Pluggable multitask diffractive neural networks based on cascaded metasurfaces

Optical neural networks have significant advantages in terms of power consumption,parallelism,and high computing speed,which has intrigued extensive attention in both academic and engineering communities.It has been considered as one of the powerful tools in promoting the fields of imaging processing and object recognition.However,the existing optical system architecture cannot be reconstructed to the realization of multi-functional artificial intelligence systems simultaneously.To push the development of this issue,we propose the pluggable diffractive neural networks(P-DNN),a general paradigm resorting to the cascaded metasurfaces,which can be applied to recognize various tasks by switching internal plug-ins.As the proof-of-principle,the recognition functions of six types of handwritten digits and six types of fashions are numerical simulated and experimental demonstrated at near-infrared regimes.Encouragingly,the proposed paradigm not only improves the flexibility of the optical neural networks but paves the new route for achieving high-speed,low-power and versatile artificial intelligence systems.

Multi-dimensional multiplexing optical secret sharing framework with cascaded liquid crystal holograms

Secret sharing is a promising technology for information encryption by splitting the secret information into different shares.However,the traditional scheme suffers from information leakage in decryption process since the amount of available information channels is limited.Herein,we propose and demonstrate an optical secret sharing framework based on the multi-dimensional multiplexing liquid crystal(LC)holograms.The LC holograms are used as spatially separated shares to carry secret images.The polarization of the incident light and the distance between different shares are served as secret keys,which can significantly improve the information security and capacity.Besides,the decryption condition is also restricted by the applied external voltage due to the variant diffraction efficiency,which further increases the information security.In implementation,an artificial neural network(ANN)model is developed to carefully design the phase distribution of each LC hologram.With the advantage of high security,high capacity and simple configuration,our optical secret sharing framework has great potentials in optical encryption and dynamic holographic display.

Logformer: Cascaded Transformer for System Log Anomaly Detection

Modern large-scale enterprise systems produce large volumes of logs that record detailed system runtime status and key events at key points.These logs are valuable for analyzing performance issues and understanding the status of the system.Anomaly detection plays an important role in service management and system maintenance,and guarantees the reliability and security of online systems.Logs are universal semi-structured data,which causes difficulties for traditional manual detection and pattern-matching algorithms.While some deep learning algorithms utilize neural networks to detect anomalies,these approaches have an over-reliance on manually designed features,resulting in the effectiveness of anomaly detection depending on the quality of the features.At the same time,the aforementioned methods ignore the underlying contextual information present in adjacent log entries.We propose a novel model called Logformer with two cascaded transformer-based heads to capture latent contextual information from adjacent log entries,and leverage pre-trained embeddings based on logs to improve the representation of the embedding space.The proposed model achieves comparable results on HDFS and BGL datasets in terms of metric accuracy,recall and F1-score.Moreover,the consistent rise in F1-score proves that the representation of the embedding spacewith pre-trained embeddings is closer to the semantic information of the log.

Cascading multi-segment rupture process of the 2023 Turkish earthquake doublet on a complex fault system revealed by teleseismic P wave back projection method

In this study,the vertical components of broadband teleseismic P wave data recorded by China Earthquake Network are used to image the rupture processes of the February 6th,2023 Turkish earthquake doublet via back projection analysis.Data in two frequency bands(0.5-2 Hz and 1-3 Hz)are used in the imaging processes.The results show that the rupture of the first event extends about 200 km to the northeast and about 150 km to the southwest,lasting~90 s in total.The southwestern rupture is triggered by the northeastern rupture,demonstrating a sequential bidirectional unilateral rupture pattern.The rupture of the second event extends approximately 80 km in both northeast and west directions,lasting~35 s in total and demonstrates a typical bilateral rupture feature.The cascading ruptures on both sides also reflect the occurrence of selective rupture behaviors on bifurcated faults.In addition,we observe super-shear ruptures on certain fault sections with relatively straight fault structures and sparse aftershocks.

