A neuro-fuzzy based combustion sensor for the control of optimal engine combustion efficiency

Thermal Science

关键词: neuro-fuzzy;    combustion sensor;    internal combustion engine efficiency;    Lolimot;    MFB50;    spark advance;   

DOI  :  10.2298/TSCI120703160M

来源: DOAJ

【 摘 要 】

Modern and advanced control systems for internal combustion engines requireaccurate feedback information from the combustion chamber. Whereas thein-cylinder pressure sensor provides this information through its closethermodynamic ties with the combustion process, drawbacks in itsimplementation push research towards other nonintrusive sensing methods. Thispaper suggests alternative methods of combustion phasing detection relying onmeasured angular crankshaft speed. Method developed, achieves sensing ofangular position of the 50% of mass fraction burned (MFB50) through twosteps: calculation of, so called, synthetic torque and its nonlineartransformation to a combustion feature estimator through local linearNeuro-fuzzy based model (LLNFM). In order to calibrate both parts of thisvirtual combustion sensor, parameters of a high-fidelity crankshaft dynamicmodel are identified, and LLNF model is trained with extensive experimentallycollected data set. Created virtual MFB50 sensor, demonstrated itsperformance, on a large test data set comprised of 70% of gathered data.

【 授权许可】

Unknown