Evaluation Of Candidate Solutions Research Articles

Multi-objective reinforcement learning-based approach for pressurized water reactor optimization

A novel method, the Pareto Envelope Augmented with Reinforcement Learning (PEARL), has been developed to address the challenges posed by multi-objective problems, particularly in the field of engineering where the evaluation of candidate solutions can be time-consuming. PEARL distinguishes itself from traditional policy-based multi-objective Reinforcement Learning methods by learning a single policy, eliminating the need for multiple neural networks to independently solve simpler sub-problems. Several versions inspired from deep learning and evolutionary techniques have been crafted, catering to both unconstrained and constrained problem domains. Curriculum Learning (CL) is harnessed to effectively manage constraints in these versions. PEARL’s performance is first evaluated on classical multi-objective benchmarks. Additionally, it is tested on two practical PWR core Loading Pattern (LP) optimization problems to showcase its real-world applicability. The first problem involves optimizing the Cycle length (LC) and the rod-integrated peaking factor (FΔh) as the primary objectives, while the second problem incorporates the mean average enrichment as an additional objective. Furthermore, PEARL addresses three types of constraints related to boron concentration (Cb), peak pin burnup (Bumax), and peak pin power (Fq). The results are systematically compared against conventional approaches from stochastic optimization. Notably, PEARL, specifically the PEAL-NdS ( ▪ ) variant, efficiently uncovers a Pareto front without necessitating additional efforts from the algorithm designer, as opposed to a single optimization with scaled objectives. It also outperforms the classical approach across multiple performance metrics, including the Hyper-volume. Future works will encompass a sensitivity analysis of hyper-parameters with statistical analysis to optimize the application of PEARL and extend it to more intricate problems.

An effective hybrid search method for the quadratic knapsack problem with conflict graphs

The quadratic knapsack problem (QKP) is a variant of the well-known knapsack problem and arises in a variety of real life applications. The quadratic knapsack problem with conflict graphs (QKPCG) further extends QKP by considering the conflicts of items. In this work, we propose an effective hybrid search method based on the framework of memetic algorithm to tackle QKPCG. The method integrates a randomized uniform-based crossover operator to generate promising offspring solutions, a multi-neighborhood tabu search to perform local optimization, and a streamline technique to speed up the evaluation of candidate solutions. The method shows a competitive performance compared to the state-of-the-art approaches in the literature. It finds 3 improved best-known solutions and matches the best-known solutions for all the remaining cases out of the 45 benchmark instances. We investigate the effects of the key ingredients of the algorithm.

Deep attention models with dimension-reduction and gate mechanisms for solving practical time-dependent vehicle routing problems

Time dependencies of travel speeds in time-dependent vehicle routing problems (TDVRPs) are usually accounted for by discretizing the planning horizon into several time periods. However, travel speeds usually change frequently in real road networks, so many time periods are needed to evaluate candidate solutions accurately in model and solution construction for practical TDVRPs, which increases substantially the computational complexity of TDVRPs. We develop two deep attention models with dimension-reduction and gate mechanisms to solve practical TDVRPs in real urban road networks. In the two models, a multi-head attention-based dimension-reduction mechanism is proposed to reduce the dimension of model inputs and obtain enhanced node representation, whereas a gate mechanism is introduced to obtain better information representation. On the basis of a travel speed dataset from an urban road network, we conduct extensive experiments to validate the effectiveness of the proposed models on practical TDVRPs with or without consideration of time windows. Experimental results show that our models can solve TDVRPs with 240 time periods and up to 250 customers effectively and efficiently and provide significantly superior overall performances over two representative heuristics and two state-of-the-art deep reinforcement learning models. Especially, compared with a recent tabu search method, our models can reduce the computation time by up to 3,540 times and improve the solution performance by up to 23%. Moreover, our models have an outstanding generalization performance. The model trained for the 30-customer TDVRP with time windows can be used directly to solve problems with up to 250 customers effectively by generating superior solutions over those generated by benchmarking methods.

