摘要
生产装置能否安全有效地运行,已经成为衡量工业生产发展水平的主要标志之一。面临日益复杂的化工过程生产装置,提高化工过程报警系统的性能有着重要的指导意义。传统的报警阈值参数设置方法局限性大,为了提升化工过程报警系统性能,需要对某些过程参数的报警阈值进行优化设置。本文针对传统粒子群算法的不足,采用了参数自适应的粒子群算法,该自适应粒子群算法通过实时调节自身的参数,使得能够较快地寻找到最优个体,且不容易陷入局部最优解。通过对一标准函数的研究,结果表明该自适应粒子群算法比传统的粒子群算法能够较快的达到最优解。随后,用该算法优化TE过程某一参数的报警阈值,降低了报警过程中误报和漏报的总次数,提高了报警系统的性能。本文所提方法为指导生产装置的安全运行提供了有效策略。
The safety and efficiency of chemical production facilities has become one of the main indicators to measure the level of development of the petrochemical industry. Faced with the increasingly complex chemical process plants, improving the performance of the alarm systems in chemical processes has important guiding significance. The traditional method of parameter setting in alarm thresholds is limited in some aspects. In order to enhance the performance of the alarm systems in large chemical processes, some alarm threshold settings of process parameters need to be optimized. A method based on Time Slice Model （TSM） is proposed for batch processes heat exchanger network synthesis by taking place. Aiming at avoiding the shortage of traditional particle swarm algorithm, a parameter-adaptive particle swarm optimization （APSO） was adopted. The APSO algorithm can adjust its parameters in every interval time, which make it possible to quickly find the best individual and not easy to get into local optimum solution. Through a standard function test, the results show that the adaptive particle swarm optimization can more quickly reach the optimal solution than the traditional PSO algorithm does. Then, the adaptive particle swarm algorithm is used to optimize certain alarm thresholds of parameters in the TE process so that the total number of alarms is decreased. So the performance of the alarm system is improved, which make the system safer and more effective operate and provide an effective strategy for the safe production of chemical processes.
出处
《计算机与应用化学》
CAS
2015年第1期103-106,共4页
Computers and Applied Chemistry
关键词
粒子群算法
自适应粒子群
TE过程
阈值优化
particle swarm optimization
adaptive particle swarm optimization
TE process
threshold optimization
作者简介
杨洪雪（1972—），女，黑龙江双城人，副教授，Email：yhxzxb@sina．com
