Multidimensional Search Space Research Articles

An artificial intelligence-based optimization framework for the optimal composition and thermomechanical processing schedule for specialized micro-alloyed multiphase steels

An artificial intelligence-based heuristic approach is presented to optimize the chemical composition and the thermomechanical processing schedule to obtain specialized micro-alloyed multiphase steels with desired mechanical properties, at minimal manufacturing cost. The optimization framework uses a modified form of genetic algorithm, called the micro-genetic algorithm (μGA), that uses a penalty-based cost function formulation operating on a multi-dimensional search space spanning 15 alloying elements, an average cooling temperature, an austenitizing temperature and eight time–temperature points from the cooling profiles of multiphase steels. With superior search speed and convergence rates to the traditional genetic algorithm, μGA uses a neural network-based reduced-order model to predict hardness. Additional correlation equations are used to determine the corresponding tensile strength and elongation. Microstructural analysis was performed using neurocomputing techniques to further validate the accuracy of the algorithm. The entire computational framework was validated using data from the literature, establishing its utility in steel design.

Dynamical Sphere Regrouping Particle Swarm Optimization: A Proposed Algorithm for Dealing with PSO Premature Convergence in Large-Scale Global Optimization

Optimizing large-scale numerical problems is a significant challenge with numerous real-world applications. The optimization process is complex due to the multi-dimensional search spaces and possesses several locally optimal regions. In response to this issue, various metaheuristic algorithms and variations have been developed, including evolutionary and swarm intelligence algorithms and hybrids of different artificial intelligence techniques. Previous studies have shown that swarm intelligence algorithms like PSO perform poorly in high-dimensional spaces, even with algorithms focused on reducing the search space. However, we propose a modified version of the PSO algorithm called Dynamical Sphere Regrouping PSO (DSRegPSO) to avoid stagnation in local optimal regions. DSRegPSO is based on the PSO algorithm and modifies inertial behavior with a regrouping dynamical sphere mechanism and a momentum conservation physics effect. These behaviors maintain the swarm’s diversity and regulate the exploration and exploitation of the search space while avoiding stagnation in optimal local regions. The DSRegPSO mechanisms mimic the behavior of birds, moving particles similar to birds when they look for a new food source. Additionally, the momentum conservation effect mimics how birds react to collisions with the boundaries in their search space or when they are looking for food. We evaluated DSRegPSO by testing 15 optimizing functions with up to 1000 dimensions of the CEC’13 benchmark, a standard for evaluating Large-Scale Global Optimization used in Congress on Evolutionary Computation, and several journals. Our proposal improves the behavior of all variants of PSO registered in the toolkit of comparison for CEC’13 and obtains the best result in the non-separable functions against all the algorithms.

Physics-informed deep learning for simultaneous surrogate modeling and PDE-constrained optimization of an airfoil geometry

We use a physics-informed neural network (PINN) to simultaneously model and optimize the flow around an airfoil to maximize its lift to drag ratio. The parameters of the airfoil shape are provided as inputs to the PINN and the multidimensional search space of shape parameters is populated with collocation points to ensure that the Navier–Stokes equations are approximately satisfied throughout. We use the fact that the PINN is automatically differentiable to calculate gradients of the lift-to-drag ratio with respect to the airfoil shape parameters. This allows us to optimize with the L-BFGS gradient-based algorithm, which is more efficient than non-gradient-based algorithms, without deriving an adjoint code. We train the PINN with adaptive sampling of collocation points, such that the accuracy of the solution improves as the solution approaches the optimal point. We demonstrate this on two examples: one that optimizes a single parameter, and another that optimizes eleven parameters. The method is successful and, by comparison with conventional CFD, we find that the velocity and pressure fields have small pointwise errors and that the method converges to optimal parameters. We find that different PINNs converge to slightly different parameters, reflecting the fact that there are many closely-spaced local minima when using stochastic gradient descent. This method can be applied relatively easily to other optimization problems and avoids the difficult process of writing adjoint codes. It is, however, more computationally expensive than adjoint-based optimization. As knowledge about training PINNs improves and hardware dedicated to neural networks becomes faster, this method of simultaneous training and optimization with PINNs could become easier and faster than using adjoint codes.

