CN-122000001-A - Intelligent hospital personnel scheduling system and method based on knowledge graph

CN122000001ACN 122000001 ACN122000001 ACN 122000001ACN-122000001-A

Abstract

The invention discloses a hospital personnel intelligent scheduling system and method based on a knowledge graph, which mainly comprises the following steps of constructing the knowledge graph of hospital personnel, tasks, environments and equipment entities; initializing a Bayesian guide continuous learning framework based on a knowledge graph, introducing GRAPHSAGE algorithm to optimize the learning mechanism of the Bayesian guide framework, updating the knowledge graph through incremental learning to optimize the embedded representation of the nodes, evaluating the task priority in real time to generate a preliminary scheduling scheme, and adjusting the scheduling scheme according to actual execution feedback. According to the invention, through optimizing the learning and updating processes of the knowledge graph, the accuracy and adaptability of the scheduling of hospital personnel are improved, the defects of task priority adjustment, personnel and environment adaptation and the like in the traditional scheduling method are effectively overcome, and the method has higher practical application value.

Inventors

GUO RUIYUAN
RONG CHAO
LI DABAO
XU XINYU
Niu Chunyun

Assignees

泽瑞科技集团有限公司

Dates

Publication Date: 20260508
Application Date: 20260126

Claims (10)

1. The intelligent hospital personnel scheduling system and method based on the knowledge graph are characterized by comprising the following steps: S1, constructing a knowledge graph containing hospital personnel, tasks, environments and equipment entities; s2, initializing a Bayesian guiding continuous learning framework based on the knowledge graph; S3, introducing an extensible embedding propagation algorithm to improve a learning mechanism in the Bayesian guiding continuous learning framework, and enhancing the propagation capability of new knowledge in the learning step of the Bayesian reasoning framework by introducing the embedding propagation mechanism to obtain an optimized Bayesian guiding continuous learning framework; S4, updating a knowledge graph through an incremental learning and reasoning mechanism based on the optimized Bayesian guiding continuous learning framework, embedding knowledge of newly added tasks, personnel and environmental changes into the graph, optimizing embedded representation of nodes in the graph, and obtaining an updated knowledge graph; s5, evaluating the priority of each task in real time according to the updated knowledge graph, and generating a preliminary scheduling scheme; and S6, after the scheduling scheme is generated, based on actual execution feedback, adjusting the scheduling scheme through a Bayesian inference mechanism to obtain the scheduling scheme after adjustment and optimization.
2. The intelligent scheduling system and method for hospital personnel based on a knowledge graph of claim 1, wherein the S1 specifically comprises: s1.1, acquiring personnel information data of a hospital from a hospital management system, wherein the personnel information data is used as personnel data in entity nodes; s1.2, task information data is obtained from a hospital task management system and used as task data in entity nodes; S1.3, acquiring real-time environment data from an environment monitoring system as environment data in an entity node; S1.4, acquiring equipment use information from a hospital equipment management system as equipment data in an entity node; And S1.5, encoding the relation among the entities according to the information of hospital personnel, tasks, environments and equipment and a preset relation model to form edges among nodes, thereby obtaining the complete knowledge graph of the hospital personnel, tasks, environments and equipment.
3. The intelligent scheduling system and method for hospital personnel based on a knowledge graph of claim 1, wherein the step S2 specifically comprises: s2.1, initializing a Bayesian guiding continuous learning framework based on the constructed knowledge graph, and setting initial parameters for the framework; s2.2, according to a preset incremental learning mechanism, taking a knowledge map as initial input, and performing initial learning through a Bayesian reasoning mechanism to generate an initial reasoning model; and S2.3, distributing initial weights for each entity node and each relation edge in the Bayesian guiding continuous learning framework, and setting reasoning rules according to the structure and the data characteristics of the atlas to obtain the Bayesian guiding continuous learning framework.
4. The intelligent scheduling system and method for hospital personnel based on a knowledge graph of claim 1, wherein the step S3 specifically comprises: S3.1, based on an initial reasoning model in the Bayesian guiding continuous learning framework, selecting GRAPHSAGE algorithm as an extensible embedding propagation algorithm, applying the algorithm to a learning step of the Bayesian reasoning framework, and optimizing an information propagation process by sampling neighbor nodes and aggregating the characteristics of the neighbor nodes; S3.2, optimizing each entity node and relation edge in the Bayesian framework by using GRAPHSAGE algorithm, updating embedded representation of the nodes by aggregating feature vectors of the neighbor nodes, and selecting a fixed number of neighbor nodes for each node by using GRAPHSAGE sampling mechanism; Aggregating the feature vectors of the selected neighbor nodes, and aggregating the feature vectors of the neighbor nodes by adopting a multi-granularity aggregation mechanism to generate an updated embedded representation of each node; Updating the embedded representation of the current node through the aggregated feature vector; S3.3, adjusting an inference mechanism in the Bayesian inference mechanism based on the updated node embedded representation; And S3.4, dynamically adjusting the relation weight among the nodes through an optimized Bayesian inference mechanism, and adjusting the learning rate according to the embedded representation of each node to obtain an optimized Bayesian guiding continuous learning framework.
