No abstract

With the increasing use of sensors and wireless communication systems, predictive maintenance is acquiring more and more importance to assess the condition of in-service equipment. Predictive maintenance presents promising cost savings, as it allows minimizing unscheduled power systems faults, which can have very costly and catastrophic consequences. Early stage detection of power system failure requires acquiring, monitoring, and periodically analyzing the condition of the elements involved, such as high-voltage power connectors, since they are critical devices which are often located in key points of power systems. This paper proposes a low-cost online system to determine the contact resistance of high-voltage direct current (DC) and alternating current (AC) power connectors, to determine their health condition in order to apply a predictive maintenance plan. The contact resistance is considered as a reliable indicator of connector's health condition. However, it cannot be directly measured, and the applied strategy differs between DC and AC power systems. Experimental results show a maximum error of 5%, thus proving the accuracy and feasibility of the approach presented in this paper, since the proposed limit of acceptable resistance increase is 20%. This approach can also be applied to many other power systems' elements.

Artificial neural networks allow the identification of black-box models. This paper proposes a method aimed at replicating the static and dynamic behavior of a DC-DC power converter based on a recurrent nonlinear autoregressive exogenous neural network. The method proposed in this work applies an algorithm that trains a neural network based on the inputs and outputs (currents and voltages) of a Buck converter. The approach is validated by means of simulated data of a realistic nonsynchronous Buck converter model programmed in Simulink and by means of experimental results. The predictions made by the neural network are compared to the actual outputs of the system, to determine the accuracy of the method, thus validating the proposed approach. Both simulation and experimental results show the feasibility and accuracy of the proposed black-box approach.

No abstract