In order to improve the accuracy and reliability of acoustic emission monitoring in bridge cable, wavelet transform was used to analyze guided wave modes, the corrosion evaluation method of steel wires based on guided wave modal characteristics was proposed. Firstly, the propagation of multi-mode guided waves in steel wires was simulated by the finite element method and wavelet transform was used to analyze modal characteristics of vibration signals. Finite element results were compared with theoretical time-frequency curves. Then, guided wave propagation experiments were carried out on steel wires with different corrosion degree, and the influence of corrosion on modal characteristics was analyzed. Finally, frequency distribution characteristics of guided wave modes were used to construct evaluation indexes for corrosion evaluation. The results show that there are different modes for vibration in different directions at the finite element node. Numerical results are in good agreement with theoretical values. Higher-order longitudinal guide waves have higher frequency and faster energy attenuation. With the increase of corrosion degree, the attenuation of high-frequency and high-order guided waves is gradually serious, and guided waves mode are gradually transferred to low-frequency and low-order. The evaluation index based on guided wave modal characteristics monotonically increases with the increase of corrosion degree. Ideally, corrosion damage with corrosion rate less than 7.6% can be effectively identified.It is concluded that wavelet transform can effectively extract guided wave modal features, and the evaluation index based on modal features can effectively identify corrosion damage