A simulated intelligent scheduling method, system, and medium for patient rooms without family members present.

By generating care task queues, calculating priorities, and matching caregivers, and combining simulation to optimize scheduling instructions, the problems of low resource utilization and difficulty in verifying scheduling strategies in wards without family members were solved, thus achieving efficient and stable care services.

CN123117586APending Publication Date: 2026-11-17HANGZHOU YUNHE TECHNOLOGY SERVICE IND CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611086658.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-07-21
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

The existing care services in wards without family members have low resource utilization, lack dynamic response mechanisms, and scheduling strategies cannot be simulated and verified in advance, resulting in task backlog and large fluctuations in service quality.

Method used

By generating a care task queue, calculating and sorting task priorities, matching target caregivers from the caregiver resource pool based on task attributes, simulating the execution process, and adjusting scheduling instructions, closed-loop optimization is achieved.

Benefits of technology

This improved resource utilization, ensured priority execution of urgent tasks, optimized human resource allocation, guaranteed consistent service quality, and enabled pre-simulation verification of scheduling strategies.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123117586A_ABST
    Figure CN123117586A_ABST
Patent Text Reader

Abstract

This application relates to an intelligent scheduling simulation method, system, and medium for patient rooms without patient companions. The method includes generating a care plan based on patient information, creating care tasks based on the care plan, and adding the care tasks to a queue of tasks to be scheduled. The method calculates the priority of each care task in the queue and sorts the care tasks according to their priority to obtain an ordered task queue. Based on the task attributes of each care task in the ordered task queue, a multi-layered screening mechanism is used to match target caregivers from a caregiver resource pool, and scheduling instructions are generated. The execution process of the care tasks is simulated based on the scheduling instructions to obtain simulation data. Evaluation indicators are calculated based on the simulation data, and the scheduling instructions are adjusted based on the evaluation indicators. This addresses the problem in related technologies of how to achieve pre-simulation verification of scheduling plans while improving resource utilization.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This application relates to the field of intelligent scheduling technology, specifically to an intelligent scheduling simulation method, system, and medium for unattended hospital wards. Background Technology

[0002] Currently, care services in patient rooms without family members primarily employ a fixed binding model of "one-on-one" dedicated companionship or "one-to-many" general companionship, meaning a relatively fixed service relationship is formed between caregivers (attendants) and patients. Existing technologies largely rely on fixed templates, doctor's orders, and nursing orders in inpatient wards for task generation. Scheduling decisions mainly depend on direct assignment based on manual experience, lacking a comprehensive modeling and dynamic matching mechanism for multi-dimensional factors such as attendant capabilities, task type, and patient risk level. Furthermore, existing systems lack simulations based on realistic probability models before actual deployment, making it difficult to detect and optimize issues such as response delays and resource mismatches during operation.

[0003] However, the existing models and scheduling technologies still have the following technical problems: First, resource utilization is low. In the "one-to-one" model, different caregivers are unevenly busy due to differences in patient needs and the randomness of order taking. In the "one-to-many" model, there is a lack of dynamic response mechanism, which can easily lead to task backlog and response delays when facing concurrent emergency needs. Second, there is a lack of pre-validation methods. The scheduling strategy is directly put into actual execution, and it is impossible to assess the impact of care tasks on response time and caregiver workload before deployment, resulting in a lag in parameter adjustment. Third, the care plan is not personalized enough. It is difficult to generate refined care tasks based on the patient's physical and mental condition, daily habits and self-care ability, and it is also impossible to achieve accurate matching of tasks and caregivers. Service quality fluctuates greatly and supervision is difficult.

[0004] Therefore, for intelligent scheduling simulation methods for unaccompanied patient rooms, no effective solution has yet been proposed to address the issue of how to improve resource utilization while simultaneously achieving pre-simulation verification of scheduling schemes. Summary of the Invention

[0005] This application provides an intelligent scheduling simulation method, system, and medium for unattended patient rooms, which at least solves the problem in related technologies of how to improve resource utilization while realizing pre-simulation verification of scheduling schemes.

[0006] In a first aspect, embodiments of this application provide an intelligent scheduling simulation method for patient rooms without accompanying family members, the method comprising: A care plan is generated based on the patient information, and a care task is created based on the care plan. The care task is then added to the task queue to be scheduled. Calculate the priority of each care task in the queue of tasks to be scheduled, and sort the care tasks in the queue of tasks to be scheduled according to the priority to obtain an ordered task queue. Based on the task attributes of each care task in the ordered task queue, a target caregiver is matched from the caregiver resource pool through a multi-layer screening mechanism, and a scheduling instruction is generated. The execution process of the care task is simulated based on the scheduling instructions to obtain simulation operation data. Evaluation indicators are calculated based on the simulation operation data, and the scheduling instructions are adjusted based on the evaluation indicators.

[0007] In one embodiment, the step of generating a care plan based on patient information, creating a care task based on the care plan, and adding the care task to a queue of tasks to be scheduled includes: Based on the patient's admission information, Bernoulli distribution is used to determine whether the patient has a care need, and the care need determination result is obtained. When the care need determination result indicates that there is a care need, a care plan is generated through a large model based on the patient's medical orders, underlying diseases, physical signs and allergy history. Based on the care plan, create corresponding care tasks; The care task is matched with a pre-established care order database, and the care task is marked with task attributes; Add the care tasks marked with the aforementioned task attributes to the queue of tasks to be scheduled.

[0008] In one embodiment, the step of calculating the priority of each care task in the queue of tasks to be scheduled, and sorting the care tasks in the queue of tasks to be scheduled according to the priority to obtain an ordered task queue, includes: Priority parameters are obtained based on the basic weight of each care task in the queue of tasks to be scheduled, the patient's risk level, the real-time waiting time, and the dependencies between tasks. Based on the priority parameters, the risk factor, waiting time factor, and dependency factor for each care task are calculated respectively. Based on the base weight, the risk factor, the waiting time factor, and the dependency factor, calculate the total priority weight for each care task; According to the total priority weight in descending order, all the care tasks in the task queue to be scheduled are sorted to obtain an initial ordered task queue. The waiting status and emergency takeover tasks of each care task in the initial ordered task queue are monitored in real time. The total priority weight of the care tasks is adjusted and the initial ordered task queue is updated to obtain an ordered task queue.

[0009] In one embodiment, the step of matching target caregivers from the caregiver resource pool based on the task attributes of each care task in the ordered task queue through a multi-layered filtering mechanism and generating scheduling instructions includes: The caregivers in the caregiver resource pool are pre-classified into Category A resident caregivers and Category B mobile caregivers; Based on the task attributes of the highest priority care task in the ordered task queue, candidate caregivers who meet the basic conditions are selected from the caregiver resource pool to obtain the first candidate caregiver set. Based on the specific conditions of the current highest priority care task, select the candidate caregivers who meet the specific conditions from the first candidate caregiver set to obtain the second candidate caregiver set. Determine whether the second candidate caregiver set is empty. If the second candidate caregiver set is not empty, select a target caregiver that meets the priority rule from the second candidate caregiver set according to the task type of the current highest priority care task. If the second candidate caregiver set is empty, activate the resource shortage handling mechanism to determine the target caregiver. Based on the target caregiver and the current highest priority care task, a scheduling instruction is generated.