An adapter-based architecture for evaluating candidate solutions in energy system scheduling

Increasing shares of volatile generation and non-steerable demand raise the need for automated control of the Energy Systems (ESs). Various solutions for management and schedule-based control of energy facilities exist today. However, the amount and diversity of applications lead to a multitude of different automated energy management solutions. Different optimization algorithms have proven more or less effective for energy management. The multitude of optimization algorithms and energy management solutions require flexible, modular, and scalable integrations. We present a novel Optimization Service (OS) for easily integrating optimization algorithms while evaluating candidate solutions in the context of ESs applications. We propose an adapter-based architecture using metadata and domain knowledge to bridge between clients, e.g. smart grid applications and optimization algorithms. The architecture interfaces different clients with optimizers in a flexible and modular way. The clients provide metadata-based descriptions of optimization jobs translated by OS. OS then interacts with optimizers and evaluates candidate solutions. A consistent definition of interfaces for clients and optimization algorithms facilitates the modular evaluation of candidate solutions. OS’s separation of client and optimization algorithms increases scalability by managing computational resources independently. We evaluate the presented architecture for scheduling a so-called Energy Hub (EH) as a test case describing a simulation scenario of a renewable EH embedded in grid scenarios from an industrial area in Karlsruhe, Germany. OS utilizes an Evolutionary Algorithm (EA) to optimize schedules for cost and strain on the electrical grid. The use case exemplifies OS’s advantages in a proof-of-concept evaluation.

MOHAQ: Multi-Objective Hardware-Aware Quantization of recurrent neural networks

The compression of deep learning models is of fundamental importance in deploying such models to edge devices. The selection of compression parameters can be automated to meet changes in the hardware platform and application. This article introduces a Multi-Objective Hardware-Aware Quantization (MOHAQ) method, which considers hardware performance and inference error as objectives for mixed-precision quantization. The proposed method feasibly evaluates candidate solutions in a large search space by relying on two steps. First, post-training quantization is applied for fast solution evaluation (inference-only search). Second, we propose the ”beacon-based search” to retrain selected solutions only and use them as beacons to estimate the effect of retraining on other solutions. We use speech recognition models on TIMIT dataset. Experimental evaluations show that Simple Recurrent Unit (SRU)-based models can be compressed up to 8x by post-training quantization without any significant error increase. On SiLago, we found solutions that achieve 97% and 86% of the maximum possible speedup and energy saving, with a minor increase in error on an SRU-based model. On Bitfusion, the beacon-based search reduced the error gain of the inference-only search on SRU-based models and Light Gated Recurrent Unit (LiGRU)-based model by up to 4.9 and 3.9 percentage points, respectively.

Optimization of mooring systems in the context of an integrated design methodology

This work describes an enhanced mooring optimization procedure, oriented towards recent floating production systems (FPS) for oil & gas exploitation in ultra-deep-water scenarios, which may present a large number of risers in an asymmetric layout. Acknowledging that the risers are the key component of an FPS, the optimization procedure is associated to an integrated mooring-riser design methodology; thus, instead of simply minimizing the platform offsets and/or the costs of the mooring system itself, one of the main objectives is to obtain a mooring configuration that ensures the integrity of the risers. Other highlights of the optimization procedure include the following aspects: Enhancements in the modeling of the optimization problem (including the definition of design variables, objective function and constraints that are relevant for such actual applications); The use of the PSO optimization algorithm associated to the ε-constrained method to efficiently handle the constraints; Enhancements in the evaluation of candidate solutions, by full nonlinear time-domain dynamic Finite Element simulations with coupled models; and the implementation in a parallel computing environment to deal with the high associated computational costs. A case study considering an FPS representative of actual applications in deepwater scenarios is presented to illustrate the practical application of the optimization tool.

