In this talk, we present a framework for the coordination of distributed energy resources (DERs) integrated into a lossy power distribution system, whose model is unknown. The DER group of DERs collectively provides active power to a bulk power grid. The proposed framework uses an input-output (IO) system model to represent the relation between the DER active power injections—the inputs—and the total resource provision—the output. Associated with the model is an estimator that estimates the parameters of the IO model and a controller that determines the DER active power injections. We formulate the estimation and the control problems as quadratic programs with box constraints and solve using the gradient projection algorithm. In order to obtain richer information about the system, random perturbations are introduced in the control algorithm. We show the estimated solutions and control schemes converge under certain conditions, i.e., the estimated parameters converge to the true parameter values and the sum of the DER outputs converge to the requested amount. We then present simulation results to illustrate the approach and discuss the effectiveness of the proposed framework.