Modeling, simulation and forecasting of wind power plants using agent-based approach

Abstract

National economy and growth rely heavily on electricity but rapid urbanization, expeditious industrialization and increased domestic use due to population growth are among the reasons for the severe energy crisis in developing countries. The extended demand-supply gaps, depleting reservoirs of fossil fuel, and the environmental hazards altogether ignite the need for wider adoption of renewable energy resources for electricity generation. A functional assessment of the engineering design for this transition is a prerequisite before proceeding to on-ground implementation due to its high impact on system sustainability. To this end, we propose an agent-based modeling and simulation framework for the rapid prototyping of wind power plants. The proposed approach abstracts active components of wind power plants using agents and implements their dynamic behavior through agent interactions. The proposed model helps in composing different model components, design valuation, and forecasting energy generation in a cost-effective and productive manner. The proposed model is demonstrated by conceptualizing the design of the Foundation Wind Energy plant, located at Sindh, Pakistan, and the development of its agent-based model. The obtained short-term and long-term electricity generation profiles are validated with the actual data. We further compared the forecasts with the time series analysis performed on the actual data, using five different time-series forecasting models. The proposed simulation model and time series analysis model fit well on the actual data with a root mean square deviation of approximately 9 MW. The proposed framework will assist the policymakers in estimating the extent of electrical energy produced at given conditions using the wind potential available at the corridors of any country. It will further aid in the realistic analysis of the future dynamics of electricity demand and supply, hence help in effective energy planning.