Channel Estimation for Reconfigurable Intelligent Surface Aided Multiuser Millimeter-Wave/THz Systems

It is assumed that reconfigurable intelligent surface(RIS)is a key technology to enable the potential of mmWave communications.The passivity of the RIS makes channel estimation difficult because the channel can only be measured at the transceiver and not at the RIS.In this paper,we propose a novel separate channel estimator via exploiting the cascaded sparsity in the continuously valued angular domain of the cascaded channel for the RIS-enabled millimeter-wave/Tera-Hz systems,i.e.,the two-stage estimation method where the cascaded channel is separated into the base station(BS)-RIS and the RIS-user(UE)ones.Specifically,we first reveal the cascaded sparsity,i.e.,the sparsity exists in the hybrid angular domains of BS-RIS and the RIS-UEs separated channels,to construct the specific sparsity structure for RIS enabled multi-user systems.Then,we formulate the channel estimation problem using atomic norm minimization(ANM)to enhance the proposed sparsity structure in the continuous angular domains,where a low-complexity channel estimator via Alternating Direction Method of Multipliers(ADMM)is proposed.Simulation findings demonstrate that the proposed channel estimator outperforms the current state-of-the-arts in terms of performance.

Conjugated microporous polymers-scaffolded enzyme cascade systems with enhanced catalytic activity

Enhancing catalytic activity of multi-enzyme in vitro through substrate channeling effect is promis-ing yet challenging.Herein,conjugated microporous polymers(CMPs)-scaffolded integrated en-zyme cascade systems(I-ECSs)are constructed through co-entrapping glucose oxidase(GOx)and horseradish peroxidase(HRP),in which hydrogen peroxide(H_(2)O_(2)) is the intermediate product.The interplay of low-resistance mass transfer pathway and appropriate pore wall-H_(2)O_(2) interactions facilitates the directed transfer of H_(2)O_(2),resulting in 2.4-fold and 5.0-fold elevation in catalytic activ-ity compared to free ECSs and separated ECSs,respectively.The substrate channeling effect could be regulated by altering the mass ratio of GOx to HRP.Besides,I-ECSs demonstrate excellent stabili-ties in harsh environments and multiple recycling.

Exploring the impacts of major events on the systemic risk of the international energy market

This study examines the systemic risk caused by major events in the international energy market(IEM)and proposes a management strategy to mitigate it. Using the tail-event driven network(TENET)method, this study constructed a tail-risk spillover network(TRSN) of IEM and simulated the dynamic spillover tail-risk process through the cascading failure mechanism. The study found that renewable energy markets contributed more to systemic risk during the Paris Agreement and the COVID-19pandemic, while fossil energy markets played a larger role during the Russia-Ukraine conflict. This study identifies systemically important markets(SM) and critical tail-risk spillover paths as potential sources of systemic risk. The research confirms that cutting off the IEM risk spillover path can greatly reduce systemic risk and the influence of SM. This study offers insights into the management of systemic risk in IEM and provides policy recommendations to reduce the impact of shock events.

Coordinated Capacitor Voltage Balancing Method for Cascaded H-bridge Inverter with Supercapacitor and DC-DC Stage

Cascaded H-bridge inverter(CHBI) with supercapacitors(SCs) and dc-dc stage shows significant promise for medium to high voltage energy storage applications. This paper investigates the voltage balance of capacitors within the CHBI, including both the dc-link capacitors and SCs. Balance control over the dc-link capacitor voltages is realized by the dcdc stage in each submodule(SM), while a hybrid modulation strategy(HMS) is implemented in the H-bridge to balance the SC voltages among the SMs. Meanwhile, the dc-link voltage fluctuations are analyzed under the HMS. A virtual voltage variable is introduced to coordinate the balancing of dc-link capacitor voltages and SC voltages. Compared to the balancing method that solely considers the SC voltages, the presented method reduces the dc-link voltage fluctuations without affecting the voltage balance of SCs. Finally, both simulation and experimental results verify the effectiveness of the presented method.

