Hybrid Optimization Scheme Research Articles

Integrating different fidelity models for process optimization: A case of equilibrium and rate-based extractive distillation using ionic liquids

We integrate equilibrium and rate-based models to formulate a hybrid optimization scheme for designing an ionic liquid-based extractive distillation process for mixed-refrigerant separation. The equilibrium model assumes vapor–liquid equilibrium at each stage but challenges arise with low-volatility, high-viscosity solvents, which drive the system away from equilibrium. The rate-based approach considers mass and heat transfer rates, giving more accurate representation. We compare the two models for separating R-410A, an azeotropic mixture of R-32 and R-125, using [EMIM][SCN] ionic liquid as entrainer. Analyzing over 4300 simulations with various dimensionality reduction and topological analysis techniques, we find that predictions from the two models exhibit similar trends, but the overestimation in equilibrium-based purities sometimes leads to infeasible process designs. The proposed optimization algorithm thus combines the strengths of the two models to locate feasible and optimal designs.

Neural dynamical operator: Continuous spatial-temporal model with gradient-based and derivative-free optimization methods

Data-driven modeling techniques have been explored in the spatial-temporal modeling of complex dynamical systems for many engineering applications. However, a systematic approach is still lacking to leverage the information from different types of data, e.g., with different spatial and temporal resolutions, and the combined use of short-term trajectories and long-term statistics. In this work, we build on the recent progress of neural operator and present a data-driven modeling framework called neural dynamical operator that is continuous in both space and time. A key feature of the neural dynamical operator is the resolution-invariance with respect to both spatial and temporal discretizations, without demanding abundant training data in different temporal resolutions. To improve the long-term performance of the calibrated model, we further propose a hybrid optimization scheme that leverages both gradient-based and derivative-free optimization methods and efficiently trains on both short-term time series and long-term statistics. We investigate the performance of the neural dynamical operator with three numerical examples, including the viscous Burgers' equation, the Navier–Stokes equations, and the Kuramoto–Sivashinsky equation. The results confirm the resolution-invariance of the proposed modeling framework and also demonstrate stable long-term simulations with only short-term time series data. In addition, we show that the proposed model can better predict long-term statistics via the hybrid optimization scheme with a combined use of short-term and long-term data.

Wheel Wear Modeling and Re-Profiling Strategy Optimization Based on the Maintenance Data of a Three-Axle Bogie Locomotive

Wear caused by wheel–rail contact forces is inevitable during vehicle operation, which has an important impact on the security and stability of train operation. Therefore, it is of great significance to study wheel wear patterns and optimize re-profiling strategies to extend service life. Based on the wheel wear data of three-axle bogie locomotives, this paper proposes a data-driven hybrid wheel wear model and optimization schemes of the re-profiling strategy. The wear model consists of a wheel flange thickness wear model, a wheel diameter wear model, and a re-profiling ratio coefficient model. Then, utilizing the above models, the optimization of the re-profiling strategy for different axle position wheels is raised, and the optimization of the complete vehicle re-profiling strategy is presented by considering the wheel diameter difference. Finally, a wheel data-driven analysis platform was developed to enable the management and utilization of wheel maintenance data. Analysis of extensive maintenance data indicates that the guide wheels wear the fastest, approximately 22.2% higher than the middle wheels, which wear the slowest. The re-profiling ratio coefficient model indicates that the ratio coefficient increases as the wheel flange thickness before re-profiling increases. Simulations demonstrate a longer expected wheel life with a flange thickness between 28 and 32.5 mm. Compared to measured values, the optimization strategy reduces complete vehicle re-profiling by 21.5%. Through the initial implementation at CRRC Dalian Locomotive Ltd, it has become evident that this methodology offers a viable solution to enhance the service longevity of locomotive wheels.

Autonomous hybrid optimization of a SiO2 plasma etching mechanism

Computational modeling of plasma etching processes at the feature scale relevant to the fabrication of nanometer semiconductor devices is critically dependent on the reaction mechanism representing the physical processes occurring between plasma produced reactant fluxes and the surface, reaction probabilities, yields, rate coefficients, and threshold energies that characterize these processes. The increasing complexity of the structures being fabricated, new materials, and novel gas mixtures increase the complexity of the reaction mechanism used in feature scale models and increase the difficulty in developing the fundamental data required for the mechanism. This challenge is further exacerbated by the fact that acquiring these fundamental data through more complex computational models or experiments is often limited by cost, technical complexity, or inadequate models. In this paper, we discuss a method to automate the selection of fundamental data in a reduced reaction mechanism for feature scale plasma etching of SiO2 using a fluorocarbon gas mixture by matching predictions of etch profiles to experimental data using a gradient descent (GD)/Nelder–Mead (NM) method hybrid optimization scheme. These methods produce a reaction mechanism that replicates the experimental training data as well as experimental data using related but different etch processes.

