We implement and test a globally convergent sequential approximate optimization algorithm based on (convexified) diagonal quadratic approximations. The algorithm resides in the class of globally convergent optimizatio...
详细信息
We implement and test a globally convergent sequential approximate optimization algorithm based on (convexified) diagonal quadratic approximations. The algorithm resides in the class of globally convergent optimization methods based on conservative convex separable approximations developed by Svanberg. At the start of each outer iteration, the initial curvatures of the diagonal quadratic approximations are estimated using historic objective and/or constraint function value information, or by building the diagonal quadratic approximation to the reciprocal approximation at the current iterate. During inner iterations, these curvatures are increased if no feasible descent step can be made. Although this conditional enforcement of conservatism on the subproblems is a relaxation of the strict conservatism enforced by Svanberg, global convergence is still inherited from the conservative convex separable approximations framework developed by Svanberg. A numerical comparison with the globally convergent version of the method of moving asymptotes and the nonconservative variants of both our algorithm and method of moving asymptotes is made.
Sequential quadratic programming (SQP) methods have been extensively studied to handle nonlinear programming problems. In this paper, a new SQP approach is employed to tackle nonlinear complementarity problems (NCPs)....
详细信息
Sequential quadratic programming (SQP) methods have been extensively studied to handle nonlinear programming problems. In this paper, a new SQP approach is employed to tackle nonlinear complementarity problems (NCPs). At each iterate, NCP conditions are divided into two parts. The inequalities and equations in NCP conditions, which are violated in the current iterate, are treated as the objective function, and the others act as constraints, which avoids finding a feasible initial point and feasible iterate points. NCP conditions are consequently transformed into a feasible nonlinear programming subproblem at each step. New SQP techniques are therefore successful in handling NCPs.
A parallel, filter-based, sequential quadratic programming algorithm is implemented and tested for typical general-purpose engineering applications. Constrained engineering test problems, including a finite element si...
详细信息
A parallel, filter-based, sequential quadratic programming algorithm is implemented and tested for typical general-purpose engineering applications. Constrained engineering test problems, including a finite element simulation, with up to 512 design variables are considered. The accuracy and serial performance of the filter-based algorithm are compared against that of a standard sequential quadratic programming algorithm. The parallel performance of the algorithm is evaluated, using up to 52 cores on a Linux Cluster. The results indicate that the filter-based algorithm competes favorably with a standard sequential quadratic programming algorithm in a serial environment. However, the filter-based algorithm exhibits much better parallel efficiency due to the lack of a one-dimensional search.
This paper proposes a procedure to construct the membership functions of the performance measures in bulk service queuing systems with the arrival rate and service rate are fuzzy numbers. The basic idea is to transfor...
详细信息
This paper proposes a procedure to construct the membership functions of the performance measures in bulk service queuing systems with the arrival rate and service rate are fuzzy numbers. The basic idea is to transform a fuzzy queue with bulk service to a family of conventional crisp queues with bulk service by applying the alpha-cut approach. On the basis of alpha-cut representation and the extension principle, a pair of parametric nonlinear programs is formulated to describe that family of crisp bulk service queues, via which the membership functions of the performance measures are derived. To demonstrate the validity of the proposed procedure, two fuzzy queues often encountered in transportation management are exemplified. Since the performance measures are expressed by membership functions rather than by crisp values, they completely conserve the fuzziness of input information when some data of bulk-service queuing systems are ambiguous. Thus the proposed approach for vague systems can represent the system more accurately, and more information is provided for designing queuing systems in real life. By extending to fuzzy environment, the bulk service queuing models would have wider applications. (C) 2004 Elsevier B.V. All rights reserved.
Non-conventional principles of locomotion for mobile robotic systems are considered, namely, snake-like locomotion of multilink systems and control of motion by means of moving internal masses. These motions are possi...
详细信息
Non-conventional principles of locomotion for mobile robotic systems are considered, namely, snake-like locomotion of multilink systems and control of motion by means of moving internal masses. These motions are possible in the presence of external resistance forces such as Coulomb's dry friction or viscous resistance. Optimal periodic motions are obtained that correspond to the maximal average velocity of the system as a whole. Evaluation of the maximal possible velocity is important for the overall estimation of operational properties of the robotic systems under consideration.
