公开(公告)号：US10132260B2

公开(公告)日：2018-11-20

申请号：US15055134

申请日：2016-02-26

Applicant: GM Global Technology Operations LLC

Inventor： Jay S. Meldrum ,  Gary Robert Cygan, Jr. ,  Layne K. Wiggins ,  Bastian Maass

IPC: F02D41/18 ,  F02D41/26 ,  G01M15/05 ,  G01M15/09 ,  F02D41/14 ,  F02D13/02 ,  F02D41/00

Abstract: Computational models and calculations relating to trapped and scavenged air per cylinder (APC) improve scavenging and non-scavenging operational modes of internal combustion engines as well as the transition there-between. Data from sensors which include engine speed, manifold air pressure, barometric pressure, crankshaft position, and valve state are provided to a pair of artificial neural networks. A first neural network utilizes this data to calculate the nominal volume of gas, i.e., air trapped in the cylinder. A second neural network utilizes this data to calculate the trapping ratio. The output of the first network is utilized with the ideal gas law to calculate the actual mass of trapped APC. The actual mass of trapped APC is also divided by the trapping ratio calculated by the second network to determine the total APC and is further utilized to calculate the scavenged APC by subtracting the trapped APC from the total APC.

公开(公告)号：US20170248093A1

公开(公告)日：2017-08-31

申请号：US15055134

申请日：2016-02-26

Applicant: GM Global Technology Operations LLC

Inventor： Jay S. Meldrum ,  Gary Robert Cygan, JR. ,  Layne K. Wiggins ,  Bastian Maass

IPC: F02D41/18 ,  G01M15/05 ,  G01M15/09 ,  F02D41/26

CPC classification number: F02D41/18 ,  F02D13/0261 ,  F02D41/0007 ,  F02D41/1405 ,  F02D41/263 ,  F02D2041/001 ,  F02D2200/0402 ,  F02D2200/0406 ,  F02D2200/0411 ,  G01M15/05 ,  G01M15/09 ,  Y02T10/144 ,  Y02T10/18

Abstract: Computational models and calculations relating to trapped and scavenged air per cylinder (APC) improve scavenging and non-scavenging operational modes of internal combustion engines as well as the transition there-between. Data from sensors which include engine speed, manifold air pressure, barometric pressure, crankshaft position, and valve state are provided to a pair of artificial neural networks. A first neural network utilizes this data to calculate the nominal volume of gas, i.e., air trapped in the cylinder. A second neural network utilizes this data to calculate the trapping ratio. The output of the first network is utilized with the ideal gas law to calculate the actual mass of trapped APC. The actual mass of trapped APC is also divided by the trapping ratio calculated by the second network to determine the total APC and is further utilized to calculate the scavenged APC by subtracting the trapped APC from the total APC.