CN-121789937-B - Rehabilitation resource collaborative scheduling method and system considering multidimensional space-time constraint

CN121789937BCN 121789937 BCN121789937 BCN 121789937BCN-121789937-B

Abstract

The invention discloses a rehabilitation resource collaborative scheduling method and system considering multidimensional space-time constraint, and relates to the technical field of medical informatization management. The method comprises the steps of responding to a scheduling request and obtaining basic data, carrying out standardized preprocessing on the data and calculating doctor-patient matching degree, constructing a mixed integer linear programming mathematical model comprising oxygenation link treatment time sequence continuity constraint, resource group sharing capacity constraint and personnel skill matching constraint, setting up a multi-objective optimization function comprising treatment amount, matching degree and load balancing, and calling a branch-and-bound algorithm solving model to generate a global optimal scheduling scheme and an abnormality analysis report. The invention effectively solves the pain points of difficult continuous guarantee, multiple conflict of shared resources, low personalized matching degree and the like of long-term projects in rehabilitation scheduling by a mathematical programming technology, and remarkably improves the resource utilization rate and the operation efficiency of a rehabilitation center.

Inventors

WANG BIN
HOU YUANYUAN
MAO FENGYU
ZHAO JIANGTAO
WANG QIANG
MENG PINGPING
ZHANG YONGXIANG
HONG FENG
YU YANWEI

Assignees

中国海洋大学

Dates

Publication Date: 20260512
Application Date: 20260306

Claims (7)

