State Aggregation Research Articles

An adaptive decision-making method with fuzzy Bayesian reinforcement learning for robot soccer

A robot soccer system is a typical complex time-sequence decision-making system. Problems of uncertain knowledge representation and complex models always exist in robot soccer games. To achieve an adaptive decision-making mechanism, a method with fuzzy Bayesian reinforcement learning (RL) is proposed in this paper. To extract the features utilized in the proposed learning method, a fuzzy comprehensive evaluation method (FCEM) is developed. This method classifies the situations in robot soccer games into a set of features. With the fuzzy analytical hierarchy process (FAHP), the FCEM can calculate the weights according to defined factors for these features, which comprise the dimensionality of the state space. The weight imposed on each feature determines the range of each dimension. Through a Bayesian network, the comprehensively evaluated features are transformed into decision bases. An RL method for strategy selection over time is implemented. The fuzzy mechanism can skillfully adapt experiences to the learning system and provide flexibility in state aggregation, thus improving learning efficiency. The experimental results demonstrate that the proposed method has better knowledge representation and strategy selection than other competing methods.

Dynamic scheduling for wireless multicast in massive MIMO HetNet

Consider a massive multiple input and multiple output (MIMO) heterogeneous network (HetNet) scenario where the macro base station is coupled with low-power nodes (LPNs) to provide wireless multicast transmission to user equipments (UEs) within a single macro cell. The Markov decision process (MDP) model is adopted to model the dynamic scheduling problem introduced by the mobility of UEs. With asymptotic conclusion for the massive MIMO multicast transmission, we propose a MDP-based off-line scheduling problem with determined system parameters and derive a linear programming (LP) model to solve the problem. As for the system with undetermined parameters, we adopt Q-learning algorithm to solve the optimal policy. To derive the low-complexity algorithm, the state aggregation method is utilized to build an series of sub-problems to estimate the original complex MDP problem. Finally we use the MDP Toolbox in MATLAB to run simulations to test the proposed algorithm. The performance of the proposed algorithm is shown and analyzed in details.

Diverse silver(i) coordination chemistry with cyclic selenourea ligands.

The coordination chemistry of two selenourea ligands (SeIMes and SeIPr) towards silver(i) triflate and silver(i) nitrate was investigated. Two aggregation modes were observed in the solid state, strongly influenced by the size of the aromatic substituents on the ligand. With mesityl groups, selenium-bridged bimetallic motifs [AgX(SeIMes)]2 were obtained, while for the bulkier diisopropylphenyl groups ion-separated species of formulae [Ag(SeIPr)2]+[X]- were obtained. Recrystallization of [Ag(NO3)(SeIMes)]2 from hot methanol resulted in the formation of a unique coordination polymer featuring three silver environments. Characterization of the complexes by NMR spectroscopy and mass spectrometry suggested all complexes adopt the ionic aggregation mode in methanol solution.

Catalyst Layer Ink Interactions That Affect Coatability

Catalyst layer inks are examples of biphasic material systems composed of a solid material, a polymer, and a solvent. Nanoscale interactions between the individual constituents can alter macroscopic properties that are relevant for coating and manufacturing processes (i.e. viscosity, surface tension, aggregation, and rheology). Control over these macroscale properties are important for controlled electrode formation during scalable roll-to-roll manufacturing. The underlying interactions include polymer|particle, particle|solvent, and polymer|solvent interactions. In this work we systematically investigate polymer|particle interactions via studying a range of formulated inks composed of different solvents (methanol, isopropyl alcohol, octanol, and water), varying polymer loadings, and particles with different surface charges. Ink aging is also addressed and over short time periods (<1 hr shelf life) the addition of a perfluorosulfonic acid ionomer was shown to stabilize the ink and also decrease the aggregation size. However, over long time periods (168 hrs) the aggregation size is independent of polymer loading, and approaches a steady state aggregation size around 350 nm. This equilibrium point suggests that the polymer is free to diffuse, adsorb, and relax within the excluded volume region. Furthermore, these results suggest that primary aggregates can be broken up with the addition of very low polymer loadings (15% I:C). A semi-empirical model is used to describe polymer|particle interactions within the ink, and the polymer coverage at the surface of the carbon was found to be the most sensitive parameter dictating ink stability. Finally, coating and rheology experiments are completed on all inks.