Spheres of interest: Space and social cognition in Phola deixis

Abstract After a period of space-centred description of demonstratives, recent research has highlighted the role of attention, psychological proximity and shared knowledge in determining deictic choice. While convincing evidence has been presented that mental states may define deictic reference (e.g. in Turkish, Jahai or Kogi), there is also neuroscientific data suggesting that spatial cognition is often drawn upon in the process and that spatial and attentional perspectives may interact with each other. Pragmatic analysis of deictic usage in some languages (e.g. Yucatec or Lao) suggests that demonstrative systems may respond to multidimensional search spaces that include not only spatial but also embodied, perceptual and social access to referents. On the basis of observational data from Phola, a Tibeto-Burman language of Southwest China, the present article contributes to these research endeavours by explicitly exploring how speaker and addressee demonstratives may independently respond to both spatial and sociocognitive modes of access to a referent. Advancing the notion of spheres of interest as a descriptive heuristics to capture this fluidity, it is shown how deictic choice not only passively reflects aspects of context but also actively projects intersubjective appraisals and expectations onto material and social reality.

Particle Swarm Optimization of Resonant Sonic Crystals Noise Barriers

The study of technological materials made by meticulously arranging acoustic elements has received a lot of attention over the past three decades with the goal of generating improved acoustic properties, often going beyond the behavior of materials found in nature. These are frequently referred to as acoustic metamaterials, and because of the way wave propagation phenomena is managed, they exhibit unusual properties. Improvements in noise mitigation techniques of acoustic systems based on metamaterials principles have been made effectively and precisely using combined or hybrid numerical methods and improved numerical formulations. These noise mitigation properties should be optimized by modifying topology and inner elements properties, which requires a huge search space. This work focuses on metamaterials called sonic crystals with Helmholtz resonators, and its innovative insulation properties as noise barriers are optimized with the Particle Swarm Optimization (PSO). This evolutionary algorithm provides a mono objective solution for the multi-dimensional search space inspired in animal behavior looking for food and communicating between each other. In order to assess the results obtained by the proposed approach, the presented PSO algorithm is compared with a Genetic Algorithm (GA): the results show that the PSO algorithm provides a better solution that pursues the objective of satisfying the acoustic comfort without exceeding the imposed practical constraints.

Optimal Composition of Li Argyrodite with Harmonious Conductivity and Chemical/Electrochemical Stability: Fine‐Tuned Via Tandem Particle Swarm Optimization (Adv. Sci. 28/2022)

Tandem Particle Swarm Optimization In article number 2201648 by Kee-Sun Sohn, Myoungho Pyo, and co-workers, a two-step optimization algorithm (tandem particle swarm optimization) is applied to rapidly identify the optimal argyrodite with the highest attainable σion in a multidimensional search space involving a huge number of compositions and processing conditions. Each particle (bee) is searching for the best candidate by exchanging information through the social behavior. The identified argyrodite exhibits enhanced σion and improved stability against moisture and high voltage and, as an all-solid-state battery, would deliver higher capacities, particularly during fast charge–discharge cycles.

Energy-Efficient Multidimensional Trajectory of UAV-Aided IoT Networks With Reinforcement Learning

This article proposes a multidimensional search space (or directional space) with more degrees-of-freedom (DOFs) to increase the energy efficiency of limited-battery-powered unmanned aerial vehicle (UAV) in the Internet of Things (IoT) data collection scenario. In this article, the UAV navigates from the initial to the goal point while collecting data from IoT sensors on the ground. Owing to the limited battery power of UAVs, an optimized trajectory is a crucial practical problem. Based on the available directional space, the direction of the UAV related to the navigation trajectory is optimized using reinforcement learning (RL). The objective of RL is to maximize the energy efficiency of the UAV as a long-term reward by selecting the optimal direction. Moreover, a practical energy consumption model and environment are presented in this article. Simulation results verified that the proposed multidimensional trajectory for UAV achieves higher energy efficiency compared with benchmark models.

