【摘要】 火力发电厂中广泛使用的化石燃料发电机组(FFPU)具有非线性、强耦合、参数随工况变化等复杂特性,依据单个工作点处的线性模型设计的多变量控制策略常常导致控制效果不佳,甚至系统不能稳定运行。针对该问题,本文提出了一种有效的多模型控制策略,以改善FFPU大工况范围运行时的控制性能。首先,通过定义归一化特征值作为系统非线性测度,将FFPU工作运行区间划分成多个局部模型区域,在每个区域用线性模型表征系统特性;随后,根据每个局部区域的线性模型,实施一种基于FFPU特性而设计的包含静态前馈、PD前馈和模糊解耦的混合解耦补偿,进而采用模糊PID结构设计分散控制器;最后,依照电力负载跟随调度策略,模糊加权局部控制器得到全局控制器输出,使得系统在大工况范围运行时平滑切换。为实现该控制策略,采用了双层控制结构:上层为监督控制层,作用是根据负载要求进行设定点调度以及进行局部控制器的模糊切换;下层为直接控制层,实施混合解耦补偿和各回路的模糊PID控制。仿真表明,所提出的控制策略性能良好且实用有效。更多还原

【Abstract】 Fossil fuel power unit(FFPU) widely used in power generations is a multivariable system with the characteristics of severe nonlinearity,strong coupling,and complex dynamics with multiple changing operating conditions.Some multivariable control schemes designed for FFPU based on linearization model around a single operation point can seldom get good control performance,and even result in unstable control system.In order to deal with this problem,a practical multi-model control approach was developed in this paper to improve FFPU control performance across its wide operation range.First,through using a nonlinearity measure of normalized eigenvalues,the operation space of a FFPU system was divided into several appropriate local regimes around the respective static state operation points,in which linear local models were obtained respectively.Second,based on these local models and the process properties of the FFPU,a hybrid decoupling technique(including static-state feed-forward,PD feed-forward,and fuzzy decoupling) was used in the control loops,and then decentralized fuzzy PID controllers were designed after pre-compensating decoupling.Finally,a fuzzy switching logic was to implement the fuzzy weight computation of different local controller outputs in terms of a practical scheduling strategy so that a smoothly switching control may be attained.The whole hierarchical structure had two layers:supervisory control level,which was the upmost level and to implement set-points scheduling and fuzzy switch scheduling in response to operation demand;direct control level,which was in the bottom to realize the hybrid decoupling and fuzzy PID control for all process loops.The good control performance of simulation demonstrated the effectiveness of the proposed approach.更多还原