Common Manifold Research Articles

(Invited) Shunt-currents in Alkaline Water-Electrolyzers and Renewable Energy

Shunt-currents in bipolar cell stack design with a common electrolyte-feed and -outlet is an inevitable physical phenomenon governed by Ohms law that causes some extra challenges when alkaline water-electrolysers shall operate on renewable energy that is both dynamic and intermittent. Shunt-currents are also referred to as bypass-current or creep-current. The shunt-currents are to much degree governed by the electrolyte inside the cell stack manifold system that transports lye in and out of the cells. The electrolyte inside the manifold system also electrically connects the individual cells together and enables transport of ions/electricity in parallel to the cell stack. Thus, a small portion of the rectifier current is being shunted outside the cells and does not contribute to any production. Previous work on shunt-current modelling brought new insight on how to design the manifolds and raised the awareness on shunt-current and the use of metallic manifolds which both reduced the ohmic resistivity of the manifold system and was a source of secondary electrolysis.Classic alkaline water-electrolysers are typically using an internal manifold system where the inlet ports are located at the bottom of the cell stack and the outlet ports are located at the top, and where the ports connecting the cell-interior and the common manifold channel are short and straight. Such design has in the past worked satisfactory for alkaline water-electrolysers that have been working on a high nominal load and only being shut-down for maintenance a few times over the stack-lifetime, mainly causing a modest reduction in the current efficiency. Membrane-chlorine electrolysers on the other hand, are designed for very small shunt-currents by using an external manifold system, which enables a current protection system that protects electrodes from corrosion under shutdown conditions.Shunt-currents bypassing the cell stack does not contribute to any product and therefore constitutes a loss in current efficiency and, hence, an accordingly loss in the energy efficiency (the shunt-currents still adds to the electricity bill). The atmospheric alkaline electrolyser has a modest loss in current efficiency at nominal load where the high gas volume blocks much of the current path in the rather open outlet ports. The high-pressure alkaline electrolysers on the other hand, where the gas volumes are much smaller, the current efficiency can be as low as 89% at nominal load due to shunt-currents when using simple internal manifold system. For an electrolyser with already low current efficiency at the nominal load, the current efficiency will drop dramatically as the electrolyser is taken to lower load, severely compromising the energy efficiency. The impact on shunt-currents also dramatically increases for increased number of cells in a cell stack, and eventually limits the number of cells that can be assembled in one single cell stack operating on the same common lye system.Shutdown and discharge of the electrodes may further lead to corrosion and degradation of the electrodes, strongly influenced by the shunt-currents and the manifold system. Large cell stacks will discharge faster and deeper, eventually causing corrosion of the electrodes. As the cell stack is discharged the current in the cell stack is reversed where the hydrogen electrode is being polarized to anodic potentials, and the oxygen electrode is polarized to low cathodic potentials which eventually may challenging the material stability. Thus, evaluation of electrode potentials must be an integral part of development of industrial electrodes [LeRoy] and especially in intermittent operation where frequent shutdown will occur. A good integration of the manifold system into the cell stack can potentially mitigate both the loss of current efficiency under dynamic operation and the electrode corrosion under intermittent operation.

A weighted prior tensor train decomposition method for community detection in multi-layer networks

Community detection in multi-layer networks stands as a prominent subject within network analysis research. However, the majority of existing techniques for identifying communities encounter two primary constraints: they lack suitability for high-dimensional data within multi-layer networks and fail to fully leverage additional auxiliary information among communities to enhance detection accuracy. To address these limitations, a novel approach named weighted prior tensor training decomposition (WPTTD) is proposed for multi-layer network community detection. Specifically, the WPTTD method harnesses the tensor feature optimization techniques to effectively manage high-dimensional data in multi-layer networks. Additionally, it employs a weighted flattened network to construct prior information for each dimension of the multi-layer network, thereby continuously exploring inter-community connections. To preserve the cohesive structure of communities and to harness comprehensive information within the multi-layer network for more effective community detection, the common community manifold learning (CCML) is integrated into the WPTTD framework for enhancing the performance. Experimental evaluations conducted on both artificial and real-world networks have verified that this algorithm outperforms several mainstream multi-layer network community detection algorithms.

