Photovoltaic (PV) systems are becoming more vulnerable to cyber threats. In response to this emerging concern, developing cyber-secure power electronics converters has received increased attention from the IEEE Power Electronics Society that recently launched a cyber-physical-security initiative. This letter proposes a deep sequence learning based diagnosis solution for data integrity attacks on PV systems in smart grids, including dc-dc and dc-ac converters. The multilayer long short-term memory networks are used to leverage time-series electric waveform data from current and voltage sensors in PV systems. The proposed method has been evaluated in a PV smart grid benchmark model with extensive quantitative analysis. For comparison, we have evaluated classic data-driven methods, including K-nearest neighbor, decision tree, support vector machine, artificial neural network, and convolutional neural network. Comparison results verify performances of the proposed method for detection and diagnosis of various data integrity attacks on PV systems.

Pantograph-catenary systems are the key equipment for current conduction from power grids to electric locomotives, and within such a system, the property of the slide, especially the contact resistance ( R_j ), directly affects its current collection performance. Compared to previous studies, this paper focuses on the microscopic roughness parameters which affects R_j greatly. The changing rule of R_j under different roughness conditions is studied on 128 groups of data collected from experiments. More importantly, the second-order partial correlation coefficient is introduced to quantify the impact strength of different roughness parameters on R_j . Based on this, a multi-parameter influence model of R_j is established and verified by non-linear fitting. The results show that the impact strength of the relevant roughness parameters is ranked as R_pc > R_p > R_sm . The conclusion provides a theoretical basis for minimizing pantograph-catenary R_j and improving current collection quality. Moreover, it facilitates slide material selection and contributes to new prototype design.

Marine electric propulsion is an important topic in the research of modern ships and underwater vehicles. The propulsion motor drives based on model predictive control (MPC) are becoming increasingly popular in marine propulsion systems as an emerging technology. However, the multi-objective optimization in conventional MPC requires cumbersome weighting factor tuning. The relatively large computational cost is also detrimental to the industrial application of MPC. Aiming at reducing the computational complexity of multi-objective optimization without weighting factors, this paper proposes an enhanced ranking-based MPC method for induction motor drives in marine electric power propulsion. The presented control set pre-optimization aims to reduce the computational complexity of enumeration and ranking. Based on the sign of torque prediction deviation, the proposed method avoids enumerating all fundamental voltage vectors. Consequently, the number of candidate elements in the initial control set are reduced to four without excessively excluding feasible solutions. By converting predicted numerical errors into ranking results, the proposed MPC seeks the optimal solution among the candidates through improved ranking evaluation. Considering the situation of simultaneous optimal ranking, the normalization error judgment is developed to further optimize the optimal solution selection process. The simulation and experimental results confirm that the proposed MPC is simple and effective. Without the involvement of tuning the weighting factors, the proposed method achieves good performance.

In this paper, a sensorless rotor position estimation method based on the field current harmonic for a load-commutated inverter (LCI)-fed electrically excited synchronous motor (EESM) operating at zero or low speed is proposed. A detailed theoretical analysis and mathematical deduction are presented. The field current harmonic is used instead of the extra high-frequency signal injection. A reasonable harmonic frequency is selected, and a realizable field current control method is given. To simplify the complex signal processing and improve the measurement accuracy, a specific bandpass filter and a normalized quadrature phase-locked loop are designed. The phase change caused by the filters is compensated properly. The proposed approach is demonstrated through the simulation and experiment with a 15-kW LCI-fed EESM drive platform, confirming its feasibility and accuracy.

In the pantograph-catenary system, the change of the surface roughness of pantograph sliding plate directly affects the current collection quality of electric locomotive. In this paper, a large number of current-carrying friction experiments have been carried out on the self-developed high-performance sliding electric contact experimental machine, and the effects of the surface roughness of the sliding friction pair on important performance parameters have been studied, including pantograph-catenary contact resistance ( $R_{j}$ ), current carrying efficiency ( $\eta$ ), current collection stability ( $\delta$ ) and the friction coefficient ( $\mu $ ). Further, the current conduction mechanism is revealed by the microscopic approach. In order to quantitatively reveal the effect of roughness on each evaluation index of current collecting performance, the paper formulates the functional relationship between $R_{a}$ (arithmetic mean height of contour) and $R_{j}$ , $\eta $ , $\delta $ , $\mu $ by nonlinear fitting. On this basis, a comprehensive evaluation equation is established by the entropy weight method, and then the optimal $R_{a}$ value is derived for the best pantograph-catenary current collecting performance. The experimental results show that with the increase of the surface roughness $R_{a}$ of the sliding plate, $R_{j}$ decreases first and then increases, while $\eta $ increases first and then decreases, $\delta $ increases monotonically, and $\mu $ gradually increases. The optimal pantograph-catenary current collection condition can be achieved by reconciliation of these indices. The research results of this paper provide a theoretical basis for improving the current collection quality of pantograph-catenary systems, and is useful for material selection or new type design of pantograph sliding plates.

A zero input current ripple DC-DC converter is proposed in the paper. In the proposed converter, a zero current ripple branch is adopted, which consists of extra winding coupled with the three-winding inductor, an capacitor and an discrete inductor. Based on the zero current ripple branch, the current variation rates of the branch and the input inductor are reverse. Further, the proposed converter can operate in soft-switching condition by adding resonant capacitor and optimal parameter design. Experimental results verify the correctness of the proposed topology and effectiveness of reducing input current ripple and soft-switching characteristics.