Multidimensional resilience decision-making for complex and substructured systems

Complex systems, such as infrastructure networks, industrial plants and jet engines, are of paramount importance to modern societies. However, these systems are subject to a variety of different threats. Novel research focuses not only on monitoring and improving the robustness and reliability of systems, but also on their recoverability from adverse events. The concept of resilience encompasses precisely these aspects. However, efficient resilience analysis for the modern systems of our societies is becoming more and more challenging. Due to their increasing complexity, system components frequently exhibit significant complexity of their own, requiring them to be modeled as systems, i.e., subsystems. Therefore, efficient resilience analysis approaches are needed to address this emerging challenge.This work presents an efficient resilience decision-making procedure for complex and substructured systems. A novel methodology is derived by bringing together two methods from the fields of reliability analysis and modern resilience assessment. A resilience decision-making framework and the concept of survival signature are extended and merged, providing an efficient approach for quantifying the resilience of complex, large and substructured systems subject to monetary restrictions. The new approach combines both of the advantageous characteristics of its two original components: A direct comparison between various resilience-enhancing options from a multidimensional search space, leading to an optimal trade-off with respect to the system resilience and a significant reduction of the computational effort due to the separation property of the survival signature, once a subsystem structure has been computed, any possible characterization of the probabilistic part can be validated with no need to recompute the structure.The developed methods are applied to the functional model of a multistage high-speed axial compressor and two substructured systems of increasing complexity, providing accurate results and demonstrating efficiency and general applicability.

Iterative sequential robust optimization of quantity and location of wells in field development under subsurface, operational and economic uncertainty

Determination of optimal quantity and location of wells is a crucial step in any field development project. Multiple challenges exist in the real-world projects, such as gigantic multidimensional search space and complex parameterization of the problem. To address these challenges, we present an iterative sequential optimization framework that consists of multiple stages for optimizing number and location of producers and injectors. Once the initial guesses have been established, the first optimization stage begins in order to determine the optimal number of producers and injectors. Next, a two-stage process is used to determine the best location for each well. By optimizing the number and location of wells, well type is optimized implicitly and does not need to be included in the optimization process as an individual decision variable. To consider the dependency and coupling between parameters, the entire optimization process is repeated iteratively up to a stopping point. The proposed approach is highly flexible and can be extended to include other potential decision variables of field development projects. We test the proposed methodology on UNISIM–II–D, a realistic benchmark case that represents pre-salt reservoirs in deep offshore Brazil, which are light-oil fractured carbonates with large volumes of CO2-rich associated gas. In our application, all producers and injectors are equipped with internal control valves (ICVs), all produced gas is reinjected into the reservoir, and water-alternating-gas with CO2 (WAG-CO2) is used to increase oil recovery and ensure that the produced CO2, as a harmful greenhouse gas, can be safely discarded underground. We use the iterative discrete Latin hypercube (IDLHC) optimization algorithm. To achieve a robust decision, we consider subsurface, operational and economic uncertainties by considering nine representative reservoir models and three economic scenarios. Based on the results, the proposed framework can provide an appropriate solution containing optimal quantity of wells, their locations and types in realistic cases with large search spaces and complex parameterizations.

Improved Fitness-Dependent Optimizer for Solving Economic Load Dispatch Problem.

