Power curves prediction using empirical data regression on small scale compressed air energy storage

Abstract

The key to optimizing the system is to know the operating point of the system at the time of loading, or it is known as the power curve. However, to identify the power curve, the existing method is to model the mathematical of the system. Therefore some component characteristics need to be known and need additional observations if the component variable is unknown, and it becomes a long identification process. So, in this exploratory research will be presented the way to find out the power curve of a system without modeling mathematical of the system, but by using the polynomial regression technique. This regression technique form is using the empirical data of the power curve form parameter on SS-CAES prototype. The method is based on five approach model in which is the variation of loading sampling data to be used with the purpose is to find the best sampling of prediction. The data will be analyzed in the form of statistical parameters and the graph to show the evaluation process of this technique. From the results of the regression can be concluded that the power curve of SS-CAES can be identified with a high correlation value of 0.997 (99,745% accuracy) and the best way to take samples of data to be used in this technique is presented in the paper.

MIMO Fuzzy to Improve Search Speed and Reduce Oscillation in Maximum Efficiency Point Tracking of Small-Scale Compressed Air Energy Storage

“…In CAES-related research, both large and small scales, several development topics have been published including thermodynamic issues [2,14,15], modeling [16][17][18], economic analysis [19,20], and the most popular issue is system optimization in the discharged phase [10,18,[21][22][23].…”

“…Several attempts were made to improve process efficiency in this discharged phase, either by combining SS-CAES with other energy storage media to develop soft computing methods to operate more optimally [10,18,[21][22][23][24][25][26][27].…”