A universal numerical approach for nonlinear mathematic programming problems is presented with an application of ratios of first-order differentials/differences of objective functions to constraint functions with resp...A universal numerical approach for nonlinear mathematic programming problems is presented with an application of ratios of first-order differentials/differences of objective functions to constraint functions with respect to design variables. This approach can be efficiently used to solve continuous and, in particular, discrete programmings with arbitrary design variables and constraints. As a search method, this approach requires only computations of the functions and their partial derivatives or differences with respect to design variables, rather than any solution of mathematic equations. The present approach has been applied on many numerical examples as well as on some classical operational problems such as one-dimensional and two-dimensional knap-sack problems, one-dimensional and two-dimensional resource-distribution problems, problems of working reliability of composite systems and loading problems of machine, and more efficient and reliable solutions are obtained than traditional methods. The present approach can be used without limitation of modeling scales of the problem. Optimum solutions can be guaranteed as long as the objective function, constraint functions and their First-order derivatives/differences exist in the feasible domain or feasible set. There are no failures of convergence and instability when this approach is adopted.展开更多
This paper works on a modified simplex algorithm for the local optimization of Continuous Piece Wise Linear（CPWL） programming with generalization of hinging hyperplane objective and linear constraints. CPWL programm...This paper works on a modified simplex algorithm for the local optimization of Continuous Piece Wise Linear（CPWL） programming with generalization of hinging hyperplane objective and linear constraints. CPWL programming is popular since it can be equivalently transformed into difference of convex functions programming or concave optimization. Inspired by the concavity of the concave CPWL functions, we propose an Objective Variation Simplex Algorithm（OVSA）, which is able to find a local optimum in a reasonable time. Computational results are presented for further insights into the performance of the OVSA compared with two other algorithms on random test problems.展开更多
This paper deals with the continuous time Markov decision programming (briefly CTMDP) withunbounded reward rate.The economic criterion is the long-run average reward. To the models withcountable state space,and compac...This paper deals with the continuous time Markov decision programming (briefly CTMDP) withunbounded reward rate.The economic criterion is the long-run average reward. To the models withcountable state space,and compact metric action sets,we present a set of sufficient conditions to ensurethe existence of the stationary optimal policies.展开更多
In order to recover a signal from its compressive measurements, the compressed sensing theory seeks the sparsest signal that agrees with the measurements, which is actually an l;norm minimization problem. In this pape...In order to recover a signal from its compressive measurements, the compressed sensing theory seeks the sparsest signal that agrees with the measurements, which is actually an l;norm minimization problem. In this paper, we equivalently transform the l;norm minimization into a concave continuous piecewise linear programming,and propose an optimization algorithm based on a modified interior point method. Numerical experiments demonstrate that our algorithm improves the sufficient number of measurements, relaxes the restrictions of the sensing matrix to some extent, and performs robustly in the noisy scenarios.展开更多
文摘A universal numerical approach for nonlinear mathematic programming problems is presented with an application of ratios of first-order differentials/differences of objective functions to constraint functions with respect to design variables. This approach can be efficiently used to solve continuous and, in particular, discrete programmings with arbitrary design variables and constraints. As a search method, this approach requires only computations of the functions and their partial derivatives or differences with respect to design variables, rather than any solution of mathematic equations. The present approach has been applied on many numerical examples as well as on some classical operational problems such as one-dimensional and two-dimensional knap-sack problems, one-dimensional and two-dimensional resource-distribution problems, problems of working reliability of composite systems and loading problems of machine, and more efficient and reliable solutions are obtained than traditional methods. The present approach can be used without limitation of modeling scales of the problem. Optimum solutions can be guaranteed as long as the objective function, constraint functions and their First-order derivatives/differences exist in the feasible domain or feasible set. There are no failures of convergence and instability when this approach is adopted.
基金supported by the National Natural Science Foundation of China (Nos. 61473165 and 61134012)the National Key Basic Research and Development (973) Program of China (No. 2012CB720505)
文摘This paper works on a modified simplex algorithm for the local optimization of Continuous Piece Wise Linear（CPWL） programming with generalization of hinging hyperplane objective and linear constraints. CPWL programming is popular since it can be equivalently transformed into difference of convex functions programming or concave optimization. Inspired by the concavity of the concave CPWL functions, we propose an Objective Variation Simplex Algorithm（OVSA）, which is able to find a local optimum in a reasonable time. Computational results are presented for further insights into the performance of the OVSA compared with two other algorithms on random test problems.
基金This paper was prepared with the support of the National Youth Science Foundation
文摘This paper deals with the continuous time Markov decision programming (briefly CTMDP) withunbounded reward rate.The economic criterion is the long-run average reward. To the models withcountable state space,and compact metric action sets,we present a set of sufficient conditions to ensurethe existence of the stationary optimal policies.
基金supported by the National Natural Science Foundation of China(Nos.61473165 and 61134012)the National Key Basic Research and Development(973)Program of China(No.2012CB720505)
文摘In order to recover a signal from its compressive measurements, the compressed sensing theory seeks the sparsest signal that agrees with the measurements, which is actually an l;norm minimization problem. In this paper, we equivalently transform the l;norm minimization into a concave continuous piecewise linear programming,and propose an optimization algorithm based on a modified interior point method. Numerical experiments demonstrate that our algorithm improves the sufficient number of measurements, relaxes the restrictions of the sensing matrix to some extent, and performs robustly in the noisy scenarios.
