Impact of measurement accuracy in a multi-agent control system for active distribution networks

The control of active demand, electric vehicles, distributed generation and storage devices is the key to offer services that increase security and quality of supply of power systems, improve energy efficiency and reduce the cost for energy. Decentralised control systems have the merit to allow the operation of many small customers with a reduced information flow by exploiting local information gathered from the field with intelligent meters. Smart metering and ICT are the enabling technologies to offer market opportunities to the consumers, and to provide the Distribution System Operator (DSO) information regarding the behaviour of the prosumers. In this paper, an accurate Multi-Agent System (MAS) for the direct control of active demand in the distribution networks has been carried out. In particular, different levels of knowledge regarding the load behaviour have been considered. The impact of such knowledge in the optimal pattern of load demand that minimizes costs and improves the Power Quality of the distribution system has been evaluated. The discussed results prove the utility of the proposed analysis for the optimal design of MAS control.