Systems and methods for operating a vehicle that includes an engine and an electric machine are described. In one example, a speed of the engine may be adjusted so that the engine provides power to drive the electric machine without generating numerous rapid engine speed changes in a short amount of time.

Systems and methods for operating a vehicle that includes an engine and an electric machine are described. In one example, a speed of the engine may be adjusted so that the engine provides power to drive the electric machine without generating numerous rapid engine speed changes in a short amount of time.

An estimation method to determine the concentration of recirculated exhaust gas present in a cylinder of an internal combustion engine; the concentrations of recirculated exhaust gas in a gas mixture flowing through an intake duct are periodically stored in a buffer; a first instant is determined, in which a programming of a following combustion in the cylinder is carried out; an advance time is determined, which elapses between the first instant and a second instant in the future, in which air will be taken into the cylinder for the following combustion in the cylinder; a transport time is determined; a third instant in the past is determined by subtracting from the first instant an amount of time which is equal to the difference between the transport time and the advance time; and the concentration of recirculated exhaust gas present in the cylinder in the second instant is estimated depending on a concentration of recirculated exhaust gas contained in the buffer (30) and corresponding to the third instant.

An estimation method to determine the concentration of recirculated exhaust gas present in a cylinder of an internal combustion engine; the concentrations of recirculated exhaust gas in a gas mixture flowing through an intake duct are periodically stored in a buffer; a first instant is determined, in which a programming of a following combustion in the cylinder is carried out; an advance time is determined, which elapses between the first instant and a second instant in the future, in which air will be taken into the cylinder for the following combustion in the cylinder; a transport time is determined; a third instant in the past is determined by subtracting from the first instant an amount of time which is equal to the difference between the transport time and the advance time; and the concentration of recirculated exhaust gas present in the cylinder in the second instant is estimated depending on a concentration of recirculated exhaust gas contained in the buffer (30) and corresponding to the third instant.

Methods and systems are provided for detecting cylinder-to-cylinder air-fuel ratio (AFR) imbalance in engine cylinders. In one example, a method may include detecting an AFR imbalance of an engine cylinder based on an individual crankshaft acceleration of the cylinder relative to a mean crankshaft acceleration produced by all cylinders of the engine, and correcting a fuel amount of the cylinder via a fuel multiplier value, the fuel multiplier value selected from a plurality of fuel multiplier values based on an imbalance source. In this way, the AFR imbalance may be accurately detected and correcting using existing engine system sensors.

Methods and systems are provided for detecting cylinder-to-cylinder air-fuel ratio (AFR) imbalance in engine cylinders. In one example, a method may include detecting an AFR imbalance of an engine cylinder based on an individual crankshaft acceleration of the cylinder relative to a mean crankshaft acceleration produced by all cylinders of the engine, and correcting a fuel amount of the cylinder via a fuel multiplier value, the fuel multiplier value selected from a plurality of fuel multiplier values based on an imbalance source. In this way, the AFR imbalance may be accurately detected and correcting using existing engine system sensors.

A controller according to the present disclosure, in each combustion cycle that composes a change cycle, calculates the average .sub.n of control amounts from the first combustion cycle to the nth (1<=n<=N) combustion cycle and calculates the error .sub.n-.sub.o of the average .sub.n with respect to the average .sub.o of a reference normal population. Also, the controller sets both a positive threshold Z.sub./2*.sub.o/n.sup.1/2 and a negative threshold Z.sub./2*.sub.o/n.sup.1/2 based on the standard error .sub.o/n.sup.1/2 of the reference normal population in the case where the number of data is n. Then, the controller chooses an operation amount to be changed from a plurality of operation amounts, based on a comparison between a series of the errors .sub.n-.sub.o and a series of the positive thresholds Z.sub./2*.sub.o/n.sup.1/2 and a comparison between the series of the errors .sub.n-.sub.o and a series of the negative thresholds Z.sub./2*.sub.o/n.sup.1/2.

A controller according to the present disclosure, in each combustion cycle that composes a change cycle, calculates the average .sub.n of control amounts from the first combustion cycle to the nth (1<=n<=N) combustion cycle and calculates the error .sub.n-.sub.o of the average .sub.n with respect to the average .sub.o of a reference normal population. Also, the controller sets both a positive threshold Z.sub./2*.sub.o/n.sup.1/2 and a negative threshold Z.sub./2*.sub.o/n.sup.1/2 based on the standard error .sub.o/n.sup.1/2 of the reference normal population in the case where the number of data is n. Then, the controller chooses an operation amount to be changed from a plurality of operation amounts, based on a comparison between a series of the errors .sub.n-.sub.o and a series of the positive thresholds Z.sub./2*.sub.o/n.sup.1/2 and a comparison between the series of the errors .sub.n-.sub.o and a series of the negative thresholds Z.sub./2*.sub.o/n.sup.1/2.

A control device of an internal combustion engine is configured to output a predicted value of an output parameter by using a learning model if actually measured values of input parameters are input, control the internal combustion engine based on the predicted value of the output parameter, learn the learning model by using a gradient method and by using a combination of actually measured values of the input parameters and an actually measured value of the output parameter as teacher data, and adjust the learning rate so that the learning is performed by a smaller learning rate when an amount of noise superposed on an actually measured value of at least one parameter among the input parameters and the output parameter is relatively large compared with when the amount of noise superposed on the actually measured value of the parameter is relatively small.

A control device of an internal combustion engine is configured to output a predicted value of an output parameter by using a learning model if actually measured values of input parameters are input, control the internal combustion engine based on the predicted value of the output parameter, learn the learning model by using a gradient method and by using a combination of actually measured values of the input parameters and an actually measured value of the output parameter as teacher data, and adjust the learning rate so that the learning is performed by a smaller learning rate when an amount of noise superposed on an actually measured value of at least one parameter among the input parameters and the output parameter is relatively large compared with when the amount of noise superposed on the actually measured value of the parameter is relatively small.