An ambient intelligence system for water release scheduling in cascade hydropower systems

Abstract

This study designed an ambient intelligence (AmI)-based decision support system that combined an electromagnetism-like mechanism (EM) and sensory data to aid human operators in making decisions regarding the management of cascade hydropower systems. Currently, cascade hydropower systems are the primary source of hydropower in Taiwan, and the Dajia Hydropower System is the largest one in Taiwan. The operational considerations of the various plants in this system include water levels constraints, minimal water release constraints, and maximal water release constraints. These considerations must be emphasized to plan the hourly water release rate and to maximize the power generation profitability of the overall system. Although extant operations are viable, an accurate calculation method for maximizing the economic efficiency of hydropower generation is yet to be developed. The AmI-based system proposed in this study can be used to determine the periodic water release at each plant in the system, thereby maximizing the company’s power-generation profit. In addition, the proposed system was compared with the scheduling system currently employed by the Dajia Hydropower System. The scheduling operations of the Dajia Hydropower System for 23 nonsummer days and a whole year were examined, respectively, and the hydropower plant’s original scheduling method was compared with that of the proposed system. Test results indicated that the scheduling plan obtained from the proposed system increased the profit of the Taiwan Power Company.