Economic load dispatch depicts a fundamental role in the operation of power systems, as it decreases the environmental load, minimizes the operating cost, and preserves energy resources. The optimal solution to economic load dispatch problems and various constraints can be obtained by evolving several evolutionary and swarm-based algorithms. The major drawback to swarm-based algorithms is premature convergence towards an optimal solution. Fitness-dependent optimizer is a novel optimization algorithm stimulated by the decision-making and reproductive process of bee swarming. Fitness-dependent optimizer (FDO) examines the search spaces based on the searching approach of particle swarm optimization. To calculate the pace, the fitness function is utilized to generate weights that direct the search agents in the phases of exploitation and exploration. In this research, the authors have used a fitness-dependent optimizer to solve the economic load dispatch problem by reducing fuel cost, emission allocation, and transmission loss. Moreover, the authors have enhanced a novel variant of the fitness-dependent optimizer, which incorporates novel population initialization techniques and dynamically employed sine maps to select the weight factor for the fitness-dependent optimizer. The enhanced population initialization approach incorporates a quasi-random Sabol sequence to generate the initial solution in the multidimensional search space. A standard 24-unit system is employed for experimental evaluation with different power demands. The empirical results obtained using the enhanced variant of the fitness-dependent optimizer demonstrate superior performance in terms of low transmission loss, low fuel cost, and low emission allocation compared to the conventional fitness-dependent optimizer. The experimental study obtained 7.94E−12, the lowest transmission loss using the enhanced fitness-dependent optimizer. Correspondingly, various standard estimations are used to prove the stability of the fitness-dependent optimizer in phases of exploitation and exploration.

Iterative Dynamic Programming Strategy for Electric Vehicle Battery Thermal Management Optimization

AbstractLithium‐ion batteries are extensively used in electric vehicles because of their superior performance in the general environment of global energy shortage. However, lithium‐ion batteries generate large amounts of heat as the battery packs are charged and discharged. If this accumulated heat is not dissipated timely, then the power supply system will fail, resulting in safety accidents. Therefore, a high‐efficiency and energy‐saving optimization strategy is required for both ensuring operating temperature of the battery and saving energy. Herein, a lumped thermal model is established on the basis of the heat generation characteristics of batteries and Newton's law of cooling, and its exactitude is proven by comparison with a liquid cooling system established using AMESim. In addition, to reduce computation time and enhance the real‐time capability, an iterative dynamic programming (IDP) method is proposed to determine the optimal values iteratively in a multidimensional search space, in view of the high nonlinearity and time‐variability of a battery thermal management system as well as the complexity of dynamic programming (DP). Thus, the results from co‐simulations performed in Matlab and AMESim verify that the proposed strategy has high cooling efficiency and can save considerable energy for cooling system compared with PID and DP.

Optimal design of second generation current conveyor using craziness-based particle swarm optimisation

In this paper, a population-based meta-heuristic search algorithm called craziness-based particle swarm optimisation (CRPSO) has been employed for the optimal design of positive second-generation current conveyor based on a trans-linear loop (CCII+). CRPSO accepts several random variables for encompassing improved and quicker search and utilisation in multidimensional search space. The main goal of this work is to optimally size the CCII+ building block to attain the suitable aspect ratios of the MOS transistors. To enhance the performance of a CCII+ design, it needs to increase the higher cut-off frequency of the current signal (fCI) as well as to decrease the input X-port parasitic resistance (RX). Accordingly, the optimisation problem is developed as a bi-objective problem of minimisation to obtain the least value of RX and the superior value of fCI. The current conveyor is simulated and validated using UMC 180 nm CMOS technology with a power supply of ±2.5 V in Cadence Virtuoso XL Design Environment.

Traçado automático de adutoras de abastecimento de água

ABSTRACT This paper presents an algorithm capable of calculating the optimal route for pipelines that traverse terrain with or without additional displacement constraints to the difference in ground elevations and head losses to be overcome by pumping. The criterion used to determine the pipeline diameter and calculate feasible routes was minimizing the annual cost, which results from the sum of (i) annualized costs related to the acquisition of the pipeline and (ii) costs of payment for electric power to operate the system throughout its useful life. The geometry of the shortest routes in the multidimensional search space is calculated by the proposed algorithm, called BAGDA (Busca pelo Ajuste Geométrico da Despesa Anual, in Portuguese, or Search for the Annual Cost of Geometric Tuning, in English), thus obtaining the optimal combination of length and manometric head of the pipeline. The performed applications show the efficiency of the algorithm in providing subjectivity-free routes sensitive to the most important variables considered in the design of piping systems.

