Two-stage Optimization Strategy Research Articles

Computing planar and volumetric B-spline parameterizations for IGA by robust mapping fitting

We present a novel method to compute bijective domain parameterizations with low distortion for isogeometric analysis. Our key insight is that instead of solving the difficult mapping optimization problem, we fit the spline function to a piecewise linear map between computational and parametric domains while ensuring bijection. The basis of our key insight is that the bijection of the piecewise linear mapping is theoretically guaranteed for 2D Tutte (1963) and 3D Campen et al. (2016). Due to the continuity difference between the spline and piecewise linear maps, we develop a two-stage optimization strategy for robust map fitting. The first stage enforces the Jacobian similarity and the second optimizes the boundary difference. Besides, these two stages always guarantee bijection by explicit checks in combination with line search. We demonstrate the efficacy of our method on various complex models in both 2D and 3D. Compared to state-of-the-art methods, our method achieves higher robustness for computing bijective and low distortion domain parameterizations.

A two-stage optimization strategy for large-scale oil field development

The optimization of the locations of a large number of wells in an oil or gas field represents a challenging computational problem. This is because the number of optimization variables scales with the maximum number of wells considered. In this work, we develop and test a new two-stage strategy for large-scale oil field optimization problems. In the first stage, wells are constrained to lie in repeated patterns, and the reduced set of optimization variables defines the pattern type and geometry (e.g., well spacing, orientation). For this component of the optimization, we introduce several important modifications, including optimization of the drilling sequence, to an existing well-pattern optimization procedure. The solutions obtained in the first stage are then used to initialize the second stage optimization. In this stage we apply comprehensive field development optimization, where the well location, type (injection or production well), drill/do not drill decision, completion interval for 3D models, and drilling time variables are determined for each well. Pattern geometry is no longer enforced in this stage. Specialized treatments (consistent with actual drilling practice) are introduced for cases where multiple geomodels, used to capture geological uncertainty, are considered. In both stages optimization is achieved using a particle swarm optimization-mesh adaptive direct search (PSO-MADS) method. The two-stage procedure is applied to 2D and 3D models corresponding to different geological scenarios. Both deterministic and geologically uncertain systems are considered. Optimization results using the new procedure are shown to clearly outperform those from the single-stage comprehensive field development optimization approach. Specifically, for the same number of function evaluations, the two-stage treatment provides net present values that exceed those of the single-stage approach by about 15–18% for the cases considered. This suggests that this optimization strategy may indeed lead to improved results in practical problems.

Short-term oil production global optimization with operational constraints: A comparative study of nonlinear and piecewise linear formulations

The short-term oil production optimization of offshore platforms which use continuous gas lift and consider several operational constraints is a challenging task, specially when decision variables include both continuous and integer ones. The intrinsic non-linearity of the Gas Lift Performance Curves (GLPC) naturally yields to Mixer-Integer Nonlinear Program formulation. Picewise linear approximations of the GLPC are usually applied in order to reduce the complexity of the optimization problem and to ensure an approximate solution, leading to Mixer-Integer Linear Program formulation. However, a comprehensive and comparative study of these two formulations for the oil production optimization problem is scarce in the literature. This paper provides a deep performance evaluation of these two approaches, showing the limitations of both strategies, and proposes a hybrid two-stage optimization strategy to achieve the global optimal operating condition which maximizes the oil production subject to safety, environmental and processing capacity constraints.

Two-stage optimization for energy-efficient bipedal walking

This paper proposes a two-stage optimization strategy for energy-efficient gait generation. At the first stage, by tracking the reference zero moment point (ZMP) trajectory, the optimal center of mass (CoM) trajectory, which contributes to the minimal unit energetic cost (UEC) of one step, is solved analytically by using an unconstrained optimization method. At the second stage, to minimize the multi-joint mechanical work, the ZMP reference during the single support phase is optimized by a constrained optimization method. As a result, by considering the feasibility constraints such as the limitation on ZMP movement, the energy-efficient walking patterns can be generated in real-time. Furthermore, the energetic performances under different step parameter configurations, which consist of step length, step duration, and time ratio of double support, are discussed. Simulations and hardware experiments have demonstrated the energetic benefits of the proposed strategy when compared with other state-of-the-art works.