Hybrid optimization schemes for solving the piezoresistive inversion problem in self-sensing materials

Materials with electrically conductive nanofillers have the ability to ‘sense’ changes to their mechanical state. When these materials are deformed, the embedded nanofiller networks are disturbed causing a measurable change in the electrical conductivity of the material. This self-sensing property, known as piezoresistivity, has been leveraged in numerous engineering venues. Although this property has been thoroughly explored, prevailing self-sensing techniques provide little-to-no information about the underlying mechanical state of the material, such as the displacement and strain. This information must be indirectly obtained from the conductivity change. This limitation exists because obtaining mechanics from conductivity is an under-determined inverse problem with many possible mathematically feasible solutions. Previous work in this area used metaheuristic algorithms and imposed mechanics-based constraints to solve the piezoresistive inversion problem. Although this approach was successful, it was computationally inefficient due to the stochastic search process and the need to perform multiple searches to find a converged solution. To overcome this limitation, we herein propose a hybrid optimization scheme for solving the piezoresistive inversion problem. This scheme is implemented in two steps. In the first step, a metaheuristic algorithm performs a single search for a suitable solution to the inverse problem. In the second step, a gradient descent algorithm searches for the final solution using the solution from the previous step as the starting point. We explore different norms for the fitness function of the metaheuristic search and demonstrate using experimental data that the proposed hybrid optimization scheme can accurately and efficiently calculate displacements and strains from conductivity changes. This exploration significantly advances the state of the art by enabling computationally efficient and highly accurate predictions of full-field mechanical condition in self-sensing materials for the first time, thereby paving the way for greater use of these principles in practice.

HybridCap: Inertia-Aid Monocular Capture of Challenging Human Motions

Monocular 3D motion capture (mocap) is beneficial to many applications. The use of a single camera, however, often fails to handle occlusions of different body parts and hence it is limited to capture relatively simple movements. We present a light-weight, hybrid mocap technique called HybridCap that augments the camera with only 4 Inertial Measurement Units (IMUs) in a novel learning-and-optimization framework. We first employ a weakly-supervised and hierarchical motion inference module based on cooperative pure residual recurrent blocks that serve as limb, body and root trackers as well as an inverse kinematics solver. Our network effectively narrows the search space of plausible motions via coarse-to-fine pose estimation and manages to tackle challenging movements with high efficiency. We further develop a hybrid optimization scheme that combines inertial feedback and visual cues to improve tracking accuracy. Extensive experiments on various datasets demonstrate HybridCap can robustly handle challenging movements ranging from fitness actions to Latin dance. It also achieves real-time performance up to 60 fps with state-of-the-art accuracy.

Optimization enabled deep learning approach with probabilistic fusion for image inpainting

ABSTRACT Nowadays, various deep learning (DL) approaches have been devised for image inpainting, which provided a substantial improvement in image quality. However, these approaches have failed to reconstruct the accurate structure of the original image. Hence, this research devised a novel and effective image inpainting approach, namely Autoregressive Flower Pollination Student Psychology Optimization (ArFPSPO). The image inpainting is carried out based on a newly devised hybrid context DL with the hybrid optimization scheme. The designed hybrid context DL approach adopts three DL techniques, such as Context encoder (CE), Context-Conditional Generative Adversarial Networks (CC-GAN), and Partial convolutional layer to complete the image inpainting process so that the outcomes are fused using probabilistic fusion with maximum entropy, thereby the final inpainted image is attained. Each of these three DL techniques is separately trained using a developed hybrid optimization technique. The experimental outcome reveals that the devised model gives optimal performance.

