Optimization of Batch Centralized Health Examination Scheduling Considering Inter-Item Transition Time and Sex...

With increasing health awareness among Chinese residents, the demand for health examinations has continued to grow, placing mounting pressure on hospitals and health examination centers. This problem is particularly evident in group batch health examinations, where a large instantaneous flow of examinees and numerous examination items can easily lead to disorder in the examination process. This study aimed to optimize scheduling for this type of health examination, improve examination efficiency, and shorten total examination time.

A health examination scheduling optimization model was developed with the objective of minimizing the maximum examination duration. The model incorporated path-dependent transition time between examination items and sex-related differences, and a genetic algorithm was designed to obtain scheduling solutions. For special group examination populations with precedence constraints among examination items, the model was modified, and a controlled experiment was designed to quantitatively evaluate the effects of introducing constraints on algorithm efficiency and optimization performance. Finally, experimental analyses were conducted, including model verification using small-scale benchmark examples, case analysis, AnyLogic simulation, and sensitivity analysis, to validate the effectiveness of the model and algorithm and the stability of the solutions.

Introducing constraints had only a minor effect on algorithm efficiency, although it reduced the optimization effect. Model verification showed that the differences in examination duration and transfer time were both 0, and the scheduling results satisfied the inter-item constraints, indicating that the model correctly implemented the predefined scheduling logic. In the case analysis, the optimized scheduling scheme yielded a maximum examination duration of 197.06 min, while the simulation result was 195.57 min, with a difference of only 1.49 min. Compared with the pre-optimization duration of 220 min, the simulation result represented an 11.10% reduction (24.43 min/220 min), confirming the effectiveness of the model and algorithm. Sensitivity analysis showed that, when service durations varied within ±5% and ±10%, most item-duration changes except ultrasound could be absorbed by the system buffer and had limited influence on the total examination duration, demonstrating a certain degree of stability in the model and algorithm solutions.

The proposed mathematical programming model and genetic algorithm can significantly improve scheduling efficiency for batch centralized health examinations, substantially shorten overall examination duration, and have strong practical application value.