A2G-QFL: Adaptive Aggregation with Two Gains in Quantum Federated learning