ABSTRACT
The article describes autoregressive one-day average and maximum methane concentration forecast models developed in Silesian University of Technology. In both presented autoregresive models, ex-post forecasts are performed based on linear equations with parameters estimated for each day of the week. As descriptive variable was used preceding day average methane concentration. As input data for research were used continuous methane measurement results and daily exploitation volumes from the one of Silesian coal mine longwalls and covers 152 days. Analyzed case study was carried out to check the accuracy of the forecasts for “Y” ventilated longwall. Such forecasts could be applied as a supporting tool for selection and adjustment of short-term methane prevention measures during exploitation.
