Search

Aiming at the problems of low recognition accuracy and low real-time in the existing human motion recognition in the complex background, a registration model recognition algorithm of human posture limit learning machine(ELM) is proposed, which uses Kinect measurement to calculate the joint association angle. Through Kinect, the spatial coordinates of human joints are obtained. Then, three-point method is used to calculate the correlation angle of human joints, and the elm registration model corresponding to the correlation angle of the target attitude joints is defined. Finally, the movement recognition test of the elm registration model of human body posture is carried out with the joint correlation angle calculated based on Kinect real-time measurement. The experimental results show that this method has high real-time and accuracy, strong robustness and generalization ability, and can identify more pose by defining more joint angles of target pose to establish elm registration model of human pose.

Multi-string pattern matching is a crucial building block for many network security applications, and thus of great importance. Since every byte of a packet has to be inspected by a large set of patterns, it often becomes a bottleneck of these applications and dominates the performance of an entire system. Many existing works have been devoted to alleviate this performance bottleneck either by algorithm optimization or hardware acceleration. However, neither one provides the desired scalability and costs that keep pace with the dramatic increase of the network bandwidth and network traffic today. In this paper, we present BOLT, a scalable and cost-efficient multi-string pattern matching system leveraging the capability of emerging programmable switches. BOLT combines the following two techniques, a smart state encoding scheme to fit a large number of strings into the limited memory on the programmable switch, and a variable k-stride transition mechanism to increase the throughput significantly with the same level of memory costs. We implement a prototype of BOLT and make its source code publicly available. Extensive evaluations demonstrate that BOLT could provide orders of magnitude improvement in throughput which is scalable with pattern sets and workloads, and could also significantly decrease the number of entries and memory requirement.

Accurate forecasting of power load is of great significance to economic dispatch and safe operation of power grid. A short-term load forecasting model based on EEMD- LN-GRU is proposed according to the characteristics of uncertainty and nonlinearity of power load. In order to solve the problem of power load fluctuation and mode aliasing caused by empirical mode decomposition (EMD), the original load time series signal is decomposed into multiple intrinsic mode functions (IMF) and residual error component by ensemble empirical mode decomposition (EEMD). Each component signal is predicted by the gating recurrent unit (GRU) after layer normalization (LN). Finally, the components are predicted. The results are recombined, and then the peak valley value of the prediction sequence obtained by peak valley value correction strategy is modified to obtain the final load series. Taking the real load data of power plant of Slovak power company as an example, this method is compared with LSTM, GRU and other methods, and the results show that the proposed method has higher prediction accuracy for load forecasting.

In order to solve the problem that a single attacking missile cannot effectively intercept the maneuvering target when its detection ability is suppressed, the unmanned aerial vehicle (UAV) acting as a detection role is considered and the multi-aircraft cooperative guidance scenario with the detection role is proposed based on the cooperative interception geometry relationship. Firstly, according to the relative kinematics and dynamics equations of particles, a cooperative interception model is established. Secondly, under the condition that the detection capability of the attacking missile is limited, the cooperative guidance requirement is analyzed and a feasible cooperative guidance scenario is derived. Finally, the cooperative guidance law is designed based on the proportional and differential control method which can achieve the goal of stabilizing the line-of-sight axis and cooperative guidance by modulating the angle between the Detector-to-Missile and Detector-to-Target LOS axis. Mathematical simulations show that when the target performs sinusoidal and uniform acceleration maneuvering, the cooperative guidance law can both ensure that the attacking missile intercepts the target effectively under the guidance of the detector, which confirms the availability and advantage of the cooperative guidance law.

This paper presents a new solution for SAR image target recognition by designing a light level convolutional neural network (CNN), and raises an unsupervised detection method by taking the advantages of convolutional features. Firstly, we train a shallow convolutional network by using Moving and Stationary Target Recognition (MSTAR) dataset to classify SAR targets. Then, we extract the outputs of the convolutional layers, and locate the position of the target in the image by the max sampling and clustering. Tested on the MSTAR data set, our method gets satisfied accuracy in classify task, and realizes unsupervised target detection effectively.

In order to improve the attack effect from the top of maneuver target for air-to-ground missile, three-dimensional proportional navigation model with over-gravity compensation is established based on the relative motion equations. Three-dimensional ballistic trajectory of the missile attacking target is simulated based on the guidance and control closed-loop circuit. The effect of the over-gravity compensation proportional navigation parameters on the flight characteristics is evaluated. The simulation results indicate the effectiveness of the proposed navigation law. In addition, the theoretical analysis results are also verified.

