Multi-Objective Optimization of Back-to-Back Starting for Pumped Storage Plants under Low Water Head Conditions Based on the Refined Model

Abstract

Pumped storage plants (PSP) must switch frequently between various working conditions. Moreover, PSPs can fall easily into an S-shaped zone under low water head conditions, especially during back-to-back starting (BTBS), which reduces the stability and safety of unit operation. In this paper, a nonlinear PSP model for BTBS is established by combining an electrical subsystem with a refined hydraulic-mechanical subsystem. The influences of the hydraulic, mechanical, and electrical factors on the BTBS process are investigated quantitatively. Choosing the speed overshoot and speed stable time as the optimization objectives, and considering a variety of constraints, the multi-objective particle swarm optimization (MOPSO) algorithm is introduced to study and optimize two typical startup strategies. The results show that: (1) The parameters of a hydraulic-mechanical-electrical system have a significant impact on BTBS process, and the most unfavorable working condition corresponds to the lowest water head; (2) In the control schemes, a novel constant excitation voltage strategy is proposed based on the multi-objective optimization scheme. Compared with the constant excitation current strategy or single-objective, the optimization strategy proposed can considerably improve the speed overshoot and the speed stable time by at least 68.27% and 3.22% under the worst working conditions. (3) It is further verified that the problem of trapping in the S-shaped region under various working conditions may be avoided by the obtained optimal control scheme. The results give prominence to the effectiveness of the proposed optimization strategy for maintaining the safety and stabilization of PSP operation.