Interactive data simulation method and system for medical operation management

Abstract

The application belongs to the technical field of data simulation, and particularly relates to an interactive data simulation method and system for medical operation management, which comprises the following steps: obtaining hospital historical archive data and real-time monitoring data flow, and determining patient flow characteristics; constructing patient agents and path selection rules based on the patient flow characteristics, updating the patient agent state, and obtaining an interactive simulation result; adjusting path priority when congestion is over the limit, generating a virtual flow scene, and determining department load statistical results; performing resource redistribution when the load is over the limit; generating patient movement trajectory logs according to the resource configuration results, and iteratively adjusting simulation parameters to obtain bottleneck area positioning results. The application can realize patient flow process reconstruction, department load analysis, resource configuration adjustment and bottleneck node positioning.

Description

Technical Field

[0001] This invention belongs to the field of data simulation technology, specifically an interactive data simulation method and system for medical operations management. Background Technology

[0002] With the continuous growth of healthcare services, medical institutions are placing higher demands on refined operational scheduling, resource coordination and allocation, and healthcare service efficiency evaluation. Existing technologies typically rely on static statistics, single prediction models, or rule-based analysis methods. Although some solutions have introduced twin simulation, model fusion, and big data mining technologies, they still generally suffer from problems such as insufficient connection between historical data and real-time status, inadequate characterization of individual patient flow processes, weak ability to analyze path conflicts and departmental load linkages, and difficulty in accurately reconstructing bottleneck nodes and abnormal periods. These issues result in insufficient relevance and executability of scheduling results. Summary of the Invention

[0003] To address the above problems, this invention provides an interactive data simulation method and system for medical operations management, which at least solves the problems of how to accurately depict the dynamic flow of patients, identify departmental load imbalances, and locate bottleneck areas.

[0004] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0005] In a first aspect, the present invention provides an interactive data simulation method for medical operations management, the method comprising: Acquire historical hospital records and real-time monitoring data streams, and determine patient flow characteristics based on the historical hospital records data; Based on the characteristics of patient traffic, patient agents and path selection rules are constructed in the virtual environment, and the patient agent status is updated according to the real-time monitoring data stream to obtain interactive simulation results; When the congestion level exceeds the congestion threshold, the path priority of the patient agent is adjusted based on the interactive simulation results to generate a virtual traffic scenario and determine the department load statistics. If any department has a load value exceeding the load threshold in the department load statistics, resource reallocation is performed based on the department load statistics to obtain the resource allocation result. Based on the resource allocation results, run the patient agent simulation to generate patient movement trajectory logs, and iteratively adjust the simulation parameters of the virtual traffic scenario based on the patient movement trajectory logs to obtain the bottleneck area location results.

[0006] In one possible implementation, patient flow characteristics are determined based on hospital historical archive data, including: extracting registration time, arrival time, department visit records, service start time, service end time, and discharge time from the hospital historical archive data; statistically analyzing the extracted results according to a preset time window to obtain the distribution of patient arrival intensity, department visit order, and stay duration; and correcting the distribution of patient arrival intensity, department visit order, and stay duration based on historical fluctuation amplitude and date type to obtain patient flow characteristics.

[0007] In one possible implementation, patient agents and path selection rules in a virtual environment are constructed based on patient traffic characteristics, and the patient agent status is updated based on real-time monitoring data streams. This includes: configuring the patient agent with arrival time, target department sequence, triage level, movement speed range, and maximum waiting time based on patient traffic characteristics; constructing a virtual environment including department nodes and access nodes, and generating path selection rules based on the waiting status of department nodes and the access capacity of access nodes; updating the patient agent's node location, queuing status, and service status based on real-time monitoring data streams to obtain interactive simulation results.

[0008] In one possible implementation, the path priority of the patient agent is adjusted based on the interactive simulation results, including: calculating the queue length, average waiting time, and flow density of department nodes and passage nodes based on the interactive simulation results; comparing the queue length, average waiting time, and flow density with queue length thresholds, waiting time thresholds, and flow density thresholds, respectively; increasing the path priority of the patient agent with the higher triage level and increasing the selection priority of alternative paths in the path selection rules when the queue length exceeds the queue length threshold, the average waiting time exceeds the waiting time threshold, or the flow density exceeds the flow density threshold; wherein, the congestion thresholds include queue length thresholds, waiting time thresholds, and flow density thresholds, which are determined based on historical operation statistics, node capacity limits, and preset operation rules.

[0009] In one possible implementation, generating virtual traffic scenarios and determining departmental load statistics includes: adjusting the path priority of patient agents, adjusting the patient arrival intensity in patient traffic characteristics within a preset disturbance range, and generating multiple rounds of virtual traffic scenarios; calculating the average waiting time, service volume per unit time, personnel service saturation, and equipment utilization rate corresponding to each round of virtual traffic scenarios; and determining the departmental load statistics based on the average waiting time, service volume per unit time, personnel service saturation, and equipment utilization rate.

[0010] In one possible implementation, resource reallocation is performed based on departmental load statistics, including: determining target departments based on departmental load statistics, where the target departments are those whose load values ​​exceed the load threshold; determining associated departments based on referral relationships or resource sharing relationships between the target departments and other departments; and adjusting the number of personnel, number of open clinics, or equipment appointment times in the target departments and associated departments, under constraints of matching personnel qualifications and job requirements, minimum on-call configuration, and equipment availability, to obtain resource allocation results.

[0011] In one possible implementation, a patient agent simulation is run based on the resource configuration results to generate a patient movement trajectory log, including: recording the patient agent's entry time, exit time, queuing time, service time, number of route changes, and department transfer order, thus generating the patient movement trajectory log.

[0012] In one possible implementation, the simulation parameters of the virtual traffic scenario are iteratively adjusted based on the patient's movement trajectory log to obtain the bottleneck area location result. This includes: extracting abnormal waiting nodes and transfer delay nodes from the patient's movement trajectory log; abnormal waiting nodes are nodes whose waiting time exceeds a preset time, and transfer delay nodes are nodes whose stay during transfer exceeds a preset time; adjusting the simulation parameters representing patient arrival intensity, path priority, and service duration based on the abnormal waiting nodes and transfer delay nodes; and in the adjusted virtual traffic scenario, if the difference between the waiting time corresponding to the abnormal waiting node and transfer delay node and the corresponding waiting time in the patient's movement trajectory log is less than a preset error threshold, then the abnormal waiting node or transfer delay node in the adjusted virtual traffic scenario is identified as the target bottleneck node, and the time window corresponding to the target bottleneck node is identified as the abnormal time period, thus obtaining the bottleneck area location result.