[0010] In one embodiment, the step of selecting candidate caregivers from the caregiver resource pool who meet basic conditions based on the task attributes of the highest priority care task in the ordered task queue, to obtain a first candidate caregiver set, includes: Extract the ward information and task difficulty coefficient information of the current highest priority care task; Caregivers are selected from the caregiver resource pool who match the ward information, are available, have a skill level that matches the task difficulty coefficient, and whose real-time fatigue level is below a preset threshold. The selected caregivers were sorted in ascending order of real-time fatigue level and descending order of service star rating to obtain the first set of candidate caregivers.

[0011] In one embodiment, the step of simulating the execution process of the care task based on the scheduling instructions to obtain simulation operation data, calculating evaluation indicators based on the simulation operation data, and adjusting the scheduling instructions based on the evaluation indicators includes: The simulation clock is triggered when a preset random event occurs; wherein, the random event includes patient arrival event, care task generation event, and caregiver attendance status change event. According to the scheduling instructions, the process of caregivers performing care tasks is simulated, and simulation operation data is recorded; the simulation operation data includes the response time, waiting time, execution duration, caregiver working hours, and task completion status for each care task. The evaluation indicators are calculated based on the simulation data, and the evaluation indicators include service quality indicators and resource utilization rate indicators. The parameters configured during the simulation process form a parameter space, and the evaluation index is optimized through a multi-objective optimization framework to obtain the target configuration parameter set; The generation rules for the scheduling instructions are adjusted according to the target configuration parameter set.

[0012] In one embodiment, triggering the simulation clock when a preset random event arrives includes: Modeling of patient arrival events: Poisson process and exponential distribution are used to generate patient arrival time intervals to determine the arrival time of each patient; Bernoulli distribution is used to determine whether a patient has a care need; discrete distribution is used to generate the clinical risk level of the patient; log-normal distribution is used to generate the length of hospital stay of the patient. Modeling of attendance status change events: The Bernoulli distribution is used to generate the daily attendance status of each caregiver in the caregiver resource database; Modeling of care task generation events: Using a negative binomial distribution to generate the number of on-demand, immediate care tasks; using a PERT distribution to generate the actual execution time of each care task; Based on the models corresponding to various preset random events, when any preset random event reaches its corresponding occurrence time, the random event is triggered and the simulation clock is advanced.

[0013] In one embodiment, the evaluation index is optimized using a multi-objective optimization framework to obtain a target configuration parameter set, including: All configuration parameters and their value ranges during the simulation are obtained to form a parameter space; Multiple sets of configuration parameters are sampled within the parameter space, and simulations are performed on each set to obtain the evaluation index vector corresponding to each set of configuration parameters. Based on a multi-objective optimization algorithm, with the evaluation index vector as the optimization objective, the Pareto front is obtained by searching the non-dominated solution set. Clustering algorithms are used to cluster the solutions on the Pareto front to identify robust regions where the fluctuation range of the evaluation index is lower than a preset threshold. The configuration parameters corresponding to the center point of the robust region are selected as the target configuration parameter set.

[0014] Secondly, embodiments of this application provide an intelligent scheduling simulation system for patient rooms without accompanying family members. The system is used to execute the above-described method and includes a task queue module, an ordered task queue module, a scheduling instruction generation module, and a simulation module; wherein: The task queue module is used to generate a care plan based on patient information, create care tasks based on the care plan, and add the care tasks to the task queue. The ordered task queue module is used to calculate the priority of each care task in the task queue to be scheduled, and sort the care tasks in the task queue to be scheduled according to the priority to obtain an ordered task queue. The module for generating scheduling instructions is used to match target caregivers from the caregiver resource pool based on the task attributes of each care task in the ordered task queue through a multi-layer filtering mechanism, and generate scheduling instructions. The simulation module is used to simulate the execution process of the care task based on the scheduling instructions, obtain simulation operation data, calculate evaluation indicators based on the simulation operation data, and adjust the scheduling instructions based on the evaluation indicators.

[0015] Thirdly, embodiments of this application provide a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements an intelligent scheduling simulation method for unattended patient rooms as described in the first aspect above.

[0016] Fourthly, embodiments of this application provide a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements an intelligent scheduling simulation method for unattended patient rooms as described in the first aspect above.

[0017] The intelligent scheduling simulation method, system, and medium for unattended hospital wards provided in this application embodiment have at least the following technical effects.

[0018] First, care plans and tasks are generated based on patient information, shifting from a "personnel-bound" to a "task-driven" approach and avoiding resource misallocation in traditional fixed models. Then, task priorities are calculated and sorted to ensure urgent tasks are executed first. Next, a multi-layered screening mechanism matches target caregivers from the caregiver resource pool based on task attributes, optimizing human resource allocation and reducing labor costs. Finally, simulation execution is performed, recording operational data and calculating evaluation indicators. Scheduling instructions are then adjusted based on these indicators, forming a closed-loop optimization. This allows the effectiveness of the scheduling strategy to be verified before actual deployment, ensuring consistent service quality and enabling data-driven continuous improvement. This addresses the challenge in related technologies of how to improve resource utilization while simultaneously achieving pre-simulation verification of scheduling schemes.

[0019] Details of one or more embodiments of this application are set forth in the following drawings and description to make other features, objects and advantages of this application more readily apparent. Attached Figure Description

[0020] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a flowchart of an intelligent scheduling simulation method for wards without family members accompanying patients; Figure 2 This is a flowchart illustrating a simulation based on an exemplary embodiment; Figure 3 This is a system structure block diagram of an intelligent scheduling simulation system for unattended hospital rooms, according to an exemplary embodiment. Figure 4 This is a structural block diagram of an electronic device according to an exemplary embodiment. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of this application clearer, the application is described and illustrated below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application. All other embodiments obtained by those skilled in the art based on the embodiments provided in this application without inventive effort are within the scope of protection of this application.

[0022] Obviously, the accompanying drawings described below are merely some examples or embodiments of this application. Those skilled in the art can apply this application to other similar scenarios based on these drawings without any inventive effort. Furthermore, it is understood that although the efforts made in this development process may be complex and lengthy, for those skilled in the art related to the content disclosed in this application, any changes to design, manufacturing, or production based on the technical content disclosed in this application are merely conventional technical means and should not be construed as insufficient disclosure of the content of this application.

