Abstract
Uncertainty should be taken into account when establishing multiobjective task assignment models for multiple unmanned combat aerial vehicles (UCAVs) due to errors in the target information acquired by sensors, implicit preferences of the commander for operational objectives, and partially known weights of sensors. In this paper, we extend the stochastic multicriteria acceptability analysis-2 (SMAA-2) method and combine it with integer linear programming to achieve multiobjective task assignment for multi-UCAV under multiple uncertainties. We first represent the uncertain target information as normal distribution interval numbers so that the values of criteria (operational objectives) concerned can be computed based on the weighted arithmetic averaging operator. Thus, we obtain multiple criteria value matrices for each UCAV. Then, we propose a novel aggregation method to generate the final criteria value matrix based on which the holistic acceptability indices are computed by the extended SMAA-2 method. On this basis, we convert the task assignment model with uncertain parameters into an integer linear programming model without uncertainty so as to implement task assignment using the integer linear programming method. Finally, we conduct a case study and demonstrate the feasibility of the proposed method in solving the multiobjective task assignment problem multi-UCAV under multiple uncertainties.
Highlights
Unmanned combat aerial vehicles (UCAVs), compared with airborne weapons and airborne platforms, have demonstrated advantages in tactical flexibility and multitasking ability due to having zero casualties, excellent stealth performance, high flight height, long endurance, multiple hard points for missiles, and a relatively low life-cycle cost [1]
The algorithms mentioned above—whether belonging to exact algorithms or heuristic algorithms—have demonstrated the ability to provide optimal or suboptimal solutions for task assignment problems of UAVs or UCAVs in various mission scenarios but may become difficult to apply when there are uncertain parameters related to mission scenarios
Uncertainties in the task assignment model for multi-UCAV exist in at least three aspects: (1) the target information obtained by various sensors, (2) the weights of operational objectives, and (3) the weights of these sensors
Summary
Unmanned combat aerial vehicles (UCAVs), compared with airborne weapons and airborne platforms (e.g., cruise missiles, manned fighters, and bombers), have demonstrated advantages in tactical flexibility and multitasking ability due to having zero casualties, excellent stealth performance, high flight height, long endurance, multiple hard points for missiles, and a relatively low life-cycle cost [1]. The algorithms mentioned above—whether belonging to exact algorithms or heuristic (metaheuristic) algorithms—have demonstrated the ability to provide optimal or suboptimal solutions for task assignment problems of UAVs or UCAVs in various mission scenarios but may become difficult to apply when there are uncertain parameters related to mission scenarios. Uncertainties in the task assignment model for multi-UCAV exist in at least three aspects: (1) the target information obtained by various sensors, (2) the weights of operational objectives, and (3) the weights of these sensors. In order to reflect the combat scenarios as accurately as possible, the above three aspects of uncertainties need to be taken into account when formulating the task assignment problem In this case, the problem would become extremely complicated and can neither be solved by exact algorithms nor by heuristic (metaheuristic) algorithms. The optimal or suboptimal task assignment scheme under severe uncertainties can be found
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