[0013] In one possible implementation, the bottleneck area location result includes the target bottleneck node and the abnormal time period corresponding to the target bottleneck node. The target bottleneck node is an abnormal waiting node or a transfer delay node.

[0014] Secondly, the present invention provides an interactive data simulation system for medical operations management, used to implement an interactive data simulation method for medical operations management, the system comprising: The traffic analysis module is used to acquire historical hospital records data and real-time monitoring data streams, and to determine patient traffic characteristics based on the historical hospital records data. The agent simulation module is used to construct patient agents and path selection rules in a virtual environment based on patient traffic characteristics, and update the patient agent status based on real-time monitoring data streams to obtain interactive simulation results; The path adjustment module is used to adjust the path priority of the patient agent based on the interactive simulation results when the congestion level exceeds the congestion threshold, generate virtual traffic scenarios and determine the department load statistics. The resource allocation module is used to perform resource reallocation based on the department load statistics results when there are department load values ​​that exceed the load threshold, and obtain the resource allocation results. The bottleneck location module is used to run patient agent simulation based on resource configuration results, generate patient movement trajectory logs, and iteratively adjust the simulation parameters of the virtual traffic scenario based on the patient movement trajectory logs to obtain the bottleneck area location results.

[0015] Compared with the prior art, the advantages and beneficial effects of the present invention are as follows: By integrating historical hospital data with real-time monitoring data streams, the system dynamically depicts patient flow characteristics, enhancing data timeliness. Through collaborative simulation of patient agents and path selection rules, it reconstructs individual flow processes, supporting congestion propagation analysis. By linking departmental load statistics with resource reallocation, it adjusts resource allocation. And by using patient movement trajectory logs to deduce bottleneck nodes, it locates target bottleneck nodes and abnormal time periods, improving the interpretability of scheduling data. Attached Figure Description

[0016] Figure 1 This is a schematic flowchart of the method of the present invention; Figure 2 This is a block diagram of the module composition of the system of the present invention. Detailed Implementation

[0017] To enable those skilled in the art to better understand the technical solution, the present invention will be described in detail below with reference to embodiments. The description in this part is only exemplary and explanatory, and should not be used to limit the scope of protection of the present invention in any way.

[0018] Interactive data simulation emphasizes the continuous updating, dynamic extrapolation, and feedback-based correction of complex business processes under data-driven conditions. Its core lies not in single-time prediction output, but in unifying historical data, real-time status, rule constraints, and scenario interventions within a single simulation framework, enabling simulation results to adjust synchronously with changes in external inputs. Based on this, twin simulation is used to construct a virtual evolutionary environment corresponding to the actual operation process; model fusion is used to coordinate the connections between different types of data and different analysis mechanisms; and big data mining is used to extract traffic change patterns, node correlation characteristics, and abnormal evolution clues from multi-source operation records. Based on these ideas, an interactive data simulation method for healthcare operations management can be further developed. By linking and extrapolating patient flow, node congestion, resource allocation, and bottleneck areas, it provides a technical foundation for subsequent dynamic analysis and scheduling.

[0019] like Figure 1 As shown, an interactive data simulation method for healthcare operations management includes: Acquire historical hospital records and real-time monitoring data streams, and determine patient flow characteristics based on the historical hospital records data; After acquiring historical hospital records and real-time monitoring data streams, the historical records are first organized to form a basic record set for subsequent processing. This basic record set covers at least the time and departmental records related to patient flow, such as registration, arrival, waiting, consultation, examination, and discharge. Then, based on the historical hospital records data, patient flow patterns are statistically analyzed across different date types, time periods, and treatment paths to obtain patient flow characteristics. These characteristics characterize the arrival rhythm, departmental flow, and dwell distribution formed during the hospital's historical operations, providing initial input for subsequent patient agent generation, path selection rule setting, and departmental load estimation. The real-time monitoring data stream is integrated and time-aligned at this stage, enabling dynamic updates of patient agent status in subsequent stages.

[0020] The patient flow characteristics are determined based on the hospital's historical archive data, including: extracting registration time, arrival time, department visit records, service start time, service end time, and discharge time from the hospital's historical archive data; statistically analyzing the extracted results according to a preset time window to obtain the distribution of patient arrival intensity, department visit order, and stay duration; and correcting the distribution of patient arrival intensity, department visit order, and stay duration based on historical fluctuation amplitude and date type to obtain the patient flow characteristics.

[0021] In one embodiment, hospital historical data can come from registration systems, triage systems, electronic medical record systems, examination appointment systems, and billing and discharge record systems. To avoid inconsistencies in time formats, department names, and patient identifiers across different systems that could distort subsequent statistical results, the hospital historical data is first standardized. Standardization may include: converting registration time, arrival time, service start time, service end time, and discharge time to the same time base; mapping aliases, abbreviations, or historical names of the same department to a unified department name; and associating the identifiers of the same patient across different systems to create continuous medical records.

[0022] After standardization, the following data is extracted from the hospital's historical records: registration time, arrival time, departmental visit record, service start time, service end time, and discharge time. Registration time reflects the moment of appointment or registration; arrival time reflects the moment the patient actually enters the hospital service process; departmental visit record reflects the patient's flow between departments; service start time and service end time determine the duration of a single service; and discharge time determines the end point of the complete medical process. Then, the extracted results are statistically analyzed according to a preset time window. The preset time window can be set to 15 minutes, 30 minutes, or one hour, depending on the hospital's operational management needs. The setting principle is to reflect changes in patient flow during peak periods without amplifying occasional fluctuations due to excessively short time spans.

[0023] Based on preset time windows, the number of patients arriving within each time window can be counted to obtain patient arrival intensity; the order of department visits can be obtained according to the order in which patients visit departments during a single visit; and the distribution of patient dwell time during waiting, consultation, examination, and the overall visit can be calculated based on arrival time, service start time, service end time, and departure time. To make the patient flow characteristics more consistent with the actual operation of the hospital, it is also necessary to make corrections based on historical fluctuation amplitude and date type. Historical fluctuation amplitude can be determined based on the range of changes in the number of patients arriving on the same date within the same time window over a recent period, and is used to reflect the combined impact of rhythmic and sporadic fluctuations. Date type is used to distinguish between weekdays, weekends, statutory holidays, and special dates with significant changes in hospital workload; different date types correspond to different patient arrival patterns and department flow patterns.