5. The intelligent scheduling system and method for hospital personnel based on a knowledge graph of claim 4, wherein the GRAPHSAGE algorithm specifically comprises: For each entity node, selecting K neighbor nodes from all neighbor nodes of the entity node according to the ascending order of node numbers by a GRAPHSAGE neighbor sampling mechanism, wherein K is a preset positive integer; aggregating the feature vectors of the selected neighbor nodes to obtain the aggregated feature vectors of the neighbor nodes through an aggregation method; splicing the feature vectors of the aggregated neighbor nodes with the feature vectors of the current entity nodes to form extended feature vectors; inputting the extended feature vector to a group of full connection layers, performing linear mapping, and generating an updated embedded representation of the entity node; And carrying out embedded updating on the relation edge, and taking the element-by-element average value of the updated embedded representation of the nodes at the two ends of the relation edge as the embedded representation of the relation edge.
6. The intelligent scheduling system and method for hospital personnel based on knowledge graph as claimed in claim 4, wherein the multi-granularity aggregation mechanism specifically comprises: each node divides the characteristics of the neighbor nodes into a plurality of granularity levels according to a preset granularity strategy, and each granularity level contains different numbers of neighbor nodes and reflects neighbor information in different ranges; In each granularity level, selecting neighbor nodes through a sampling mechanism, aggregating the characteristics of the nodes, integrating the characteristics of the neighbor nodes through weighted average, and generating node representation of the granularity level; and generating node characteristic representation of each granularity level by performing independent aggregation operation on the neighbor nodes of the level.
7. The intelligent scheduling system and method for hospital personnel based on a knowledge graph of claim 1, wherein the step S4 specifically comprises: S4.1, optimizing the embedded representation of each node through a Bayesian inference mechanism based on the optimized Bayesian guiding continuous learning framework, reasoning the embedded representation of each node through the inference mechanism, adjusting the relation weight among the nodes, and updating the characteristics of each node in the atlas; S4.3, updating embedded representations of all nodes in the map, and optimizing the characteristic representation of each node through an incremental learning algorithm; and S4.4, dynamically adjusting the relation between the nodes by using an inference rule based on the updated node embedded representation to obtain an updated knowledge graph.
8. The intelligent scheduling system and method for hospital personnel based on a knowledge graph of claim 1, wherein the step S5 specifically comprises: S5.1, evaluating the priority of each task based on the updated knowledge graph, calculating the priority value of each task according to the emergency degree, the resource requirement, the personnel skill, the equipment availability and the environmental change of the task through a preset priority calculation rule, and distributing a priority score for each task; s5.2, sorting the tasks according to the priority score of each task from high to low to generate a priority list of the tasks; and S5.3, according to the priority list of the tasks, combining the availability, skills and health states of hospital personnel, performing personnel matching, and generating a preliminary scheduling scheme meeting actual conditions.
9. The intelligent scheduling system and method for hospital personnel based on a knowledge graph of claim 1, wherein the step S6 specifically comprises: S6.1, acquiring relevant data of the execution state of a task, the working efficiency of personnel, the running state of equipment and the environmental change in the actual scheduling execution process by real-time execution feedback, and inputting the relevant data as feedback information into a Bayesian inference mechanism; s6.2, based on the feedback information, reasoning the completion condition of each task by using a Bayesian reasoning mechanism, and adjusting the priority of the task and personnel allocation; S6.3, reassigning tasks according to actual task execution conditions, working capacity of personnel, availability of equipment and environmental conditions, and optimizing a scheduling plan; And S6.4, based on the updated task allocation and personnel arrangement, adjusting the learning rate in the Bayesian inference mechanism to obtain a final adjustment optimized scheduling scheme.
10. The intelligent scheduling system and method for hospital personnel based on the knowledge graph as claimed in claim 1, wherein the intelligent scheduling system and method comprise the following modules: the knowledge graph construction module is used for constructing a knowledge graph comprising hospital personnel, tasks, environments and equipment entities; The Bayesian guide continuous learning framework initialization module is used for initializing the Bayesian guide continuous learning framework based on the knowledge graph and setting initial parameters for the framework; the embedded propagation mechanism optimization module is used for introducing an extensible embedded propagation algorithm and improving a learning mechanism in the Bayesian guided continuous learning framework; the knowledge map updating module is used for updating the knowledge map through an incremental learning and reasoning mechanism based on the optimized Bayesian guiding continuous learning framework, embedding the knowledge of the newly added tasks, personnel and environmental changes into the map, and optimizing the embedded representation of the nodes in the map; the scheduling scheme generating module is used for evaluating the priority of each task in real time according to the updated knowledge graph and generating a preliminary scheduling scheme by combining task, personnel and resource information; And the scheduling scheme adjusting module is used for adjusting the scheduling scheme through a Bayesian inference mechanism based on actual execution feedback after the scheduling scheme is generated, so as to obtain the scheduling scheme after adjustment and optimization.

