Adaptive Configuration Research Articles

Efficient very large-scale integration architecture design of proportionate-type least mean square adaptive filters

The effectiveness of adaptive filters are mainly dependent on the design techniques and the algorithm of adaptation. The most common adaptation technique used is least mean square (LMS) due its computational simplicity. The application depends on the adaptive filter configuration used and are well known for system identification and real time applications. In this work, a modified delayed μ-law proportionate normalized least mean square (DMPNLMS) algorithm has been proposed. It is the improvised version of the µ-law proportionate normalized least mean square (MPNLMS) algorithm. The algorithm is realized using Ladner-Fischer type of parallel prefix logarithmic adder to reduce the silicon area. The simulation and implementation of very large-scale integration (VLSI) architecture are done using MATLAB, Vivado suite and complementary metal–oxide– semiconductor (CMOS) 90 nm technology node using Cadence register transfer level (RTL) Genus Compiler respectively. The DMPNLMS method exhibits a reduction in mean square error, a higher rate of convergence, and more stability. The synthesis results demonstrate that it is area and delay effective, making it practical for applications where a faster operating speed is required.

Modularity Implications of an Overground Exoskeleton on Plantar Pressures, Strength, and Spasticity in Persons with Acquired Brain Injury.

This study explored the effects of a modular overground exoskeleton on plantar pressure distribution in healthy individuals and individuals with Acquired Brain Injury (ABI). The research involved 21 participants, including ABI patients and healthy controls, who used a unique exoskeleton with adaptable modular configurations. The primary objective was to assess how these configurations, along with factors such as muscle strength and spasticity, influenced plantar pressure distribution. The results revealed significant differences in plantar pressures among participants, strongly influenced by the exoskeleton's modularity. Notably, significant distinctions were found between ABI patients and healthy individuals. Configurations with two modules led to increased pressure in the heel and central metatarsus regions, whereas configurations with four modules exhibited higher pressures in the metatarsus and hallux regions. Future research should focus on refining and customizing rehabilitation technologies to meet the diverse needs of ABI patients, enhancing their potential for functional recovery.

DSTEA: Improving Dialogue State Tracking via Entity Adaptive pre-training

Dialogue State Tracking (DST) is critical for comprehensively interpreting user and system utterances, thereby forming the cornerstone of efficient dialogue systems. Despite past research efforts focused on enhancing DST performance through alterations to the model structure or integrating additional features like graph relations, they often require additional pre-training with external dialogue corpora. In this study, we propose DSTEA, improving Dialogue State Tracking via Entity Adaptive pre-training, which can enhance the encoder through by intensively training key entities in dialogue utterances. DSTEA identifies these pivotal entities from input dialogues utilizing four different methods: ontology information, named-entity recognition, the spaCy toolkit, and the flair library. Subsequently, it employs selective knowledge masking to train the model effectively. Remarkably, DSTEA only requires pre-training without the direct infusion of extra knowledge into the DST model. This approach results in substantial performance improvements of four robust DST models on MultiWOZ 2.0, 2.1, and 2.2, with joint goal accuracy witnessing an increase of up to 2.69% (from 52.41% to 55.10%). Comparative experiments considering various entity types and different entity adaptive pre-training configurations, such as masking strategy and masking rate, further validated the efficacy of DSTEA.

MVIndEmo: a dataset for micro video public-induced emotion prediction on social media

Distinct from the realm of perceived emotion research, induced emotion pertains to the emotional responses engendered within content consumers. This facet has garnered considerable attention and finds extensive application in the analysis of public social media. However, the advent of micro videos presents unique challenges when attempting to discern the induced emotional patterns exhibited by content consumers, owing to their free-style representation and other factors. Consequently, we have put forth two novel tasks concerning the recognition of public-induced emotion on micro videos: emotion polarity and emotion classification. Additionally, we have introduced a accessible dataset specifically tailored for the analysis of public-induced emotion on micro videos. The data corpus has been meticulously collected from Tiktok, a burgeoning social media platform renowned for its trendsetting content. To construct the dataset, we have selected eight captivating topics that elicit vibrant social discussions. In devising our label generation strategy, we have employed an automated approach characterized by the fusion of multiple expert models. This strategy incorporates a confidence measure method that relies on three distinct models for effectively aggregating user comments. To accommodate adaptable benchmark configurations, we provide both binary classification labels and probability distribution labels. The dataset encompasses a vast collection of 7,153 labeled micro videos. We have undertaken an extensive statistical analysis of the dataset to provide a comprehensive overview composition. It is our earnest aspiration that this dataset will serve as a catalyst for pioneering research avenues in the analysis of emotional patterns and the understanding of multi-modal information.

