At present, the most widely used control algorithm in the field of Active Noise Control (ANC) is the classic FxLMS algorithm and its improved methods. This type of algorithm has the characteristics of simple structure, low computational complexity, easy implementation, and good stability. However, when applied to large space and large range noise control such as the cockpit of turboprop aircraft, as the number of channels in the ANC system increases, the computational complexity of the algorithm will rapidly expand, The real-time performance of the algorithm is difficult to meet, and the noise reduction effect of the system is greatly compromised or even ineffective. The Sequential Partial Update FxLMS (SPU-FxLMS) algorithm effectively solves this problem, but its convergence performance is weaker than the FxLMS algorithm. This article makes improvements to the problem of slow convergence of the SPU-FxLMS algorithm, allowing it to converge at a faster speed in the initial stage and continue to run with low computational complexity after converging to a stable state. Theoretical derivation and simulation analysis were conducted on the algorithm, and the simulation results showed that the method not only significantly reduces computational complexity, but also has good noise reduction performance and robustness.