<p>Transient violent dc-link voltage fluctuations present a significant challenge for electrolytic capacitor-less (ECL) drive systems under dynamic operation conditions. In stark contrast to conventional single-variable control paradigms, which suffer from a trade-off between response speed and voltage fluctuation suppression capability, this paper presents a novel power tracking model predictive current control (PTMPCC) strategy that employs a multi-dimensional dynamic coupling model to overcome this challenge. The proposed approach establishes a theoretical framework beginning with the construction of a virtual error function for dc-link voltage regulation. A Lyapunov-based virtual control function is subsequently developed to establish the relationship among motor power variations, dc-link voltage error, and grid current. By integrating this virtual control function into a reconstructed cost function in place of conventional current controllers, the proposed strategy enables rapid and precise tracking of motor power through grid power adjustment, particularly under dynamic conditions. The stability of the proposed method is mathematically guaranteed through the incorporation of the Lyapunov function, where the dc-link voltage, which is treated as the control objective of the stability function, satisfies the Lyapunov stability criteria. This theoretical foundation ensures swift tracking of the dc-link voltage reference values and robust performance during sudden load changes. Furthermore, a predictive current model for dual decoupling inductors is developed to achieve high-precision complementary control of the asymmetric split-capacitor voltages. The experimental results demonstrate that the proposed control strategy effectively minimizes dc-link voltage ripple and ensures rapid voltage stabilization under dynamic conditions, achieving over a 30% improvement in both respond time and fluctuation suppression. These results validate its practical effectiveness in ECL drive system applications.</p>