[0024] During the correction process, statistical results from multiple historical periods under the same date type can be used as a baseline to weight and correct patient arrival intensity, departmental visit order, and dwell time distribution, thus obtaining patient flow characteristics. These patient flow characteristics are not single statistical values, but rather a set of features for subsequent simulations, including at least the patient arrival intensity, typical departmental visit order, and dwell time distribution corresponding to the time window. This processing method ensures that subsequently generated patient agents retain historical operational patterns while possessing a flow foundation that matches the actual date type.

[0025] Based on the characteristics of patient traffic, patient agents and path selection rules are constructed in the virtual environment, and the patient agent status is updated according to the real-time monitoring data stream to obtain interactive simulation results; After extracting patient flow characteristics, patient agents are generated in the virtual environment based on these characteristics, and path selection rules corresponding to the actual spatial layout of the hospital are established simultaneously. The virtual environment includes at least the outpatient entrance area, waiting area, treatment area, examination area, and the paths between these areas, used to describe the patient's movement and treatment process within the hospital. The patient agent is a simulation object in the virtual environment corresponding to a single patient, carrying behavioral information about the patient's movement, waiting, reception, examination, and transfer within the hospital. Different patient agents reflect individual differences in the treatment process through attributes and status information such as arrival time, target department sequence, triage level, waiting tolerance time, current node, queuing status, and service status. Triage level and waiting tolerance time are determined based on historical data statistics or preset rules. Path selection rules characterize the flow of patient agents between different department nodes and passage nodes. Subsequently, real-time monitoring data streams are connected to the virtual environment to dynamically correct the current location, queuing status, and service status of the patient agents, ensuring that the patient flow process in the virtual environment continuously closely reflects the current operating state of the hospital, thus obtaining interactive simulation results.

[0026] Based on patient flow characteristics, a patient agent and path selection rules are constructed in a virtual environment, and the patient agent status is updated according to real-time monitoring data streams. This includes: configuring the patient agent with arrival time, target department sequence, triage level, movement speed range, and maximum waiting time according to patient flow characteristics; constructing a virtual environment including department nodes and access nodes, and generating path selection rules based on the waiting status of department nodes and the access capacity of access nodes; updating the patient agent's node location, queuing status, and service status according to real-time monitoring data streams to obtain interactive simulation results.

[0027] In one embodiment, the construction of the patient agent is not a simple replication of historical records, but rather a structured configuration of patient behavior elements within the hospital based on patient flow characteristics. Specifically, the arrival time of the patient agent can be configured according to the patient arrival intensity within a time window, the target department sequence can be configured according to the department visit order obtained from historical statistics, the triage level can be configured according to triage records or reception priority rules, and the movement speed range and maximum waiting time can be configured according to the historical performance of different patient groups in the hospital's passage and waiting. The arrival time is used to determine the point in time when the patient agent enters the virtual environment, the target department sequence is used to determine the treatment nodes and examination nodes that the patient agent may pass through subsequently, the triage level is used to determine the priority passage order of the patient agent in congestion, the movement speed range is used to limit the movement rhythm of the patient agent in the passage nodes, and the maximum waiting time is used to determine whether the patient agent triggers path switching, node transfer, or abnormal stay during the waiting process. To avoid the patient agent configuration being too idealistic, random sampling can be performed according to the historical distribution results under the same date type and the same time window, so that different patient agents still maintain individual differences under the same overall pattern.

[0028] The virtual environment is constructed using a combination of nodes and paths. Department nodes represent spatial units that provide services or create spaces for people to stay, such as registration windows, waiting areas, consultation rooms, examination rooms, payment areas, and pharmacy areas. Access nodes represent corridors, corners, entrances / exits, elevator entrances, and connections between different functional areas. Each department node needs to record the current number of people waiting, the number of people served, the average service time, and the capacity. Each access node needs to record its access capacity, current traffic flow, and whether there are any access restrictions. Path selection rules are generated based on the waiting status of department nodes and the access capacity of access nodes. Specifically, they can be set according to the following principles: when there are multiple accessible paths to the same target department, the path with the shorter estimated arrival time and lower traffic load is prioritized; when the number of people waiting at a department node continues to increase and approaches its capacity limit, the priority of paths leading to that department node is reduced; when the traffic flow at a access node approaches its capacity limit, the priority of paths passing through that access node is reduced, and the selection weight of alternative access nodes is increased. The path priority here is not an abstract concept, but rather a ranking criterion used to drive the patient agent to choose among multiple candidate paths. Traffic capacity can be set based on hospital corridor width, historical number of people passing through per unit time, and on-site management rules. Waiting status can be determined jointly based on the current number of people in the queue, average service time, and the number of people already in the service queue.

[0029] After the real-time monitoring data stream is connected, it is used to continuously update the patient agent status. The real-time monitoring data stream can include one or more of the following: access control count data, waiting list call data, department queuing data, examination appointment status data, and video passenger flow statistics. Data from different sources are first time-aligned and location-mapped before entering the virtual environment, so that each real-time monitoring record can correspond to a specific department node or passage node. Then, the patient agent's node, queuing status, and service status are updated according to the real-time monitoring data stream. If the real-time monitoring data shows that a patient has entered the waiting area, the corresponding patient agent's node is updated to the department node corresponding to the waiting area, and the status is updated to queuing status; if the real-time monitoring data shows that a patient has been called and entered the examination room, the corresponding patient agent's node is updated to the examination room node, and the status is updated to service status; if the real-time monitoring data shows that a passage node has a significant accumulation of people within the current time window, the passage time for patient agents passing through that passage node is increased simultaneously. For real-time monitoring data that cannot be mapped one-to-one to a specific patient, a group correction method can be used. This involves batch correcting the status of patient agents within the same node and time window based on real-time statistical changes in the number of people at each node. After this processing, patient agents in the virtual environment no longer statically evolve based solely on historical traffic characteristics but dynamically change under the drive of the current hospital operation, ultimately forming interactive simulation results. These results include at least the number of people queuing at each department node, the flow status at each access node, the in-hospital location of each patient agent, and the service progress of each patient agent, used for congestion assessment and path priority adjustment in the next stage.

