In this paper, the fault diagnosis problem for power converters is considered. Given that the existing fault diagnosis models rarely address the problems of the data noise and the new faults that are never emerged in the database, thus, an optimized fault diagnosis model for power converters based on temporal convolutional network (TCN) is proposed. Our contributions include: 1) unknown faults can be efficiently distinguished with an optimized classifier; 2) the proposed model has good robustness and reliability under noisy environment without any subsidiary pre-denoising algorithm; 3) it can realize adaptive feature extraction, and the parameters are small. Experimental results on a three-phase voltage inverter platform demonstrate that the proposed approach is efficient and can be adaptively applied to various real applications.

High quality online uninterruptible power supplies (UPS) are widely used in applications e.g. precision medical equipment with critical loads. This paper proposes a low-cost and high performance online UPS system. Firstly, the performance improvement of a bidirectional switch based bridgeless power factor (PFC) converter, which is regulated by a hybrid control strategy with load current feedforward is studied. Secondly, the proposed closed-loop controller integrates the improved predictive current control with predictive voltage control and thereby it applies to the second-stage neutral-point-clamped (NPC) three-level inverter to reduce harmonics of the output ac voltage. Moreover, a push-pull DC-DC converter to boost the voltage as the backup energy source interfaces a 24V battery bank. Considering the complexity of the UPS system as well as cost reduction, we implement a multi-task decomposition method for control on a monolithic low-cost digital signal processor (DSP) embedded with Control Law Accelerator (CLA). Finally, the experimental results from a 1kVA a prototype can effectively verify validly of the theoretical analysis and design.

The grid-connected inverter is a key device in the renewable energy power generation system and large-scale energy storage system, which the operational stability and reliability are the basis for the efficient and safe application of electrical energy. A real-time fault diagnosis method of a three-phase for grid-connected application combining a mixed logic dynamic (MLD) model and finitecontrol set model predictive control (FCS-MPC) is proposed. This paper not only realizes the open circuit fault diagnosis and location of the switching devices in the main power circuit, but also discusses the threshold issues and post-fault operations. The advantage of the proposed method is that it directly uses the control data and measurement signals of the controller without extra sensors and calculation, which will shorter the fault diagnosis time and occupy less calculation resource of the main processor. Simulation results illustrate the quickness of the fault identification and accurate position with robustness to the interference of the diagnosis method. Finally, the effectiveness of the diagnosis method was verified by a 1500 W experimental prototype in a laboratory.

Eddy current testing (ECT), as a non-contact and non-destructive testing (NDT) technology, has been widely applied to various industrial fields for pipeline defect detection. In this paper, a combined defect parameter estimation method using a novel arc-shaped differential probe with orthogonal receivers is proposed to detect circumferential defects in pipelines. The sensitivity between coils is analyzed to optimize the probe parameters. The scanning signals of defects on the complex plane is characterized by the shell curves (SC), and a Lissajous curve (LC) model is employed for curve fitting. It is observed that a monotonous relationship holds between fitting parameters and defect dimensions. From the fitted scanning curve, the defect depth and length can be estimated by employing the K-nearest neighbors (KNN) algorithm. Experiments have been conducted to verify the proposed defect parameter estimation method by testing aluminum pipelines with cracks. The experimental results show that the estimation of defect parameters achieves with the average accuracy of more than 96.78% for various dimensions of cracks.