Journal of Industrial & Management Optimization
September 2020 , Volume 16 , Issue 5
Select all articles
The topic of investment timing in multi-stage public-private partnership (PPP) projects has not been received much attention so far. This study investigates optimal expansion timing decisions in multi-stage PPP projects under an uncertain demand and where the first-stage greenfield project involving a dedicated asset is immediately implemented as the PPP contract is closed, whereas the timing of the later expansion is flexibly decided according to the demand. In this setting, the optimal multiple stopping timing theory is adopted to model the expansion framework. Furthermore, we integrate a government subsidy, including an investment subsidy and revenue subsidy, into the expansion timing decisions. Through a hypothetical three-stage investment plan for a sanitary sewerage project, the optimal expansion strategies and the value of the multi-stage project before and after the subsidy are provided using a least squares Monte Carlo simulation. Also, the influences of a dedicated asset on the expansion strategies and project value are illustrated. In addition, we compare the incremental value before and after the subsidy and earlier expansion derived from two types of subsidies. The comparisons show that there is more incremental value for the revenue subsidy, and that the investment subsidy brings an earlier expansion.
In this paper, we propose a novel low-dimensional semidefinite programming (SDP) relaxation for convex quadratically constrained nonconvex quadratic programming problems. This new relaxation is derived by simultaneous matrix diagonalization method under the difference of convex decomposition scheme. The highlight of the relaxation is the low dimensionality of the positive semidefinite constraint, which only depends on the number of negative eigenvalues in the objective function. We prove that a mixed SOCP and SDP relaxation appeared in the literature is equivalent to the proposed relaxation, while the proposed relaxation has fewer constraints. We also provide conditions under which the proposed relaxation is as tight as the classical SDP relaxation and provides an optimal value for the original problem. For general cases, a spatial branch-and-bound algorithm is designed for finding a global optimal solution. Extensive numerical experiments support that the proposed algorithm outperforms two cutting-edge algorithms when the problem size is medium or large and the number of negative eigenvalues in the nonconvex objective function is relatively small.
Microwave heating has been widely used in various fields during recent years. However, it also has a common problem of uneven heating. In this paper, optimal frequency control problem for microwave heating process is considered. The cost function is defined such that the temperature profile at the final stage has a relative uniform distribution in the field. The controlled system is a coupled by Maxwell equations with nonlinear heating equation. The existence of a weak solution for coupled system is proved. The weak continuity of the solution operator is also shown. Moreover, the existence of a global minimizer of the optimal frequency control problems is proved.
In the past one decade, an increasing number of motor vehicles necessarily results in huge amounts of end-of-life vehicles (ELVs) in the future. From the view point of environment protection and resource utilization, government subsidy and public awareness of environmental protection play a critical role in promoting the formal recycle enterprises to recycle the ELVs as many as possible. Different from the existing similar models, a mixed integer nonlinear optimization model is established in this paper to formulate the management problems of recycling ELVs as a centralized decision-making system, where damaged and aging degrees, correlation between the recycled quantity and take-back price of ELVs, and the public environmental protection awareness are considered. Unlike the results available in the literature, take-back prices of the ELVs are the endogenous variables of the model (decision variables), which affect the collected quantity of ELVs and the profit of recycling system. Additionally, due to distinct damaged and aging degrees of the ELVs, the refurbished or dismantled amounts of ELVs are also regarded as the decision variables so that the recycle system is more applicable. By case study and sensitivity analysis, validity of the model is verified and impacts of the governmental subsidy and environmental awareness are analyzed. By the proposed model, it is revealed that: (1) Distinct treatment of ELVs with different damaged and aging degrees can increase the profit of recycling ELVs; (2) Compared with the transportation cost, higher processing cost is a main obstacle to the profit growth. Advanced processing technology plays the most important role in improving the ELV recovery efficiency. (3) Both of government subsidy and environmental awareness seriously affect decision-making of recycle enterprises.
