Off-campus South Dakota State University users: To download campus access theses, please use the following link to log into our proxy server with your South Dakota State University ID and password.

Non-South Dakota State University users: Please talk to your librarian about requesting this thesis through interlibrary loan.

Document Type

Thesis - University Access Only

Award Date


Degree Name

Master of Science (MS)

Department / School

Electrical Engineering and Computer Science

First Advisor

Reinaldo Tonkoski


Solar energy is a good alternative in remote microgrids because it is readily available in remote area and also reduces pollution due to conventional sources of energy. However, due to the presence of clouds, the solar energy available is intermittent in nature and therefore, it is not completely reliable. The variation of solar energy is compensated using conventional sources such as diesel generator. The operation of diesel generator is not economical when it needs to operate to compensate the variation. Therefore, forecasting of renewable energy of day ahead helps to schedule the generation economically. Different forecast technologies currently available utilize different weather information and satellite information for day ahead forecasting which is not readily available in remote areas. Therefore, an easy to implement forecasting method for day ahead forecasting in remote areas for remote microgrids is required. The objective of this thesis was to develop a forecasting method for utilization in remote microgrid. The Markov Switching Model considers the past solar irradiance data available, clear sky irradiance and Fourier basis functions to create linear models which consider three regimes for a day. Those three regimes are high energy regime when there is no cloud interrupting the sun, medium energy regime when the cloud is medium and low energy regime when the sky is highly cloudy. The transition from one energy regime to another regime is dependent upon the probability transition matrix which shows high probability xiv of the solar irradiance remaining in the same energy level. Therefore, considering the beginning four hours of sunlight, the solar irradiance for the day ahead was determined. This forecasted value of solar irradiance was utilized for energy management in the remote microgrid. The cost of operation of energy management for a year using forecasted solar data was compared against a baseline energy management. The cost of operation of energy management using forecasted solar irradiance was found to be less than the energy management without solar irradiance forecasting. This shows that the irradiance forecasting can be done using freely available resources using the Markov Switching Model. And forecasted irradiance could decrease the cost of operation of the microgrids by economical operation of the generation units.

Library of Congress Subject Headings

Solar radiation--Forecasting
Distributed generation of electric power
Renewable energy sources
Markov processes
Microgrids (Smart power grids)


Includes bibliographical references (pages 95-97)



Number of Pages



South Dakota State University



Rights Statement

In Copyright