(Invited) Shunt-currents in Alkaline Water-Electrolyzers and Renewable Energy

Shunt-currents are inevitably an integrated part of a bipolar electrolyzer with a common manifold system for distribution of the electrolyte in and out of the different cell compartments. Shunt-currents represent some extra challenges when operating alkaline water-electrolyzers (AWE) with renewable energy, both with respect to energy efficiency and reduced lifetime of electrodes. These challenges multiply for increasing length of the electrolyzer-stack and limits the number of cells in the stack. This paper will explain the challenges in more detail, how it limits the operation with renewables and what are the important design criteria to mitigate the impact of shunt-currents.

An improved transfer learning approach based on geodesic flow kernel for multiphase batch process soft sensor modeling

For multiphase batch process, the characteristics of process data under various batches differ. Consequently, the soft sensor model built for a particular working condition is inapplicable to other working conditions. Besides, each batch can be divided into several phases whose characteristics are probably different. To address the above problems, a soft sensor model based on phase division and transfer learning strategy is proposed. First, transfer learning strategy is adopted to construct a soft sensor model applicable to various working conditions. Specifically, geodesic flow kernel based on linear local tangent space alignment (LLTSA-GFK) algorithm is designed. By projecting process data to the common manifold subspace, the distribution difference of data between various batches is reduced and the performance of the soft sensor model is enhanced. In addition, sequence-based fuzzy clustering and just-in-time learning (JITL) are adopted to solve the multistage characteristic for batch process. The root-mean-square error ( RMSE), coefficient of determination [Formula: see text], and mean absolute error ( MAE) are adopted to compare the conventional soft sensing approach (i.e., partial least-square regression based on JITL, support vector regression, and back propagation neural network) with the proposed approach. The superiority of the proposed model is verified by a fed-batch penicillin fermentation process.

How do kernel-based sensor fusion algorithms behave under high-dimensional noise?

Abstract We study the behavior of two kernel based sensor fusion algorithms, nonparametric canonical correlation analysis (NCCA) and alternating diffusion (AD), under the nonnull setting that the clean datasets collected from two sensors are modeled by a common low-dimensional manifold embedded in a high-dimensional Euclidean space and the datasets are corrupted by high-dimensional noise. We establish the asymptotic limits and convergence rates for the eigenvalues of the associated kernel matrices assuming that the sample dimension and sample size are comparably large, where NCCA and AD are conducted using the Gaussian kernel. It turns out that both the asymptotic limits and convergence rates depend on the signal-to-noise ratio (SNR) of each sensor and selected bandwidths. On one hand, we show that if NCCA and AD are directly applied to the noisy point clouds without any sanity check, it may generate artificial information that misleads scientists’ interpretation. On the other hand, we prove that if the bandwidths are selected adequately, both NCCA and AD can be made robust to high-dimensional noise when the SNRs are relatively large.

Full-scale investigation of the short-term continuous and intermit-tent operation of an energy piled wall section

