A Dynamic Scheduling Method for Medical Delivery in Comprehensive Hospitals

Abstract

This invention provides a dynamic scheduling method for medical delivery in comprehensive hospitals. The method includes: acquiring dynamic information about the hospital; constructing a scheduling decision model; inputting the hospital's dynamic information into the optimized scheduling decision model; and the scheduling decision model outputting the optimal scheduling scheme. This application features intelligent decision-making, high scheduling efficiency, and can effectively reduce task delays and balance personnel workload.

Description

Technical Field

[0001] This invention relates to the field of medical information technology, and in particular to a dynamic scheduling method for medical delivery in general hospitals. Background Technology

[0002] In recent years, the development of my country's medical and health services has attracted much attention, and the country is vigorously promoting the smart healthcare strategy. In large general hospitals, the scheduling of support staff and nursing assistants is a crucial aspect of hospital operations. However, the current central distribution and scheduling systems in most Chinese hospitals are relatively outdated, with significant room for optimization and improvement in terms of operational efficiency and cost.

[0003] Medical delivery in general hospitals refers to the delivery of medical supplies and specimens by support staff, nursing assistants, and nurses within the hospital, or accompanying patients for examinations or surgeries. Large general hospitals have medical delivery dispatch systems covering the entire hospital, with cross-deliveries between different departments, resulting in a large number of daily delivery tasks and a large number of personnel involved. However, most hospitals currently rely primarily on walkie-talkies, telephones, WeChat, and other communication methods, with delivery tasks manually assigned by a central dispatch center. This traditional dispatching method has the following two problems:

[0004] The scheduling efficiency is poor. Currently, decisions are mainly made based on human experience, generally assigning tasks according to the principle of proximity, which lacks scientific basis and easily leads to problems such as untimely handling of emergency tasks, unbalanced workloads among workers, and serious task delays.

[0005] Low scheduling efficiency. When there are many delivery tasks, manually arranging them one by one requires a lot of communication time, mainly because communication facilities limit the efficiency of transmitting scheduling information and cannot achieve real-time and effective monitoring of central scheduling. Summary of the Invention

[0006] The purpose of this invention is to provide a dynamic scheduling method for medical delivery in comprehensive hospitals. This method features intelligent decision-making, high scheduling efficiency, and can effectively reduce task delays and balance personnel workload.

[0007] A dynamic scheduling method for medical delivery in general hospitals includes:

[0008] Obtain dynamic information about the hospital;

[0009] Construct a scheduling decision model;

[0010] The hospital's dynamic information is input into the optimized scheduling decision model, which then outputs the optimal scheduling scheme.

[0011] The hospital's dynamic information includes:

[0012] Current hospital task information: The set of tasks that need to be completed, the starting department of each task, the ending department of each task, the earliest delivery time, and the latest delivery time;

[0013] Current hospital personnel information: real-time location of personnel, and current cumulative workload of each personnel.

[0014] The construction of the scheduling decision model includes:

[0015] Define the decision variables for the scheduling decision model;

[0016] Establish the objective function of the scheduling decision model;

[0017] Establish the constraints of the scheduling decision model.

[0018] The decision variables for defining the scheduling decision model include:

[0019] Indicates task a∈A t and task b∈A t Does {a} consist entirely of personnel k∈K, where K is the number of hospital staff? t Let be the set of hospital tasks at time t, and let a be the adjacent position before task b.

[0020] Introduce a virtual task 0 as the predecessor task of the first task and the successor task of the last task;

[0021] For any person k∈K, the endpoint of their virtual start task is f0 = ID. k ID k This indicates the department where worker k is currently located. The virtual end point and end point of the task are the end point of the last non-virtual task. That is, after completing the last task, the worker stays in place and waits for further instructions.

[0022] The objective function for establishing the scheduling decision model includes:

[0023] Let r a ∈{0,1} represents a non-virtual task a∈A t The relationship between whether it can be completed within 2 hours of becoming executable and its completion time is as follows:

[0024]

[0025] Therefore, the number of tasks executed in more than 2 hours is:

[0026]

[0027] Let z a ∈{0,1} represents task a∈A t The relationship between whether there is a delay and the completion time is as follows:

[0028]

[0029] Therefore, the weighted average number of delayed tasks is:

[0030] W a Weighting based on the urgency of the task

[0031] The maximum time required to complete all tasks is

[0032] C a The completion time of task a

[0033] Maximum difference in workload between personnel

[0034]

[0035] Among them, TW k This represents the cumulative workload of personnel k∈K.

