Fast Algorithm Research Articles

Background-Oriented Schlieren For The Detection Of Thermoacoustic Oscillations In Flames

Thermoacoustic oscillations are well known to combustion engineers. They are not only a cause of concern, but also give hope to be able to operate aircraft engines with hydrogen, due to flame stabilization and significant reduction of NOx emissions when thermoacoustically excited. The aim of this work is to investigate the potential of using a heterodyne background-oriented schlieren (HBOS) technique to detect thermoacoustic oscillations in the density field of an acoustically excited flame. It also seeks to address the issue of accuracy due to the small amplitude of thermoacoustic oscillations compared to turbulent density fluctuations in a flame. The experiment uses an unconfined swirl-stabilized methane flame and investigates thermoacoustic oscillations at about 3.4 kW power at ambient conditions excited by a siren at 225 Hz. The HBOS technique recently presented by the authors uses a carrier fringe system in background-oriented schlieren recordings, with subsequent analysis using evaluation techniques known from holographic interferometry. These fast algorithms reduce turbulence noise by phase averaging a large number of images. To derive local data from the line-of-sight projections, an inverse Abel transform is applied. Laser interferometric vibrometry is used to calibrate to the observed oscillations. A comparison with data from chemiluminescence is also presented to better demonstrate the application of this efficient technique to the detection of heat release oscillations. Future tasks in this project will deal with full-scale test benches and machine learning tools to address the limited observations in them.

Dynamics simulation-based packing of irregular 3D objects

The 3D packing problem has a wide range of applications. However, the complex geometry of irregular objects leads to a sharp increase in the number of placement combinations, making it a challenging problem. In this paper, we propose a packing pipeline based on rigid body dynamics simulation to deal with two types of 3D packing problems. One is the variant bin packing problem, which involves placing more objects into a container of given dimensions to maximize space utilization. The other is the open dimension problem, where the goal is to minimize the container that can accommodate all objects. We first use heuristic placement strategies and a fast collision detection algorithm to efficiently obtain initial packing results. Then, we simulate the shaking of the container according to the dynamic principle. Combined with the vacant space filling operation, shaking the container drives the movement of objects in the container to make the arrangement of objects more compact. For the open dimension packing, the container height is optimized by adjusting the constraints of simulation in the basic pipeline. Experimental results show that our method has advantages over existing methods in both speed and packing density.

New global multi-objective optimization of compact heat exchanger

This paper introduces a robust and global algorithm, the multi-objective version of the Set Trimming method (MOST). The application of MOST is demonstrated through optimization of a plate and fin heat exchanger. The results are compared with NSGA-II (Fast and elitism non-dominated sorting Genetic Algorithm) and an exhaustive search. The optimization focuses on effectiveness and annual cost, with seven selected design parameters and for six different cases. Results indicate that the Pareto optimal fronts obtained with NSGA-II are entirely dominated by those achieved using MOST in the all studied cases. Furthermore, MOST improves annual cost for the best economic point solution by up to 13.44% as compared to NSGA-II. Simultaneously, effectiveness is enhanced up to 0.04% when using MOST compared to NSGA-II. The computational time is also improved with the presented optimization algorithm compared to NSGA-II in the range of 78.6–88.8%.

A real-time CNN–BiLSTM-based classifier for patient-centered AR-SSVEP active rehabilitation exoskeleton system

