为改善控制器模块与在线建模模块之间的残余信号相互影响，以及当次级路径突然发生剧烈变化时，自适应滤波器的过度更新问题，提出了一种基于范数方法的在线建模有源噪声控制算法，该算法通过监测残余信号平均能量判断收敛性，并根据收敛情况动态调节建模步长。此外，为自适应滤波器系数设定阈值，当自适应滤波器的范数超出阈值时，采用范数方法限制滤波器过度更新。分别以单频、复合频率和带限白噪声作为噪声源，并与已有在线建模有源噪声控制算法相比，当次级路径信息发生突变时，所提算法能够有效降低建模误差，并提高降噪量，达到稳态时，建模误差分别是为-43 dB、-38 dB和-39 dB，降噪量分别可达到32 dB、35 dB和25 dB。仿真结果表明，所提算法能够在保证有源噪声控制系统稳定性的同时提高建模误差的收敛速度，在稳态时取得良好的降噪效果。

To mitigate the mutual influence of residual signals between the controller module and the online modeling module， and to address the issue of excessive updates of the adaptive filter when the secondary path undergoes abrupt changes. A norm-based online modeling active noise control algorithm is proposed， which assesses convergence by monitoring the average energy of the residual signal and dynamically adjusts the modeling step size accordingly. Additionally， a threshold is set for the norm of the adaptive filter coefficients， when the norm of the adaptive filter exceeds this threshold， a norm-constrained method is employed to limit excessive updates of the filter， ensuring stability and preventing over-adjustment. Compared to existing online modeling active noise control algorithms， the proposed algorithm effectively reduces modeling error and improves noise reduction performance. At steady state， the modeling errors for single frequency， composite frequency， and band-limited white noise are -43 dB， -38 dB， and -39 dB， respectively， wachieving noise reduction leve of 32 dB， 35 dB， and 25 dB. The single frequency， composite frequency， and narrowband white noise are used as the primary noise sources in the simulation. When there is sudden large change in the secondary path information， results show the proposed algorithm can maintain the stability of the active noise control system， enhance the convergence speed of the modeling error， and achieve superior noise reduction performance at steady state.