供应链二级可靠性评价的BP神经网络应用研究

Application Research of BP Neural Network to Two Level Reliability Evaluation Model of Supply Chain

以制造业供应链为研究对象,基于不确定性提出二级供应链可靠性评价模型,通过算例仿真验证模型可靠性诊断结果的合理性,最后提出针对结果的反向改进方案。二级模型是在一级模型的基础上,针对具有不确定性因素的一级评价指标,提出决定其性能的二级评价指标的优化模型,具有更高精度和更好的反向改进能力。BP神经网络作为一种成熟的工具具有强大的学习能力,经过大量历史数据的学习能够很好应用在诊断领域,应用算例得出结果与专家系统评价结果相符。

Evaluation model for reliability of two level supply chain is taking manufacturing supply chain as the research object. It’s validated the rationality of the reliability of diagnosis model by simulating an example, and present ability of reverse improvement solution. Two level model, base on the one level model, focus on uncertainty factors evaluation index. The secondary optimization model of evaluation index with determining performance, higher precision and better ability of reverse improvement. BP neural network as a mature tool has strong ability in learning. Through studying massive historical data which well applied in field of diagnosis. The results of calculation of BP neural network was highly similar with expert system.