Adaptive Network Configuration for Efficient and Accurate Neural Video Inference

Adaptive Network Configuration for Efficient and Accurate Neural Video Inference

Collaborative Path Planning of Multiple AUVs Based on Adaptive Multi-Population PSO

Collaborative operations of multiple AUVs have been becoming increasingly popular and efficient in underwater tasks of marine applications. Autonomous navigation capability and cooperative control stability of multiple AUVs are crucial and challenging issues in underwater environments. To address the collaborative problem of path planning for multiple AUVs, this paper proposes an adaptive multi-population particle swarm optimization (AMP-PSO). In AMP-PSO, we design a grouping strategy of multi-population and an exchanging mechanism of particles between groups. We separate particles into one leader population and various follower populations according to their fitness. Firstly, in the grouping strategy, particles within the leader population are updated by both the leader population and follower populations so as to keep global optimization, while particles within the follower population are updated by their own group so as to keep local priority. Secondly, in the exchanging mechanism, particles are exchanged between the leader population and follower populations so as to improve multi-population diversity. To accommodate multi-population characteristics, an adaptive parameter configuration is also included to enhance the global search capability, convergence speed, and complex environment adaptability of AMP-PSO. In numerical experiments, we simulate various scenarios of collaborative path planning of multiple AUVs in an underwater environment. The simulation results convincingly demonstrate that AMP-PSO can obtain feasible and optimal path solutions compared to classic PSO and other improved PSO, which enable multiple AUVs to effectively achieve objectives under the conditions of collision avoidance and navigation constraint.

Elastic foundation-introduced defective phononic crystals for tunable energy harvesting

Defective phononic crystals offer the advantage of concentrating elastic waves, thereby enhancing the potential for piezoelectric energy harvesting (PEH). However, a key limitation is their reliance on a fixed operating frequency, rendering them susceptible to the prevailing vibration environment. To surmount this constraint, this study introduces a novel approach involving a tunable elastic foundation system for defective phononic crystal structures. The newly developed phononic crystal is fashioned by integrating a periodically elastic foundation beneath a uniform beam. A defect is induced by selectively removing specific elastic foundations and integrating piezoelectric components. Explicit analytical solutions are established through the transfer matrix method and the spectral element method, which are subsequently corroborated via comparison with finite element results. The findings underscore that the periodic elastic foundations impart bandgaps in the elastic wave band structure. The absence of specific elastic foundations results in the emergence of distinct defect modes. Additionally, frequency response analysis exposes the potential for energy enhancement, albeit with inherent variations. Noteworthy is the revelation that manipulating the stiffness of the elastic foundation triggers shifts in the resonant frequency of the output voltage. Therefore, the proposed tunable elastic foundation system exhibits promising potential to engender versatile and adaptive phononic crystal configurations, thereby advancing the domain of PEH.

Fishing risky behavior recognition based on adaptive transformer, reinforcement learning and stochastic configuration networks

The fishing behavior in a high-voltage environment may lead to electric shock accidents. This paper proposes a fishing risky behavior recognition method based on adaptive Transformer, reinforcement learning and stochastic configuration networks (SCNs). Firstly, the fishing behavior image sample training set is evaluated using the actor network based on linear entropy in the reinforcement learning module to select suitable training samples that possess abundant straight-line elements. Subsequently, the selected training samples are fed into a multi-scale spatial deep convolution to extract continuous spatial features of elongated objects. The deformable Transformer network, enhanced with adaptive encoding and decoding layers through SCNs, is used to obtain the position and detection results of the fishing rod. The trained Transformer model is evaluated by a defined credibility evaluation metric for acquiring rewards to update the training sample set iteratively. Then, an adaptive adjustment mechanism is constructed for the encoding and decoding layers of the adaptive Transformer network to establish a library of adaptive Transformer models with different encoding and decoding layer levels to accommodate the feature requirements of training samples in various scenarios. Finally, the blending learning method for combining the detection results from the model library is integrated with human body object detection and pose estimation methods, and a predefined expert system is utilized for logic reasoning on fishing risky behavior. Experimental results demonstrate the effectiveness of the proposed method in this paper.

Energy-Efficient Federated Training on Mobile Device

On-device deep learning technology has attracted increasing interest recently. CPUs are the most common commercial hardware on devices and many training libraries have been developed and optimized for them. However, CPUs still suffer from poor training performance (i.e., training time) due to the specific asymmetric multiprocessor. Moreover, the energy constraint imposes restrictions on battery-powered devices. With federated training, we expect the local training to be completed rapidly therefore the global model converges fast. At the same time, energy consumption should be minimized to avoid compromising the user experience. To this end, we consider energy and training time and propose a novel framework with a machine learning-based adaptive configuration allocation strategy, which chooses optimal configuration combinations for efficient ondevice training. We carry out experiments on the popular library MNN and the experimental results show that the adaptive allocation algorithm reduces substantial energy consumption, compared to all batches with fixed configurations on off-the-shelf CPUs.