Optimal location determination for an EVPL and capacitors in distribution network considering power loss and voltage profile: SGOS2A technique

This paper proposes a hybrid optimization system depending on optimal location for electric vehicles parking lot (PL) and capacitors on distribution system to maintain voltage profile with electricity loss. The proposed system is the consolidation of Seagull optimization algorithm (SGO) and salp swarm algorithm (SSA). The migration and attacking behaviors of seagull is empowered through SSA method. By this manner, the proposed hybrid optimization scheme is known as SGOS2A method. Here, parking zone allocation with capacitor is considered to congestion management in conjunction through the compensation of reactive energy. So, one can optimally decide the size of automobile parking space, SGOS2A method is followed. Moreover, parking lot with capacitor allocation is considered to congestion control at the side of reactive power compensation. By this proper manipulate, the capacitors exact location, automobile parking space of electric vehicles on the grid, lessening of active with reactive power loss, voltage profile conversion is selected optimally. Besides, the proposed SGOS2A scheme is activated on MATLAB/Simulink site, then the efficiency is examined with different techniques. The mean, median and standard deviation of the proposed approach achieves 1.0593, 1.0915 and 0.1050.

A novel hybrid Chaotic Aquila Optimization algorithm with Simulated Annealing for Unmanned Aerial Vehicles path planning

Unmanned Aerial Vehicles path planning is one of the critical issues in terms of guaranteeing good performance in real-world applications. Basically, it is responsible for determining and ensuring a short, smooth, and collision-free path between two positions from a source to a destination point. This paper presents a hybrid optimization scheme based on the hybridization of Chaotic Aquila Optimization with Simulated Annealing for solving the Unmanned Aerial Vehicles path planning problem. The main purpose of using Simulated Annealing is to provide a more suitable balance between exploitation/exploration. The performance of the proposed algorithm is assessed using three benchmark maps with various numbers of threats. Compared to nine well-known meta-heuristics, the results of simulated experiments demonstrate that the proposed algorithm performs better in the majority of the scenarios by obtaining a significant enhancement in minimizing the fitness value and the path cost up to 36% and 19%, respectively.

Low-dimensional Manifold Constrained Disentanglement Network for Metal Artifact Reduction.

Deep neural network based methods have achieved promising results for CT metal artifact reduction (MAR), most of which use many synthesized paired images for supervised learning. As synthesized metal artifacts in CT images may not accurately reflect the clinical counterparts, an artifact disentanglement network (ADN) was proposed with unpaired clinical images directly, producing promising results on clinical datasets. However, as the discriminator can only judge if large regions semantically look artifact-free or artifact-affected, it is difficult for ADN to recover small structural details of artifact-affected CT images based on adversarial losses only without sufficient constraints. To overcome the illposedness of this problem, here we propose a low-dimensional manifold (LDM) constrained disentanglement network (DN), leveraging the image characteristics that the patch manifold of CT images is generally low-dimensional. Specifically, we design an LDM-DN learning algorithm to empower the disentanglement network through optimizing the synergistic loss functions used in ADN while constraining the recovered images to be on a low-dimensional patch manifold. Moreover, learning from both paired and unpaired data, an efficient hybrid optimization scheme is proposed to further improve the MAR performance on clinical datasets. Extensive experiments demonstrate that the proposed LDM-DN approach can consistently improve the MAR performance in paired and/or unpaired learning settings, outperforming competing methods on synthesized and clinical datasets.

An optimization planning framework for allocating multiple distributed energy resources and electric vehicle charging stations in distribution networks

In developing a sustainable and efficient power systems network while reducing carbon footprint, renewable energy (RE)-based Distribution Generation (DG) units are highly recommended. Furthermore, Battery Energy Storage Systems (BESS) and other passive electronic units are adopted to improve grid performance and mitigate the effects of high variability from RE power. Hence, planning frameworks are developed to optimally allocate these units to distribution networks. However, current planning mechanisms do not consider the relative effect of different allocated units in planning frameworks. To bridge this gap, this paper presents a novel comprehensive planning framework for allocating DG units, BESS units, and Electric Vehicle Charging Station (EVCS) facilities in a distribution network while optimizing its technical, economic, and environmental benefits. The proposed framework uses a recombination technique to generate more solutions by dynamically updating the DG and BESS units’ locations in one iteration. A Reinforcement Learning (RL)-based algorithm is introduced to coordinate EV charging that suggests the optimal EVCS location in relation to other units’ locations. To cope with the complexity ensuing from searching a larger solution space, a multi-stage, hybrid optimization scheme is developed to produce optimal allocation variables. A category-based multiobjective framework is further developed to simultaneously optimize many objective functions — power loss, voltage stability, voltage deviation, installation and operation cost, and emission cost. Through numerical simulations on the IEEE 33- and 118-bus distribution network, it is shown that the proposed optimization scheme achieves higher metric values than the adopted benchmark optimization schemes. A validation process was also carried out on the proposed multiobjective optimization approach, comparing it with other approaches. Using the Spacing metric, the proposed approach proves to be efficient, depicting a good spread of Pareto optimal solutions.

