Using full factorial experiment to simulate electricity generation by solar batteries

Abstract

The amount of electrical energy generated by solar batteries (SB) depends significantly on their position relative to the Sun. The presence of dependencies of the initial parameter of the SB on the factors determining this position allows finding the optimal orientation of fixed or partially fixed SBs. This paper presents the results of the study on the conditions of using a full factorial experiment for creating mathematical models — regression equations — to simulate the operation of solar batteries under the condition of reducing the amount of field tests necessary for this. It is shown that the simulation accuracy significantly depends on the range of factor values. The conducted study demonstrates that the use of a full factorial experiment in the range of factor values approaching all possible is appropriate for preliminary simulation of the solar battery operation and an approximate assessment of the simulation results in connection with significant errors. Narrowing the range of factor values significantly reduces the deviations of simulation results from experimental data, which in many cases can be accepted, especially if the simulation results are used for specific operating conditions of the solar cell. In order to create mathematical models that would more accurately approach the actual ones obtained experimentally, it is necessary to use dependencies for concretized values of factors.