Assessing the performance of polyphosphate accumulating organisms in a full-scale side-stream enhanced biological phosphorous removal.

Phosphorous (P) removal in wastewater treatment is essential to prevent eutrophication in water bodies. Side-stream enhanced biological phosphorous removal (S2EBPR) is utilized to improve biological P removal by recirculating internal streams within a side-stream reactor to generate biodegradable carbon (C) for polyphosphate accumulating organisms (PAOs). In this study, a full-scale S2EBPR system in a water resource recovery facility (WRRF) was evaluated for 5 months. Batch experiments revealed a strong positive correlation (r = 0.91) between temperature and C consumption rate (3.56-8.18 mg-COD/g-VSS/h) in the system, with temperature ranging from 14°C to 18°C. The anaerobic P-release to COD-uptake ratio decreased from 0.93 to 0.25 mg-P/mg-COD as the temperature increased, suggesting competition between PAOs and other C-consumers, such as heterotrophic microorganisms, to uptake bioavailable C. Microbial community analysis did not show a strong relationship between abundance and activity of PAO in the tested WRRF. An assessment of the economic feasibility was performed to compare the costs and benefits of a full scale WRRF with and without implementation of the S2EBPR technology. The results showed the higher capital costs required for S2EBPR were estimated to be compensated after 5 and 11 years of operation, respectively, compared to chemical precipitation and conventional EBPR. The results from this study can assist in the decision-making process for upgrading a conventional EBPR or chemical P removal process to S2EBPR. PRACTITIONER POINTS: Implementation of S2EBPR presents adaptable configurations, exhibiting advantages over conventional setups in addressing prevalent challenges associated with phosphorous removal. A full-scale S2EBPR WRRF was monitored over 5 months, and activity tests were used to measure the kinetic parameters. The seasonal changes impact the kinetic parameters of PAOs in the S2EBPR process, with elevated temperatures raising the carbon demand. PAOs abundance showed no strong correlation with their activity in the full-scale S2EBPR process in the tested WRRF. Feasibility assessment shows that the benefits from S2EBPR operation can offset upgrading costs from conventional BPR or chemical precipitation.

Integrated Design of Hydraulic Coupling Bolts for Large Shaft Systems

Hydraulic coupling bolts are used in a wide range of industrial sectors, where ensuring the structural integrity and reliability of shaft systems is of paramount importance. However, research on systematic design methods for these specialized bolts is still lacking. In this study, an integrated design was used on hydraulic coupling bolts with M72 specification, considering both the tightening and untightening processes. First, the second tightening load was determined based on material strength, stress concentration and safety factors. Subsequently, the upper and lower limits of the first tightening loads were calculated using the thick-walled cylinder theory. An appropriate first tightening load maintaining the tightening force in both the axial and radial directions was obtained using finite element analysis. Design variables for the adapter configuration and oil passage shape were selected, and their impact on structural integrity was assessed using finite element analysis. The optimal combination of these design variables was derived. Finally, operational experiments for hydraulic coupling bolts were conducted according to the design results. The test results demonstrated a successful tightening and untightening process, confirming the effectiveness of the proposed design method.

Modular Approach to Selected Configuration Interaction in an Arbitrary Spin Basis: Implementation and Comparison of Approaches.

A modular selected configuration interaction (SCI) code has been developed that is based on the existing Monte-Carlo configuration interaction code (MCCI). The modularity allows various selection protocols to be implemented with ease and allows for fair comparison between wave functions built via different criteria. We have initially implemented adaptations of existing SCI theories, which are based on either energy- or coefficient-driven selection schemes. These codes have been implemented not only in the basis of Slater determinants (SDs) but also in the basis of configuration state functions (CSFs) and extended to state-averaged regimes. This allows one to take advantage of the reduced dimensionality of the wave function in the CSF basis and also the guarantee of pure spin states. All SCI methods were found to be able to predict potential energy surfaces to high accuracy, producing compact wave functions, when compared to full configuration interaction (FCI) for a variety of bond-breaking potential energy surfaces. The compactness of the error-controlled adaptive configuration interaction approach, particularly in the CSF basis, was apparent with nonparallelity errors within chemical accuracy while containing as little as 0.02% of the FCI CSF space. The size-to-accuracy was also extended to FCI spaces approaching one billion configurations.

