在低延时矢量励线性预测 (LD -CELP)编码算法的基础上 ,提出直接矢量量化的低延迟矢量激励线性预测编码方案 (DVQ -LD -CELP) ,采用误差反向传播神经网络 (BPNN)来实现DVQ -LD -CELP方法中的码本搜索算法 ,实现在保持原编码方法 (LD -CELP)生成语音的音质和码率的同时 ,降低码本搜索的复杂度 ,提高码本搜索算法的效率 Based on low-delay vector linear prediction (LD-CELP) coding algorithm, direct vector quantization (LDQVLD) linear low latency vector predictive coding scheme (DVQ-LD-CELP) is proposed. Error backpropagation neural network (BPNN) To realize the codebook search algorithm in the DVQ-LD-CELP method, so as to reduce the complexity of the codebook search and improve the performance of the codebook search algorithm while maintaining the sound quality and the code rate of the original code method (LD-CELP) effectiveness