Computational Time Requirements Research Articles

Multicounty fuel constraint‐based dynamic economic dispatch problem using horse herd optimization

AbstractDue to the diminution of fossil fuels, the profitable usage of fuel for the generation of power is reducing. It is an extremely vital concern for power generation companies. The fossil fuel‐based power plant must operate within its fuel limits and contractual restrictions. Therefore, economic optimization is required to optimize the fuel cost for generation under fuel constraints. To address this issue, in this work, a horse herd optimization algorithm (HOA) is developed to solve multicounty fuel‐constrained dynamic economic dispatch with demand‐side management incorporating wind turbine generators, solar photovoltaic plants, and pumped hydro storage plants. The efficiency of the suggested HOA algorithm has been revealed on a selected test system. The HOA technique shows better convergence as compared with the other methods. Moreover, the computation time requirement of the HOA technique is also lower than that of other compared methods. From numerical results, it is observed that the consumption of fuel is sufficiently high if constraints are incorporated into the problem. Simulation results are matched up to those obtained from hierarchical particle swarm optimization with time‐varying acceleration coefficients, fast convergence evolutionary programming, and differential evolution. After comparison, it is observed that the recommended HOA offers a better‐quality solution.

A Multi-Objective Tri-Level Algorithm for Hub-and-Spoke Network in Short Sea Shipping Transportation

Hub-and-Spoke (H&S) network modeling is a form of transport topology optimization in which network joins are connected through intermediate hub nodes. The Short Sea Shipping (SSS) problem aims to efficiently disperse passenger flows involving multiple vessel routes and intermediary hubs through which passengers are transferred to their final destination. The problem contains elements of the Hub-and-Spoke and Travelling Salesman, with different levels of passenger flows among islands, making it more demanding than the typical H&S one, as the hub selection within nodes and the shortest routes among islands are internal optimization goals. This work introduces a multi-objective tri-level optimization algorithm for the General Network of Short Sea Shipping (GNSSS) problem to reduce travel distances and transportation costs while improving travel quality and user satisfaction, mainly by minimizing passenger hours spent on board. The analysis is performed at three levels of decisions: (a) the hub node assignment, (b) the island-to-line assignment, and (c) the island service sequence within each line. Due to the magnitude and complexity of the problem, a genetic algorithm is employed for the implementation. The algorithm performance has been tested and evaluated through several real and simulated case studies of different sizes and operational scenarios. The results indicate that the algorithm provides rational solutions in accordance with the desired sub-objectives. The multi-objective consideration leads to solutions that are quite scattered in the solution space, indicating the necessity of employing formal optimization methods. Typical Pareto diagrams present non-dominated solutions varying at a range of 30 percent in terms of the total distance traveled and more than 50 percent in relation to the cumulative passenger hours. Evaluation results further indicate satisfactory algorithm performance in terms of result stability (repeatability) and computational time requirements. In conclusion, the work provides a tool for assisting network operation and transport planning decisions by shipping companies in the directions of cost reduction and traveler service upgrade. In addition, the model can be adapted to other applications in transportation and in the supply chain.

FEM Modeling of Electro-Acoustic Nonlinearities in Surface Acoustic Wave Devices: A Methodological Review

In the framework of electro-elasticity theory and the finite element method (FEM), a model is set up for the computation of quantities in surface acoustic wave (SAW) devices accounting for nonlinear effects. These include second-order and third-order intermodulations, second and third harmonic generation and the influence of electro-acoustic nonlinearity on the frequency characteristics of SAW resonators. The model is based on perturbation theory, and requires input material constants, e.g., the elastic moduli up to fourth order for all materials involved. The model is two-dimensional, corresponding to an infinite aperture, but all three Cartesian components of the displacement and electrical fields are accounted for. The first version of the model pertains to an infinite periodic arrangement of electrodes. It is subsequently generalized to systems with a finite number of electrodes. For the latter version, a recursive algorithm is presented which is related to the cascading scheme of Plessky and Koskela and strongly reduces computation time and memory requirements. The model is applied to TC-SAW systems with copper electrodes buried in an oxide film on a LiNbO3 substrate. Results of computations are presented for the electrical current due to third-order intermodulations and the displacement field associated with the second harmonic and second-order intermodulations, generated by monochromatic input tones. The scope of this review is limited to methodological aspects with the goal to enable calculations of nonlinear quantities in SAW devices on inexpensive and easily accessible computing platforms.

Machine learning based lineage tree reconstruction improved with knowledge of higher level relationships between cells and genomic barcodes.

