心理声学模型是感知编码的核心部分,它是一个模拟人耳听觉的数学模型,其效果好坏直接影响音频编码的质量。为了降低当前使用的心理声学模型的运算复杂度,提高其运算精度,文中采用综合归纳和规范分析的方法,扼要阐述了常用心理声学模型的类型,详细分析了现有模型的不足,并提出了改进的方法,使其更具科学性和实用价值。 Psychoacoustic model is the core part of perceptual coding. It is a mathematical model that simulates the auditory sense of the human ear, and its effect is directly affecting the quality of audio coding. In order to reduce the computational complexity of the psychoacoustic model currently used and to improve its computational accuracy, this paper briefly summarizes the types of commonly used psychoacoustic models by comprehensive induction and normative analysis, and analyzes the shortcomings of the existing models in detail. Improved method to make it more scientific and practical value.