No abstract

Abstract. This paper proposes the application of neural network as a controller to isolate engine vibration in an active engine mounting system. It has been shown that the NARMA-L2 neurocontroller has the ability to reject disturbances from a plant. The disturbance is assumed to be both impulse and sinusoidal disturbances that are induced by the engine. The performance of the neural network controller is compared with conventional PD and PID controllers tuned using Ziegler-Nichols. From the result simulated the neural network controller has shown better ability to isolate the engine vibration than the conventional controllers.

No abstract

No abstract

No abstract

This paper discusses terminal sliding mode control for active engine mounting (AEM) system to isolate vehicle engine body vibration. The engine vibration may occur due to road irregularities, at low frequency, and also reciprocating mechanism of the piston in the engine, at high frequency about 20 -40 Hz. Active engine mounting system is designed to deal with isolation of vibration in high level frequency. A number of controllers for AEM system have been proposed. One of them is sliding mode control. However, there is no systematic method to determine sliding surface for conventional sliding mode technique to assure the system will be asymptotically stable. The main advantage of the proposed controller compared to the conventional (linear) sliding mode control is that the sliding motion to the origin (stable condition) can be reached in finite time by introducing the nonlinear sliding surface based on the concept of terminal attractors. The result shows that terminal sliding mode control not only able to attenuate the vibration but also robust to the different type of disturbance and parametric uncertainties such as mass, stiffness, and damping coefficient.

Piezoelectric tube scanner is a major component that used in nanoscale imaging tools such as atomic force microscopy (AFM). This is because it can provide precise nanoscale positioning. However the precision is limited by vibration and some nonlinear drawbacks represented by creep and hysteresis. Hysteresis problem appears when positioning is needed at wide range. In this paper, a feed forward multi-layer neural network (MLNN) is trained to shape a proper control signal based on reference input and actual output signals. The experimental results show that the developed neural network scheme improves the performance of the system by significantly minimizing the effect of hysteresis.

No abstract