Optimization of Groundwater Supply under Consideration of Capacity Expansion in Unconfined Aquifer
Hsiao Chung Chin
Liang Cheng Chang
|關鍵字:||微分動態規劃;遺傳演算法;非拘限含水層;Constrained Differential Dynamic Programming;Genetic Algorithm;Unconfined Aquifer|
Generally speaking, the demand of water resources is gradually increasing according to the increase of population and economic development. Therefore, a cost-effective investment strategy is to expand the capacity of a water resources project with the augment of water demand. The system capacity of a groundwater supply system depends on the total number of wells and the capacity of each well. Since each well is installed independently, the total capacity of the system is convenient to expand according the increase of demand. However, since the problem of capacity expansion is a discrete, nonlinear and transient optimization problem,no optimization algorithms have been developed to resolve the problem. This study utilizes dynamic optimal control and Genetic Algorithms (GAs) to solve the groundwater management problem considering the system capacity expansion. The total cost of the problem including fixed and operating costs, and both are computed under the consideration interest rates. In GAs, one chromosome represents a possible network design which consists the well number and the location and developing schedule for each production well. The fixed cost can then be computed according to the network design.The optimal operating cost is evaluated by using the dynamic optimal control algorithm in which the pumping rates are the decision variables. According to this study, the consideration of system capacity expansion have significant impact on the fixed cost and the network design. This work also demonstrates that fixed costs of drilling well may significantly influence a design of groundwater production network. Therefore, the strategy of capacity expansion and fixed costs should be explicitly incorporated into a groundwater management model.
|Appears in Collections:||Thesis|