Simulation of Large Customer Price Response Under Time-of-Use Electricity Pricing Based on Multi-Agent System

The mechanism of customer TOU response is the key for success of time-of-use pricing. However different customers have different price response behavior and no simple analytic model is possible for it. Therefore multi-agents simulation is introduced. In simulation research TOU price response is taken as a decentralized decision process of all related customers, while power supply corporation can dynamically adjust its policy according to customer response. The paper discusses in details with the design principle and framework of the simulation system. In the system customer agent is implemented as responsive agent with its response rules built on fuzzy logic, utility agent is implemented as learning agent with classifier learning algorithm as its learning rules, and environment management module manages the interaction among them. Then the simulation flow is given and finally a case study based on a real power supply corporation is given and simulation results reported.