In this paper we assume the insurance wealth process is driven by the compound Poisson process. The discounting factor is modelled as a geometric Brownian motion at first and then as an exponential function of an integrated Ornstein-Uhlenbeck process. The objective is to maximize the cumulated value of expected discounted dividends up to the time of ruin. We give an explicit expression of the value function and the optimal strategy in the case of interest rate following a geometric Brownian motion. For the case of the Vasicek model, we explore some properties of the value function. Since we can not find an explicit expression for the value function in the second case, we prove that the value function is the viscosity solution of the corresponding HJB equation.
This paper investigates the impact of competition and the strategic inventories on the performance of a supply chain comprising two competing suppliers and one retailer. Existing literature has shown that the retailer's optimal strategy in equilibrium is to carry inventories, and the suppliers are unable to prevent this. In contrast, our results show that the suppliers will prevent the retailer from carrying strategic inventories when the degree of competition between suppliers is high, and the retailer's carrying strategic inventory is not necessary to force suppliers to lower the future wholesale price. We also find the substitutable relationship between the effect of strategic inventories and the effect of competition. When the degree of competition increases, the suppliers are worse off but the retailer and the total supply chain are both better off when carrying strategic inventories. The retailer could introduce profit sharing contracts so as to encourage suppliers to support strategic inventories which enhance the entire performance of the supply chain.
We develop a stochastic model of contagion with two individual types by extending an archetypal SIS CTMC model. Our results include the analyses of the contagion duration and the number of individual afflictions. Numerical applications with the minority and majority types are provided to consider various contagions.
This paper studies an optimal investment-reinsurance problem for an insurance company which is subject to a dynamic Value-at-Risk (VaR) constraint in a Markovian regime-switching environment. Our goal is to minimize its ruin probability and control its market risk simultaneously. We formulate the problem as an infinite horizontal stochastic control problem with the constrained strategies. The dynamic programming technique is applied to derive the coupled Hamilton-Jacobi-Bellman (HJB) equations and the Lagrange multiplier method is used to tackle the dynamic VaR constraint. Furthermore, we propose an efficient numerical method to solve those HJB equations. Finally, we employ a practical example from the Korean market to verify the numerical method and analyze the optimal strategies under different VaR constraints.
A physical-based numerical algorithm using Kirchhoff circuit is detailed for modelling the free vibration of moderate thick symmetrically laminated plates based on the first order shear deformation theory (FSDT). With the help of multidimensional passivity of analog circuit and nonlinear optimization solvers, the philosophy gives rise to a nonlinear programming (NLP) model that can apply further to explore stability characteristics and optimum performance of the resultant multidimensional wave digital filtering network representing the FSDT plate. Various optimization methods exploiting gradient-based and direct search methods are adopted with efficient broad search power to tackle the NLP model. As a result, the necessary Courant-Friedrichs-Levy stability criterion can be fully satisfied at all time with least restriction on the spatially discretized geometry of the scattering problem. With full stability guaranteed, the waveform is analyzed by the power cepstrum for spectra peaks detection, which has led to more accurate estimate of various vibration effects in predicting nature frequencies with different fiber orientations, stacking sequences, stiffness ratios and boundary conditions. These results have shown in excellent agreement with early published works based on the finite element solutions of the high-order shear deformation theory and other well known numerical techniques.
This paper looks at a stochastic variance reduced gradient (SVRG) method for minimizing the sum of a finite number of smooth convex functions, which has been involved widely in the field of machine learning and data mining. Inspired by the excellent performance of two-point stepsize gradient method in batch learning, in this paper we present an improved SVRG algorithm, named stochastic two-point stepsize gradient method. Under some mild conditions, the proposed method achieves a linear convergence rate
This paper analyzes the approximate augmented Lagrangian dynamical systems for constrained optimization. We formulate the differential systems based on first derivatives and second derivatives of the approximate augmented Lagrangian. The solution of the original optimization problems can be obtained at the equilibrium point of the differential equation systems, which lead the dynamic trajectory into the feasible region. Under suitable conditions, the asymptotic stability of the differential systems and local convergence properties of their Euler discrete schemes are analyzed, including the locally quadratic convergence rate of the discrete sequence for the second derivatives based differential system. The transient behavior of the differential equation systems is simulated and the validity of the approach is verified with numerical experiments.
