Warehouse stock;Time series;Expert system;Fuzzy logic;Analysis;Optimization;Prediction

Actually, a lot of companies have tried to optimize their systems of warehouse stock management to minimize the production costs. The main goalis evident ? not to spend too much money for stock. To predict the behaviour of the system, there are usually used the methods of time series analysis. They are able to determine the main trend very well, seasonal influences, etc. But they donot take into account the internal and external influences acting on the system. For their destination there is appropriate to use the expert knowledge from companies that are often vague. As an appropriate tool, therefore appears to be a fuzzy expert system. Its use, however, causes problems if the system exhibits a significant trend. It turns out that the system for determining the main trend is appropriate to use methods known from the time series analysis and then followedby taking advantage of the expert system. This paper presents a fuzzy expert system that combines expert knowledge with the analysis of the trend ofthe system. The presented expert system was also verified in a practical application.