Indonesia Conference Directory

The main equipment in the transportation system is a driving element, such as a DC motor on an electric car. Sensors and actuators used in DC motors are components that are directly related to the environment so that both components are susceptible to damage. Currently the field of Fault Tolerant Control (FTC) has developed, which is a controlling system that can tolerate errors that occur and maintain system performance as desired. One of the FTC methods is Passive Fault Tolerant Control (PFTC). The PFTC method is designed as a robust controlling system from a component error. Robust control works by taking into uncertainties and disturbances in the system. In this case the component error is seen as an unobserved disturbance. However, robust control assumes the measurement data is in accordance with the real conditions, or there is no error in the measuring instrument or sensor. So that when there is an error in the measurement, robust control is not able to maintain the stability and performance specified. In addition, the operating conditions of DC motors with varying load torque exacerbate the DC motor speed response because load torque variations are considered a disturbance that must be muted if using a robust control scheme. Therefore, the control algorithm should be derived from a model involving load torque. However, this requires information about the real condition of the torque load. This research proposes a robust control strategy that is able to resolve actuator errors and sensor errors in the PFTC scheme on DC motor servo systems. The variable that is controlled is the speed with the changing torque load. The control structure used is state feedback with the integrator. To accommodate errors in sensors and actuators, these controllers are equipped with an observer that has a robust characters. Information about changes in load torque is obtained not from the measurement of torque, but through estimation so that the model used by the observer must accommodate this problem. Determination of gain observer is done by linear matrix inequality (LMI) formulation and is treated as a multiobjective optimization problem. PFTC system testing is carried out both simulation and experimental by providing an offset error. The bias error in the actuator can be a disturbance that blocks the current from the servo amplifier (as an actuator) from entering the DC motor. Error bias in speed sensor occurs in tachogenerator. The PFTC system is tested in real time on the MS150DC modular servo system. The PFTC system is better able to maintain performance when there are errors in the actuators and sensors compared to systems without PFTC so that the system response can return to the set point. The PFTC system designed to correct errors that occur in the system with each error value that can be estimated at 70% on the actuator and 73% on the sensor.