The rehabilitation of stroke has brought a serious social burden to many countries. The augmented reality steady-state visually evoked potential (AR-SSVEP) exoskeleton rehabilitation system can effectively help stroke patients recover motor function and reduce the burden on society and rehabilitation physicians. However, the decoding accuracy and efficiency of AR-SSVEP in short time stimulus are low, and the types of trajectory decoded are few. It has a negative impact on the patient’s rehabilitation interest and efficiency. In this article, a patient-centered AR-SSVEP active rehabilitation exoskeleton system is developed based on the CNN–BiLSTM fast classification algorithm. The algorithm uses CNN architecture to improve model generalization, and uses BiLSTM to improve classification accuracy. The patient-centered interactive trajectory decoding logic based on a task-driven strategy is proposed to achieve patient-defined coordinated movement trajectory of the left and right upper limb shoulder elbow joints through the 6-class brain–computer interface (BCI). The offline and online experiments were conducted on the system. In offline experiments, under a 0.5 s time window, the classification accuracy rate was 98.5%, and the information transfer rate (ITR) reached 210 bits/min. In online experiment, the patient controlled the exoskeleton of the upper limb to perform four non-repetitive trajectories, with an average time of 319 s, which was 20% more than the standard reference time. The experimental results have shown the system can effectively improve the accuracy and efficiency of motion intention identification within short time stimulus, and decode patient-defined motion trajectory. In conclusion, this study will provide a new rehabilitation way for stroke patients.

An efficient algorithm for time-domain acoustic scattering in three dimensions by layer potentials

In this paper, we develop a simple and fast algorithm for the time-domain acoustic scattering by a sound-soft or an impedance obstacle in three dimensions. We express the solution to the scattering problem by layer potentials, and then a time-domain boundary integral equation is derived. To numerically solve the resulting boundary integral equation, we propose a full discretization scheme by combining the convolution splines with a Galerkin method. In time, we approximate the density in a backward manner in terms of the convolution splines. In space, we project the density at each time onto the space of spherical harmonics, and then use the spatial discretization of a Nyström type on the surface of an obstacle which is homeomorphic to a sphere. A gallery of numerical examples are presented to show the efficiency of our algorithm. The stability, convergence and accuracy of the algorithm are discussed.

A novel fast robust optimization algorithm for intensity-modulated proton therapy with minimum monitor unit constraint.

Intensity-modulated proton therapy (IMPT) optimizes spot intensities and position, providing better conformability. However, the successful application of IMPT is dependent upon addressing the challenges posed by range and setup uncertainties. In order to address the uncertainties in IMPT, robust optimization is essential. This study aims to develop a novel fast algorithm for robust optimization of IMPT with minimum monitor unit (MU) constraint. The study formulates a robust optimization problem and proposes a novel, fast algorithm based on the alternating direction method of multipliers (ADMM) framework. This algorithm enables distributed computation and parallel processing. Ten clinical cases were used as test scenarios to evaluate the performance of the proposed approach. The robust optimization method (RBO-NEW) was compared with plans that only consider nominal optimization using CTV (NMO-CTV) without handling uncertainties and PTV (NMO-PTV) to handle the uncertainties, as well as with conventional robust-optimized plans (RBO-CONV). Dosimetric metrics, including D95, homogeneity index, and Dmean, were used to evaluate the dose distribution quality. The area under the root-mean-square dose (RMSD)-volume histogram curves (AUC) and dose-volume histogram (DVH) bands were used to evaluate the robustness of the treatment plan. Optimization time cost was also assessed to measure computationalefficiency. The results demonstrated that the RBO plans exhibited better plan quality and robustness than the NMO plans, with RBO-NEW showing superior computational efficiency and plan quality compared to RBO-CONV. Specifically, statistical analysis results indicated that RBO-NEW was able to reduce the computational time from to s ( ) and reduce the mean organ-at-risk (OAR) dose from % of the prescription dose to % of the prescription dose ( ) compared toRBO-CONV. This study introduces a novel fast robust optimization algorithm for IMPT treatment planning with minimum MU constraint. Such an algorithm is not only able to enhance the plan's robustness and computational efficiency without compromising OAR sparing but also able to improve treatment plan quality andreliability.