Performance optimization of non-equilibrium ionization simulations from MapReduce and GPU acceleration

We propose a two-stage optimization strategy to accelerate non-equilibrium ionization (NEI) calculation that is crucial to various high energy astrophysical phenomena, by using methods of MapReduce modeling and GPU acceleration. First, we construct a parallel pipeline based on the MapReduce model that processes massive particles trajectories on a separate mesh decoupled from that has been used by other equations in the multiphysics simulations. Second, we accelerate the calculation of massive NEI equations by taking full advantage of heterogeneous multicore architecture of GPUs. The approach has been prototyped and tested in simulations using FLASH code and AtomDB atomic database. Our results show that the method can improve the end-to-end performance by 3-fold with less computing resources and reduce the overhead significantly. For standalone tests, the GPU-accelerated NEI solver can achieve a maximum 212-fold speedup compared to the CPU-based solver. With the capability to support nonintrusive simulation-time data analysis, our approach can be also applied to other multiphysics processes such as reactive flow simulations.

A multi-objective optimization strategy for the optimal control scheme of pumped hydropower systems under successive load rejections

Pumped hydropower is the most important sustainable energy in a power grid, which is converted into electricity through pumped storage hydropower systems. The energy conversion process in a pumped storage hydropower system, however, may be greatly threatened by extreme energy conversion conditions, especially successive load rejections. Although successive load rejections have been addressed in previous studies, the optimal control scheme for successive load rejections has not yet been reported. In this study, a refined nonlinear model for successive load rejections is constructed by incorporating a guaranteed calculation for regulation (GCR), and the effects of the rejection parameters on the transient processes of pumped storage hydropower systems are investigated quantitatively based on the refined model. The results show that: (1) the rejection parameters significantly impact the successive load rejections, and the worst working condition of a successive load rejection can be identified by the worst combination of rejection parameters; (2) the rotational speed and the water pressure of the pump storage unit are conflicting objectives with variation in the control schemes. Moreover, a novel two-stage multi-objective optimization strategy for the optimal control scheme of pumped hydropower systems is proposed and a case study is conducted based on a real pumped storage hydropower system in China. Compared with the on-site measurements, the proposed optimization strategy can improve the maximum volute water pressure and the minimum draft tube water pressure by at most 7.6% and 17.4% under the worst working condition of a successive load rejection. Furthermore, it is verified that the obtained optimal control scheme meets the GCR criterion under various working conditions that are not the worst case. These results highlight the effectiveness and availability of the proposed optimization strategy for sustaining the safe operation of pumped storage hydropower systems.

Two-Stage Optimal Scheduling Strategy for Large-Scale Electric Vehicles

This paper proposes a two-stage scheduling strategy for large-scale electric vehicles to reduce the adverse impact of the uncontrolled charging of the electric vehicles on the grid. Based on the statistical data of private car travel, the uncontrolled charging demand of individual electric vehicles and their aggregation are simulated. In the first stage, the electric vehicles and thermal power units are jointly scheduled. To minimize the total cost and standard deviation of the total load curve, the charging and discharging load guiding curve of the electric vehicles and the optimal output plans of the thermal power units in each period of the scheduled day are formulated. In the second stage, the electric vehicle load management and control centre formulates specific charging and discharging plans for the users through rolling optimization to follow the guiding load curve. The cost of vehicle discharge compensation is considered to improve the willingness of users to participate in scheduling and the user satisfaction. To avoid the “dimension disaster” caused by the centralized dispatching of large numbers of electric vehicles, the K-means clustering algorithm is used to divide the vehicles into different groups. Next, each group is scheduled as a unit, and the model is solved by using the particle swarm optimization algorithm. By comparing the optimization results of different scenarios, the feasibility and effectiveness of the proposed strategy are verified.

Hybrid Energy Management System for Operation of Wind Farm System Considering Grid-Code Constraints

In this paper, a hybrid energy management system is developed to optimize the operation of a wind farm (WF) by combining centralized and decentralized approaches. A two-stage optimization strategy, including distributed information sharing (stage 1); and centralized optimization (stage 2) is proposed to find out the optimal set-points of wind turbine generators (WTGs) considering grid-code constraints. In stage 1, cluster energy management systems (CEMSs) and transmission system operator (TSO) interact with their neighboring agents to share information using diffusion strategy and then determine the mismatch power amount between the current output power of WF and the required power from TSO. This amount of mismatch power is optimally allocated to all clusters through the CEMSs. In stage 2, a mixed-integer linear programming (MILP)-based optimization model is developed for each CEMS to find out the optimal set-points of WTGs in the corresponding cluster. The CEMSs are responsible for ensuring the operation of WF in accordance with the requirements of TSO (i.e., grid-code constraints) and also minimizing the power deviation for the set-points of WTGs in each cluster. The minimization of power deviation helps to reduce the internal power fluctuations inside each cluster. Finally, to evaluate the effectiveness of the proposed method, several case studies are analyzed in the simulations section for operation of a WF with 20 WTGs in four different clusters.

