地铁备品备件的需求量预测有利于备件储备决策质量的提高，但是大量少样本备件的需求量预测不能采用常规的统计方法，经典贝叶斯预测模型有效利用先验信息进行预测，但是需要较多的数据，操作复杂，结果误差值较大。文章基于样本数据，通过实验的方式调整贝叶斯预测模型参数的计算方法，降低了数据要求，简化了贝叶斯模型操作的复杂程度，也提高了预测的精确度，在一定程度上提高了地铁备件管理的效率。%Metro spare parts demand forecast is conducive to the improvement of the quality of spare parts reserves decision, but the general statistics method can't be used in the demand forecast of large less sample spare parts, prior information is used effectively in classic Bayesian forecast model to forecast, but much data is needed, the operation is complex, and the errors value of the results is large. Based on the sample data, this paper adjusted the calculating method of Bayesian forecast model parameters by experiment, reduced the data requirements, simplified the operating complexity of Bayesian model, improved the accuracy of the forecasts, and to a certain extent improved the efficiency of metro spare parts management.
展开▼
