Fault Diagnosis Scheme for Railway Switch Machine Using Multi-Sensor Fusion Tensor Machine

Railway switch machine is essential for maintaining the safety and punctuality of train operations.A data-driven fault diagnosis scheme for railway switch machine using tensor machine and multi-representation monitoring data is developed herein.Unlike existing methods,this approach takes into account the spatial information of the time series monitoring data,aligning with the domain expertise of on-site manual monitoring.Besides,a multi-sensor fusion tensor machine is designed to improve single signal data’s limitations in insufficient information.First,one-dimensional signal data is preprocessed and transformed into two-dimensional images.Afterward,the fusion feature tensor is created by utilizing the images of the three-phase current and employing the CANDE-COMP/PARAFAC(CP)decomposition method.Then,the tensor learning-based model is built using the extracted fusion feature tensor.The developed fault diagnosis scheme is valid with the field three-phase current dataset.The experiment indicates an enhanced performance of the developed fault diagnosis scheme over the current approach,particularly in terms of recall,precision,and F1-score.

Machine-Learning Based Packet Switching Method for Providing Stable High-Quality Video Streaming in Multi-Stream Transmission

Broadcasting gateway equipment generally uses a method of simply switching to a spare input stream when a failure occurs in a main input stream.However,when the transmission environment is unstable,problems such as reduction in the lifespan of equipment due to frequent switching and interruption,delay,and stoppage of services may occur.Therefore,applying a machine learning(ML)method,which is possible to automatically judge and classify network-related service anomaly,and switch multi-input signals without dropping or changing signals by predicting or quickly determining the time of error occurrence for smooth stream switching when there are problems such as transmission errors,is required.In this paper,we propose an intelligent packet switching method based on the ML method of classification,which is one of the supervised learning methods,that presents the risk level of abnormal multi-stream occurring in broadcasting gateway equipment based on data.Furthermore,we subdivide the risk levels obtained from classification techniques into probabilities and then derive vectorized representative values for each attribute value of the collected input data and continuously update them.The obtained reference vector value is used for switching judgment through the cosine similarity value between input data obtained when a dangerous situation occurs.In the broadcasting gateway equipment to which the proposed method is applied,it is possible to perform more stable and smarter switching than before by solving problems of reliability and broadcasting accidents of the equipment and can maintain stable video streaming as well.

Evaluation of Switched-Reluctance Machine through Generalized Sizing Equations

With the evolution of converter fed machines (CFMs), it becomes important to evaluate the power potential of such machines of vastly different topologies with a variety of waveforms of back emf and current. It is based on the generalized sizing equations and permits the evaluation of the main dimensions with respect to the power of those machines. In this paper, a general approach is presented to extend the evaluation method of machine power density to the switched reluctance (SR) machine, and furthermore to compare the power production capability between the SR machine and the well known squirrel cage induction machine.

Four-quadrant Force Control with Minimal Ripple for Linear Switched Reluctance Machines

Linear switch reluctance machine(LSRM)has been tried to act as an alternative generator for direct drive linear wave energy converter(WEC).Many researchers have proposed new topologies of LSRM to improve the power density,efficiency and reliability.However,the control methods for LSRM applied in direct drive WEC have been paid little attention,especially control methods considering the wave energy generator operating characteristics.In this paper,according to the generator control requirements of the direct drive WEC,force control algorithm for LSRM operating in four quadrants without a speed closed loop is put forward.The force ripple of LSRM is suppressed using force sharing function method.The four-quadrant control is easy to realize requiring only phase currents information.Simulation results validate the proposed method and indicate that LSRM is able to be used as the generator for direct drive WEC.

Study on Magnetic Shielding for Performance Improvement of Axial-Field Dual-Rotor Segmented Switched Reluctance Machine

A category of permanent-magnet-shield(PM-shield)axial-field dual-rotor segmented switched reluctance machines(ADS-SRMs)are presented in this paper.These topologies are featured by using the magnetic material to shield the flux leakage in the stator and rotor parts.Besides,the deployed magnets weaken the magnetic saturation in the iron core,thus increasing the main flux.Hence,the torque-production capability can be increased effectively.All the PM-shield topologies are proposed and designed based on the magnetic equivalent circuit(MEC)model of ADS-SRM,which is the original design deploying no magnet.The features of all the PM-shield topologies are compared with the original design in terms of the magnetic field distributions,flux linkages,phase inductances,torque components,and followed by their motion-coupled analyses on the torque-production capabilities,copper losses,and efficiencies.Considering the cost reduction and the stable ferrite-magnet supply,an alternative proposal using the ferrite magnets is applied to the magnetic shielding.The magnet demagnetization analysis incorporated with the thermal behavior is performed for further verification of the motor performance.

