Aiming at the poor effect of deep learning algorithm on small objects detection, SSD object detection method based feature fusion is proposed. The reasons for low rate and robustness classical methods are analysed; through theoretical analysis comparative experiments, characteristic layer was shallow layers with high resolution strong semantics fused structure; finally, a complete structure designed residual block to increase width depth network. contrast experiment PASCAL VOC dataset...

With the rapid development of economy and emission a lot polluted gases, air pollution is increasingly serious. Air quality prediction an effective way to provide early warning harmful pollutants, which can protect public health. A hybrid model uses K-Means clustering deep neural network proposed in this paper. The with capacity regressive computation consists bidirectional LSTM (Long Short-Term Memory) fully connected network. First all, historical meteorological monitoring data Qingdao...

Purpose This paper addresses the limitations of current graph neural network-based recommendation systems, which often neglect integration side information and modeling complex high-order interactions among nodes. The research motivation stems from need to enhance performance by effectively utilizing all available data. We propose a novel method called MSHCN, leverages hypergraph networks integrate model interactions, thereby improving user item representations. Design/methodology/approach...

Abstract The use of Kikuchi lines is very familiar as a part electron microscopic technique; it not so widely recognized that bend contours may be used in the same way. two techniques are fact similar and generally complementary. differences between methods examined advantages each discussed. has number convenience, especially for complicated specimens.

In order to solve the problem which is leader- follower formation control for robots, a dynamic surface method proposed robots whose wheel axis center and geometric don't coincide. A virtual vehicle introduced transform into trajectory tracking problem. The avoid differential explosion caused by backstepping. Using Lyapunov function theory, it proved that all signals of closed-loop system are consistent ultimately bounded. simulation results also prove effectiveness method.

Data sparsity is the main challenge that recommendation algorithms have been facing. E-commerce mainly focus on explicit interaction between users and items to alleviate this problem, while ignoring implicit feature of hidden sentiment information in review texts. To address above issues, a deep collaborative filtering model based analysis reviews (SA-NCF) proposed. First, user extracted for quantitative processing. Then matrix constructed by combining with ratings. Finally, whole nonlinear...

This paper considers the problem of robust control synthesis such that time-delay multi-agent systems are consensus and capture finite frequency domain H∞ performance. Using a linear transformation, is first transformed into partial stability problem. Then, based on dual version generalised Kalman–Yakubovich–Popov (GKYP) lemma multiplier expansion method, new sufficient conditions guaranteeing existence static output feedback controller proposed in terms matrix inequalities (LMIs) without...

With the rapid development of urbanization, environmental pollution has drawn worldwide attention. Accurate air quality prediction is very significant for alleviating severe conditions and human healthy life. A novel hybrid model that combines a spatiotemporal correlation analysis method an effectively simple network proposed to forecast concentrations pollutants. Firstly, new calculation considering distance <a:math xmlns:a="http://www.w3.org/1998/Math/MathML" id="M1"> <a:msub> <a:mrow>...

The finite frequency domain H∞ consensus control problem of neutral multi-agent systems with the input delay is investigated in this paper for first time. According to linear transformation, converted into a set stability problems. Through Lyapunov theory, combined time-delay interval decomposition method, less conservative controller design method state feedback presented. And obtained can ensure closed-loop fulfil consensus. Utilizing generalised Kalman-Yakubovich-Popov (GKYP) lemma,...

<title>Abstract</title> Since the objects are usually small and densely distributed, current mainstream algorithms have poor performance in task of object detection remote sensing, prone to produce missed false detection. Moreover, sensing images(RSI) contain complex background noise, which further exacerbates this phenomenon. To settle these problems, paper proposes a novel model named RS-YOLO. In order suppress noise enhance attention objects, by integrating Biformer, dynamic mechanism, we...

Abstract Synchronization widely exists in brain networks and plays a significant role the transmission processing of signals cognitive processes. Abnormal synchronization can lead to neurological psychiatric disorders such as schizophrenia Alzheimer’s disease. It is an essential challenging task design controller achieve between different regions. This paper aims investigate problem two nonidentical neural populations. Error feedback are proposed In addition, sufficient conditions provided...

We consider the fuzzy control problem of a chaotic system, based on genetic algorithm. apply algorithm to optimize parameters membership function controller, which integrates global searching and high precision celerity controller. The results simulation show effectiveness this method.

This paper focuses on uncertain pendulum-like systems subject to norm constraint structured uncertainties, and addresses the robust gradient-like behaviour analysis synthesis problems for such systems. By applying singular value of an operator uncertainty, results are obtained. Based results, using KYP lemma, a dynamic output feedback controller closed loop system guaranteeing is designed parameters explicitly expressed by set feasible solutions corresponding linear matrix inequalities....

This paper investigates the H <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">∞</sub> stability and consensus of discrete-time nonlinear complex dynamical network systems based on T-S fuzzy model. The system consists N subsystems. Parallel distributed compensation algorithm Lyapunov method are applied to design an decentralized controller. By considering characteristic interconnection matrix network, test a is separated in several systems....

In this paper, a novel formation control scheme is presented by integrating the dynamic surface for multi-quadrotor subject with model uncertainties and external disturbances. First, according to requirements of control, target trajectories followers are obtained designed virtual quadrotors. Then, problem transformed into trajectory tracking problem. Second, saturation function auxiliary system developed make up nonlinear terms arising from input saturation. A extended state observer (ESO)...

Bend or extinction contours maps in titanium and mica are interpreted indexed. They were constructed from a series of electron micrographs taken using beam energy 100 keV epitaxial film about 700 Å thickness, thin (muscovite) 1000 thickness prepared by cleavage air.

This paper investigates the H <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">∞</sub> consensus problem for networks of multi-agent systems subject to external disturbances. In particular, a distributed dynamic output feedback controller with time-varying delays is provided, which based on relative outputs neighboring agents. By using Schur complement lemma, Newton Leibniz formula and Jensen's inequality, sufficient conditions are derived...