An active-exploration-based method has been developed for an engine combusting fuels with occasional rapid changes in a mixed ratio of various fuel types. To apply learning-based control to the engine, training data should be acquired and combustion models should be generated during actual operation. Occasional rapid changes in fuels lead to a small amount of data for each fuel type and ratio, which is susceptible to noise and results in slow convergence in exploring training data. Furthermore, slow convergence leads to extended control using no-longer appropriate parameters, causing faults. The method reduces the negative effect of noise by discretizing the combustion state space and naturally avoids engine faults by predicting the no-fault condition (i.e. stable combustion with higher thermal efficiency). Numerical experiments of typical optimization problems showed that the proposed method is robust against noise during the optimization process. By applying the method to an actual engine combusting a changing mixture of hydrogen and ethanol, thermal efficiency after combustion-model generation increased up to 7.9% and the coefficient of variation of the indicated mean effective pressure was less than 3% for all ratios. The increase in thermal efficiency corresponds to continuous predictions of the no-faults condition. These results demonstrate the potential practical use of engines with occasional rapid changes in mixed ratios of various fuel types during operation.
ShimadaAShirakawaYIshikawaT. Improved thermal efficiency using hydrous ethanol reforming in advanced spark-ignition engines. SAE technical paper 2016-01-2262, 2016.
2.
AkansuSODulgerZKahramanNVeziroğluTN. Internal combustion engines fueled by natural gas—hydrogen mixtures. Int J Hydrogen Energy2004; 29(14): 1527–1539.
3.
YaodongHZhouTYaoCYangFWangJOuyangM, et al. Accommodation closed-loop control of dieseline fueled flexible fuel engine based on in-cylinder pressure sensor. IFAC-PapersOnLine2016; 49(11): 202–209.
4.
ZhaoJWangJ. Energy-based and oxygen-based biodiesel blend level estimation methods for diesel engines. In: American control conference, Montreal, 2012, pp.4975–4980.
5.
MirheidariSMohammadpourJGrigoriadKM. Biodiesel blend estimation based on fuel consumption and engine power. In: American control conference, Baltimore, 2010, pp.3021–3026.
6.
BishopJDKStettlerMEJMoldenNBoiesAM.Engine maps of fuel use and emissions from transient driving cycles. Appl Energy2016; 183(1): 202–217.
7.
AmorimRJBaetaJGCValleRMBarrosJCarvalhoR. Experimental analyses of flexible fuel systems in spark ignition engine. SAE technical paper 2005-01-2183, 2005.
8.
MalcolmJSAleiferisPGCairnsAToddARHumeABlaxillH, et al. A study of alcohol blended fuels in a new optical spark-ignition engine. In: International conference on internal combustion engines: performance, fuel economy and emissions. London: IMechE, 2007, pp.223–234.
9.
HsiehWDChenRHWuTLLinTH. Engine performance and pollutant emission of an SI engine using ethanol–gasoline blended fuels. Atmos Environ2002; 36(3): 403–410.
10.
TogunNBaysecSKaraT. Nonlinear modeling and identification of a spark ignition engine torque. Mech Syst Sig Process2012; 26: 294–304.
11.
WongPKTangLMVongCM. Engine idle-speed system modelling and control optimization using artificial intelligence. Proc Inst Mech Eng Part D: J Automobile Eng2010; 224(1): 55–72.
12.
Rahimi-GorjiMGhajarMKakaeeAHGanjiDD. Modeling of the air conditions effects on the power and fuel consumption of the SI engine using neural networks and regression. J Braz Soc Mech Sci Eng2017; 39(2): 375–384.
13.
SayinCErtuncHMHosozMKilicaslanICanakciM. Performance and exhaust emissions of a gasoline engine using artificial neural network. Appl Therm Eng2007; 27(1): 46–54.
14.
ZhaiYJYuDLTafreshiRAl HamidiY. Fast predictive control for air-fuel ratio of SI engines using a nonlinear internal model. Int J Eng Sci Technol2011; 3(6): 1–17.
15.
ColinGChamaillardYBlochGCordeG. Neural control of fast nonlinear systems—application to a turbocharged SI engine with VCT. IEEE Trans Neural Networks2007; 18(4): 1101–1114.
16.
WongKLVongCMWongPKLuoJ. Sparse Bayesian extreme learning machine and its application to biofuel engine performance prediction. Neurocomputing2015; 149: 397–404.
17.
WuBFilipiZAssanisDKramerDMOhlGLPruckaMJ, et al. Using artificial neural networks for representing the air flow rate through a 2.4 liter VVT engine. SAE Trans2004; 113: 1676–1686.
18.
WuBPruckaRGFilipiZSKramerDMOhlGL. Cam-phasing optimization using artificial neural networks as surrogate models—maximizing torque output. SAE Trans2005; 114: 1586–1599.
19.
BeltramiCChamaillardYMilleriouxGHigelinPBlochG. AFR control in SI engine with neural prediction of cylinder air mass. In: American control conference, Denver, 2003, pp.1404–1409.
20.
