Mutual information of cylinder pressure and combustion phase estimation in spark ignition engines

关键词:  Information entropy, mutual information, combustion phase estimation, in-cylinder pressure, extreme learning machine

Key words:  Information entropy, mutual information, combustion phase estimation, in-cylinder pressure, extreme learning machine

摘要:

For the study of internal combustion engines, combustion control is an important method to achieve high efficiency and low emissions. Currently, in-cylinder pressure sensor-based closed-loop control strategies have become the preferred solution. However, their productional application in automotive industries is limited due to the cost of intensive pressure acquisition for a whole cycle and the calculation load of combustion phase indicators. This paper proposes a method of combustion phase estimation for spark ignition (SI) engines. In this method, the combustion phase is estimated only based on pressure measurements at several crank angles. Information entropy and mutual information are introduced to analyze the feasibility and accuracy of the combustion phase estimation, which shows that the pressure measurements at selected points contain most of the information for the estimation. As a result, only pressure measurements at 3 points and ELM estimation models are required to obtain the combustion phase, instead of intensive data acquisition and calculation.

Abstract:

For the study of internal combustion engines, combustion control is an important method to achieve high efficiency and low emissions. Currently, in-cylinder pressure sensor-based closed-loop control strategies have become the preferred solution. However, their productional application in automotive industries is limited due to the cost of intensive pressure acquisition for a whole cycle and the calculation load of combustion phase indicators. This paper proposes a method of combustion phase estimation for spark ignition (SI) engines. In this method, the combustion phase is estimated only based on pressure measurements at several crank angles. Information entropy and mutual information are introduced to analyze the feasibility and accuracy of the combustion phase estimation, which shows that the pressure measurements at selected points contain most of the information for the estimation. As a result, only pressure measurements at 3 points and ELM estimation models are required to obtain the combustion phase, instead of intensive data acquisition and calculation.