The optimization of both power consumption and balance of Smart Grid is required for online power control considering the use of alternative energy sources. A concept of “generating consumers” has been introduced for active consumers, leading to flexibly regulating the power flows, equalizing the load schedule, and minimizing the financial costs of consumed power. This study aimed to develop a new mathematical model of optimal both power consumption and balance considering the participation of “generating consumer” in the form of wind resources, which presents the intelligent system with a bidirectional power flow. The approach proposes a priority system to choose the optimal generation source, minimizing its material and financial costs based on the particle swarm optimization algorithm and holon concept. A high potential of own wind sources has been examined in a coastal zone of the Far East and on the Russky and Popov Islands. As result, Russky Island doesn't receive any additional income (max −433 EUR) since its power consumption cannot be covered even with bidirectional power flows while the Popov Island gains income (max 468 EUR) due to existing power excess. The efficiency of the developed software has been validated by using the gradient descent algorithm.