L1 adaptive control design for the rigid body launch vehicle

This paper investigates the use of an L1-adaptive controller to improve the performance of the Vega launch vehicle because the classical controller does not guarantee stability and tracking of the system in the transient. The L1-AC ensures uniformly bounded transient and steady-state tracking for both systems’ signals, input, and output. In this paper, we used the equations of the adaptation and the L1-norm with two filters, the first one is first-order order and the second filter is third-order, we used the large adaptive gain with the first filter, also used the low adaptive gain with the second filter, and after the analysis the result numerically we found the lambda with the first filter less than 1 and the lambda with second filter larger than lambda with the first filter. The L1 adaptive controller can generate a stable system response to track the control input and the system output, both in transient and steady-state because we selected the adaptive gain large with minimize lambda. It is noted that the system response for the L1 adaptive control configuration with the first filter, as compared with the system response with the second filter, has much better performances, both from the point of view of the overshoot and rise time.

Adaptive and Contract Aware Configuration for Industrial Wireless Networks

Planning the transmission routing and scheduling is a challenging issue in industrial wireless networks (IWN) due to high quality of service (QoS) requirements. Moreover, in many applications we face the problem of node mobility, changes of transmission conditions, node resource and powering limitations. This paper initially formulates the problem of searching optimal network configuration covering the outlined features and subsequently proposes an effective solution. It is designed for typical industrial FDMA and TDMA networks with horizontal communication. The goal of our studies is to find IWN configuration (routing, time slot and channel allocation) that maximizes the number of successfully concluded contracts. The presented method is compatible with the ISA100.11a standard and integrates relevant QoS requirements for different periodic and aperiodic contracts. Our approach is based on original deterministic algorithms for planning the traffic. They are optimized with a genetic algorithm and facilitate the adaptation to changing operation conditions of network nodes. Extensive simulation experiments have been used in the development and evaluation of the proposed algorithms. The introduced test environment and benchmarks confirmed the advantages of our approach and revealed drawbacks of other approaches.

An explainable deep reinforcement learning algorithm for the parameter configuration and adjustment in the consortium blockchain

In recent years, consortium blockchains have attracted considerable interest from the business communities and academia. To satisfy specific application requirements, appropriately configuring or adjusting the parameters is essential but challenging for developers when building the consortium blockchain. In this paper, the parameter configuration and adjustment for the consortium blockchain is transformed as a multi-objective optimization problem, and a novel explainable deep reinforcement learning (DRL) algorithm is proposed to solve the problem. On considering that existing DRL algorithms cannot be directly used in the consortium blockchain as they suffer from lacking of explainability, a causal model for configuring and adjusting the consortium blockchain’s parameters is proposed and integrated into the DRL algorithm. The causal model can be used to derive causal explanations of the DRL algorithm to increase its trustworthiness. Furthermore, the causal model-based DRL (C-DRL) algorithm can perform causal inference before taking action to eliminate unreasonable exploration and improve the DRL algorithm’s performance. The experimental results demonstrate the proposed algorithm provides the consortium blockchain with adaptive parameter configuration and adjustment to achieve sustainable high performance and security. Furthermore, the proposed algorithm increases the convergence speed of the vanilla algorithm by 49.3% and is more trustworthy than the vanilla DRL algorithm.

Innovative Advancements in Network Topologies: A Comprehensive Investigation of Mesh Network, Tree Topology, and Hypercube Network

In the dynamic digital era, characterized by diverse and evolving communication demands, the necessity for adept network architectures has reached a paramount juncture. This paper responds to this imperative by exploring the novel architectures poised to redefine the contours of modern networking. The study employed a comprehensive methodology through theoretical analysis, simulation studies, and real-world applications. It rigorously assesses the topology structure, scalability, latency and throughout, fault tolerance, and application of the three innovative network topologies namely: Mesh Network, Tree Topology, and Hypercube Network. The results showed that Mesh Network is more suitable for dynamic IoT environment due to high resilience and fault tolerance. It emerged as a robust choice for managing the complex communication patterns inherent in IoT deployments. However, Mesh Network must strengthen its security measures by incorporating encryption protocols and intrusion detection systems. The Tree topologies showcase scalability but may be challenged in fault-tolerant scenarios. This requires tree topologies to focus on enhancing the scalability of hierarchical structures, perhaps through adaptive configurations. Hypercube networks, inspired by parallel processing, demonstrate efficiency in edge computing environments. The efficiency of hypercube networks in edge computing environments suggests the need for continued optimization. The research also elucidates potential benefits and delve into the opportunities and challenges associated with each topology. It does not only underscore their strengths and limitations but also provides actionable recommendations for their effective implementation across various scenarios. The findings of this study can contribute to the collective knowledge base and serve as a catalyst for future developments guiding researchers, practitioners, and decision-makers in their pursuit of resilient and adaptive network designs in an ever-changing digital landscape