Two-Stage Optimal Operation Strategy of Isolated Microgrid With TSK Fuzzy Identification of Supply Security

Due to the uncertainty of intermittent energy and system load, it is a big challenge to optimally operate an isolated power system. This article proposes a two-stage optimal operation strategy with a Takagi–Sugeno–Kang (TSK) fuzzy system to address the supply security under uncertainty circumstance. For proper analysis of the uncertainty characteristics, adjustable uncertainty parameters of intermittent energy resource and system load are taken as fuzzy sets; with the consideration of the robustness of these uncertainty parameters on isolated power system, it creates a supply-security identification model with the TSK fuzzy approach under radial basis function (RBF) neural network, and deduces optimal weight values with a recursive least square method. For properly avoiding potential risks, security index is classified into several degrees, each degree of risk can switch a different operation model, which can ensure the supply security of an isolated power system. For properly solving the optimization model, gradient descent-based multiobjective cultural differential evolution is employed to minimize economic cost and emission rate simultaneously. With simulations on isolated regional network, the obtained results reveal that the proposed method can be a viable alternative for optimal operation in isolated power systems.

Thermomechanical Coupling Analysis and Optimization of Metallic Thermal Protection System

The metallic thermal protection system (MPTS) is a key technology for reducing the cost of reusable launch vehicles, offering the combination of increased durability and competitive weights when compared with other systems. A two‐stage optimization strategy is proposed in this paper to improve thermal and mechanical performance of MTPS while minimizing its weight. Sensitivity analysis is conducted to determine the influence of various design variables, and these variables are classified into two groups according to the results. Three main parameters are optimized in the first‐stage optimization, and others are optimized in the second stage. Combining the response surface method with genetic algorithm ensures the computational efficiency of this strategy. Optimum results show that the thermal insulation capability of MTPS is improved by 20% with little mass cost, validating the feasibility of this optimization strategy.

Eagle Strategy based Crow Search Algorithm for solving Unit Commitment Problem

<p>Eagle strategy is a two-stage optimization strategy, which is inspired by the observation of the hunting behavior of eagles in nature. In this two-stage strategy, the first stage explores the search space globally by using a Levy flight; if it finds a promising solution, then an intensive local search is employed using a more efficient local optimizer, such as hillclimbing and the downhill simplex method. Then, the two-stage process starts again with new global exploration, followed by a local search in a new region. One of the remarkable advantages of such a combina-tion is to use a balanced tradeoff between global search (which is generally slow) and a rapid local search. The crow search algorithm (CSA) is a recently developed metaheuristic search algorithm inspired by the intelligent behavior of crows.This research article integrates the crow search algorithm as a local optimizer of Eagle strategy to solve unit commitment (UC) problem. The Unit commitment problem (UCP) is mainly finding the minimum cost schedule to a set of generators by turning each one either on or off over a given time horizon to meet the demand load and satisfy different operational constraints. There are many constraints in unit commitment problem such as spinning reserve, minimum up/down, crew, must run and fuel constraints. The proposed strategy ES-CSA is tested on 10 to 100 unit systems with a 24-h scheduling horizon. The effectiveness of the proposed strategy is compared with other well-known evolutionary, heuristics and meta-heuristics search algorithms, and by reported numerical results, it has been found that proposed strategy yields global results for the solution of the unit commitment problem.</p><p> </p>

More is Better: Precise and Detailed Image Captioning Using Online Positive Recall and Missing Concepts Mining.

Recently, a great progress in automatic image captioning has been achieved by using semantic concepts detected from the image. However, we argue that existing concepts-to-caption framework, in which the concept detector is trained using the image-caption pairs to minimize the vocabulary discrepancy, suffers from the deficiency of insufficient concepts. The reasons are two-fold: 1) the extreme imbalance between the number of occurrence positive and negative samples of the concept and 2) the incomplete labeling in training captions caused by the biased annotation and usage of synonyms. In this paper, we propose a method, termed online positive recall and missing concepts mining, to overcome those problems. Our method adaptively re-weights the loss of different samples according to their predictions for online positive recall and uses a two-stage optimization strategy for missing concepts mining. In this way, more semantic concepts can be detected and a high accuracy will be expected. On the caption generation stage, we explore an element-wise selection process to automatically choose the most suitable concepts at each time step. Thus, our method can generate more precise and detailed caption to describe the image. We conduct extensive experiments on the MSCOCO image captioning data set and the MSCOCO online test server, which shows that our method achieves superior image captioning performance compared with other competitive methods.

