Determining a sensor error of a sensor in an exhaust gas system of a motor vehicle

1. A method for identifying a sensor fault of a sensor for an exhaust-gas system of a motor vehicle, the method comprising the steps of:

determining at least one actual sensor signal of the sensor;

determining at least one setpoint sensor signal of the sensor by way of at least one fault model; and

identifying the sensor fault of the sensor in a manner dependent on a deviation between the actual sensor signal of the sensor and the setpoint sensor signal of the sensor, wherein the at least one fault model comprises at least one of the following fault models:

(i) time delay of the actual sensor signal relative to the setpoint sensor signal in the event of a change of the combustion air ratio from a rich mixture to a lean mixture, and no time delay of the actual sensor signal relative to the setpoint sensor signal in the event of a change of the combustion air ratio from a lean mixture to a rich mixture,

(ii) time delay of the actual sensor signal relative to the setpoint sensor signal in the event of a change of the combustion air ratio from a lean mixture to a rich mixture, and no time delay of the actual sensor signal relative to the setpoint sensor signal in the event of a change of the combustion air ratio from a rich mixture to a lean mixture,

(iii) low-pass filtering of the actual sensor signal relative to the setpoint sensor signal in the event of a change of the combustion air ratio from a rich mixture to a lean mixture, and no low-pass filtering of the actual sensor signal relative to the setpoint sensor signal in the event of a change of the combustion air ratio from a lean mixture to a rich mixture,

(iv) low-pass filtering of the actual sensor signal relative to the setpoint sensor signal in the event of a change of the combustion air ratio from a lean mixture to a rich mixture, and no low-pass filtering of the actual sensor signal relative to the setpoint sensor signal in the event of a change of the combustion air ratio from a rich mixture to a lean mixture,

(v) time delay of the actual sensor signal relative to the setpoint sensor signal in the event of a change of the combustion air ratio from a lean mixture to a rich mixture, and time delay of the actual sensor signal relative to the setpoint sensor signal in the event of a change of the combustion air ratio from a rich mixture to a lean mixture, or

(vi) low-pass filtering of the actual sensor signal relative to the setpoint sensor signal in the event of a change of the combustion air ratio from a lean mixture to a rich mixture, and low-pass filtering of the actual sensor signal relative to the setpoint sensor signal in the event of a change of the combustion air ratio from a rich mixture to a lean mixture.

2. The method according to claim 1 , wherein identifying the deviation between the actual sensor signal of the sensor and the setpoint sensor signal of the sensor comprises the steps of:

determining at least one fault sensor signal of the sensor by linking the setpoint sensor signal of the sensor with the at least one fault model; and

identifying the sensor fault of the sensor in a manner dependent on at least one deviation between the actual sensor signal of the sensor and in each case one fault sensor signal of the sensor.

3. The method according to claim 2 , wherein

the at least one fault model is a parameterizable fault model and has at least one parameter, and

the identifying of the at least one deviation between the actual sensor signal of the sensor and in each case one fault sensor signal of the sensor comprises the steps of:

estimating the at least one parameter of the fault model such that the deviation between the actual sensor signal of the sensor and in each case one fault sensor signal of the sensor is minimized; and

identifying the sensor fault in a manner dependent on at least one parameter of the fault model.

4. The method according to claim 2 , the method further comprising the steps of:

providing at least two fault models for the actual sensor signal of the sensor;

determining at least two fault sensor signals of the sensor by linking the setpoint sensor signal of the sensor with in each case one fault model;

identifying the deviation between the actual sensor signal of the sensor and in each case one fault sensor signal of the sensor;

selecting one of the fault models in a manner dependent on the identified deviations between the actual sensor signal of the sensor and the respective fault sensor signal of the sensor; and

identifying the sensor fault in a manner dependent on the selected fault model.

5. The method according to claim 1 , wherein

the sensor is a lambda probe arranged in the exhaust-gas system upstream of a catalytic converter in relation to an exhaust-gas flow.

6. The method according to claim 1 , wherein

the model for determining the at least one setpoint sensor signal of the sensor is a neural network.

7. The method according to claim 1 , wherein the sensor is a lambda probe arranged in the exhaust-gas system downstream of a catalytic converter in relation to an exhaust-gas flow, wherein the method comprises the steps of:

determining at least one actual sensor signal of the sensor,

determining at least two setpoint sensor signals of the sensor by way of in each case one model, wherein each model is characteristic of a specific state of aging of the catalytic converter;

identifying in each case one deviation between the actual sensor signal of the sensor and one of the setpoint sensor signals of the sensor;

selecting one of the setpoint sensor signals of the sensor in a manner dependent on the identified deviations; and

identifying the sensor fault in a manner dependent on a deviation between the actual sensor signal of the sensor and the selected setpoint sensor signal of the sensor.

8. A device for identifying a sensor fault of a sensor in an exhaust-gas system of a motor vehicle, comprising:

a processor and associated memory configured to:

receive at least one actual sensor signal of the sensor,

determine at least one setpoint sensor signal of the sensor by way of at least one fault model, and

identify the sensor fault of the sensor in a manner dependent on a deviation between the actual sensor signal of the sensor and the setpoint sensor signal of the sensor, wherein the at least one fault model comprises at least one of the following fault models:

(i) time delay of the actual sensor signal relative to the setpoint sensor signal in the event of a change of the combustion air ratio from a rich mixture to a lean mixture, and no time delay of the actual sensor signal relative to the setpoint sensor signal in the event of a change of the combustion air ratio from a lean mixture to a rich mixture,

(ii) time delay of the actual sensor signal relative to the setpoint sensor signal in the event of a change of the combustion air ratio from a lean mixture to a rich mixture, and no time delay of the actual sensor signal relative to the setpoint sensor signal in the event of a change of the combustion air ratio from a rich mixture to a lean mixture,

(iii) low-pass filtering of the actual sensor signal relative to the setpoint sensor signal in the event of a change of the combustion air ratio from a rich mixture to a lean mixture, and no low-pass filtering of the actual sensor signal relative to the setpoint sensor signal in the event of a change of the combustion air ratio from a lean mixture to a rich mixture,

(iv) low-pass filtering of the actual sensor signal relative to the setpoint sensor signal in the event of a change of the combustion air ratio from a lean mixture to a rich mixture, and no low-pass filtering of the actual sensor signal relative to the setpoint sensor signal in the event of a change of the combustion air ratio from a rich mixture to a lean mixture,

(v) time delay of the actual sensor signal relative to the setpoint sensor signal in the event of a change of the combustion air ratio from a lean mixture to a rich mixture, and time delay of the actual sensor signal relative to the setpoint sensor signal in the event of a change of the combustion air ratio from a rich mixture to a lean mixture, or

(vi) low-pass filtering of the actual sensor signal relative to the setpoint sensor signal in the event of a change of the combustion air ratio from a lean mixture to a rich mixture, and low-pass filtering of the actual sensor signal relative to the setpoint sensor signal in the event of a change of the combustion air ratio from a rich mixture to a lean mixture.