This paper presents the study of permanent magnet synchronous machines (PMSM) running with eccentricity and bearings damage. The objective is to detect and identify the fault through the current signature analysis. The stator current has been analyzed by means of both Fourier (FFT) and Discrete Wavelet (DWT) transforms. Simulations have been carried out with a two-dimensional (2-D) finite element analysis (FEA), and they have been also compared with experiments. The results prove that the method proposed can be used to identify mechanical faults in a PMSM.

No abstract

This paper proposes a predictive real time energy management strategy for plug-in-hybrid electric vehicles (PHEV) based on an adaptation of Dynamic Programming (DP). The computational load of predictive real time strategies increases with the trip length. Therefore, for online computation by the onboard computer, they strongly depend on an efficient implementation. To reduce computation cost, current approaches for predictive strategies rely on strongly simplified intern vehicle models. The here proposed energy management strategy (EMS) uses a different approach, which is based on the use of precalculated lookup tables for the different operating points of the powertrain. This precalculation make the use of more exact vehicle models possible by using more detailed loss models of the powertrain components. The proposed EMS separates the optimization process, i.e. the calculation of the power distribution to engine and electric motor and gear in two calculation steps. The first step, which is computationally more intensive, has only to be executed once for a certain vehicle configuration. The obtained results are saved in lookup tables to avoid a later recomputation. In the second step, which is done online in the vehicle, a shortest path search algorithm is employed which is based on the predicted vehicle speed and rode slope of the trip. Techniques are integrated which decrease the rounding error caused by the use of lookup tables. The resulting difference of the consumed fuel mass between the lookup table based DP and standard DP is smaller than 0.03% by an approximately 50 times faster calculation. Using the proposed algorithm, even complex intern vehicle models do not affect the online computation cost and can be implemented by real time strategies.

No abstract

In this paper an evolutive algorithm is used to train an adaptative-network-based fuzzy inference system (ANFIS), particularly a genetic algorithm (GA). The GA is able to train the antecedent and consequent parameters of an ANFIS, which is used for energy load profile forecasting in an automated factory. This load forecasting is useful to support an intelligent energy management system (IEMS), which enables the user to optimize the energy consumptions by means of getting the optimal work points, scheduling the production according to these points, etc. The proposed training algorithm showed excellent results with complex plants like industrial energy consumers in the user-side, where the randomness of the loads is higher than in utility loads. Real data from an automated car factory were used to test the presented algorithms. Appropriated results were obtained.Peer ReviewedPostprint (published version