Identification and Utility of a Short Form of the Pediatric Symptom Checklist-Youth Self-Report (PSC-17-Y)

Abstract. The Pediatric Symptom Checklist-Youth self-report (PSC-Y) is a 35-item measure of adolescent psychosocial functioning that uses the same items as the original parent report version of the PSC. Since a briefer (17-item) version of the parent PSC has been validated, this paper explored whether a subset of items could be used to create a brief form of the PSC-Y. Data were collected on more than 19,000 youth who completed the PSC-Y online as a self-screen offered by Mental Health America. Exploratory factor analyses (EFAs) were first conducted to identify and evaluate candidate solutions and their factor structures. Confirmatory factor analyses (CFAs) were then conducted to determine how well the data fit the candidate models. Tests of measurement invariance across gender were conducted on the selected solution. The EFAs and CFAs suggested that a three-factor short form with 17 items is a viable and most parsimonious solution and met criteria for scalar invariance across gender. Since the 17 items used on the parent PSC short form were close to the best fit found for any subsets of items on the PSC-Y, the same items used on the parent PSC-17 are recommended for the PSC-Y short form.

Minimizing Expected Cycle Time of Stochastic Customer Orders Through Bounded Multi-Fidelity Simulations

This paper considers the scheduling of stochastic customer orders to minimize expected cycle time. Customer orders dynamically arrive at a machine station, and each order consists of multiple product types. Random workloads are required by each product type, and the workloads are assigned to a set of unrelated parallel machines in the station to be processed. The objective is to obtain the minimal long-run expected order cycle time through proper workload assignments. In view of the difficulty in evaluating the objective function, this paper models the targeted problem as a simulation optimization problem and proposes to solve it under the multifidelity model framework. To improve the efficiency in evaluating candidate solutions, a low-fidelity model is constructed to select solutions with better performances for high-fidelity simulations. The effectiveness of this low-fidelity model is demonstrated through a series of theoretical evidences. A simulation optimization algorithm, named Bound-Multi-fidelity Optimization with Ordinal Transformation and Optimal Sampling Bound-(MO2TOS), is developed by taking advantage of the properties of the low-fidelity model. Numerical experiments are conducted to evaluate the performance of the proposed algorithm against three other well-known simulation optimization algorithms in the literature. Results indicate that the Bound-MO2TOS outperforms all the other tested algorithms, and its performance is robust against changes in problem scale. Note to Practitioners —Customer order scheduling models have wide applications in industries where every single order may contain multiple product types and the entire order requires one shipment delivery. In many applications, such as pigments dyes manufacturing, auto repair, and grocery, consideration of customer orders rather than individual jobs is often preferred. Given the stochastic nature and the synchronization constraints of the problem, a multifidelity simulation algorithm named Bound-Multi-fidelity Optimization with Ordinal Transformation and Optimal Sampling Bound-(MO2TOS) is proposed. Experimental results suggest that the proposed algorithm outperforms three other popular simulation optimization algorithms under a variety of scenarios. It is admitted that further improvement can be made in refining the computation resource distribution process of Bound-MO2TOS. In the future research, such an improvement will be more thoroughly elaborated so that practitioners can utilize it to coordinate their productions better.

A Random Forest-Assisted Evolutionary Algorithm for Data-Driven Constrained Multiobjective Combinatorial Optimization of Trauma Systems.

Many real-world optimization problems can be solved by using the data-driven approach only, simply because no analytic objective functions are available for evaluating candidate solutions. In this paper, we address a class of expensive data-driven constrained multiobjective combinatorial optimization problems, where the objectives and constraints can be calculated only on the basis of a large amount of data. To solve this class of problems, we propose using random forests (RFs) and radial basis function networks as surrogates to approximate both objective and constraint functions. In addition, logistic regression models are introduced to rectify the surrogate-assisted fitness evaluations and a stochastic ranking selection is adopted to further reduce the influences of the approximated constraint functions. Three variants of the proposed algorithm are empirically evaluated on multiobjective knapsack benchmark problems and two real-world trauma system design problems. Experimental results demonstrate that the variant using RF models as the surrogates is effective and efficient in solving data-driven constrained multiobjective combinatorial optimization problems.