MagnerSJankovicM. Delta air-charge anticipation for mass air flow and electronic throttle control based systems. In: American control conference, Anchorage, 2002, pp.1407–1412.
21.
MajorsMStoriJDiC. Neural network control of automotive fuel-injection systems. IEEE Control Syst Mag1994; 14(3): 31–36.
YanRLiHDongZYTangH. Nonlinear control approaches for SI engine model with uncertainties. In: The 48h IEEE conference on decision and control (CDC) held jointly with 2009 28th Chinese control conference, Shanghai, 2009, pp.5440–5445.
24.
HowellMNBestMC. On-line PID tuning for engine idle-speed control using continuous action reinforcement learning automata. Control Eng Pract2000; 8(2): 147–154.
25.
TangHWengLDongZYYanR. Adaptive and learning control for SI engine model with uncertainties. IEEE/ASME Trans Mechatron2009; 14(1): 93–104.
26.
SaraswatiSChandS. Online linearization-based neural predictive control of air–fuel ratio in SI engines with PID feedback correction scheme. Neural Comput2010; 19(6): 919–933.
27.
SlepickaCKochCR. Iterative learning on dual-fuel control of homogeneous charge compression ignition. IFAC-PapersOnLine2016; 49(11): 347–352.
28.
WongPKWongHCVongCMXieZHuangS. Model predictive engine air-ratio control using online sequential extreme learning machine. Neural Comput Appl2016; 27(1): 79–92.
29.
ZhangYShenT.On-board map learning-based combustion phase control in spark ignition engines. In: IEEE conference on control technology and applications, Mauna Lani, 2017, pp.287–292.
30.
MohammadiAManzieCNešićD. Online optimization of spark advance in alternative fueled engines using extremum seeking control. Control Eng Pract2014; 29: 201–211.
31.
GuardiolaCPlaBBlanco-RodriguezDCabreraP. A learning algorithm concept for updating look-up tables for automotive applications. Math Comput Modell2013; 57(7–8): 1979–1989.
32.
ZhangYGaoJShenT. Probabilistic guaranteed gradient learning-based spark advance self-optimizing control for spark-ignited engines. IEEE Trans Neural Networks Learn Syst2018; 29(10): 4683–4693.
33.
ThrunSBMöllerK.Active exploration in dynamic environments. In: Neural information processing systems, Denver, 1991, pp.531–538.
34.
BaranesAOudeyerPY. Intrinsically motivated goal exploration for active motor learning in robots: a case study. In: IEEE/RSJ international conference on intelligent robots and systems, Taipei, 2010, pp.1766–1773.
35.
NgoHLuciwMForsterASchmidhuberJ. Learning skills from play: artificial curiosity on a Katana robot arm. In: International joint conference on neural networks, Brisbane, 2012, pp.1–8.
36.
PapeLOddoCMControzziMCiprianiCFörsterACarrozzaMC, et al. Learning tactile skills through curious exploration. Front Neurorobot2012; 6: 6.
37.
BaranesAOudeyerPY. Active learning of inverse models with intrinsically motivated goal exploration in robots. Rob Auton Syst2013; 61(1): 49–73.
38.
KahnGSujanPPatilSBopardikarSRydeJGoldbergK, et al. Active exploration using trajectory optimization for robotic grasping in the presence of occlusions. In: IEEE international conference on robotics and automation, Seattle, 2015, pp.4783–4790.
39.
DelmericoJMuegglerENitschJScaramuzzaD. Active autonomous aerial exploration for ground robot path planning. IEEE Rob Autom Lett2017; 2(2): 664–671.
40.
AndersenGKonidarisG. Active exploration for learning symbolic representations. In: Neural information processing systems, Long Beach, 2017, pp.5009–5019.
41.
Martinez-HernandezUDoddTJPrescottTJ. Feeling the shape: active exploration behaviors for object recognition with a robotic hand. IEEE Trans Syst Man Cybern Syst2018; 48(12): 2339–2348.
42.
JinYLeeM. Enhancing binocular depth estimation based on proactive perception and action cyclic learning for an autonomous developmental robot. IEEE Trans Syst Man Cybern Syst2019; 49(1): 169–180.
43.
EsakiKShirakawaYNaitoHMatsumotoKItoKIshikawaT. Engine controller using implicit fault-avoidance learning of control parameters for mixed-fuel combustion. In: 44th Annual conference of the IEEE industrial electronics society, Washington, DC, 2018, pp.2311–2316.
44.
BackTHammelU.Evolution strategies applied to perturbed objective functions. In: IEEE conference on evolutionary computation. IEEE world congress on computational intelligence, Orlando, 1994, pp.40–45.
45.
AvrielM.Nonlinear programming: analysis and methods. Mineola: Dover Publications, 2003.
46.
NocedalJWrightS. Numerical optimization. New York: Springer, 2006.
47.
NelderJAMeadR. A simplex method for function minimization. Comput J1965; 7(4): 308–313.
48.
WrightMH. Direct search methods: once scorned, now respectable. In GriffithsDFWatsonGA (eds) Numerical analysis. Harlow: Addison-Wesley, 1996, pp.191–208.