Optimal control of a quantum sensor: A fast algorithm based on an analytic solution

Quantum sensors can show unprecedented sensitivities, provided they are controlled in a very specific, optimal way. Here, we consider a spin sensor of time-varying fields in the presence of dephasing noise, and we show that the problem of finding the pulsed control field that optimizes the sensitivity (i.e., the smallest detectable signal) can be mapped to the determination of the ground state of a spin chain. We find an approximate but analytic solution of this problem, which provides a lower bound for the sensitivity and a pulsed control very close to optimal, which we further use as initial guess for realizing a fast simulated annealing algorithm. We experimentally demonstrate the sensitivity improvement for a spin-qubit magnetometer based on a nitrogen-vacancy center in diamond.

Optimal pipe-sizing design of water distribution networks using modified Rao-II algorithm

ABSTRACT Several evolutionary algorithms (EAs) have been suggested in the last three decades for the least-cost design of water distribution networks (WDNs). EAs generally worked well to identify the global/near-global optimal solutions for small- to moderate-size networks in a reasonable time and computational effort. However, their applications to large-size networks are still challenging due to large computational effort. Recent developments in EAs are towards parameter-less techniques that avoid fine-tuning of case-specific parameters to reduce the computational effort. Further, several self-adaptive penalties and search-space reduction methodologies have been suggested to reduce the computational effort. A fast, efficient, and parameter-less Rao-II algorithm has been used earlier with penalty-based approaches for the optimal design of WDNs. In this study, the application of a Rao-II algorithm is further explored with three self-adaptive penalty approaches to compare the convergence characteristics. The Rao-II algorithm is observed to converge at an infeasible solution in cases that the applied penalty to an infeasible solution is not so substantial to make it inferior to the feasible solutions. Modifications are suggested to improve the Rao-II algorithm, named the modified Rao-II algorithm. The modified Rao-II algorithm with the self-adaptive penalty methods resulted in better solutions than those obtained earlier.

NeuroMotion: Open-source platform with neuromechanical and deep network modules to generate surface EMG signals during voluntary movement.

Neuromechanical studies investigate how the nervous system interacts with the musculoskeletal (MSK) system to generate volitional movements. Such studies have been supported by simulation models that provide insights into variables that cannot be measured experimentally and allow a large number of conditions to be tested before the experimental analysis. However, current simulation models of electromyography (EMG), a core physiological signal in neuromechanical analyses, remain either limited in accuracy and conditions or are computationally heavy to apply. Here, we provide a computational platform to enable future work to overcome these limitations by presenting NeuroMotion, an open-source simulator that can modularly test a variety of approaches to the full-spectrum synthesis of EMG signals during voluntary movements. We demonstrate NeuroMotion using three sample modules. The first module is an upper-limb MSK model with OpenSim API to estimate the muscle fibre lengths and muscle activations during movements. The second module is BioMime, a deep neural network-based EMG generator that receives nonstationary physiological parameter inputs, like the afore-estimated muscle fibre lengths, and efficiently outputs motor unit action potentials (MUAPs). The third module is a motor unit pool model that transforms the muscle activations into discharge timings of motor units. The discharge timings are convolved with the output of BioMime to simulate EMG signals during the movement. We first show how MUAP waveforms change during different levels of physiological parameter variations and different movements. We then show that the synthetic EMG signals during two-degree-of-freedom hand and wrist movements can be used to augment experimental data for regressing joint angles. Ridge regressors trained on the synthetic dataset were directly used to predict joint angles from experimental data. In this way, NeuroMotion was able to generate full-spectrum EMG for the first use-case of human forearm electrophysiology during voluntary hand, wrist, and forearm movements. All intermediate variables are available, which allows the user to study cause-effect relationships in the complex neuromechanical system, fast iterate algorithms before collecting experimental data, and validate algorithms that estimate non-measurable parameters in experiments. We expect this modular platform will enable validation of generative EMG models, complement experimental approaches and empower neuromechanical research.

