Agent-based energy management system for remote community microgrid

Abstract

Rural communities are often unable to access electrical energy due to their distant location away from the national grid. Renewable energy sources (RESs) make it possible to provide electrical energy to these isolated areas. Sustainable generation is possible at a local level and is not dependant on connection to a national power grid. Microgrids are small scale, stand-alone electricity networks that harness energy at its geographical location, from natural resources. These small scale power grids are either connected to a national grid or operate separately by obtaining their power from an RES. Microgrids are becoming increasingly popular because they can provide electricity, independently of the national grid. The size of microgrid systems are dependent on the amount of energy that needs to be drawn and the amount of energy that has to be stored. Mechanical and electrical system component sizes become bigger due to increased operational energy requirements. Increases in component sizes are required on growing power networks when higher current levels are drawn. Energy management of microgrids must thus be introduced to prevent overloading the power grid network and to extend the operational life of the storage batteries. Energy management systems consist of different components which are seen as operational units. Operational units are responsible for measurement, communication, decision–making and power supply switching control, to manipulate the power output to meet the energy demands. Due to the increasing popularity of DC home appliances, it is important to explore the possibility of keeping these microgrids on a DC voltage basis. Electrical generation equipment such as photovoltaic panels can be used to generate DC at designed voltage levels. The energy management system connects the user loads and generation units together to form the microgrid. The aim of this study was to carry out the design of an agent–based energy management system for rural and under-developed communities. It investigates how the control of the output of the energy management system can be carried out to service the loads. The simulations were done using the following software packages: Simulink, Matlab, and SimPowerSystems. PV sources, energy management system (EMS) and user load parameters are varied in the simulation software to observe how the control algorithm executes load shedding. A stokvel-type charge share concept is dealt with where the state-of-charge (SOC) of batteries and user consumption will determine how grid loads are managed. Load shedding within the grid is executed by monitoring energy flow and calculating how much energy is allowed to be used by each consumer. The energy management system is programmed to always provide the largest amount of energy to the consumer with the lowest energy consumption for each day. The batteries store surplus electrical energy during the day. Load shedding starts at 18:00 each day. Users will be disconnected from the grid whenever their allotted energy capacity were depleted.