1. A rehabilitation resource collaborative scheduling method considering multidimensional space-time constraint is characterized by comprising the following steps: acquiring rehabilitation scheduling request data, and acquiring associated clinical diagnosis and treatment data and resource allocation data based on the request data; performing data cleaning and preprocessing based on the acquired data; Constructing a scheduling optimization mechanism based on the preprocessed data, and constructing a scheduling decision objective function based on the scheduling optimization mechanism; generating a scheduling variable solution of a scheduling decision objective function by utilizing the improved branch delimitation; generating a rehabilitation scheduling plan according to the scheduling variable solution; the data cleaning and preprocessing based on the acquired data comprises constructing a mathematical model input set containing patient requirements, treatment item attributes, resource capacity matrix and doctor-patient matching relation based on the acquired data, and loading a data association input table into a scheduling optimization mechanism configured at a server side, wherein a patient set is generated through a first data construction unit And a treatment project collection , Represents any one patient in the patient set, For any one treatment in the treatment item set, For patients Generating a set of time slices by a second data construction unit , Setting the time length of each time slice as a fixed value for any standard period in the time slice set, and generating a resource set by a third data construction unit And a set of physical therapists , Is common equipment resource Generating a set of oxygenation time blocks for any one of the physical therapists by a fourth data construction unit , For any one preset fixed time block of oxygenation treatment, each time block Comprising a series of successive time periods Calculating the matching degree matrix of the doctor-patient skill by a fifth data construction unit Generating a score based on the matching results of the patient diagnosis keywords and the physical therapist skill priority list; the method for constructing the scheduling optimization mechanism based on the preprocessed data comprises the steps of generating an allocation variable set for scheduling variable control through an allocation variable mechanism based on a mathematical model input set obtained after preprocessing, wherein a period occupation variable is generated through a first allocation unit Wherein For determining patients Whether or not in a time period Proceeding with the project Generating block selection variables by a second allocation unit Wherein For determining patients Whether or not to select the first Generating personnel assigned variables by a third dispensing unit Wherein For determining patients During a time period Receive treatment items Whether or not to be treated by therapists Providing service, generating load deviation variable by fourth distribution unit Wherein For indicating time periods A bias value for resource loading exceeding an average level; The preprocessing-based data construction scheduling optimization mechanism further comprises a constraint control mechanism for constructing a condition constraint set containing the oxygenation continuous block treatment time sequence continuity constraint based on an allocation variable set to constraint the scheduling variable, wherein the constraint control mechanism comprises the steps of carrying out oxygenation continuous block treatment time sequence continuity constraint through a first constraint unit, and specifically comprises the steps of selecting a uniqueness constraint by a construction block: Limiting the daily selection of at most one oxygenation time block for each patient, and constructing a time period-block linkage constraint: For constraining if a certain time block is selected, all continuous time periods covered by the block must be occupied synchronously and can not be interrupted, constructing a cabin capacity constraint: For restraining each time block while the number of persons treated does not exceed the maximum volume of the bunk Performing resource group sharing capacity constraint through a second constraint unit, wherein the resource group sharing capacity constraint is used for constraining the uninterruptible project set needing to share the total capacity The total concurrency number in the same time period is expressed as the following formula: wherein An upper limit of the maximum number of concurrent people allowed for the resource group in a single time period; personnel-project binding constraints, including task-personnel consistency constraints, by a third constraint unit For determining the items requiring manual intervention if the scheduling decision is to be performed Then one therapist must be and can only be assigned, as well as skill qualification constraints For determining the assigned therapist The device comprises items Performing patient single-period mutual exclusion constraint by a fourth constraint unit for determining that the same patient is in the same time slice The maximum treatment is carried out in the formula of The medical advice compliance constraint is carried out by the fifth constraint unit, and is used for determining that only the medical advice recommendation list of the patient is arranged And each non-oxygenated item is performed at most once a day.
2. The method for collaborative scheduling of rehabilitation resources with consideration of multidimensional space-time constraints according to claim 1, wherein the constructing a scheduling decision objective function based on a scheduling optimization mechanism includes constructing sub-objective functions including total therapy output, matching quality and load smoothing based on a scheduling optimization mechanism, forming a scheduling decision objective function by weighted combination, and scheduling decision objective function Is Maximize of Wherein, the method comprises the steps of, For the total number of treatment projects completed all day, the formula is used for maximizing treatment output ; The sum of the degree of matching scores of the skills of the traditional Chinese medicine for all the arranged physical treatment projects is used for maximizing the personalized treatment quality, and the formula is that ; For each time period, the resource load and the average load For minimizing load fluctuations to achieve balanced scheduling, by ; Respectively corresponding weight coefficients.
3. The method for collaborative scheduling of rehabilitation resources taking into account multidimensional space-time constraints according to claim 2 is characterized in that the method for generating a scheduling variable solution of a scheduling decision objective function by utilizing improved branch-and-bound comprises the steps of solving the scheduling variable solution of the scheduling decision objective function by utilizing an improved branch-and-bound algorithm, firstly configuring solver parameters including maximum solving time and relative error tolerance, carrying out iterative search on an MILP (minimum likelihood solution) scheduling optimization model based on the improved branch-and-bound algorithm, determining upper and lower boundaries of the solution by utilizing a relaxed linear programming, triggering a conflict resolution mechanism if no feasible solution exists under preset constraint conditions, sequentially relaxing non-critical constraint conditions or temporarily rejecting scheduling requirements of low-priority patients according to preset priority strategies, and recording conflict constraint sets which lead to no solution.
4. The method of claim 3, wherein the iterative search is performed on an MILP scheduling optimization model based on an improved branch-and-bound algorithm, the upper and lower bounds of the solution are determined by using a relaxed linear programming, the method comprises solving a conventional branch-and-bound algorithm which is easy to fall into dimension disasters when processing large-scale binary variables based on a constructed MILP model by adopting the improved branch-and-bound algorithm which fuses heuristic preprocessing and a specific domain pruning strategy, wherein, in order to accelerate search convergence, an initial feasible solution is generated by adopting a multi-level priority rule greedy algorithm As an initial lower bound of the branch-and-bound tree, it is defined that the degree of matching according to the doctor-patient relationship Arranging in descending order, filling the oxygen-added connection block requirement into a plurality of preset candidate continuous time blocks, assigning priority to the projects of the specific scarce equipment or the specific qualification therapist, ensuring that the key resources are not occupied by the common project redundancy, filling greedy to the common project, and utilizing Filling the remaining idle period, expressed as if the greedy algorithm finds a feasible solution Then set global lower bound Otherwise Then based on heuristic variable selection and branching strategy of service weight, selecting node to be expanded in search tree When the branch decision variables are selected, a scoring function is constructed by discarding conventional random selection, adopting a branch strategy based on service sensitivity and adopting a variable selection operator based on pseudo cost Quantitative evaluation of candidate variables: , Wherein, the And Respectively as variables Average unity gain produced when the current node branches down and up; Is a balance coefficient; preferential selection of the service by algorithm for the weight of service relevance The maximized variables are split to force the model to collapse toward the solution space with the highest medical value gain.
5. The method for collaborative scheduling of rehabilitation resources taking into account multi-dimensional space-time constraints according to claim 4, wherein the iterative search is performed on an MILP scheduling optimization model based on an improved branch-and-bound algorithm, the upper and lower bounds of the solution are determined by using a relaxed linear programming, and the method further comprises linear relaxation and upper bound estimation, node-pair Linear relaxation is performed, and the local upper bound of the node is calculated by using a simplex method Then pruning is carried out, if Namely, the optimal potential of the current node is not as good as the known feasible solution, the node and all the child nodes thereof are directly cut off, then logic decision pruning aiming at a rehabilitation scene is introduced, invalid branches are removed in advance before the solution, and a logic conflict detection operator is defined in the solution process Aiming at the oxygenation link block constraint, constructing a logic judgment pruning judgment formula: , Wherein, the In order to indicate the function, Representing nodes The current definition field of the lower variable, if Judging that the consistency semantics of the current local solution space and the oxygen-added connection block are conflicted, according to the logic implication principle, all the sub-topology spaces of the node do not contain feasible solutions, immediately executing forced pruning Where by evaluating shared resource groups The feasibility of the current search subspace is prejudged, and a resource conflict detection operator is constructed The mathematical form of (a) is defined as follows: , Wherein, the In order to indicate the function, Representing decision variables At the node The lower bound of the value.
6. The method for collaborative scheduling of rehabilitation resources with consideration of multidimensional space-time constraints according to claim 5, wherein triggering a conflict resolution mechanism if there is no feasible solution under preset constraint conditions includes converting an original MILP model into lagrangian relaxation form, finding a suboptimal feasible solution, introducing an elastic variable when a search space is closed and there is no solution Rewriting an objective function: , Wherein, the To identify the set of core conflict constraints by the irreducible inconsistent subsystem IIS detection algorithm, And in the iterative process, if the scheduling decision objective function value of a new candidate rank Cheng Jie is better than the scheduling decision objective function value of the current optimal solution, updating the new candidate rank Cheng Jie as the current optimal scheduling variable solution, otherwise, utilizing a loose linear programming to determine the upper and lower bounds of the solution, cutting off the inferior branches and avoiding invalid searching.
7. A rehabilitation resource collaborative scheduling system considering multidimensional space-time constraints, which performs a rehabilitation resource collaborative scheduling method considering multidimensional space-time constraints as claimed in claim 1, comprising: The data acquisition module is configured to acquire rehabilitation scheduling request data and acquire associated clinical diagnosis and treatment data and resource configuration data based on the request data; the preprocessing module is configured to perform data cleaning and preprocessing based on the acquired data; a decision module configured to construct a scheduling optimization mechanism based on the preprocessed data, and to construct a scheduling decision objective function based on the scheduling optimization mechanism; A calculation module configured to generate a scheduling variable solution of a scheduling decision objective function using the modified branch-and-bound; The generation module is configured to generate a rehabilitation schedule according to the schedule variable solution.

