چكيده به لاتين
The use of unmanned aerial vehicles has grown dramatically in the last decade due to the high capabilities of these systems. Aerial manipulator that belong to the unmanned aerial vehicle family have wide applications in civilian industries such as physical inspection of infrastructure, intelligent response to natural disasters, pest spraying in agriculture and meteorological fields. Aerial manipulators come in two helicopter and multi-rotor platforms. To increase the efficiency of these systems, they use the ability to collaborate and network. Collaboration with such systems has always been a challenge. The main purpose of this thesis is to co-operate a network of Aerial manipulators with a multi-rotor platforme that enables the communication network to have a delay in communication as well as a switching communication topology. Multi-rotor structure is one of the under-actuated systems, meaning that the number of control inputs is less than the number of variables required for configuration. It also introduces the use of a mechanical arm to configure redundant air arm configuration. Therefore, controlling and networking the aerial manipulator presents serious challenges. It is assumed that all the variables of the state of the aerial manipulators are available and these variables are shared with a delay in the between the agents. To obtain the dynamic model of the system, the total potential and kinetic energy of the aerial manipulator is calculated then using the Lagrange equation to reach the system model. Maple software was used to obtain the mass, spring and damping matrices of dynamic equations. The obtained model is simulated for validation and the results show the accuracy of the dynamic model.Feedback linearization algorithm is used to control the aerial manipulator. With this algorithm, the problem of being under-actuated system is considered. The proposed control algorithm consists of two main layers. A layer is provided to solve the kinematic inverse of the aerial manipulator to calculate the desier state values. The second layer is responsible for achieving these desier values despite the associated delay and switching topology, and finally, to investigate the performance of the proposed algorithm, the equations are simulated in MATLAB software, which results in acceptable controller performance.
