IEEE Transactions on Power Electronics. Sep2019, Vol. 34 Issue 9, p8715-8723. 9p.
RAPID prototyping, CURRENT-voltage characteristics, and FEEDBACK (Psychology)
Using a photovoltaic (PV) emulator (PVE) simplifies the testing of the PV generation system. However, conventional controllers used for PVEs suffer from oscillating output voltage, requiring a high number of iterations, or being too complex to be implemented. This paper proposes a controller based on a resistance feedback control strategy that produces a stable and fast converging operating point for the PVE. The resistance feedback control strategy requires a new type of PV model, which is the current–resistance (I–R) PV model. This model is computed using a binary search method at a fast convergence rate. It is combined with a closed-loop buck converter using a proportional-integral controller to form the resistance feedback control strategy. The PVE's controller is implemented into dSPACE ds1104 hardware platform for experimental validation. The acquired experimental results show that the proposed PVE is able to follow the current–voltage characteristic of the PV module accurately. In addition, the PVE's efficiency is more than 90% under maximum power point operation. The transient response of the proposed PVE is similar to the PV panel during irradiance changes. [ABSTRACT FROM AUTHOR]
TARGET costing, PROTOTYPES, PRODUCT design, RAPID prototyping, and SUPPLIERS
Prototyping allows firms to evaluate the technical feasibility of alternative product designs and to better estimate their costs. We study a collaborative prototyping scenario in which a manufacturer involves a supplier in the prototyping process by letting the supplier make detailed design choices for critical components and provide prototypes for testing. While the supplier can obtain private information about the costs, the manufacturer uses target costing to gain control over the design choice. We show that involving the supplier in the prototyping process has an important influence on the manufacturer's optimal decisions. The collaboration results in information asymmetry, which makes parallel prototyping less attractive and potentially reverses the optimal testing sequence under sequential prototyping: It may be optimal to test designs in increasing order of attractiveness to avoid that the supplier does not release technically and economically feasible prototypes for strategic reasons. We also find that the classical target costing approaches (cost‐ and market‐based) need to be adjusted in the presence of alternative designs: Due to the strategic behavior of suppliers, it is not always optimal to provide identical target costs for designs with similar cost and performance estimates, nor to provide different target costs for dissimilar designs. Furthermore, the timing is important: While committing upfront to carefully chosen target costs reduces the supplier's strategic behavior, in some circumstances, the manufacturer can take advantage of this behavior by remaining flexible and specifying the second prototype's target costs later. [ABSTRACT FROM AUTHOR]