You can fit the ID-VG curve with a field effect formula which include a fitting parameter n0. It is also well seen off plotted in semi-log ID=f(log(VG-VD)) . A good example is shown in paper https://arxiv.org/pdf/0707.1807.pdf
My doubt is that, if it is residual charge density then we should be able to calculate it directly from the value of Dirac voltage without any curve fitting.
Beware of some possibility of confusion here. There are two different quantities : the Dirac voltage give access to the doping level of the Graphene transistor (which you can express in charge carrier density) but this is different from the residual charges at charge neutrality (corresponding to average depth of electron/hole puddles) which basically is given by the sharpness of the ID-VG curve. the graphene resistance R can be fitted by the formula :
R=L/w*1/(mu*sqrt(n0*e+C*|VG-VD|)+Rc
L/W aspect ratio of the transistor
mu mobility
C gate capacitance
in that formula VD and n0 are indeed two independent parameters.