[0036] CW k S represents the current workload of worker k. a For task a, the starting department is F. b For the department at the end of task b, P b The processing time after sampling and delivery for task b. This indicates the time required to travel from the starting department of task a to the ending department of task b;

[0037] Finally, weighting coefficients are used to merge multiple objectives into a single overall objective.

[0038]

[0039] Where H = |A t |w max That is, the number of tasks multiplied by the maximum task weight.

[0040] The constraints for establishing the scheduling decision model include establishing the lower bounds of the constraints, specifically:

[0041] make The following constraints will be used to standardize personnel scheduling criteria and constraints:

[0042] At the very end of each personnel list is only one non-virtual task:

[0043]

[0044] At the very beginning of each personnel list is only one non-virtual task:

[0045]

[0046] Each non-virtual task can only be handled by one person:

[0047]

[0048] Each task can have only one preceding and one succeeding task:

[0049]

[0050] For each task, a time window is defined, and the completion time of each task is obtained through conditional constraints, where M is the maximum value:

[0051] Lower bound of the task completion time variable:

[0052] e a The earliest start time for task a

[0053] After non-virtual task A is completed, task B begins:

[0054]

[0055] Non-virtual task b is first handled by person k:

[0056] IT k For time t

[0057] The earliest free time for person K.

[0058] Establishing constraints for the scheduling decision model includes establishing upper limits for these constraints, specifically:

[0059] If the preceding task of non-virtual task b is non-virtual task a, then we have

[0060]

[0061] Define a new binary variable σ b , and The relationship is as follows:

[0062]

[0063]

[0064]

[0065] If the preceding task of non-virtual task b is virtual task 0, and is executed by person k∈K, then:

[0066]

[0067] like Then σ b =1, otherwise σ b =0, that is:

[0068]

[0069]

[0070]

[0071] r a ∈{0,1} represents a non-virtual task a∈A t The relationship between whether it can be completed within 2 hours of becoming executable and the completion time is as follows:

[0072] If it is completed in two hours, force r a =1

[0073]

[0074] If completed within two hours, a mandatory order will be issued. a =0

[0075]

[0076] z a ∈{0,1} represents task a∈A t The relationship between whether there is a delay and the completion time is as follows:

[0077] If the deadline is delayed, a mandatory order will be issued. a =1

[0078]

[0079] If there is no delay, a mandatory order will be issued. a =0

[0080] LF a This represents the latest completion time for task a.

[0081] After constructing the scheduling decision model, the process also includes optimizing the scheduling decision model using the simulated annealing algorithm, specifically including:

[0082] By simulating the annealing process and iterating continuously, a new solution is generated each time and compared with the objective function of the current solution;

[0083] If the difference is less than 0, then accept the new solution;

[0084] If the difference is greater than 0, the probability of acceptance is represented by generating a random number r between 0 and 1;

[0085] If: e -(Δf / T)>r accepts the new solution; otherwise, it does not accept the new solution.

[0086] Ultimately, an optimal task scheduling scheme is obtained.

[0087] The hospital's dynamic information is input into the optimized scheduling decision model, which outputs the optimal scheduling scheme, including:

[0088] After the task scheduling is obtained at time t, the scheduling decision model calculates the task delay by calculating the difference between the completion time of each task and the latest specified completion time of the task.

[0089] It also calculates the number of tasks that are more than two hours late and assigns different weighting coefficients based on the task priority.

[0090] Calculate the workload of each person and add it to the current cumulative workload. Calculate the difference between the maximum and minimum workloads of different people to ultimately balance the workload of each person.

[0091] The objective function formula is as follows:

[0092]

[0093] Among them, A t ω represents the total number of tasks at time t. max A represents t The maximum weight value of the task. This indicates the number of tasks whose execution time exceeds 2 hours. C represents the weighted number of delayed tasks. max This represents the total time required to complete all tasks, TW k This represents the workload of the kth person, with α and β corresponding to the respective weights.

[0094] A dynamic scheduling system for medical delivery in a comprehensive hospital includes:

[0095] The data acquisition module is used to acquire dynamic information about the hospital.

