Use of machine learning for detecting cylinder intake and/or exhaust valve faults during operation of an internal combustion engine

Patent metadata

Field	Value
Publication number	US-11434839-B2
Application number	US-202017137955-A
Country	US
Kind code	B2
Filing date	Dec 30, 2020
Priority date	Dec 30, 2020
Publication date	Sep 6, 2022
Grant date	Sep 6, 2022

Abstract

A system and method for the use of machine learning for detecting faults for cylinder intake and/or exhaust valves that do not properly open or close as commanded and for generating a flag for such faults.

First claim

What is claimed is:

1. An engine system for a vehicle, comprising:
an internal combustion engine having a plurality of cylinders; and
an engine system controller configured to:
operate the internal combustion engine in a skip fire mode where at least one cylinder is fired, skipped and either fired or skipped over successive working cycles while the internal combustion engine is operating at an effective reduced displacement that is less than full displacement of the internal combustion engine;
generate a command to either open or close a valve associated with a select cylinder, selected among the plurality of cylinders, for a given working cycle while the internal combustion engine is operating in the skip fire mode, the valve commanded to be either opened or closed during the given working cycle for any of the following purposes:
(i) the select cylinder is commanded to be fired during the given working cycle while the internal combustion engine is operating in the skip fire mode;
(ii) the select cylinder is commanded to be skipped during the given working cycle while the internal combustion engine is operating in the skip fire mode; and
(iii) for controlling air pressure within a combustion chamber of the select cylinder when skipped during the given working cycle;
feed one or more signals indicative of one or more operational parameters of the vehicle to a model, the model specifying weighted variables for the one or more operational parameters of the vehicle respectively, the weighted variables of the model derived from machine training;
compare the command to an output of the model, the output of the model being a probability of a behavior of the valve associated with the select cylinder during the given working cycle; and
generate a valve fault flag if the comparison yields that the probability of the behavior of the valve does not match the command.

2. The engine system of claim 1 , wherein the model is further configured to generate the probability by:
generating an estimated score for the behavior of the valve associated with the select cylinder during the given working cycle by propagating the one or more operational parameters of the vehicle through the model; and
generating the probability of the behavior of the valve by applying a sigmoid function to the estimated score;
wherein the probability of the behavior of the valve is indicative of either a first state indicating that the valve opened during the given working cycle or a second state indicative that the valve remained closed during the given working cycle.

3. The engine system of claim 1 , wherein the fault flag is generated if the command was to open the valve during the valve event, but the probability indicates that the valve remained closed during the given working cycle.

4. The engine system of claim 1 , wherein the fault flag is generated if the command was to close the valve during the given working cycle, but the probability indicates that the valve opened during the given working cycle.

5. The engine system of claim 1 , wherein the engine system controller is further configured to generate a success flag if the command was to open the valve during the given working cycle and the probability indicates that the valve opened during the given working cycle.

6. The engine system of claim 1 , wherein the engine system controller is further configured to generate a success flag if the command was to close the valve and the probability indicates that the valve remained closed during the given working cycle.

7. The engine system of claim 1 , wherein the operational parameters include one of the following:
(a) exhaust manifold pressure;
(b) intake manifold pressure;
(c) speed of the internal combustion engine;
(d) brake torque of the internal combustion engine;
(e) an indicated torque estimate;
(f) a vane position of a Variable Turbo Geometry (VTG) of a turbocharger associated with the internal combustion of the vehicle;
(g) Exhaust EGR valve position;
(h) charge flow;
(i) previous cylinder status;
(j) next cylinder status; or
(k) any combination of (a) through (j).

8. The engine system of claim 1 , wherein the internal combustion engine is one of the following:
(a) a Diesel engine;
(b) a gasoline engine;
(c) a compression-ignition engine;
(d) a spark-ignition engine.

9. The engine system of claim 1 , wherein the engine system controller is further configured to operate the select cylinder as an Air Spring (AS) when the select cylinder is skipped during the given working cycle.

10. The engine system of claim 1 , wherein the engine system controller is further configured to operate the select cylinder as a Low-Pressure Exhaust Spring (LPES) when the select cylinder is skipped during the given working cycle.

11. The engine system of claim 1 , wherein the engine system controller is further configured to operate the select cylinder as a High-Pressure Exhaust Spring (HPES) when the select cylinder is skipped during the given working cycle.

12. The engine system of claim 1 , wherein the engine system controller controls the air pressure within the combustion chamber of the select cylinder by commanding the valve to either open or close for the purpose of inducting an air-charge into the combustion chamber of the select cylinder when skipped during the given working cycle.

13. The engine system of claim 1 , wherein the engine system controller controls the air pressure within the combustion chamber of the select cylinder by commanding the valve to either open or close for the purpose of maintaining pressure within a desired pressure range in the combustion chamber of the select cylinder when skipped during the given working cycle.

14. The engine system of claim 13 , wherein the desired pressure range further comprises one of the following:
(a) a desired low-pressure range; or
(b) a desired high-pressure range.

15. The engine system of claim 1 , wherein the valve associated with the select cylinder is an intake valve.

16. The engine system of claim 1 , wherein the valve associated with the select cylinder is an exhaust valve.

17. The engine system of claim 1 , wherein the engine system controller is further configured to generate multiple commands to either open or close a multiplicity of valves associated with the plurality of cylinders respectively, the multiple commands configured to separately and individually control the multiplicity of valves to either open or close for any of the purposes (i), (ii), and (iii) for each of the plurality of cylinders during operation of the internal combustion engine in the skip fire mode respectively.

18. The engine system of claim 1 , wherein the engine system controller is further configured to operate the internal combustion engine in a dynamic skip fire mode wherein decisions to either fire or skip each of the plurality of cylinders is dynamically made either (i) on a firing opportunity-by-firing opportunity basis or (ii) on an engine cycle-by-engine cycle basis.

19. The engine system of claim 1 , wherein the select cylinder includes an intake valve and an exhaust valve, and the engine system controller is further configured to separately and independently command the intake valve and the exhaust valve to be either opened or closed during the given working cycle,
wherein the valve associated with the select cylinder can be either the intake valve or the exhaust valve.

20. The engine system of claim 1 , wherein the output of the model is derived from the one or more signals

Assignees

• Tula Technology Inc

Inventors

• Chen Shikui Kevin
• Mandal Aditya
• Serrano Louis J
• Cai Xiaoping

Classifications

• G06N3/08

  Learning methods · CPC title

• G06N3/0499

  Feedforward networks · CPC title

• G06N3/09

  Supervised learning · CPC title

• F01N2430/02

  by cutting out a part of engine cylinders · CPC title

• F02D41/0087 Primary

  Selective cylinder activation, i.e. partial cylinder operation (deceleration cut-off F02D41/123) · CPC title