Iet Electric Power Applications最新文献

Electrical machine optimisation is normally a high-dimensional non-linear multi-objective optimisation problem. A multi-level optimisation (MO) strategy is currently used to improve efficiency, where sensitivity analysis is required for dividing design parameters into different groups. However, the conventional MO strategy cannot handle ultra-high-dimensional optimisation problems. In this paper, a sensitivity analysis method with variable weighted intervals is proposed to calculate the sensitivity coefficient in the parameter design range. Moreover, three improved multi-level optimisation strategies based on different optimisation algorithms, sequential sensitivity strategies, and machine learning models are proposed, analysed, and compared with the conventional MO strategy. Through a case study of a synchronous reluctance machine, it can be seen that the proposed optimisation strategies can improve the optimisation results and efficiency of ultra-high-dimensional optimisation of electrical machines.

The Permanent Magnet Linear Motor (TPMLM) is widely used in different industrial fields. TPMLMs with slots and iron cores have high power density, but their thrust fluctuations and copper losses are significant. Due to the nonlinearity and saturation of magnetic circuits, their electromagnetic models are complex and the accuracy of numerical methods is very inferior. Substantially accurate modelling is crucial for motor optimisation design. In this paper, a data-driven modelling method based on Bayesian optimisation deep neural network (DNN) is proposed to improve the accuracy of the electromagnetic field. The finite element (FE) modelling under different structural parameters is analysed and provides a training dataset for DNN. Then, a multi-objective optimisation problem for the slotted TPMLM is carried out based on the multi-objective black hole algorithm. Compared to the original design, the average thrust of TPMLM increased by 49.37%, the thrust fluctuation percentage decreased by 9.59%, and the coil copper consumption percentage decreased by 2.64%. The results show that the improved DNN model has very high modelling accuracy, providing a new way for motor design and optimisation.

The fault signal characteristics of motor rolling bearings for more electric aircraft are easily masked by strong background noise. Directly using machine learning, deep learning, or other methods results in a lower accuracy in fault recognition. In this article, a Northern Goshawk algorithm using a fusion subtraction optimiser and Cauchy strategy (SCNGO) is proposed to optimise the number of white noise additions and amplitude weights in the improved full set empirical mode decomposition method based on adaptive noise (ICEEMDAN). Then, a multi-scale convolutional neural network (MCNN) is used to extract the time–frequency domain features of the de-noised signal and perform information fusion. Finally, the bidirectional long short-term memory network (BiLSTM) was used to learn the faults' fusion features and complete the faults' recognition at different speeds. The research results show that the SCNGO-ICEEMDAN and MCNN-BiLSTMT model shows significant advantages in bearing fault recognition with an average recognition accuracy of 98.67% at various speeds.

The cogging torque is usually considered as an adverse effect of the PM machine servo system. This paper presents a new perspective that the cogging torque can be exploited in many practical situations as long as the speed ripple can be suppressed. By the proposed cogging torque design scheme, the local core saturation on the stator yoke can be constructed by specially designing the yoke structure. With this idea, the cogging torque distribution can be designed, with the transmission ratio, generating a series of designated auxiliary positioning forces on the load side. This weak positioning force will be amplified by the reduction gears and make the terminal servo load prone to stop at the designated stable equilibrium positions. This feature can help humans conveniently move the load to some designated positions by manual mode with a certain precision. Considering the negative impact of this constructed cogging effect, a speed ripple suppression method is applied to guarantee that the cogging torque machine can provide a good servo performance. The designated cogging torque and performances are verified by prototype machines, testing, and motion control experiments.

To the challenges posed by the complex structure and low reliability of existing hydraulic actuators used for more electric aircraft landing gear, this paper proposes an optimised design method for an electric actuator based on the principles of tubular permanent magnet linear motor employing a Halbach array. For this electric actuator, a subdomain model considering the longitudinal end effect is established by setting virtual permanent magnet regions at both ends of the Halbach array. The effectiveness of the proposed method is verified through a comparison with the finite element method. Subsequently, a surrogate model is developed using the proposed subdomain model, and an uncertain optimisation model considering machining and assembly errors is developed based on the interval possibility and the interval order theory. Compared to its initial design, the optimised structure increased the average thrust by 13.4% and reduced the thrust ripple rate by 87.2%, significantly improving the overall performance of the electric actuator. Finally, prototype experiments are carried out to verify the effectiveness of the proposed subdomain method and optimised approach.

