Key points are not available for this paper at this time.
The purpose of this paper is to describe the model and the design of a neural network (NN) that optimizes a DC-DC converter power management system. The DC-DC converter plays a crucial role in maintaining the output voltage within the desired range. However, the conventional power management system has limitations in handling nonlinearities and uncertainties in the system. The proposed approach exploits the capability of a NN to accurately estimate certain parameters of the circuit that are closely associated with the functionality of the feedback control loop. In the case under analysis, we use Proportional-Integral-Derivative (PID) control to regulate the output voltage. Precise estimation of the real value of analog components involved in the computation of PID parameters can improve the control loop and overall performance. Moreover, estimation of parasitic resistances of capacitors and inductors provides information about the functional state of the DC-DC converter and forecasting future fails.
Radhakrishnan et al. (Sun,) studied this question.