Resource Allocation Optimization Based on Mixed Integer Linear Programming in the Multi-Cloudlet Environment

Cloudlet provides services with low latency and high bandwidth. Some research addresses the resource allocation problem in the cloudlet-based mobile cloud computing environment. However, there are few works considering how to optimize resource allocation while satisfying users’ requirements in multi-cloudlet situations. To solve this problem, a two-stage optimization strategy is proposed. First, a cloudlet selection model based on mixed integer linear programming (MILP) is proposed to obtain the cloudlet for mobile users by optimizing latency and mean reward. Second, a resource allocation model based on MILP is presented to allocate resources in the selected cloudlet by optimizing reward and mean resource usage. A comparison of resource allocation is analyzed with a cloudlet selection model based on MILP and an existing cloudlet selection strategy in the multi-cloudlet environment. From the simulation, our proposed strategy has better performance in terms of the access latency, the system reward and the resource usage.

Robust optimization for integrated scrap steel charge considering uncertain metal elements concentrations and production scheduling under time-of-use electricity tariff

Under the background of time-of-use electricity tariff, the problems of scrap steel charge optimization considering metal elements concentrations uncertainty and scrap steel melt shop production scheduling optimization are integrally studied in this paper using robust optimization approaches. First, taking the parallel electric arc furnaces melt shop as the research object, this paper constructs a robust optimization model about integrated scrap steel charge considering uncertain metal elements concentrations and production scheduling under time-of-use electricity tariff applying a block modeling method. Then, based on the assumptions that the random parameter values of scrap steel metal elements concentrations are distributed in the "bounded uncertainty" set and "bounded symmetric uncertainty" set respectively, the robust optimization model is transformed into two deterministic counterpart models. Finally, Cplex software is used to solve the nominal model and two deterministic counterpart models, the results of which are also compared with two-stage optimization strategy results, indicating the feasibility and effectiveness of the models. Meanwhile, the impacts of uncertainty of metal elements concentrations on production costs, the impacts of time-of-use electricity tariff on production scheduling, and the impacts of electricity tariff fluctuations on the consumption of scrap steels and ferroalloys and the number of setups are analyzed.

Research on Operation Strategy for Bundled Wind-thermal Generation Power Systems Based on Two-Stage Optimization Model

Wind power has the advantages of being clean and non-polluting and the development of bundled wind-thermal generation power systems (BWTGSs) is one of the important means to improve wind power accommodation rate and implement “clean alternative” on generation side. A two-stage optimization strategy for BWTGSs considering wind speed forecasting results and load characteristics is proposed. By taking short-term wind speed forecasting results of generation side and load characteristics of demand side into account, a two-stage optimization model for BWTGSs is formulated. By using the environmental benefit index of BWTGSs as the objective function, supply-demand balance and generator operation as the constraints, the first-stage optimization model is developed with the chance-constrained programming theory. By using the operation cost for BWTGSs as the objective function, the second-stage optimization model is developed with the greedy algorithm. The improved PSO algorithm is employed to solve the model and numerical test verifies the effectiveness of the proposed strategy.

A Two-Stage Optimization Strategy for Fuzzy Object-Based Analysis Using Airborne LiDAR and High-Resolution Orthophotos for Urban Road Extraction

In the last decade, object-based image analysis (OBIA) has been extensively recognized as an effective classification method for very high spatial resolution images or integrated data from different sources. In this study, a two-stage optimization strategy for fuzzy object-based analysis using airborne LiDAR was proposed for urban road extraction. The method optimizes the two basic steps of OBIA, namely, segmentation and classification, to realize accurate land cover mapping and urban road extraction. This objective was achieved by selecting the optimum scale parameter to maximize class separability and the optimum shape and compactness parameters to optimize the final image segments. Class separability was maximized using the Bhattacharyya distance algorithm, whereas image segmentation was optimized using the Taguchi method. The proposed fuzzy rules were created based on integrated data and expert knowledge. Spectral, spatial, and texture features were used under fuzzy rules by implementing the particle swarm optimization technique. The proposed fuzzy rules were easy to implement and were transferable to other areas. An overall accuracy of 82% and a kappa index of agreement (KIA) of 0.79 were achieved on the studied area when results were compared with reference objects created via manual digitization in a geographic information system. The accuracy of road extraction using the developed fuzzy rules was 0.76 (producer), 0.85 (user), and 0.72 (KIA). Meanwhile, overall accuracy was decreased by approximately 6% when the rules were applied on a test site. A KIA of 0.70 was achieved on the test site using the same rules without any changes. The accuracy of the extracted urban roads from the test site was 0.72 (KIA), which decreased to approximately 0.16. Spatial information (i.e., elongation) and intensity from LiDAR were the most interesting properties for urban road extraction. The proposed method can be applied to a wide range of real applications through remote sensing by transferring object-based rules to other areas using optimization techniques.