The growing demand for highly efficient renewable energy technologies has positioned shallow geothermal energy as an attractive alternative. The initial high capital cost of traditional ground heat exchangers (GHEs), mainly associated with drilling, has extensively hindered their implementation. Conventional geostructures overcome this drawback, and their use as GHEs is getting more attention. Piles [4, 8, 10], pavements [7], and walls [1, 5] are some of the structures that are being used. Energy piles have been the most popular among them, and their implementation has increased steadily since the 1980's [5]. After more than two decades of intensive research, a large dataset of full-scale energy pile foundation observations is available [6], and a thermo-mechanical conceptual framework [2, 3] has been formulated.
 Piles are not limited to foundations; they are used in retaining walls. In contrast to foundations, retaining walls mainly work under lateral stress conditions instead of axial loads. In principle, energy piled retaining wall's initial state, circuits configuration and boundary conditions control their response. Nevertheless, in the context of (energy) piles, observations appear limited to a few full-scale thermo-mechanical studies to date [1, 5]. More attention to this energy geostructure from the scientific and engineering practice communities is required, and the documentation of additional case studies is needed to promote the energy piled walls implementation more broadly.
 In this work, a series of full-scale thermo-mechanical field testing has been performed in an energy piled wall section. The retaining wall system section consists of three piles equipped with pipes to work as heat exchangers. Each houses 4 U-loops in series made of high-density polyethylene (HDPE) pipe with an outer diameter of 25 mm and HDR of 11. The pipe circuits are connected in parallel through a common header manifold operated on the surface. Two of these piles have been instrumented with 10 pairs of vibrating wire rebar strainmeters with temperature sensors located on diametrically opposing sides of the pile. One is close to the excavation, whereas the other is to the ground side - see more details in [10].
 The thermal activation of the wall section occurred at 34.7 m depth and was achieved through an in-situ Thermal Response Test (TRT) unit. The unit heats a carrier fluid through constantly powered electrical heating components and pumps the fluid in the pipe loops within the GHEs. To emulate a conventional thermal load (e.g., energy piles 30-70 W/m, [9]), the entire wall section is supplied with a heating power of 5.5 kW over 5 days active period, and the thermal load is provided under either continuous or intermittent conditions (i.e., 5 thermal cycles, 12 hours on/off). Five recovery days allowed the ground to return partially to its initial thermal equilibrium. The testing period lasted 26 days of constant monitoring.
 The obtained measurements in one pile are presented in Figure 1. It compares the measured strains at different construction stages (i.e., excavation to 32.5, 36.4, and 34.7 m depths) against those observed at the end of both operation modes (i.e., 5 days, 5th cycle). It is noticed that the temperature-induced strains recorded by the available sensors (i.e., 11/20 in this case) are, in most cases, within the axial strain's envelope derived from previous excavation and construction stages. Only one sensor at 33 m indicates tensile axial strains larger than those observed at different excavation depths; nevertheless, it represents a minor deviation compared with the increment recorded at the same depth during the last excavation and anchors removal. Based on these observations and assuming a thermoelastic response, thermally induced axial stresses remain within the design ranges delimited by design envelopes.

A Quantum-Behaved Particle Swarm Optimization Algorithm on Riemannian Manifolds

The Riemannian manifold optimization algorithms have been widely used in machine learning, computer vision, data mining, and other technical fields. Most of these algorithms are based on the geodesic or the retracement operator and use the classical methods (i.e., the steepest descent method, the conjugate gradient method, the Newton method, etc.) to solve engineering optimization problems. However, they lack the ability to solve non-differentiable mathematical models and ensure global convergence for non-convex manifolds. Considering this issue, this paper proposes a quantum-behaved particle swarm optimization (QPSO) algorithm on Riemannian manifolds named RQPSO. In this algorithm, the quantum-behaved particles are randomly distributed on the manifold surface and iteratively updated during the whole search process. Then, the vector transfer operator is used to translate the guiding vectors, which are not in the same Euclidean space, to the tangent space of the particles. Through the searching of these guiding vectors, we can achieve the retracement and update of points and finally obtain the optimized result. The proposed RQPSO algorithm does not depend on the expression form of a problem and could deal with various engineering technical problems, including both differentiable and non-differentiable ones. To verify the performance of RQPSO experimentally, we compare it with some traditional algorithms on three common matrix manifold optimization problems. The experimental results show that RQPSO has better performance than its competitors in terms of calculation speed and optimization efficiency.

Learning Brain Dynamics of Evolving Manifold Functional MRI Data Using Geometric-Attention Neural Network.

Functional connectivities (FC) of brain network manifest remarkable geometric patterns, which is the gateway to understanding brain dynamics. In this work, we present a novel geometric-attention neural network to characterize the time-evolving brain state change from the functional neuroimages by tracking the trajectory of functional dynamics on high-dimension Riemannian manifold of symmetric positive definite (SPD) matrices. Specifically, we put the spotlight on learning the common state-specific manifold signatures that represent the underlying cognition. In this context, the driving force of our neural network is tied up with the learning of the evolution functionals on the Riemannian manifold of SPD matrix that underlies the known evolving brain states. To do so, we train a convolution neural network (CNN) on the Riemannian manifold of SPD matrices to seek for the putative low-dimension feature representations, followed by an end-to-end recurrent neural network (RNN) to yield the time-varying mapping function of SPD matrices which fits the evolutionary trajectories of the underlying states. Furthermore, we devise a geometric attention mechanism in CNN, allowing us to discover the latent geometric patterns in SPD matrices that are associated with the underlying states. Notably, our work has the potential to understand how brain function emerges behavior by investigating the geometrical patterns from functional brain networks, which is essentially a correlation matrix of neuronal activity signals. Our proposed manifold-based neural network achieves promising results in predicting brain state changes on both simulated data and task functional neuroimaging data from Human Connectome Project, which implies great applicability in neuroscience studies.

