Akmeliawati, RiniGrainger, StevenLu, Tien-Fu (De Facto Principal Supervisor)Al-Khulaidi, Rami Ali Ghaleb Ahmed2025-07-292025-07-292025https://hdl.handle.net/2440/146402This thesis investigates the development of integrated trajectory planning and control systems for hyper-redundant manipulators operating in semi-structured environments. The research objectives include developing and optimising a hyper-redundant manipulator for semi-structured environments, developing and implementing an integrated trajectory planning and passivity-based control (PBC) system utilising the Port-Controlled Hamiltonian (PCH) framework, and developing a collision-free trajectory planning method using Non-Linear Model Predictive Control (NLMPC) combined with an artificial bee colony optimisation algorithm. These objectives are addressed through theoretical development, simulation-based validation, and integration of innovative strategies into manipulator design and control frameworks. The research begins with the structural optimisation of a hyper-redundant manipulator featuring rigid short links, designed to enhance dexterity and manipulability. This design proves particularly effective in confined semi-structured environments, such as agricultural greenhouses. Building upon this foundation, an integrated trajectory planning and control system is developed using the Port-Controlled Hamiltonian (PCH) framework. This system employs a novel passivity-based control approach to achieve precise trajectory tracking and improved joint stability in semi-structured environments. To address the challenge of collision-free navigation, the thesis presents a hybrid method combining Non-Linear Model Predictive Control (NLMPC) with artificial bee colony optimisation. This approach demonstrates robust performance in navigating both static and dynamic obstacles, while effectively avoiding self-collisions. Additionally, the research explores an emotion-driven reinforcement learning model, which integrates human-like emotions into the decision-making process. This novel framework significantly enhances manipulators’ adaptability in decision-making, enabling efficient navigation and real-time adjustments in semi-structured and unknown environments. By addressing these research objectives, this thesis makes significant contributions to the field of robotics, offering innovative solutions to the challenges faced by hyper-redundant manipulators in dynamic and semi-structured settings. The findings provide a foundation for advancing manipulator design, trajectory planning, and control strategies, with implications for a wide range of applications in robotics and automation.enhyper-redundant manipulatorssemi-structured environmentstrajectory planningpassivity-based controlNLMPChuman learning behaviouremotion-based reinforcement learningThesis