基于贝叶斯优化的自适应循环发动机性能寻优控制

朱鑫宇; 徐思远; 肖红亮; 魏鹏飞; 符江锋

doi:10.13224/j.cnki.jasp.20240112

基于贝叶斯优化的自适应循环发动机性能寻优控制

doi: 10.13224/j.cnki.jasp.20240112

1.
西北工业大学动力与能源学院，西安 710072
2.
中国航天科工集团有限公司北京动力机械研究所，北京 100074
3.
长安大学能源与电气工程学院，西安 710064

基金项目: 国家科技重大专项（J2019-Ⅴ-0016-0111）；国家自然科学基金面上项目（52372396）

详细信息

作者简介:
朱鑫宇（2001-），女，硕士生，从事航空发动机控制方向的研究。E-mail：zhuxinyu@mail.nwpu.edu.cn

通讯作者:
符江锋（1984-），男，研究员、博士生导师，博士，从事航空发动机控制方向的研究。E-mail：fjf@nwpu.edu.cn

中图分类号: V233.7
计量
- 文章访问数: 714
- HTML浏览量: 405
- PDF量: 62
- 被引次数: 0
出版历程
- 收稿日期: 2024-02-29
- 网络出版日期: 2024-12-15

Performance seeking control of adaptive cycle engine based on Bayesian optimization

1.
School of Power and Energy，Northwestern Polytechnical University，Xi’an 710072，China
2.
Beijing Power Machinery Institute，China Aerospace Science and Industry Corporation Limited，Beijing 100074，China
3.
School of Energy and Electrical Engineering，Chang’an University，Xi’an 710064，China

摘要

摘要:
针对自适应循环发动机多工作模式下的性能寻优控制需求，为减少发动机模型调用次数和计算时长，避免局部最优问题，给出一种基于代理模型和贝叶斯优化的发动机性能寻优控制策略。该方法以高精度部件法模型为基础，基于高斯过程回归理论构建自适应循环发动机代理模型；采用罚函数和指示函数，将带有约束条件的寻优问题转化为无约束问题，解决性能寻优控制中的非线性约束问题；最后以自适应循环发动机过渡态推力、耗油率和涡轮前温度为寻优目标进行验证。研究结果表明：少量样本构建的高斯代理模型显著减少了发动机模型调用次数，有效降低了计算量，且能够避免发动机模型调用过程中陷入局部循环的问题；贝叶斯优化算法采用主动学习策略，根据收敛条件评估代理模型自主增加样本点并更新模型，提高了模型计算精度；贝叶斯优化算法具有全局搜索特性，能够克服发动机性能寻优算法依赖人工经验的缺点，为发动机性能寻优提供了一种有效的解决方案；分别对带有核心机驱动风扇的双外涵自适应循环发动机最大推力模式，最低耗油率模式和最低涡轮前温度模式优化，推力优化了1501.27 N，耗油率优化了0.38%，涡轮前温度优化了7.9 K。
- 航空航天推进系统 /
- 自适应循环发动机 /
- 性能寻优控制 /
- 贝叶斯优化算法 /
- 主动学习
Abstract:
According to the performance seeking control of adaptive cycle engine in multiple operating modes, an engine performance optimization control strategy based on surrogate model and Bayesian optimization was proposed to reduce the number of engine model calls and calculation time, and avoid local optimization problems. This method built an adaptive cycle engine surrogate model based on the high-precision component model and the Gaussian process regression. It used penalty functions and indicator functions to transform the optimization problem with constraints into an unconstrained problem. The performance seeking control system based on the adaptive cycle engine model was verified with three optimization modes: minimum fuel flow at constant thrust, minimum turbine temperature at constant thrust, and maximum thrust at maximum dry and full afterburner throttle settings. The simulation results showed that the Gaussian surrogate model constructed with a small number of samples significantly reduced the number of engine model calls, effectively cut down the amount of calculation, and avoided the problem of falling into a local loop during the engine model calling process; the Bayesin optimization algorithm used an active learning strategy to independently increase sample points and update the model based on the convergence condition evaluation agent model; the Bayesian optimization algorithm with global search characteristics can overcome engine performance problems. The optimization algorithm overcame the disadvantage of relying on manual experience, providing an effective solution for engine performance optimization. Optimization results showed 1501.27 N enhancement in maximum thrust mode, 0.38% reduction in minimum fuel consumption mode and 7.9 K reduction in minimum turbine temperature mode for adaptive cycle engine with a core-driven fan respectively.
- aerospace propulsion system /
- adaptive cycle engine /
- performance seeking control /
- Bayesian optimization /
- active learning

