Zhifeng Dou
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Zhifeng Dou has a total of 20 co-authors. Here are the top co-authors by number of shared publications:
- Falong Lu
College of Electrical and Information Engineering, Zhengzhou University of Light Industry, Zhengzhou, China
3 shared publications - Qian Wang
Zhengzhou University of Light Industry
3 shared publications Xiaoyu An
2 shared publicationsChongyan Zhao
College of Electric and Information Engineering, Zhengzhou University of Light Industry, Zhengzhou, China
1 shared publicationFalong Lu
1 shared publication- Guangzhao Cui
Zhengzhou University of Light Industry
1 shared publication
Latest Publications
- 13 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Power Electronics
- Vol. 41 (9) , 16209-16222
A Reverse Design Paradigm: Modular Constraint-Driven Reinforcement Learning for Automated Derivation of Switched-Capacitor Multilevel Inverters
Artificial intelligence (AI), with its advanced capabilities in multi-objective optimization and search, offers a promising solution to the combinatorial explosion problem in power electronics. This problem has long hindered traditional multilevel topology design, which is constrained by intricate component integration and complex high-dimensional modeling. This article introduces a modular, scalable, and requirement-driven derivation framework for deriving complex multilevel converter topologies using reinforcement learning (RL). The framework decouples topology constraints into two core constraint rewards via a hierarchical reward mechanism: 1) the basic constraint reward (BCR) for forward design, and 2) customized functional constraint reward (FCR) for reverse design. This framework was successfully applied to derive multilevel inverter topologies, resulting in 20 distinct five-level common-grounded switched-capacitor multilevel inverter (CGSCMLI) structures, including multiple novel configurations. All constraints are encoded as concise graph-theoretic formulations, avoiding complex numerical calculations. Finally, experimental verification of a representative novel topology verified its functional completeness and demonstrated the efficacy of the framework.
- 13 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Power Electronics
- Vol. 41 (9) , 16209-16222
A Reverse Design Paradigm: Modular Constraint-Driven Reinforcement Learning for Automated Derivation of Switched-Capacitor Multilevel Inverters
Artificial intelligence (AI), with its advanced capabilities in multi-objective optimization and search, offers a promising solution to the combinatorial explosion problem in power electronics. This problem has long hindered traditional multilevel topology design, which is constrained by intricate component integration and complex high-dimensional modeling. This article introduces a modular, scalable, and requirement-driven derivation framework for deriving complex multilevel converter topologies using reinforcement learning (RL). The framework decouples topology constraints into two core constraint rewards via a hierarchical reward mechanism: 1) the basic constraint reward (BCR) for forward design, and 2) customized functional constraint reward (FCR) for reverse design. This framework was successfully applied to derive multilevel inverter topologies, resulting in 20 distinct five-level common-grounded switched-capacitor multilevel inverter (CGSCMLI) structures, including multiple novel configurations. All constraints are encoded as concise graph-theoretic formulations, avoiding complex numerical calculations. Finally, experimental verification of a representative novel topology verified its functional completeness and demonstrated the efficacy of the framework.
- 18 March 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Power Electronics
- Vol. 40 (8) , 11369-11379
A Novel Simultaneous Diagnosis Method for IGBT Open-Circuit Faults and Current Sensor Faults of Three-Phase SPWM Inverter
This article presents a novel fault diagnosis method for open-circuit faults of insulated gate bipolar transistors (IGBTs) and current sensor faults in three-phase sinusoidal pulsewidth modulation inverters. The proposed method facilitates the simultaneous diagnosis of multiple IGBT and current sensor faults. A combination of Park transform and fast Fourier transform (FFT) is employed to extract fault features. The Park transform converts three-phase current into two-phase current, thereby reducing the dimensionality of the data. The FFT integrates time-domain and frequency-domain analyses to enhance the representation of fault characteristics. Additionally, fault features are extracted from the samples post-FFT to minimize the data volume of individual sample input by the model. For fault diagnosis, a temporal convolutional neural network model incorporating stacked residual blocks is developed to train and evaluate potential faults and demonstrate the model's validity and accuracy. The results of online simulations and hardware-in-the-loop experiments indicate that the proposed method can identify and locate the IGBT and sensor faults within 32 ms, and the accuracy is 98.75%. This method achieves higher training and testing accuracy and faster training and testing speed, which is superior to the existing diagnosis technologies.
- 18 March 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Power Electronics
- Vol. 40 (8) , 11369-11379
A Novel Simultaneous Diagnosis Method for IGBT Open-Circuit Faults and Current Sensor Faults of Three-Phase SPWM Inverter
This article presents a novel fault diagnosis method for open-circuit faults of insulated gate bipolar transistors (IGBTs) and current sensor faults in three-phase sinusoidal pulsewidth modulation inverters. The proposed method facilitates the simultaneous diagnosis of multiple IGBT and current sensor faults. A combination of Park transform and fast Fourier transform (FFT) is employed to extract fault features. The Park transform converts three-phase current into two-phase current, thereby reducing the dimensionality of the data. The FFT integrates time-domain and frequency-domain analyses to enhance the representation of fault characteristics. Additionally, fault features are extracted from the samples post-FFT to minimize the data volume of individual sample input by the model. For fault diagnosis, a temporal convolutional neural network model incorporating stacked residual blocks is developed to train and evaluate potential faults and demonstrate the model's validity and accuracy. The results of online simulations and hardware-in-the-loop experiments indicate that the proposed method can identify and locate the IGBT and sensor faults within 32 ms, and the accuracy is 98.75%. This method achieves higher training and testing accuracy and faster training and testing speed, which is superior to the existing diagnosis technologies.
- 1 March 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 2330-2336
Finite states model predictive direct power control for phase leg faults tolerant operation of bidirectional AC/DC converter
Based on the operation analysis of bidirectional AC/DC converter with leg faults, a power predictive model of four-switch three-phase (FSTP) in αβ coordinates is established. The finite states model predictive direct power control (MPDPC) method of fault-tolerant FSTP converter is designed for continuous operation, even if the phase leg of bidirectional AC/DC converter has a fault. According to the power predictive model and cost function, the optimal space voltage vector is selected to achieve flexible seamless switching between inverter and rectifier mode with direct power control. DC-link split capacitor voltage balance control is achieved by injecting a dc bias current into the faulty phase current. The proposed method does not need pulse width modulation signals and current control loop. It has the advantages of only minor calculations are required and it is easy to be implemented. The simulation and experiment results show good steady state and dynamic performance of proposed control method for bidirectional AC/DC converter with leg faults.
- 1 March 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 2330-2336
Finite states model predictive direct power control for phase leg faults tolerant operation of bidirectional AC/DC converter
Based on the operation analysis of bidirectional AC/DC converter with leg faults, a power predictive model of four-switch three-phase (FSTP) in αβ coordinates is established. The finite states model predictive direct power control (MPDPC) method of fault-tolerant FSTP converter is designed for continuous operation, even if the phase leg of bidirectional AC/DC converter has a fault. According to the power predictive model and cost function, the optimal space voltage vector is selected to achieve flexible seamless switching between inverter and rectifier mode with direct power control. DC-link split capacitor voltage balance control is achieved by injecting a dc bias current into the faulty phase current. The proposed method does not need pulse width modulation signals and current control loop. It has the advantages of only minor calculations are required and it is easy to be implemented. The simulation and experiment results show good steady state and dynamic performance of proposed control method for bidirectional AC/DC converter with leg faults.