Abstract: A novel method for coalition formation of multiple heterogeneous unmanned aerial vehicles in cooperative search and attack in unknown environment is presented to improve the cooperative search and attack effectivenesses of multiple heterogeneous unmanned aerial vehicles. A coalition formation model is established based on the minimum target attack time and the minimum coalition scale with the constraints of required resources and simultaneous strike. A multistage sub-optimal coalition formation algorithm (MSOCFA) that has low computational complexity is proposed to solve the optimization problem of coalition formation. The performances of MSOCFA and partical swarm optimization algorithms are compared in terms of mission performance, complexity of algorithm and time taken to form the coalitions. In order to enable the multiple cooperative unmanned aerial vehicles to accomplish the search and attack missions autonomously, a distributed autonomous control strategy is proposed, which is based on the finite-state machine (FSM). The simulation results show the rationality, validity and high real-time performance of the proposed method for the coalition formation of multiple heterogeneous unmanned aerial vehicles in cooperative search and attack in the unknown environment. Monte Carlo method is employed to validate the impact of the number of unmanned aerial vehicles and targets on coalition formation. The reduced average mission completion time relates to the decreased number of targets and the increased the number of unmanned aerial vehicles.

Key words: control science and technology, multi-unmanned aerial vehicle, cooperative search and prosecute, coalition formation, finite-state machine, Monte Carlo method