Progress toward larger molecular simulation on a quantum computer: Simulating a system with up to 28 qubits accelerated by point-group symmetry

The exact evaluation of the molecular ground state in quantum chemistry requires an exponentially increasing computational cost. Quantum computation is a promising way to overcome the exponential problem using polynomial-time quantum algorithms. A quantum-classical hybrid optimization scheme known as the variational quantum eigensolver is preferred for noisy intermediate-scale quantum devices. However, the circuit depth becomes one of the bottlenecks of its application to large molecules of more than 20 qubits. In this work, we employ point-group symmetry to reduce the number of operators in constructing ansatz so as to achieve a more compact quantum circuit. We illustrate this methodology with a series of molecules ranging from LiH (12 qubits) to ${\mathrm{C}}_{2}{\mathrm{H}}_{4}$ (28 qubits). A significant reduction of up to 82% of the operator numbers is reached on ${\mathrm{C}}_{2}{\mathrm{H}}_{4}$. This also sheds light onto further work in this direction to construct even shallower ansatz with enough expressive power and simulate even larger scale systems.

Research on Lane Change Motion Planning Steering Input Based on Optimal Control Theory

Aiming at traditional control methods are not suitable for solving the multiconstraint problem of the vehicle steering and lane-change maneuvering process, a hybrid optimization scheme based on dynamic adaptive salp swarm algorithm and Gaussian pseudospectral method (DASSA-GPM) is proposed for its strong global search capability. Based on a steering inverse dynamics model, the lane change motion planning steering problem is converted into an optimal control problem which is then converted into a nonlinear programming problem by applying the adaptive salp group algorithm which is solved through the sequential quadratic programming (SQP) method finally. The simulation results verify the effectiveness and feasibility of the proposed dynamic adaptive salp swarm algorithm and hybrid optimization scheme. In addition, compared with GPM which the absolute error of steering wheel angle is 8.2 × 10−4 degree, the proposed method has higher computational accuracy, which absolute error of steering wheel angle is 5.5 × 10−4 degree. At the same time, under the same calculation accuracy (10−3), compared with GPM which CPU time is 4.81 s, the proposed method has higher calculation efficiency which CPU time is 3.86 s when solving non-smooth problems. The proposed method can provide a reference value into the active safety of manned and unmanned vehicles.

A computationally efficient hybrid optimization‐based model predictive control for inductive power transfer systems

Model predictive control (MPC) has been actively researched in recent years for power electronics applications due to its intuitive concept, flexibility, and superior dynamic performance. However, current research on MPC for inductive power transfer (IPT) systems is rather limited. Several emerging applications such as dynamic electric vehicle charging, vehicle-to-grid services, and ad hoc power transfer between mobile devices require IPT systems to possess fast dynamic response characteristic. Here, a new MPC method based on computationally efficient hybrid optimization scheme is proposed to meet the needs of these applications. The proposed MPC method adaptively selects the moving discretized control set-based optimization or the newly proposed group-based optimization under different system's states to minimize the number of iterations required for determining the optimum control variable, thus offering the advantage of low computational burden. The paper also proposes a new prediction error compensation scheme that effectively improves the control accuracy of the proposed MPC method. All the proposed works are experimentally verified on a laboratory prototype.

An elitist control scheme for power flow management in smart grid system: a hybrid optimization scheme

ABSTRACT In this manuscript, an elitist power flow management (PFM) scheme of PV/FC/battery/SCAP in smart grid (SG) system is proposed. The proposed structure has photovoltaic (PV) generator (Maximum Power Point Tracking (MPPT) and power limitation mode), battery storage, fuel cell (FC), and super capacitor (SCAP). The proposed method is the consolidation of binary adaptation of ant lion optimizer (BALO) and squirrel search optimizer (SSO), and hence it is known BALSO method. Here, BALO is generated the inverter control pulses depends on power transfer an assortments among source and load side. The formulation of multiobjective function in terms of available source power through grid required generated active with reactive power variety. Here, SSO is utilized to obtain that online control pulse using power flow varieties. Moreover, a power flow management method as per the global state of charge (SOC) of energy storage and load demand is proposed. Finally, the BALSO system is performed on MATLAB/Simulink work site. The efficiency of the BALSO system is analyzed through other existing systems. The elapsed time with proposed and existing technique is evaluated.