In this paper, we apply two nonparametric approaches to mean absolute deviation (MAD) portfolio selection model. The first one is to use the nonparametric kernel mean estimation to replace the returns of assets with five different kernel functions. Then, we construct the nonparametric kernel mean estimation-based MAD portfolio model. The second one is to utilize the nonparametric kernel median estimation to replace the returns of assets with five different kernel functions. Then, we construct the nonparametric kernel median estimation-based MAD portfolio model. We also extend the two kinds of nonparametric approach to mean-Conditional Value-at-Risk portfolio model. Finally, we give the in-sample and out-of-sample analysis of the proposed strategies and compare the performance of the proposed models by using actual stock returns in Shanghai stock exchange of China. The experimental results show the nonparametric estimation-based portfolio models are more efficient than the original portfolio model.
Active suspension control strategy design in vehicle suspension systems has been a popular issue in road vehicle applications. In this paper, we consider a quarter-car suspension problem. A nonlinear objective function together with a system of state-dependent ODEs is involved in the model. A differential equation approximation method, together with the control parametrization enhancing transform (CPET), is used to find the optimal proportional-integral-derivative (PID) feedback gains of the above model. Hence, an approximated optimal control problem is obtained. Proofs of convergences of the state and the optimal control of the approximated problem to those of the original optimal control problem are provided. A numerical example is solved to illustrate the efficiency of our method.
The paper considers a split inverse problem involving component equilibrium problems in Hilbert spaces. This problem therefore is called the split equilibrium problem (SEP). It is known that almost solution methods for solving problem (SEP) are designed from two fundamental methods as the proximal point method and the extended extragradient method (or the two-step proximal-like method). Unlike previous results, in this paper we introduce a new algorithm, which is only based on the projection method, for finding solution approximations of problem (SEP), and then establish that the resulting algorithm is weakly convergent under mild conditions. Several of numerical results are reported to illustrate the convergence of the proposed algorithm and also to compare with others.
In this paper, a new concave reformulation is proposed on a convex hull of some given points. Based on its properties, we attempt to solve DC Programming problems globally under uncertain environment by using Robust optimization method and CVaR method. A global optimization algorithm is developed for the Robust counterpart and CVaR model with two kinds of special convex hulls: simplex set and box set. The global solution is obtained by solving a sequence of convex relaxation programming on the original constraint sets or divided subsets with branch and bound method. Finally, numerical experiments are given for DC programs under uncertain environment with two kinds of constraints: simplex and box sets. Simulation results show the feasibility and efficiency of the proposed global optimization algorithm.
Emergency logistics is crucial to ameliorate the impact of large earthquakes on society. We present a modeling framework to assist decision makers in strategic and tactical planning for effective relief operations after an earthquake's occurrence. The objective is to perform these operations quickly while keeping its total expenses under a budget. The modeling framework locates/allocates resources in potentially affected zones, and transportation capacity is dynamically deployed in those zones. Demand uncertainty is directly incorporated through an impulse stochastic process. The novelty of this approach is threefold. It incorporates temporo-spatial dependence and demands heterogeneity. It incorporates the availability of transportation capacity at different zones. It incorporates tight budget constraints that precludes the total satisfaction of demands. The resulting model is a large size stochastic mixed-integer programming model, which can be approximately solved through Sample Average Approximation. An example is provided and a thorough sensitivity analysis is performed. The numerical results suggest that that the response times are highly sensitive to the availability of inventory at each period. In addition, all logistics parameters (except for inventory capacity) have approximately the same impact on the total response time. The elasticity for all these parameters indicate constant returns to scale.
