摘要
针对火电厂目前主蒸汽温度数学模型不准确、不能有效地反应实际生产过程、辨识结果偏差较大等问题,使用学习因子动态调整及随机惯性权重策略改进的二阶振荡粒子群优化(ISO-PSO)算法进行建模。首先,采用学习因子动态调整、随机惯性权重策略对二阶振荡粒子群算法进行改善。测试结果表明,改进算法拥有更好的收敛速度及寻优精度。其次,处理山西某电厂350 MW超临界循环流化床(CFB)锅炉在300 MW工况下采集到的主蒸汽温度数据。最后,对比ISO-PSO算法和粒子群优化(PSO)算法。辨识结果及验证结果表明,ISO-PSO算法辨识的主蒸汽温度模型与实际输出的拟合度更高,能够更加准确、快速地反映主蒸汽温度的变化。所建立的主蒸汽温度模型可以及时跟踪实际曲线。该研究为后续350 MW超临界CFB锅炉主蒸汽温度优化控制研究奠定了良好的基础。
To address the problems of inaccurate mathematical model of main steam temperature in thermal power plants,inability to effectively reflect the actual production process,and large deviation of identification results,the improved secondorder oscillatory particle swarm optimization(ISO-PSO)algorithm with improved learning factor dynamic adjustment and randominertia weighting strategy is used for modeling.Firstly,the learning factor dynamic adjustment and random inertia weighting strategy is used to improve the second-order oscillatory particle swarm algorithm.The test results show that the improved algorithm has better convergence speed and better finding accuracy.Secondly,the main steam temperature data collected from a 350 MW supercritical circulating fluidized bed(CFB)boiler in a Shanxi power plant at 300 MW operating conditions are processed.Finally,comparing the ISO-PSO algorithm and the particle swarm optimization(PSO)algorithm.The identification results and the validation results show that the main steam temperature model identified by the ISO-PSO algorithm has a better fit with the actual output and can respond to the main steam temperature changes more accurately and quickly,and the established main steam temperature model can track the actual curve in time.This study lays a good foundation for the subsequent research on the optimal control of main steam temperature in 350 MW supercritical CFB boilers.
作者
王琦
张力文
王丽婕
钟义
WANG Qi;ZHANG Liwen;WANG Lijie;ZHONG Yi(School of Automation and Software Engineering,Shanxi University,Taiyuan 030000,China)
出处
《自动化仪表》
CAS
2023年第3期20-25,33,共7页
Process Automation Instrumentation
基金
国家自然科学基金资助项目(61803244)
山西省研究生创新基金资助项目(2021Y158)。
关键词
循环流化床锅炉
主蒸汽温度建模
粒子群优化算法
测试函数
对比试验
Circulating fluidized bed(CFB)boiler
Main steam temperature modeling
Particle swarm optimization(PSO)algorithm
Test function
Comparison test