Augmented reality navigation for liver surgery: an enhanced coherent point drift algorithm based hybrid optimization scheme

Augmented reality navigation for liver surgery: an enhanced coherent point drift algorithm based hybrid optimization scheme

Fast Computation of Latent Correlations

Latent Gaussian copula models provide a powerful means to perform multi-view data integration since these models can seamlessly express dependencies between mixed variable types (binary, continuous, zero-inflated) via latent Gaussian correlations. The estimation of these latent correlations, however, comes at considerable computational cost, having prevented the routine use of these models on high-dimensional data. Here, we propose a new computational approach for estimating latent correlations via a hybrid multilinear interpolation and optimization scheme. Our approach speeds up the current state of the art computation by several orders of magnitude, thus allowing fast computation of latent Gaussian copula models even when the number of variables p is large. We provide theoretical guarantees for the approximation error of our numerical scheme and support its excellent performance on simulated and real-world data. We illustrate the practical advantages of our method on high-dimensional sparse quantitative and relative abundance microbiome data as well as multi-view data from The Cancer Genome Atlas Project. Our method is implemented in the R package mixedCCA, available at https://github.com/irinagain/mixedCCA.

Multiphase Trajectory Optimization of a Lunar Return Mission to an LEO Space Station

Lunar exploration architecture can be made more flexible and reliable with the support of a low-Earth orbit (LEO) space station. This study therefore evaluated a proposed hybrid optimization scheme to design the entire trajectory of a reusable spacecraft starting from trans-Earth injection (EI) at the perilune and ending at an LEO space station. As such a trajectory has multiple constraints and multiple dynamical models, it is divided into the trans-Earth phase, aerocapture phase, and postatmospheric phase. The optimization scheme is performed at two levels: sublevel and top level. At the sublevel, two novel pseudo rules are proposed to optimize the trans-Earth trajectory so that it satisfies the coplanar constraints of the space station. Then, in the aerocapture phase, the bank angle is optimized to satisfy the mission constraints, and in the atmospheric phase, the one-impulsive maneuver is performed and optimized to insert the spacecraft into the target space station orbit. The multiple phases are connected to each other by boundary conditions where the terminal state of the previous phase is transformed into the initial state of the following phase. At the top level, the vacuum perigee height is selected as the mission design variable based on problem characteristics analysis and a hybrid optimization scheme is conducted to minimize the total velocity increment. The simulation results demonstrate that the proposed hybrid optimization method is effective for the design of an entire trajectory with acceptable velocity cost which is less than that in the previous study. The coplanar constraints of the space station and other mission constraints in each phase are also satisfied. Furthermore, the proposed trajectory design method is shown to be applicable to a reusable spacecraft returning to an LEO space station parked in any arbitrary orbital plane.

Investigation of real-time optimization based on CAN bus

Aiming at the problem of CAN bus network transmission delay which affects the transmission rate, a bus network transmission delay model is established. The main indicators of network transmission delay are modeled. A hybrid optimization combining message offset compensation and identifier optimization is proposed. In terms of overall optimization, an optimization method for message segmentation and encoding scheme is designed, and a method for offset compensation of delayed messages is designed for partial parts. Symtavision software is used to simulate network conditions of the two CAN buses. Analyse the main indicators of delay after the optimization plan is adopted. The simulation results show that the hybrid optimization scheme can reduce the delay by more than 16% compared with before, which significantly improves the network communication quality, and improves the network transmission rate and scalability.

Cardinality constrained portfolio optimization with a hybrid scheme combining a Genetic Algorithm and Sonar Inspired Optimization

The constraints and the vast solution space of operational research optimization problems make them hard to cope with. However, Computational Intelligence, and especially Nature-Inspired Algorithms, has been a useful tool to tackle hard and large space optimization problems. In this paper, a very consistent and effective hybrid optimization scheme to tackle cardinality constrained portfolio optimization problems is presented. This scheme consists of two nature-inspired algorithms, i.e. Sonar Inspired Optimization algorithm and Genetic Algorithm. Also, the incorporation of heuristic information, i.e. an expert’s knowledge, etc., to the overall performance of the hybrid scheme is tested and compared to previous studies. More specifically, under the framework of a financial portfolio optimization problem, the heuristic information-enhanced hybrid scheme manages to reach a new optimal solution. Additionally, a comparison of the proposed hybrid scheme with other hybrid schemes applied to the same problem with the same data is performed.