Joint slot allocation and dynamic pricing of container sea–rail multimodal transportation

The container sea–rail multimodal transport system faces complex challenges with demand uncertainties for joint slot allocation and dynamic pricing. The challenge is formulated as a two-stage optimal model based on revenue management (RM) as actual slots sale of multi-node container sea–rail multimodal transport usually includes contract sale to large shippers and free sale to scattered shippers. First stage in the model utilizes an origin-destination control approach, formulated as a stochastic integer programming equation, to settle long-term slot allocation in the contract market and empty container allocation. Second stage in the model is formulated as a stochastic nonlinear programming equation to solve a multiproduct joint dynamic pricing and inventory control problem for price settling and slot allocation in each period of free market. Considering the random nature of demand, the methods of chance constrained programming and robust optimization are utilized to transform stochastic models into deterministic models. A numerical experiment is presented to verify the availability of models and solving methods. Results of considering uncertain/certain demand are compared, which show that the two-stage optimal strategy integrating slot allocation with dynamic pricing considering random demand is revealed to increase the revenue for multimodal transport operators (MTO) while concurrently satisfying shippers' demand. Research resulting from this paper will contribute to the theory and practice of container sea–rail multimodal transport revenue management and provide a scientific decision-making tool for MTO.

Optimal posturing in day-ahead market clearing for uncertainties considering anticipated real-time adjustment costs

This paper proposes a market-clearing mechanism which explicitly takes into account the impact of uncertainties in wind power generation and load forecast. Since market clearing is a multisettlement process-day ahead and real time (RT), a strategy is proposed, which provides best-fit day-ahead (DA) schedule, which minimizes the twin (both DA and RT adjustment) costs/maximizes social welfare, under all possible scenarios in RT. This two-stage optimization strategy consists of a genetic algorithm (GA) based DA market clearing and a two-point estimate-based probabilistic RT optimal power flow (OPF). The former generates sample schedules, while the latter provides mean adjustment costs. Two commonly employed standard market practices to incorporate wind energy into wholesale electricity markets have been presented. The results for a sample system with GA and two-point estimate OPF, and GA and Monte Carlo simulation have been obtained to ascertain the effectiveness of the proposed method.

Semi-supervised learning with nuclear norm regularization

Integrating new knowledge sources into various learning tasks to improve their performance has recently become an interesting topic. In this paper we propose a novel semi-supervised learning (SSL) approach, called semi-supervised learning with nuclear norm regularization (SSL-NNR), which can simultaneously handle both sparse labeled data and additional pairwise constraints together with unlabeled data. Specifically, we first construct a unified SSL framework to combine the manifold assumption and the pairwise constraints assumption for classification tasks. Then we provide a modified fixed point continuous algorithm to learn a low-rank kernel matrix that takes advantage of Laplacian spectral regularization. Finally, we develop a two-stage optimization strategy, and present a semi-supervised classification algorithm with enhanced spectral kernel (ESK). Moreover, we also present a theoretical analysis of the proposed ESK algorithm, and derive an easy approach to extend it to out-of-sample data. Experimental results on a variety of synthetic and real-world data sets demonstrate the effectiveness of the proposed ESK algorithm.

Using a two-stage optimization strategy for the active alignment of multifiber optical devices

The alignment of optical components is a key factor when designing and manufacturing multifiber optical systems. This problem can be treated as a standard multiobjective optimization problem and solved by numerical optimization methodologies. The core diameter of a single-mode fiber is ~9 µm, and any slight misalignment during manufacturing will cause signification optical losses in connections. Previous studies have shown that the currently used alignment methods for multifiber devices can increase the optical power summation of all fibers, but the results are not very accurate. This study first compares different numerical optimization methodologies that can be used to find the ideal connection position. Two indices are used to judge the performances of different methods: the required time and the optical power. Next, a two-stage optimization strategy is proposed to obtain a fast and accurate result. In the first stage, the Nelder-Mead simplex method is used to move toward the optimum position quickly. In the second stage, the steepest descent method with polynomial interpolation is applied to improve the accuracy because of the stability of the method.