面向光纤法布里-珀罗传感器的腔长解调，提出一种结合幅值归一化与最小均方差的联合解调算法.采用幅值归一化实现返回光谱信号与模拟光谱信号的等幅，提高解调精度.仿真结果表明，相较于传统最小方差法，该算法的解调精度显著提升.实验结果表明，采用该算法可实现优于0.72 nm的高分辨率.

A combined algorithm of amplitude normalization and minimum mean square error for the cavity length interrogation of fiber-optic Fabry-Perot sensors is proposed. An amplitude normalization method is introduced to realize the equiamplitude of both the reflected and the simulated spectral signals, which helps improving the interrogating accuracy. Significant increasing of interrogating accuracy compared with conventional algorithms is verified through numerical simulations. And, a 0.72 nm resolution of cavity length interrogation is achieved in real experiment.