RESUMEN
The cold start of fuel cells limits their wide application. Since the water produced by fuel cells takes up more space when it freezes, it may affect the internal structure of the stack, causing collapse and densification of the pores inside the catalytic layer. This paper mainly analyzes the influence of different startup strategies on the stack cold start, focusing on the change in the stack temperature and the ice volume fraction of the catalytic layer. When designing a startup strategy, it is important to focus not only on the optimization of the startup time, but also on the principle of minimizing the damage to the stack. A lumped parameter cold-start model was constructed, which was experimentally verified to have a maximum error of 8.9%. On this basis, a model predictive control (MPC) algorithm was used to control the starting current. The MPC cold-start strategy reached the freezing point at 17 s when the startup temperature was -10 °C, which is faster than other startup strategies. Additionally, the time to ice production was controlled to about 20 s. Compared with the potentiostatic strategy and maximum power strategy, MPC is optimal and still has great potential for further optimization.
RESUMEN
In remote measurement systems, the lead wire resistance of the resistance sensor will produce a large measurement error. In order to ensure the accuracy of remote measurement, a novel lead-wire-resistance compensation technique is proposed, which is suitable for a two-wire resistance temperature detector. By connecting a zener diode in parallel with the resistance temperature detector (RTD) and an interface circuit specially designed for it, the lead-wire-resistance value can be accurately measured by virtue of the constant voltage characteristic of the zener diode when reverse breakdown occurs, and compensation can thereby be made when calculating the resistance of RTD. Through simulation verification and practical circuit testing, when the sensor resistance is in 848-2120 Ω scope and the lead wire resistance is less than 50 Ω, the proposed technology can ensure the measuring error of the sensor resistance within ±1 Ω and the temperature measurement error within ±0.3 °C for RTDs performing 1000 Ω at 0 °C. Therefore, this method is able to accurately compensate the measurement error caused by the lead wire resistance in two-wire RTDsand is suitable for most applications.
