| 基于反馈快速学习网的热力系统混合预测模型 |
| |
| 引用本文： | 唐立力,陈国彬,刘 超,牛培峰.基于反馈快速学习网的热力系统混合预测模型[J].热能动力工程,2018,33(11):113. |
| |
| 作者姓名： | 唐立力 陈国彬 刘 超 牛培峰 |
| |
| 作者单位： | 重庆工商大学 融智学院,重庆 401320,重庆工商大学 融智学院,重庆 401320,贵州航天电器股份有限公司,贵州 贵阳 550009,燕山大学 电气工程学院,河北 秦皇岛 066004 |
| |
| 基金项目： | 国家自然科学基金(61403331，61573306)；重庆市教委科学技术研究项目(KJQN201802102） |
| |
| 摘 要： | 针对汽轮机机组功率难以精确计算的问题,提出一种热力系统混合预测模型以改善机组功率的计算精度。混合预测模型以热平衡方程为基本模型,以反馈快速学习网(feedback FLN,B-FLN)为补偿模型,将B-FLN的功率预测值作为热平衡方程功率计算值的补偿值。以某300 MW机组为研究对象,采用提出的混合预测模型对机组的功率进行预测,将其结果与热平衡方程计算值进行比较。仿真结果表明:混合模型输出功率误差小于±2 MW,降低了热平衡方程计算误差,为机组功率的可靠计算提供了一种新的解决思路。
|
| 关 键 词： | 热力系统 快速学习网 热平衡方程 汽轮机 预测 |
Hybrid Prediction Model of Thermal System based on Feedback Fast Learning Neural Network |
| |
| Tang Li-li,CHEN Guo-bin,LIU Chao and NIU Pei-feng.Hybrid Prediction Model of Thermal System based on Feedback Fast Learning Neural Network[J].Journal of Engineering for Thermal Energy and Power,2018,33(11):113. |
| |
| Authors: | Tang Li-li CHEN Guo-bin LIU Chao and NIU Pei-feng |
| |
| Affiliation: | Rongzhi College of Chongqing Technology and Business University,Chongqing,China,Post Code: 401320,Rongzhi College of Chongqing Technology and Business University,Chongqing,China,Post Code: 401320,Guizhou Aerospace Electronics Co.,LTD.,Guiyang,China,Post Code: 550009 and Institute of Electrical Engineering,Yanshan University,Qinhuangdao,China,Post Code: 066004 |
| |
| Abstract: | In order to establish the power prediction model of turbine units accurately, a hybrid prediction model of thermal system is proposed to improve the calculation accuracy of turbine unit power. The hybrid prediction model uses the thermal balance equation as the basic model and feedback FLN (B-FLN) as the compensation model, in which, the power prediction value of B-FLN is taken as the compensation value of the power calculation value of the thermal balance equation. With a 300MW unit as the study object, the power of the unit is predicted by using the hybrid prediction model proposed, and the result is compared with the calculated value of the thermal balance equation. The simulation results show that the power error of the hybrid model is less than 2 MW, which reduces the calculation error of the thermal balance equation and provides a new solution for the reliable calculation of unit power. |
| |
| Keywords: | thermal system fast learning network thermal balance equation steam turbine prediction |
| 本文献已被 CNKI 等数据库收录！ |
| 点击此处可从《热能动力工程》浏览原始摘要信息 |
|
点击此处可从《热能动力工程》下载全文 |
|