Active magnetic bearing (AMB) is a key technology in high-speed rotating machines for rotor suspension, where the displacement sensors play a crucial role in detecting and controlling the rotor position. However, the traditional displacement sensors have the problems of high cost, large volume and poor reliability. To solve these problems, this paper proposes an innovative solution that utilises a recurrent neural network (RNN) to estimate the rotor displacement from the current in the AMB controller. The proposed method offers high-quality prediction performance for the rotor displacement which is close to the high precision eddy current displacement sensors. The mathematical model of AMB is analysed to provide guidance in historical current data acquisition and design of RNN. The input dimensions and the architecture of the neural network are optimised to improve both prediction accuracy and computational complexity. Experimental results validate the effectiveness of the algorithm and demonstrate that the proposed method has high accuracy, robustness and generalisation ability.

The vibration signals of a circuit breaker (CB) contain important action timing information. The optimisation of features extraction for vibration signals generated during the operation process of CBs is crucial for rapid defect location and identification for CB. The authors propose a new feature optimisation method, defined as energy trajectory entropy of vibration signals via tracing the action characteristics of the main shaft of CBs. Firstly, an image tracking algorithm is employed to dynamically capture the key frames of the high-speed image sequence of the main shaft and accurately divide the action sequence. The “cluster” instantaneous energy waveforms of the vibration signal, divided by zones, are then diffused by the twiddle factor in the polar coordinate sub grid area, while the energy trajectory entropy algorithm (ETE) is utilised to investigate the subtle changes in the energy release process. The ETE scale parameters are optimised using a support vector machine identification model. This enables rapid location of defective components in CBs, resulting in a significant reduction in time cost. The experiment has confirmed the strong feature correlation between CBs action and vibration signals, offering new ideas for achieving non-invasive defect identification of CBs.

Electric machines form an essential part of a wide range of modern systems. When speed control is required, the use of pulse width modulation-based inverters is generally the solution of choice. It is also usual to connect the machine to the inverter using a cable. The combination of these three elements produces the potential for voltages which exceed the dc link voltage to occur at the machine terminals. Methods for predicting the terminal voltage exist; however, these methods assume that the pulses applied to the system can be considered as isolated, discrete events. The authors highlight an issue with this assumption. When a switching event occurs, it will cause a voltage disturbance in the unswitched phases of the system due to the mutual coupling between the phases. If a second switching event occurs within a short time of this event the resultant voltage will interact with the previous switching event resulting in a higher terminal voltage than would be the case for an isolated event. This effect can be problematic for insulation design if it is not considered. This issue is demonstrated, with the worst-case scenarios identified and potential methods of reducing terminal voltage being proposed.

Parallel inverters have the advantages of low-output harmonics and high-parallel power, making them very suitable as the topology structure for an electric motor emulator. However, parallel inverters can also bring about circulating current issues, especially when using carrier phase shifted sinusoidal PWM (CPS-SPWM) control mode, which significantly increases circulating current, leading to increased output harmonics and losses in the system. In order to suppress circulating currents, this paper provides a detailed analysis from both high-frequency and low-frequency perspectives in CPS-SPWM control mode. A circulating current suppression method that combines hardware and software systems, adopting a cascaded coupled output network structure along with a corresponding error voltage space vector control method to achieve circulating current suppression, is proposed. By constructing a 6-branch parallel output experimental platform, the feasibility and effectiveness of the proposed method have been verified, suppressing current circulating currents and improving current quality.

Harmonic state estimation, which highly depends on accurate synchronised phasors, is a crucial basis for the location, traceability, and governance of harmonics in power systems. However, most existing synchronous phasor measurement units (PMUs) provide power frequency phasors rather than harmonic phasors. In addition, as the dominant voltage signal provider for PMUs in power systems, capacitive voltage transformers (CVTs) have complex frequency responses and cannot be directly used for harmonic measurement. Therefore, accurate harmonic state estimation is difficult to achieve in applications. A method for harmonic voltage phasor reconstruction and harmonic state estimation based on the measurement data of a CVT is proposed. First, the reasons for harmonic voltage measurement errors were analysed. Second, the scattering parameter method was used to measure the wideband voltage transfer characteristics of CVT. Third, a voltage reconstruction method was proposed to improve the measurement accuracy and expand the frequency range of synchronous phasor measurement. Fourth, this method was applied to harmonic state estimation, and the performance of harmonic state estimation before and after voltage reconstruction was quantitatively compared. Results showed that the proposed method can remarkably improve the accuracy of harmonic state estimation. Finally, some factors, such as state observability, complex error disturbances, CVT structure, and harmonic order, influencing the performance of the proposed method were investigated. Results further demonstrated that the method has high accuracy, strong anti-interference, and adaptability, and can provide basic information for harmonic localisation and traceability.