基于PSO-PF算法的SVM识别方法 及其在异常声音中的应用

针对异常声音识别率低和算法复杂度高等技术难题，提出了一种基于粒子群优化粒子滤波（PSO-PF）算法优化支持向量机（SVM）的识别方法.将PSO算法引入粒子滤波中，通过不断更新粒子速度和位置，使粒子群向高似然后验概率区域移动，提高粒子滤波的参数估计精度.将PSO-PF算法应用于SVM参数优化中，可解决现有SVM参数优化算法易陷入局部最优值等问题.实验结果表明，将所提方法应用于多类异常声音识别，能够有效提高识别率，降低算法复杂度.

Recognition and Application of Abnormal Sound Via SVM Based on PSO-PF

In order to solve the problems of low recognition accuracy and high computation complexity in abnormal sound signals, a particle filter based on particle swarm optimization (PSO-PF) algorithm is proposed to optimize the parameters of support vector machine (SVM). To improve the estimation precision of particle filter, the particle swarm optimization is applied to drive all the particles to the regions in which their likelihoods are high, by updating the velocity and position of particles constantly. And the algorithm can avoid falling into local optimum in SVM parameter optimization. The experimental results show that the new algorithm can achieve higher recognition accuracy and lower computation complexity for abnormal sounds recognition.