Tensor-based multi-view clustering with consistency exploration and diversity regularization

How to make good use of information from all views is the core problem in multi-view clustering (MVC), and widely recognized experience of attaining this goal is to leverage consistent and complementary principles. However, current methods do not take full advantage of these two principles. On the one hand, some existing methods ignore the connection between multiple views and the influence of manifold information on consensus representation. On the other hand, some others simply sum all the view-specific representations together as the final representation, which result in insufficient mining of complementary information. In this manuscript, we propose a novel tensor-based multi-view clustering approach that significantly improves consistency exploration and diversity capturing. In particular, we utilize the low-rank tensor learning with tensor-based singular value decomposition (t-SVD), aiming at capturing the correlations among different views. Moreover, we further split the low-rank tensor into a consistent matrix and a set of view-specific matrices, where the former is responsible for exploring the common manifold information via graph regularization, and the latter is used to mine complementary information through learning the view-specific representations as various as possible. We integrate low-rank tensor learning, consistency exploration and diversity regularization into a whole framework and use an alternating direction minimization strategy to optimize it. Experiments conducted on eight benchmark datasets show that our approach gains superior performance over several state-of-the-arts.

Multi-manifold discriminant local spline embedding

Manifold learning reveals the intrinsic low-dimensional manifold structure of high-dimensional data and has achieved great success in a wide spectrum of applications. However, traditional manifold learning methods assume that all the data lie on a common manifold, hence fail to capture the complicated geometry structure of the real-world data lying on multiple manifolds. This paper proposes a novel Multi-manifold Discriminant Local Spline Embedding (MDLSE) algorithm for high-dimensional classification, which considers a more realistic scenario where data of the same class lies on the same manifold. On the basis of this assumption, MDLSE seeks to reconstruct multiple manifolds for different classes of data in the embedding and separate them as apart as possible. In order to preserve the geometry structure of all the manifolds, MDLSE employs thin plate splines to align the local patches within each manifold compatibly in the global embedding. Meanwhile, to separate the different manifolds, MDLSE utilizes discriminative information to ensure the neighboring data from different manifolds to be mapped far from each other. Extensive experiments on both synthetic and real-world datasets demonstrate the superiority of MDLSE over the other representative manifold learning algorithms. The advantage of MDLSE is often more obvious on smaller size of training data and in lower embedding dimensions.

Procedures for cognitive and synergetic observation and proactive control of power capacity of nickel-hydrogen storage batteries onboard a geostationary spacecraft

The paper presents methods of providing power resources for the mission of the Yamal-100 spacecraft from the secondary onboard power supply sources — nickel-hydrogen storage batteries. For the first time in the world practice of using metal-hydrogen electrochemical batteries installed in a common gas manifold a ~10-years operational life in orbit has been achieved for batteries operated onboard a geostationary satellite. A decline in power capacity of storage batteries could result in early termination of the spacecraft mission. Therefore, the energy resource of the storage battery was maintained through proactive management consisting of prediction and active control. The methods that were developed for continuous observation in the form of monitoring and diagnostics of the storage batteries status including estimation of the current discharge energy capacity made it possible to predict further degradation loss under conditions of insufficient power for the storage batteries cooling equipment. The active control consisted in using methods aimed at providing optimal operating conditions, fending off a potential development of an anomalous thermal situation within the electrochemical reaction zone, and regenerating the energy capacity. The control furthermore used the synergy effect produced by secondary energy interrelationships between onboard systems. This made it possible to meet the batteries operational life requirements under exposure to an unfavorable external factor in the form of an increasing heat impact on the batteries temperature control equipment. The increase in thermal load was caused by re-reflection of a portion of the luminous flux from structural elements of the spacecraft and an increase in solar absorption factor (aging) of thermal control coatings. Key words: spacecraft, control process, nickel-hydrogen battery, provision of resources, energy capacity, proactive control, synergy.

