A dynamic soft constraint method for thermostatically controlled appliances scheduling

Abstract

Demand side attracts large attention in smart grid for energy efficiency and demand response, and thermostatically controlled appliances (TCAs) are the most potential resources. TCAs scheduling is a basic tool to utilize the flexibility of TCAs to achieve payment savings or peak reduction for users. Current research usually models the users' comfort a hard constraint in TCAs scheduling, which may fail to generate scheduling commands in abnormal situations, i.e. the hard constraint is deviated. To overcome this problem, a dynamic soft constraint method is proposed to enable the scheduling tool work in any situation. This proposed method considers correction process and consists of two techniques: a soft constraint and a dynamic comfort zone. In normal situations, the dynamic soft constraint is equivalent to the hard constraint, while in abnormal situations, it can still work and pull the temperature back to the comfort zone to limit the deviations. Therefore, the dynamic soft constraint method makes the TCAs scheduling more robust and adaptive to various situations. Both normal and abnormal situations are simulated to verify this new method.