[0030] When the congestion level exceeds the congestion threshold, the path priority of the patient agent is adjusted based on the interactive simulation results to generate a virtual traffic scenario and determine the department load statistics. After updating the patient agent status, congestion in the virtual environment is identified based on the interactive simulation results. When the queue length, average waiting time, or patient density at a department node or access node exceeds the corresponding threshold, a path priority adjustment process is initiated to prioritize the passage and reception of patients with higher triage levels and increase the weight of alternative paths. After path priority adjustment, the patient arrival intensity is varied within a preset disturbance range to generate multiple rounds of virtual traffic scenarios. The average waiting time, service volume per unit time, personnel service saturation, and equipment utilization rate in each round of scenarios are statistically analyzed to form departmental load statistics. The congestion threshold and preset disturbance range are set based on historical operational statistics, node capacity limits, and hospital operating rules, ensuring that the load assessment results reflect the current simulation state while remaining consistent with the actual operational boundaries of the hospital.

[0031] Adjusting the path priority of patient agents based on interactive simulation results includes: calculating the queue length, average waiting time, and flow density of department nodes and passage nodes based on the interactive simulation results; comparing the queue length, average waiting time, and flow density of nodes with queue length thresholds, waiting time thresholds, and flow density thresholds, respectively; increasing the path priority of patient agents with higher triage levels and increasing the selection priority of alternative paths in the path selection rules when the queue length exceeds the queue length threshold, the average waiting time exceeds the waiting time threshold, or the flow density of nodes exceeds the flow density threshold; wherein, the congestion thresholds include queue length thresholds, waiting time thresholds, and flow density thresholds, which are determined based on historical operation statistics, node capacity limits, and preset operation rules.

[0032] In one embodiment, when adjusting the path priority of patient agents based on the interactive simulation results, the calculations for each department node and each passage node are performed one by one according to a preset statistical period. For department nodes, the focus is on counting the current number of people waiting, the number of people in the service queue, the average waiting time over a recent period, and the number of patients completed per unit time. For passage nodes, the focus is on counting the current number of people passing through, the number of people passing through per unit time, and the number of people temporarily staying. The queue length can be represented by the number of patient agents in a waiting state within the node, the average waiting time can be represented by the average waiting time of patients who have completed services within the same statistical period, and the node flow density can be determined based on the ratio between the current number of people at the node and the node's allowed capacity.

[0033] Queue length threshold, waiting time threshold, and crowd density threshold are not fixed constants, but are set by combining historical operational statistics, node capacity limits, and preset operational rules. Historical operational statistics reflect the common operational status of the node under the same date type and time window; the node capacity limit restricts the number of people that can be accommodated in waiting areas, outer buffer zones of examination rooms, corridors, elevator entrances, etc., under safe conditions; preset operational rules reflect the hospital's management requirements regarding priority for emergency care, examination appointments, and special patients. In actual setting, historical statistical values ​​from several recent dates of the same type can be read to obtain the common queuing and waiting levels of the node during peak periods. Then, combined with the node's capacity and the hospital's on-site crowd control rules, queue length threshold, waiting time threshold, and crowd density threshold are formed respectively. Subsequently, the queue length, average waiting time, and node crowd density obtained in the current statistical period are compared with the corresponding thresholds item by item.

[0034] When any of the three indicators exceeds the corresponding threshold, the node is considered to be in a congested state, and path priority adjustment is initiated. Path priority adjustment is carried out in two parts. The first part focuses on the patient agents themselves, adjusting the passage priority of patient agents according to their triage level, allowing critically ill patients, patients with tight examination deadlines, or patients who have been waiting for a long time to pass through congested nodes first. The second part focuses on alternative paths, reassessing the access time, current congestion level, and the acceptance capacity of target nodes for different paths, increasing the selection priority of low-load alternative paths, and decreasing the selection priority of paths passing through congested nodes. If a passage node is approaching its capacity limit, the priority of all paths passing through that passage node is temporarily reduced, and patient agents are guided to other reachable paths. If the number of people waiting at a certain department node continues to increase, but the related departments still have acceptance capacity, some patient agents are allowed to be transferred to alternative service nodes according to established treatment rules. In this way, the flow order and flow path of patient agents in the virtual environment can be reordered without changing the actual departmental relationships and existing service rules of the hospital, resulting in updated results for subsequent scenario simulations.

[0035] The process of generating virtual traffic scenarios and determining departmental load statistics includes: adjusting the path priority of patient agents, adjusting the patient arrival intensity in patient traffic characteristics within a preset disturbance range, and generating multiple rounds of virtual traffic scenarios; calculating the average waiting time, service volume per unit time, staff service saturation, and equipment utilization rate corresponding to each round of virtual traffic scenarios; and determining the departmental load statistics based on the average waiting time, service volume per unit time, staff service saturation, and equipment utilization rate.

[0036] In one embodiment, when generating a virtual traffic scenario and determining the departmental load statistics, the patient agent distribution state after path priority adjustment is used as the baseline scenario. Then, multiple rounds of perturbation simulations are performed, combining this with the patient arrival intensity from the patient traffic characteristics. A preset perturbation range is used to limit the boundary of changes in patient arrival intensity. The setting principle is to cover both normal fluctuations that have occurred in historical operations and to reserve room for potential upward fluctuations during holidays, peak outpatient periods, or peak examination periods. The preset perturbation range can be determined based on historical fluctuation amplitudes and date types; for example, different intensity change ranges can be set for different date types such as weekday mornings, weekend afternoons, and the first day of holidays.

[0037] After setting the disturbance range, in each round of the virtual traffic scenario, the intensity of patient arrivals is adjusted by varying degrees, while maintaining the patient's departmental visit order, the basic attributes of the patient agent, and the path priority adjusted in the previous step. This ensures that multiple rounds of virtual traffic scenarios reflect different traffic pressures under the same operating rules. Each round of the scenario runs a complete patient agent simulation process until the operating status of each department node and each passage node within the corresponding time window is obtained. Subsequently, key indicators in each round of the scenario are statistically analyzed. Average waiting time reflects the overall waiting level of patients before waiting for their appointment or examination; service volume per unit time reflects the number of appointments or examinations actually completed by the department within a given statistical period; personnel service saturation reflects the proportion of time that medical staff are in service within the statistical period relative to the available service time; and equipment utilization reflects the proportion of time that examination equipment, treatment equipment, or window equipment are occupied within the statistical period relative to the available scheduling time.

