Data files: all data were collected from Korea Data portal (www.data.go.kr) and K-water (www.kwater.or.kr)

 - DC_training_original.xlsx: reservoir inflow and upstream water level data collected from Data.go.kr
 - DC_training_wavelet.xlsx: reservoir inflow data transformed by a wavelet filter
   * Details of wavelet filter can be found in our previous research, 
     titled:Comparison of scenario reduction approaches for reservoir inflow scenarios generated by a Bayesian Neural Network
     (https://data.4tu.nl/datasets/e343331b-496f-40ab-83eb-f546df6dffa6)
 - wl_down_events.xlsx: downstream water level data collected from Data.go.kr


Python files

  Scenario generation and reduction;
    - Input_scenario.py: Senario generation by a MC dropout BNN model and reduction by K-means and so on.
      * Details of scenario generation and reduction can be found in our previous research,
        titled:Comparison of scenario reduction approaches for reservoir inflow scenarios generated by a Bayesian Neural Network
        (https://data.4tu.nl/datasets/e343331b-496f-40ab-83eb-f546df6dffa6)

  Stochastic MPC experiment
    - SMPC_formulation_CVaR.py: MPC formulation with CVaR-type soft constraints
    - SMPC_CVaR.py: Numerical experment of stochastic MPC with CVaR constraints
