Code underlying: Flood control of reservoir systems: Learning-based explicit and switched model predictive control approaches

Python codes to implement explicit and switched MPC using data-driven surrogate models.The python files starting with PDMPC are for generating PDMPC results to train surrogate models.O_results_check and W_results_check files are for arranging results from the explicit MPC surrogate model and switched MPC surrogate model, respectively.W_ML.py is to build and test the switched MPC surrogate model, and O_DNN_hyper_opt.py is to find the optimal hyperparameters for the explicit MPC surrogate model as well as to train it.The datasets for this research are included.

用于实现基于数据驱动代理模型（surrogate models）的显式模型预测控制（explicit Model Predictive Control, MPC）与切换式模型预测控制的Python代码。以PDMPC开头的Python文件用于生成训练代理模型所需的PDMPC结果。O_results_check与W_results_check文件分别用于整理显式MPC代理模型与切换式MPC代理模型的输出结果。W_ML.py用于构建并测试切换式MPC代理模型；O_DNN_hyper_opt.py则用于寻优显式MPC代理模型的最优超参数并完成模型训练。本研究使用的数据集已包含在内。