基于火电燃烧机组中固态变压器改进LADRC协同优化控制策略

Improvement of LADRC collaborative optimization control strategy based on solid state transformers in thermal power combustion units

在火电燃烧机组中,固态变压器面临惯性大、滞后多和耦合强等问题,为提升其控制性能,研制了一种基于改进线性自抗扰的协同燃烧优化控制策略。该策略以参数解耦型观测器为核心,结合BP神经网络算法进行参数调节,将固态变压器逆变级数学模型改写为解耦自抗扰形式。经数字仿真实验验证,该策略能有效提升抵抗扰动的能力,稳定误差下降2.5%,控制误差在3%以内,优于传统的PI控制和LADRC控制方法。

In thermal power combustion units,solid-state transformers face problems such as large inertia,multiple lags,and strong coupling.To improve their control performance,a collaborative combustion optimization control strategy based on improved linear self disturbance rejection has been developed.This strategy adopts a parameter decoupling observer as the core,combined with BP neural network algorithm for parameter adjustment,and rewrites the mathematical model of the solid-state transformer inverter stage into a decoupled self disturbance rejection form.Through digital simulation experiments,it has been verified that this strategy can effectively improve the ability to resist disturbances,reduce stability errors by 2.5%,and control errors within 3%,which is superior to traditional PI control and LADRC control methods.