超（超）临界燃煤机组非线性压制与抗干扰受限控制方法研究

Low speed of load response, insufficient ability to eliminate perturbation, inability to adapt to different coal qualities, and large variation of control inputs are the four main problems practically faced on by the (ultra-)supercritical unit in power plant. .To solve these problems, this project proposes a new type of constrained control strategy for (ultra-)supercritical unit through the investigations on the following three scientific topics such as dynamic modeling, disturbance rejection and nonlinear domination controller design. Firstly, with some assumptions, we establish grey dynamic models of (ultra-)supercritical unit in form of state space equation so as to trade-off the accuracy and structure complexity, required by controller design. Meanwhile, state observer design method is investigated to estimate some unmeasured states of the unit. Secondly, a constrained controller design method is proposed via idea of nonlinear domination in the framework of Lyapunov stability theory. This method applies, instead of cancelling, sufficiently the nonlinearities of unit to design controller, and therefore not only can significantly enhance the transient response speed of the unit but also can depress some perturbations that are caused by some operations such as soot blowing, startup and shutdown of ball mill, and so on. Thirdly, a nonlinear disturbance observer is designed for the unit. This novel observer can be used to estimate the lumped disturbances including mismatched model, un-modeled dynamics and unmeasured disturbances. Finally, with the estimated lumped disturbances, a hybrid form of constrained coordinated control strategy, called "nonlinear domination feedback-disturbance compensation feed forward", is proposed. To validate the performance of such hybrid control strategy, some numerical and experimental simulations are conducted on a control platform. .This project provides an effective control strategy used to solve tracking control problems for multi-inputs multi-output nonlinear systems in thermal process, and is meaningfully important to guarantee the safety and economical efficiency of the (ultra-)supercritical unit.

升降负荷速率低、消除扰动工况能力不足、煤种适应能力差以及给煤给水控制量波动大是现役大型燃煤机组面临的主要问题。针对上述问题，本项目围绕动态特性建模、干扰估计补偿、非线性压制控制等科学问题开展新的控制方法研究，主要包括：首先，在一定假设条件下，开展机组状态空间灰箱建模及状态观测方法研究，以折中满足控制器设计所需精度和复杂度的要求并实现不可测状态观测；其次，在Lyapunov稳定性理论范畴内开展非线性压制受限控制方法研究，通过充分利用机组非线性特性设计控制器以提高机组瞬态响应性能和抑制扰动工况能力；第三，开展非线性抗干扰观测方法研究，用于估计机组模型失配、未建模态、煤种变化等不可测综合干扰；最后，引入综合干扰估计，开展“非线性压制反馈——干扰估计补偿前馈”新型受限协调控制策略、仿真和性能试验研究。本项目为解决复杂热工过程多变量跟踪控制问题提供了一种有效方法，对保证机组经济高效运行具有重要意义。

升降负荷速率低、消除扰动工况能力不足、煤种适应能力差以及给煤给水控制量波动大是现役大型燃煤机组面临的主要问题。针对上述问题，本项目围绕动态特性建模、干扰估计补偿、非线性控制器设计等科学问题开展研究，主要包括：.首先，在一定合理的简化条件下，基于机理分析建立了超超临界燃煤机组的状态空间动态数学模型，并在此基础上结合典型负荷工况下的实验数据和全工况实时运行数据，采用进化算法辨识动态数学模型中的结构参数，使得建立的动态数学模型输出与机组动态响应在宽负荷范围内具有较好的一致性和较高的精度。.其次，为了适应煤质等未知扰动以及工况大范围变化对协调系统控制性能的影响，提出了一种基于高阶滑模干扰观测器的燃煤机组未知扰动估计方法。同时，从机理和稳定性分析的角度，给出了一种简单有效的入炉煤量在线观测方法。在此基础上结合机组自身运行特性，创新性地提出了一种“干扰估计补偿+非线性反馈压制+机理前馈”抗干扰制方法，可实现大范围负荷的快速、小偏差跟踪。.第三，考虑到实际控制作用受限的约束条件，将提出的抗干扰控制方法融入到最优控制框架，提出了一种可直接基于机组连续状态空间动态数学模型的滚动最优受限控制方法。该方法可折中实现机组协调的快速响应性能和控制作用受限之间的矛盾，避免了控制器增益的反复试凑整定方式，但也牺牲了大量的计算资源。.第四，考虑到过热汽温控制品质对机组协调控制性能的影响，提出了一种非线性压制控制器设计方法。该方法能够有效抑制一定强度的模型参数摄动等未知扰动对协调系统控制性能的影响，具有较好的稳定性和鲁棒性。.最后，构建了数据采集控制仿真平台，通过大量的模拟仿真以及从现场获取的大量运行数据对提出的控制方法进行了理论验证和分析。.研究结果表明，本项目提出的方法从理论上解决燃煤机组负荷快速跟踪控制问题，具有普适性，具有重要的理论研究意义和应用价值。