Assessing the mechanical and static aeroelastic performance of cellular Kirigami wingbox designs

Adaptive wings configurations have been evaluated for morphing airframe applications during the last two decades. Constructions with flexible hinges can be in particular a solution for small top medium-scale air vehicles, while novel Kirigami technologies help to produce flexible and complex structures by enabling novel geometric paradigms. In this study, a cellular Kirigami wingbox with an adaptive hinge is designed and manufactured. The mechanical properties of the wingbox are numerically evaluated, considering the shear modulus of the cellular elements patterning the wingbox. Thus, the equivalent torsional, flexural stiffness, as well as the shear center location of the whole wingbox structure are calculated. The analysis is parametrized against various possible internal cell angles and cell thickness values that define the Kirigami cellular tessellation of the wingbox. The static divergence speed is also evaluated by means of the same parametrization. This study shows the feasibility of using a Kirigami wingbox concept for morphing/adaptive small to medium-scale from a structural and aeroelastic perspective.

Modular stochastic configuration network based prediction model for NOx emissions in municipal solid waste incineration process

The accurate prediction of the nitrogen oxides (NOx) emissions is extremely important for pollutant control in municipal solid waste incineration (MSWI) process. Modular neural network (MNN) provides a support for predicting NOx emissions to overcome the limitations of a single model in nonlinear processes and different climatic conditions. However, the design of sub model for MNN is a challenge. We propose a method based on MNN and adaptive ensemble stochastic configuration network (m-AESCN) in this work. First, fuzzy C-means algorithm decomposes the task into several sub datasets with similar characteristics, and an evaluation function is used to guarantee the optimal decomposition result. Second, for each sub dataset, an adaptive ensemble stochastic configuration network as the sub model of modular neural network. The optimal output for sub model can be obtained by an adaptive weighting method based on variance. Third, in testing phase, a sub model activation method based on cluster centroid is proposed to select a suitable sub model. Besides, the m-AESCN method is validated by the real data of an MSWI plant and shows considerable performance in two different datasets. And the proposed method is compared with several modeling methods, such as stochastic configuration network, support vector regression, random forest and so on. The experimental results under two datasets prove that the maximum improvements in accuracy of m-AESCN are 39.74% and 31.67%, respectively.

Approximate Excited-State Geometry Optimization with the State-Averaged Adaptive Sampling Configuration Interaction Algorithm with Large Active Spaces.

The selected configuration interaction (SCI) wave function is a useful approximation to the full configuration interaction (FCI) one. The adaptive sampling CI (ASCI) method is a deterministic SCI method. By combining ASCI and orbital optimization, the ASCI self-consistent field (ASCI-SCF) method, which is an approximation of the complete active space self-consistent field (CASSCF) method, can be formulated as well. However, their applicability has been tested mainly on the systems in their electronically ground states. In this work, we implement the state-average (SA) ansatz in ASCI-SCF calculations to calculate excited states. We also derive expressions for the approximate analytical gradient and implement them as a computer program. We demonstrate the applicability of the current method for calculating vertical and adiabatic excitation energies and optimizing the molecular geometries of thermally activated delayed fluorescence (TADF) molecules.

Nonlinear model predictive control for yaw rate and body motion control through semi-active and active suspensions

Active and semi-active suspensions for passenger cars traditionally enhance comfort through body control, and vehicle handling by reducing the tyre load variations induced by road irregularities. Active suspensions can also be designed to track a desired yaw rate profile through the control of the lateral load transfer distribution between the front and rear axles. This paper considers an integrated system including semi-active and active suspension actuation to control the yaw, roll, pitch and heave dynamics excited by the driving actions. To this purpose, two novel real-time-capable implicit nonlinear model predictive control (NMPC) formulations, excluding and including cost function weight adaptation, are proposed and compared with the passive vehicle, and the controlled vehicle with two combinations of skyhook and active roll control, the first based on a pseudoinverse decoupling transformation for obtaining the damping force contributions, and the second using an inverse formulation. The algorithms are assessed through an experimentally validated simulation model, along manoeuvres corresponding to sub-limit and limit handling operation, to analyse the trade-off between body motion reduction and cornering response enhancement. The results show that the adaptable NMPC configuration provides the best performance in all scenarios, also for significant variations of the main vehicle and tyre parameters.