Method for fault detection in an actuator

A method for fault detection in an actuator (2), in which a control value (1), generated by a control unit (11), being supplied as an input to the actuator (2) and also as an input to a modeling unit (3), an output of the actuator (2) being connected to an input of a signal conditioning unit (5), an output of the signal conditioning unit (5), being connected as a first input to a computer unit (8), and an output of the modeling unit (3) being connected as a second input to the computer unit (8), the method comprising the steps of:

• generating the control value (1) via a control unit (11);
• transmitting the control value (1) to both the actuator (2) and a modeling unit (3);
• sensing a measured actuator position (4) with a sensor of the actuator (2), the measured actuator position (4) being relative to the control value (1) and transmitting the measured actuator position (4) to the signal conditioning unit (5);
• determining a measured actuator movement (6) over a defined time interval from the measured actuator position (4), which was transmitted to the signal conditioning unit (5), and transmitting the measured actuator movement (6) to the computer unit (8) as the first input;
• calculating from the control value (1), in the modeling unit (3) without any feedback from the signal conditioning unit (5) and the computer unit (8) into the modeling unit (3) over the defined time interval, an expected actuator movement (7) and transmitting the expected actuator movement (7) to the computer unit (8) as the second input;
• summing the measured actuator movement (6) and the expected actuator movement (7) over the defined time interval between 10 and 20 calculation cycles to prevent a drifting separation of the measured actuator movement (6) and the expected actuator movement (7);
• and
• producing a status signal (9), via the computer unit (8), that relates to a difference between the measured actuator movement (6) and the expected actuator movement (7) and corresponds to a functioning status of at least one of the actuator (2) and the sensor for determining the actuator position (4).

A method for fault detection in an actuator (2) in which a control value (1), generated by a control unit (11), being applied at an input to the actuator (2) and also as an input to a modeling unit (3) for calculating an expected actuator movement (7) based upon the control value (1), the actuator (2) having a sensor for producing a signal indicative of a measured actuator position (4), and the actuator (2) outputting the signal indicative of the measured actuator position (4) to a signal conditioning unit, (5) for deriving a measured actuator movement (6) from the position signal of the sensor, the signal conditioning unit(5) sending the measured actuator movement (6) as a first input to a computer unit (8), and the expected actuator movement (7) being sent, by the modeling unit (3), as a second input to the computer unit (8), the method comprising the steps of:

• generating the control value (1) via the control unit (11);
• transmitting the control value (1) to both the actuator (2) and the modeling unit (3);
• sensing the measured actuator position (4) with the sensor of the actuator (2) with the measured actuator position (4) being relative to the control value (1), and transmitting the measured actuator position (4) to the signal conditioning unit (5);
• determining the measured actuator movement (6) over a defined time interval from the measured actuator position (4), which was transmitted to the signal conditioning unit (5), and transmitting the measured actuator movement (6) to the computer unit (8) as the first input;
• calculating from the control value (1), in the modeling unit (3), any expected actuator movement (7) without any feedback from the signal conditioning unit (5) and the computer unit (8) into from the modeling unit (3) over the defined time interval, and transmitting the expected actuator movement (7) to the computer unit (8) as the second input;
• summing the measured actuator movement (6) and the expected actuator movement (7) over the defined time interval between 10 and 20 calculation cycles for preventing a drifting separation of the measured actuator movement (6) and the expected actuator movement (7);
• and
• producing a status signal (9), via the computer unit (8), that indicates any excessive difference between the measured actuator movement (6) and the expected actuator movement (7) which corresponds to a operating status of the actuator (2) and a functionality of the sensor.