Description

Intelligent hospital personnel scheduling system and method based on knowledge graph Technical Field The invention relates to the technical field of intelligent scheduling of hospital personnel, in particular to an intelligent scheduling system and method for hospital personnel based on a knowledge graph. Background The intelligent scheduling system for the hospital staff plays an important role in optimizing allocation of hospital staff resources, improving working efficiency and improving service quality. Along with the continuous improvement of the requirements of the medical industry on personnel scheduling efficiency and flexibility, the traditional scheduling mode gradually exposes a plurality of defects, and particularly under the conditions of frequent task priority change, personnel skill and health status fluctuation, equipment availability change and the like, the traditional scheduling method is difficult to effectively cope with the dynamic changes, so that the resource allocation is unreasonable, and the operation efficiency of a hospital is influenced. Currently, intelligent scheduling systems for hospital staff mainly perform task allocation based on a rule engine or a simple optimization algorithm. Traditional rule engine methods rely on static rules and preset constraints, while capable of handling some simple scheduling tasks, have poor scheduling efficiency and adaptability to complex hospital environments, especially when faced with personnel shortages, bursty tasks or special needs. In addition, the rule-based system cannot dynamically adjust the scheduling scheme, resulting in a scheduling result that cannot respond in time to real-time changes. In recent years, a data-driven intelligent scheduling method is attracting attention. By utilizing machine learning, reinforcement learning and other technologies, some new scheduling systems can optimize resource allocation to a certain extent, automatically generate a scheduling scheme and adjust emergency conditions to a certain extent. However, most of the existing intelligent shift scheduling methods have the problems of insufficient utilization of knowledge patterns, lack of comprehensive consideration of multi-source data and the like, so that the performance and the adaptability of the methods are limited. Many existing systems rely only on static information of personnel and tasks, neglecting the influence of dynamic information such as task changes, personnel health conditions, environmental conditions and the like, and making it difficult for a scheduling system to adjust in real time and make accurate decisions. In the existing method, the knowledge graph is used as a powerful data structure, can effectively represent complex relations among personnel, tasks, environments and equipment in a hospital, and provides richer and dynamic data support. However, most of the current systems lack sufficient reasoning capability when processing knowledge graphs, cannot make full use of the rich information in the graphs to perform effective reasoning and decision, and particularly cannot make rapid response and optimization according to the information in the graphs when facing sudden demands. Therefore, how to fully utilize the knowledge graph and the advanced algorithm technology and improve the dynamic adaptability and the real-time decision making capability of the intelligent scheduling system of the hospital staff becomes a key problem to be solved urgently in the current field. Therefore, how to provide a system and a method for intelligent scheduling of hospital personnel based on a knowledge graph is a problem to be solved by the person skilled in the art. Disclosure of Invention The invention aims to provide a hospital personnel intelligent scheduling system and method based on a knowledge graph, and provides a Bayesian inference framework which is optimized by constructing a dynamically updated knowledge graph and combining GRAPHSAGE algorithm and a multi-granularity aggregation mechanism to realize more accurate task priority assessment and personnel resource scheduling aiming at the problem that a hospital personnel scheduling system in the prior art cannot effectively process dynamic task change, personnel health state fluctuation and equipment availability change. The invention has the technical advantages of being capable of updating the scheduling scheme in real time, dynamically coping with sudden demands, improving the resource utilization efficiency and optimizing the operation of a hospital, and has stronger practical application value. According to the embodiment of the invention, the intelligent scheduling system and method for hospital personnel based on the knowledge graph comprise the following steps: S1, constructing a knowledge graph containing hospital personnel, tasks, environments and equipment entities; s2, initializing a Bayesian guiding continuous learning framework based on the knowledge graph; S3, introduci