A5100434101- Machine Fault Diagnosis Techniques
- Electric Motor Design and Analysis
- Metaheuristic Optimization Algorithms Research
- Sensorless Control of Electric Motors
- Artificial Immune Systems Applications
- Advanced Algorithms and Applications
- Industrial Vision Systems and Defect Detection
- Advanced Decision-Making Techniques
- Magnetic Properties and Applications
- Advanced Memory and Neural Computing
- Evaluation and Optimization Models
- Remote Sensing and Land Use
- Color Science and Applications
- Industrial Technology and Control Systems
- High-Voltage Power Transmission Systems
- Surface Roughness and Optical Measurements
- Advanced Image Fusion Techniques
- Remote-Sensing Image Classification
- Engineering Diagnostics and Reliability
- Non-Destructive Testing Techniques
- Power Systems Fault Detection
- Evaluation Methods in Various Fields
- Fault Detection and Control Systems
- Hearing, Cochlea, Tinnitus, Genetics
- Smart Grid and Power Systems
Hunan University of Science and Technology
2015-2025
Taiyuan University of Science and Technology
2024
Shanghai University of Electric Power
2012-2024
Electric Power University
2024
Shanghai Electric (China)
2024
China Electronics Technology Group Corporation
2024
Jiangsu University
2008-2023
Wuhan University of Technology
2023
Sichuan University
2022
First Affiliated Hospital of Zhengzhou University
2021
A global parameter estimation method for a permanent magnet synchronous machines (PMSM) drive system is proposed, where the electrical parameters, mechanical and voltage-source-inverter (VSI) nonlinearity are regarded as whole formulated single optimization model. dynamic learning estimator proposed tracking VSI of PMSM by using self-learning particle swarm (DSLPSO). In DSLPSO, novel movement modification equation with exemplar strategy designed to ensure its diversity achieve reasonable...
Fault diagnosis of rolling bearings is an essential process for improving the reliability and safety rotating machinery. It always a major challenge to ensure fault accuracy in particular under severe working conditions. In this article, deep adversarial domain adaptation (DADA) model proposed bearing diagnosis. This constructs network solve commonly encountered problem numerous real applications: source target are inconsistent their distribution. First, stack autoencoder (DSAE) combined...
Fault diagnosis plays an indispensable role in prognostics and health management of rotating machines. In recent years, intelligent fault methods based on domain adaptation technology have attracted the attention researchers. However, a more extensive application scenario - partial (PDA) has not been well-resolved. this article, for first time, novel stacked auto-encoder adversarial (SPADA) model is proposed to solve problem PDA situations. Two deep stack auto-encoders are designed extract...
With the use of sensorless control strategy, mechanical position sensors can be removed from gearbox, so as to decrease maintenance costs and enhance system robustness. In this paper, a switching PI based model reference adaptive (MRAS) observer using Fuzzy-Logic-Controller (FLC) is introduced for permanent magnet synchronous motor (PMSM) drives. The main work innovation paper include: 1) A disturbance voltage compensation method proposed solve problems distortion in source inverter (VSI)...
A hierarchical fast parallel co-evolutionary immune particle swarm optimization (PSO) algorithm, accelerated by graphics processing unit (GPU) technique (G-PCIPSO), is proposed for multiparameter identification and temperature monitoring of permanent magnet synchronous machines (PMSM). It composed two levels developed based on compute unified device architecture (CUDA). In G-PCIPSO, the antibodies (Abs) higher level memory are selected from lower swarms improved clonal-selection operator....
In this paper, a coevolutionary particle-swarm-optimization (PSO) algorithm associating with the artificial immune principle is proposed. proposed algorithm, whole population divided into two kinds of subpopulations consisting one elite subpopulation and several normal subpopulations. The best individual each will be memorized during evolution process. A hybrid method, which creates new individuals by using three different operators, presented to ensure diversity all Furthermore, simple...
Remaining useful life (RUL) prediction for condition-based maintenance decision making plays a key role in prognostics and health management (PHM). Accurately predicting RUL of the rotating components complex machines becomes challenging task PHM. For many existing methods, current error may be accumulated into future predictions, thus can lead to superposition problem. In this article, formation mechanism is analyzed, first time deep adversarial long short-term memory (LSTM) prognostic...
In this paper, an accurate parameter estimation model of surface permanent magnet synchronous machines (SPMSMs) is established by taking into account voltage-source-inverter (VSI) nonlinearity. A fast dynamic particle swarm optimization (DPSO) algorithm combined with a receptor editing (RE) strategy proposed to explore the optimal values estimations. This combination provides accelerated implementation on graphics processing unit (GPU), and method is, therefore, referred as G-DPSORE....
