Vehicular Ad Hoc network（ VANET） will lose data packet when suffered from the black hole attack

which may affect the normal network communications. Put forward a selection method of feature vector

aiming at the black hole attack

and establish a fleet simulation model under the normal condition and the black hole attack

and use the method of support vector machine（ SVM） to detect the intrusion of black hole of the VANET.Simulation results show that for the nodes which are not source nodes

destination nodes and attacking nodes

using this detection method can maintain high detection rate and low false detecting rate

and can achieve a good effect in judging whether the VANET suffered from the black hole attack based on communication data from single node.