Given the scarcity of spectrum resources and the interference among multiple users in wireless communication, a multi-unmanned aerial vehicle (UAV) integrated sensing and communication (ISAC) system supported by non-orthogonal multiple access (NOMA) technology has been studied. On the constraints of communication quality and effective sensing power thresholds, the maximize the sum of the minimum throughput in all NOMA user groups by jointly optimizing user grouping association, sensing target area association allocation, power allocation, horizontal trajectories of multiple UAVs, and vertical trajectories of multiple UAVs. A more efficient alternating optimization algorithm is adopted to solve the resultant non-convex integer optimization problem with highly coupled variables. First, an improved clustering algorithm obtains a feasible set of user grouping associations for the optimization problem. Under a certain user grouping association scheme, four sub-problems: the association allocation of sensing target areas, power allocation, horizontal trajectories of multiple UAVs, and vertical trajectories of multiple UAVs, are optimized to obtain the optimization results. Simulation results show that compared with several benchmark schemes, the proposed optimization algorithm can exhibit better performance in terms of common throughput.