Effective multi-objective discrete optimization of Truss-Z layouts using a GPU

Truss-Z (TZ) is an Extremely Modular System for creating skeletal free-form ramps and ramp networks. The TZ structures are comprised of four variations of two types of basic unit subjected to rotation. The two types of units are: R and L being a mirror reflection of each other. This paper presents a novel method based on image processing, evolutionary algorithm and intensive parallelization of multi-objective optimization of TZ layouts.The algorithm returns a sequence of modules. The result guarantees a TZ connection between two given points (regions) and minimizes the fitness function representing certain costs associated with setting up the TZ structure.The fitness function depends on the cost of TZ structure as well as the variety of costs related to the environment where the it is to be placed. E.g.: the earthworks, vegetation removal, obstacles avoidance, etc. There are no restrictions on the fitness function definition. It can depend on any variable which can be represented by a two-dimensional map of any property of the environment.The formulation of the presented method is suited for application of well-established image processing methods which efficiently evaluate candidate solutions on a GPU. As a result, the employed genetic algorithm efficiently probes the search space. The practical applicability of this approach is demonstrated with three case-studies:1simultaneous paving of a path with congruent units in a hilly environment with trees & bushes and finding the best location for a pier over an existing river;2constructing of a TZ connector spanning over a mountain valley with lakes (where supports can not be placed);3retrofitting of an existing railway station with a large wheelchair TZ ramp of over 10 m elevation while preserving trees and minimizing the earthworks.

A metaheuristic solution method for optimizing vessel fleet size and mix for maintenance operations at offshore wind farms under uncertainty

Maintenance operations at offshore wind farms are challenging due to the offshore element; maintenance technicians and spare parts need to be transported from an onshore port or offshore station to the individual wind farm components in need of maintenance. The vessel resources needed to support these maintenance tasks constitute a major part of the total maintenance costs, and hence up-keeping an optimal vessel fleet and corresponding deployment is essential to reduce cost-of-energy. This paper introduces a metaheuristic solution method to determine cost-efficient vessel fleets to support maintenance tasks at offshore wind farms under uncertainty. It considers weather conditions and failures leading to corrective maintenance tasks as stochastic parameters, and evaluates candidate solutions by a simulation program. The solution method has been incorporated in a decision support tool. Computational experiments, including comparison of results with an exact solution method, illustrate that the decision support tool can be used to provide near-optimal solutions within acceptable computational time.

Simulation-based multimodal optimization of decoy system design using an archived noise-tolerant genetic algorithm

The difficulty of warship decoy system design problem is twofold. First, we need to find not just one but as many optimal solutions as possible. Second, it demands a heavy computation to evaluate a candidate solution through a long series of underwater warfare simulations. The previous approach tried to reduce the amount of search by heuristically selecting a set of plausible starting points for the search by a simulated annealing algorithm. However, it shows only limited success and cannot easily scale up to larger problems. This paper proposes an efficient and easy-to-scale-up multimodal optimization algorithm named A-NTGA that is based on a genetic algorithm. A-NTGA quickly evaluates candidate solutions by conducting only a small number of simulations, but instead copes with these inaccurate or noisy fitness values by using a noisy optimization technique. To further enhance the efficiency of search by promoting the population diversity, A-NTGA is provided with an archive to which some good-looking solutions are migrated in order to prevent the population from being too crowded with similar solutions. Usually at the end of the search, many optimal solutions are retrieved from the archive as well as the population. The experimental results show that our method can find multiple optimal solutions more efficiently compared to other methods and can be easily scaled up to larger problems.

A New Representation in PSO for Discretization-Based Feature Selection.

