This paper reports on the advantages of introducing a novel adaptive congestion control algorithm on physical networks for multimedia communication at virtual channels and virtual paths' levels. The adaptive control begins when congestion appears on the network. Knowledge representation for this algorithm based on the Petri net theory is proposed. The control strategy is based on the information of congestion that is monitored and revised according to congestion status at any time. As a result, we can obtain the prospective trends in intelligent flow control to avoid the congestion of multimedia information on the network.