In recent years, small-size commercial drones have become one of the catchiest stars on the stage of science and technology. They consist of multiple rotors and can move neatly in every direction, i.e., they are omni-directional. Although the basic flight control technology of these drones is mature, the rise of drones brings a series of new problems. Users want the drones to use as little energy as possible and have as little shaking as possible while smoothly moving along the self-optimized path according to the preset air route smoothly. All these requirements are based on the trajectory generation and control technology.

This thesis work focuses on the study of a quadrotor. A quadrotor model is needed to prove the differential flatness, and be a part of constraint of algorithm....[ Read more ]

In recent years, small-size commercial drones have become one of the catchiest stars on the stage of science and technology. They consist of multiple rotors and can move neatly in every direction, i.e., they are omni-directional. Although the basic flight control technology of these drones is mature, the rise of drones brings a series of new problems. Users want the drones to use as little energy as possible and have as little shaking as possible while smoothly moving along the self-optimized path according to the preset air route smoothly. All these requirements are based on the trajectory generation and control technology.

This thesis work focuses on the study of a quadrotor. A quadrotor model is needed to prove the differential flatness, and be a part of constraint of algorithm. Thus, the modelling of a dynamic system and the evaluation of the trajectory generation algorithm are carried out. The algorithms need to be tested in many ways. In the first stage, mathematical performance evaluations were performed using MATLAB simulated model. Then, in the second stage, real system behavior were evaluated on a quadrotor platform.

In this thesis, the author built a system to verify the algorithm, compared and analyzed two algorithms to choose the better one for engineering implementation, optimized the algorithm for real application and proposed a method of time allocation.