Zeroth‐order spatial electromagnetic force wave has been regarded as one of the main sources for vibration and noise integer slot multipole permanent magnet synchronous machine (PMSM) used in electric vehicles. In this study, 48‐slot‐8‐pole built‐in motor a vehicle was taken an example. The characteristic parameters (frequency amplitude) zeroth‐order PMSM were presented. Based on this, method stator rotor structure optimisation proposed to effectively reduce magnitudes by increasing natural...
A novel d-axis variable-structure control (VSC) current regulator is proposed for realizing fast and accurate sensorless in the surface-mounted permanent magnet synchronous motor (SPMSM) this article. In regulator, sigmoid function employed as a switching law to reduce chattering effect increase degrees of freedom tuning regulator. The performance under VSC testified by using model reference adaptive system (MRAS) sliding-mode (SMC) method. MRAS observer, speed resistance designed so that...
Rotating machinery working under changing operation conditions is prone to failure. In recent years, domain adaptation has been successfully used for fault diagnosis. However, the existing diagnosis methods based on have two main disadvantages: 1) with these methods, it difficult precisely measure and estimate differences between source target domains 2) they only consider discrepancies in feature space, but not label space. this article, a new optimal transport (OT)-based deep model...
In recent years, most existing domain-adapted bearing fault diagnoses for rotating machinery have been designed to decrease domain drifts various operating conditions with an assumption that sufficient tag data are available. To overcome scarcity, a possible solution is use information of other machines the same category diagnose status target machine (i.e., cross-machine diagnosis). This article proposes variational auto-encoder (VAE)-based multisource deep adaptation model using optimal...
Abstract Background The enlarged vestibular aqueduct (EVA) is the commonest malformation of inner ear accompanied by sensorineural hearing loss in children. Three genes SLC26A4 , FOXI1 and KCNJ10 have been associated with EVA, among them being most common. Yet, hotspot mutation screening can only diagnose a small number patients. Methods Thus, this study, we designed new molecular diagnosis panel for EVA based on multiplex PCR enrichment next‐generation sequencing exon flanking regions . A...
The identification of physical parameters is crucial for control system designs, condition monitoring and fault diagnosis industrial drive systems. article brings multicore architecture based parallel computing technology bioinspired intelligent optimisation algorithm insight into designing parameter estimation models.models. In this study, a implementation using an immune-cooperative dynamic learning particle swarm optimization (PSO) with computation architectures presented permanent magnet...
Accurate real-time object recognition from sensory data has long been a crucial and challenging task for autonomous driving. Even though deep neural networks (DNNs) have successfully applied in this area, most existing methods still heavily rely on the pre-processing of pulse signals derived LiDAR sensors, therefore introduce additional computational overhead considerable latency. In paper, we propose an approach to address problem directly with raw temporal pulses utilizing spiking network...
This article presents a novel strategy for improving the well-established component substitution-based multispectral image fusion methods, because fused results obtained by substitution methods tend to exhibit significant spectral distortion. The main cause of distortion is analyzed and discussed based on method's general model. An improved scheme derived from sensitivity imaging model refine approximate spatial detail obtain one that almost ideal. experimental two data sets show when it has...
Objective This study was performed to evaluate the association of preoperative anxiety with inflammatory indicators and postoperative complications in patients undergoing scheduled aortic valve replacement surgery. Methods A prospective cohort performed. The Hamilton Anxiety Scale used assess anxiety. serum white blood cell (WBC) count concentrations C-reactive protein, interleukin (IL)-6, IL-8 were measured 1 day preoperatively 3 7 days postoperatively. Postoperative also recorded. Results...
Spiking neural network (SNN) is interesting both theoretically and practically because of its strong bio-inspiration nature potentially outstanding energy efficiency. Unfortunately, development has fallen far behind the conventional deep (DNN), mainly difficult training lack widely accepted hardware experiment platforms. In this paper, we show that a temporal-coded SNN can be trained easily directly over benchmark datasets CIFAR10 ImageNet, with testing accuracy within 1% DNN equivalent size...
Event-driven sensors such as LiDAR and dynamic vision sensor (DVS) have found increased attention in high-resolution high-speed applications. A lot of work has been conducted to enhance recognition accuracy. However, the essential topic delay or time efficiency is largely under-explored. In this paper, we present a spiking learning system that uses neural network (SNN) with novel temporal coding for accurate fast object recognition. The proposed scheme maps each event's arrival data into SNN...