Multi-variable Bayesian optimization for a new composition with high Na+ conductivity in the Na3PS4 family

A Bayesian optimization (BO) algorithm is employed to discover new Na3PS4 compositions with a high σion in a multi-dimensional co-doped search space.

Autotuning of Exascale Applications With Anomalies Detection.

The execution of complex distributed applications in exascale systems faces many challenges, as it involves empirical evaluation of countless code variations and application runtime parameters over a heterogeneous set of resources. To mitigate these challenges, the research field of autotuning has gained momentum. The autotuning automates identifying the most desirable application implementation in terms of code variations and runtime parameters. However, the complexity and size of the exascale systems make the autotuning process very difficult, especially considering the number of parameter variations that have to be identified. Therefore, we introduce a novel approach for autotuning exascale applications based on a genetic multi-objective optimization algorithm integrated within the ASPIDE exascale computing framework. The approach considers multi-dimensional search space with support for pluggable objective functions, including execution time and energy requirements. Furthermore, the autotuner employs a machine learning-based event detection approach to detect events and anomalies during application execution, such as hardware failures or communication bottlenecks.

Solving Multiphysics, Multiparameter, Multimodal Inverse Problems: An Application to NMR Relaxation in Porous Media

A general and robust Bayesian optimization framework for the extraction of intrinsic physical properties from an integration of pore-scale forward modeling and experimental measurements of macroscopic system responses is developed. The efficiency of the scheme, which utilizes Gaussian process regression, enables the simultaneous extraction of multiple intrinsic physical properties with a minimal number of function evaluations. Here it is applied to nuclear magnetic resonance (NMR) relaxation responses, paving the way for inverse problem approaches to digital rock physics given its general nature. NMR relaxation responses of fluids in porous media may be described by sums of multiexponential decays resulting in a relaxation time distribution. The shape of this distribution is dependent on intrinsic physical system properties, but also effects like diffusion coupling between different relaxation regimes in heterogeneous porous materials. Forward models based on high-resolution images are employed to naturally incorporate structural heterogeneity and diffusive motion without limiting assumptions. Extracting the required multiple intrinsic parameters of the system poses an ill-conditioned multiphysics multiparameter inverse problem where multiple scales are covered by the underlying microstructure. Exploration of the multidimensional search space given an expensive cost function makes an efficient solution strategy mandatory. We propose a workflow to match experimental measurements with simulations via Bayesian optimization, with special attention paid to the multimodal nature of the topography of the objective function using solution space partitioning. A multimodal search strategy using state-of-the-art evolutionary algorithms and gradient-based optimization algorithms guarantees that the multimodal nature is captured. The workflow is demonstrated on ${T}_{2}$ relaxation responses of Bentheimer sandstone, extracting three physical parameters simultaneously: the surface relaxivity of quartz grains, the effective transverse relaxation time, and the effective diffusion coefficient in clay regions. Multiple mathematically sound and physically plausible solutions corresponding to global minimum and multiple local minima of the objective function are identified within a limited number of function evaluations. Importantly, the shape of the experimental ${T}_{2}$ distribution is recovered almost perfectly, enabling the use of classical interpretation techniques and local analysis of responses based on numerical simulation.

A Novel Approach to Model and Optimize Qualities of Castings Produced by Differential Pressure Casting Process

The present study attempts to promote the quality of differential pressure casting component by two-stage optimization, orthogonal virtual casting and BP neural network. In the first stage, the critical parameters determined by the numerical model of casting procedure indicate that the qualities of castings, including casting solidification time, secondary dendrite spacing and porosity, are highly affected by the die temperature, pouring temperature and cooling medium temperature. In the second stage, the input–output relationship developed by utilizing BP neural network is found to be statistically adequate and yielded better prediction accuracy. Artificial fish swarm algorithm (AFSA) performs multi-directional search in multi-dimensional space simultaneously and performs desirability function approach. The results of nonlinear neural network-based models and the performance of artificial fish swarm algorithm optimization technique are summarized.