Novel online portfolio selection algorithm using deep sequence features and reversal information

Computational finance combines machine learning with financial needs to provide more efficient solutions for investment analysis and automated trading. In previous studies, traditional online portfolio selection (OLPS) algorithms were found to be overly reliant on artificially designed, subjective financial features. To address this issue, we propose a new predictive price tracking algorithm based on deep sequence features and reversal information (DSF-RI-PPT) for OLPS, extending a hybrid stock prediction algorithm to a multi-asset trading algorithm. We respectively employ the complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN), principal component analysis (PCA) algorithms and long short-term memory (LSTM) network to perform decomposition, feature extraction and prediction on financial data. Further, we supplement the reversal information by modifying the predicted prices with a reversal indicator-rate of change (ROC). Finally, we introduce a fast error back-propagation algorithm to integrate the predictive information into the investment ratio using gradient projection. Through empirical comparison and statistic analysis of the DSF-RI-PPT algorithm, price-tracking algorithms with similar prediction models, and nine classic OLPS algorithms in nine portfolio data sets under three financial indexes, it can be found that the DSF-RI-PPT algorithm is profitable and generalizable.

A simple and fast measurement algorithm for power system flicker severity evaluation

A simple and fast measurement algorithm for power system flicker severity evaluation

Early-Season Crop Mapping by PRISMA Images Using Machine/Deep Learning Approaches: Italy and Iran Test Cases

Despite its high importance for crop yield prediction and monitoring, early-season crop mapping is severely hampered by the absence of timely ground truth. To cope with this issue, this study aims at evaluating the capability of PRISMA hyperspectral satellite images compared with Sentinel-2 multispectral imagery to produce early- and in-season crop maps using consolidated machine and deep learning algorithms. Results show that the accuracy of crop type classification using Sentinel-2 images is meaningfully poor compared with PRISMA (14% in overall accuracy (OA)). The 1D-CNN algorithm, with 89%, 91%, and 92% OA for winter, summer, and perennial cultivations, respectively, shows for the PRISMA images the highest accuracy in the in-season crop mapping and the fastest algorithm that achieves acceptable accuracy (OA 80%) for the winter, summer, and perennial cultivations early-season mapping using PRISMA images. Moreover, the 1D-CNN algorithm shows a limited reduction (6%) in performance, appearing to be the best algorithm for crop mapping within operational use in cross-farm applications. Machine/deep learning classification algorithms applied on the test fields cross-scene demonstrate that PRISMA hyperspectral time series images can provide good results for early- and in-season crop mapping.

TORONTO: A trial-oriented multidimensional psychometric testing algorithm.

Bayesian adaptive methods for sensory threshold determination were conceived originally to track a single threshold. When applied to the testing of vision, they do not exploit the spatial patterns that underlie thresholds at different locations in the visual field. Exploiting these patterns has been recognized as key to further improving visual field test efficiency. We present a new approach (TORONTO) that outperforms other existing methods in terms of speed and accuracy. TORONTO generalizes the QUEST/ZEST algorithm to estimate simultaneously multiple thresholds. After each trial, without waiting for a fully determined threshold, the trial-oriented approach updates not only the location currently tested but also all other locations based on patterns in a reference data set. Since the availability of reference data can be limited, techniques are developed to overcome this limitation. TORONTO was evaluated using computer-simulated visual field tests: In the reliable condition (false positive [FP] = false negative [FN] = 3%), the median termination and root mean square error (RMSE) of TORONTO was 153 trials and 2.0dB, twice as fast with equal accuracy as ZEST. In the FP = FN = 15% condition, TORONTO terminated in 151 trials and was 2.2 times faster than ZEST with better RMSE (2.6 vs. 3.7dB). In the FP = FN = 30% condition, TORONTO achieved 4.2dB RMSE in 148 trials, while all other techniques had > 6.5dB RMSE and terminated much slower. In conclusion, TORONTO is a fast and accurate algorithm for determining multiple thresholds under a wide range of reliability and subject conditions.