Risk-Balanced Routing Strategy for Service Function Chains of Cyber-Physical Power System Considering Cross-Space Cascading Failure

Cyber-physical power system(CPPS)has significantly improved the operational efficiency of power systems.However,cross-space cascading failures may occur due to the coupling characteristics,which poses a great threat to the safety and reliability of CPPS,and there is an acute need to reduce the probability of these failures.Towards this end,this paper first proposes a cascading failure index to identify and quantify the importance of different information in the same class of communication services.On this basis,a joint improved risk-balanced service function chain routing strategy(SFC-RS)is proposed,which is modeled as a robust optimization problem and solved by column-and-constraint generation(C-CG)algorithm.Compared with the traditional shortest-path routing algorithm,the superiority of SFC-RS is verified in the IEEE 30-bus system.The results demonstrate that SFC-RS effectively mitigates the risk associated with information transmission in the network,enhances information transmission accessibility,and effectively limits communication disruption from becoming the cause of cross-space cascading failures.

Simulation and Analysis of Cascading Faults in Integrated Heat and Electricity Systems Considering Degradation Characteristics

Cascading faults have been identified as the primary cause of multiple power outages in recent years.With the emergence of integrated energy systems(IES),the conventional approach to analyzing power grid cascading faults is no longer appropriate.A cascading fault analysis method considering multi-energy coupling characteristics is of vital importance.In this study,an innovative analysis method for cascading faults in integrated heat and electricity systems(IHES)is proposed.It considers the degradation characteristics of transmission and energy supply com-ponents in the system to address the impact of component aging on cascading faults.Firstly,degradation models for the current carrying capacity of transmission lines,the water carrying capacity and insulation performance of thermal pipelines,as well as the performance of energy supply equipment during aging,are developed.Secondly,a simulation process for cascading faults in the IHES is proposed.It utilizes an overload-dominated development model to predict the propagation path of cascading faults while also considering network islanding,electric-heating rescheduling,and load shedding.The propagation of cascading faults is reflected in the form of fault chains.Finally,the results of cascading faults under different aging levels are analyzed through numerical examples,thereby verifying the effectiveness and rationality of the proposed model and method.

Model Predictive Control for Cascaded H-Bridge PV Inverter with Capacitor Voltage Balance

We designed an improved direct-current capacitor voltage balancing control model predictive control(MPC)for single-phase cascaded H-bridge multilevel photovoltaic(PV)inverters.Compared with conventional voltage balanc-ing control methods,the method proposed could make the PV strings of each submodule operate at their maximum power point by independent capacitor voltage control.Besides,the predicted and reference value of the grid-connected current was obtained according to the maximum power output of the maximum power point tracking.A cost function was con-structed to achieve the high-precision grid-connected control of the CHB inverter.Finally,the effectiveness of the proposed control method was verified through a semi-physical simulation platform with three submodules.

Adaptive Output-Feedback Stabilization for PDE-ODE Cascaded Systems with Unknown Control Coefficient and Spatially Varying Parameter

This paper investigates the adaptive stabilization for a class of uncertain PDE-ODE cascaded systems. Remarkably, the PDE subsystem allows unknown control coefficient and spatially varying parameter, and only its one boundary value is measurable. This renders the system in question more general and practical, and the control problem more challenging. To solve the problem,an invertible transformation is first introduced to change the system into an observer canonical form,from which a couple of filters are constructed to estimate the unmeasurable states. Then, by adaptive technique and infinite-dimensional backstepping method, an adaptive controller is constructed which guarantees that all states of the resulting closed-loop system are bounded while the original system states converging to zero. Finally, a numerical simulation is provided to illustrate the effectiveness of the proposed method.

A distributed VSG control method for a battery energy storage system with a cascaded H-bridge in a grid-connected mode

With the high penetration of renewable energy,new challenges,such as power fluctuation suppression and inertial support capability,have arisen in the power sector.Battery energy storage systems play an essential role in renewable energy integration.In this paper,a distributed virtual synchronous generator(VSG)control method for a battery energy storage system(BESS)with a cascaded H-bridge converter in a grid-connected mode is proposed.The VSG is developed without communication dependence,and state-of-charge(SOC)balancing control is achieved using the distributed average algorithm.Owing to the low varying speed of SOC,the bandwidth of the distributed communication networks is extremely slow,which decreases the cost.Therefore,the proposed method can simultaneously provide inertial support and accurate SOC balancing.The stability is also proved using root locus analysis.Finally,simulations under different conditions are carried out to verify the effectiveness of the proposed method.

