YFYao Fang

Yao Fang

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Yao Fang has a total of 18 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Koopman operator-augmented physics-informed neural network with SEI kinetics for stable and explainable SOH estimation in battery packs

    by Yao Fang, Chunlan Yang, Xiuchao Chang, Yinlong Zhuang

    Accurate estimation of State of Health (SOH) is pivotal for battery management, yet existing data-driven methods often lack physical interpretability and struggle with generalization under complex operating conditions. To bridge the gap between data-driven flexibility and electrochemical mechanism consistency, this paper proposes a Koopman Operator-Augmented Physics-Informed Neural Network (Koopman-PINN) tailored for battery pack SOH estimation. Unlike traditional black-box models, the proposed method embeds the nonlinear SOH degradation into a high-dimensional Koopman observation space where it evolves linearly. Crucially, the Solid Electrolyte Interphase (SEI) growth kinetics and boundedness conditions are embedded as physical loss terms. A spectral domain constraint is further imposed on the Koopman operator matrix, which forces all eigenvalues inside the unit circle (|λi| < 1) and constrains their phase angles to guarantee the physical consistency and long-term stability of the degradation trajectory. Furthermore, the SEI degradation rate coefficient α is treated as an independent trainable parameter, enabling the end-to-end learning of physical parameters. Validated on extensive battery pack datasets under quasi-constant temperature (804 cycles) and variable temperature (667 cycles) conditions, the Koopman-PINN reduces the MAE by 46.0% and RMSE by 48.9% compared to standard PINNs in quasi-constant temperature. Under variable temperature conditions, Koopman-PINN reduces MAE by 30.0% and RMSE by 29.6% compared to the traditional PINN. Notably, the learned activation energy (≈30 kJ/mol) aligns well with the physical laws of SEI growth, and the spectral analysis reveals distinct dynamical modes governing the degradation process, providing a novel interpretable perspective on battery aging mechanisms.

  • Open Access

    Nonlinear Disturbance Observer-Based Sliding Mode Control for the PMSM with Matched and Mismatched Disturbances

    by Yao Fang, Huifang Kong, Tiankuo Liu

    High-efficiency permanent-magnet synchronous motor (PMSM) is a key technology to improve the driving range of batteries in electric vehicles, while the mismatched disturbance that is caused by external disturbances and parameter perturbation may easily result in speed fluctuations and overshoot of the PMSM, which further deteriorate the performance and efficiency of batteries. To solve the problem, a novel nonlinear disturbance observer-based sliding mode control (NDO-SMC) is proposed. Compared with the traditional SMC method, the NDO-SMC scheme has better disturbance rejection ability in the presence of matched and mismatched uncertainties and disturbances by introducing the estimation value of the nonlinear disturbance observer in the sliding surface. Furthermore, owing to the compensation of the disturbance observer, the switching gain is only required to be greater than the bound of the disturbance estimation error rather than that of the disturbance; thus, the chattering problem is substantially alleviated. A rigorous stability proof of the whole closed-loop system is given in detail using Lyapunov theory by designing an appropriate Lyapunov function. The simulation results demonstrate the feasibility and superiority of the proposed NDO-SMC strategy.


  • Open Access

    Torque Distribution Strategy Based on Deep Learning for Parallel Hybrid Electric Vehicle

    by Yao Fang, Huifang Kong, Lei Fan

    In order to improve the fuel economy of hybrid electric vehicle (HEV), an improved deep recurrent neural network-based energy management strategy model (IDRNN-EMS) is proposed to learn optimal torque distribution strategy for the single-axle parallel hybrid electric vehicle (HEV). Specifically, we formulate EMS as time series prediction problem to meet the requirement of deep recurrent neural network (DRNN). Then, skip connection structure is introduced into deep recurrent neural network to improve the training efficiency of our proposed model. We provide comprehensive simulation evidence showing that the improved model with skip connection is easier to optimize in public platform TensorFlow. Comparing with the DRNN-EMS model, our improved model with ship connection yields remarkable performance in terms of fuel economy and accuracy.

