Decision support model for optimal reservoir operation using association rule mining

Abstract

Reservoir operation has often been treated as a multistage stochastic control problem offering substantial increase in benefits even for relatively small improvements in operating efficiency Inspite of the development of new techniques implicit stochastic optimization models solved using linear programming remain one of the most readily applicable techniques for complex reservoir systems In ISO models finding dependencies from data were based on statistical analysis such as correlation analysis and multivariate regression Application of these techniques however finds a limitation due to the necessity to define explicitly the mathematical form of the link between independent and dependent variables For highly nonlinear problems and for problems with high uncertainty this method may not yield satisfactory results This places the focus on new generation techniques and tools of Machine Learning emerging to intelligently newlineassist humans in analysing data finding useful knowledge and in some cases performing analysis automatically Also the detailed literature reveals that specific system level operating policy improves the system performance which can be implemented easily using latest methodologies newline