[0096] The data processing module is used to calculate the optimal scheduling plan based on the hospital's dynamic information;

[0097] The data output module is used to output the optimal scheduling scheme.

[0098] This invention establishes a scheduling decision model, which is a multi-objective optimization mathematical model and solution algorithm for efficient scheduling decisions based on dynamically changing delivery tasks and personnel information. This enables scientific and efficient dynamic scheduling of medical delivery in comprehensive hospitals, solves the scheduling decision problem of medical delivery in comprehensive hospitals, reduces task delays, shortens delivery time, and balances the workload of workers. Attached Figure Description

[0099] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0100] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0101] Figure 1 This is a flowchart of a dynamic scheduling method for medical delivery in a comprehensive hospital proposed in this invention;

[0102] Figure 2 This is a diagram illustrating the architecture of a dynamic scheduling system for medical delivery in a comprehensive hospital, as proposed in this invention.

[0103] Figure 3 This is a simulation model diagram of a dynamic scheduling method for medical delivery in a comprehensive hospital proposed in this invention;

[0104] Figure 4 This is a comparison chart of experimental parameters for a dynamic scheduling method for medical delivery in a comprehensive hospital proposed in this invention.

[0105] Figure 5 This is an Insert strategy diagram for a dynamic scheduling method for medical delivery in a comprehensive hospital proposed in this invention.

[0106] Figure 6 Figure 1 shows the SWAP strategy of a dynamic scheduling method for medical delivery in a comprehensive hospital proposed in this invention;

[0107] Figure 7 Figure 2 shows the SWAP strategy of a dynamic scheduling method for medical delivery in a comprehensive hospital proposed in this invention. Detailed Implementation

[0108] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0109] It should be noted that all directional indications (such as up, down, left, right, front, back, etc.) in the embodiments of the present invention are only used to explain the relative positional relationship and movement of each component in a certain specific posture (as shown in the figure). If the specific posture changes, the directional indication will also change accordingly.

[0110] Furthermore, the use of terms such as "first" and "second" in this invention is for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, features defined with "first" and "second" may explicitly or implicitly include at least one of those features. Additionally, the technical solutions of the various embodiments can be combined with each other, but only on the basis of being achievable by those skilled in the art. When the combination of technical solutions is contradictory or impossible to implement, such a combination of technical solutions should be considered non-existent and not within the scope of protection claimed by this invention.

[0111] Example 1

[0112] A dynamic scheduling method for medical delivery in general hospitals includes:

[0113] S100, obtain dynamic information about the hospital;

[0114] S200, constructing a scheduling decision model;

[0115] S300 inputs dynamic information from the hospital into the optimized scheduling decision model, which then outputs the optimal scheduling scheme.

[0116] The hospital's dynamic information includes:

[0117] Current hospital task information: The set of tasks that need to be completed, the starting department of each task, the ending department of each task, the earliest delivery time, and the latest delivery time;

[0118] Current hospital personnel information: real-time location of personnel, and current cumulative workload of each personnel.

[0119] The S200 scheduling decision model includes:

[0120] S201, Define the decision variables for the scheduling decision model;

[0121] S202, Establish the objective function of the scheduling decision model;

[0122] S203, establish the constraints of the scheduling decision model.

[0123] S201 defines the decision variables for the scheduling decision model as including:

[0124] Indicates task a∈A t and task b∈A t Does {a} consist entirely of personnel k∈K, where K is the number of hospital staff? t Let be the set of hospital tasks at time t, and let a be the adjacent position before task b.

[0125] Introduce a virtual task 0 as the predecessor task of the first task and the successor task of the last task;

[0126] For any person k∈K, the endpoint of their virtual start task is f0 = ID. k ID k This indicates the department where worker k is currently located. The virtual end point and end point of the task are the end point of the last non-virtual task. That is, after completing the last task, the worker stays in place and waits for further instructions.

[0127] S202 establishes the objective function of the scheduling decision model, which includes:

[0128] Let r a ∈{0,1} represents a non-virtual task a∈A t The relationship between whether it can be completed within 2 hours of becoming executable and its completion time is as follows:

[0129]

[0130] Therefore, the number of tasks executed in more than 2 hours is:

[0131]

[0132] Let z a ∈{0,1} represents task a∈A t The relationship between whether there is a delay and the completion time is as follows:

[0133]

[0134] Therefore, the weighted average number of delayed tasks is:

[0135] W a Weighting based on the urgency of the task

[0136] The maximum time required to complete all tasks is

[0137] C a The completion time of task a

[0138] Maximum difference in workload between personnel

[0139]

[0140] Among them, TW k This represents the cumulative workload of personnel k∈K.

