Dynamic programming with successive approximation and relaxation strategy for long-term joint power generation scheduling of large-scale hydropower station group

The joint optimal operation of large-scale hydropower station group (LHSG) is faced with the higher dimension than that of cascade hydropower station, the demand for the efficient optimization techniques of the above problem is urgent. Integrating the characteristics of problem into optimization techniques is an effective way. Therefore, based on some previous research results, the approximate concavity and monotonicity characteristics of power generation utility function of dynamic programming with successive approximation (DPSA) in each stage is analyzed. Then, an improved DPSA with relaxation strategy (named DPSARS) based on the above mathematical derivations is proposed to solve the long-term joint power generation scheduling (LJPGS) of LHSG. Compared with DPSA, the time complexity exhibits quadratic increase with the number of discrete states, while DPSARS only exhibits linear increase. Then, in order to further test the convergence accuracy and efficiency of the proposed DPSARS, the model of the LJPGS problem of LHSG, composed of 61 hydropower stations in the upper reaches of the Yangtze River, is established. The experimental results show that DPSARS represents its competitive performance in solving the LJPGS problem of LHSG compared with other methods. (c) 2021 Elsevier Ltd. All rights reserved.

Automated summary

The research analyzed the characteristics of the power generation utility function of dynamic programming with successive approximation (DPSA) and proposed an improved strategy named DPSARS based on mathematical derivations to solve the long-term joint power generation scheduling problem of LHSG. Experimental results demonstrate that DPSARS shows competitive performance in solving the long-term joint power generation scheduling problem of LHSG compared to other methods.