Description

Rehabilitation resource collaborative scheduling method and system considering multidimensional space-time constraint Technical Field The invention relates to the technical field of resource collaborative scheduling, in particular to a rehabilitation resource collaborative scheduling method and system considering multidimensional space-time constraint. Background With the rapid development of rehabilitation medicine and the increasing diversification of patient rehabilitation demands, the diagnosis receiving amount and treatment project category of the rehabilitation center show explosive growth. In this context, rehabilitation programs are becoming increasingly complex and challenging as a key element in linking patient needs with medical resources. Rehabilitation scheduling is essentially a typical multi-constraint, multi-objective, strongly coupled resource scheduling optimization problem, and has the core difficulty of cross coupling of three dimensions of treatment requirement isomerism, resource constraint pluripotency and individuality matching necessity. In the prior art, the traditional rehabilitation scheduling mode mainly relies on manual experience or a simple greedy algorithm based on a 'first come first serve' rule. The manual scheduling is low in efficiency, is difficult to deal with large-scale concurrency demands, and is extremely easy to cause resource conflict or scheduling violation due to human negligence. Although the speed of the simple greedy algorithm is improved, the greedy algorithm lacks a global field of view, is difficult to process the continuous requirements of long-term continuous block projects such as oxygenation and the like, cannot effectively cope with complex constraints of resource sharing groups such as musculoskeletal rehabilitation and the like, cannot consider multiple targets such as treatment amount, doctor-patient matching degree and load balancing, often causes the coexistence of high-value equipment idling and key resource congestion, and severely restricts the service efficiency and treatment quality of a rehabilitation center. Therefore, a rehabilitation scheduling method capable of comprehensively considering time continuity, resource sharing and personalized matching requirements and realizing all-weather, multi-resource and multi-objective global optimization is needed. Disclosure of Invention In order to solve the above-mentioned problems, the present invention provides a rehabilitation resource collaborative scheduling method and system considering multidimensional space-time constraints. Through various medical data of the deep integration rehabilitation center, an MILP mathematical model comprising oxygenation link block treatment time sequence continuity constraint, resource group sharing capacity constraint and personnel skill matching constraint is constructed, a multi-objective optimization function comprising treatment amount, matching degree and load balancing is established, and global optimal solution is realized by utilizing a branch-and-bound algorithm, so that the utilization efficiency of rehabilitation resources is remarkably improved, the scheduled medical compliance is guaranteed, and the patient hospitalizing experience is optimized. In a first aspect, the present invention provides a rehabilitation resource collaborative scheduling method considering multidimensional space-time constraint, which adopts the following technical scheme: a rehabilitation resource collaborative scheduling method considering multidimensional space-time constraint comprises the following steps: acquiring rehabilitation scheduling request data, and acquiring associated clinical diagnosis and treatment data and resource allocation data based on the request data; performing data cleaning and preprocessing based on the acquired data; Constructing a scheduling optimization mechanism based on the preprocessed data, and constructing a scheduling decision objective function based on the scheduling optimization mechanism; generating a scheduling variable solution of a scheduling decision objective function by utilizing the improved branch delimitation; And generating a rehabilitation scheduling plan according to the scheduling variable solution. Further, the data cleaning and preprocessing based on the acquired data comprises constructing a mathematical model input set containing patient requirements, treatment item attributes, resource capacity matrix and doctor-patient matching relation based on the acquired data, and loading a data association input table into a scheduling optimization mechanism configured at a server side, wherein the first data construction unit generates a patient setAnd a treatment project collection,Represents any one patient in the patient set,For any one treatment in the treatment item set,For patientsGenerating a set of time slices by a second data construction unit,Setting the time length of each time slice as a fixed value for any stand