Tracking cells as they divide and progress through differentiation is a fundamental step in understanding many biological processes, such as the development of organisms and progression of diseases. In this study, we investigate a machine learning approach to reconstruct lineage trees in experimental systems based on mutating synthetic genomic barcodes. We refine previously proposed methodology by embedding information of higher level relationships between cells and single-cell barcode values into a feature space. We test performance of the algorithm on shallow trees (up to 100 cells) and deep trees (up to 10 000 cells). Our proposed algorithm can improve tree reconstruction accuracy in comparison to reconstructions based on a maximum parsimony method, but this comes at a higher computational time requirement.

Analyzing Big EHR Data—Optimal Cox Regression Subsampling Procedure with Rare Events

Massive sized survival datasets become increasingly prevalent with the development of the healthcare industry, and pose computational challenges unprecedented in traditional survival analysis use cases. In this work we analyze the UK-biobank colorectal cancer data with genetic and environmental risk factors, including a time-dependent coefficient, which transforms the dataset into “pseudo-observation” form, thus, critically inflating its size. A popular way for coping with massive datasets is downsampling them, such that the computational resources can be afforded by the researcher. Cox regression has remained one of the most popular statistical models for the analysis of survival data to-date. This work addresses the settings of right censored and possibly left-truncated data with rare events, such that the observed failure times constitute only a small portion of the overall sample. We propose Cox regression subsampling-based estimators that approximate their full-data partial-likelihood-based counterparts, by assigning optimal sampling probabilities to censored observations, and including all observed failures in the analysis. The suggested methodology is applied on the UK-biobank for building a colorectal cancer risk-prediction model, while reducing the computation time and memory requirements. Asymptotic properties of the proposed estimators are established under suitable regularity conditions, and simulation studies are carried out to evaluate their finite sample performance. Supplementary materials for this article are available online.

Optimal and efficient planning of charging stations for electric vehicles in urban areas: formulation, complexity and solutions

The deployment of charging infrastructure for electric vehicles is crucial in extending their range. Many studies on the charging infrastructure deployment adopt the Mixed Integer Linear Programming (MILP) method to optimize various objectives. However, as the number of integer variables and constraints increases, the computational time and memory requirements of MILP models increase exponentially. This makes it impractical to use MILP models to solve large-scale optimization problems. In this paper, we formulate and prove that the Planning of Electric Vehicle Charging Stations (PEVCS) is an NP-complete combinatorial optimization problem. We also prove that PEVCS has a significant effect, that is, submodularity. Additionally, we propose two efficient methods that use submodularity to improve the conventional methodology for PEVCS. Furthermore, we provide a provable guarantee for the performance of our proposed methods. Our experimental results demonstrate the efficiency and effectiveness of these methods on small-scale and large-scale datasets, especially in realistic large-scale situations.

Short-Term Wind Power Prediction Based on a Hybrid Markov-Based PSO-BP Neural Network

Wind power prediction is an important research topic in the wind power industry and many prediction algorithms have recently been studied for the sake of achieving the goal of improving the accuracy of short-term forecasting in an effective way. To tackle the issue of generating a huge transition matrix in the traditional Markov model, this paper introduces a real-time forecasting method that reduces the required calculation time and memory space without compromising the prediction accuracy of the original model. This method is capable of obtaining the state probability interval distribution for the next moment through real-time calculation while preserving the accuracy of the original model. Furthermore, the proposed Markov-based Back Propagation (BP) neural network was optimized using the Particle Swarm Optimization (PSO) algorithm in order to effectively improve the prediction approach with an improved PSO-BP neural network. Compared with traditional methods, the computing time of our improved algorithm increases linearly, instead of growing exponentially. Additionally, the optimized Markov-based PSO-BP neural network produced a better predictive effect. We observed that the Mean Absolute Percentage Error (MAPE) and Mean Absolute Error (MAE) of the prediction model were 12.7% and 179.26, respectively; compared with the existing methods, this model generates more accurate prediction results.

Design approach for plasmonic filter using Q factor analysis

An efficient design method for plasmonic devices composed of a periodic slit array based on extraordinary optical transmission is proposed using Q factor analysis. The proposed method significantly reduces the required computation time for cavity resonance in slits by calculating the Q factor at a single frequency that is an eigenfrequency of the wave equation instead of computing the entire transmissions. The computation time is reduced by 20.60 times, and the device performance is increased by 2.43 times through a design based on Q factor analysis. The figure of merit for the overall performance is defined as a product of the transmission peak value and the Q factor. The design variables are the slit width, slit height, and period of slit array. The optical field distributions are calculated numerically using the finite element method, and the Q factor is obtained by solving an eigenfrequency for the wave equation for dispersive media.