Effects of high hydrostatic pressure on Rhizopus chinensis lipase: II. Intermediate states during unfolding

Abstract High hydrostatic pressure (HHP) is currently considered a well-established technology for processing food and biological materials and there is an interest in investigating the changes in the structural and functional properties of these materials after high pressure treatment. Therefore, the changes in the structure of Rhizopus chinensis lipase (RCL) after high hydrostatic pressure treatment were investigated. Far-UV circular dichroism (CD) spectra showed that the secondary structure of RCL is maintained at pressures below 400 MPa and becomes gradually disordered after higher pressures are applied. Near-UV CD spectra showed that the RCL begins to lose its tertiary structure at pressure over 400 MPa. Fluorescence quenching and the binding of 1-anilinonaphtha-lene-8-sulfonate confirmed that a partially unfolded intermediate, with loosely compacted conformation and hydrophobic regions, is formed at a pressure of 600 MPa. These results also suggest that RCL maintains a native-like state at pressures below 400 MPa. Above 500 MPa RCL molecules showed characteristics of being in a molten globule state. Dynamic light scattering (DLS) and atomic force microscopy (AFM) measurements indicated that RCL molecules at these pressures are aggregating. The addition of (NH4)2SO4 to the protein solution could prevent the aggregation, and at 600 MPa the molecule had a hydrodynamic radius approximately 8% larger than that observed for the control sample, which was regarded as being in the molten globule state. The observations suggest that at increasing pressures, the unfolding mechanism of RCL follows well-defined steps from a native state via a native-like structure ending in molten globular state or molecular aggregation.

Tracking hydrophobicity state, aggregation behaviour and structural modifications of pork proteins under the influence of assorted heat treatments

Structural modifications of pork proteins under an assortment of industrial heat treatments were studied. With raw as control, assorted heat treatments involved were 58, 80, 98 and 160°C for 72min, 118°C for 8min and 58°C for 17h, resembling most common processing procedures. Protein denaturation, surface protein hydrophobicity state and protein aggregation behaviour were investigated. Modifications and molecular chemistry in protein structures were tracked by Fourier Transform Infrared Spectroscopy in order to extract relative proportions of β-sheet, α-helix and residual conformations. In comparison to uncooked samples, cooked ones showed more than two-fold increase in hydrophobicity and larger particles. Thermograms from differential scanning calorimetry showed endothermic transitions (positive enthalpy) indicating a different pattern of protein denaturation as a result of varied cooking temperatures and cooking times. Deconvolution and curve fitting procedures (R2=0.99) provided information on rise of the β-sheet to α-helix ratio that further confirmed aggregation with thermal rise and longer cooking time.

Analysis of In Situ Microscopy Images of Flocculated Sediment Volumes

AbstractAn in situ microscopy method was employed to study the sediment volumes that result from batch settling of polymer‐flocculated kaolin suspensions. Statistical methods were applied including chord length distributions, perimeter‐based fractal dimension of the void areas, white area image ratio, and gray level co‐occurrence (GLCM) features. Several GLCM features have a good linear relationship with aggregation state, and the parameters of aggregate and void chord distributions can be interpreted to help explain the flocculation results. The methods applied here are suggested for the measurement of the aggregation state or relative aggregate volume on bench or industrial scale when the settled suspension interface is not available, such as in a continuous process or reaction.

An Operational State Aggregation Technique for Transmission Expansion Planning Based on Line Benefits

This paper provides a novel technique to represent in a reduced, or compact, way temporal variability in transmission expansion planning (TEP). This reduction is handled by means of “snapshot selection.” Instead of taking into account all the possible operational states and their associated optimal power flow, a reduced group of them is selected that is representative of all the states that should have an influence on investment decisions. Considering this reduced group of operational states should lead to the same investment decisions as if all the snapshots in the target year were considered. Original operational states are compared in the space of the benefits produced by potential reinforcements considered, which are relevant drivers for investment decisions. The benefits produced by these potential reinforcements are computed based on the incremental change in operation costs resulting from their installation. A clustering algorithm is used to group together those operational states where similar line benefits are realized. Our algorithm has been tested on a modified version of the standard IEEE 24 bus system. The method produces promising results and proves to be more efficient than other snapshot selection methods used until now in computing an accurate enough selection of snapshots representing system operation variability in TEP.