Compartmentalized dynamics within a common multi-area mesoscale manifold represent a repertoire of human hand movements

The human hand is unique in the animal kingdom for unparalleled dexterity, ranging from complex prehension to fine finger individuation. How does the brain represent such a diverse repertoire of movements? We evaluated mesoscale neural dynamics across the human "grasp network," using electrocorticography and dimensionality reduction methods, for a repertoire of hand movements. Strikingly, we found that the grasp network represented both finger and grasping movements alike. Specifically, the manifold characterizing the multi-areal neural covariance structure was preserved during all movements across this distributed network. In contrast, latent neural dynamics within this manifold were surprisingly specific to movement type. Aligning latent activity to kinematics further uncovered distinct submanifolds despite similarities in synergistic coupling of joints between movements. We thus find that despite preserved neural covariance at the distributed network level, mesoscale dynamics are compartmentalized into movement-specific submanifolds; this mesoscale organization may allow flexible switching between a repertoire of hand movements.

Learning sufficient scene representation for unsupervised cross-modal retrieval

In this paper, a novel unsupervised Cross-Modal retrieval method via Sufficient Scene Representation (CMSSR) is proposed. Distinguished from the existing methods which mainly focus on simultaneously preserving the mutually-constrained intra- and inter-modal similarity relation, CMSSR considers data of different modalities as the descriptions of a scene from different views and accordingly integrates information of different modalities to learn a complete common representation containing sufficient information of the corresponding scene. To obtain such common representation, Gaussian Mixture Model (GMM) is firstly utilized to generate statistic representation of each uni-modal data, while the uni-modal spaces are accordingly abstracted as uni-modal statistical manifolds. In addition, the common space is assumed to be a high-dimensional statistical manifold with different uni-modal statistical manifolds as its sub-manifolds. In order to generate sufficient scene representation from uni-modal data, a representation completion strategy based on logistic regression is proposed to effectively complete the missing representation of another modality. Then, the similarity between different multi-modal data can be more accurately reflected by the distance metric in common statistical manifold. Based on the distance metric in common statistical manifold, Iterative Quantization is utilized to further generate binary code for fast cross-modal retrieval. Extensive experiments on three standard benchmark datasets fully demonstrate the superiority of CMSSR compared with several state-of-the-art methods.

Upper extremity prosthesis 3D bio-pneumatic self-adjusting clamping to non-homogeneous surfaces under the principle of communicating vessel

There are several models that emulate the mechanical behavior of biological tissues. Its automation, new materials and manufacturing techniques will enable its use in the near future. This research is based on the use of compressed air in pneumatic muscles that actuate the phalanges. The research is important because it proposes a system that adapts to the needs and economic accessibility or people with limited resources. The problem to solve is that it meets certain standards such as: stable grip and pressure and that it adapts to irregularly shaped objects with more natural movements. The principle of communicating vessels and the force exerted by a fluid on the walls of the container that is used. Consequently, the fluid exerts pressure in all directions. The prosthesis with the design of the Flexy Hand 2 is manufactured by inserting pneumatic muscles to each of the fingers connecting them by means of nylon ropes that are attached to the homemade pneumatic muscles connected to a common manifold, regulating the pressure by means of a valve and a degree of freedom. As a result of the above, it can be concluded that the prototype worked favorably.

Modeling the Effects of the Inlet Manifold Design on the Performance of a Diesel Oxidation Catalytic Converter

The effect of inlet flow distribution on the performance characteristics of a catalytic converter is analyzed for two common inlet manifold designs. Three-dimensional steady-state computational flu...

Modeling pressure relief devices mounted on a common inlet manifold

AbstractLarge equipment items, such as distillation column systems, compressors, or major pressure vessels, are commonly protected by multiple pressure relief devices mounted on a common inlet manifold. In selecting this type of design, the potential exists to inadvertently overlook the flow characteristics associated with such a common inlet manifold. The methodology discussed in this paper should effectively model flow through multiple pressure relief devices mounted on common inlet manifolds. This approach ensures accurate representation of flow through each pressure relief device while avoiding the potential pitfalls of taking shortcuts by only analyzing flow through one device.