Transforming a Computational Model from a Research Tool to a Software Product: A Case Study from Arc Welding Research

Arc welding is a thermal plasma process widely used to join metals. An arc welding model that couples fluid dynamic and electromagnetic equations was initially developed as a research tool. Subsequently, it was applied to improve and optimise industrial implementations of arc welding. The model includes the arc plasma, the electrode, and the workpiece in the computational domain. It incorporates several features to ensure numerical accuracy and reduce computation time and memory requirements. The arc welding code has been refactored into commercial-grade Windows software, ArcWeld, to address the needs of industrial customers. The methods used to develop ArcWeld and its extension to new arc welding regimes, which used the Workspace workflow platform, are presented. The transformation of the model to an integrated software application means that non-experts can now run the code after only elementary training. The user can easily visualise the results, improving the ability to analyse and generate insights into the arc welding process being modelled. These changes mean that scientific progress is accelerated, and that the software can be used in industry and assist welders’ training. The methods used are transferrable to many other research codes.

Decentralized Distributionally Robust Coordination Between Distribution System and Charging Station Operators in Unbalanced Distribution Systems

Coordination between the distribution system operator (DSO) and charging station operators (CSOs) is crucial for ensuring stable and economical operation of distribution grids and electric vehicle charging stations (EVCSs). We propose a computationally efficient and privacy-preserving decentralized DSO-CSO coordination framework for robust operation of distribution grids and EVCSs under uncertainties in active unbalanced distribution systems. A centralized model predictive control (MPC) is developed based on distributionally robust optimization (DRO), which characterizes the uncertainties in photovoltaic generation outputs, loads, and electricity prices using the Wasserstein metric. Two scalable optimization approaches are presented to address the high computation time requirement of the proposed centralized DRO-based MPC method through the reduction of constraints: i) a robust optimization method including a distributionally robust bound on the uncertainties, and ii) a bi-level optimization method that determines critical buses associated with voltage violations. Furthermore, to preserve data privacy of electric vehicles, the centralized DRO-based MPC problem is transformed through an alternating direction multiplier method into a decentralized DRO-based MPC problem comprising the DSO and CSO problems. The efficiency and scalability of the proposed DSO-CSO coordination framework are analyzed using the IEEE 37-bus and IEEE 123-bus systems.

Lidar Pose Tracking of a Tumbling Spacecraft Using the Smoothed Normal Distribution Transform

Lidar sensors enable precise pose estimation of an uncooperative spacecraft in close range. In this context, the iterative closest point (ICP) is usually employed as a tracking method. However, when the size of the point clouds increases, the required computation time of the ICP can become a limiting factor. The normal distribution transform (NDT) is an alternative algorithm which can be more efficient than the ICP, but suffers from robustness issues. In addition, lidar sensors are also subject to motion blur effects when tracking a spacecraft tumbling with a high angular velocity, leading to a loss of precision in the relative pose estimation. This work introduces a smoothed formulation of the NDT to improve the algorithm’s robustness while maintaining its efficiency. Additionally, two strategies are investigated to mitigate the effects of motion blur. The first consists in un-distorting the point cloud, while the second is a continuous-time formulation of the NDT. Hardware-in-the-loop tests at the European Proximity Operations Simulator demonstrate the capability of the proposed methods to precisely track an uncooperative spacecraft under realistic conditions within tens of milliseconds, even when the spacecraft tumbles with a significant angular rate.

COMPARISON OF FEATURE SELECTION BASED ON COMPUTATION TIME AND CLASSIFICATION ACCURACY USING SUPPORT VECTOR MACHINE

The goal of this research to compare Chi-Square feature selection with Mutual Information feature selection based on computation time and classification accuracy. In this research, people's comments on Twitter are classified based on positive, negative, and neutral sentiments using the Support Vector Machine method. Sentiment classification has the disadvantage that it has many features that are used, therefore feature selection is needed to optimize a sentiment classification performance. Chi-square feature selection and mutual information feature selection are feature selections that both can improve the accuracy of sentiment classification. How to collect the data on twitter taken using the IDE application from python. The results of this study indicate that sentiment classification using Chi-Square feature selection produces a computation time of 0.4375 seconds with an accuracy of 78% while sentiment classification using Mutual Information feature selection produces an accuracy of 80% with a required computation time of 252.75 seconds. So that the conclusion are obtained based on the computational time aspect, the Chi-Square feature selection is superior to the Mutual Information feature selection, while based on the classification accuracy aspect, the Mutual Information feature selection is more accurate than the Chi-Square feature selection. The recommendations for further research can use mutual information feature selection to get high accuracy results on sentiment classification

