Inventory Optimization Research For Critical Spare Parts Based On Risk-measure Criterion

With integrated product design further developed,critical spare parts in complicated industrial system have bigger influence on the system's operation cost and utilization.Due to the high-value,high backorder cost,high demand uncertainty characteristics of critical spare parts,its inventory decision requires more delicate tools.Risk-neutral and simple assumptions for demand distributions no longer apply,and risk-averse inventory strategy with simulation for stochastic demand can be one of the answers for the challenges met in critical spare parts inventory management.The research uses conditional value-at-risk as the risk criterion to be involved in objective function,and establishes the risk-averse inventory model which minimizes the CVaR for inventory cost loss.The research gives the closed-form solution to the risk-averse model,and analyzes the characters of optimal solution under a high backorder cost,high demand uncertainty scenario.The analysis includes exploring the influence of backorder cost and risk confidence on optimal solution,as well as the solution's sensitivity to heavy-tail demand distribution.The research also considers numerical solution when in practice the demand distribution function is not available.The research transforms the original model into a linear optimization problem,which can be solved by using demand sample.Based on reliability theory,the research obtains a demand estimation formula and generates demand samples.Using the numerical model and generated demand sample,the research is able to conduct a series of numerical experiments,regarding the character of loss distribution under risk-averse strategy,the decreasing of loss uncertainty and mean of large losses by risk-averse strategy.