Locating an inconspicuous object is critical in remote distance space-based measurement. An indirect location method is presented based on the locations of the salient regions, from which the location of the inconspicuous object is transferred indirectly. The location problem of inconspicuous object is considered via affine invariability, which existed in the remote measurement. Salient region detection and matching are carried out first to match the real-time images and template images, and then affine transform is established related to the two images via affine function. The ultimate aim of location is achieved by homography matrix, which is calculated by homologous feature points between the real time image and template image. At the end of this paper, experiment is performed to prove the validity of this method.

A metric direct (feature-less) online monocular SLAM algorithm is proposed, which, different from the traditional monocular SLAM algorithm, can obtain the absolute scale of the world. A calibrate object is used to calculate the relative pose of the first key frame and initialize the depth map, introducing the metric scale into the reconstruction meanwhile. Highly accurate pose estimation is achieved by direct image alignment, and the depth of image points is obtained by filtering over a large number of pixel-wise small-baseline stereo comparisons. With those, the environment is reconstructed in real-time as pose-graph of key frames with associated semi-dense depth maps. Also an explicitly scale-drift aware formulation is proposed to align the scale between key frames. The algorithm is tested online, and is proved to run in real-time on a CPU.

In this paper, the infrared small target detection in video sequence is investigated. A collaborative structured sparse coding model which incorporates the L 1,2 and L 2,1 regularization terms is proposed to detect the infrared small target in video sequence. Further, online dictionary learning is embedded into the model and temporal information is incorporated to eliminate the clutters and noises. Finally, four simulation datasets are constructed to test the proposed method and the experimental validation shows promising results.

Considering on an aircraft threatened by an interceptor, one of the effective penetration strategies is to release a defender from the aircraft to confront the interceptor. In this case, the aircraft, the defender and the interceptor constitute the three-body guidance relationship, and the cooperation of the aircraft and its defender to achieve the best tactical effects turns into the concerned problem. This paper studies the triangle interception guidance problem via the finite-time theory. The paper presents linear system Input-Output Finite-time Stabilization (IO-FTS) method. A state feedback controller is designed, via Linear Matrix Inequalities (LMIs), to guarantee the system IO-FTS. The proposed methods are applied to the guidance design of triangle interception guidance problem. The simulation result illustrates the effectiveness of the proposed method.

Similarity measure is the most basic and important element of scene matching algorithm. In this paper, the similarity measures for scene matching are summarized. According to the requirements of scene matching system, the performances of mean absolute difference, mean square difference, and normalized product correlation are analyzed through matching simulation experiments. Based on researched of space geometric relationship between the reference image vectror and the real-time image vector, a novel projective similarity measure (Proj) is proposed. Moreover, three scene matching algorithms are designed based on the projective measure. The simulation experimental results demonstrate the good performance of the three algorithms and the validity and efficiency of Proj.

Thought is the forerunner of action in the new period, to improve college students' ideological and political education work for college students, as the implementation of ideological and political education is one of the main position will outline of modern Chinese teachers should be guided by the scientific development concept, properly, strengthen occupational role to the overall development of self consciousness and service consciousness, vision consciousness, dedicated awareness and innovation awareness and actively cultivate correct role consciousness, may raise the outline of modern Chinese history teaching effects of college students, develop a new era of suitable for moral character

For the code tracking loop, two new code phase discriminator algorithms based on DLL structure were proposed. The performances of the two algorithms were studied by using theoretical method and experimental method with analog IF signal data and actual IF signal data respectively. The experimental results indicate that both the two algorithms discriminate and track the code phase offset effectively. At the same time, it points out their own application environment, which is very useful for the design and implementation of GPS signal tracking loop.

This study was aimed to assess the levels of selected heavy metals and herbicides in soils under different land usage and plant products from an agricultural area of Tieling district, Liaoning Province, China. Cd, Pb, atrazine, acetochlor and butachlor were analyzed from maize, paddy and vegetable field. Cd concentrations in 40% of vegetable soils and 6.67% of rice soils were higher than i°Environmental quality standard for soilsi± (GB15618-1995), respectively. Cd concentrations in20% of vegetable samples were higher than i°Pollution-Freevegetables standardi± (GB18406.1-2001) and Pb concentrations of all vegetable samples were lower than the standard. The Cd concentrations in maize and rice samples did not exceedi°Maximum levels of contaminants in foodsi±(GB2762-2005), Pb concentrations in 26.7% of maize samples and 13.3% of rice samples were higher than levels. The vegetables affected by Cd pollution were more serious than maize and rice, but the latter affected by Pb pollution were more serious than the former. Residues of three herbicides were detected in soils, with levels lower than the relevant limit. The residue levels of atrazine, acetochlor and butachlor in soils were vegetable soil> maize soil> paddy soil, maize soil> vegetable soil> paddy soil, paddy soil > vegetable soil > maize soil, separately. Atrazine was detected in all vegetable samples due to high application rates of previous periods in vegetable. Acetochlor and butachlor were detected in rice samples, with levels lower than the relevant limit. Detection rate (83.7%) of acetochlor was much higher than that (10.0%) of butachlor. Detection values of butachlor in rice were 1.92, 2.60 and 3.99ƒEg.kg-1, respectively. The content of atrazine, acetochlor and butachlor in rice and vegetable should be monitored continuously, as they are well known to be persistent and accumulative chemicals.