An innovative cluster-based prediction approach for mass solar site management

The scarcity of energy resources and global warming over the past few decades have prompted the widespread adoption of renewable energy sources. Among the potential renewable energy sources, solar energy has emerged as one of the most promising renewable energy sources. However, the uncertainty and fluctuations of solar power generation create negative impacts on the stability and reliability of the electric grid, planning of operation, economic feasibility, and investment. Therefore, accurate prediction of solar power generation is crucial to ensure the stability of the power grid and promote a large-scale investment in a solar energy system. A large number of research studies have been conducted on predicting solar power generation under different perspectives. However, no existing study analyses and predicts power generation of multi-solar energy sites by only one prediction model. The integration of multiple sites into one predictive model will reduce the number of required models for each site, thereby saving the computing resources and required calculation time. This paper proposes a novel methodology to group multiple solar sites and develop an integrated model by using a machine learning algorithm to predict power generation of each group. Firstly, the K-means clustering algorithm is utilized to cluster multiple solar sites which have similar power generation properties into one group. Then, a machine learning algorithm has been developed to predict power generation in a computationally fast and reliable manner. The proposed approach is verified by the real data of 223 solar sites in Taiwan. The experimental results show that the training time can be reduced by 93.2% without reducing the accuracy of the prediction model. Therefore, the cluster-based prediction approach gives better performance as compared with existing models.

On the use of common random numbers in activity-based travel demand modeling for scenario comparison

ABSTRACT Activity-based travel demand models provide a high level of detail when modeling complex travel behavior. Since stochastic simulation is used, however, this high level may induce large random fluctuations in the output, necessitating many model reruns to produce reliable output. This may become prohibitive in terms of computation time when comparing travel behavior between multiple scenarios, in which case each scenario requires its own simulation. To alleviate this issue, we study the use of common random numbers, which is a technique that reuses the same random numbers for choices made by travelers between scenarios. This ensures that any observed difference in output across scenarios cannot be attributed to mutual differences in drawn random numbers, eliminating an important source of random fluctuation. We demonstrate by a numerical study that common random numbers can greatly reduce the number of runs needed, and thus also the required computation time, to obtain reliable output.

Equivalent Magnetic Circuit Model of IPMSM

This paper proposes a magnetic equivalent circuit model for an interior permanent magnet synchronous motor (IPMSM) for the optimal design of the entire system. The computational speed of the proposed model is greater than that of the finite element method (FEM), and the model is suitable for large scale optimization problems. Unlike conventional magnetic circuit models, the proposed model can reflect the dq-axis saliency, which enables analysis under operating conditions using the reluctance torque, which is essential in IPMSMs. In the proposed model, rotor rotation is represented by an AC magnetomotive source, and the saliency characteristics are represented by a multi-phase-q-axis transducer that simulates a q-axis magnetic path connected in parallel to the magnets. The effectiveness of the proposed method is verified by comparisons with the FEM, and the results reveal good agreement under several conditions. Furthermore, the computation of the N-T characteristics and efficiency maps is verified, and consequently, the proposed method is deemed capable of obtaining these maps with sufficient accuracy for the initial design, and the required computation time is four orders of magnitude lower.

Creation of a Walloon Pasture Monitoring Platform Based on Machine Learning Models and Remote Sensing

The use of remote sensing data and the implementation of machine learning (ML) algorithms is growing in pasture management. In this study, ML models predicting the available compressed sward height (CSH) in Walloon pastures based on Sentinel-1, Sentinel-2, and meteorological data were developed to be integrated into a decision support system (DSS). Given the area covered (>4000 km2 of pastures of 100 m2 pixels), the consequent challenge of computation time and power requirements was overcome by the development of a platform predicting CSH throughout Wallonia. Four grazing seasons were covered in the current study (between April and October from 2018 to 2021, the mean predicted CSH per parcel per date ranged from 48.6 to 67.2 mm, and the coefficient of variation from 0 to 312%, suggesting a strong heterogeneity of variability of CSH between parcels. Further exploration included the number of predictions expected per grazing season and the search for temporal and spatial patterns and consistency. The second challenge tackled is the poor data availability for concurrent acquisition, which was overcome through the inclusion of up to 4-day-old data to fill data gaps up to the present time point. For this gap filling methodology, relevancy decreased as the time window width increased, although data with 4-day time lag values represented less than 4% of the total data. Overall, two models stood out, and further studies should either be based on the random forest model if they need prediction quality or on the cubist model if they need continuity. Further studies should focus on developing the DSS and on the conversion of CSH to actual forage allowance.