In machine learning, discretization and feature selection (FS) are important techniques for preprocessing data to improve the performance of an algorithm on high-dimensional data. Since many FS methods require discrete data, a common practice is to apply discretization before FS. In addition, for the sake of efficiency, features are usually discretized individually (or univariate). This scheme works based on the assumption that each feature independently influences the task, which may not hold in cases where feature interactions exist. Therefore, univariate discretization may degrade the performance of the FS stage since information showing feature interactions may be lost during the discretization process. Initial results of our previous proposed method [evolve particle swarm optimization (EPSO)] showed that combining discretization and FS in a single stage using bare-bones particle swarm optimization (BBPSO) can lead to a better performance than applying them in two separate stages. In this paper, we propose a new method called potential particle swarm optimization (PPSO) which employs a new representation that can reduce the search space of the problem and a new fitness function to better evaluate candidate solutions to guide the search. The results on ten high-dimensional datasets show that PPSO select less than 5% of the number of features for all datasets. Compared with the two-stage approach which uses BBPSO for FS on the discretized data, PPSO achieves significantly higher accuracy on seven datasets. In addition, PPSO obtains better (or similar) classification performance than EPSO on eight datasets with a smaller number of selected features on six datasets. Furthermore, PPSO also outperforms the three compared (traditional) methods and performs similar to one method on most datasets in terms of both generalization ability and learning capacity.

Accounting for population exposure to vehicle-generated pollutants and environmental equity in the toll design problem

ABSTRACTThe emissions generated by motor vehicles remain a major source of air pollutants that affect public health and contribute to anthropogenic climate change. These negative externalities can be reduced, in part, with the implementation of environmentally oriented road pricing schemes, which can be designed using optimization-based approaches. In this paper, a toll design problem is proposed for determining toll locations and levels that minimize the expected human exposure to air pollutants and the related environmental inequalities, subject to constraints on pollutant concentration levels and implementation costs. The practical use of the proposed problem is hindered in most real-world applications by the computational costs associated with the evaluation of candidate solutions, as is common for network design problems. Furthermore, the problem cannot be expressed analytically given the multiple types of models (e.g., traffic assignment, emissions, air dispersion models) that would be required to evaluate a single design alternative. For these reasons, a derivative-free surrogate-based solution algorithm is proposed for mixed integer problems like the ones considered here. Numerical examples are used to illustrate possible applications of the proposed model and to test the performance of the surrogate-based algorithm. Relative to a joint simulated annealing-genetic algorithm heuristic and a genetic algorithm-based approach, the proposed algorithm found better solutions in fewer function evaluations.

Applications of Color in Design for Color-Deficient Users

Color vision deficiency is common, affecting one in every 12 men. Despite its prevalence, displays are seldom designed to accommodate color-vision-deficient (CVD) users, who confront daily challenges interpreting color in a broad range of applications, whether weather displays, informational graphics, road signs, or computer interfaces. In this article we discuss the prevalence of color deficiency, its effects, and the availability of tools that enable design teams to evaluate candidate solutions that meet the needs of CVD users, thereby ensuring universal accessibility.

Distributed Renewable Generation and Storage System Sizing Based on Smart Dispatch of Microgrids

This paper considers the contribution of independent owners (IOs) operating within microgrids (MGs) toward green power generation in deregulated energy markets. An optimization scheme is introduced for sizing distributed renewable generation (DRG) and a distributed energy storage system (DESS) based on a novel energy management system (EMS) that accounts for demand response (DR), DESS dispatch and performance degradation, dynamic pricing environments, power distribution loss and irregular renewable generation. The proposed EMS utilizes an iterative Newton-Raphson linear programming algorithm that schedules resources in order to minimize the objective function, to deal with the complicated nonlinear nature of the problem and to enable efficient long-term assessments. The EMS is used to evaluate candidate solutions that are generated by a genetic algorithm (GA) to determine the optimal combination of DRG and DESS. A case study for IEEE 34-bus distribution MG in Okinawa, Japan, is used for testing the algorithm and analyzing the potential for IO/MG investments and their strategies.

