquisition module is connected to the skills information recording module by signal. It extracts the medical staff who can be on duty in the next shift cycle from the personnel skills table and generates a candidate personnel skills table. The department demand acquisition module is connected to the department workload entry module to obtain the estimated workload table for each department in the next scheduling cycle; The scheduling design module establishes a scheduling allocation model with the constraint of meeting the expected workload of each department in the next scheduling cycle. It extracts the minimum number of medical staff from the candidate personnel skill list as the fitness function of the scheduling allocation model; solves the scheduling allocation model to generate the scheduling plan for the next scheduling cycle. The task assignment module assigns corresponding work tasks to medical staff based on their work skills when they execute the shift schedule. The shift allocation model is as follows: Department code i, i∈{1, 2, 3, ... I}, i represents the department index, and I represents the total number of departments; The job skill is encoded as k, k∈{1, 2, 3, ... K}, where k represents the job skill index and K represents the total number of job skills. The medical staff code j, j∈{1, 2, 3, ... J}, j represents the medical staff index, and J represents the total number of medical staff; The job skill code for medical staff member J is A. j , , The label represents the skill level of the j-th medical worker. =0 indicates that medical staff j has not mastered the k-th job skill. =1 indicates that medical staff j has mastered the kth job skill; Workload sequence D of department i i , ; This represents the quantity of job skill k required by department i. It is a non-negative integer; Define the allocation status X of medical staff; , This indicates the allocation status of the j-th medical staff member; ,when When = 0, it indicates that the j-th medical staff member has not been assigned a shift. When =i, it indicates that the j-th medical staff member will be assigned to the i-th department for duty in the next shift cycle; Constraint 1: ; Constraint 2: ; Indicates the length of the scheduling cycle. This represents the total average time taken for all medical staff in department i to perform work skill k during the current scheduling cycle; The main objective function is: ; It is an indicator function; its value is 1 when the condition is true, and 0 otherwise. Where J represents the total number of medical staff, and j represents the index of medical staff. The label represents the skill level of the j-th medical worker. This represents the number of job skills k required by department i, where k represents the job skill index, K represents the total number of job skills, and i represents the department index. This represents the number of medical staff assigned to department i who possess job skill k. This represents the total number of medical staff assigned to department i; The solution process for the shift allocation model includes the following steps: Step 1: Randomly generate N particles that simultaneously satisfy both constraint 1 and constraint 2. Each particle corresponds to a medical staff assignment state X. Use these N particles as the initial population. Step 2: Calculate the fitness function value F(X) for each particle; （X） ； L0 represents the primary objective function, L1 represents the first secondary objective function, L2 represents the second secondary objective function, and λ0, λ1, and λ2 represent the first weighting factor, the second weighting factor, and the third weighting factor, respectively. Step 3: Sort all particles in descending order according to their fitness function values ​​to obtain a particle count. Step 4: Set the fitness function value dividing line F0, and use particles with fitness function values ​​less than F0 as the diffusion population, and use particles with fitness function values ​​not less than F0 as the update population; Step 5: For all particles in the updated population, randomly diffuse with the center point of all particles in the diffusion population as the target. For all particles in the diffusion population, randomly diffuse with the global optimal position as the direction. Step 6: Repeat steps 3 to 5 until the termination condition is met. The termination condition is reaching the maximum number of iterations or the gradient of the fitness function value of the global optimal solution is less than a preset threshold.

2. The medical staff scheduling system based on medical staff skills according to claim 1, characterized in that, The medical skills input device also includes: The skills information update module records the execution time of each medical staff member for each general skill and the execution time for each specialized skill, and updates the average execution time of each general skill and the average execution time of each specialized skill.

3. The medical staff scheduling system based on medical staff skills according to claim 2, characterized in that, The department's patient entry device also includes: The workload update module monitors patient admission information in real time, obtains the number of patients entering each department in the next scheduling cycle based on the newly entered admission information, and updates the estimated workload table for each department.

4. The medical staff scheduling system based on medical staff skills according to claim 3, characterized in that, The workload update module includes: The outpatient information monitor is used to monitor outpatient information that will be available for the next scheduling cycle in real time. The emergency information monitor is used to predict which emergency cases will need to enter the department in the next scheduling cycle based on historical emergency data. The information updater updates the projected workload table based on outpatient and emergency information.

5. The medical staff scheduling system based on medical staff skills according to claim 1, characterized in that, Department codes i are arranged in adjacent order according to the physical location of the departments; The scheduling model also includes a first sub-objective function L1: ; ； Indicates the parameter factor, Indicates an intermediate function; This represents the sum of human resources allocated to department i. This represents the sum of human resources allocated to department i+1. This represents the sum of human resources allocated to department i-1. This represents the sum of the workload that department i-1 needs to perform in the next scheduling cycle. This represents the sum of the workload that department i needs to perform in the next scheduling cycle. This represents the sum of the workload that department i+1 needs to perform in the next scheduling cycle.

6. The medical staff scheduling system based on medical staff skills according to claim 1, characterized in that, The scheduling model also includes a second sub-objective function L2: ; ; ； , These represent the first and second intermediate parameters, respectively. This indicates the number of work skills k that need to be performed in department i during the previous scheduling cycle. This represents the quantity of job skill k required by department i. This represents the sum of the number of all job skills k in department i. This represents the sum of the number of work skills k that need to be performed in department i during the previous scheduling cycle. This represents the sum of human resources allocated to department i. This represents the sum of the workload that department i needs to perform in the next scheduling cycle.

7. The medical staff scheduling system based on medical staff skills according to claim 1, characterized in that, ； Where t represents the index of the iteration number. Indicates the maximum number of iterations. Let represent the minimum fitness function value among all particles at the t-th iteration. This represents the maximum fitness function value among all particles at the t-th iteration.

Images