Research And Application Of Appointment Scheduling Method Considering Service Duration Uncertainty

Appointment scheduling is a work that is often carried out in the practice of social production and life.Compared with traditional resource scheduling problem,the appointment scheduling decides material,manpower and other resource allocation plans by analyzing known reservation requirements to guide the actual service process.And then inform the reservation party to come in the order which they are scheduled in advance.However,service duration uncertainty often occurs in the actual process,so it is necessary to make adjustment in time.Thus,designing the scheduling method under this uncertainty can be an effective decision for the problem.This paper studies the method of solving the appointment scheduling problem under service duration uncertainty and applies it to a related practical business field.Considering the existing limitation of knowledge management work about scheduling model and method,the characteristics and process of the problem are analyzed in detail based on the knowledge-unit model.And inspired by the thoughts of data-driven,the frameworks of the model and method to solve the appointment scheduling problem are constructed.On this basis,this paper improves the traditional multi-objective evolutionary algorithm based on decomposition and constructs multiple different scenarios to tackle the uncertainty of the problem,thus enhancing the quality and robustness of the decision.Furthermore,in order to solve the high computational cost of a large number of scenario simulation processes,a surrogate model is introduced and trained in the evolution process to directly assist the evaluation of the fitness value.As a result,the efficiency and effect of the solving algorithm can be improved.After that,considering the important usage of the appointment scheduling in the field of medical service management,a practical study on the appointment scheduling of the MRI equipment in hospital is carried out.Aiming at the duration uncertainty of the examination,we will optimize the MRI operation cost and patients' waiting time by setting out a reasonable service plan for the examination order and the start time of different types of appointments.Finally,the real data of MRI problem is used to test the proposed method.The results are compared with two classical multi-objective evolution algorithms and the scientificity and validity of the designed method are evaluated.At the same time,from the perspective of model management,a unified knowledge representation of the MRI equipment scheduling is provided,which helps to store,manage,reuse,and expand it in the decision support system.It is finally proved that the model and method proposed in this paper can not only provide decision support method for the management of medical examination equipment scheduling,but also give theoretical support for applications in other field of appointment scheduling and knowledge management.