Multi-View Representation Learning via Dual Optimal Transportation

Recently, multi-view representation learning has gained rapid growth in various fields. Most of previous multi-view learning methods rely on strong notions of distances that often provide no useful gradients in deep network training, which greatly degrades the performance in merging complementary information of views. To address this challenge, a multi-view representation learning network with dual optimal transportation (MDOT-Net) is proposed to capture fusion representations embedded in a common manifold with the weak topology. In MDOT-Net, the multi-view representation learning is modelled as an optimal transportation (OT) problem in manifold fitting, which is further decomposed into the intra-view OT and the inter-view OT. The intra-view OT is implemented by a view-specific adversarial variational network, which accurately captures local manifold structures within views by leveraging fusion knowledge. The inter-view OT is implemented by a view-fusion adversarial inference network, which models fusion representations compatible with diversities of sub-manifolds by utilizing view-specific knowledge. The two OTs boost mutually by providing prior knowledge to each other in multi-view representation learning. Finally, numerous experiments are conducted on four benchmark datasets, and the results demonstrate MDOT-Net achieves the state-of-the-art performance.

Coordination of a Flywheel Energy Storage Matrix System: An External Model Approach

This paper studies the coordination of a heterogenous flywheel energy storage matrix system aiming at simultaneous reference power tracking and state-of-energy balancing. It is first revealed that this problem is solvable if and only if the state-of-energy of all the flywheel systems synchronize to a common time-varying manifold governed by a nonautonomous dynamic system. Next, by treating this nonautonomous dynamic system as an external model, the coordination problem can be decoupled into two separate problems, namely, the global double layer estimation problem and the local tracking problem. Then, a distributed control scheme is proposed to solve the coordination problem by integrating the adaptive distributed observer approach and the certainty equivalence control method. Comprehensive case studies are provided to show the performance of the proposed control scheme.

Semisupervised charting for spectral multimodal manifold learning and alignment

For one given scene, multimodal data are acquired from multiple sensors. They share some similarities across the sensor types (redundant part of the information, also called coupling part) and they also provide modality-specific information (dissimilarities across the sensors, also called decoupling part). Additional critical knowledge about the scene can hence be extracted, which is not extractable from each modality alone. For the processing of multimodal data, we propose in this paper a model to simultaneously learn the underlying low-dimensional manifold in each modality, and locally align these manifolds across different modalities. For each pair of modalities we first build a common manifold that represents the corresponding (redundant) part of information, ignoring non-corresponding (modality specific) parts. We propose a semi-supervised learning model, using a limited amount of prior knowledge about the coupling and decoupling components of the different modalities. We propose a localized version of Laplacian eigenmaps technique specifically designed to handle multimodal manifold learning, in which the ideas of local patching of the manifolds, also known as manifold charting, is combined with the joint spectral analysis of the graph Laplacians of the different modalities. The limited given supervised information is then extending on the manifold of each modality. The idea of functional mapping is finally used to align the different manifolds across modalities. The evaluation of the proposed model using synthetic and real-world multimodal problems shows promising results, compared to several related techniques.

Cross-domain transfer person re-identification via topology properties preserved local fisher discriminant analysis

Person re-identification (Re-ID) systems aim to identify a person appeared in non-overlapping cameras. However, a sufficient amount of pairwise cross-camera-view person images are often not available in a new scenario. Transfer learning can assist the new Re-ID system through leveraging knowledge from other related scenarios. Since the images from existed scenarios may not be the exact representative samples, how to learn a robust transfer Re-ID model with limited labeled person images is still a challenge so far. To solve this problem, a novel cross-domain transfer person Re-ID via topology properties preserved local Fisher discriminant analysis (TPPLFDA) method is proposed in this paper. Making an assumption that all person images in the new and related scenarios share common manifold, TPPLFDA projects all cross-domain images into a low dimensional linear subspace, while preserves the topology properties according to the geodesic distances on manifold. Then, multiple cross-domain datasets as source domains are considered and kernel TPPLFDA for multi-source domain transfer is proposed, so that TPPLFDA can handle more complex cross-domain transfer Re-ID tasks. Extensive experiments on several transfer Re-ID datasets show that TPPLFDA is effective.