A case study showing highly traceable sources of bacteria on surfaces of university buildings

University students predominantly spend their time indoors, where prolonged exposure raises the risk of contact with microorganisms of concern. However, our knowledge about the microbial community characteristics on university campus and their underpinnings is limited. To address it, we characterized bacterial communities from the surfaces of various built environments typical of a university campus, including cafeterias, classrooms, dormitories, offices, meeting rooms, and restrooms, in addition to human skin. The classrooms harbored the highest α-diversity, while the cafeterias had the lowest α-diversity. The bacterial community composition varied significantly across different building types. Proteobacteria, Actinobacteria, Firmicutes, Bacteroidetes, and Cyanobacteria were common phyla in university buildings, accounting for more than 90 % of total abundance. Staphylococcus aureus was the most abundant potential pathogen in classrooms, dormitories, offices, restrooms, and on human skin, indicating a potential risk for skin disease infections in these buildings. We further developed a new quantitative pathogenic risk assessment method according to the threat of pathogens to humans and found that classrooms exhibited the highest potential risk. The fast expectation-maximization algorithm identified 59 %-86 % of bacterial sources in buildings, with the human skin as the largest bacterial source for most buildings. As the sources of bacteria were highly traceable, we showed that homogeneous selection, dispersal limitation, and ecological drift were major ecological forces that drove community assembly. Our findings have important implications for predicting the distribution and sources of indoor dust bacterial communities on university campus.

Fast Screening Algorithm for Satellites Based on Multi-Constellation System

This paper proposes a fast satellite screening algorithm aimed at the problem of balancing between positioning accuracy and system computing efficiency in a multi-constellation system environment under the Global Navigation Satellite System (GNSS). The algorithm constructs an observation model based on a positioning error for the larger number of satellites under a multi-constellation. The space region is divided based on elevation and azimuth angles to implement the screening algorithm for the solution of the point to be determined. An analysis of the experimental data shows that the average GDOP value of this scheme is 1.835, and the position error of the point to be determined is controlled within 2.5 m when the cut-off altitude angle is 5° and the screening ratio is more than 70%.

The Heston–Queue-Hawkes process: A new self-exciting jump–diffusion model for options pricing, and an extension of the COS method for discrete distributions

We propose a new self-exciting jump–diffusion process, the Heston–Queue-Hawkes (HQH) model, which integrates the well-known Heston model with the recently introduced Queue-Hawkes (Q-Hawkes) jump process. Similar to the Heston–Hawkes process (HH), the HQH model effectively captures both the slow and continuous evolution of prices, and the sudden and impactful market movements due to self-excitation and contagion. But a significant advantage of the HQH model is that its characteristic function is available in closed form, allowing for the efficient application of Fourier-based fast pricing algorithms, such as the COS method (which we extend to deal with discrete distributions). We also demonstrate that, by leveraging partial integrals of the characteristic function, which are explicitly known for the HQH process, we can reduce the dimensionality of the COS method, thereby decreasing its numerical complexity. Our numerical results for pricing European and Bermudan options indicate that the HQH model provides a broader range of volatility smiles compared to the Bates model, while maintaining a substantially lower computational burden than the HH process.

Rapid variable-step computation of dynamic convolutions and Volterra-type integro-differential equations: RK45 Fehlberg, RK4

