高空舱飞行高度模拟串级LADRC鲁棒控制技术
Robust cascade LADRC technology for flight altitude simulation of high altitude cell
但志宏 1张松 1张和洪 2钱秋朦 3王信 3赵伟1
作者信息
- 1. 中国航发四川燃气涡轮研究院高空模拟技术重点实验室,四川绵阳 621703;中国航发四川燃气涡轮研究院,四川绵阳 621703
- 2. 福州大学计算机与大数据学院,福州 350108
- 3. 中国航发四川燃气涡轮研究院,四川绵阳 621703
- 折叠
摘要
针对航空发动机高空舱飞行高度模拟控制系统存在的强非线性、高度不确定性、强外部扰动等实际工程问题,对抑制系统不确定性影响的主要控制方法进行了分析和讨论,设计了一种基于降阶线性扩张状态观测器(RLESO)的串级线性自抗扰(LADRC)鲁棒控制方法.分析了被控对象的主要特性和控制难点,并将广义受控对象分为蝶阀位置回路和飞行高度回路.对两个回路分别设计降阶扩张状态观测器和控制器并组建串级控制系统.通过控制仿真并与经典PID控制方法进行了对比分析,结果显示在推力瞬变试验控制仿真中,被控压力的最大波动值从3.5 kPa减小至0.8 kPa,表明了基于RLESO的串级LADRC技术能够显著提升高空台飞行高度模拟的控制品质,获得了较为理想的鲁棒控制性能和抗扰性能.
Abstract
With strong nonlinearity,high uncertainty and strong external disturbance for flight altitude simulation of altitude test facility,a robust cascade linear active disturbance rejection control(LADRC)algorithm via the reduced order linear extended state observer(RLESO)was proposed.In particular,the main characteristics and control issues of the controlled plant were analyzed,and the generalized controlled plant can be included by the butterfly valve position loop and flight altitude loop.The RLESO and corresponding controller were designed for the two loops respectively,meanwhile the cascade control system was constructed.The robust cascade LADRC was realized in the simulation environment and compared with the classical PID control scheme.When the engine was subject to the thrust transient test,the maximum fluctuation value of the controlled pressure was reduced from 3.5 kPa to 0.8 kPa.This indicated that the robust cascade LADRC technology via RLESO can significantly improve the dynamic control quality.The ideal robust control performance and anti-disturbance ability for flight altitude simulation were obtained.
关键词
高空舱/飞行高度模拟/鲁棒性/降阶线性扩张状态观测器/线性自抗扰控制器Key words
high altitude cell/flight altitude simulation/robustness/reduced order linear extended state observer/cascade linear active disturbance rejection control引用本文复制引用
基金项目
中国航发四川燃气涡轮研究院稳定支持项目(GJCZ-0011-19)
福建省自然科学基金(2021J02008)
国家自然科学基金(62003088)
出版年
2024
NETL
NSTL
万方数据