HTML

图 1 优化流程图

Figure 1. Optimization flowcharts

下载: 全尺寸图片幻灯片

图 2 最大推力寻优过程及结果

Figure 2. Training details and result of maximum thrust mode

下载: 全尺寸图片幻灯片

图 3 最大推力寻优结果图

Figure 3. Optimization results for maximum thrust

下载: 全尺寸图片幻灯片

图 4 最小耗油率寻优过程及结果

Figure 4. Training details and result of minimum fuel flow mode

下载: 全尺寸图片幻灯片

图 5 最小耗油率寻优结果图

Figure 5. Optimization results for minimum fuel flow

下载: 全尺寸图片幻灯片

图 6 最低涡轮前温度寻优过程及结果

Figure 6. Training details and result of minimum turbine temperature mode

下载: 全尺寸图片幻灯片

图 7 最低涡轮前温度寻优结果图

Figure 7. Optimization results for minimum turbine temperature

下载: 全尺寸图片幻灯片

表 1 最优控制变量计算结果

Table 1. Calculation results of control variates

参数	额定状态	最大推力状态	最小耗油率状态	最低涡轮前温度状态	下限	上限
$ \zeta_{\mathrm{sfc}} $/（kg/（N·h））	0.0787	0.0790	0.0784	0.0789
$ {F_{\text{n}}} $/N	94209.58	95710.85	94211.96	94209.63
$ {T_{41}} $/K	1740.36	1761.59	1744.32	1732.46
$ {W_{{\text{fb}}}} $/（kg/s）	2.0589	2.1000	2.0539	2.0839	1.2250	2.4000
$ {A_8} $/m²	0.2299	0.2345	0.2345	0.2193	0.1954	0.2345
$ {A_{163}} $/m²	0.1506	0.1280	0.0896	0.1573	0.0896	0.1689
$ {n_{\text{l}}} $	0.9500	0.9733	0.9600	0.9300		1.03
$ {n_{\text{h}}} $	0.9971	0.9994	0.9964	0.9918		1.03
$ S_{{\text{fan}}} $	0.3896	0.4262	0.4155	0.3106	0.1
$ S_{{\text{cdfs}}} $	0.1449	0.1477	0.1471	0.1364	0.1
$ S_{{\text{com}}} $	0.2296	0.2255	0.2304	0.2373	0.1

下载: 导出CSV

表 2 不同方法的优化结果

Table 2. Optimization results of different methods

模式	方法	设计点（$ {W_{{\text{fb}}}},{A_8},{A_{163}} $）/ （$ {\text{kg/s}},{{\text{m}}^{\text{2}}},{{\text{m}}^{\text{2}}} $）	优化值	变异性系数/10⁻⁴	样本数	时间/s
最大推力模式/N	遗传算法	（2.1000, 0.2345, 0.1318）	95703.23		4927	8.1
	粒子群优化算法	（2.4000, 0.2185, 0.0896）	105251.17		3000	33.9
	贝叶斯优化算法	（2.1000, 0.2345, 0.1280）	95710.85	0.2	11+0	0.9
最小耗油率模式/ （kg/s）	遗传算法	（2.0589, 0.2299, 0.1506）	0.0787		6868	9.5
	粒子群优化算法	（2.0572, 0.2345, 0.1422）	0.0786		3000	27.4
	贝叶斯优化算法	（2.0539, 0.2345, 0.0896）	0.0784	5.7	11+5	4.2
最低涡轮前温度模式/K	遗传算法	（2.0874, 0.2178, 0.1291）	1730.43		32488	39.4
	粒子群优化算法	（2.0937, 0.2157, 0.0896）	1732.10		3000	28.3
	贝叶斯优化算法	（2.0839, 0.2193, 0.1573）	1732.46	8	11+6	6.0