We introduce a novel, time-efficient adaptive Runge-Kutta computational scheme tailored for systematically solving linear and nonlinear Volterra-type Integro-Differential Equations (VTIDEs). This scheme is particularly effective for equations featuring a specific class of convolution memory, K(t,s)=aκ(t−s) within convolution integrals, where a,κ∈R. Such equations frequently arise in fields like viscoelasticity, population dynamics, epidemiology, control systems, financial mathematics, and neuroscience, where current states are influenced by historical data, especially in dynamical systems within physics. Our proposed scheme demonstrates remarkable computational efficiency, achieving a cost-effective computation time of O(Nt) for Nt iterations, a significant improvement over existing schemes for VTIDE, which typically exhibit complexities of O(Ntlog⁡Nt) and O(Nt(log⁡Nt))2. The key feature of our scheme lies in its ability to elegantly and efficiently handle convoluted integrals of varying complexity such as those including nonlinearities and derivatives. Specifically, our approach eliminates the need for backstage integration, allowing the main Runge-Kutta scheme to operate efficiently with only a few data points at each time step. This not only saves time but also prevents potential data storage issues while rendering numerical implementation straightforward and tractable. Moreover, this article introduces flexible discretizations for integration and differentiation, accommodating unequally spaced data points. We implement a fast RK45-Fehlberg algorithm (TE-RK45-Fehlberg), which we use to solve three examples of nonlinear VTIDE presented in this study. We further demonstrate the effectiveness of our approach by using an example, where we employ a novel parabolic step size generator function to efficiently generate step sizes for a computationally fast RK4 scheme (Flex-RK4). For all examples tackled using TE-RK45-Fehlberg and Flex-RK4, we generate plots that compare numerical data with exact solutions, providing insight into the accuracy and reliability of our approach. We also show how the scheme provides straightforward ways of transforming it into implicit forms and obtaining stability curves.

An Improved Frequent Directions Algorithm for Low-Rank Approximation via Block Krylov Iteration.

Frequent directions (FDs), as a deterministic matrix sketching technique, have been proposed for tackling low-rank approximation problems. This method has a high degree of accuracy and practicality but experiences a lot of computational cost for large-scale data. Several recent works on the randomized version of FDs greatly improve the computational efficiency but unfortunately sacrifice some precision. To remedy such an issue, this article aims to find a more accurate projection subspace to further improve the efficiency and effectiveness of the existing FDs' techniques. Specifically, by utilizing the power of the block Krylov iteration and random projection technique, this article presents a fast and accurate FDs algorithm named r-BKIFD. The rigorous theoretical analysis shows that the proposed r-BKIFD has a comparable error bound with original FDs, and the approximation error can be arbitrarily small when the number of iterations is chosen appropriately. Extensive experimental results on both synthetic and real data further demonstrate the superiority of r-BKIFD over several popular FDs algorithms both in terms of computational efficiency and accuracy.

Fast Algorithm for Noninvasive SAR Prediction Based on Artificial Neural Networks

Fast Algorithm for Noninvasive SAR Prediction Based on Artificial Neural Networks

Optimization of urban transport vehicle tasks for large, mixed fleet of vehicles and real-world constraints

Planning the operation of urban public transport vehicles is the first stage of operational planning and consists in combining timetable tips, which are input data, into blocks that constitute the daily tasks of vehicles. For a large mixed fleet of vehicles of various types, especially those with battery power that requires recharging, operating from many depots, with numerous requirements and rolling stock constraints, the problem is a major engineering challenge, even for an experienced team of planners. IT solutions based on realistic, mathematical decision-making models and fast optimization algorithms can be of great assistance. For the problem formulated this way, a mathematical decision model with a multi-criteria objective function was built, taking into account technical, economic, and ecological criteria, natural, and binary decision variables. The model takes into account the real requirements and constraints, a mixed fleet of different types vehicles, including electric buses, multiple depots, technical trips (dead heads), and battery charging. The considered problem is an NP-hard combinatorial optimization problem. The use of classical, exact algorithms to solve this problem is not possible for schedules with many thousands of line trips and fleets of hundreds or thousands of vehicles. This research proposes an original, dedicated heuristic, enabling to obtain an acceptable, but still suboptimal solution, in a very short time. The tests of the proposed heuristic algorithm were carried out on real databases of public transport systems of the two selected medium and large Polish cities. In particular, multiple depots, a mixed fleet of different types of vehicles, and real-world constraints were taken into account. The results of computer experiments carried out using the developed heuristic were compared with the results obtained manually by a team of experienced and expert planners. For the developed multi-criteria decision-making model results comparable to and better than those prepared manually by experts were obtained in a very short time using the proposed heuristics. It is the basis for further development works on expanding the model and improving the optimization algorithm.