A new neural network system for object recognition is proposed which is invariant to translation, scaling and rotation. The system consists of two parts. The first is a preprocessor which obtains projection from the input image plane such that the projection features are translation and scale invariant, and then adopts the Rapid Transform which makes the transformed outputs rotation invariant. The second part is a neural net classifier which receives the outputs of preprocessing part as the input signals. The most attractive feature of this system is that, by using only a simple shift invariant transformation (Rapid transformation) in conjunction with the projection of the input image plane, invariancy is achieved and the system is of reasonably small size. Experiments with six geometrical objects with different degrees of scaling and rotation shows that the proposed system performs excellent when the neural net classifier is trained by the Cascade-correlation learning algorithm proposed by Fahlman and Lebiere.

Voltage Regulator Module, called VRM is a dedicated module for supplying power to microprocessor units. Recently, significant improvement of microprocessor units arises new challenges for supplying stable power. For stable and efficient control, multiphase interleaved topology is often used in today's VRM. To achieve high performance VRM, a current sensing circuit with both high efficiency and high accuracy is demanded. To achieve high accuracy, thermal dependency is a problem to be solved. In this paper, a novel alternating voltage controlled current sensing method is proposed for suppressing thermal dependency. In the proposed method, a high frequency AC voltage is superposed on the gate-ON-voltage. Then, the AC channel current is generated, and its amplitude becomes proportional to inductor current. The AC channel current is detected through a LC filter. The proposed current sensing method is very effective for realizing a current mode control DC-DC converter. In first, we simulated the relationship between our proposed current sensing method and a electrical characteristic of a power MOSFET. We used a power MOSFET device model published by a manufacture in this simulation. From the results, we find the gate parasitic capacitance of power MOSFET effects on the sensitivity of the current sensing circuit. Besides, the power dissipation in a power MOSFET increases by the frequency of applied gate ac voltage. Moreover, the proposed current sensing circuit based on the proposed method was designed and simulated the operations by Hspice. From the results, the designed current sensing circuit based on the proposed method has enough wide sensing window from 3A to 30A for VRM applications. Moreover, comparing to the conventional current sensing circuits with the MOSFET ON-resistance, the error of the proposed current sensing circuit can be decreased over 25% near 100°C.