下载: 导出CSV

参考文献(26)

[1]	樊思齐,李华聪,樊丁,等. 航空发动机控制[M]. 西安: 西北工业大学出版社,2008. FAN Siqi,LI Huacong,FAN Ding,et al. Aero engine control[M]. Xi’an: Northwestern Polytechnical University Press. 2008. (in Chinese FAN Siqi, LI Huacong, FAN Ding, et al. Aero engine control[M]. Xi’an: Northwestern Polytechnical University Press. 2008. (in Chinese)
[2]	廉筱纯,吴虎. 航空发动机原理[M]. 西安: 西北工业大学出版社,2005. LIAN Xiaochun,WU Hu. Aviation engine principle[M]. Xi’an: Northwestern Polytechnical University Press. 2005. (in Chinese LIAN Xiaochun, WU Hu. Aviation engine principle[M]. Xi’an: Northwestern Polytechnical University Press. 2005. (in Chinese)
[3]	朱之丽,陈敏,唐海龙,等. 航空燃气涡轮发动机工作原理及性能[M]. 3版. 上海: 上海交通大学出版社, 2023. ZHU Zhili,CHEN Min,TANG Hailong,et al. Working principle and performance of aviation gas turbine engines[M]. 3nd ed. Shanghai: Shanghai Jiao Tong University Press, 2023. (in Chinese ZHU Zhili, CHEN Min, TANG Hailong, et al. Working principle and performance of aviation gas turbine engines[M]. 3nd ed. Shanghai: Shanghai Jiao Tong University Press, 2023. (in Chinese)
[4]	李永进. 智能航空发动机性能优化控制技术研究[D]. 南京: 南京航空航天大学,2018. LI Yongjin. Research on intelligent aero-engine performance seeking control technology[D]. Nanjing: Nanjing University of Aeronautics and Astronautics,2018. (in Chinese LI Yongjin. Research on intelligent aero-engine performance seeking control technology[D]. Nanjing: Nanjing University of Aeronautics and Astronautics, 2018. (in Chinese)
[5]	薛益春. 变循环发动机多变量控制及性能寻优[D]. 南京: 南京航空航天大学,2012. XUE Yichun. Multivariable control and performance optimization of variable cycle engine [D]. Nanjing: Nanjing University of Aeronautics and Astronautics,2012. (in Chinese XUE Yichun. Multivariable control and performance optimization of variable cycle engine [D]. Nanjing: Nanjing University of Aeronautics and Astronautics, 2012. (in Chinese)
[6]	陈玉春,贾琳渊,任成,等. 变循环发动机稳态控制规律设计的新方法[J]. 推进技术,2017,38(10): 2262-2270. CHEN Yuchun,JIA Linyuan,REN Cheng,et al. An innovative method for design of steady state control law for variable cycle engines[J]. Journal of Propulsion Technology,2017,38(10): 2262-2270. (in Chinese CHEN Yuchun, JIA Linyuan, REN Cheng, et al. An innovative method for design of steady state control law for variable cycle engines[J]. Journal of Propulsion Technology, 2017, 38(10): 2262-2270. (in Chinese)
[7]	GILYARD G,ORME J. Performance seeking control: program overview and future directions[R]. AIAA-1993-3765,1993.
[8]	SIMMONS R J. Design an control of a variable geometry turbofan with an independently modulated third stream[D]. Columbus,Ohio,US: The Ohio State University,2009.
[9]	王芳. 航空发动机性能寻优控制研究[D]. 西安: 西北工业大学,2005. WANG Fang. Research on performance optimization control of aero-engine[D]. Xi’an: Northwestern Polytechnical University,2005. (in Chinese WANG Fang. Research on performance optimization control of aero-engine[D]. Xi’an: Northwestern Polytechnical University, 2005. (in Chinese)
[10]	朱玉斌. 航空发动机数控系统仿真技术研究[D]. 西安: 西北工业大学,2004. ZHU Yubin. Research on simulation technology of aero-engine numerical control system[D]. Xi’an: Northwestern Polytechnical University. 2004. (in Chinese ZHU Yubin. Research on simulation technology of aero-engine numerical control system[D]. Xi’an: Northwestern Polytechnical University. 2004. (in Chinese)