[0023] In this application, the reference to "embodiment" means that a specific feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment that is mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described in this application may be combined with other embodiments without conflict.

[0024] Unless otherwise defined, the technical or scientific terms used in this application shall have the ordinary meaning understood by one of ordinary skill in the art to which this application pertains. The terms “a,” “an,” “an,” “the,” and similar words used in this application do not indicate quantity limitation and may indicate singular or plural. The terms “comprising,” “including,” “having,” and any variations thereof used in this application are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or device that includes a series of steps or modules (units) is not limited to the listed steps or units, but may also include steps or units not listed, or may include other steps or units inherent to these processes, methods, products, or devices. The terms “connected,” “linked,” “coupled,” and similar words used in this application are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. “Multiple” used in this application refers to two or more. “And / or” describes the relationship between related objects, indicating that three relationships may exist; for example, “A and / or B” can represent: A alone, A and B simultaneously, and B alone. The character " / " generally indicates that the preceding and following objects are in an "or" relationship. The terms "first," "second," and "third" used in this application are merely to distinguish similar objects and do not represent a specific ordering of the objects.

[0025] In this document, it should be understood that the terms used may be technical means used to implement part of the present invention or other summary technical terms. For example, the terms may include: Emergency Responsibility Task: A special care task that arises from a patient's sudden clinical emergency, has the highest priority, can forcibly interrupt low-priority care tasks that are currently being performed, and immediately obtains caregiver resources.

[0026] Firstly, embodiments of this application provide an intelligent scheduling simulation method for wards without patient companions. Figure 1 This is a flowchart of an intelligent scheduling simulation method for hospital wards without patient companions, such as... Figure 1 As shown, the method includes: Step S101: Generate a care plan based on the patient information, create a care task based on the care plan, and add the care task to the queue of tasks to be scheduled.

[0027] Step S102: Calculate the priority of each care task in the task queue to be scheduled, and sort the care tasks in the task queue to be scheduled according to the priority to obtain an ordered task queue.

[0028] Step S103: Based on the task attributes of each care task in the ordered task queue, a target caregiver is matched from the caregiver resource pool through a multi-layer screening mechanism, and a scheduling instruction is generated.

[0029] Step S104: Simulate the execution process of the care task based on the scheduling instructions to obtain simulation operation data, calculate the evaluation index based on the simulation operation data, and adjust the scheduling instructions based on the evaluation index.

[0030] In summary, this application provides an intelligent scheduling simulation method for patient rooms without family members present. First, it generates care plans and creates care tasks based on patient information, shifting from a "personnel-bound" to a "task-driven" approach, avoiding resource misallocation in traditional fixed models. Then, it calculates and sorts task priorities to ensure urgent tasks are executed first. Next, based on task attributes, a multi-layered screening mechanism matches target caregivers from the caregiver resource pool, optimizing human resource allocation and reducing labor costs. Finally, it simulates execution, records operational data, calculates evaluation indicators, and adjusts scheduling instructions based on these indicators, forming a closed-loop optimization. This allows for verification of the scheduling strategy's effectiveness before actual deployment, ensuring consistent service quality and enabling data-driven continuous improvement. This addresses the problem in related technologies of how to improve resource utilization while simultaneously achieving pre-simulation verification of scheduling schemes.

[0031] In one embodiment, step S101 involves generating a care plan based on patient information, creating a care task based on the care plan, and adding the care task to the task queue to be scheduled. Specifically, this includes the following steps: Step S1011: Based on the patient's admission information, determine whether the patient has a care need through Bernoulli distribution, and obtain the care need determination result; Step S1012: When the care needs assessment result indicates that there is a care need, a care plan is generated through a large model based on the patient's medical orders, underlying diseases, physical signs and allergy history. Step S1013: Create corresponding care tasks according to the care plan; Step S1014: Match the care task with the pre-established care order database and mark the task attributes for the care task; Step S1015: Add the care tasks marked with task attributes to the queue of tasks to be scheduled.

[0032] Optionally, upon patient admission, a probability value following the Bernoulli distribution is dynamically generated based on the patient's admission information (such as age, medical history, and admitting department) and the hospital's preset care request rate. If the value falls within the care request probability range, care is deemed necessary; otherwise, no care is required. For patients deemed to have no care need, a ward is directly assigned and the ward status is updated to "admitted," ending the process. For patients deemed to have care need, a ward is assigned, and the corresponding caregiver queue status is updated to "awaiting service." Subsequently, the patient's electronic medical record information (including medical orders, underlying diseases, vital signs, allergy history, etc.) is input into a pre-trained large language model, which automatically generates a personalized care plan covering multiple dimensions such as daily living, behavior management, and medical cooperation. It should be noted that the large language model uses a pre-trained language model based on the Transformer architecture, including but not limited to any one of the GPT series models, LLM models, ChatGLM series models, Qwen series models, and Wenxin Yiyan large model.

[0033] After generating a care plan, each care item in the plan (such as assisting with eating, toileting care, and rehabilitation exercises) is instantiated as an independent care task. Each task records its associated patient, planned execution time window, and content description. Next, each care task is matched against a pre-established nursing order database in the hospital. This database is pre-classified and labeled by the agent, with each order item accompanied by tags such as priority principle, execution scenario (inside / outside the ward), task difficulty (high / medium / low), and estimated time. After matching, each care task is automatically assigned these tags as its task attributes. Finally, all care tasks with labeled attributes are sequentially added to the task queue for scheduling, awaiting subsequent priority calculation and scheduling allocation. This ensures the automation and standardization of the entire process from patient admission to task scheduling availability.

[0034] Specifically, the patient care needs generation algorithm includes: generating care needs based on Bernoulli distribution, with probabilities adjustable according to patient characteristics; zero-shot cold start: initializing basic parameters, weights, and probabilities based on domain knowledge and statistical yearbooks, and modeling multi-level care needs; few-shot generation adversarial: introducing scheduling quality benefits as an additional loss based on a small number of real care records to improve the realism of the generated needs; and a decoupled matching of needs and tasks with a dedicated nursing order database: before deployment in the hospital, the list of nursing orders is quickly classified and labeled according to the above classification structure to form a hospital-specific nursing order database. An agent quickly labels tags and enumeration values, such as priority principles, execution scenarios, task difficulty, and estimated time, which are then manually reviewed and fine-tuned. During daily care work, unrecorded items are created in a separate table, and the hospital periodically evaluates whether to update the dedicated nursing order database based on the custom database.

[0035] Steps S1011-S1015 enable personalized generation of care services, which can automatically generate more comprehensive and refined care plans that take into account both daily life and behavior management, compared to traditional plans that rely on human experience and fixed templates. This can be achieved by integrating multi-dimensional information such as medical orders, underlying diseases, physical signs and allergy history.