[0141] CW k S represents the current workload of worker k. a For task a, the starting department is F.b For the department at the end of task b, P b The processing time after sampling and delivery for task b. This indicates the time required to travel from the starting department of task a to the ending department of task b;

[0142] Finally, weighting coefficients are used to merge multiple objectives into a single overall objective.

[0143]

[0144] Where H = |A t |w max That is, the number of tasks multiplied by the maximum task weight.

[0145] S203 establishes the constraints of the scheduling decision model, including establishing the lower bounds of the constraints, specifically:

[0146] make The following constraints will be used to standardize personnel scheduling criteria and constraints:

[0147] At the very end of each personnel list is only one non-virtual task:

[0148]

[0149] At the very beginning of each personnel list is only one non-virtual task:

[0150]

[0151] Each non-virtual task can only be handled by one person:

[0152]

[0153] Each task can have only one preceding and one succeeding task:

[0154]

[0155] For each task, a time window is defined, and the completion time of each task is obtained through conditional constraints, where M is the maximum value:

[0156] Lower bound of the task completion time variable:

[0157] e a The earliest start time for task a

[0158] After non-virtual task A is completed, task B begins:

[0159]

[0160] Non-virtual task b is first handled by person k:

[0161] IT k For time t

[0162] The earliest free time for person K.

[0163] S203 establishes the constraints of the scheduling decision model, including establishing the upper limit of the constraints, specifically:

[0164] If the preceding task of non-virtual task b is non-virtual task a, then we have

[0165]

[0166] Define a new binary variable σ b , and The relationship is as follows:

[0167]

[0168]

[0169]

[0170] If the preceding task of non-virtual task b is virtual task 0, and is executed by person k∈K, then:

[0171]

[0172] like Then σ b =1, otherwise σ b =0, that is:

[0173]

[0174]

[0175]

[0176] r a ∈{0,1} represents a non-virtual task a∈A t The relationship between whether it can be completed within 2 hours of becoming executable and the completion time is as follows:

[0177] If it is completed in two hours, force r a =1

[0178]

[0179] If completed within two hours, a mandatory order will be issued. a =0

[0180]

[0181] z a ∈{0,1} represents task a∈A t The relationship between whether there is a delay and the completion time is as follows:

[0182] If the deadline is delayed, a mandatory order will be issued. a =1

[0183]

[0184] If there is no delay, a mandatory order will be issued. a =0

[0185] LF a This represents the latest completion time for task a.

[0186] After S200 constructs the scheduling decision model, S210 further optimizes the scheduling decision model using the simulated annealing algorithm, specifically including:

[0187] By simulating the annealing process and iterating continuously, a new solution is generated each time and compared with the objective function of the current solution;

[0188] If the difference is less than 0, then accept the new solution;

[0189] If the difference is greater than 0, the probability of acceptance is represented by generating a random number r between 0 and 1;

[0190] If: e -(Δf / T) >r accepts the new solution; otherwise, it does not accept the new solution.

[0191] Ultimately, an optimal task scheduling scheme is obtained.

[0192] The neighborhood generates new arrangement schemes (three schemes for generating new solutions):

[0193] (1) Insect strategy, such as Figure 5 :

[0194] Step 1: Randomly select one person from the task assignment;

[0195] Step 2: Randomly select another person and randomly choose any position in the task arrangement;

[0196] Step 3: Insert the extracted task into the selected position;

[0197] Step 4: Create a new task arrangement;

[0198] (2) SWAP Strategy 1: Randomly select a person's task and swap positions with the previous person's task, such as... Figure 6

[0199] Step 1: Randomly select one person;

[0200] Step 2: Randomly select a task from the personnel's task assignment;

[0201] Step 3: Swap the position of this task with the previous task;

[0202] Step 4: If there is no available slot for the task, draw a new task.

