Research On The Problem Of Operation Scheduling Considering The Uncertainty Of Operation Duration And Recovery Time

Finding appropriate scheduling methods which have the ability in effective mamagement of medical resources performs significant roles in enhancing the competitiveness of medical institution and satisfaction of patients.But conventional scheduling methods have some disadvantages such as long time consuming,low efficiency and lack of the ability in handling uncertainty factors in scheduling process.These defects lead to thereduction of practicability for schedules even the risk of out of schedules.In this thesis,a scheduling method which has better practicability is proposed so that hospital administrators can obtain schedules which conform to expectations in a short time.Due to the interactions among sections in the process of operation scheduling,the resource distribution and accommodation of upstream and downstream of operating room also has great impact on scheduling effect.In this thesis,post-operative recovery secitonsis added to the structure of schedule and research on scheduling methods which consider both duration and recovery time is performed.The uncertainty of duration and recovery time is also taken into consideration and different methods used to deal with the uncertainty,and the effectiveness of these methods in handling uncertainty is analyzed.Above all,the research problem of this thesis is the operation scheduling methods which take the uncertainty of duration and recovery time into consideration.Based on the literature review among the related papers,the main research work is summaried as follows:(1)Taking the problem of operation scheduling as the research objectivewith considering the restriction of operating room and recovery bed,by synthetically considering the influence on schedule effect of duration and recovery time,an integer programming model is built with the objective of minimizing the total cost of operating rooms and recovery beds.By aiming at deepening understanding of model and its solution process,complexity analysis is conducted and ILOG CPLEX is applied to solve the model.At last the historical scheduling data is used to verify the feasibility and solving scale of the established model.The results show that the feasibility of the model and the ability of obtaining operation schedules which meet the actual operation scale requirements in a reasonable running time.(2)Based on integer programming,an expectation stochastic programming model which has the objective of minimizing the total cost is built under the consideration of uncertainty of duration and recovery time.After analyzing the complexity of the model,transformation method is used to reduce its complexity by transforming stochastic programming model to deterministic ones.Transformation method firstly obtains mean value and variance which represent digital feature in accordance with the analyzing of historical data.Secondly it finds distribution function which represents distribution law based on related literature.Monte Carlo simulation is used to generate the expectation for replacing random variable which is used to reduce the complexity of model.By contrasting objective values of integer programming model and stochastic programming model which are solved by ILOG CPLEX,the robustness of the stochastic programming model is verified;the contrasting results also show that considering some uncertainty factors instead of only one is helpful for improving the ability of resistance the variation of external environment.(3)Integrating the above methods for handling uncertainty factors,interval numbers is used to represent the uncertainty of duration and recovery time and a two-stage robust optimization model is established.Two-stage robust optimization is hard to be solved.So the two-stage model is transformed into normal ones by using linearization,slack and dual to reduce the difficulty in solving.The experiments verify the robustness of robust optimization model by contrasting the objective valuesbetweenthe robust optimization model and stochastic programming model.At the same time,the influence of conservative parameter on objective value is analyzed by changing conservative parameter value and observes the change of objective values.The results show that conservative parameters and objective value have the link of positive correlation and the conservative parameter which represents the degree of deviation between actual recovery time and its lower bound has greater influence on theobjective value.