[0036] In one embodiment, step S102 involves calculating the priority of each care task in the task queue to be scheduled, and sorting the care tasks in the task queue according to their priorities to obtain an ordered task queue. Specifically, this includes the following steps: Step S1021: Obtain priority parameters based on the basic weight of each care task in the task queue to be scheduled, the patient risk level, the real-time waiting time, and the dependencies between tasks; Step S1022: Based on the priority parameters, calculate the risk factor, waiting time factor, and dependency factor for each care task; Step S1023: Calculate the total priority weight of each care task based on the basic weight, risk factor, waiting time factor, and dependency factor; Step S1024: Sort all care tasks in the task queue to be scheduled according to the total priority weight from high to low to obtain the initial ordered task queue. Step S1025: Monitor the waiting status of each care task and the emergency takeover task in the initial ordered task queue in real time, adjust the total priority weight of the care tasks and update the initial ordered task queue to obtain the ordered task queue.

[0037] Optionally, after obtaining the queue of tasks to be scheduled, each care task in the queue is first traversed. Based on the pre-marked base weights of the task in the nursing order database (P0=55, P1=40, P2=30, P3=20), the clinical risk level of the patient to which the task belongs (high risk +15, medium risk +8, low risk +0), and the dependencies between tasks (single dependent task +5, multiple dependent tasks +15), its base weight, risk factor, and dependency factor are determined respectively. Simultaneously, the waiting time for each care task from its entry into the queue to the current time is recorded, and the waiting time factor is calculated according to segmentation rules: if the waiting time is less than 30 minutes, the waiting time factor = waiting minutes × 0.5; if the waiting time is ≥30 minutes, the waiting time factor = 15 + waiting minutes × 0.2. Then, the above four factors are summed to obtain the total priority weight of each care task (total weight = base weight + risk factor + waiting time factor + ...). The task queue is sorted in descending order according to the total priority weight (dependency factor) to form an initial ordered task queue.

[0038] Next, during simulation or actual operation, the waiting status of each task in the queue is continuously monitored: if the waiting time of any task exceeds a preset threshold (e.g., 10 minutes) and has not yet been accepted, its overall priority weight is automatically increased (e.g., by adding an extra increment) or a reallocation is directly triggered. In addition, the system also monitors for emergency preemptive tasks generated by patients (e.g., sudden emergency medical needs). Once such a task is detected, the system immediately raises its overall priority weight to the highest value, placing it at the top of the queue, interrupting currently executing regular tasks, prioritizing the emergency preemptive task, and resuming the interrupted tasks after its completion. Through these dynamic adjustments, the system updates the initial ordered task queue in real time, ultimately obtaining an ordered task queue for scheduling.

[0039] Steps S1021-S1025 can quantify the total priority weight of each care task based on the task's basic weight, patient risk level, real-time waiting time, and inter-task dependencies, thus avoiding the problem of insufficient caregiver resources caused by multiple tasks stacking up during peak demand periods. In one embodiment, step S103 involves matching target caregivers from the caregiver resource pool based on the task attributes of each care task in the ordered task queue using a multi-layered filtering mechanism, and generating scheduling instructions. Specifically, this includes the following steps: Step S1031: Pre-classify the caregivers in the caregiver resource pool into Category A resident caregivers and Category B mobile caregivers; Step S1032: Based on the task attributes of the highest priority care task in the ordered task queue, select candidate caregivers who meet the basic conditions from the caregiver resource pool to obtain the first candidate caregiver set; specifically including: Extract ward information and task difficulty coefficient information for the current highest priority care task; Caregivers are selected from the pool of caregivers who match the ward information, are available, have a skill level that matches the task difficulty coefficient, and whose real-time fatigue level is below a preset threshold. The selected caregivers were sorted in ascending order of real-time fatigue level and descending order of service star rating to obtain the first set of candidate caregivers.

[0040] Step S1033: Based on the special conditions of the current highest priority care task, select candidate caregivers who meet the special conditions from the first candidate caregiver set to obtain the second candidate caregiver set. Step S1034: Determine whether the second candidate caregiver set is empty. If the second candidate caregiver set is not empty, select the target caregiver that meets the priority rule from the second candidate caregiver set according to the task type of the current highest priority care task. If the second candidate caregiver set is empty, start the resource shortage handling mechanism to determine the target caregiver. Step S1035: Generate scheduling instructions based on the target caregiver and the current highest priority care task.

[0041] Optionally, after obtaining the ordered task queue, all caregivers in the caregiver resource pool are first pre-divided into Category A resident caregivers (fixed responsibility for specific wards, undertaking duties such as safety monitoring and emergency response) and Category B mobile caregivers (mobile across wards, undertaking standardized tasks such as cleaning care and excretion care). Then, the highest priority care task in the ordered task queue is retrieved, its ward information and task difficulty coefficient are extracted, and caregivers meeting the basic conditions are selected from the categorized caregiver resource pool: ward matching, available status, skill level matching the task difficulty coefficient, and real-time fatigue level below a preset threshold (hereinafter referred to as the fatigue threshold, e.g., 0.8). The selected caregivers are then sorted in ascending order of real-time fatigue level and descending order of service star rating to form the first candidate caregiver set. Next, it is determined whether the highest priority care task belongs to a specific task type (such as excretion care or cleaning care). If so, caregivers meeting special conditions such as gender matching are further selected from the first candidate caregiver set to obtain the second candidate caregiver set. The system first selects a set of caregivers. If the current task has no special requirements, the first set of candidate caregivers is used as the second set. The system then checks if the second set of candidate caregivers is empty. If it is not empty, the system selects the best-matched target caregiver based on the task type (e.g., clinical support tasks prioritize matching resident caregivers of type A, standardized execution tasks prioritize matching mobile caregivers of type B, or strategies such as skill priority or proximity priority). If the second set of candidate caregivers is empty, the system initiates a resource shortage handling mechanism, allowing caregivers with fatigue levels exceeding the fatigue threshold to participate in the allocation. The system then checks if the difficulty of the current task is lower than the preset difficulty threshold and if the estimated time is shorter than the preset duration threshold. If so, a caregiver with excessive fatigue is directly assigned as the target caregiver. Otherwise, the priority weight of the current task is increased and it is pushed to the waiting queue. The response time is recorded for subsequent analysis. Finally, a scheduling instruction is generated based on the determined target caregiver and the current highest priority care task, instructing the target caregiver to execute the task.

[0042] It is worth noting that the core of the feedback-based adaptive scheduling algorithm is to trigger real-time decisions through key events and dynamically adjust based on the real-time system status to achieve dual optimization of task completion efficiency and caregiver resource utilization. The algorithm includes three key modules: First, event-driven, which automatically triggers matching or reallocation at critical moments such as task generation, caregiver idleness, or task alerts to ensure timely response; second, closed-loop correction, where the system will initiate a state reset and reallocation mechanism when the task waiting time exceeds a preset threshold, re-execute the matching process according to the latest situation, and record the reasons for subsequent algorithm optimization; and third, load balancing, which monitors the real-time fatigue level and current workload of caregivers, sets a fatigue threshold (e.g., fatigue level 0.8), and prioritizes scheduling caregivers with lower workloads to achieve dynamic and balanced task allocation in the short term.

