Exploration And Maneuver Control Methods In Unknown Indoor Environment For Agile Aircraft

Agile aircraft, small and smart, has such features as small size, light weight, goodmaneuverability and the ability to vertical takeoff and landing. These characteristics maymake itself suit for several missions, such as, reconnaissance, surveillance, target acquisitionand even some more complex missions in near-surface environment. Meanwhile, agileaircraft is the first choice for the indoor complex mission. However, it has not been solvedwell that the problem how it can work in the rapid and flexible maneuver under wide rangecomplex indoor environment as has become one of the main limiting for development ofagile aircraft. Therefore, this thesis focuses more on exploration and rapid maneuver problemin agile aircrafts, such as small quadrotors, in unknown indoor environment. The mainresearch works have been made as follows.First of all, it analyzes the agile aircraft’s body structure and flight principle. The6DOFmotions nonlinear dynamic model and quaternion rotational kinematics model have beenestablished with the part of forces and moments. By analyzing the characteristics of aircraftpropulsion system, including propeller, electronic speed controller and battery, we establishthe nonlinear model of propulsion system.Furthermore, we design and implement a flight controller suit for agile aircraft. Basedon the controller, we build a quadrotor flight research testbed, and redesign its internal flightcontrol algorithm meeting the needs of autonomous flight control. We also design a newintegrated testbed that fits for solving out the undetermined coefficients of propulsionsystems of aircraft. We could determine parameters through designing propulsion systems ofrelevant experiments according to integrated testbed. Thus, the theoretical and experimentalfundamentals had be established for later researches.Secondly, we propose an attitude and altitude control approach based on PC-BEPSmodel for quadrotor accurate control. Applying integrated testbed, the experiments andanalyzes on the propulsion system characteristics and boundary effects had been made. Theexperiment results shows that, in the effective working time of battery, the decline ratio ofthe propulsion system output is reached to about20%when the battery voltage drops.Especially, in the second half of the effective working time of battery, the nonlinearcharacteristics of propulsion system will seriously affect flight control performance. Whileboundary effects can increase the propulsion system output, and the rising ratio is up to33%.As a consequence, we establish PCPS compensation model and BEPS model by means of the experimental and analytical results, and then design the attitude and altitude controller toweaken the power characteristics and the boundary effects of propulsion system. In order tosolve another problem about the thrust estimation of propulsion system, we design anobserver to estimate the thrust in the hover state. Simulation results show that it caneffectively estimate thrust in the hover state. And not only does the accurate control methodimpair the power characteristics of propulsion system, but also uses the enhanced efficiencyof boundary effects in thrust as far as possible, and extends the effective working time ofquadrotor as well, and the work time increased nearly double.Thirdly, we propose a nonlinear maneuvering flight control approach based on DI-DSMC, realizing the aircraft’s large curvature and rapid maneuver. We use quaternion todescribe aircraft’s system model. And also, we design a new DSMC feedback controller byselecting the switching function and constructing a new integral sliding mode surface. Thecontroller is not only insensitive to the system model parameter perturbation and disturbance,but also could eliminate the system chattering phenomenon caused by traditional slidingmode switching function. With this, we design the quaternion DI system of aircraft using Liederivative inverse system design method, and we use the system’s output to be thefeedforward compensation. These methods can both improve the dynamic response speed,realize the decoupling and weaken the nonlinear level of the system. Finally, the simulationresults proves that the nonlinear maneuvering flight control method realizes the fast trackingof large angle input, thus, the quadrotor owns a large curvature and rapid change flight ability.Finally, in view of the localization in unknown indoor environment, we successfully doan aided laser inertial navigation by combining Laser scanning rangefinder instead of GPSwith inertial navigation system. The experiment results prove that the navigation algorithmcan help to effectively control the aircraft’s position. Furthermore, we design a LAMOV-AOD system according to the unknown indoor environment characteristics. Then we proposeLOLE-AC algorithm and DEOA algorithm for autonomous obstacle detection. Theexperiment results prove the LAMOV-AOD system can identify obstacles even in the lowlight environment. On this basis of LAMOV-AOD system, we propose SEB-LDPRM andSEB-DLPA*based on sensors with the partial unknown static environment and wholeunknown static environment. The simulation results prove the method of path planning caneffectively plan and re-plan the path based on the sensor information.