基于物理模型的声音合成理论研究

Theoretical Research on Sound Synthesis Technique Based on Physical Model

基于物理模型的声音合成技术是一种可以精确描述物体发声的方法 ,其实质是以物体的振动PDE方程为基础来直接合成声音 ,该方法已逐步成为目前声音合成技术领域内的研究热点 .本文从理论上和试验上证明了一个被研究者广为引用的一维基础PDE方程存在的局限性 ,利用力学理论推导出了新的振动PDE方程 ,并从理论上和试验上证明了其正确性 ,从而提出了更具普遍意义的新一维声音合成基础方程 ,克服了传统方程的缺陷 .

Sound synthesis technique based on physical model is a kind of accurate method to describe object's sounds. The method was partial differential object's vibration equation to synthesize sound directly. We find that the PDE cited by many researchers for a long time has some limitations, and prove this by theory and experimentation. And then, we derive a new PDE by dynamics elements. We prove that our new PDE is correct. A basic and universal equation has been proposed by the paper for sound synthesis based on physical model.