[0043] Steps S1031-S1035, through the aforementioned funnel-shaped multi-layer screening mechanism, achieve refined matching of care tasks and optimized allocation of human resources, significantly reducing labor costs.

[0044] In one embodiment, Figure 2 This is a flowchart illustrating a simulation based on an exemplary embodiment, such as... Figure 2 As shown, step S104 involves simulating the execution process of a care task based on scheduling instructions to obtain simulation data. Evaluation indicators are then calculated based on the simulation data, and scheduling instructions are adjusted accordingly. Specifically, this includes the following steps: Step S1041: Trigger the simulation clock when a preset random event occurs; wherein, the random event includes patient arrival event, care task generation event, and caregiver attendance status change event. Specifically, it includes: Modeling of patient arrival events: Poisson process and exponential distribution are used to generate patient arrival time intervals to determine the arrival time of each patient; Bernoulli distribution is used to determine whether a patient has a care need; discrete distribution is used to generate the clinical risk level of the patient; log-normal distribution is used to generate the length of hospital stay of the patient. Modeling of attendance status change events: The Bernoulli distribution is used to generate the daily attendance status of each caregiver in the caregiver resource database; Modeling of care task generation events: Using a negative binomial distribution to generate the number of on-demand, immediate care tasks; using a PERT distribution to generate the actual execution time of each care task; Based on the models corresponding to various preset random events, when any preset random event reaches its corresponding occurrence time, the random event is triggered and the simulation clock is advanced.

[0045] Optionally, firstly, stochastic process modeling is performed: initially, based on domain knowledge and statistical yearbook analysis, the characteristics of patient information, caregiver information, and task information are analyzed to make reasonable assumptions and models, achieving a zero-sample cold start. Subsequently, a fit test is performed based on sample data during operation to select the optimal distribution. An event-driven mechanism is adopted, that is, the clock is triggered only when each nursing task reaches the corresponding time point according to the preset probability model, thereby evaluating the full-cycle operation effect after actual use. The generated task arrival information is used to advance the simulation clock, simulating the many-to-many scheduling process in the scenario of care in a patient room without a caregiver.

[0046] a. Random generation modeling of patient entities and their behaviors: Dynamically generate diverse virtual patient queues in chronological order.

[0047] Patient arrival: A Poisson process is used to model patient arrival, with the time intervals of patient arrival generated by an exponential distribution. The parameter λ is set according to the average monthly admission frequency of the client hospital to simulate the characteristics of random and independent patient arrival.

[0048] Patient clinical risk: Three risk levels (high, medium, and low) and their probabilities are preset. The risk level of the patient is directly generated by a custom discrete distribution (in order to simplify the model and focus on the core scheduling logic, patient characteristics and other attributes are merged into the connotation of the risk level). The distribution probability of each level is a configurable item and is set according to the historical statistical data of the client hospital.

[0049] Patient hospital stay duration: Considering that hospital stay duration is usually right-skewed and can be regarded as the product of multiple independent positive random factors, a log-normal distribution is used for modeling. The distribution parameters (mean, variance) are set based on publicly available statistical data such as the average length of stay in public tertiary hospitals.

[0050] Caregiver Demand Determination: A conditional probability model based on Bernoulli distribution is used to determine whether the patient has a caregiver demand upon admission. The caregiver rate is set as a configurable item based on historical data from the client's hospital or expert experience. To focus on key aspects and avoid overly complex models, caregiver demand is strongly correlated with the admission assessment node. This simplification does not affect the verification of the core logic for caregiver scheduling.

[0051] The main patient information generated is shown in Table 1 below: Table 1: Patient Information Table

[0052] b. Stochastic Perturbation Model of Caregiver Resource Availability: Caregiver attendance status (in-hospital / out-of-hospital) is a stochastic event in system operation, and its uncertainty has a significant impact on the execution of task allocation and scheduling strategies. Completely ignoring the impact of attendance status during modeling, while increasing caregiver redundancy during actual deployment to address potential absenteeism, may lead to significant deviations between actual system performance indicators and simulation results, reducing the model's prediction accuracy and decision reliability. If historical operational data is available, caregiver attendance rates can be calculated based on statistical data, and the total number of attendance days over 30 consecutive days can be counted to provide empirical evidence for the model. To balance model simplification with real-world relevance, this solution presets a daily caregiver attendance rate of p=0.95 and uses a Bernoulli distribution to probabilistically model daily attendance status. Subsequently, the attendance rate can be dynamically adjusted based on actual operational feedback or different scenario requirements to verify the robustness and adaptability of the caregiver allocation and scheduling system, and to observe whether the system can still operate stably and respond efficiently in changing environments.

[0053] c. Random Event Modeling for Nursing Order Generation: Daily Task List – In simulating the acquisition of a patient's daily nursing order list, the list of care items must first be determined, followed by the construction of a complete daily nursing order list. Formal Process for Generating the Daily Nursing Order List: The structured patient information is input into the [Nursing Order List Generation Agent]. The system automatically generates a daily nursing order task list that meets the patient's actual needs based on clinical logic and a rule engine, ensuring the accuracy and completeness of the nursing orders. Simulation Process: To support system simulation and algorithm verification, in the absence of real data, two types of tasks are selected from the nursing order simulation library to form the nursing order list: one is "daily routine" tasks (all selected), and the other is "on-demand immediate" tasks (generated by randomly selecting several items). The number of "on-demand immediate" tasks is modeled using a negative binomial distribution to reflect the strong heterogeneity of the patient population and the complex dynamic changes in their condition, thus more realistically depicting the uncertainty and volatility of nursing orders during hospitalization.

[0054] It should be noted that the parameters for the negative binomial distribution are set as follows: Mean μ: Represents the expected number of daily nursing orders. Based on clinical practice and expert advice, it is set as 18 routine mandatory items plus 8 on-demand immediate items (approximately 26 items in total). Discrete parameter k: Used to control the degree of data dispersion. The smaller the value of k, the greater the difference in nursing care needs among patients. Considering the diversity and complexity of the multi-specialty patient group in the hospital, the preset value of k is 1~2 to reflect the large variability in overall nursing care needs.

[0055] New tasks added instantly: An independent Poisson process is used to simulate new emergency or temporary needs that may arise randomly during a patient's hospitalization, with the incidence rate set with reference to the nursing intensity.

[0056] Actual task execution time: The PERT distribution was used for modeling to simulate the uncertainty of the actual execution time of similar care tasks caused by factors such as caregiver proficiency and patient cooperation. The three estimation parameters (upper limit, lower limit and most likely value) were set by care experts.

