Embedded Files
Prof. Dr. Mohammad Osman Tokhi
Prof. Dr. Mohammad Osman Tokhi
Speech Topic: Swarm intelligence control of hybrid dynamic systems
The control of hybrid dynamic systems poses certain challenges pertaining to their dynamic behaviour which covers a broad range of frequencies. A promising approach is to realise decoupled control loops for flexible and rigid dynamics of the system. Such an approach provides the flexibility of designing a range of controller options suitable for the system dynamics. Traditional approaches of controller design are model based, where the quality of the control system depends on the quality of the system model, which is often affected due to modelling errors. The emergence of swarm intelligence approaches has led to the adoption of non-model-based control design strategies. Such strategies lead to potentially more accurate control designs compared to model-based approaches, by removing the system modelling errors from the design process. The focus of this presentation is the design of a range of controller types for hybrid dynamic systems using swarm intelligence. The control types include traditional adaptive control, inverse model controls, command shaping with feedback control. Single-objective and multi-objective requirements are incorporated in the design processes. The designs and system performances are exemplified in the control of a flexible robotic manipulator system and a twin-rotor system.
Biography
Professor Tokhi received his BSc in Electrical Engineering from Kabul University, Afghanistan in 1978 and PhD degree from Heriot-Watt University, UK in 1988. He is a Chartered Engineer, Fellow of IET (Institution of Engineering and Technology), Life Senior Member of IEEE (Institute of Electronic and Electrical Engineering), Member of IIAV (International Institute of Acoustics and Vibration) and Member of CLAWAR (Climbing and Walking Robots) Association. He has worked for industry and at various higher education establishments.
His research interests include Active Noise and Vibration Control, Adaptive/intelligent Control, Soft-computing Modelling and Control of Dynamic Systems, and Assistive Robotics. He has published extensively and has completed numerous projects in these areas.
He is Editor-in-Chief of Journal of Low-Frequency Noise, Vibration and Active Control, Editor of book series on Service Robotics (published by World Scientific Publishing Company), Co-editor of book series on Mobile Service Robotics (published by Elsevier Publishing Company), member of Editorial Advisory/Editorial Board of Industrial Robot Journal, Artificial Intelligence and Technology, Mechanical System Dynamics, Sensors, and Modelling, Identification and Control. He has acted as Editor-in-Chief, Associate Editor, and member of editorial board of several other international journals.
He is Chair of CLAWAR Association, Chair of BSI (British Standards Institute) AMT10/1 Sub-Committee on Ethics of Robots and Autonomous Systems, Convenor of ISO technical committee TC299 (Robotics) working group WG2 – Service Robot Safety, Member of several working groups of ISO/TC299–Robotics Committee and of BSI AMT/010 Robotics Committee, Member of IFAC Technical Committees TC 3.1 (Computers for Control) and TC 3.2 (Computational Intelligence for Control).
Assoc. Prof. Ts. Dr. Mohammad Fadhil Abas
Assoc. Prof. Ts. Dr. Mohammad Fadhil Abas
Speech Topic: Machine Learning and Optimization in Drone Anomalies Detection
Ensuring the safety and reliability of drones is critical as they become integral to industries like logistics, agriculture, and infrastructure inspection. Leveraging advancements in machine learning, AI, and optimization, drone anomaly detection systems are evolving from traditional rule-based methods to intelligent, proactive solutions. Unlike conventional systems that rely on fixed thresholds, AI-based methods analyze large-scale data, detect subtle patterns, and adapt to dynamic environments. AI techniques, such as advanced pattern recognition, intelligent decision-making, and swarm intelligence, enable real-time fault detection, resilience, and collaboration among drones. Optimization further enhances these systems by efficiently allocating resources and computing optimal flight paths to minimize risks. Emerging trends, including Edge AI, adaptive learning, and 5G integration, promise even greater real-time capabilities and efficiency. While challenges like noisy data and adversarial attacks remain, innovations such as hybrid models and federated learning are paving the way for robust solutions. This convergence of AI and optimization is transforming drones into intelligent collaborators, ensuring safer, more reliable operations across industries and shaping a smarter, safer future.
Biography:
Ts Dr Mohammad Fadhil Bin Abas is Associate Professor in Universiti Malaysia Pahang Al-Sultan Abdullah (UMPSA), Malaysia in Electrical and Electronics Engineering Technology. He has been with UMPSA since 2004. He received Philosophy Doctor in Artificial System Science from Chiba University, JAPAN in 2013 under the topic Unmanned Aerial Vehicle Formation Control, Master of Science in Power System from Universiti Putra Malaysia, Malaysia in 2006 and Bachelor in Electric, Electronics, and System from Universiti Kebangsaan Malaysia, Malaysia in 2001. His main interests include fault detection and identification, Unmanned Aerial Vehicle, underactuated mobile robot, navigation and positioning system, optimization technique, machine learning and artificial Intelligence, and water quality systems. He is the investigator of 29 research grants including Fundamental Research Grant Scheme (FRGS). He has won numerous Gold Medal in ITEX and CITREX. Currently, he holds a position of Head of Drone Research Network in UMPSA.
Page updated
Report abuse