Mach-Zehnder Interferometers with Asymmetric Modulation Arms in Applications of High Speed Silicon-on-Insulator Based Optical Switches

Modulation arms with different widths are introduced to Mach-Zehnder interferometers （MZIs） to obtain improved performance. Theoretical analysis and numerical simulation have shown that when tile widths of tile two arms are properly designed to achieve an inherent mπ/2 （m is an odd integer） optical phase difference between the arms, the asymmetric MZI presents higher modulation speed. Furthermore, the carrier-absorption induced divergence of insertion losses in silicon-on-insulator （SOI） based MZI optical switches can be obviously improved.

Analysis and Comparison of Power Electronic Converters for Conventional and Toroidal Switched Reluctance Machines

Different power electronic converter topologies are introduced in this paper for both Conventional Switched Reluctance Machine (CSRM) and Toroidal Switched Reluctance Machine (TSRM) drive systems. Their commutation, switch and diode currents, power losses, and efficiencies under over modulation operation are analyzed and compared for converter characteristics study, performance evaluation and topology selection for CSRM and TSRM drive systems. The switch and diode silicon volumes required for each CSRM and TSRM drives are also compared according to their corresponding currents at the equivalent machine torque versus speed operating points.

Temporal and spatial evolution of air-spark switch plasma investigated by the Mach–Zehnder interferometer

An air-spark switch plasma was diagnosed by the Mach–Zehnder laser interferometer with ultra-high spatial and temporal resolution. The interferograms containing plasma phase shift information at different time were obtained. The phase shift distributions of the plasma were extracted by numerically processing the interferograms. The three-dimensional(3 D) electron density distributions of the air-spark switch plasma were then obtained. The working process of the air-spark switch was described by analyzing the temporal and spatial evolution of the plasma electron density.

Generic meminductive characteristics of switched reluctance machines

The meminductive system can be regarded as the generalization of the meminductor. This paper focuses on exploring the generic meminductive characteristics of the switched reluctance machine （SRM）. The dynamical equations of SRM sys- tems are derived and discussed in comparison with the typical constitutive relation equations of the meminductive system. Memory ability and pinched hysteresis loop （PHL） are taken as the indicative fingerprints to draw forth the theoretically comparative analysis. Based on the theoretical analysis, in addition to simulation and experimental confirmation, it can be concluded that from the viewpoint of circuit, SRM can be considered as a generic meminductive system.

Development of a New Flux Switching Transverse Flux Machine with the Ability of Linear Motion

This paper proposes a new rotary flux switching transverse flux machine with the ability of linear motion(FSTFMaLM),in which both the stator and the rotor cores are made by using soft magnetic composite(SMC)materials.With the special design pattern,for the rotary motion model,the proposed machine can combine both the advantages of the flux switching permanent magnet machine(FSPMM)and the transverse flux machine(TFM).It can output with relatively high torque density,and as there is no windings or the magnets on the rotor cores,the proposed machine can operate in the high speed region to improve the output power.With the adoption of the SMC materials,the manufacturing of this machine can be quite easy.By stacking the rotor core together and prolong it with the determined length in the axial direction,in addition with the special control algorithm,the proposed machine can have the ability of the linear motion.In this paper,the operation principle of this machine has been explained and the design methods are also presented.To seek the better performance,the main dimension of the machine is optimized,and for the performance evaluation,the finite element method(FEM)is adopted.The proposed machine can be used for the electric driving systems,robotic systems or other applications where the linear motion ability is required.

Electromagnetic Performance Analysis of Flux-Switching Permanent Magnet Tubular Machine with Hybrid Cores

The performance of traditional flux switching permanent magnet tubular machine(FSPMTM)are improved by using new material and structure in this paper.The existing silicon steel sheet making for all mover cores or part of stator cores are replaced by soft magnetic composite(SMC)cores,and the lamination direction of the silicon steel sheet in stator cores have be changed.The eddy current loss of the machine with hybrid cores will be reduced greatly as the magnetic flux will not pass through the silicon steel sheet vertically.In order to reduce the influence of end effect,the unequal stator width design method is proposed.With the new design,the symmetry of the permanent magnet flux linkage has been improved greatly and the cogging force caused by the end effect has been reduced.Both 2-D and 3-D finite element methods(FEM)are applied for the quantitative analysis.

