摘要
为了准确地建立汽轮机热耗率预测模型,提出了一种基于反向学习自适应的鲸鱼优化算法(AWOA)和快速学习网(FLN)综合建模的方法。首先将改进后的鲸鱼算法与经典改进的粒子群、差分进化算法和基本鲸鱼算法进行比较,结果证明其具有更高的收敛精度和更快的收敛速度;然后采用某热电厂600 MW超临界汽轮机组现场收集的运行数据建立汽轮机热耗率预测模型,并将改进后的鲸鱼算法优化的快速学习网模型的预测结果与基本快速学习网及经典改进的粒子群、差分进化算法和基本鲸鱼算法优化的快速学习网模型预测结果相比较。结果表明,AWOA-FLN预测模型具有更高的预测精度和更强的泛化能力,更能准确地预测汽轮机的热耗率。
In order to establish an accurate prediction model for heat consumption rate of steam turbines, an integrated modeling method was proposed by combination of oppositely adaptive whale optimization algorithm（AWOA） and fast learning network（FLN）. Compared to basic whale algorithm, improved particle swarm optimization algorithm, and differential evolution algorithm, the improved whale algorithm had higher convergence accuracy and faster convergence speed. A prediction model for heat consumption rate of a 600 MW supercritical steam turbine generator set in a thermal power plant was established from the collected operation data, which was also compared to FLN model, improved particle swarm optimization, differential evolution algorithm, and whale optimization algorithm. The results show that the AWOA-FLN prediction model had higher prediction accuracy and stronger generalization ability, which therefore could predict heat consumption rate of steam turbine more accurately.
出处
《化工学报》
EI
CAS
CSCD
北大核心
2017年第3期1049-1057,共9页
CIESC Journal
基金
国家自然科学基金项目(61573306
61403331)~~
关键词
汽轮机
热耗率
鲸鱼优化算法
快速学习网
反向学习算法
steam turbine
heat consumption rate
whale optimization algorithm
fast learning network
opposition-based learning