[11]	朱玉斌,樊思齐,李华聪,等. 航空发动机性能寻优控制混合优化算法[J]. 航空动力学报,2006,21(2): 421-426. ZHU Yubin,FAN Siqi,LI Huacong,et al. Research of hybrid optimization method for aeroengine performance seeking control[J]. Journal of Aerospace Power,2006,21(2): 421-426. (in Chinese doi: 10.3969/j.issn.1000-8055.2006.02.033 ZHU Yubin, FAN Siqi, LI Huacong, et al. Research of hybrid optimization method for aeroengine performance seeking control[J]. Journal of Aerospace Power, 2006, 21(2): 421-426. (in Chinese) doi: 10.3969/j.issn.1000-8055.2006.02.033
[12]	张海波,孙健国. 可行SLP法在发动机在线优化中的应用[J]. 航空学报,2010,31(4): 663-670. ZHANG Haibo,SUN Jianguo. Application of feasible descent sequential linear programming to aero-engine online optimization[J]. Acta Aeronautica et Astronautica Sinica,2010,31(4): 663-670. (in Chinese ZHANG Haibo, SUN Jianguo. Application of feasible descent sequential linear programming to aero-engine online optimization[J]. Acta Aeronautica et Astronautica Sinica, 2010, 31(4): 663-670. (in Chinese)
[13]	赵琳,樊丁,陕薇薇. 航空发动机过渡态全局寻优控制方法研究[J]. 航空动力学报,2007,22(7): 1200-1203. ZHAO Lin,FAN Ding,SHAN Weiwei. A global optimization approach to aero-engine under transient conditions[J]. Journal of Aerospace Power,2007,22(7): 1200-1203. (in Chinese ZHAO Lin, FAN Ding, SHAN Weiwei. A global optimization approach to aero-engine under transient conditions[J]. Journal of Aerospace Power, 2007, 22(7): 1200-1203. (in Chinese)
[14]	聂友伟,李秋红,王元,等. 基于SQCQP算法的变循环发动机性能寻优控制[J]. 北京航空航天大学学报,2017,43(12): 2564-2572. NIE Youwei,LI Qiuhong,WANG Yuan,et al. Variable cycle engine performance seeking control based on SQCQP algorithm[J]. Journal of Beijing University of Aeronautics and Astronautics,2017,43(12): 2564-2572. (in Chinese NIE Youwei, LI Qiuhong, WANG Yuan, et al. Variable cycle engine performance seeking control based on SQCQP algorithm[J]. Journal of Beijing University of Aeronautics and Astronautics, 2017, 43(12): 2564-2572. (in Chinese)
[15]	王元. 变循环发动机建模及性能寻优控制技术研究[D]. 南京: 南京航空航天大学,2015. WANG Yuan. Research on modeling techniques and performance seeking control of variable cycle engine[D]. Nanjing: Nanjing University of Aeronautics and Astronautics,2015. (in Chinese WANG Yuan. Research on modeling techniques and performance seeking control of variable cycle engine[D]. Nanjing: Nanjing University of Aeronautics and Astronautics, 2015. (in Chinese)
[16]	贾琳渊. 变循环发动机控制规律设计方法研究[D]. 西安: 西北工业大学,2017. JIA Linyuan. Research on variable cycle engine control schedule design[D]. Xi’an: Northwestern Polytechnical University,2017. (in Chinese JIA Linyuan. Research on variable cycle engine control schedule design[D]. Xi’an: Northwestern Polytechnical University, 2017. (in Chinese)
[17]	ZHENG Qiangang,FU Dawei,WANG Yong,et al. A study on global optimization and deep neural network modeling method in performance-seeking control[J]. Proceedings of the Institution of Mechanical Engineers,Part I: Journal of Systems and Control Engineering,2020,234(1): 46-59.