FROM CLINICAL ASSESSMENTS TO POPULATION DATA: FROM VISION TO REALITY

Sidney Katz believed that structuring clinical assessments determines the quality of information and this was reflected in the 1987 Nursing Home Reform Act, which mandated all residents be comprehensively assessed using a standardized instrument. Over the last 30 years these assessments, now computerized and linked to Medicare and Medicaid claims, have transformed our knowledge of long term care populations, services and policies. Enlightened facilities stratify risk, target treatment and conduct quality improvement interventions. Researchers link assessment and utilization data, longitudinally tracking patients and create facility, county, state and year aggregates for public use. These data facilitate policy evaluation, pharmaco-epidemiology and comparative effectiveness studies. Most recently, pragmatic cluster randomized trials of quality innovations are undertaken, rivaling research being done by hospital systems. This lecture documents a career shaping this transformation of knowledge and practice, revealing the great opportunities that remain.

Insights into the anion effect on the self assembly of perylene bisimide diimidazolium salts

We studied how changing the anion affects the self-assembly of three diimidazolium salts bearing a perylene bisimide spacer. In particular we used [BF4−] and [NTf2−] based salts together with a mixed salt bearing both [I−] and [BF4−] anions. Thermal as well as concentration and temperature dependent spectroscopic investigations revealed that changing the anion impacts on solid state and solution aggregation, leading to articulate thermotropic behaviors and contrasting solvent effects, respectively. We also studied the electrochemical behavior of the salts to verify if changing the anion alters the redox properties of the salts. Finally, different anions also lead to very different aggregate morphologies, as evidenced by SEM and TEM measurements.

Optimal Power Allocation and Scheduling Under Jamming Attacks

In this paper, we consider a jammed wireless scenario where a network operator aims to schedule users to maximize network performance while guaranteeing a minimum performance level to each user. We consider the case where no information about the position and the triggering threshold of the jammer is available. We show that the network performance maximization problem can be modeled as a finite-horizon joint power control and user scheduling problem, which is NP-hard. To find the optimal solution of the problem, we exploit dynamic programming techniques. We show that the obtained problem can be decomposed, i.e., the power control problem and the user scheduling problem can be sequentially solved at each slot. We investigate the impact of uncertainty on the achievable performance of the system and we show that such uncertainty leads to the well-known exploration-exploitation tradeoff. Due to the high complexity of the optimal solution, we introduce an approximation algorithm by exploiting state aggregation techniques. We also propose a performance-aware online greedy algorithm to provide a low-complexity sub-optimal solution to the joint power control and user scheduling problem under minimum quality-of-service requirements. The efficiency of both solutions is evaluated through extensive simulations, and our results show that the proposed solutions outperform other traditional scheduling policies.

Oxide films state analysis by IR spectroscopy based on the simple oscillator approximation

Stabilization of structure-phase state in a wide temperature range is one of the most important problems of improving properties of oxide compounds. As such, the search of new effective methods for obtaining metal oxides with desired physic-chemical, electro-physical and thermal properties and their control is important and relevant. The aim of this work is identification features state of the oxide films of some metals Be, Al, Fe, Cu, Zr on the metal surface of the polycrystalline samples by infrared spectroscopy. To identify the resonance emission bands the algorithm of IR-spectra processing was developed and realized on the basis of table processor EXCEL-2010, which allow revealing characteristic resonance bands successfully and identification of inorganic chemical compounds. In the frame of simple oscillator model, resonance frequencies of normal vibrations of water and some inorganic compounds: metal oxides – Be, Al, Fe, Cu, Zr were calculated and characteristic frequencies for different states (aggregate, deformation, phase) were specified. By means of IR-spectroscopy fundamental possibility of revealing oxides films on metal substrate features state is shown, that allow development and optimization of the technology for production of the oxide films with desired properties.

Discrete-time chaotic systems synchronization based on vector norms approach

Suitable stabilization conditions obtained for continuous chaotic systems are generalized to discrete-time chaotic systems. The proposed approach, leading to these conditions for complete synchronization is based on the use of state feedback and aggregation techniques for stability studies associated with the arrow form matrix for system description. The results are successfully applied for two identical discrete-time hyper chaotic Henon maps with different orders and also for non-identical discrete-time chaotic systems with same order namely the Lozi and the Ushio maps.

