中文摘要：

针对模糊神经网络训练采用BP算法比较依赖于网络的初始条件,训练时间较长,容易陷入局部极值的缺点,利用粒子群优化算法（PSO）的全局搜索性能,将PSO用于模糊神经网络的训练过程.由于基本PSO算法存在一定的早熟收敛问题,引入一种自适应动态改变惯性因子的PSO算法,使算法具有较强的全局搜索能力.将此算法训练的模糊神经网络应用于语音识别中,结果表明,与BP算法相比,粒子群优化的模糊神经网络具有较高的收敛速度和识别率.

英文摘要：

The paper proposes fuzzy neural network trained by particle swarm optimization （PSO） algorithm which has global search characteristic,in order to overcome shortage of traditional BP algorithm which replies on initial conditions,has the longer training time,and is easy to be trapped into the local extremum in fuzzy neural network.The paper introduces the inertia factor that changed adaptively and dynamically because of premature convergence of the basic PSO algorithm,which make the algorithm have stronger global search ability. Fuzzy neural network trained by this algorithm is applied to speech recognition system, and the experimental result proves that PSO algorithm has better recognition results and convergence speed than BP algorithm in the fuzzy neural network training.