Estimation of the air-fuel mixture ratio, based on the signal from the optical fibre interference sensor, using artificial neuron network

• Autorzy: Kamiński T.
• Języki publikacji: EN

Abstrakty

EN

There are a lot of definitions for the "intelligence", however according to the Prof. Jan Strelau the intelligence can only be attributed to a human. Thus, programs or devices thought out by a man can only imitate intelligence. The phras - artificial neuron networks describes programs or electronic devices running mathematical models of pseudo parallel data processing, consisting of many interconnected neurons imitating actions of biological structures of a brain. The neuron networks are used, amongst other things, for the sound and picture recognition, for the predicting, objects classification, data analysis, matching and optimisation. This work describes construction of an artificial neuron network in use. The design and operation principles of a fibre optics interference side - hole sensor are presented for the pressure measurement inside the engine combustion chamber and the data range used during a „teaching" process of an artificial neuron network. The gas pressure in the engine combustion chamber carries a lot of information which can be used to characterize the working cycle process. Knowing the pressure curve it is possible to estimate the air-fuel mixture ratio, detect lack of mixture ignition in the cylinder, explosive knocking combustion, unevenness of the following engine working cycles and even estimate the combustion chamber walls' temperature. Construction of a sensor ensures high pressure sensitivity with a low temperature sensitivity. The measuring head has been placed in the threaded opening made in the engine cylinder head. The paper presents results for two examples of the air-fuel mixture ratio estimations using already developed network. In the first example the measurement data, used during the network educating process, has deliberately been changed at random by 3%. In the second case, the original measurement data has been used. This allowed the initial assessment of the measurement noise influence on the mixture content estimation using fibre optics pressure sensor, in the combustion chamber, combined with the artificial neuron networks. The mixture content estimation results, together with the data obtained during measurements using wide range oxygen sensor, which are presented on the diagram, allowed the conclusions to be formulated.

Słowa kluczowe

EN

transport; fibre optics sensor; neuron networks

• Wydawca: Łukasiewicz Research Network - Institute of Aviation ; European Science Society of Powertrain and Transport KONES Poland ; Wydawnictwo Instytutu Technicznego Wojsk Lotniczych
• Czasopismo: Journal of KONES
• Rocznik: 2008
• Tom: Vol. 15, No. 2
• Strony: 153--158
• Opis fizyczny: Bibliogr. 9 poz., rys.

Twórcy

autor

Kamiński T.

• Motor Transport Institute Jagiellońska Street 80, 03- 301 Warszawa tel.: [phone] ext. 129, fax.: [phone],

Bibliografia

• [1] Arsie, I., Flauti, G., Pianese, C., Rizzo, G., Cylinder Thermal State Detection from Pressure Cycle in SI Engine, Proc. of 4th International Conference on Internal Combustion Engines: Experiments and Modeling, Capri, pp. 581-588, ISBN 88-900399-06, sept. 12-16, 1999.
• [2] Dues, S., Adams, J., Shinkle, G, Combustion Knock Sensing. Sensor Selection and Application Issues, SAE Technical Paper No 900488, pp. 93 - 103, 1990.
• [3] Kamiski, T., Mitraszewska, I., Nowacki, G., Wendeker, M., Wojciechowski, A., Godula, A., Metrology Properties of Optical Fibre Sensor of The Pressure of The Side - Hole Type, Materiały XXXII Międzynarodowego Kongresu Naukowego Napędów i Środków Transportu (KONES 2006), Nałęczów 2006.
• [4] Kamiński, T., Wendeker, M., Koncepcja światłowodowego czujnika side - hole do pomiaru ciśnienia w komorze spalania silnika spalinowego, Journal of KONES, Vol. 9, No. 3-4, Warszawa - Gdynia 2002.
• [5] Muller, N., Isermann, R., Control of Mixture Composition using Cylinder Pressure Sensors, SAE Technical Paper No 01ATT-283, SAE, 2001.
• [6] Nawrocka, M., Dysertacja nt. światłowodowy czujnik interferencyjny do pomiaru szybkich zmian ciśnienia. Politechnika Wrocławska, Instytut Fizyki, Wrocław 2001.
• [7] Pineda, F., Generalization of backpropagation to reccurent neural networks, Phys. Rev. Letters, Vol. 18, 1987.
• [8] Sellnau, M. C., Matekunas, F. A., Battiston, P. A., Chang, C. F., Lancaster, D., Cylinder- Pressure-Based Engine Control Using Pressure-Ratio-Management and Low-Cost Non-Intrusive Cylinder Pressure Sensors, SAE, Paper No. 2000-01-0932, 2000.
• [9] Tadeusiewicz, R., Sieci neuronowe, Oficyna Wydawnicza Politechniki Warszawskiej, Warszawa 1994.