【作者】 华志刚；

【导师】 吕剑虹；

【作者基本信息】 东南大学 ， 热能工程， 2006， 博士

【摘要】 国内大型热力设备的自动控制系统已逐步采用先进的分散控制系统,这为先进控制理论的应用提供了良好的实现条件。但是目前大型热力设备的自动控制几乎都仍采用常规的线性PID控制系统,难以在整个负荷变动范围内均取得优良的控制品质。绝大多数大型发电厂的关键控制系统,在机组负荷大范围变化时,均未能取得理想的控制品质,严重影响了热力设备的安全、经济运行。目前国内大型火电机组的锅炉过热汽温和再热汽温控制,几乎仍采用常规的串级控制系统,而不少电厂汽温被控对象的滞后很大,且喷水阀存在严重的非线性,使得当机组负荷变化时,汽温往往偏离设定值8～10℃,超温十分频繁。由于过热汽温和再热汽温的频繁超温,容易导致锅炉爆管事故的发生,严重影响了火电机组的安全、经济运行。为此,有必要研究基于现代控制策略的新型汽温控制方案。同样,大型火电机组的负荷系统是一个多变量非线性动态系统,它的动态特性随工况的变化而大范围变化。基于传统PID控制系统的机、炉协调控制系统,只是根据大型火电机组在某一负荷点上的对象特性来设计的,但当机组的负荷大范围变化时,被控对象的动态特性往往变化很大,且被控对象存在严重的非线性。因此,常规PID控制系统在机组的整个负荷范围内不可能是全局最优的,使大型火电机组的负荷跟踪能力及机组的稳定性均难以取得良好的控制效果。由于PID控制规律是线性的,而热工被控对象是非线性的,因此,若仍采用PID控制策略来对电站热工控制系统进行优化,总是突破不了用线性控制器来控制非线性对象的这一局限性,控制品质提高必定会受到限制。要进一步提高控制品质,应该尝试从方法上改变目前的这个常规控制模式,而研究采用基于火电机组整体非线性模型的全局非线性优化控制系统。为了解决电厂热工过程控制中存在的一些实际问题,本文针对大型火电机组汽温被控对象大惯性、大滞后的特点,以及负荷系统的非线性、时变性等特点,结合非线性系统模糊模型和预测控制、状态变量控制的方法,提出了一系列基于非线性模糊模型的状态变量控制方法和预测控制方法的先进控制策略,为复杂热力系统的控制提供了新的思路和方法。论文的第一章主要介绍了论文的选题意义和背景,以及状态变量控制、预测控制、基于线性矩阵不等式(LMI)控制的研究现状,分析了当前火电机组热工过程控制中存在的问题,并介绍了论文的主要研究工作。考虑到状态变量控制技术和预测控制技术是目前两种比较适合大滞后被控过程的控制方法,论文第二章重点介绍了状态变量控制和预测控制的基本原理,针对大型火电厂再热汽温被控对象大惯性、大滞后的特点,提出了基于状态变量-预测控制技术的再热汽温控制方法,即先采用状态反馈理论来补偿再热汽温被控对象的滞后和惯性,然后通过预测控制来对补偿后的广义被控对象进行控制。仿真试验及现场运行结果表明,该控制方法具有优良的控制品质,是一种对大滞后过程较为有效的控制策略。论文第三章根据目前国内外在非线性系统控制研究上的现状及存在的问题,文中针对被控对象的仿射TSK模糊模型,结合现代控制理论中的状态变量控制原理,研究提出了仿射模糊状态变量控制方法,并从理论上证明了闭环系统的稳定性,该控制方法具有优良的全局更多还原

【Abstract】 Advanced DCS (Distributed Control Systems) have been used in large scale thermal dynamic facility’s automatic control system. It provides good conditions for the application of the advanced control theory. However, because almost all of the large scale thermal facilities are still using conventional linear PID control system, it is very difficult to acquire excellent controlling performance in the global load varying range. Almost all of the critical control systems of the large scale power plant cannot acquire satisfying control quality when unit load varies in great ranges, it threatens the secure and economical operation of thermal dynamic facility severely.Most of the reheat steam temperature and superheat steam temperature control systems are the common cascade control system. But many of the controlled objects have large inertia and lag characteristic, and the spray valve have great nonlinearity. The steam temperature will deviate 8～10℃from the set-point when unit load changes, and always exceeds the steam temperature limits. Because the steam temperature usually exceeds the setpoint, the accident of blowing out the pipeline happened frequently. It threatens the unit’s security and economy severely. So it is necessary to research new style steam temperature control strategy based on the modern control theory.Load system of the large scale power plant’s unit as well, is a multi-variable nonlinear dynamic system, and the dynamic characteristic will change with the different operating condition. The traditional boiler-turbine coordinated control system is based on PID. This kind of coordinated control system is designed on a specific load condition. When the unit load varies in a large range, the unit’s characteristic would also change significantly. The characteristic of the unit’s load system has great nonlinearity, so it is impossible that the common PID control system have the global optimization in the whole load varying range. And it is difficult that the unit load tracking ability and the unit’s stability have good control performance.Because the PID controlling law is linear while the controlled object is nonlinear, if we still use the traditional PID control theory to optimize the thermal power plant control system, we are using the linear controller to control the nonlinear object. Then the control performance would be confined. In order to improve the control performance, we must change the conventional control strategy thoroughly. We can research the global nonlinear optimizing control system which based on the power plant’s global nonlinear model.In order to resolve the practical problem in the thermal process control of power plant, aimed at the large inertia and lag characteristics of the steam temperature object, and the time-variant and nonlinearity of the load system, combine with the nonlinear fuzzy model and predictive control, state variable control methods, this paper put forwards a series of advanced control strategies based on nonlinear fuzzy model state variable control theory, state feedback control theory and predictive control theory. It is a new idea for the controlling of the complex thermal-dynamical systems.The purpose of the selected researching aspect and the background of this paper are introduced in the first part of the paper. It also introduces the researching status of the state variable control, predictive control and Linear Matrix Inequality, and analyzes the existing更多还原