Intelligent Shipyard Inventory Non-Surplus Inventory Control Algorithm and Empirical Research

Abstract

There is still a big gap between China and other advanced shipbuilding countries, such as digital fields, networking and intellectualization. The inventory management leads to more inventories, growing cost, slower construction progress, and so on. Therefore, the idea of "no surplus" inventory management was put forward, which was established on scientific and effective prediction of the stock allowance of shipbuilding materials. Then, the appropriate algorithms were studied to serve the idea above. This study considered two-stage predictor construction. First stage 3 single predictors were constructed, including ARIMA model, improved PSO (particle swarm optimization)-SVR (support vector machine) and Wavelet Neural Network. Then the second stage constructed a strong predictor formed by the integrated algorithm. The effectiveness of the algorithm is verified by the contrast test. Then an empirical case was used to show how a non-surplus inventory scheme came to be. At last, the relevant countermeasures for the above scheme were proposed.