关键词: 供应风险, 装配系统, 订购决策, 最优策略

Abstract: For the complex components ordering problem of an assembler with two types of supply uncertainties (component 1 has uncertain production capacity and component 2 has random yield rate), using stochastic nonlinear programming, we present a multidimensional optimization model in which the object is to maximize the assembler’s expected profit. Under deterministic demand, the optimal ordering policyis characterized and sensitivity analysis is provided. We show that theassembler’s expected profit function is concave in a specific zone, in which the optimal ordering amounts are located. The ratio of two components’ optimal ordering amounts happens to be a constant, which depends on component 2’s random yield distribution and component 1’s production cost. We prove that improving component 1’s production capacity (stochastically) generally leads to increase in both components’ optimal ordering amounts. However, surprisingly, improving component 2’s random yield may lead to non-monotone changes in the components’ optimal ordering amounts. Numerical experiments are conducted to justify the model and to further explore the effect of supply uncertainties. Valuable management insights are generated to help the assembler in making ordering decisions and managing supply risks.

Key words: supply uncertainties, assemblysystem, ordering decision, optimal policy