以信息熵理论为基础,针对马氏体相变过程声发射信号,利用基于频域分析的小波包能谱熵,得到相变过程声发射信号频率空间的能量分布信息。提取出不同类型信号的能谱熵特征值,并利用基于时域分析的吉布斯概率熵,以能谱熵特征值作为幅度谱的子区间划分参数,实现马氏体相变过程声发射信号吉布斯概率熵变化过程的准确描述。分析表明,这些信息熵特征值可以准确评价马氏相变过程的声发射信号变化规律。 Based on the information entropy theory, according to the acoustic emission signal of the martensitic transformation process, the energy distribution of the acoustic emission signal in the phase transition process is obtained by using the spectral envelope entropy of the wavelet packet based on the frequency domain analysis. The energy entropy eigenvalues ​​of different types of signals are extracted, and Gibbs probabilistic entropy based on time-domain analysis is used as the sub-interval dividing parameter of amplitude spectrum to realize the acoustic emission signal of martensitic transformation Accurate Description of Gibbs’ Probability Entropy Change Process. The analysis shows that these information entropy eigenvalues ​​can accurately evaluate the variation law of acoustic emission signal during the process of phase transition.