| 一种基于改进KFCM算法的火电厂煤耗特性模型的建立方法 |
| |
| 引用本文： | 东南大学 能源热转换及其过程测控教育部重点实验室,江苏 南京,内蒙古岱海发电有限责任公司,内蒙古 乌兰察布.一种基于改进KFCM算法的火电厂煤耗特性模型的建立方法[J].热能动力工程,2017,32(3):69-74. |
| |
| 作者姓名： | 东南大学 能源热转换及其过程测控教育部重点实验室 江苏 南京 内蒙古岱海发电有限责任公司 内蒙古 乌兰察布 |
| |
| 作者单位： | 为提高煤耗特性模型的实时性与可靠性，给出了从系统历史数据库中建立煤耗特性模型的基本流程。提出了基于标准 K 均值（Kmeans）与传统 FCM（模糊 C 均值）的改进 KFCM（模糊聚类）算法，引入有效性指标来评价聚类结果的有效性，迭代获得合理的聚类个数与聚类中心，以此作为 FCM 算法的初始条件，并完成测试数据的聚类分析。测试结果表明：该算法具有较高的精度。以某电厂1 000 MW 机组为研究对象，重点考虑环境温度、机组负荷对供电煤耗率的影响程度，建立煤耗特性模型。引入能耗知识库的离线更新，使得煤耗特性模型更符合机组实际运行过程，从而改善了当前电厂煤耗特性模型长期不变的现状。 |
| |
| 基金项目： | 国家自然科学基金资助项目(51176030) |
| |
| 摘 要： | 为提高煤耗特性模型的实时性与可靠性,给出了从系统历史数据库中建立煤耗特性模型的基本流程。提出了基于标准K均值(Kmeans)与传统FCM(模糊C均值)的改进KFCM(模糊聚类)算法,引入有效性指标来评价聚类结果的有效性,迭代获得合理的聚类个数与聚类中心,以此作为FCM算法的初始条件,并完成测试数据的聚类分析。测试结果表明:该算法具有较高的精度。以某电厂1 000 MW机组为研究对象,重点考虑环境温度、机组负荷对供电煤耗率的影响程度,建立煤耗特性模型。引入能耗知识库的离线更新,使得煤耗特性模型更符合机组实际运行过程,从而改善了当前电厂煤耗特性模型长期不变的现状。
|
| 关 键 词： | 环境温度 煤耗特性模型 有效性指标 KFCM 能耗知识库 |
A Coal Consumption Modeling Method for Power Plant based on Improved KFCM Algorithm |
| |
| Key Laboratory of Energy Thermal Conversion and Control of Ministry of Education,Southeast University,Nanjing,Chin,Post Code: and Inner Mongolia Daihai Electric Power Generation Co. Ltd.,Wulanchabu,Chin,Post Code:.A Coal Consumption Modeling Method for Power Plant based on Improved KFCM Algorithm[J].Journal of Engineering for Thermal Energy and Power,2017,32(3):69-74. |
| |
| Authors: | Key Laboratory of Energy Thermal Conversion and Control of Ministry of Education Southeast University Nanjing Chin Post Code: and Inner Mongolia Daihai Electric Power Generation Co Ltd Wulanchabu Chin Post Code: |
| |
| Affiliation: | The coal consumption characteristics of coal-fired units have evident variation during the operating period. As time goes by,the unit performance characteristics change,deviating from the curve obtained from the thermal test data or the performance parameters provided by the manufacturer. In order to make coal consumption model more real-time and reliable,this paper established a new coal consumption model. An improved fuzzy clustering algorithm based on Kmeans and fuzzy C-means was proposed. An index was introduced to evaluate the validity of clustering results and the reasonable clustering number and cluster centers resulted from iteration calculations were used as the initial conditions for the FCM algorithm. A case study test has verified the satisfied accuracy of the proposed algorithm. With a 1 000 MW unit as an object,the paper focused on the impact of the environment temperature and unit load on power generation coal consumption rate and established a new coal consumption model. With the addition of off-line updating of the coal consumption knowledge base,the presented coal consumption model can better respond to the actual operation process. |
| |
| Abstract: | |
| |
| Keywords: | environment temperature coal consumption model validity index KFCM knowledge base of coal consumption |
| 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 |