Machine Learning and Micromagnetic Studies of Magnetization Switching

Magnetization switching is one of the most fundamental topics in the field of magnetism.Machine learning(ML)models of random forest(RF),support vector machine(SVM),deep neural network(DNN)methods are built and trained to classify the magnetization reversal and non-reversal cases of single-domain particle,and the classification performances are evaluated by comparison with micromagnetic simulations.The results show that the ML models have achieved great accuracy and the DNN model reaches the best area under curve(AUC)of 0.997,even with a small training dataset,and RF and SVM models have lower AUCs of 0.964 and 0.836,respectively.This work validates the potential of ML applications in studies of magnetization switching and provides the benchmark for further ML studies in magnetization switching.

Static Simulation of a 12/8 Switched Reluctance Machine (Application: Starter-Generator)

Because its high efficiency, its simple stator and rotor structures, the low cost and high reliability, speed operation combined with robust and low cost construction, the switched reluctance machines have represented. In recent years, an interesting alternative to other machine types has been chosen for traction applications especially starter-generator. Their rotors do not generate significant heat, resulting in easy cooling. Their unidirectional flux and current may generate lower core losses and require a simple converter design. Moreover, the switched reluctance machines are known for their high reliability and capability of operating in four quadrants for a variable speed drive. Despite those merits, switched reluctance machine has not been extensively used until recently because of its problems of torque ripples and noise. Additionally, researchers have faced many difficulties to build a SRM model because it is inherently multivariable. It has strong coupling and especially a high nonlinearity. In this paper, we deal with many modeling methods. Numerical, analytical and intelligent approaches are studied. The important aim in this research is to use static results from FEMM simulation as flux-linkage, co-energy, static torque to form a dynamic model of a switched reluctance machine used next as a starter-generator of a hybrid vehicle.

Torque Sharing Function Control of Switched Reluctance Machines with Reduced Current Sensors

This paper presents a Torque Sharing Function(TSF)control of Switched Reluctance Machines(SRMs)with different current sensor placements to reconstruct the phase currents.TSF requires precise phase current information to ensure accurate torque control.Two proposed methods with different chopping transistors or a new PWM implementation require four or two current sensors to replace the current sensors on each phase regardless of the phase number.For both approaches,the actual phase current can be easily extracted during the single phase conducting region.However,how to separate the incoming and outgoing phase current values during the commutation region is the difficult issue to deal with.In order to derive these two adjacent currents,the explanations and comparisons of two proposed methods are described.Their effectiveness is verified by experimental results on a four-phase 8/6 SRM.Finally,the approach with a new PWM implementation is selected,which requires only two current sensors for reducing the number of sensors.The control system can be more compact and cheaper.

Optimal Designs of Wound Field Switched Flux Machines with Different DC Windings Configurations

Wound field switched flux(WFSF)machines exhibits characteristics of the simple robust rotor,flexible flux-adjustable capability,and no risk of demagnetization.However,they suffer from a poor torque density compared with permanent magnet machines due to the saturation.Therefore,in this paper,two WFSF machines with single-and double-layer DC windings,respectively,are optimized for the maximum torque.The end-winding(EW)lengths differ in these two machines,which can affect the optimal design.Design parameters including the DC to armature winding copper loss ratio,slot area ratio and split ratio are optimized when two machines have the same copper loss and overall sizes.In addition,the influence of the flux density ratio,total copper loss,air-gap length and aspect ratio on the optimal split ratio is investigated using the finite element method and results are explained through the analytical model accounting for the saturation.It is discovered that the EWs have no effect on the optimal copper loss ratio,which is unity.In terms of the slot area ratio,the machine with single-layer DC windings prefers smaller DC slot areas than armature slot areas.In the WFSF machine with longer EWs,the optimal split ratio becomes smaller.Moreover,compared with other parameters,the flux density ratio can significantly affect the optimal split ratio.

Design and Analysis of a Novel Rotor-Segmented Axial-Field Switched Reluctance Machine

As excitation strategy is influenced by the parity of phase number,and serious end effect is caused by full pitch windings adopted by the conventional rotor-segmented switched reluctance machine(RS-SRM),a novel rotor-segmented axial-field switched reluctance machine(RS-AFSRM),which adopts concentrated windings,is proposed in this paper,whose control strategy is independent of the parity of phase number.The stator of the proposed RS-AFSRM consists of inner and outer flux-conductive ring,excitation poles,and stator yoke,while the rotor consists of fan-shape cylinder-type segments.The structure and principle of the proposed machine is introduced through the magnetic circuit model of the machine.The fringing effect on air-gap field is considered.The simplified magnetic circuit considering the fringing effect is presented and the fringing effect coefficient is introduced.The design procedure considering the fringing effect through iterative modification is presented in this paper and is validated through 3D finite-element method(3D-FEM).Results show that both magnetic field distribution and operation are reasonable and satisfy the requirements.