  • Encoder–decoder with double spatial pyramid for semantic segmentation

    by Huifang Kong, Jie Hu, Lei Fan, Xiaoxue Zhang, Yao Fang

    Semantic segmentation, as a dense pixelwise classification task, is of great significance to scene understanding. Many approaches based on convolutional neural network still suffer from two kinds of challenges: (1) insufficient semantic information results in semantic obfuscation between similar categories, (2) loss of spatial information leads to inaccurate location of inconspicuous objects. To tackle these challenges, we design a network with an encoder–decoder architecture based on two proposed modules: global pyramid attention module (GPAM) and pyramid decoder module (PDM). Specifically, GPAM exploits an attention mechanism as global prior knowledge to adaptively capture discriminative features for enhancing semantic representation, and PDM employs small convolutions connected in parallel to predict adjacent position relationships for refining spatial information. A series of ablation experiments are conducted to demonstrate the effectiveness of our designs, and our network achieves a mean intersection over union score of 83.4% on PASCAL VOC 2012 dataset and 78.5% on Cityscapes dataset.


  • Open Access

    A Novel Torque Distribution Strategy Based on Deep Recurrent Neural Network for Parallel Hybrid Electric Vehicle

    by Huifang Kong, Yao Fang, Lei Fan, Hai Wang, Xiaoxue Zhang, Jie Hu

    In this paper, energy management strategy (EMS) model based on deep recurrent neural network is presented to learn optimal torque distribution for the single-axle parallel hybrid electric vehicle. The model has two distinguishing properties: 1) Because EMS is formulated as a time series prediction problem, taking historical data as input of EMS model captures the input-and-output dynamic characteristics and enhances the prediction capability. 2) EMS model based on end-to-end framework directly generates torque distribution results without extracting features of driving cycles and other artificial interference. Extensive simulations are conducted to demonstrate the accuracy and generalization capability of the EMS model in public platform TensorFlow. Comparing with other energy management strategies, our proposed model yields better performance in terms of fuel economy and accuracy. Simulation results show that our proposed EMS model provide a novel way to study the energy management strategy.


  • Open Access

    Effect of Slot Wall Jet on Combustion Process in a 660 MW Opposed Wall Fired Pulverized Coal Boiler

    by Yong Zhang, Yao Fang, Baosheng Jin, Youwei Zhang, Chunlei Zhou, Farooq Sher

    Numerical investigations of an anti-corrosion design and the combustion process (original conditions and optimal conditions) were conducted for a 660 MW opposed wall fired boiler. In order to solve high-temperature corrosion of the side wall, a scheme was proposed: slotting in the side wall and introducing air (closing-to-wall air) from the secondary air. The effect of anti-corrosion was disclosed in detail by varying the structures of slotting, gas velocities from nozzles and jet inclination angles. The temperature and NOx distribution in the furnace at optimized conditions were compared with those at the original operating conditions. Simulation results showed that the structures of the slot and gas velocities from the nozzles had a marked effect on anti-corrosion of the side wall. When the gas velocity was 4 m/s, an inclination angle of the gas velocity was not conducive to anti-corrosion of the side wall. When the gas velocity increased at the middle and bottom of the side wall, the anti-corrosion effect increased significantly. When the optimal scheme was adopted, the corrosion area of the side wall decreased obviously, but the furnace temperature and the NOx emission increased slightly. The detailed results of this work promote a full understanding of closing-to-wall air and could help to reduce the corrosive area in pulverized-coal furnaces or boilers.


  • Open Access

    Neural Network PID Algorithm for a Class of Discrete-Time Nonlinear Systems

    by Huifang Kong, Yao Fang

    The control of nonlinear system is the hotspot in the control field. The paper proposes an algorithm to solve the tracking and robustness problem for the discrete-time nonlinear system. The completed control algorithm contains three parts. First, the dynamic linearization model of nonlinear system is designed based on Model Free Adaptive Control, whose model parameters are calculated by the input and output data of system. Second, the model error is estimated using the Quasi-sliding mode control algorithm, hence, the whole model of system is estimated. Finally, the neural network PID controller is designed to get the optimal control law. The convergence and BIBO stability of the control system is proved by the Lyapunov function. The simulation results in the linear and nonlinear system validate the effectiveness and robustness of the algorithm. The robustness effort of Quasi-sliding mode control algorithm in nonlinear system is also verified in the paper.