Optimal Power Systems Planning for IEEE-14 Bus Test System Application

As renewable energy sources become more prevalent in the grid, grid operators face new issues in maintaining system reliability in the face of changing conditions while also maximizing renewable energy usage. The need for system robustness increases as more renewable energy, such as wind power, is used in the system. In this paper, short-term power system planning and control performance is being investigated; dynamic programming (DP) as a robust optimization technique is being applied to decrease power system operational expenses. The proposed methodology is being tested on an IEEE 14-bus test system. Gated recurrent unit (GRU) and hybrid gated recurrent unit with long short term memory (GRU/LSTM) are two machine learning algorithms being utilized to forecast the performance of short-term wind generation and load demand. The prediction results show that GRU/LSTM outperforms GRU with a mean square error of 0.045 and 0.043 for load and wind prediction, respectively, to achieve the plan of UC with minimum production costs of 508933.8$ for the day ahead.