Research On UAV Path Planning And Formation Control Method In Earth Observation

Earth observation aims to monitor the space-time information on surface in real time using various types of sensing equipment,including space,air and ground.It provides scientific basis for humans to study the geospace environment and making decisions on environmental protection.It also supplies geospatial information to support national defense construction and national strategic deployment.Moreover,it offers positioning services for the economic construction and daily life of humans.Many countries have accordingly formulated strategy for further earth observation,aiming to explore and develop new technologies for ground observation.Developing advanced ground observation technology has also been integrated into scientific development strategy by the Chinese government,and it was included in the Guidelines for the Medium-and Long-Term National Science and Technology Development Program(2006-2020)where they aim to build up an observation system including space-based,air-based and ground-based observation platforms.As the core of low-altitude ground observation platform,UAVs have characteristics of small size,high maneuverability,low cost,low detectability and no need to consider airborne life.However,lightweight structural design,unpredictable observing environment and complex mission requirements have led to problems such as limited energy utilization,low observation efficiency,and inconsistency of observation resolution.While how to plan an optimal path to shorten the mission trajectory,how to gynamically control the flight to guarantee the observation results and how to control multi-UAV cooperation to achieve resources complementarity as well as information fusion are effective to solve the above problems and they are also challenges for UAV ground observation.Path planning is the basis and core of UAV ground observation mission.The latency obstacles such as mountains,buildings,and towers from environment and some unpredictable threats would increase the flight risk of an UAV.Thus,an effective path must be planed to guarantee the safe of UAV.Meanwhile,the timing plot when the sub-tasks are executing is particularly determined by flight paths,resulting that the exactness of the designed path has decisive effect on the efficiency and success rates of a mission.Additionally,the change of 3D terrain brings dimamically changing of observation cross section,inconsistency of observation resolution and uneven energy consumption.Thus,a covering path planning with dynamic adjustment ability is needed to improve observation quality and energy utilization.With the limitation of load weight,the battery capacity and fuel quantity on UAVs are restricted,which immensely influnces the flight and observation ability of single UAV.Furthermore,single UAV execute a mission time after time will extend the task period and increases the security threat,reducing the execution efficiency of system and even leading to task failure.Meanwhile,single UAV excure the same task back and forth will cause time-space discontinuum of the information flow,reducing the observation information integrity.On the contrary,multi-UAV cooperation can achieve the fusion of heterogeneous resource,improving the time-space discontinuum of the information as well as increasing the integrity and robustness of system.However,the complexity of a UAV's dynamic and kinematic models brings path planning and cooperative control a great deal of difficulty and challenge.Since the six-degree of freedom movement of a UAV needs to be described by multivariate differential equations and the state variables are tightly coupled,it is arduous to analyze and acquire the complete solutions.With the constraints of the motion itself,task requirements and environmental factors,cooperative control of multi-UAVs is and challenging.This is mainly because that the restrictions on UAV dynamics and task schedules result in difficulty in uniform description and solving complexity.In summary,it is of great theoretical and application significance to research on path planning and formation of UAVs.Based on the UAV's kinematics theory,graph theory and combination-optimization theory,this thesis focuses on path planning and formation controlling problems.The main contents are presented as follows.(1)In order to solve the problem of compatibility for path feasibility and accuracy,a continuous and bounded curvature path planning algorithm is proposed.The planning workflow is divided into two phases:path smoothing and path re-planning.The problem of curvature discontinuity has been solved firstly through the insertion and translation of the path point,on the basis of which,the path whose curvature exceeds a threshold has been local re-planned using the minimum curvature circle connection algorithm and accomplish the bounded control of a curvature value and reduce the computational complexity effectively.The results show that the actual flight path of the UAV has a good contact ratio with the design path and passes through all the path points by importing the resulting path parameters into the UAV physical model(2)Aiming at the problem of observation resolution consistency and energy efficiency,a coverage path planning algorithm is proposed which maintains observed resolution with resource consumption optimization.Through the construction of the digital terrain surface model,the flight path of the UAV is optimized on the basis of the dynamics of the UAV,and the flight control parameters of the UAV can also be obtained,which can realize the optimal control of the UAV.The results show that the proposed coverage path planning method can reduce the energy consumption by 30%compared with the traditional method.(3)Aiming at the observation and control problems caused by the lack of satellite information loss,a coverage path planning method based on position estimation is proposed.Based on the Bayesian filter theory and the UAV cooperative turning principle,the Markov position and attitude estimation model is established,and the quantitative relationship between the position,attitude and observation coverage of UAV is deduced.Then,we propose a coverage path planning algorithm based on particle swarm optimization(Particle Swarm Optimization,PSO).The simulation results show that the algorithm can effectively elimilate the non-Gaussian white noise in the nonlinear system,and has good position estimation ability,which can realize the regional coverage observation according to the expected target.(4)To solve the problem of low efficiency and inconsistency between the spatial and temporal resolution of single UAV during observation,the thesis proposes a multi-UAV formation keeping and energy-optimal formation adaption control method.A ?-?-? control law and dynamic visual distance estimation method are put forward.For the first time,the eigenvalue law of the associated matrix is deduced based on the singular value feature decomposition.The equivalent size calculation method of the Cartesian space coordinate system is given,and the adaptive transform algorithm based on the switching energy is proposed,which achieve the optimization of the cluster energy under the condition of the synchronize arrival of controlling many unmanned aerial vehicles.The experimental results show that the algorithm can effectively improve estimation accuracy when the attitude of a camera changes.Also,stable formation topology and flexible formation reconfiguration can be achieved by using the proposed method.The proposed path planning algorithm and the cooperative fleet control algorithm can be able to form a complete UAV mission system key technical link incoperated with the existing obstacle avoidance and task allocation technologies,which can achieve a closed-loop structure for UAV-based ground observation platform,and that effectively improves the reliability and efficiency of the unmanned aerial observation system.Moreover,it plays an important role in promoting the research and application of the three-dimensional gorund observation.