A novel intrusion detection system based on deep learning and random forest for digital twin on IOT platform

Digital Twin (DT) has bright prospects for a broad spectrum of applications, such as supply chain, healthcare, predictive maintenance, etc. On one hand, the intrinsic properties of DT make it highly relevant to real-world applications, while on the other hand, they render it vulnerable to cyber threats. So, the security of DT is of utmost importance for the security of the communication and computational infrastructure underneath it. As cybercriminals are always coming up with new attack techniques, it is essential to safeguard the digital twin against malicious attacks. This can be accomplished by putting in place an efficient security mechanism, one of which is an intrusion detection system (IDS). The work done in this paper introduces a novel hybrid model for IDS in digital twins that integrates deep learning (DL) and random forest (RF). The effectiveness of the developed model is evaluated in comparison to that of k-Nearest Neighbors (KNN), Naive Bayes, Logistic Regression (LR), Support Vector Machine (SVM), Random Forest (RF), and Long Short Term Memory (LSTM), and it’s evident from the outcome that the proposed method has higher accuracy than the competing models by at least 5%. In addition to its high accuracy, our proposed model also promises impressively low computing time requirements when compared to other competing models.

Predictive Energy Management for Hybrid Electric Aircraft Propulsion Systems

We present a model predictive control (MPC) algorithm for energy management in aircraft with hybrid electric propulsion systems consisting of gas turbine and electric motor components. Series and parallel configurations are considered. By combining a point-mass aircraft dynamical model with models of electrical losses and losses in the gas turbine, the fuel consumed over a given future flight path is minimized subject to constraints on the battery, electric motor, and gas turbine. The optimization is formulated as a convex problem under mild assumptions and its solution is used to define a predictive energy management control law that takes into account the variation in aircraft mass during flight. We investigate the performance of algorithms for solving this problem. An alternating direction method of multipliers (ADMM) algorithm is proposed and compared with a general purpose convex interior point solver. We also show that the ADMM implementation reduces the required computation time by orders of magnitude in comparison with a general purpose nonlinear programming solver, making it suitable for real-time supervisory energy management control.

A parallel workflow framework using encoder-decoder LSTMs for uncertainty quantification in contaminant source identification in groundwater

Groundwater contamination is a major problem and it can be contaminated by natural or manmade activities. Contaminant sources cannot be identified promptly after a contamination event due to slow groundwater movement in aquifers. To limit the further spread of the plume and to take appropriate remedial measures, it is essential to determine the release rates associated with source locations as soon as a contaminant is detected in one of the observation wells in the aquifer. Conventional techniques for contaminant source identification such as the simulation–optimization (SO) uses a simulation model, which is based on estimated or, measured hydrogeological parameters, for replicating the aquifer response to contamination at source locations. However, the associated intrinsic uncertainty in the parameters can affect the estimated release histories at the source locations. To assess confidence in the results, it is important to consider stochasticity in the hydrogeological parameters for contaminant release history estimation requiring numerous runs of the SO model for corresponding breakthrough curves. The present study proposes a novel surrogate simulation optimization (SSO) framework which utilizes an Encoder-Decoder Long Short-Term Memory (ED-LSTM) based surrogate model along with Multiverse Optimization (MVO) for contaminant source identification. The comparison of ED-LSTM with other surrogate models shows its better accuracy and significantly low computational time requirement compared to the simulation model. To enhance the computational efficiency further, the SSO model is embedded in a parallel workflow for the quantification of uncertainty associated with the inverse modelling. The study shows that the proposed workflow can be efficiently used for uncertainty quantification in contaminant source identification due to stochastic parameters with a reduced computational cost.

A Fast Forward and Inversion Strategy for Three-Dimensional Gravity Field

Obtaining a three-dimensional (3D) density distribution within a reasonable time is one of the most critical problems in gravity exploration. In this paper, we present an efficient 3D forward modeling and inversion method for gravity data. In forward modeling, the 3D model is discretized into multiple horizontal layers, with the gravity field at a point on the surface being the sum of the gravity fields from all layers. To calculate the gravity field from each horizontal layer, we use the fast Fourier transform (FFT) method and the Block Toeplitz with Toeplitz Blocks (BTTB) matrix, which dramatically reduces both the computation time and storage requirement. In the inversion, the observed gravity data are separated into multiple gravity components of different depths using the cutting separation method. An iterative method is used to adjust the model to fit the above gravity component for each cutting radius. The initial model is constructed from the transformation of gravity components. These methods were applied to both synthetic data and field data. The numerical simulation validated the proposed methods, and the inversion results of field data were consistent with information obtained from well logging. The computational time and memory usage were also reasonable.