基于超限学习机与涡流搜索算法的锅炉燃烧优化策略

Boiler combustion optimization strategy based on extreme learning machine and eddy current search algorithm

设计一种基于超限学习机(Extreme Learning Machine,ELM)算法的涡流搜索控制策略,调用锅炉燃烧控制物联网探头系统中的12个高清红外单通道探头数据,实现对每秒597.1968MB大宗数据的深度卷积超限学习挖掘,对锅炉中的涡流状态给出[0,1]区间上的评价值。通过将该系统投并到锅炉控制系统中,将该系统反馈值控制到0.200以下作为控制目标之一,使锅炉总功率较投并前提升1.73%,平均煤耗节约1.66%。

An eddy current search control strategy based on extreme learning machine(ELM)algorithm is designed.The data of 12 high-definition infrared single channel probes in the internet of things probe system for boiler combustion control are called to realize the deep convolution over limit learning mining of 597.1968 MB bulk data per second,and the evaluation value in[0,1]interval of eddy current state in the boiler is given.By putting the system into the boiler control system,the feedback value of the system is controlled below 0.200 as one of the control objectives.The total power of the boiler is increased by 1.73%and the average coal consumption is saved by 1.66%.