Automatic Shimming Method Using Compensation of Magnetic Susceptibilities and Adaptive Simplex for Low-Field NMR

During the past several decades, inexpensive compact nuclear magnetic resonance (NMR) instruments have been widely used for on-site detections of chemical identity and sample analyses in industrial applications. In general, via shim coils and control strategies, automatic search shimming methods are capable of improving the magnetic field uniformity so that optimized signals can be acquired and particularly suited for rapid detections in compact NMR instruments. However, because these methods inherently endure the inefficiency in multidimensional shims and the sensitivity of starting points, it is time-consuming to obtain acceptable results in a field infected with distortions and strays. Here, two remedies that improve the shimming speed for the search shimming method are proposed. First, analyzing and compensating for magnetic susceptibility effects, we relocate starting points closely to optima and reduce the complexity of global exploration. Second, modifying the simplex movement with adaptive scale factors and gradient-like centroid, we have improved the convergence of the simplex shimming algorithm in a multidimensional search space. Via theoretical analyses and lab experiments, we find that the proposed shimming method outperforms regular shimming methods in terms of handling complex magnetic fields and reducing shimming times and hope that our study can help further improve NMR shimming methods in the future.

Emended salp swarm algorithm for multiobjective electric power dispatch problem

This paper presents an emended salp swarm algorithm (ESSA) which is basically the extension of the basic salp swarm algorithm (SSA) to solve multiobjective electric power load dispatch problem. The main inspiration behind ESSA is the swarming behavior and reproduction cycle of salps. Salps, in the chain, move across a multi-dimensional search space to aim for the food source (global solution). Owing to the searching behavior of SSA that makes the algorithm prone to premature convergence, the solitary and colonial reproduction phase of salp has been introduced to improve the convergence behavior along with their swarming behavior The multiobjective optimization problem is firstly converted into scalar objective exploiting fuzzy set theory and the conflicting nature of objectives is resolved by cardinal priority ranking. The variable elimination method with exterior penalty is used to handle the physical and operational constraints of generating units. The validation of the proposed ESSA has been examined on the standard benchmark functions and seven EcLD test systems including both scalar and multiple objectives. The statistical analysis based on Wilcoxon sign rank test, supports that results achieved by the algorithm are superior to the other competing algorithms. So, the proposed algorithm (ESSA) is found a promising algorithm.

Bass Re-visited: Quantifying Multi-Finality

As forecasting the diffusion of new technologies results in unreliable predictions when only few trend data are available, scholars and practitioners alike tend to favor foresight efforts of a more qualitative nature to gauge foreseeable futures. We argue and demonstrate that it is feasible to reconcile quantitative forecasting models with (qualitative) scenario development efforts by advancing a set of heuristics that build on known diffusion curves. Rather than focusing on the ultimate set of parameter estimates, we explore an exhaustive range of combinations signaling potential outcomes. We demonstrate this approach for Battery Electric Vehicles in Europe and the US. A multi-dimensional search space is developed consisting of a wide range of values for each parameter implied. Resulting outcomes are assessed in terms of how well they explain current observations (2011-2017) by means of a loss function. The loss function allows us to evaluate the presence of different end states (multi-finality), their frequency of occurrence, and the implied time frames. The insights obtained inform foresight exercises in a number of distinctive ways. First, by incorporating multi-finality in the design of the heuristics, it becomes feasible to qualify multiple pathways ex ante and simultaneously. This allows us to assess the likelihood of - and the time horizon implied by - different scenarios. Second, allowing for multiple pathways and end states directs our attention to the antecedents required for different scenarios to unfold, which can inspire back-casting efforts of a more qualitative nature. Finally, applying the grid logic on more fine-grained levels of analysis (e.g. countries, states, or regions) helps us assess the (differential) impact of policies oriented towards stimulating diffusion. As such, quantitative heuristics become a valuable complement to scenario development exercises, rather than an inferior or even neglected substitute.