基于智能优化算法的热工大惯性对象模型辨识研究

发布时间：2018-08-14 13:40

【摘要】：热工大惯性对象的模型辨识是研究热工控制问题的基础,确定模型的结构和参数是进行控制系统设计、调试和投运的重要一环。在确定或辨识出对象模型之后,可以对实际控制系统进行富有针对性的参数优化,有效提升控制系统的质量和效率,保障生产过程的安全。本文以模型辨识技术、PID参数优化技术为背景,以典型的热工大惯性对象的实际运行数据为依据,采用改进粒子群算法对热工对象进行辨识,分析影响系统辨识结果的主要因素以及在辨识过程中应注意的问题,利用辨识结果实现PID控制器参数的优化。采用PSO算法对热工对象进行模型辨识具有速度快、灵活方便的特点。但PSO算法存在易陷入局部最优解、易早熟及搜索精度不高等问题,因此需要在PSO算法的基础上,研究对该算法改进与提高的方法。首先,在PSO中引入混沌搜索和模拟退火思想,使算法有更好的多样性和跳出局部最优的能力。其次,结合细菌趋化算法,引入排斥操作,用以增强粒子的搜索能力。用典型函数对改进后的算法进行验证,可以看出:算法的收敛性、稳定性及全局搜索能力得到了明显改善。根据对热工模型结构的认识,将改进粒子群算法用于单入单出的实验室锅炉温度对象和多入单出的超超临界锅炉主蒸汽温度对象的模型辨识中,并对几种改进算法的辨识结果进行分析和对比,结果表明:针对不同的热工控制系统,细菌趋化粒子群算法的适应能力更好,辨识所得的模型更精确。
[Abstract]:The model identification of large inertia object is the basis of studying the thermal control problem. Determining the structure and parameters of the model is an important part of the design, debugging and commissioning of the control system. After the object model is determined or identified, the actual control system can be optimized for parameters, which can effectively improve the quality and efficiency of the control system and ensure the safety of the production process. In this paper, based on the model identification technology and pid parameter optimization technology, based on the actual operation data of typical large inertia thermal engineering object, the improved particle swarm optimization algorithm is used to identify the thermal object. The main factors affecting the system identification results and the problems that should be paid attention to in the identification process are analyzed. The parameters of the PID controller are optimized by using the identification results. The model identification of thermal objects using PSO algorithm has the advantages of high speed, flexibility and convenience. However, PSO algorithm is easy to fall into local optimal solution, easy to prematurity and low searching precision. Therefore, it is necessary to study the method of improving and improving the algorithm based on PSO algorithm. Firstly, chaotic search and simulated annealing are introduced into PSO, which makes the algorithm have better diversity and the ability to jump out of the local optimum. Secondly, combined with bacterial chemotaxis algorithm, repulsive operation is introduced to enhance the searching ability of particles. The improved algorithm is verified by the typical function, and the convergence, stability and global search ability of the algorithm are obviously improved. Based on the understanding of the structure of the thermal model, the improved particle swarm optimization algorithm is applied to the model identification of the temperature object of the laboratory boiler and the main steam temperature object of the supercritical boiler with multiple input and single output. The identification results of several improved algorithms are analyzed and compared. The results show that the bacterial chemoattractant particle swarm optimization algorithm has better adaptability and more accurate identification model for different thermal control systems.
【学位授予单位】：上海电力学院
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TM621

【参考文献】