Abstract One of the most attractive features of multiphase machines is the fault-tolerant capability due to the higher number of phases. Different post-fault control strategies based on hysteresis, PI-resonant and predictive techniques have been recently proposed. They all proved their capabilities to withstand fault situations and to preserve the fundamental component of the air-gap field, while achieving minimum losses, maximum torque per-ampere and reducing torque vibrations. Nonetheless, due to their recent introduction, no thorough study has yet appeared comparing the performance of these controllers. In this paper two open-phase fault-tolerant control schemes are experimentally compared in a real five-phase induction machine. The controllers being compared are based on PI-resonant and predictive control techniques, respectively. The experiments include pre-and postfault situations. Obtained results show that both control methods offer nearly the same performance. When compared, predictive control provides faster control response and superior performance at low speed operation but is found to be less resilient to fault detection delays and to have higher current ripple. Regarding the controller implementation, it is shown that the transition from preto post-fault operation involves modelling the non-linear effects observed when an open-phase fault occurs for the predictive controller, and the proper retuning of the PI trackers for the PIresonant controller, in order to ensure post-fault operation.

No abstract

No abstract

No abstract

No abstract

No abstract

This paper addresses the problem of the greenhouse crop growth control using a hierarchical control approach. The proposed control scheme consists of two layers. In the lower one, adaptive and predictive controllers are used. The system dynamics at this level are described with hybrid models that arise due to the different modes of operating/controlling the greenhouse climate (heating and ventilation). As a result, different choices for the state of the system can be considered where inner temperature, solar radiation, and optimized set-points will act as logical conditions in the description of the hybrid system. The hybrid dynamics are described both in a general way and using a MLD representation, this result being useful for control purposes in the greenhouse climate control community. The upper layer calculates optimal climate set-points based on economic criteria. An optimization algorithm is used with a receding horizon strategy to provide the desired climate to the lower layer during the whole campaign. Representative experimental results of the implementation of the hierarchical control architecture are presented and discussed in the paper.

No abstract