Flexible goal-directed behavior requires monitor-control networks to detect the need for behavioral adjustments and to implement the required regulations. Among event-related brain potentials related to the function of such networks is the feedback-related negativity (FRN), which is detected in trial-and-error learning tasks. Conflict monitoring theory (CMT) as one of the influential theories of such networks cannot describe the FRN. Recently, we have proposed a cost-conflict monitoring system that extends the CMT. The cost-conflict monitoring holds that the monitoring system can detect conflict signal, but the conflict is over the costs of alternative outcomes of the selected action rather than the response conflict as proposed by the CMT. In the cost-conflict monitoring, cost functions are computed based on waiting times from the response to feedback delivery and from these quantities a conflict signal is derived. Here, we present a computational realization of such cost-conflict monitor-controller network. We utilize this computational model to simulate existing human performance and ERP data of a trial-and-error learning task. The model successfully simulated the behavioral data and FRN signals under different conditions in this task.