Deep learning for predictive mechanical properties of hot-rolled strip in complex manufacturing systems

Higher requirements for the accuracy of relevant models are put throughout the transformation and upgrade of the iron and steel sector to intelligent production.It has been difficult to meet the needs of the field with the usual prediction model of mechanical properties of hotrolled strip.Insufficient data and difficult parameter adjustment limit deep learning models based on multi-layer networks in practical applications;besides,the limited discrete process parameters used make it impossible to effectively depict the actual strip processing process.In order to solve these problems,this research proposed a new sampling approach for mechanical characteristics input data of hot-rolled strip based on the multi-grained cascade forest(gcForest)framework.According to the characteristics of complex process flow and abnormal sensitivity of process path and parameters to product quality in the hot-rolled strip production,a three-dimensional continuous time series process data sampling method based on time-temperature-deformation was designed.The basic information of strip steel(chemical composition and typical process parameters)is fused with the local process information collected by multi-grained scanning,so that the next link’s input has both local and global features.Furthermore,in the multi-grained scanning structure,a sub sampling scheme with a variable window was designed,so that input data with different dimensions can get output characteristics of the same dimension after passing through the multi-grained scanning structure,allowing the cascade forest structure to be trained normally.Finally,actual production data of three steel grades was used to conduct the experimental evaluation.The results revealed that the gcForest-based mechanical property prediction model outperforms the competition in terms of comprehensive performance,ease of parameter adjustment,and ability to sustain high prediction accuracy with fewer samples.

State feedback stabilization of cascaded nonlinear systems with discontinuous connection

In practical engineering, many phenomena are described as a discontinuous function of a state variable, and the discontinuity is usually the main reason for the degradation of the control performance. For example, in the set-point control problem of mechanical systems, the static friction （described by a sgn function of velocity of the contacting faces） causes undesired positioning error. In this paper, we will investigate the stabilization problem for a class of nonlinear systems that consist of two subsystems with cascaded connection. We will show the basic idea with a special case first, and then the result will be extended to more general cases. Some interesting numerical examples will be given to demonstrate the effectiveness of the proposed design approach.

A Two-Stage Approach for Large-Scale Cascaded Hydropower System Operations

The paper presents a two-stage approach to cope with the long-term optimal operation of cascaded hydropower systems. This approach combines progressive optimality algorithm (POA) with quadratic programming (QP) to improve the optimization results. POA is used at the first stage to generate a local optimal result, which will be selected as the initial feasible solution of QP method employed at the second stage. Around the initial solution, a rational local search range for QP method is then determined, where the nonlinear water level function and tailrace level function can be linearized nearly with high accuracy. The simplified optimization problem is formulated as a QP model with a quadratic generation function and a linear set of constraints, and solved using the available mathematic optimization software package. Simulation is performed on the long term operation of Hongshui River hydropower system which is located in southwest China and consists of 9 built hydropower plants. Results obtained from the proposed approach show a significant increase in the total energy production compared to the results from POA.

Optimal Scheduling of Cascaded Hydrothermal Systems Using a New Improved Particle Swarm Optimization Technique

Optimum scheduling of hydrothermal plants generation is of great importance to electric utilities. Many evolutionary techniques such as particle swarm optimization, differential evolution have been applied to solve these problems and found to perform in a better way in comparison with conventional optimization methods. But often these methods converge to a sub-optimal solution prematurely. This paper presents a new improved particle swarm optimization technique called self-organizing hierarchical particle swarm optimization technique with time-varying acceleration coefficients (SOHPSO_TVAC) for solving short-term economic generation scheduling of hydrothermal systems to avoid premature convergence. A multi-reservoir cascaded hydrothermal system with nonlinear relationship between water discharge rate, power generation and net head is considered here. The performance of the proposed method is demonstrated on two test systems comprising of hydro and thermal units. The results obtained by the proposed methods are compared with other methods. The results show that the proposed technique is capable of producing better results.