Characterization of Protein Aggregation by Solid-State Nanopore

In this report, we use solid-state nanopore device to characterize native state aggregation of s-lactoglobulin variant A (βLGa) protein as a function of salt concentration, temperature, pH, and electric field strength. Silicon nitride nanopores used in this work, 5-20 nm in diameter, are fabricated by low energy ion beam sculpting method. The main component of a nanopore device is a nanopore supported by a silicon substrate which separates two PDMA chambers containing salt solution. When a charged protein molecule or a protein aggregate passes through a nanopore driven by an applied voltage across the nanopore chip, a protein aggregate that has larger volume than a single protein molecule,will generate larger current blockage, therefore a nanopore device can be used to characterize protein aggregation. We present data of βLGa self-association and aggregation at pH 4.6 as a function of temperature measured in 2M and 0.1M salt. The nanopore geometry is calibrated by a standard that has known geometry such as a long rod shaped dsDNA molecule. The volume of a translocating proteins or protein aggregates are estimated using the calibrated nanopore. We show that by using a reference dsDNA molecule, solid-state nanopore method is capable of measuring protein aggregation number and the aggregation number distribution in the conditions that is close to their native aqueous solution environment.

Controlled Markov Decision Processes with AVaR criteria for unbounded costs

In this paper, we consider the control problem with the Average-Value-at-Risk (AVaR) criteria of the possibly unbounded L1-costs in infinite horizon on a Markov Decision Process (MDP). With a suitable state aggregation and by choosing a priori a global variable s heuristically, we show that there exist optimal policies for the infinite horizon problem for possibly unbounded costs.

Escaping Depressions in LRTS Based on Incremental Refinement of Encoded Quad-Trees

In the context of robot navigation, game AI, and so on, real-time search is extensively used to undertake motion planning. Though it satisfies the requirement of quick response to users’ commands and environmental changes, learning real-time search (LRTS) suffers from the heuristic depressions where agents behave irrationally. There have introduced several effective solutions, such as state abstractions. This paper combines LRTS and encoded quad-tree abstraction which represent the search space in multiresolutions. When exploring the environments, agents are enabled to locally repair the quad-tree models and incrementally refine the spatial cognition. By virtue of the idea of state aggregation and heuristic generalization, our EQ LRTS (encoded quad-tree based LRTS) possesses the ability of quickly escaping from heuristic depressions with less state revisitations. Experiments and analysis show that (a) our encoding principle for quad-trees is a much more memory-efficient method than other data structures expressing quad-trees, (b) EQ LRTS differs a lot in several characteristics from classical PR LRTS which represent the space and refine the paths hierarchically, and (c) EQ LRTS substantially reduces the planning amount and curtails heuristic updates compared with LRTS on uniform cells.

Platelets from Fresh and Cold-Stored Whole Blood Contribute to Clot Formation after Transfusion in Rats with Acute Traumatic Coagulopathy

