The heterogeneous cluster of unmanned platforms can accomplish more complex tasks, and has better adaptability to harsh battlefield environments compared with a single or homogeneous platform. To solve the problem of how to enable the cluster to complete tasks in the rapidly changing battlefield environment efficiently, this paper designs and validates a heterogeneous unmanned cluster platform. The heterogeneous individuals in the cluster carry different initial resource amounts and require different amounts of resources to complete the same task. Specifically, compared to unmanned aerial vehicles (UAV), unmanned ground vehicles (UGV) often carry more resources and have a longer range. For homogeneous individuals, differences in relative position and velocity can also lead to different situations in the task. In addition, the contribution of different individuals in the execution of past tasks (individual reputation) should also be considered in the task allocation decision factor. Therefore, this paper comprehensively considers the initial position, speed, resource carrying capacity, individual reputation, and other factors of the cluster and establishes a task allocation model based on coalition game algorithm. Subsequently, the software and hardware platform built in this article is introduced. Finally, based on the simulated battlefield environment built, the proposed algorithm and the built heterogeneous unmanned cluster platform are experimentally verified. The verification results show that the platform can find the optimal task allocation method, and guide multiple heterogeneous unmanned platforms to complete task objectives by comprehensively considering the battlefield situation.