学位论文简介

The doctoral dissertation focuses on discrete-time filter approximations of integrators and differentiators of integer and fractional-order orders. It proposes several design methods for discrete-time infinite-impulse-response (IIR) wideband approximations of integrators, differentiators, and fractional-order differ-integrators to address the need for accurate modeling of integrators and differentiators in various digital signal processing applications. The key innovation points of the dissertation can be summarized as follows:

1. Closed-Form Design Method for IIR Integrators: A novel closed-form design method   for IIR linear-phase integrators using numerical integration rules constructed via trigonometric interpolation is proposed. This method allows for the design of IIR integrators that satisfy prescribed frequency-domain specifications. The effectiveness of the proposed method is demonstrated through simulation comparisons and experimental verification in real applications.

2. Closed-Form Design Method for Fractional-Order Differ-Integrators: A new closed-form design method for IIR fractional-order differ-integrators is proposed. This method utilizes a modified indirect discretization strategy based on continued fraction expansion to minimize truncation errors around a predefined low frequency. The proposed method shows potential for achieving low-order accurate models suitable for hardware implementations.

3. Efficient Realization of variable fractional-order differ-integrator: A new ladder structure for the implementation of variable fractional-order differ-integrator is proposed. The new structure uses either direct discretization scheme or indirect discretization scheme. To reduce the implementation complexity, new discretization operators are proposed. The new structure is further adapted for the implementation of generalized fractional-order differ-integrator whose magnitude and phase can be controlled independently.

4. Iterative Design Method for IIR differentiators and integrators: A novel objective function based on a weighted L1-norm, in which the weight is inversely proportional to the magnitude, is combined to a metaheuristic algorithm called salp swarm algorithm for the optimal design of stable, minimum phase and wideband IIR differentiators and integrators. The proposed weighting function improves the phase linearity in the vicinity of the zero frequency, and promotes to a distribution of the error such that the lower the magnitude, the lower is the error. The performance of the proposed approach is extensively investigated by comparisons with well-known methods.

5. Iterative design method for IIR fractional-order differ-integrator: A novel design method based on L1-norm optimally criterion and a metaheuristic technique, namely, multi-verse optimizer, for IIR fractional-order differentiators and integrators. The proposed method enables an accurate implementation of stable, minimum phase and wideband IIR fractional-order differentiators and integrators with flat magnitude response. The performance of the proposed method is extensively investigated against that of well-known methods.

Overall, the doctoral dissertation contributes to the advancement of discrete-time filter approximations for integrators and differentiators, providing innovative design methods and demonstrating their superiority over existing approaches.

主要学术成果

1. Talal Ahmed Ali Ali, Zhu Xiao, Jingru Sun, Seyedali Mirjalili, Vincent Havyarimana, Hongbo Jiang, Optimal design of IIR wideband digital differentiators and integrators using salp swarm algorithm, Knowledge-Based Systems, Volume 182, 2019, 104834. (SCI, 第一作者)

2. Talal Ahmed Ali Ali, Zhu Xiao, Seyedali Mirjalili, Vincent Havyarimana, Efficient design of wideband digital fractional order differentiators and integrators using multi-verse optimizer, Applied Soft Computing, Volume 93, 2020, 106340. (SCI, 第一作者)

3. Talal Ahmed Ali Ali, Zhu Xiao, Hongbo Jiang, Bo Li, A Class of Digital Integrators Based on Trigonometric Quadrature Rules, IEEE Transactions on Industrial Electronics, (Early Access). DOI: 10.1109/TIE.2023.3290247 (SCI, 第一作者)

4. Talal Ahmed Ali Ali, Zhu Xiao, Ahmed Jawad Albdairi, Hongbo Jiang, A Closed-form IIR Approximation of Fractional Operator $s^{\nu}$ Around A Prescribed Low Frequency, IEEE Signal Processing Letters. (SCI, 第一作者)

5. Talal Ahmed Ali Ali, Zhu Xiao, Hongbo Jiang, IIR Ladder Realization of Variable Fractional-Order Differintegrators, IEEE Transactions on Circuits and Systems I: Regular papers. (Under review)

6. Bai, Jing, Wei Shi, Zhu Xiao, Amelia C. Regan, Talal Ahmed Ali Ali, Yongdong Zhu, Rui Zhang, and Licheng Jiao. "Hyperspectral image classification based on superpixel feature subdivision and adaptive graph structure." IEEE Transactions on Geoscience and Remote Sensing 60 (2022): 1-15. (SCI, 第五作者)

7. Bai, Jing, Wei Shi, Zhu Xiao, Talal Ahmed Ali Ali, Fawang Ye, and Licheng Jiao. "Achieving better category separability for hyperspectral image classification: A spatial–spectral approach." IEEE Transactions on Neural Networks and Learning Systems (2023). (SCI, 第四作者)

8. Taresh, Mundher Mohammed, Ningbo Zhu, Talal Ahmed Ali Ali, Mohammed Alghaili, Asaad Shakir Hameed, and Modhi Lafta Mutar. "KL-MOB: automated COVID-19 recognition using a novel approach based on image enhancement and a modified MobileNet CNN." PeerJ Computer Science 7 (2021): e694. (SCI, 第三作者)

9. Bai, Jing, Junjie Ren, Zhu Xiao, Zheng Chen, Chengxi Gao, Talal Ahmed Ali Ali, and Licheng Jiao. "Localizing From Classification: Self-Directed Weakly Supervised Object Localization for Remote Sensing Images." IEEE Transactions on Neural Networks and Learning Systems (2023). (SCI, 第六作者)

10. Taresh, Mundher Mohammed, Ningbo Zhu, Talal Ahmed Ali Ali, Asaad Shakir Hameed, and Modhi Lafta Mutar. "Transfer learning to detect covid-19 automatically from x-ray images using convolutional neural networks." International Journal of Biomedical Imaging 2021 (2021): 1-9. (ESCI, 第三作者)

11. Taresh, Mundher Mohammed, Ningbo Zhu, Talal Ahmed Ali Ali, Mohammed Alghaili, and Weihua Guo. "Using a novel fractional-order gradient method for CNN back-propagation." arXiv preprint arXiv:2205.00581 (2022). (Preprint, 第三作者)