A method and system for predicting the number of spare parts based on the combination of multiple models

Abstract

The invention discloses a method and system for predicting the quantity of spare parts based on a combination of multiple models. The method includes: constructing a database of the historical use quantity of spare parts, selecting a training set, constructing time series features for each training sample, and training the GPR model, GMR model and RBFN model, according to the sample prediction deviation, the optimal model label calibration is performed on the training samples, the GMM model training is performed on the calibrated data sets, and the timing characteristics of the samples to be tested are input into different GMM models to obtain three probabilities Value, compare the probability value to select the optimal model label, input the timing characteristics of the sample to be tested into the corresponding optimal model for retraining, and use the retrained optimal model to predict the number of samples to be tested in the next month. The invention improves data robustness through time series features, and proposes a spare parts prediction method based on a GMR model and an RBFN model, which is simple to implement, and can effectively improve prediction accuracy by selecting an optimal prediction model from multiple models.

Description

technical field [0001] The invention belongs to the field of artificial intelligence, and more specifically, relates to a method and system for predicting the quantity of spare parts based on multi-model combination. Background technique [0002] Spare parts are spare parts of the equipment. The lack of reasonable planning for the purchase and storage of spare parts may cause great troubles and even losses to the business and development of the enterprise. Too many spare parts reserves will occupy a large amount of working capital and storage warehouse space of the enterprise, which is not conducive to the long-term development of the enterprise; too few spare parts reserves will cause maintenance or production operations to be unable to be completed in time, resulting in economic losses for the enterprise. [0003] The existing spare parts quantity prediction methods mainly include two types, one is based on statistical models such as ARMA, and the other is based on machine...

AI Technical Summary

Problems solved by technology

However, these prediction methods have some defects: (1) The usage patterns of different spare parts are not similar, for example, some parts are prone to periodic failures while some parts basically do not fail, so the selected model is not suitable for all parts ; (2) The choice of prediction model depends on human experience, rather than the rules reflected by the spare parts sample itself; (3) The process may be adjusted during the production process of spare parts, and the source may be adjusted during the procurement process of spare parts. The adjustment will cause the spare parts to be different before and after, and the old model cannot be applied to the new spare parts

Images