[0057] The above modeling aims to improve the rationality of the nursing order generation process and the realism of the simulation environment, providing a reliable data foundation for the evaluation of subsequent caregiver allocation and scheduling strategies.

[0058] The specific content of the generated nursing care task is shown in Table 2 below: Table 2: Nursing Order Task Sheet

[0059] Step S1042: According to the scheduling instructions, simulate the process of caregivers performing care tasks and record the simulation operation data; the simulation operation data includes the response time, waiting time, execution duration, caregiver working hours and task completion status of each care task; Optionally, during the simulation, the process of caregivers performing care tasks is simulated item by item according to the scheduling instructions generated in step S103, and the simulation operation data of each care task is recorded synchronously. Specifically, the generation time, acceptance time, start execution time, and completion time of each care task are recorded; the response time is generated based on the difference between the start execution time and the generation time; the waiting time is generated based on the difference between the acceptance time and the generation time; the execution duration is generated based on the difference between the completion time and the acceptance time; and the caregiver's working hours are generated based on the difference between the completion time and the start execution time. The completion status (completed, timed out, or interrupted) of each care task is also recorded. The above simulation operation data provides the data basis for the calculation of evaluation indicators in the subsequent step S1043.

[0060] Step S1043: Calculate the evaluation indicators based on the simulation operation data. The evaluation indicators include service quality indicators and resource utilization rate indicators.

[0061] Optionally, the following steps involve indicator calculation and analysis: recording task execution and collecting patient satisfaction feedback, transforming service quality and scheduling efficiency into measurable indicators, including service quality indicators, resource utilization indicators, and system efficiency indicators. This, in turn, drives the continuous optimization and iteration of the algorithm model and operational strategies. The core assessment indicators of quality factors are analyzed to evaluate the overall quality of care. The core assessment indicators are shown in Table 3: Table 3: Core Evaluation Indicators

[0062] Simultaneously, the quantitative analysis of the ratios between caregiver allocation, patient volume, and task efficiency ensures that the system optimizes operating costs and maximizes resource utilization while meeting service quality and response time requirements. Based on the core evaluation indicators of the efficiency factors, inefficient caregivers are identified, overall operational efficiency is assessed, and ROI trends are predicted. The core evaluation indicators of the efficiency factors are shown in Table 4. Table 4: Benefit Factor Table

[0063] Step S1044: Construct a parameter space from the parameters configured during the simulation, and optimize the evaluation index using a multi-objective optimization framework to obtain the target configuration parameter set; specifically including: All configuration parameters and their value ranges during the simulation are obtained to form a parameter space; Multiple sets of configuration parameters are sampled in the parameter space, and simulations are performed on each set to obtain the evaluation index vector corresponding to each set of configuration parameters. Based on a multi-objective optimization algorithm, with the evaluation index vector as the optimization objective, the Pareto front is obtained by searching the non-dominated solution set. Clustering algorithms are used to cluster solutions on the Pareto front to identify robust regions where the fluctuation of evaluation indicators is below a preset threshold. The configuration parameters corresponding to the center point of the robust region are selected as the target configuration parameter set.

[0064] Optionally, after calculating the evaluation indicators, all configurable parameters during the simulation (such as patient arrival rate, caregiver ratio, fatigue threshold, priority weight, etc.) are used to construct a multi-dimensional parameter space. The system employs a multi-objective optimization algorithm (such as NSGA-II) to search this parameter space: first, multiple sets of configuration parameters are sampled within the parameter space, and complete simulations are performed for each set to obtain the evaluation indicator vector corresponding to each set of configuration parameters; then, using the evaluation indicator vector as the optimization objective, a non-dominated solution set is searched to obtain the Pareto front. Subsequently, a clustering algorithm is used to cluster all solutions on the Pareto front, identifying regions where the fluctuation amplitude of the evaluation indicators is lower than a preset threshold as robust regions, and selecting the configuration parameters corresponding to the center point of this robust region as the target configuration parameter set. The target configuration parameter set is used to adjust the generation rules of scheduling instructions.

[0065] Step S1045: Adjust the generation rules of scheduling instructions according to the target configuration parameter set.

[0066] Optionally, after obtaining the target configuration parameter set, the system updates the generation rules of the scheduling instructions accordingly with the values ​​of each configuration parameter in the target configuration parameter set. Specifically, if the target configuration parameter set contains an adjustment value for the priority calculation weight, the system updates the basic weight, risk factor assignment, or segmented calculation coefficient of the waiting time factor in step S102; if it contains an adjustment value for the fatigue threshold, the system updates the fatigue threshold in the basic screening conditions of the multi-layer screening mechanism in step S103; if it contains an adjustment value for the ratio of A-class and B-class caregivers, the system adjusts the configuration ratio of the two types of caregivers in the caregiver resource pool. Through the above adjustments, a closed-loop feedback from simulation evaluation to scheduling rule optimization is completed, enabling the subsequently generated scheduling instructions to achieve a better balance between response speed, service quality, and operating costs.

[0067] Steps S1041-S1045, through the above event-driven stochastic process modeling and closed-loop simulation optimization, realize the pre-verification and parameter optimization of the scheduling strategy for unaccompanied patients before actual deployment, forming a closed loop of execution-evaluation-feedback-improvement, and continuously improving the robustness and economy of the scheduling strategy.

[0068] In summary, this application provides an intelligent scheduling simulation method for patient rooms without family members present. First, it generates care plans and creates care tasks based on patient information, shifting from a "personnel-bound" to a "task-driven" approach, avoiding resource misallocation in traditional fixed models. Then, it calculates and sorts task priorities to ensure urgent tasks are executed first. Next, based on task attributes, a multi-layered screening mechanism matches target caregivers from the caregiver resource pool, optimizing human resource allocation and reducing labor costs. Finally, it simulates execution, records operational data, calculates evaluation indicators, and adjusts scheduling instructions based on these indicators, forming a closed-loop optimization. This allows for verification of the scheduling strategy's effectiveness before actual deployment, ensuring consistent service quality and enabling data-driven continuous improvement. This addresses the problem in related technologies of how to improve resource utilization while simultaneously achieving pre-simulation verification of scheduling schemes.

[0069] This application also includes the following beneficial effects: Optimize human resource allocation: Increase the proportion of effective working hours for caregivers to avoid uneven workloads; To ensure consistent service quality, tasks are performed by the most suitable caregiver, reducing reliance on individual experience. Enhance system flexibility and robustness: Support dynamic adjustment of the ratio of A / B category caregivers (e.g., increase the proportion of A category caregivers in the ICU and increase the proportion of B category caregivers in the rehabilitation department) to adapt to the needs of different departments; Reduced labor costs: With the same service scale, the total number of caregivers required is reduced, and highly skilled caregivers focus on high-value tasks, improving overall efficiency.