[0203] (3) SWAP Strategy 2: Randomly select two tasks for two people and swap their positions, such as... Figure 7

[0204] Step 1: Randomly select one person;

[0205] Step 2: Randomly select a task for the person;

[0206] Step 3: Randomly select another person;

[0207] Step 4: Randomly select a task for this person;

[0208] Step 5: Swap the positions of the two extracted tasks to generate a new arrangement scheme;

[0209] The S300 inputs dynamic information from the hospital into an optimized scheduling decision model, which outputs the optimal scheduling scheme, including:

[0210] After the task scheduling is obtained at time t, the scheduling decision model calculates the task delay by calculating the difference between the completion time of each task and the latest specified completion time of the task.

[0211] It also calculates the number of tasks that are more than two hours late and assigns different weighting coefficients based on the task priority.

[0212] Calculate the workload of each person and add it to the current cumulative workload. Calculate the difference between the maximum and minimum workloads of different people to ultimately balance the workload of each person.

[0213] The objective function formula is as follows:

[0214]

[0215] Among them, A t ω represents the total number of tasks at time t. max A represents t The maximum weight value of the task. This indicates the number of tasks whose execution time exceeds 2 hours. C represents the weighted number of delayed tasks. maxThis represents the total time required to complete all tasks, TW k This represents the workload of the kth person, with α and β corresponding to the respective weights.

[0216] Based on the MIP model, the project team wrote C# code to call the Gurobi solver to solve the model and verify the scheduling algorithm. A total of 9 simulation examples were tested (Gurobi solved the problem using a single core with a time limit of 100 seconds). The test results are compared between the two methods, as shown in the table below:

[0217] Table 1-1 Comparison of Test Results

[0218]

[0219] As can be seen, the scheduling algorithm closely approximates Gurobi's optimal solution. However, with a slightly larger problem size, Gurobi struggles to find the optimal solution within 100 seconds, while the scheduling algorithm is computationally very fast. When the problem size increases further (e.g., 4 workers with 20 tasks, or 5 workers with 25 tasks), the scheduling algorithm's results significantly outperform Gurobi's best result within the 100-second time limit. This demonstrates that the proposed scheduling algorithm exhibits excellent optimization performance while meeting the timeliness requirements of practical decision-making.

[0220] The scheduling optimization algorithm was embedded into a simulation model, and simulation experiments were set up to verify the optimization effect of the algorithm. The simulation experiment took the problem objective as the research premise and used four indicators as experimental factors: the number of tasks not completed within 2 hours (Obj1_C2hrs), the cumulative weighted sum of delayed task weights (Obj2_TWN), the time required to complete all tasks (Obj3_Cmax), and the difference in workload between the logistics personnel with the largest cumulative workload and the logistics personnel with the smallest cumulative workload (Obj4_G). The algorithm scheduling time period (actual worker scheduling rules, 1 minute / time, 2 minutes / time, 5 minutes / time, 10 minutes / time) was used as experimental variables. Through simulation, the changes of the four indicators under different factor levels were obtained, and comparative experimental analysis was carried out to obtain the optimal solution.

[0221] Based on the application scenarios of the model and the functional characteristics of the simulation software, we divide the main functions implemented by the basic model into the following two main modules: order information determination (delivery workers, etc.) and order delivery processing. According to the basic layout of the surveyed hospital and the proportion of departments and logistics personnel, a basic Plant Simulation model is established. A WorkerPool object simulates the main delivery station, two FootPath objects simulate the staircases between hospital floors, nine FootPath objects simulate the walkways on each floor, and two Source objects, one Station object, and five Buffer objects combine to simulate a hospital department. The final model is as follows: Figure 3 The image shown is a simulation model of a hospital (simulating three floors, with three departments on each floor, and a delivery team of four medical staff).

[0222] The simulation experiment takes the problem objective as the research premise and uses four major indicators as experimental factors: the number of tasks not completed within 2 hours (Obj1_C2hrs), the cumulative weighted sum of delayed task weights (Obj2_TWN), the time required to complete all tasks (Obj3_Cmax), and the difference in workload between the logistics personnel with the largest cumulative workload and the logistics personnel with the smallest cumulative workload (Obj4_G). The algorithm scheduling time cycle (actual worker scheduling rules, 1 minute / time, 2 minutes / time, 5 minutes / time, 10 minutes / time) is used as experimental variables. Through simulation, the changes of the four major indicators under different factor levels are obtained, and comparative experimental analysis is carried out to obtain the optimal solution.