[0038] To ensure that the statistical results can be directly used for operational judgment, the above-mentioned indicators for each department in each round of scenarios can be compiled into a department load record table, further providing peak, average, and fluctuation data. If a department consistently exhibits high average waiting times, near-maximum staff service saturation, and consistently high equipment utilization in most virtual traffic scenarios, it is marked as a high-load department. If a department only experiences short-term overload in a few high-intensity scenarios, it is marked as a fluctuating-load department. If a department maintains low waiting times and low service usage across multiple rounds of scenarios, it is marked as an adjustable department. The resulting department load statistics not only reflect the congestion spread trend under the current interactive simulation state but also the pressure boundaries of each department under different patient arrival intensities, providing a direct basis for subsequent resource reallocation.

[0039] If any department has a load value exceeding the load threshold in the department load statistics, resource reallocation is performed based on the department load statistics to obtain the resource allocation result. After completing the departmental load statistics, it is first determined whether any department has a load value exceeding the load threshold within the current simulation period. If a department's load value exceeds the threshold, the corresponding department is included in the resource reallocation process. Resource reallocation is not simply adding resources, but rather rearranging adjustable resources within the hospital's existing personnel, clinics, and equipment. During the reallocation process, the referral relationships, examination collaboration relationships, and shared resource relationships between high-load departments and surrounding departments are considered to redetermine the number of personnel, the number of open clinics, and equipment appointment times, resulting in a new resource allocation. The resource allocation results reflect the actual service arrangements that each department can perform during peak hours, providing a basis for subsequent patient agency operations and bottleneck area identification.

[0040] Resource reallocation is performed based on departmental load statistics, including: determining target departments based on departmental load statistics (the target departments are those whose load values ​​exceed the load threshold); determining associated departments based on referral relationships or resource sharing relationships between the target departments and other departments; and adjusting the number of personnel, number of open clinics, or equipment appointment times in the target departments and associated departments, under constraints of matching personnel qualifications and job requirements, minimum on-call configuration, and equipment availability, to obtain the resource allocation results.

[0041] In one embodiment, when performing resource reallocation, the departmental load statistics are first used as the basis for judgment. A comprehensive check is performed on each department's average waiting time, service volume per unit time, staff service saturation, and equipment utilization rate within the current time window. Load thresholds can be set based on the hospital's daily operational experience and historical statistical results. The threshold corresponding to average waiting time is used to determine whether patients are experiencing continuous waiting; the threshold corresponding to staff service saturation is used to determine whether medical staff are approaching a continuous full-load working state; and the threshold corresponding to equipment utilization rate is used to determine whether examination or treatment equipment is approaching full occupancy. If a department consistently exceeds its corresponding threshold for at least one key indicator within the current time window, or if multiple indicators simultaneously approach the upper limit of the threshold, then that department is identified as the target department.

[0042] Once the target department is identified, related departments are then identified based on the hospital's actual operational relationships. Related departments are not limited to those on the same floor or in adjacent locations; they can also be departments with referral relationships, collaborative examination relationships, or resource-sharing relationships with the target department. For example, outpatient clinics and auxiliary examination departments may have collaborative examination relationships, multiple examination areas may have physician support relationships, and multiple examination units may have equipment substitution relationships. By identifying these operational connections, resource reallocation can be avoided by limiting it to a single department, and cross-departmental collaboration can be achieved within the hospital's existing organizational structure.

[0043] After identifying the target and related departments, each available resource is checked. When reallocating personnel resources, first verify the current staffing gaps in the target department and the number of personnel available in related departments. Then verify the professional qualifications, job skills, and scheduling status of the personnel to be reallocated, ensuring that personnel qualifications match job requirements. If a position requires specialized patient reception, examination operation, or nursing qualifications, personnel with the corresponding qualifications will be reallocated only. Minimum on-call staffing limits the basic number of staff each department must maintain within any time window, preventing operational instability in related departments due to resource transfer. When reallocating consultation room resources, first check the number of consultation rooms available in the target department within the current time window, their usage status, and the availability of auxiliary equipment. Then determine if related departments have vacant consultation rooms or consultation rooms that can be temporarily repurposed. When reallocating equipment resources, first check the current occupancy status, expected release time, and appointment backlog of equipment used in the target department. Then determine if related departments have alternative equipment or adjustable appointment slots. If alternative equipment is available, priority will be given to alleviating equipment congestion through reservation and diversion methods; if no alternative equipment is available, equipment pressure will be alleviated by adjusting reservation intervals, extending opening hours, or temporarily adding concurrent use periods.

[0044] The resource reallocation can be executed sequentially at three levels: personnel, clinics, and equipment. Alternatively, it can address the most pressing resource type first, based on hospital management needs. If the target department primarily experiences a high number of waiting patients and insufficient capacity, priority should be given to increasing the number of qualified medical staff or opening more clinics. If the target department primarily experiences a significant backlog of examination appointments, priority should be given to adjusting equipment appointment times and usage order. If the target department has both insufficient capacity for patient care and insufficient capacity for examinations, the key bottlenecks in patient stays should be addressed sequentially. After reallocation, the adjusted personnel allocation, number of open clinics, and equipment appointment times are recorded in the resource allocation results. The resource allocation results should include at least the newly added or adjusted personnel arrangements for the target department, the number of available clinics, equipment opening times, and resource changes for related departments. This results provide a clear source, clear execution boundaries, and clear allocation targets, aligning with the hospital's actual operating conditions and facilitating subsequent mapping of resource changes back to the virtual environment for further patient flow simulation and bottleneck area identification.

[0045] Based on the resource allocation results, run the patient agent simulation to generate patient movement trajectory logs, and iteratively adjust the simulation parameters of the virtual traffic scenario based on the patient movement trajectory logs to obtain the bottleneck area location results.

[0046] After resource reallocation is completed, the resource configuration results are written back to the virtual environment, and the patient agent simulation is rerun with the updated personnel arrangements, clinic opening status, and equipment appointment times as the current simulation conditions. During the simulation, the entry time, exit time, queuing time, service time, number of path changes, and transfer order of the patient agent at each department node and access node are continuously recorded to form a patient movement trajectory log. Then, based on the patient movement trajectory log, the patient arrival intensity parameters, path priority parameters, and service time parameters in the virtual traffic scenario are iteratively adjusted so that the simulation results gradually approximate the abnormal waiting and transfer delays reflected in the log, and the bottleneck area location result is determined when the error meets the set conditions. The bottleneck area location result includes at least the target bottleneck node and the abnormal time period corresponding to the target bottleneck node.

