With the extensive application of computer technology in all areas of society, the computer has become indispensable to people's lives, an important part. Computer voice digital signal is more due to its simple, direct, and easy to be accepted into many areas of the characteristics of human society. In the popularization of computer technology, the popularity of the trend, subject to a number of computer-aided voice processing software for real-time signal processing is based on an important development direction of generalpurpose computer signal processing simulation system. This article focuses on the genetic algorithm to improve computer assisted voice digital signal processing technology, focusing on the effect of genetic algorithms in terms of speaker recognition, proposed a set of related technologies using genetic algorithm to improve computer-aided optimization of voice processing solutions. One can easily LBG algorithm for speech processing plays an important role in vector quantization techniques used in the design process codebook into local optimum problem, will produce genetic algorithm (GA) and its combination of GA-LBG algorithm;Second, for poorly performing computer-aided speech processing problems play a very important role in neural network RBF network obtained in the clustering process, combined with improved adaptive genetic algorithm to optimize the design of the network training algorithms. Examples of the computer-aided speaker recognition by voice processing applications through improved genetic algorithm LBG algorithm and RBF neural networks were trained and identified.Experimental results show a good effect on genetic algorithm optimized generated.