System for detecting abnormality in blow-by gas feed path in internal combustion engine
AssigneeToyota Jidosha Kabushiki Kaisha
InventorsHarufumi Muto, Naoya Okubo, Akihiro Katayama, Yosuke Hashimoto
Life of this patent
Estimating the term…
Abstract
A learned neural network learned in weights using an engine load, an engine speed, and an intake pressure inside the engine intake passage downstream of the throttle valve (19) as input parameters of the neural network and using leakage of blow-by gas from a blow-by gas feed path (20) as a truth label is stored. At the time of operation of the vehicle, the learned neural network is used to detect the abnormality of leakage of blow-by gas from the blow-by gas feed path (20) from the above input parameters.
Claims7
A system for detecting an abnormality in a blow-by gas feed path in an internal combustion engine, comprising:
- an intake air amount detector arranged in an engine intake passage,
- a throttle valve arranged in the engine intake passage downstream of the intake air amount detector, blow-by gas in an engine crankcase being fed through the blow-by gas feed path into the engine intake passage downstream of the throttle valve, and
- a PCV valve arranged in the blow-by gas feed path to control an amount of flow of blow-by gas, wherein
- a learned neural network learned in weights using an engine load, an engine speed, and an intake pressure in the engine intake passage downstream of the throttle valve as input parameters of the neural network and using an abnormality of a leakage of blow-by gas from the blow-by gas feed path and an abnormality of a stuck opening of the PCV valve continuing to remain open as truth labels respectively is stored, and
- the learned neural network is used to detect the abnormality of the leakage of blow-by gas from the blow-by gas feed path and the abnormality of the stuck opening of the PCV valve from the input parameters.
A system for detecting an abnormality in a blow-by gas feed path in an internal combustion engine, comprising:
- an intake air amount detector arranged in an engine intake passage,
- a throttle valve arranged in the engine intake passage downstream of the intake air amount detector, blow-by gas in an engine crankcase being fed through the blow-by gas feed path into the engine intake passage downstream of the throttle valve, and
- a PCV valve arranged in the blow-by gas feed path to control an amount of flow of blow-by gas, wherein
- a learned neural network learned in weights using an engine load, an engine speed, a measured value of an intake pressure in the engine intake passage downstream of the throttle valve, and an estimated value of the intake pressure in the engine intake passage downstream of the throttle valve as input parameters of the neural network and using an abnormality of a leakage of blow-by gas from the blow-by gas feed path as a truth label is stored, and
- the learned neural network is used to detect the abnormality of the leakage of blow-by gas from the blow-by gas feed path from the input parameters.
A system for detecting an abnormality in a blow-by gas feed path in an internal combustion engine, comprising:
- an intake air amount detector arranged in an engine intake passage,
- a throttle valve arranged in the engine intake passage downstream of the intake air amount detector, blow-by gas in an engine crankcase being fed through the blow-by gas feed path into the engine intake passage downstream of the throttle valve, and
- a PCV valve arranged in the blow-by gas feed path to control an amount of flow of blow-by gas, wherein
- a learned first neural network learned in weights using an engine load, an engine speed, and an intake pressure in the engine intake passage downstream of the throttle valve at a time of engine low load steady operation as input parameters of the first neural network and using an abnormality of a leakage of blow-by gas from the blow-by gas feed path as a truth label is stored,
- a learned second neural network learned in weights using an engine load, an engine speed, and an intake pressure in the engine intake passage downstream of the throttle valve at a time of engine medium load steady operation as input parameters of the second neural network and using an abnormality of a stuck closing of the PCV valve remaining closed as a truth label is stored,
- at the time of engine low load steady operation, the learned first neural network is used to detect the abnormality of the leakage of blow-by gas from the blow-by gas feed path from the input parameters at the time of engine low load steady operation, and
- at the time of engine medium load steady operation, the learned second neural network is used to detect the abnormality of the stuck closing of the PCV valve from the input parameters at the time of engine medium load steady operation.
Dependencies are inferred from the claim text.