[0223] like Figure 4 Analysis of the experimental results showed that when the scheduling period was 1 minute, the number of unfinished tasks within 2 hours was reduced by 22.58%, the cumulative weight of delayed tasks was reduced by 10.92%, the time required to complete all tasks was reduced by 6.7%, and the workload of workers was more balanced, making it the scheduling period that best fits the model's objective. Therefore, the algorithm parameters were adjusted based on the experimental results.

[0224] Simulation results show that the scheduling algorithm has a faster computation speed than the solver and its scheduling efficiency is also better than the simulation results of actual manual scheduling.

[0225] The present invention can effectively solve the problem of dynamic scheduling of medical delivery in general hospitals.

[0226] Example 2

[0227] A dynamic scheduling system for medical delivery in a comprehensive hospital includes:

[0228] The data acquisition module is used to acquire dynamic information about the hospital.

[0229] The data processing module is used to calculate the optimal scheduling plan based on the hospital's dynamic information;

[0230] The data output module is used to output the optimal scheduling scheme.

[0231] This invention establishes a scheduling decision model, which is a multi-objective optimization mathematical model and solution algorithm for efficient scheduling decisions based on dynamically changing delivery tasks and personnel information. This enables scientific and efficient dynamic scheduling of medical delivery in comprehensive hospitals, solves the scheduling decision problem of medical delivery in comprehensive hospitals, reduces task delays, shortens delivery time, and balances the workload of workers.

[0232] The above description is merely a specific embodiment of the present invention, enabling those skilled in the art to understand or implement the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features claimed herein.

Claims

1. A dynamic scheduling method for medical delivery in a comprehensive hospital, characterized in that, include: Obtain dynamic information about the hospital; Construct a scheduling decision model; Input the hospital's dynamic information into the optimized scheduling decision model. The scheduling decision model outputs the optimal scheduling scheme; The decision variables for defining the scheduling decision model include: Indicates task and tasks Are they all staff? deal with, For the number of hospital staff, Let be the set of hospital tasks at time t, and In the task Previous adjacent positions; Introduce a virtual task 0 as the predecessor task of the first task and the successor task of the last task; For any person The virtual start point of the task , This indicates the department where worker k is currently located. The start and end points of its virtual task are the end points of the last non-virtual task. That is, after completing the last task, it stays in place and waits for orders. The objective function for establishing the scheduling decision model includes: make Indicates non-virtual task The relationship between whether it can be completed within 2 hours of becoming executable and its completion time is as follows: ； Therefore, the number of tasks executed in more than 2 hours is: ； make Indicates task The relationship between whether there is a delay and the completion time is as follows: ； Therefore, the weighted average number of delayed tasks is: , Weights based on the urgency of the task; The maximum time required to complete all tasks is , The completion time of task a; Maximum difference in workload between personnel ； in Personnel Cumulative workload , For worker k's current workload, The starting department for task a, For the department at the end of task b, The processing time after sampling and delivery for task b. This indicates the time required to travel from the starting department of task a to the ending department of task b; Finally, weighting coefficients are used to merge multiple objectives into a single overall objective. ； in That is, the number of tasks multiplied by the maximum task weight.

2. The method for dynamic scheduling of medical delivery in a comprehensive hospital according to claim 1, characterized in that, The hospital's dynamic information includes: Current hospital task information: The set of tasks that need to be completed, the starting department of each task, the ending department of each task, the earliest delivery time, and the latest delivery time; Current hospital personnel information: real-time location of personnel, and current cumulative workload of each personnel.

3. The method for dynamic scheduling of medical delivery in a comprehensive hospital according to claim 1, characterized in that, The construction of the scheduling decision model includes: Define the decision variables for the scheduling decision model; Establish the objective function of the scheduling decision model; Establish the constraints of the scheduling decision model.

4. The method for dynamic scheduling of medical delivery in a comprehensive hospital according to claim 3, characterized in that, The constraints for establishing the scheduling decision model include establishing the lower bound of the constraints, specifically: make The following constraints are used to uniformly define the staffing standards and constraints: At the very end of each personnel list is only one non-virtual task: ； At the very beginning of each personnel list is only one non-virtual task: ； Each non-virtual task can only be handled by one person: ； Each task can have only one preceding and one succeeding task: ； For each task, a time window is defined, and the completion time of each task is obtained through conditional constraints, where M is the maximum value: Lower bound of the task completion time variable: , The earliest start time for task a; Non-virtual task After completion start: ； Non-virtual task By personnel First step: , Let t be the earliest idle time for worker k.