Background: The hemostatic and platelet functions of fresh whole blood (FWB) are well preserved when stored at 4°C for 14 days (human) or 7 days (rats). Resuscitation of either FWB or whole blood stored for 7-days at 4°C (SWB) in rats with acute traumatic coagulopathy (ATC) induced by polytrauma and hemorrhage (T/H), shows identical correction of hemodynamic and metabolic function, as well as the equivalent impact on platelet aggregation in vitro. Platelet hemostatic function is influenced by platelet number, aggregation, and platelet-leukocyte interactions. Therefore, the objective of this study is to identify the in-vivo fate and activity of platelets from FWB and SWB, and determine whether transfused platelets contribute to clot formation in tissue injury after transfusion.Method: Isoflurane anesthetized Sprague-Dawley rats underwent polytrauma, followed by 40% hemorrhage. The rats were resuscitated (20% of bled volume) 1hr after trauma using either FWB or SWB (n=5 each) collected from green fluorescence (GFP) donor rats. SWB was kept in mini-blood-bag and refrigerated at 40C for 7 days, and warmed up to 370C prior to transfusion. Sham rats had no trauma, but underwent 20% blood replacement by FWB or SWB (n=4 each). Immediately after transfusion, a liver incision was made, followed by the measurement of kidney bleeding time (both standardized incisions was made by a Surgicutt adult template device). The incisional liver sample was taken 10min after injury for immunohistochemistry stained for CD61 (platelets) and GFP. Blood samples taken from donor rats (FWB and SWB) and recipient rats at baseline, immediately before resuscitation, and 0min and 10min after transfusion were analyzed by flow cytometry using CD42d, CD45, and CD62P antibodies for defining platelets, leukocytes and activated platelets respectively. Lactadherin was used to detect phosphatidylserine (PS). Platelet-leukocyte aggregates were defined as CD42d+/CD45+. The transfused and native platelet or leukocyte was differentiated by GFP signal.Results: GFP signal intensity was equally expressed in platelets derived from either FWB or SWB. Storage at 4°C led to a significant increase in number of platelets expressing CD62P (76±2% vs. 11±4%) and PS (8.6±0.7% vs. 1.1±0.4%) as compared to FWB. SWB also showed a significant rise in intensity, but not percentage, of platelet-leukocyte aggregates (45781±6235 vs. 10978±2345 MFI). After transfusion, recipient rats showed a significant elevation in the percentage of GFP+ platelets after transfusion of FWB as compared to SWB (Sham:13.0±1.0% vs. 4.9±0.4%; T/H: 17.0±0.8% vs. 6.8±0.%). Similarly, GFP+ leukocytes aggregates from FWB were three times higher than SWB (Sham: 6.15±0.5% vs. 1.9±0.5%; T/H: 4.4±0.4% vs. 1.2±0.2%). There was greater percentage of GFP+ platelet-leukocyte aggregates in both T/H and sham rats transfused with SWB than FWB (Sham: 11.7±3.3% vs. 33.8±2.3%; T/H: 9.8±1.6% vs. 47.2±5.1%). Transfusion with SWB led to a significant increase in percentage of activated GFP+ platelets in T/H rats as compared to FWB (30.0±2.7% vs. 2.5±0.5%). However, the activity of native platelets was not significantly different between SWB and FWB in T/H rats after transfusion (3.9±1.2% vs. 3.7±0.3%). The kidney bleeding time was not significantly different in T/H rats receiving FWB and SWB (131±4 vs.127±7 seconds) under equivalent mean arterial blood pressure (82±7 vs 85±9 mmHg). The clot that formed at the site of liver incision was identified by platelet aggregates stained by the CD61 antibody. Using co-localization of CD61 and GFP, we found that the platelets from both FWB and SWB equivalently incorporated into the clot at incisional site.Conclusion: The platelets from SWB are higher in platelet activation state, clearance rate and platelet-leukocyte aggregates than FWB after transfusion in both sham rats and rats with polytrauma and hemorrhage. However, platelets from both fresh and cold-stored whole blood contribute to hemostasis of tissue injury after transfusion. This study suggests that cold-stored whole blood is an alternative resource to treat trauma patients for restoration of hemostatic function. Future study is necessary to optimize the storage of whole blood to prolong the platelet survival rate and optimize hemostatic function. This project was funded by US Army Medical Research and Material Command. DisclosuresNo relevant conflicts of interest to declare.

Amplified Fluorescence from Polyfluorene Nanoparticles with Dual State Emission and Aggregation Caused Red Shifted Emission for Live Cell Imaging and Cancer Theranostics.

A newly synthesized polyfluorene derivative with pendant di(2-picolyl)amine (PF-DPA) shows dual state emission and aggregation caused red shifted emission that was utilized for cell imaging and cancer theranostics. PF-DPA was nontoxic to normal cells but showed cytotoxicity against cancer cells, suggesting its utility for cancer therapy. PF-DPA exhibits a large and unique red shifted emission at 556 nm at higher water ratio of THF:H2O (10:90) due to the formation of polymer nanoparticles or PDots spontaneously by intra- and intermolecular self-assembly induced aggregation. Dual state emission and aggregation caused red shifted emission (>100 nm) in PF-DPA homopolymer nanoparticles is very unique and attributed to the combined effect of intramolecular planarization and J-type aggregate formation in the PDots (25 ± 5 nm). The PF-DPA PDots exhibit bright green and orange fluorescence with exceptional live cell imaging properties and potential applications in cancer theranostics due to their selective cytotoxic nature toward cancer cells.

State aggregation for fast likelihood computations in molecular evolution.