[0047] Based on the resource allocation results, run the patient agent simulation to generate a patient movement trajectory log, including: recording the patient agent's entry time, exit time, queuing time, service time, number of route changes, and department transfer order, thus generating the patient movement trajectory log.

[0048] In one embodiment, when running the patient agent simulation and generating patient movement trajectory logs, the resource configuration results are first mapped to various service nodes in the virtual environment. The number of personnel is used to determine the number of people who can provide reception, triage, examination, or nursing services simultaneously within the same time window; the number of open clinics is used to determine the number of clinics that can provide services in parallel; and the equipment reservation time slot is used to determine the availability status of examination or treatment equipment in different time periods.

[0049] After mapping is completed, the simulation is re-executed according to the patient agent's current arrival time, target department sequence, triage level, and path priority. Each patient agent experiences the processes of entry, waiting, consultation, examination, department transfer, and departure in the updated virtual environment. To ensure logs are usable for subsequent location analysis, key events are continuously recorded during the simulation. Entry time indicates the moment a patient agent first enters a department node or passage node; departure time indicates the moment a patient agent leaves that node; queuing time indicates the duration a patient agent spends waiting before entering service; service duration indicates the duration a patient agent actually receives medical treatment, examination, or payment services within the node; path change count indicates the number of times a patient agent adjusts their route during a complete visit; and department transfer order indicates the order in which a patient agent passes through multiple departments.

[0050] To avoid fragmented log content, a unique log identifier can be created for each patient agent during log generation, and the logs can be written into an event sequence in chronological order. The event sequence includes at least a node identifier, event type, and event occurrence time. Event types can be categorized as entry events, exit events, queue start events, queue end events, service start events, service end events, and path switching events. If a patient agent changes their original route due to path priority adjustments during simulation, a path switching record is added to the log, and the node before the switch, the node after the switch, and the switch time are simultaneously written. If a patient agent repeatedly queues at the same node, the start and end times of each queue are recorded separately to distinguish between normal waiting and abnormal dwelling. This patient movement trajectory log not only records which nodes the patient agent passes through but also how long they stayed at each node, when they entered the service, and whether a path change occurred, thus comprehensively reflecting the patient flow process under the influence of resource allocation results. By adopting this log generation method, simulation results can be transformed into traceable and comparable time-series records, providing a direct basis for subsequent identification of abnormal waiting nodes, transfer delay nodes, and target bottleneck nodes.

[0051] The simulation parameters of the virtual traffic scenario are iteratively adjusted based on the patient's movement trajectory log to obtain the bottleneck area location result. This includes: extracting abnormal waiting nodes and transfer delay nodes from the patient's movement trajectory log; abnormal waiting nodes are nodes whose waiting time exceeds a preset time, and transfer delay nodes are nodes whose stay during transfer exceeds a preset time; adjusting the simulation parameters representing patient arrival intensity, path priority, and service duration based on the abnormal waiting nodes and transfer delay nodes; in the adjusted virtual traffic scenario, if the difference between the waiting time corresponding to the abnormal waiting node and transfer delay node and the corresponding waiting time in the patient's movement trajectory log is less than a preset error threshold, the abnormal waiting node or transfer delay node in the adjusted virtual traffic scenario is identified as the target bottleneck node, and the time window corresponding to the target bottleneck node is identified as the abnormal time period, thus obtaining the bottleneck area location result.

[0052] In one embodiment, when iteratively adjusting the simulation parameters of the virtual traffic scenario based on patient movement trajectory logs, the patient movement trajectory logs are first processed at the node level. The purpose of this node-level processing is to re-aggregate the waiting, service, and passage information of each patient agent at different nodes to the node dimension, forming waiting statistics and transfer statistics for each node within different time windows. Then, abnormal waiting nodes and transfer delay nodes are extracted from the node-level processing results. Abnormal waiting nodes are used to indicate nodes with significantly higher waiting times in the current simulation cycle. Specifically, a preset duration can be set according to the average waiting time of nodes, the percentile of waiting time of nodes, or the longest waiting time allowed by hospital on-site management. If the waiting time of a node exceeds the preset duration in several consecutive time windows, the node is marked as an abnormal waiting node.

[0053] The "Transfer Delay Node" indicates nodes where patients spend an unusually long time during transfers. The preset duration can be set based on historical average transfer times, path length, and on-site traffic rules. If a patient's stay before reaching the next target department exceeds the preset duration after leaving the previous service node, the corresponding passage node or intermediate waiting node is marked as a transfer delay node. After extracting abnormal nodes, the simulation parameters of the virtual traffic scenario are iteratively adjusted around these nodes. The "Patient Arrival Intensity Parameter" corrects for the number of patients entering the virtual environment within a specific time window, suitable for handling short-term traffic spikes caused by holiday fluctuations, concentrated registration, or batch examinations. The "Path Priority Parameter" corrects the selection order of patient agents among multiple candidate paths, suitable for handling local channel congestion or underutilization of alternative paths. The "Service Duration Parameter" corrects for the average processing time of reception, examination, or payment nodes, suitable for handling situations where the actual service efficiency of some nodes is lower than the initial estimate.

[0054] After each round of parameter adjustments, the virtual traffic scenario is rerun, and the waiting times for the adjusted abnormal waiting nodes and transfer delay nodes are compared with the corresponding waiting times in the patient movement trajectory logs. A preset error threshold is used to determine whether the simulation scenario has approximated the actual congestion state reflected in the logs. This threshold can be set based on hospital management needs, historical statistical fluctuation ranges, and positioning accuracy requirements. If the waiting time difference is still greater than the preset error threshold, the patient arrival intensity parameters, path priority parameters, and service duration parameters are adjusted again in the next round. If the waiting time difference is lower than the preset error threshold, the iteration stops, and the current scenario is identified as the target scenario that can truly reflect the bottleneck formation process. In this way, bottleneck localization does not rely solely on a single statistical judgment, but gradually approximates the actual operating state through log comparison and parameter callback, making the localization results more interpretable and reproducible.