[0070] Secondly, embodiments of this application provide an intelligent scheduling simulation system for wards without accompanying family members. Figure 3 This is a system architecture block diagram illustrating an intelligent scheduling simulation system for patient rooms without accompanying family members, according to an exemplary embodiment. Figure 3 As shown, the system includes a task queue module 310 to be scheduled, an ordered task queue module 320, a scheduling instruction generation module 330, and a simulation module 340; wherein: The task queue module 310 is used to generate a care plan based on patient information, create care tasks based on the care plan, and add the care tasks to the task queue. The ordered task queue module 320 is used to calculate the priority of each care task in the task queue to be scheduled, and sort the care tasks in the task queue to be scheduled according to the priority to obtain the ordered task queue. The scheduling instruction generation module 330 is used to match target caregivers from the caregiver resource pool based on the task attributes of each care task in the ordered task queue through a multi-level filtering mechanism, and generate scheduling instructions. The simulation module 340 is used to simulate the execution process of care tasks based on scheduling instructions, obtain simulation operation data, calculate evaluation indicators based on the simulation operation data, and adjust scheduling instructions based on the evaluation indicators.

[0071] In summary, the intelligent scheduling simulation system for unattended hospital wards provided in this application solves the problem in related technologies of how to achieve pre-simulation verification of scheduling schemes while improving resource utilization through the synergistic effect of the task queue module 310, the ordered task queue module 320, the scheduling instruction generation module 330, and the simulation module 340.

[0072] It should be noted that the intelligent scheduling simulation system for patient rooms without accompanying family members provided in this embodiment is used to implement the above-described embodiments, and details already described will not be repeated. As used above, the terms "module," "unit," "subunit," etc., can refer to a combination of software and / or hardware that performs a predetermined function. Although the device described in the above embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0073] Thirdly, embodiments of this application provide an electronic device, Figure 4 This is a block diagram illustrating an electronic device according to an exemplary embodiment. (e.g.) Figure 4 As shown, the electronic device may include a processor 41 and a memory 42 storing computer program instructions.

[0074] Specifically, the processor 41 may include a central processing unit (CPU), an application specific integrated circuit (ASIC), or one or more integrated circuits that can be configured to implement the embodiments of this application.

[0075] The memory 42 may include a large-capacity memory for data or instructions. For example, and not limitingly, the memory 42 may include a hard disk drive (HDD), a floppy disk drive, a solid-state drive (SSD), flash memory, an optical disk drive, a magneto-optical disk drive, magnetic tape, or a Universal Serial Bus (USB) drive, or a combination of two or more of these. Where appropriate, the memory 42 may include removable or non-removable (or fixed) media. Where appropriate, the memory 42 may be internal or external to a data processing device. In a particular embodiment, the memory 42 is non-volatile memory. In a particular embodiment, the memory 42 includes read-only memory (ROM) and random access memory (RAM). Where appropriate, the ROM may be a mask-programmed ROM, a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), an electrically alterable read-only memory (EAROM), or flash memory, or a combination of two or more of these. Where appropriate, the RAM can be Static Random-Access Memory (SRAM) or Dynamic Random-Access Memory (DRAM). DRAM can be Fast Page Mode Dynamic Random-Access Memory (FPMDRAM), Extended Data Out Dynamic Random-Access Memory (EDODRAM), Synchronous Dynamic Random-Access Memory (SDRAM), etc.

[0076] The memory 42 can be used to store or cache various data files that need to be processed and / or used for communication, as well as possible computer program instructions executed by the processor 41.

[0077] The processor 41 reads and executes the computer program instructions stored in the memory 42 to implement any of the intelligent scheduling simulation methods for unattended patient rooms in the above embodiments.

[0078] In one embodiment, an intelligent scheduling simulation device for patient rooms without accompanying family members may further include a communication interface 43 and a bus 40. Wherein, as Figure 4 As shown, the processor 41, memory 42, and communication interface 43 are connected through bus 40 and complete communication with each other.

[0079] The communication interface 43 is used to enable communication between the various modules, devices, units, and / or equipment in the embodiments of this application. The communication interface 43 can also enable data communication with other components such as external devices, image / data acquisition devices, databases, external storage, and image / data processing workstations.

[0080] Bus 40 includes hardware, software, or both, that couples together components of an intelligent scheduling simulation device for patient rooms without accompanying family members. Bus 40 includes, but is not limited to, at least one of the following: Data Bus, Address Bus, Control Bus, Expansion Bus, and Local Bus. For example, and not as a limitation, bus 40 may include an Accelerated Graphics Port (AGP) or other graphics bus, an Extended Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), a HyperTransport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an InfiniBand interconnect, a Low Pin Count (LPC) bus, a memory bus, a Micro Channel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local Bus (VLB) bus, or other suitable buses, or a combination of two or more of these. Where appropriate, bus 40 may include one or more buses. Although specific buses are described and illustrated in the embodiments of this application, this application considers any suitable bus or interconnection.

[0081] Fourthly, embodiments of this application provide a computer-readable storage medium storing a program thereon, which, when executed by a processor, implements the intelligent scheduling simulation method for unattended patient rooms provided in the first aspect.

[0082] The readable storage medium may be more specifically adopted, including but not limited to: portable disk, hard disk, random access memory, read-only memory, erasable programmable read-only memory, optical storage device, magnetic storage device, or any suitable combination thereof.

[0083] In a possible implementation, the present invention can also be implemented as a program product comprising program code, which, when the program product is run on a terminal device, causes the terminal device to perform steps implementing the intelligent scheduling simulation method for unattended hospital rooms provided in the first aspect.

[0084] The program code for executing the present invention can be written in any combination of one or more programming languages. The program code can be executed entirely on the user device, partially on the user device, as a standalone software package, partially on the user device and partially on a remote device, or entirely on a remote device.

[0085] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0086] The above embodiments merely illustrate several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A smart scheduling simulation method for hospital wards without accompanying family members, characterized in that, The method includes: A care plan is generated based on the patient information, and a care task is created based on the care plan. The care task is then added to the task queue to be scheduled. Calculate the priority of each care task in the queue of tasks to be scheduled, and sort the care tasks in the queue of tasks to be scheduled according to the priority to obtain an ordered task queue. Based on the task attributes of each care task in the ordered task queue, a target caregiver is matched from the caregiver resource pool through a multi-layer screening mechanism, and a scheduling instruction is generated. The execution process of the care task is simulated based on the scheduling instructions to obtain simulation operation data. Evaluation indicators are calculated based on the simulation operation data, and the scheduling instructions are adjusted based on the evaluation indicators.