MotivationCodon models are widely used to identify the signature of selection at the molecular level and to test for changes in selective pressure during the evolution of genes encoding proteins. The large size of the state space of the Markov processes used to model codon evolution makes it difficult to use these models with large biological datasets. We propose here to use state aggregation to reduce the state space of codon models and, thus, improve the computational performance of likelihood estimation on these models.ResultsWe show that this heuristic speeds up the computations of the M0 and branch-site models up to 6.8 times. We also show through simulations that state aggregation does not introduce a detectable bias. We analyzed a real dataset and show that aggregation provides highly correlated predictions compared to the full likelihood computations. Finally, state aggregation is a very general approach and can be applied to any continuous-time Markov process-based model with large state space, such as amino acid and coevolution models. We therefore discuss different ways to apply state aggregation to Markov models used in phylogenetics.Availability and ImplementationThe heuristic is implemented in the godon package (https://bitbucket.org/Davydov/godon) and in a version of FastCodeML (https://gitlab.isb-sib.ch/phylo/fastcodeml).Supplementary information Supplementary data are available at Bioinformatics online.

Deposition of Uranium Oxide Following the Reduction in Weak Acid Solution Using Electrochemical Quartz Crystal Microbalance (EQCM)

To evaluate the safety of the geological disposal of high-level radioactive waste, it is important to understand the behavior of actinide (An) in nature environment. An ions having large electric charge easily form complexes and colloids. The migration behaviors of An species differ from one another, hence formation reaction and chemical state of aggregates such as colloids should be clarified. While it was reported that the reduction of U(VI) to U(IV) leads to the formation of UO2 nanoparticles, the scheme of nanoparticulation following the reduction has not been understood. In previous study (Y. Kitastuji et al., Electrochim. Acta, 2014, 141, 6-12.), the authors investigated the reduction of U(VI) in a weak acid solution. It was found that aggregates of U(IV) formed in reduction process of U(V) enhanced the rate of the disproportionation and the electrolytic reduction of U(V). In this study, we investigated the aggregation of U(IV) following the valence change of U as electrodeposition using electrochemical quartz crystal microbalance (EQCM). EQCM (ALS420C, BAS) were measured by using the 8 MHz AT-cut quartz crystal (diameter 13.7 mm) with Au electrodes (diameter 5.1 mm) on both sides in 1 M NaClO4 solutions containing 1 × 10-3 M UO2 2+. The pHs of the solution were adjusted to be 2.0 and 4.0 by adding NaOH. A platinum wire and a silver-silver chloride electrode (SSE) with 1 M LiCl were employed as the counter electrode and the reference electrode, respectively. The applied potential was -0.35 V at which the current for the reduction of U(VI) to U(V) was observed. When EQCM measurements of the reduction of U(VI) in a weak acid solution were performed, the frequency was changed negatively. This indicates that deposits of U(IV) were formed on the electrode surface. We consider that the deposition is divided into the three steps according to deposition rates; I) the induction step which is until starting deposition, II) metastable aggregation step of which reaction rate is temporary fast, and III) stable aggregation step of which rate is constant. The oxidative potential of deposits at each aggregation step was investigated by stripping voltammetry measurement. The oxidative potential of deposit at stable aggregation step was higher than that at metastable aggregation step. This indicates that deposits are stabilized with growth. The pH dependence of the deposition rate was investigated. The deposits of U(IV) formed at more than pH 3.1. When pH was higher, the induction period was shorter, and deposition rates at metastable and stable aggregation steps were high. This suggests that a U(IV) hydroxide relates to the formation of deposits. In addition, the mass change of deposits was measured to investigate the chemical reaction after stopping the electrolysis. Mass increase was observed in the a little time from the end of electrolysis. This increase proposed that the slow chemical reaction, i.e. disproportionation, makes progress and U(IV) aggregates grow up. It presumes that this reaction leads to the deposition of U(IV) aggregates. The oxidative potential of deposits after stopped electrolysis was shifted to positive potential and also increased in the residual amount of deposited species on the electrode surface. It is considered that formed deposits change more stable one with time. From these results, we proposed the following aggregation mechanism (see Figure). In the induction step, U(IV) is formed by the disproportionation following the reduction of U(VI) to U(V). In metastable aggregation step, U(IV) produced a U(IV) hydroxide. The U(IV) hydroxide aggregates as a core and form into the metastable aggregate of various stable states. Also, the increase of reduction current at metastable aggregation step was observed due to reduction of U(V) by deposits on the electrode surface. We consider that decrease of deposition rate at stable aggregation step depends on dehydration of metastable aggregates. In stable aggregation step, those aggregates are dehydrated, and more stable U oxide is finally formed. This work was supported by JSPS KAKENHI Grant Number 15H04247. Figure 1