Context. High resolution X-ray spectra of black hole X-ray binaries (BHBs) show blueshifted absorption lines suggesting the presence of outflowing winds. Furthermore, observations show that the disk winds are equatorial and they occur in the Softer (disk dominated) states of the outburst and are less prominent or absent in the Harder (power-law dominated) states. Aims. We want to test whether the self-similar magneto-hydrodynamic (MHD) accretion-ejection models can explain the observational results for accretion disk winds in BHBs. In our models, the density at the base of the outflow from the accretion disk is not a free parameter. This mass loading is determined by solving the full set of dynamical MHD equations without neglecting any physical term. Thus, the physical properties of the outflow depend on and are controlled by the global structure of the disk. Methods. We studied different MHD solutions characterized by different values of the disk aspect ratio (ε) and the ejection efficiency (p). We also generate two kinds of MHD solutions depending on the absence (cold solution) or presence (warm solution) of heating at the disk surface. Such heating could be either from dissipation of energy due to MHD turbulence in the disk or from illumination of the disk surface. Warm solutions can have large (>0.1) values of p, which would imply larger wind mass loading at the base of the outflow. We use each of these MHD solutions to predict the physical parameters (distance, density, velocity, magnetic field, etc.) of an outflow. Motivated by observational results, we have put limits on the ionization parameter (ξ), column density, and timescales. Further constraints were derived for the allowed values of ξ from thermodynamic instability considerations, particularly for the Hard SED. These physical constraints were imposed on each of these outflows to select regions within it, which are consistent with the observed winds. Results. The cold MHD solutions are found to be inadequate and cannot account for winds because of their low ejection efficiency. On the contrary, warm solutions can have sufficiently high values of p( 0.1), which are required to explain the observed physical quantities in the wind. From our thermodynamic equilibrium curve analysis for the outflowing gas, we find that in the Hard state a range of ξ is unstable. This constraint makes it impossible to have any wind at all in the Hard state. Conclusions. Using the MHD outflow models we are able to explain the observed trends, i.e. that the winds are equatorial and that they are observable in the Soft states (and not expected in the Hard state) of the BHB outbursts.