Abstract For indirect-drive robot manipulators, it is necessary to experimentally tune the controller parameters to compensate for the uncertainties in the actual system and to obtain the desired performance. Since ma...
详细信息
Abstract For indirect-drive robot manipulators, it is necessary to experimentally tune the controller parameters to compensate for the uncertainties in the actual system and to obtain the desired performance. Since manual tuning of controllers is very time-consuming, much effort has been invested in developing systematic tuning methods. Also, in standard industrial applications of robot manipulators, the performance is evaluated based on load-side information, while the feedback control loop is closed on the motor side. Therefore, the load-side information needs to be considered during controller tuning process. Iterative controller tuning is a method that tunes controllers in a repetitive process using data collected in experiments. In this paper, a model-free method that automatically optimizes an arbitrary multi-variable cost function is used. In the proposed method, the variables are updated while the gradient of the cost function is continuously estimated by a perturbation in real-time. The effectiveness of the controller tuning method is demonstrated by experimental results.
The development of chance constrained optimization model for design of portal frames is presented in this paper. The developed model assumes that the ranges of variations of the input and output parameters are specifi...
详细信息
The development of chance constrained optimization model for design of portal frames is presented in this paper. The developed model assumes that the ranges of variations of the input and output parameters are specified, all variables/parameters are independent random variables and follow normal distribution. To ensure that the solutions of the developed model obey a desired structural behaviour, the slope deflection model of structural analysis in probabilistic form is used in the model. The chance constraints ensure that the probability of realizing the constraints in a random environment is greater than or equal to a specified probability. The optimization model simultaneously seeks to minimize the chances of failure and the total weight of the portal frame. The bi-objective model is demonstrated for optimal design of portal frames with built-up I-section. The performance evaluation establishes the potential of the developed model for future use.
In the present work, the problem associated to the firm energy evaluation is treated as a non linear optimization model, which allows the representation of the productivity variation of the hydro plants. The proposed ...
详细信息
In the present work, the problem associated to the firm energy evaluation is treated as a non linear optimization model, which allows the representation of the productivity variation of the hydro plants. The proposed model takes into account the individualized representation of the plants and the historical series of flows since the month of January of 1931. The proposed optimization problem will be solved using the Primal-Dual Interior Point Method. A case study will be presented including the Brazilian Interconnected National System. The results obtained show that the proposed methodology is promising, since it presents an energy market value more realistic when compared with existing methodologies.
This research proposes the change of damping coefficient regarding minimum displacement. From the mass with external forced and damper problem, when is the constant external forced transmitted to the understructure in...
详细信息
This research proposes the change of damping coefficient regarding minimum displacement. From the mass with external forced and damper problem, when is the constant external forced transmitted to the understructure in the difference angle between 30 and 60 degrees. This force generates the vibration as general known;however, the objective of this problem is to have minimum displacement. As the angle is changed and the goal is the same;therefore, the damper of the system must be varied while keeping constant spring stiffness. The problem is solved by using nonlinear programming and the suitable changing of the damping coefficient is provided.
The paper describes the design and implementation of BNB-Solver, an object-oriented framework for discrete and continuous parallel global optimization. The framework supports exact branch-and-bound algorithms, heurist...
详细信息
The paper describes the design and implementation of BNB-Solver, an object-oriented framework for discrete and continuous parallel global optimization. The framework supports exact branch-and-bound algorithms, heuristic methods and hybrid approaches. BNB-Solver provides a support for distributed and shared memory architectures. The implementation for distributed memory machines is based on MPI and thus can run on almost any computational cluster. In order to take advantages of multicore processors we provide a separate multi-threaded implementation for shared memory platforms. We introduce a novel collaborative scheme for combining exact and heuristic search methods that provides the support for sophisticated parallel heuristics and convenient balancing between exact and heuristic methods. In the experimental results section we discuss a nonlinear programming solver and a highly efficient knapsack solver that significantly outperforms existing parallel implementations.
暂无评论