Finding exact hitting set solutions for systems biology applications using heterogeneous GPU clusters

The Systems Biology field presents several complex combinatorial problems that can be in part reduced to an instance of the Hitting Set Problem (HSP), which is NP-Hard. These reduced problems often come with a large amount of data that needs to be processed, such as gene expression profiles, resulting in prohibitive computational costs for finding the exact solutions. There are some proposals to obtain exact solutions for HSP, including an approach which uses GPUs. However, such an approach is not scalable for real input sizes (thousands of variables). We propose a novel algorithm for solving HSP instances with thousands of variables by using: (i) clause sorting, which enables the efficient discarding of non-solution candidates, (ii) parallel generation and evaluation of candidate solutions through the use of GPUs, and (iii) support for multiple GPUs. To permit the execution on heterogeneous clusters, we determine the minimum kernel size that does not incur extra overhead and distribute tasks among available GPUs on demand. Our experimental results show that the combination of these techniques results in a speedup of 118.5, when using eight NVIDIA Tesla K20c in comparison with a ten-core Intel Xeon E5-2690 processor. Consequently, our algorithm can enable the usage of exact algorithms for solving the Hitting Set problem and applying it to real world problems.

A Framework for Constrained Optimization Problems Based on a Modified Particle Swarm Optimization

This paper develops a particle swarm optimization (PSO) based framework for constrained optimization problems (COPs). Aiming at enhancing the performance of PSO, a modified PSO algorithm, named SASPSO 2011, is proposed by adding a newly developed self-adaptive strategy to the standard particle swarm optimization 2011 (SPSO 2011) algorithm. Since the convergence of PSO is of great importance and significantly influences the performance of PSO, this paper first theoretically investigates the convergence of SASPSO 2011. Then, a parameter selection principle guaranteeing the convergence of SASPSO 2011 is provided. Subsequently, a SASPSO 2011-based framework is established to solve COPs. Attempting to increase the diversity of solutions and decrease optimization difficulties, the adaptive relaxation method, which is combined with the feasibility-based rule, is applied to handle constraints of COPs and evaluate candidate solutions in the developed framework. Finally, the proposed method is verified through 4 benchmark test functions and 2 real-world engineering problems against six PSO variants and some well-known methods proposed in the literature. Simulation results confirm that the proposed method is highly competitive in terms of the solution quality and can be considered as a vital alternative to solve COPs.

Smart City Wireless Connectivity Considerations and Cost Analysis: Lessons Learnt From Smart Water Case Studies

According to estimates from the United Nations, while the world population is likely to reach 9 billion by 2050 (from around 7 billion today), the water resources available to cater to this population are likely to remain similar to what they are today. Therefore, there is a growing concern to reduce water wastage and improve the efficiency of water distribution systems, in particular, urban water systems given the proliferation in both the number of cities that are springing up and the number of people moving to live in cities. Motivated by this, the European Commission under the aegis of the Framework Program (FP7) funded the Information and Communications Technology (ICT) Solutions for the Efficient Water Resources Management project. This Smart Water project is aimed at investigating the role of ICT in monitoring and efficiently managing urban water systems, in particular, exploring the deployment of sensors, communication technologies, and associated decision support systems in utility providers water networks geared toward addressing problems, such as leakage management, demand management, asset management, and so on. This article elaborates on the wireless connectivity considerations, proposes a total cost of ownership framework for evaluating candidate solutions, and highlights experiences from Smart Water case studies involving two utilities in Europe.

Robotic Design Choice Overview Using Co-Simulation and Design Space Exploration

Rapid robotic system development has created a demand for multi-disciplinary methods and tools to explore and compare design alternatives. In this paper, we present a collaborative modeling technique that combines discrete-event models of controller software with continuous-time models of physical robot components. The proposed co-modeling method utilizes the Vienna development method (VDM) and MATLAB for discrete-event modeling and 20-sim for continuous-time modeling. The model-based development of a mobile robot mink feeding system is used to illustrate the collaborative modeling method. Simulations are used to evaluate the robot model output response in relation to operational demands. An example of a load-carrying challenge in relation to the feeding robot is presented, and a design space is defined with candidate solutions in both the mechanical and software domains. Simulation results are analyzed using design space exploration (DSE), which evaluates candidate solutions in relation to preselected optimization criteria. The result of the analysis provides developers with an overview of the impacts of each candidate solution in the chosen design space. Based on this overview of solution impacts, the developers can select viable candidates for deployment and testing with the actual robot.