Optimization and Control (math.OC)

Abstract: We develop a principled approach to end-to-end learning in stochastic optimization. First, we show that the standard end-to-end learning algorithm admits a Bayesian interpretation and trains a posterior Bayes action map. Building on the insights of this analysis, we then propose new end-to-end learning algorithms for training decision maps that output solutions of empirical risk minimization and distributionally robust optimization problems, two dominant modeling paradigms in optimization under uncertainty. Numerical results for a synthetic newsvendor problem illustrate the key differences between alternative training schemes. We also investigate an economic dispatch problem based on real data to showcase the impact of the neural network architecture of the decision maps on their test performance.

Abstract: In the paper, we introduce several accelerate iterative algorithms for solving the multiple-set split common fixed-point problem of quasi-nonexpansive operators in real Hilbert space. Based on primal-dual method, we construct several iterative algorithms in a way that combines inertial technology and the self-adaptive stepsize such that the implementation of the algorithms doesn't need any prior information about bounded linear operator norm. Under suitable assumptions, weak convergence of the proposed algorithms is established. As applications, we obtain relative iterative algorithms to solve the multiple-set split feasibility problem. Finally, the performance of the proposed algorithms is illustrated by numerical experiments.

Abstract: In road traffic networks, large traffic volumes may lead to extreme delays. These severe delays are caused by the fact that, whenever the maximum capacity of a road is approached, speeds drop rapidly. Therefore, the focus in this paper is on real-time control of traffic input rates, thereby aiming to prevent such detrimental capacity drops. To account for the fact that, by the heterogeneity within and between traffic streams, the available capacity of a road suffers from randomness, we introduce a stochastic flow model that describes the impact of traffic input streams on the available road capacities. Then, exploiting similarities with traffic control of telecommunication networks, in which the available bandwidth is a stochastic function of the input rate, and in which the use of effective bandwidths have proven an effective input rate control framework, we propose a similar traffic rate control policy based on the concept of effective bandwidths. This policy allows for increased waiting times at the access boundaries of the network, so as to limit the probability of large delays within the network. Numerical examples show that, by applying such a control policy capacity violations are indeed rare, and that the increased waiting at the boundaries of the network is of much smaller scale, compared to uncontrolled delays in the network.

Abstract: This study explores the potential of using public transportation systems for freight delivery, where we intend to utilize the spare capacities of public vehicles like buses, trams, metros, and trains, particularly during off-peak hours, to transport packages within the city instead of using dedicated delivery vehicles. The study contributes {to the growing} literature on innovative strategies for performing sustainable last mile deliveries. We study an operational level problem called the Three-Tier Delivery Problem on Public Transportation, where packages are first transported from the Consolidation and Distribution Center (CDC) to nearby public vehicle stations by delivery trucks. From there, public vehicles transport them into the city area. The last leg of the delivery is performed to deliver the packages to their respective customers using green vehicles or eco-friendly systems. We propose mixed-integer linear programming formulations to study the transport of packages from the CDC to the customers, use decomposition approaches to solve them, and provide numerical experiments to demonstrate the efficiency and effectiveness of the system. Our results show that this system has the potential to drastically reduce the length of trips performed by dedicated delivery vehicles, thereby reducing the negative social and environmental impacts of existing last mile delivery systems.

Abstract: The purpose of this paper is to introduce two new classes of accelerated distributed proximal conjugate gradient algorithms for multi-agent constrained optimization problems; given as minimization of a function decomposed as a sum of M number of smooth and M number of nonsmooth functions over the common fixed points of M number of nonlinear mappings. Exploiting the special properties of the cost component function of the objective function and the nonlinear mapping of the constraint problem of each agent, a new inertial accelerated incremental and parallel computing distributed algorithms will be presented based on the combinations of computations of proximal, conjugate gradient and Halpern methods. Some numerical experiments and comparisons are given to illustrate our results.

Abstract: Linear approximation commonly used in solving alternating-current optimal power flow (AC-OPF) simplifies the system models but incurs accumulated voltage errors in large power networks. Such errors will make the primal-dual type gradient algorithms converge to the solutions at which the power networks may be exposed to the risk of voltage violation. In this paper, we improve a recent hierarchical OPF algorithm that rested on primal-dual gradients evaluated with a linearized distribution power flow model. Specifically, we propose a more accurate gradient evaluation method based on a three-phase unbalanced nonlinear distribution power flow model to mitigate the errors arising from model linearization. The resultant gradients feature a blocked structure that enables us to further develop an improved hierarchical primal-dual algorithm to solve the OPF problem. Numerical results on the IEEE $123$-bus test feeder and a $4,518$-node test feeder show that the proposed method can enhance the overall voltage safety while achieving comparable computational efficiency with the linearized algorithm.

Abstract: SeDuMi and SDPT3 are two solvers for solving Semi-definite Programming (SDP) or Linear Matrix Inequality (LMI) problems. A computational performance comparison of these two are undertaken in this paper regarding the Stability of Continuous-time Linear Systems. The comparison mainly focuses on computational times and memory requirements for different scales of problems. To implement and compare the two solvers on a set of well-posed problems, we employ YALMIP, a widely used toolbox for modeling and optimization in MATLAB. The primary goal of this study is to provide an empirical assessment of the relative computational efficiency of SeDuMi and SDPT3 under varying problem conditions. Our evaluation indicates that SDPT3 performs much better in large-scale, high-precision calculations.

Abstract: We introduce a new path-by-path approach to mean field games with common noise that recovers duality at the pathwise level. We verify this perspective by explicitly solving some difficult examples with linear-quadratic data, including control in the volatility coefficient of the common noise as well as the constraint of partial information. As an application, we establish the celebrated separation principle in the latter context. In pursuing this program, we believe we have made a crucial contribution to clarifying the notion of regular solution in the path dependent PDE literature.