2. The intelligent scheduling simulation method for unattended hospital wards as described in claim 1, characterized in that, The process of generating a care plan based on patient information, creating care tasks based on the care plan, and adding the care tasks to a queue of tasks to be scheduled includes: Based on the patient's admission information, Bernoulli distribution is used to determine whether the patient has a care need, and the care need determination result is obtained. When the care need determination result indicates that there is a care need, a care plan is generated through a large model based on the patient's medical orders, underlying diseases, physical signs and allergy history. Based on the care plan, create corresponding care tasks; The care task is matched with a pre-established care order database, and the care task is marked with task attributes; Add the care tasks marked with the aforementioned task attributes to the queue of tasks to be scheduled.

3. The intelligent scheduling simulation method for unattended hospital wards as described in claim 2, characterized in that, The process of calculating the priority of each care task in the queue of tasks to be scheduled, and sorting the care tasks in the queue of tasks to be scheduled according to the priority to obtain an ordered task queue, includes: Priority parameters are obtained based on the basic weight of each care task in the queue of tasks to be scheduled, the patient's risk level, the real-time waiting time, and the dependencies between tasks. Based on the priority parameters, the risk factor, waiting time factor, and dependency factor for each care task are calculated respectively. Based on the base weight, the risk factor, the waiting time factor, and the dependency factor, calculate the total priority weight for each care task; According to the total priority weight in descending order, all the care tasks in the task queue to be scheduled are sorted to obtain an initial ordered task queue. The waiting status and emergency takeover tasks of each care task in the initial ordered task queue are monitored in real time. The total priority weight of the care tasks is adjusted and the initial ordered task queue is updated to obtain an ordered task queue.

4. The intelligent scheduling simulation method for unattended hospital wards as described in claim 1, characterized in that, The process of matching target caregivers from the caregiver resource pool based on the task attributes of each care task in the ordered task queue through a multi-layered filtering mechanism, and generating scheduling instructions, includes: The caregivers in the caregiver resource pool are pre-classified into Category A resident caregivers and Category B mobile caregivers; Based on the task attributes of the highest priority care task in the ordered task queue, candidate caregivers who meet the basic conditions are selected from the caregiver resource pool to obtain the first candidate caregiver set. Based on the specific conditions of the current highest priority care task, select the candidate caregivers who meet the specific conditions from the first candidate caregiver set to obtain the second candidate caregiver set. Determine whether the second candidate caregiver set is empty. If the second candidate caregiver set is not empty, select a target caregiver that meets the priority rule from the second candidate caregiver set according to the task type of the current highest priority care task. If the second candidate caregiver set is empty, activate the resource shortage handling mechanism to determine the target caregiver. Based on the target caregiver and the current highest priority care task, a scheduling instruction is generated.

5. The intelligent scheduling simulation method for unattended hospital wards as described in claim 4, characterized in that, The first candidate caregiver set is obtained by selecting candidate caregivers from the caregiver resource pool based on the task attributes of the highest priority care task in the ordered task queue, which meets the basic conditions. This set includes: Extract the ward information and task difficulty coefficient information of the current highest priority care task; Caregivers are selected from the caregiver resource pool who match the ward information, are available, have a skill level that matches the task difficulty coefficient, and whose real-time fatigue level is below a preset threshold. The selected caregivers were sorted in ascending order of real-time fatigue level and descending order of service star rating to obtain the first set of candidate caregivers.

6. The intelligent scheduling simulation method for unattended hospital wards as described in claim 1, characterized in that, The process of simulating the execution of the care task based on the scheduling instructions to obtain simulation operation data, calculating evaluation indicators based on the simulation operation data, and adjusting the scheduling instructions based on the evaluation indicators includes: The simulation clock is triggered when a preset random event occurs; wherein, the random event includes patient arrival event, attendance status change event, and care task generation event; According to the scheduling instructions, the process of caregivers performing care tasks is simulated, and simulation operation data is recorded; the simulation operation data includes the response time, waiting time, execution duration, caregiver working hours, and task completion status for each care task. The evaluation indicators are calculated based on the simulation data, and the evaluation indicators include service quality indicators and resource utilization rate indicators. The parameters configured during the simulation process form a parameter space, and the evaluation index is optimized through a multi-objective optimization framework to obtain the target configuration parameter set; The generation rules for the scheduling instructions are adjusted according to the target configuration parameter set.

7. The intelligent scheduling simulation method for unattended hospital wards as described in claim 6, characterized in that, The step of triggering the simulation clock when a preset random event arrives includes: Modeling of patient arrival events: Poisson process and exponential distribution are used to generate patient arrival time intervals to determine the arrival time of each patient; Bernoulli distribution is used to determine whether a patient has a care need; discrete distribution is used to generate the clinical risk level of the patient; log-normal distribution is used to generate the length of hospital stay of the patient. Modeling of attendance status change events: The Bernoulli distribution is used to generate the daily attendance status of each caregiver in the caregiver resource database; Modeling of care task generation events: Using a negative binomial distribution to generate the number of on-demand, immediate care tasks; using a PERT distribution to generate the actual execution time of each care task; Based on the models corresponding to various preset random events, when any preset random event reaches its corresponding occurrence time, the random event is triggered and the simulation clock is advanced.

8. The intelligent scheduling simulation method for unattended hospital wards as described in claim 6, characterized in that, The evaluation metrics are optimized using a multi-objective optimization framework to obtain a set of target configuration parameters, including... All configuration parameters and their value ranges during the simulation are obtained to form a parameter space; Multiple sets of configuration parameters are sampled within the parameter space, and simulations are performed on each set to obtain the evaluation index vector corresponding to each set of configuration parameters. Based on a multi-objective optimization algorithm, with the evaluation index vector as the optimization objective, the Pareto front is obtained by searching the non-dominated solution set. Clustering algorithms are used to cluster the solutions on the Pareto front to identify robust regions where the fluctuation range of the evaluation index is lower than a preset threshold. The configuration parameters corresponding to the center point of the robust region are selected as the target configuration parameter set.

9. An intelligent scheduling simulation system for hospital wards without accompanying family members, characterized in that, The system is used to execute the method according to any one of claims 1 to 8, and the system includes a task queue module to be scheduled, an ordered task queue module, a scheduling instruction generation module, and a simulation module; wherein: The task queue module is used to generate a care plan based on patient information, create care tasks based on the care plan, and add the care tasks to the task queue. The ordered task queue module is used to calculate the priority of each care task in the task queue to be scheduled, and sort the care tasks in the task queue to be scheduled according to the priority to obtain an ordered task queue. The module for generating scheduling instructions is used to match target caregivers from the caregiver resource pool based on the task attributes of each care task in the ordered task queue through a multi-layer filtering mechanism, and generate scheduling instructions. The simulation module is used to simulate the execution process of the care task based on the scheduling instructions, obtain simulation operation data, calculate evaluation indicators based on the simulation operation data, and adjust the scheduling instructions based on the evaluation indicators.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements an intelligent scheduling simulation method for unattended patient rooms as described in any one of claims 1 to 8.