In the real-world production process, the firms need to determine the optimal production planning under minimum production quantity constraint in order to achieve economies of scale. However, the inventory cost will hugely increase when there is a very large amount of production in a period and also a large amount of total demands for the next few periods. This paper considers a single-item dynamic lot sizing problem with production-or-outsourcing decisions. In each period, the production level cannot be lower than a given quantity in order to make full use of resources, but the outsourcing is unrestricted. The demands in a period can be backlogged. The production and outsourcing costs are fixed-plus-linear, and the inventory and backlogging costs are linear. We establish a mathematical programming model according to the real problem in the firm. We explore some structural properties of the optimal solution and use them to develop a dynamic programming algorithm to solve the proposed problem. We further present a special case with stationary production and outsourcing costs which can be solved with reduced computational complexities. In the end, we use three numerical instances to show how to obtain the optimal solutions by using the dynamic programming algorithm. Furthermore, we show that the policy of backlogging or outsourcing can reduce the total cost.
Cloud computing makes it possible for application providers to provide services seamlessly and application users to receive services adaptively. By offering services that give users an initial experience, application providers can usually attract more users. This research proposes a type of sleeping mechanism-based cloud architecture where an experience service and an enrollment service are provided on one virtual machine (VM). Accordingly, we model the cloud architecture as a queue with an asynchronous multi-vacation and a selectable extra service. We also analyze the queueing model in the steady state by constructing a three-dimensional Markov chain. Following this, we evaluate the system performance of the proposed cloud architecture based on the energy conservation level of the system and the mean delay of the visitors who select the enrollment service. Moreover, we study the Nash equilibrium strategy of visitors by building an individual welfare function, and develop an improved intelligent search algorithm to investigate the socially optimal strategy of visitors. Aiming to achieve a social optimum, we formulate a pricing policy with a reasonable enrollment fee.
We consider an inverse quadratic programming (QP) problem in which the parameters in the objective function of a given QP problem are adjusted as little as possible so that a known feasible solution becomes the optimal one. We formulate this problem as a minimization problem involving $l_1$ vector norm with a positive semidefinite cone constraint. By utilizing convex optimization theory, we rewrite its first order optimality condition as a generalized equation. Under extremely simple assumptions, we prove that any element of the generalized Jacobian of the equation at its solution is nonsingular. Based on this, we construct an inexact Newton method with Armijo line search to solve the equation and demonstrate its global convergence. Finally, we report the numerical results illustrating effectiveness of the Newton methods.
We address a two-machine no-wait flowshop scheduling problem with respect to the performance measure of total completion time. Minimizing total completion time is important when inventory cost is of concern. Setup times are treated separately from processing times. Furthermore, setup times are uncertain with unknown distributions and are within some lower and upper bounds. We develop a dominance relation and propose eight algorithms to solve the problem. The proposed algorithms, which assign different weights to the processing and setup times on both machines, convert the two-machine problem into a single-machine one for which an optimal solution is known. We conduct computational experiments to evaluate the proposed algorithms. Computational experiments reveal that one of the proposed algorithms, which assigns the same weight to setup and processing times, is superior to the rest of the algorithms. The results are statistically verified by constructing confidence intervals and test of hypothesis.
In this article, we study a discrete-time MAP/PH/1 queue with finite system capacity and two-stage vacations. The two-stage vacations policy which comprises single working vacation and multiple vacations is featured by that once the system is empty during the regular busy period, the system first takes the working vacation during which the server can still provide the service but at a lower service rate. After this working vacation, if the system is empty, the server will take a vacation during which the server stops its service completely, otherwise, the server resumes to the normal service rate. For this queue, using the matrix-geometric combination solution method, we obtain the stationary probability vectors when the traffic intensity is not equal to one. In addition, we discuss the spectrum properties of the key matrices and give their decomposition results that can be used to reduce the computation loads. Further, waiting time is derived by constructing an absorbing Markov chain. Various performance measures are obtained. At last, some numerical examples are presented to show the impacts of system parameters on performance measures.
