摘要
锅炉热效率和NOx排放一直是火电机组运行中的重要指标,在大容量、高参数超超临界机组呈现良好发展势头的当前,对1000MW超超临界机组锅炉燃烧系统的建模问题进行研究具有重要意义。借助某电厂1000 MW超超临界锅炉的现场燃烧调整试验数据,分别建立锅炉热效率和NOx排放量的最小二乘支持向量机(LSSVM)模型,采用一种改进的粒子群算法对模型参数进行优化选取。实验结果表明,锅炉热效率和NOx排放量的平均预测误差分别可达到0.78%和4.22%,验证了LSSVM模型的有效性。
Boiler thermal efficiency and NOx emission are always the critical indexs in thermal power-generation units. Research on modeling for 1000MW ultra supercritical boiler combustion system has improtant significance, es- pecially along with the development of large capacity and high parameter ultra supercirtical units. Based on the com- bustion adjustment experiment data of a 1000MW ultra supercritical boiler, the LSSVM models for boiler thermal effi- ciency and NOx emission were established. And an improved particle swarm optimization was selected to optimize the model parameters. The effectiveness of LSSVM model was verified and the average prediction error for bolier efficeiency and NOx emission can reach 0.78% and 4.22% respectively.
出处
《计算机仿真》
CSCD
北大核心
2013年第5期129-132,147,共5页
Computer Simulation
关键词
超超临界
最小二乘支持向量机
粒子群优化
锅炉热效率
Ultra supercritical
Least squares support vector machine （LSSVM）
Particle swarm optimization（ PSO ）
Boiler thermal efficiency
