To achieve levitation in active magnetic bearing (AMB) systems, levitation coils must be driven by a switch power amplifier (SPA). Therefore, SPA performance is important to stable levitation. In previous studies, five-phase six-leg (5P6L) SPA topology was proposed to reduce hardware costs. However, current tracking performance degrades due to the coupling problem of the five-phase current and freewheeling loss. Then, the rotor may deviate markedly from the balance position, affecting rotor stability. The keys to solving these problems lie in the duty cycle control of the common leg and the compensation of freewheeling loss. This article proposes an improved one-cycle control (OCC) algorithm in which the duty cycle of the common leg is optimized by finite control set model predictive control (FCS-MPC), and the freewheeling loss is compensated for by a high-precision mathematical model. In addition, the parameter uncertainty of the FCS-MPC model is analyzed, and the convergence of the optimization algorithm is investigated. Simulations and experiments show that the proposed algorithm is effective, and the high dynamic response and low steady-state error performance of SPA are achieved.

A bearingless switched reluctance motor (BSRM) has the combined characteristics of a switched reluctance motor (SRM) and a magnetic bearing. The hybrid-rotor BSRM (HBSRM) discussed in the paper has a twelve-pole stator and an eight-pole hybrid rotor, which is composed of a cylindrical rotor and a salient-pole rotor. Although the asymmetry of the hybrid rotor makes the structure and magnetic field of the HBSRM more complex, it can always produce a significant amount of magnetic pulling force to levitate a rotor shaft at all the rotor angular positions of each phase, which is not available in a traditional BSRM. The classical mathematical model for a conventional BSRM is valid only when its rotor rotates from the start of the overlap position to the aligned position, and the radial force and torque derived from this model are discontinuous at the aligned positon, which is harmful to the motor’s stable operation. In this paper, a full-period mathematical model on the assumption that the gap permeance is cut apart by straight lines or improved elliptical lines for a 12/8-pole HBSRM is provided. On the basis of this mathematical model, the continuity of the radial force and torque at all the rotor angular positions can be guaranteed, and the fine characteristics of this mathematical model have been verified by simulations.

In this paper, a compact single current sensor measurement based four-phase SRM integrated drive system is developed. Based on the current measurement characteristics of this drive system, the phase current reconstruction and demagnetization current suppression strategies are proposed. To ensure reliable operation of the SRM, two key issues are considered: 1) The four-phase current should be obtained from the only measured bus current signal; 2) As the non-conducting phases are unable to demagnetize normally with traditional control algorithms, serious negative torque effects exist. For the first issue, a four-phase current reconstruction method is proposed by decoupling the adjacent two-phase currents in the overlapping conduction regions. For the second issue, high-frequency pulses are injected in the shared upper power switch of the bridge arm where the conducting phase located, providing rapid demagnetization for the other phase in the same bridge arm. The two methods are fully decoupled and can be integrated freely to achieve current reconstruction and demagnetization current suppression. Based on these methods, a hybrid control algorithm that can satisfy the operation requirement in different speed range is developed for the SRM drive. Experiments have been implemented in an 8/6 structure SRM prototype based system, and detailed experimental results verify the effectiveness of the proposed methods.

The coupling between torque and radial force, and between the two radial forces is detrimental to the efficient control and performance improvement of the bearingless switched reluctance motor (BSRM). In the article, an 18/16-pole permanent magnet (PM) biased BSRM with radial force windings driven by a three-phase full-bridge inverter is presented. To solve the coupling between two radial forces in this proposed BSRM, three of its derived decoupled topologies have been developed, and the decoupled characteristics of these structures are proved by the radial force mathematical model and finite element analysis. Further, based on the space vector pulse width modulation (SVPWM) technique for three-phase full-bridge inverters, a one-cycle digital control scheme for the radial force control applicable to four types of BSRMs has been developed. Finally, a BSRM is prototyped and tested for experimental validation of radial levitation.

This paper presents a novel magnetic bearing switched reluctance motor (SRM) (MBSRM) with two degrees of freedom suspension, consisting of an SRM and an active magnetic bearing (AMB). In the proposed MBSRM, three-phase armature windings of the SRM and a biased winding of the AMB are fed together by a modified asymmetric half-bridge converter. The rotational torque in SRM and the biased flux used for producing electromagnetic force in AMB are generated simultaneously when the MBSRM works on the traditional switched reluctance excitation mode. First, the structure and working principle of the MBSRM were introduced. Second, its theoretical formulas of suspending forces were given and validated with the finite element analysis. Third, the proposed half-bridge converter is shown, and currents in the biased winding and three-phase armature windings of MBSRM are obtained by the magnetic field-circuit coupling simulation. Moreover, a prototype is manufactured and the experiment results have proved the feasibility of the proposed converter.

Purpose: – The purpose of this paper is to analyze the network structure of technology in and between different fields, as well as the evolution of their relations. Design/methodology/approach: – Using the patent data in Derwent Innovation Index (DII) from 1991 to 2010, this paper analyzes the co-classification of Derwent Manual Code (DMC) of patents in all technology fields. Large-scaled co-classification matrices are employed to generate the DMC co-classification networks. In addition, analyses are pursued at different levels of aggregation in four five-year windows: 1991-1995, 1996-2000, 2001-2005 and 2006-2010. Using Girvan-Newman algorithm in the clustering process, the structure transformations over time are detected. Findings: – The paper identifies the key technological knowledge in certain fields and finds out how different technological fields are connected and integrated. What is more, the dynamic evolution between networks in different time periods reveals the trend of generic technology development in the macroscopic level. Originality/value: – The paper investigates a large quantity of data – all the patent data in DII from 1991 to 2010 in this paper. The paper applies Girvan-Newman algorithm in the co-classification analysis and uses co-classification networks to reveal technology network structures. Evolution coincident with the realistic technological shifts can be observed.