Personality heterogeneity is an important topic in team management. In many working groups, there exists certain type of people that are talented but under-disciplined, who could occasionally make extraordinary contributions for the team, but often have less satisfactory overall performance. It is interesting to investigate whether the existence of such people in the team does help improve the overall team performance, and if it does so, what are the conditions for their existence to be positive, and through which channel their benefits for the team are manifested. This study proposes an analytical model with a simple structure that sets up an environment to study these questions. It is shown that: (1) feedback learning could be the mechanism through which outliers' benefits to the team are established, and thus could be a prerequisite for outliers' positive existence; (2) different types of teamwork settings have different outlier-positivity conditions: a uniform round-wise punishment for teamwork failures could be the key idea to encourage outliers' existence; for two specific types of teamwork, teamwork that highlights assistance in interactions are more outliers-friendly than teamwork that consists internal competitions. These results well match empirical observations and may have further implications for managerial practice.
In this paper, we are denoted to introducing the strict feasibility of a variational inclusion problem as a novel notion. After proving a new equivalent characterization for the nonemptiness and boundedness of the solution set for the variational inclusion problem under consideration, it is proved that the nonemptiness and boundedness of the solution set for the variational inclusion problem with a maximal monotone mapping is equivalent to its strict feasibility in reflexive Banach spaces.
The present paper studies a new class of problems of optimal control theory with linear second order self-adjoint Sturm-Liouville type differential operators and with functional and non-functional endpoint constraints. Sufficient conditions of optimality, containing both the second order Euler-Lagrange and Hamiltonian type inclusions are derived. The presence of functional constraints generates a special second order transversality inclusions and complementary slackness conditions peculiar to inequality constraints; this approach and results make a bridge between optimal control problem with Sturm-Liouville type differential differential inclusions and constrained mathematical programming problems in finite-dimensional spaces.The idea for obtaining optimality conditions is based on applying locally-adjoint mappings to Sturm-Liouville type set-valued mappings. The result generalizes to the problem with a second order non-self-adjoint differential operator. Furthermore, practical applications of these results are demonstrated by optimization of some semilinear optimal control problems for which the Pontryagin maximum condition is obtained. A numerical example is given to illustrate the feasibility and effectiveness of the theoretic results obtained.
The rumor propagation analysis and important nodes detection is a hot topic in complex network under crisis situation. The traditional propagation model does not consider enough states, so it cannot intact reflect the real world. In this paper, a new rumor propagation model which considers the Wiseman and the Truth Spreader is proposed based on the Graph Theory. Then, 3 new methods are proposed to find important nodes in the new model. These methods consider the differences between nodes to evaluate the importance of the nodes. Finally, 4 networks are demonstrated to show that the 3 proposed methods are useful to control rumor propagation.
In this paper, we consider a particular class of optimal switching problem for the linear-quadratic switched system in discrete time, where an optimal switching sequence is designed to minimize the quadratic performance index of the system with a switching cost. This is a challenging issue and studied only by few papers. First, we introduce a total variation function with respect to the switching sequence to measure the volatile switching action. In order to restrain the switching magnitude, it is added to the cost functional as a penalty. Then, the particular optimal switching problem is formulated. With the positive semi-definiteness of matrices, we construct a series of exact lower bounds of the cost functional at each time and the branch and bound method is applied to search all global optimal solutions. For the comparison between different global optimization methods, some numerical examples are given to show the efficiency of our proposed method.
In this paper, we focus on some inequalities for the Fan product of
This paper addresses a investment and risk control problem with a delay for an insurer in the defaultable market. Suppose that an insurer can invest in a risk-free bank account, a risky stock and a defaultable bond. Taking into account the history of the insurer's wealth performance, the controlled wealth process is governed by a stochastic delay differential equation. The insurer's goal is to maximize the expected exponential utility of the combination of terminal wealth and average performance wealth. We decompose the original optimization problem into two subproblems: a pre-default case and a post-default case. The explicit solutions in a finite dimensional space are derived for a illustrative situation, and numerical illustrations and sensitivity analysis for our results are provided.
Add your name and e-mail address to receive news of forthcoming issues of this journal:
[Back to Top]