Inner Cascaded U^(2)-Net:An Improvement to Plain Cascaded U-Net

Deep neural networks are now widely used in the medical image segmentation field for their performance superiority and no need of manual feature extraction.U-Net has been the baseline model since the very beginning due to a symmetricalU-structure for better feature extraction and fusing and suitable for small datasets.To enhance the segmentation performance of U-Net,cascaded U-Net proposes to put two U-Nets successively to segment targets from coarse to fine.However,the plain cascaded U-Net faces the problem of too less between connections so the contextual information learned by the former U-Net cannot be fully used by the latter one.In this article,we devise novel Inner Cascaded U-Net and Inner Cascaded U^(2)-Net as improvements to plain cascaded U-Net for medical image segmentation.The proposed Inner Cascaded U-Net adds inner nested connections between two U-Nets to share more contextual information.To further boost segmentation performance,we propose Inner Cascaded U^(2)-Net,which applies residual U-block to capture more global contextual information from different scales.The proposed models can be trained from scratch in an end-to-end fashion and have been evaluated on Multimodal Brain Tumor Segmentation Challenge(BraTS)2013 and ISBI Liver Tumor Segmentation Challenge(LiTS)dataset in comparison to related U-Net,cascaded U-Net,U-Net++,U^(2)-Net and state-of-the-art methods.Our experiments demonstrate that our proposed Inner Cascaded U-Net and Inner Cascaded U^(2)-Net achieve better segmentation performance in terms of dice similarity coefficient and hausdorff distance as well as get finer outline segmentation.

基于改进的Cascade RCNN铸管字符检测算法

由于工业现场采集的铸管字符图像存在背景模糊、字符区域占比小、刻字位置不固定、油漆遮挡等问题,导致现有模型的检测精度难以满足工业现场的需求。针对上述问题,提出改进的Cascade RCNN铸管字符检测算法。首先对特征金字塔进行改进,提出融合小目标增强的特征金字塔(STE-FPN),利用多尺度特征融合的特征增强能力丰富铸管小目标字符的特征信息。其次引入自矫正/池化的ResNeSt(SCP-ResNeSt)作为特征提取网络,利用自矫正卷积和池化操作以提升背景复杂的铸管字符特征提取效率。最后对级联结构进行改进,引进Mask分支结构,可以自适应地检测字符区域并去除干扰区域,优化了检测结果。将改进后的算法在铸管数据集上进行测试,其平均检测精度mAP为99.1%,比原Cascade RCNN算法提高了2.3%,得到的精度表明改进后的性能优于原算法。

基于改进Cascade R-CNN的安全帽检测算法

针对安全帽检测中,目标形状、尺度变化大,易出现漏检、误检等问题,提出了一种基于改进级联基于区域的卷积神经网络(Cascade R-CNN)的安全帽检测算法。首先,对ResNet50进行改进形成D-ResNet50,利用可变形卷积仅增加少量参数就可增大感受野的特性,对特征提取网络的C2~C5卷积层进行重塑,提高网络对目标几何变换的适应能力和特征提取能力。其次,将D-ResNet50作为主干网络引入Cascade R-CNN,形成级联目标检测器,在每个阶段对正负样本重采样,抑制误检问题。再次,对递归特征金字塔进行改进,更高效地进行多尺度特征融合,并且基于反馈信息对特征进行二次处理,增强特征表达,提高网络的分类和定位能力。最后,使用Soft-非极大值抑制(Soft-NMS)进行后处理,进一步解决漏检问题。提出的方法在Hard hat workers数据集上的AP值相比检测基线提高了3.5%,与Sparse R-CNN、TridentNet、VFnet等先进算法相比分别提升了4.7%、5.9%、2.3%等。