绿色能源互补智能电厂云控制系统研究

Based on the theory of cloud control system, an intelligent power plant cloud control system (IPPCCS) is designed to overcome problems of complex objects, multi-sources heterogenous data, "information island" and the poor ability of overall optimization scheduling in modern electric power enterprise. To solve problems of strong fluctuation and poor disturbance resistance of green power generation, a machine learning method is used to obtain the short-term prediction value of wind and solar power based on their history data. Then in the cloud, the economic model predictive control (EMPC) algorithm is applied to provide the power predictive scheduling strategy of water turbines by real-time rolling optimization, to ensure the robustness of green energy complementary power generation, consume wind and solar power fully and reduce the frequency of starting/stopping and crossing the vibration zones of the turbines, which both provides clear and stable energy support for the users and protects the devices. The simulations show the effectiveness of the proposed method in an example of regional cloud data center.