Mohamad Bostan - Academia.edu (original) (raw)
Uploads
Papers by Mohamad Bostan
Engineering Optimization, Apr 24, 2014
This article proposes a hybrid linear programming (LP-LP) methodology for the simultaneous optima... more This article proposes a hybrid linear programming (LP-LP) methodology for the simultaneous optimal design and operation of groundwater utilization systems. The proposed model is an extension of an earlier LP-LP model proposed by the authors for the optimal operation of a set of existing wells. The proposed model can be used to optimally determine the number, configuration and pumping rates of the operational wells out of potential wells with fixed locations to minimize the total cost of utilizing a two-dimensional confined aquifer under steady-state flow conditions. The model is able to take into account the well installation, piping and pump installation costs in addition to the operational costs, including the cost of energy and maintenance. The solution to the problem is defined by well locations and their pumping rates, minimizing the total cost while satisfying a downstream demand, lower/upper bound on the pumping rates, and lower/upper bound on the water level drawdown at the wells. A discretized version of the differential equation governing the flow is first embedded into the model formulation as a set of additional constraints. The resulting mixed-integer highly constrained nonlinear optimization problem is then decomposed into two subproblems with different sets of decision variables, one with a piezometric head and the other with the operational well locations and the corresponding pumping rates. The binary variables representing the well locations are approximated by a continuous variable leading to two LP subproblems. Having started with a random value for all decision variables, the two subproblems are solved iteratively until convergence is achieved. The performance and ability of the proposed method are tested against a hypothetical problem from the literature and the results are presented and compared with those obtained using a mixed-integer nonlinear programming method. The results show the efficiency and effectiveness of the proposed method for solving practical groundwater management problems.
Engineering Optimization, 2014
This paper proposes a LP-LP methodology for simultaneous optimal design and operation of the grou... more This paper proposes a LP-LP methodology for simultaneous optimal design and operation of the groundwater utilization systems. The proposed model is an extension of an earlier LP-LP model proposed by authors for the optimal operation of a set of existing wells. The proposed model can be used to optimally determine the number, configuration and pumping rates of the operational wells out of potential wells with fixed locations to minimize the total cost of utilizing a two-dimensional (2D) confined aquifer under steady-state flow condition. The model is able to take into account the well installation cost, piping cost,and pump installation costs in addition to the operational cost including the cost of energy and maintenance. The solution of the problem is defined by well locations and their pumping rates, minimizing the total cost while satisfying a downstream demand, lower/upper bound on the pumping rates, and lower/upper bound on the water level drawdown at the wells. Discretized version of the differential equation governing the flow is first embeded into the model formulation as a set of additional constraints. The resulting mixed-integer highly constrained nonlinear optimization problem is then decomposed into two sub-problems with different set of decision variables, one with piezometric head and the other with the operational well locations and the corresponding pumping rates. The binary variables representing the well locations are approximated by a continuous variable leading to two LP sub-problems. Having started with a random value for all decision variables, the two sub-problems are solved iteratively until convergence is achieved. The performance and ability of the proposed methodis tested against a hypothetical problem from the literature and the results are 2 presented and compared with those obtained using Mixed Integer Non-Linear Programming (MINLP) method. The results show the efficiency and effectiveness of the proposed method for solving practical groundwater management problems.
Engineering Optimization, Apr 24, 2014
This article proposes a hybrid linear programming (LP-LP) methodology for the simultaneous optima... more This article proposes a hybrid linear programming (LP-LP) methodology for the simultaneous optimal design and operation of groundwater utilization systems. The proposed model is an extension of an earlier LP-LP model proposed by the authors for the optimal operation of a set of existing wells. The proposed model can be used to optimally determine the number, configuration and pumping rates of the operational wells out of potential wells with fixed locations to minimize the total cost of utilizing a two-dimensional confined aquifer under steady-state flow conditions. The model is able to take into account the well installation, piping and pump installation costs in addition to the operational costs, including the cost of energy and maintenance. The solution to the problem is defined by well locations and their pumping rates, minimizing the total cost while satisfying a downstream demand, lower/upper bound on the pumping rates, and lower/upper bound on the water level drawdown at the wells. A discretized version of the differential equation governing the flow is first embedded into the model formulation as a set of additional constraints. The resulting mixed-integer highly constrained nonlinear optimization problem is then decomposed into two subproblems with different sets of decision variables, one with a piezometric head and the other with the operational well locations and the corresponding pumping rates. The binary variables representing the well locations are approximated by a continuous variable leading to two LP subproblems. Having started with a random value for all decision variables, the two subproblems are solved iteratively until convergence is achieved. The performance and ability of the proposed method are tested against a hypothetical problem from the literature and the results are presented and compared with those obtained using a mixed-integer nonlinear programming method. The results show the efficiency and effectiveness of the proposed method for solving practical groundwater management problems.
Engineering Optimization, 2014
This paper proposes a LP-LP methodology for simultaneous optimal design and operation of the grou... more This paper proposes a LP-LP methodology for simultaneous optimal design and operation of the groundwater utilization systems. The proposed model is an extension of an earlier LP-LP model proposed by authors for the optimal operation of a set of existing wells. The proposed model can be used to optimally determine the number, configuration and pumping rates of the operational wells out of potential wells with fixed locations to minimize the total cost of utilizing a two-dimensional (2D) confined aquifer under steady-state flow condition. The model is able to take into account the well installation cost, piping cost,and pump installation costs in addition to the operational cost including the cost of energy and maintenance. The solution of the problem is defined by well locations and their pumping rates, minimizing the total cost while satisfying a downstream demand, lower/upper bound on the pumping rates, and lower/upper bound on the water level drawdown at the wells. Discretized version of the differential equation governing the flow is first embeded into the model formulation as a set of additional constraints. The resulting mixed-integer highly constrained nonlinear optimization problem is then decomposed into two sub-problems with different set of decision variables, one with piezometric head and the other with the operational well locations and the corresponding pumping rates. The binary variables representing the well locations are approximated by a continuous variable leading to two LP sub-problems. Having started with a random value for all decision variables, the two sub-problems are solved iteratively until convergence is achieved. The performance and ability of the proposed methodis tested against a hypothetical problem from the literature and the results are 2 presented and compared with those obtained using Mixed Integer Non-Linear Programming (MINLP) method. The results show the efficiency and effectiveness of the proposed method for solving practical groundwater management problems.