[18]	李立君,尹泽勇,乔渭阳. 基于多目标遗传算法的航空发动机总体性能优化设计[J]. 航空动力学报,2006,21(1): 13-18. LI Lijun,YIN Zeyong,QIAO Weiyang. Performance optimal design of aircraft engine based on multi-objective genetic algorithms[J]. Journal of Aerospace Power,2006,21(1): 13-18. (in Chinese doi: 10.3969/j.issn.1000-8055.2006.01.003 LI Lijun, YIN Zeyong, QIAO Weiyang. Performance optimal design of aircraft engine based on multi-objective genetic algorithms[J]. Journal of Aerospace Power, 2006, 21(1): 13-18. (in Chinese) doi: 10.3969/j.issn.1000-8055.2006.01.003
[19]	随阳,叶志锋,薛益春. 基于遗传算法的变循环发动机稳态性能优化[J]. 航天控制,2013,31(6): 17-21,49. SUI Yang,YE Zhifeng,XUE Yichun. Optimization of variable cycle engine steady-state performance based on genetic algorithm[J]. Aerospace Control,2013,31(6): 17-21,49. (in Chinese doi: 10.3969/j.issn.1006-3242.2013.06.003 SUI Yang, YE Zhifeng, XUE Yichun. Optimization of variable cycle engine steady-state performance based on genetic algorithm[J]. Aerospace Control, 2013, 31(6): 17-21, 49. (in Chinese) doi: 10.3969/j.issn.1006-3242.2013.06.003
[20]	李岩,聂聆聪,牟春晖,等. 自适应循环发动机性能智能在线寻优算法研究[J]. 推进技术,2021,42(8): 1716-1724. LI Yan,NIE Lingcong,MU Chunhui,et al. Online intelligent optimization algorithm for adaptive cycle engine performance[J]. Journal of Propulsion Technology,2021,42(8): 1716-1724. (in Chinese LI Yan, NIE Lingcong, MU Chunhui, et al. Online intelligent optimization algorithm for adaptive cycle engine performance[J]. Journal of Propulsion Technology, 2021, 42(8): 1716-1724. (in Chinese)
[21]	骆广琦,李游,刘琨,等. 变循环发动机组合变几何调节方案[J]. 航空动力学报,2014,29(10): 2273-2278. LUO Guangqi,LI You,LIU Kun,et al. Combined variable geometry regulation schemes for variable cycle engine[J]. Journal of Aerospace Power,2014,29(10): 2273-2278. (in Chinese LUO Guangqi, LI You, LIU Kun, et al. Combined variable geometry regulation schemes for variable cycle engine[J]. Journal of Aerospace Power, 2014, 29(10): 2273-2278. (in Chinese)
[22]	骆广琦,李游,吴涛,等. 基于蜂群算法的变循环发动机最小耗油率优化[J]. 航空发动机,2016,42(1): 1-5. LUO Guangqi,LI You,WU Tao,et al. The minimum fuel consumption optimization of variable cycle engine based on artificial bee colony algorithm[J]. Aeroengine,2016,42(1): 1-5. (in Chinese LUO Guangqi, LI You, WU Tao, et al. The minimum fuel consumption optimization of variable cycle engine based on artificial bee colony algorithm[J]. Aeroengine, 2016, 42(1): 1-5. (in Chinese)
[23]	LU Congwen,PAULSON J A. No-regret Bayesian optimization with unknown equality and inequality con-straints using exact penalty functions[J]. IFAC-PapersOnLine,2022,55(7): 895-902. doi: 10.1016/j.ifacol.2022.07.558
[24]	FRAZIER P I. A tutorial on Bayesian optimization[EB/OL]. [2024-01-02]. https://arxiv.org/abs/1807.02811v1.
[25]	LI Jie,FAN Ding,SREERAM V. SFC optimization for aero engine based on hybrid GA-SQP method[J]. International Journal of Turbo and Jet Engines,2013,30(4): 383-391.
[26]	肖红亮变循环航空发动机性能寻优控制技术研究[D]. 西安: 西北工业大学,2019. XIAO Hongliang. Research on performance seeking control technology of variable cycle engine[D]. Xi’an: Northwestern Polytechnical University,2019. (in Chinese XIAO Hongliang. Research on performance seeking control technology of variable cycle engine[D]. Xi’an: Northwestern Polytechnical University, 2019. (in Chinese)