Reduction of Cogging Torque in Permanent Magnet Flux-Switching Machines

Permanent magnet flux-switching machine (PMFSM) is a relatively new structure. Available literatures mainly focused on its general design procedure and performance analysis. In this paper, Finite Element Method (FEM) is taken to ana-lyze various design techniques to reduce the cogging torque in a prototype 12/10-pole PMFSM.

Performance Evaluation and Comparison of Multi - Objective Optimization Algorithms for the Analytical Design of Switched Reluctance Machines

This paper systematically evaluates and compares three well-engineered and popular multi-objective optimization algorithms for the design of switched reluctance machines.The multi-physics and multi-objective nature of electric machine design problems are discussed,followed by benchmark studies comparing generic algorithms(GA),differential evolution(DE)algorithms and particle swarm optimizations(PSO)on a 6/4 switched reluctance machine design with seven independent variables and a strong nonlinear multi-objective Pareto front.To better quantify the quality of the Pareto fronts,five primary quality indicators are employed to serve as the algorithm testing metrics.The results show that the three algorithms have similar performances when the optimization employs only a small number of candidate designs or ultimately,a significant amount of candidate designs.However,DE tends to perform better in terms of convergence speed and the quality of Pareto front when a relatively modest amount of candidates are considered.

Efficacy of conbercept after switching from bevacizumab/ranibizumab in eyes of macular edema secondary to central retinal vein occlusion

AIM: To explore the efficacy of conbercept after switching from bevacizumab/ranibizumab in eyes of central retinal vein occlusion(CRVO) through optical coherence tomography angiography(OCTA). METHODS: Patients with prior treatment of a minimum of three consecutive intravitreal injections of either bevacizumab or ranibizumab, followed by injection of conbercept, were recruited. The minimal follow-up period after switching was 12 mo. Central retinal thickness(CRT), best-corrected visual acuity(BCVA), the interval of injections was reviewed. Perfusion density(PD) and vascular length density(VLD) of superficial and deep capillary plexus were acquired from OCTA images before and after switching. RESULTS: Twenty-four eyes were included. CRT significantly decreased from 460.71±153.23 μm(before switching) to 283.92±38.27 μm at the end of follow-up(P<0.001). However, BCVA gained to some extent(from 0.98±0.33 to 0.76±0.42 log MAR) but the difference was not significant(P=0.070). After switching to conbercept the injection interval extended from 5.2±2.3 wk to 8.3±3.9 wk(P=0.012). At the end of follow-up, PD of deep retinal layer decreased significantly compared with before switching(from 34.62%±5.27% to 33.26%±5.82%, P=0.016), similar result was found in VLD of deep retinal layer but not in PD or VLD in superficial layer.CONCLUSION: In cases of refractory macular edema secondary to CRVO, switching to conbercept improves macular thickness and extends interval of injection. Retinal microvasculature cannot improve with treatment of conbercept.

Influence of Inner/Outer Stator Pole Ratio and Relative Position on Electromagnetic Performance of Partitioned Stator Switched Flux Permanent Magnet Machines

Based on the 6-pole outer stator(armature winding-stator),the influence of inner(permanent magnet-stator)/outer stator pole ratio n(n=NIS/NOS),stator relative positions and rotor pole number combinations on electromagnetic performance of partitioned stator switched flux permanent magnet(PM)machines(PS-SFPMMs)is investigated in this paper.Since the armature windings and PMs are located in two separated stators and PMs are stationary,PS-SFPMMs have high fault tolerance capabilities.To maximize the torque performance,the PM of inner stator pole should be aligned with outer stator pole when n is odd while the iron rib of inner stator pole should be aligned with outer stator pole when n is even.No matter what n is selected,the rotor pole number NR can be any integers except the phase number and its multiples.The analysis results indicate that the optimal NR is closed to(NIS+NOS)/2 and it is odd when n is odd while it is even when n is even.Meanwhile,symmetrical phase back-EMF waveform will be obtained when the ratio of Min(NOS,NIS)to the greatest common divisor of Min(NOS,NIS)and NR is even.Based on the optimal rotor pole numbers for 6-pole outer stator with different n and corresponding optimal relative position together with same rated copper loss,the average torque is improved by 18.4%,25.1%and 25.7%respectively in PS-SFPMMs with n equal to 2,3 and 4 when compared with PS-SFPMM with n equal to 1.The analyses are validated by experiment results of the prototype machine.