Group Role Assignment And Collaborative Optimization Of UAVs In Disaster Areas

After an earthquake,timely and effective disaster relief efforts can greatly reduce casualties and financial losses.In order to make the rescue plan twice as effective,it is crucial to know the situation when the rescue work first starts.Drones can patrol across geographical boundaries to survey the disaster area.However,at the time of rescue,equipment is often in short supply and there are many survey nodes,making for an intricate and complex set of tasks.This leads to a decrease in the effectiveness of multi-UAS collaboration.Therefore,this thesis investigates the UAV cruise assignment for disaster area rescue based on Group Role Assignment（GRA）.The thesis aims to clarify the granularity of tasks in the assignment process,improve the efficiency of assignment collaboration,and optimize the task assignment results.Since rescue in disaster areas is urgent,sudden and time-sensitive,and supplies are often scarce at the time of rescue,an efficient and reasonable assignment method is the key to the UAV cruise assignment in disaster areas.In order to solve the UAV disaster area cruise assignment problem,this thesis researches the UAV assignment problem,and the main research contents include:（1）Modeling the UAV disaster area cruise assignment problem.Role-Based Collaboration（RBC）theory and E-CARGO model are used to abstract the UAV disaster area cruise assignment problem.In this thesis,a clustering algorithm with constraint’k-mean constraint’ is proposed for the role negotiation process of RBC,and the UAV cruise tasks are combined and reorganized into new task sets as roles（Role）.The relationships,coupling and constraints between agents and roles are analyzed and formal modeling is performed.（2）UAV disaster area cruise assignment problem.To analyze the collaborative relationship between Role and Agent,this thesis uses Ant Colony algorithm to solve the cruising path of UAV to complete each task set,and establishes the cost evaluation matrix in GRA according to the path distance.Finally,an optimal assignment result is sought based on the evaluation matrix using GRA to form the completed UAV cruise assignment scheme for the disaster area.（3）Multi-to-multi UAV cruise cooperative assignment optimization.Based on the UAV disaster area cruise assignment problem,we further consider the multi-pairmultiple assignment to improve the multi-UAV cooperative efficiency and maximize the UAV utilization.In this thesis,we use Group Multi Role Assignment（GMRA）and Group Role Assignment with Cooperation and Conflict Factors（GRACCF）to solve the UAV multi problem which is solved by the Group Role Assignment with Cooperation and Conflict Factors（GRACCF）.The collaboration and conflict relationships between the roles in the problem are analyzed to form a collaboration conflict matrix.The GMRA and GRACCF pairs are used to quickly solve for the many-to-many UAV cruise collaborative assignment based on the collaborative conflict matrix.It is shown that the model and algorithm proposed in the appeal can achieve fast convergence even with large-scale data volume and can quickly assign UAV disaster area cruises within seconds.The ’k-mean constrained’ algorithm,UAV disaster area cruise assignment and many-to-many UAV cruise co-process are fast and reliable,which can efficiently and accurately deploy through UAV cruise missions and form an important support for UAV to perform relay rescue missions in disaster areas.