电动汽车充电平台智能语音识别技术及数据挖掘

Intelligent Voice Recognition Technology and Data Mining for Electric Vehicle Charging Platform

针对现有技术中电动汽车充电平台智能语音识别能力差的问题,设计了新型的电动汽车充电平台,该系统平台包括计算机网络终端、电网调度中心以及充电桩等,能够实现上层管理中心的语音识别,电路包括语音采集模块、语音辨别模块和控制驱动模块等,设计出基于UniSpeech-SDA80D51芯片的语音识别电路,提高了语音识别能力,并构建出隐马尔可夫模型(hidden Markov model,HMM)和人工神经元网络(artificial neural network,ANN)相融合的模型,实现了智能语音识别数据信息的挖掘与处理,进而增强了语音识别系统的性能。试验表明,该研究在不同噪音下的识别率,其中在20 dB的噪音下识别率为88.3%。该方法提高了语音识别和挖掘能力。

Aiming at the problem of poor intelligent voice recognition capabilities of electric vehicle charging platforms in the prior art,a new type of electric vehicle charging platform was designed. The system platform includes computer network terminals,power grid dispatching centers and charging piles,etc.,which can realize the voice recognition of the upper management center,Realize the recognition of various charging data information by constructing a voice recognition hardware circuit,which includes a voice acquisition module,a voice recognition module and a control drive module,etc. The voice recognition circuit based on the Uni Speech-SDA80 D51 chip is designed to improve the voice recognition capability. And build a hidden Markov model(HMM) and artificial neural network(ANN) fusion model,realize the mining and processing of intelligent speech recognition data information,and then enhance the performance of the speech recognition system. Experiments show that the recognition rate of this study under different noises,of which the recognition rate is 88.3% under 20 decibels of noise. This method improves speech recognition and mining capabilities.