5. The method for dynamic scheduling of medical delivery in a comprehensive hospital according to claim 3, characterized in that, The constraints for establishing the scheduling decision model include establishing upper limits for the constraints, specifically: If the preceding task of non-virtual task b is non-virtual task a, then we have ； Define a new binary variable , and The relationship is as follows: ； ； ； If the preceding task of non-virtual task b is virtual task 0, and it is performed by personnel If executed, then: ； like ,but ,otherwise ,Right now: ； ； ； Indicates non-virtual task The relationship between whether it can be completed within 2 hours of becoming executable and the completion time is as follows: If completed within two hours, a mandatory order will be issued. : ； If completed within two hours, a mandatory order will be issued. : ； Indicates task The relationship between whether there is a delay and the completion time is as follows: If the deadline is delayed, an injunction will be issued. : ； If there is no delay, an injunction : , This represents the latest completion time for task a.

6. The method for dynamic scheduling of medical delivery in a comprehensive hospital according to claim 1, characterized in that, After constructing the scheduling decision model, the method further includes optimizing the scheduling decision model using a simulated annealing algorithm, specifically including: By simulating the annealing process and iterating continuously, a new solution is generated each time and compared with the objective function of the current solution; If the difference is less than 0, then accept the new solution; If the difference is greater than 0, the probability of acceptance is represented by generating a random number r between 0 and 1; if: Accept the new solution; otherwise, do not accept the new solution. Ultimately, an optimal task scheduling scheme is obtained.

7. The method for dynamic scheduling of medical delivery in a comprehensive hospital according to claim 1, characterized in that, The process of inputting dynamic information from the hospital into the optimized scheduling decision model, and the scheduling decision model outputting the optimal scheduling scheme, includes: After the task scheduling is obtained at time t, the scheduling decision model calculates the task delay by calculating the difference between the completion time of each task and the latest specified completion time of the task. It also calculates the number of tasks that are more than two hours late and assigns different weighting coefficients based on the task priority. Calculate the workload of each person and add it to the current cumulative workload. Calculate the difference between the maximum and minimum workloads of different people to ultimately balance the workload of each person. The objective function formula is as follows: ； in, This represents the total number of tasks at time t. express The maximum weight value of the task. This indicates the number of tasks whose execution time exceeds 2 hours. This indicates the weighted number of overdue tasks. This represents the total time required to complete all tasks. Indicates the first The workload of each person Corresponding weights.

8. A dynamic scheduling system for medical delivery in a comprehensive hospital, characterized in that, include: The data acquisition module is used to acquire dynamic information about the hospital. The data processing module is used to calculate the optimal scheduling plan based on the hospital's dynamic information; The data output module is used to output the optimal scheduling scheme; The scheduling decision model outputs the optimal scheduling scheme; The decision variables for defining the scheduling decision model include: Indicates task and tasks Are they all staff? deal with, For the number of hospital staff, Let be the set of hospital tasks at time t, and In the task Previous adjacent positions; Introduce a virtual task 0 as the predecessor task of the first task and the successor task of the last task; For any person The virtual start point of the task , This indicates the department where worker k is currently located. The start and end points of its virtual task are the end points of the last non-virtual task. That is, after completing the last task, it stays in place and waits for orders. The objective function for establishing the scheduling decision model includes: make Indicates non-virtual task The relationship between whether it can be completed within 2 hours of becoming executable and its completion time is as follows: ； Therefore, the number of tasks executed in more than 2 hours is: ； make Indicates task The relationship between whether there is a delay and the completion time is as follows: ； Therefore, the weighted average number of delayed tasks is: , Weights based on the urgency of the task; The maximum time required to complete all tasks is , The completion time of task a Maximum difference in workload between personnel ； in Personnel Cumulative workload , For worker k's current workload, The starting department for task a, For the department at the end of task b, The processing time after sampling and delivery for task b. This indicates the time required to travel from the starting department of task a to the ending department of task b; Finally, weighting coefficients are used to merge multiple objectives into a single overall objective. ； in That is, the number of tasks multiplied by the maximum task weight.

Images