In this paper, we proposes a distributed algorithm to seek the generalized Nash equilibria (GNE) problem. We consider aggregative games with affine coupling constraint. In the game, the agent can only interact with neighboring agents. The cost of each agent depends on some aggregate effect of the actions of all the other agents, without the need for multiple communication between agents. The convergence of the algorithm is carried out by the transformation of consensus subspace and the invariance of aggregative estimation.