[0055] The bottleneck area location results include the target bottleneck node and the abnormal time period corresponding to the target bottleneck node. The target bottleneck node is an abnormal waiting node or a transfer delay node.

[0056] In one embodiment, the bottleneck region location result can be further refined into target bottleneck nodes and corresponding abnormal time periods. Target bottleneck nodes represent key nodes that continue to experience abnormal waiting or abnormal transfer delays after iteration; these can be either abnormal waiting nodes or transfer delay nodes. If the same node exhibits both persistently long waiting times and concentrated patient delays during transfers, it is preferentially identified as the target bottleneck node. Abnormal time periods represent the time range within which the target bottleneck node is in a significantly congested state; these periods can be determined based on the start and end times of consecutively exceeding waiting time limits, the start and end times of consistently high patient numbers at the node, or the start and end times of frequent path switching.

[0057] In practice, the waiting time, number of people at each bottleneck node, and number of path switching times within each time window can be sorted first. Then, consecutive abnormal intervals can be identified, ultimately forming abnormal time periods. To facilitate subsequent use of the location results, the node type, node location, abnormal time period, corresponding waiting time level, and main associated paths of the target bottleneck node can be recorded during output. If the target bottleneck node is a department node, the focus should be on recording the queuing situation, service occupancy, and alternative department relationships of the corresponding department during the abnormal time period. If the target bottleneck node is a passage node, the focus should be on recording the flow density, passage time, and upstream and downstream associated nodes of the corresponding passage during the abnormal time period. This recording method allows the bottleneck area location results to be directly used for subsequent operational analysis and scheduling decisions. For example, if the target bottleneck node is mainly concentrated in the waiting area, further adjustments to the patient reception rhythm and waiting diversion strategy can be considered; if the target bottleneck node is mainly concentrated in the passage node in front of the examination equipment, the focus should be on checking the rationality of the appointment time settings and examination path arrangements. The target bottleneck nodes and abnormal periods obtained in this way can not only indicate where the congestion occurs, but also when the congestion is most obvious, thus providing a clear and actionable basis for judgment in hospital operation and management.

[0058] like Figure 2 As shown, an interactive data simulation system for healthcare operations management is used to implement an interactive data simulation method for healthcare operations management. The system includes: The traffic analysis module is used to acquire historical hospital records and real-time monitoring data streams, and to determine patient traffic characteristics based on the historical records. The module is also used to complete basic data access and traffic feature extraction. Hardware-wise, it may include a data interface circuit, a communication bus, a processor, and a historical data cache unit. The data interface circuit is used to access historical hospital records and real-time monitoring data streams; the communication bus is used to transmit data between various hardware units; the processor is used to perform time alignment, record extraction, statistical calculations, and feature generation; and the historical data cache unit is used to temporarily store raw data such as registration records, arrival records, consultation records, and discharge records, as well as the extracted patient traffic characteristics.

[0059] The agent simulation module is used to construct patient agents and path selection rules in the virtual environment based on patient traffic characteristics, and update the patient agent status according to real-time monitoring data streams to obtain interactive simulation results. The agent simulation module is used to complete virtual environment construction, patient agent generation, and interactive simulation. Hardware-wise, it can include a simulation calculation unit, a scene storage unit, and a state update unit. The simulation calculation unit is used to execute patient agent generation, path rule loading, and node flow calculation; the scene storage unit is used to save department nodes, access nodes, patient agent attributes, and path selection rules; and the state update unit is used to receive real-time monitoring data streams and refresh the patient agent's location, queuing status, and service status, thereby outputting interactive simulation results.

[0060] The path adjustment module is used to adjust the path priority of patient agents based on interactive simulation results when congestion exceeds a congestion threshold, generating a virtual traffic scenario and determining the departmental load statistics. The module is responsible for congestion identification, path priority adjustment, and load statistics. Hardware-wise, it may include a threshold comparison circuit, a scheduling processing unit, and a statistical calculation unit. The threshold comparison circuit compares queue length, waiting time, and node patient density with preset thresholds. The scheduling processing unit rearranges the path priority of patient agents and switches alternative paths when trigger conditions are met. The statistical calculation unit performs multiple rounds of statistics on the adjusted virtual traffic scenario and outputs the departmental load statistics.

[0061] The resource allocation module is used to perform resource reallocation based on the departmental load statistics when there are departmental load values ​​exceeding the load threshold, thus obtaining the resource allocation result. The resource allocation module is responsible for completing both resource reallocation and configuration result generation. Hardware-wise, it may include a resource mapping unit, a constraint verification unit, and a configuration output unit. The resource mapping unit reads the target department, related departments, and their schedulable resource information. The constraint verification unit verifies constraints such as personnel qualifications, minimum shift configuration, and equipment availability. The configuration output unit generates resource allocation results corresponding to the number of personnel, the number of open clinics, and the equipment reservation time slots, and sends the resource allocation results to the subsequent simulation stage.

[0062] The bottleneck localization module is used to run patient agent simulation based on resource configuration results, generate patient movement trajectory logs, and iteratively adjust the simulation parameters of the virtual traffic scenario based on the patient movement trajectory logs to obtain the bottleneck area localization result. The bottleneck localization module is used to complete trajectory log generation, parameter iteration, and bottleneck localization. Hardware-wise, it can include a log recording unit, a parameter iteration unit, and a localization output unit. The log recording unit records the patient agent's entry time, exit time, queuing time, service time, number of path changes, and department transfer order. The parameter iteration unit iterates and corrects relevant parameters of the virtual traffic scenario based on the patient movement trajectory logs. The localization output unit outputs the target bottleneck node and corresponding abnormal time period, thus forming the bottleneck area localization result.

[0063] It should be noted that, in this document, the terms "comprising," "including," and any other variations are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Specific examples have been used in this document to illustrate the principles and implementation methods of the present invention. These examples are merely for the purpose of helping to understand the method and core ideas of the present invention. The above descriptions are only preferred embodiments of the present invention. It should be pointed out that, due to the limitations of written expression and the objective existence of infinite specific structures, those skilled in the art can make several improvements, modifications, or variations without departing from the principles of the present invention, and can also combine the above technical features in an appropriate manner. These improvements, modifications, variations, or combinations, or the direct application of the concept and technical solution of the present invention to other situations without modification, should all be considered within the scope of protection of the present invention.

Claims

1. An interactive data simulation method for healthcare operations management, characterized in that, The method includes: Acquire historical hospital records and real-time monitoring data streams, and determine patient flow characteristics based on the historical hospital records data; Based on the patient traffic characteristics, patient agents and path selection rules in the virtual environment are constructed, and the patient agent status is updated according to the real-time monitoring data stream to obtain interactive simulation results; If the congestion level exceeds the congestion threshold, the path priority of the patient agent is adjusted according to the interactive simulation results to generate a virtual traffic scenario and determine the department load statistics. If any department load value exceeds the load threshold in the department load statistics, resource reallocation is performed based on the department load statistics to obtain the resource allocation result; Based on the resource configuration results, a patient agent simulation is run to generate a patient movement trajectory log. The simulation parameters of the virtual traffic scenario are then iteratively adjusted based on the patient movement trajectory log to obtain the bottleneck area location results.

2. The method according to claim 1, characterized in that, The process of determining patient flow characteristics based on the hospital's historical records includes: Extract the registration time, arrival time, department visit record, service start time, service end time, and discharge time from the hospital's historical archive data. The extracted results were statistically analyzed according to the preset time window to obtain the distribution of patient arrival intensity, department visit order and stay duration; The patient flow characteristics are obtained by correcting the patient arrival intensity, department visit order, and stay duration distribution based on historical fluctuation amplitude and date type.

3. The method according to claim 1, characterized in that, The step of constructing patient agents and path selection rules in the virtual environment based on the patient traffic characteristics, and updating the patient agent status based on the real-time monitoring data stream, includes: Based on the patient flow characteristics, configure the patient agent with arrival time, target department sequence, triage level, movement speed range, and maximum waiting time; Construct the virtual environment including department nodes and access nodes, and generate the path selection rules based on the waiting status of the department nodes and the access capabilities of the access nodes; The interaction simulation results are obtained by updating the patient agent's node, queuing status, and service status based on the real-time monitoring data stream.

4. The method according to claim 3, characterized in that, The step of adjusting the path priority of the patient agent based on the interactive simulation results includes: The queue length, average waiting time, and pedestrian density of the department node and the passage node are calculated based on the interactive simulation results. The queue length, the average waiting time, and the node crowd density are compared with the queue length threshold, the waiting time threshold, and the crowd density threshold, respectively. If the queue length exceeds the queue length threshold, the average waiting time exceeds the waiting time threshold, or the node flow density exceeds the flow density threshold, the path priority of the patient agent with the higher triage level is increased, and the selection priority of the alternative paths in the path selection rules is increased. The congestion thresholds include the queue length threshold, the waiting time threshold, and the crowd density threshold, which are determined based on historical operational statistics, node capacity limits, and preset operational rules.

5. The method according to claim 4, characterized in that, The generation of virtual traffic scenarios and determination of departmental load statistics include: After adjusting the path priority of the patient agent, the patient arrival intensity in the patient traffic characteristics is adjusted within a preset disturbance range to generate multiple rounds of the virtual traffic scenario; The average waiting time, service volume per unit time, personnel service saturation, and equipment utilization rate for each round of virtual traffic scenarios were statistically analyzed. The departmental load statistics are determined based on the average waiting time, the service volume per unit time, the staff service saturation, and the equipment utilization rate.

6. The method according to claim 1, characterized in that, The step of performing resource reallocation based on the departmental load statistics includes: The target department is determined based on the department load statistics results. The target department is the department whose department load value exceeds the load threshold in the department load statistics results. The associated departments are determined based on the referral relationships or resource-sharing relationships between the target department and other departments; Under the constraints of matching personnel qualifications with job requirements, minimum on-call configuration, and equipment availability, the number of personnel, number of open clinics, or equipment reservation time slots in the target department and the related departments are adjusted to obtain the resource allocation result.

7. The method according to claim 1, characterized in that, The step of running the patient agent simulation based on the resource configuration result and generating a patient movement trajectory log includes: The system records the patient's entry time, exit time, queuing time, service time, number of route changes, and department transfer order to generate the patient's movement trajectory log.

8. The method according to claim 1, characterized in that, The step of iteratively adjusting the simulation parameters of the virtual traffic scenario based on the patient's movement trajectory log to obtain the bottleneck area location result includes: Based on the patient's movement trajectory log, abnormal waiting nodes and transfer delay nodes are extracted. The abnormal waiting nodes are nodes whose waiting time exceeds a preset time, and the transfer delay nodes are nodes whose stay time during transfer exceeds the preset time. Based on the abnormal waiting nodes and the transfer delay nodes, the simulation parameters characterizing patient arrival intensity, path priority, and service duration are adjusted; In the adjusted virtual traffic scenario, if the difference between the waiting time corresponding to the abnormal waiting node and the transfer delay node and the corresponding waiting time in the patient's movement trajectory log is less than a preset error threshold, the abnormal waiting node or transfer delay node in the adjusted virtual traffic scenario is identified as the target bottleneck node, and the time window corresponding to the target bottleneck node is identified as the abnormal time period, thus obtaining the bottleneck area location result.

9. The method according to claim 8, characterized in that, The bottleneck area location result includes the target bottleneck node and the abnormal time period corresponding to the target bottleneck node. The target bottleneck node is either the abnormal waiting node or the transfer delay node.

10. An interactive data simulation system for healthcare operations management, used to implement the interactive data simulation method for healthcare operations management as described in any one of claims 1-9, characterized in that, The system includes: The traffic analysis module is used to acquire historical hospital records data and real-time monitoring data streams, and to determine patient traffic characteristics based on the historical hospital records data. The agent simulation module is used to construct patient agents and path selection rules in a virtual environment based on the patient traffic characteristics, and update the patient agent status based on the real-time monitoring data stream to obtain interactive simulation results; The path adjustment module is used to adjust the path priority of the patient agent based on the interactive simulation results when the congestion level exceeds the congestion threshold, generate a virtual traffic scenario and determine the department load statistics. The resource allocation module is used to perform resource reallocation based on the department load statistics results when there are department load values ​​exceeding the load threshold, and obtain resource allocation results. The bottleneck location module is used to run a patient agent simulation based on the resource configuration results, generate a patient movement trajectory log, and iteratively adjust the simulation parameters of the virtual traffic scenario based on the patient movement trajectory log to obtain the bottleneck area location result.

Images