was established for detecting the trace arsenic in water. With the existence of strong acid, the trace concentrations of arsenic after concentrating by JM-33resin reacted with disulfoantipyl methane to form a 1iA2complex which the maximum absorption wave lied at325nm with an apparent molar absorptivity of 1.74iA104l£¯( moli¤cm)£®When the concentration of arsenic was between 0.02-0.5mg/l, the reaction conformed to Beeri¯slaw. The effects of some important factors were also investigated, which showed the method was simple, selective and with high precision and accuracy. This method had been used to determine trace arsenic in the real water samples and spiked water samples with satisfactory results.

Stand-alone electronic equipments are important parts of a satellite system, inside which remainders will bring great harm to reliability of the satellite system. Particle shock noise detection (PIND) is an effective method to detect the remainders in stand-alone electronic equipments. Signal to noise ratio (SNR) of detection signal is a bottleneck to improve processing accuracy and robustness of testing software. In this paper, key factors of wavelet threshold de-noising method were investigated and a wavelet threshold algorithm was proposed to improve the SNR of remainder detection signal of the stand-alone electronic equipments. The analytic results proved that the wavelet threshold method efficiently improved the de-noising effect.

This paper studies adaptive six Degree-of-Freedom (6DOF) coordination control of spacecraft formation with coupling time delay. By rearranging the combined translational and rotational dynamics into a unified Euler-Lagrange formulation, the developed controllers can be applied directly to maintain formation as well as desired relative attitudes. Then, based on a novel adaptive control architecture, distributed controllers are developed, which allow for parameter uncertainties and unknown external disturbances. It is shown that arbitrary desired formation and relative attitudes among spacecraft can be obtained with constant coupling time delays. Simulations are provided that demonstrate the effectiveness of our theoretical results.

This paper describes a computer vision method for reconstruction of three-dimensional object, this method can reconstruct the object real time and has high precision. In this approach the structured light is used and the light beam is projected to the surface of object. This work is composed of three parts: the first is the introduction of the light-section method and the experimental platform based on it. The information of the three-dimensional object surface can be acquired by the platform. In the second part a coordinate transformation algorithm is introduced, including camera calibration and calibration of the central axis of a turntable. And then the space information about object surface can be restored from the two-dimensional plane coordinate of the surface. Finally, a method for plotting the contour of the three-dimensional object surface using OpenGL is described in detail. In this paper, an experiment on a small flowerpot has been carried out and the three-dimensional contour of the flowerpot has been obtained. The result shows a high precision of the three-dimensional reconstruction.

We address the problem of 6DOF spacecraft formation control in a leader-follower architecture using Robust Integral of the Sign Error (RISE) based Neural Network (NN) technique. Based on 6DOF relative dynamic model of the spacecraft formation, RISE based NN is introduced to approximate the dynamics of the follower as well as various practical disturbances. It is shown that the errors of the entire formation closed-loop are guaranteed to be asymptotical stable (AS), which takes significant advantage over the typical Uniformly Upper Bounded (UUB) property of most NN controllers in high-precision formation tasks. Numerical simulation is provided to illustrate the effectiveness of the proposed algorithm.

This paper presents a combination of intelligent learning algorithm, the Support Vector Machine, and the recognition of star pattern in Celestial Navigation. Considering the star pattern recognition's character, noticing the advantages of SVM in learning competence, the paper proposes a solution to star pattern recognition with multi-kernel SVM. A multi-kernel algorithm bases on Genetic Programming is designed. Topics of multi-kernel function generation are cited in detail, and a star pattern recognition routine including an “Indexing + recognition” scheme, feature vector definition and generation, SVM training realization are designed and realized.

An uniplanar capacitance sensor is a core device of high-precision instrument applied to measure wood moisture content, permittivity, etc. In nature, it is a special case of a capacitance sensor of unspecific form. This paper describes the numerical solution to a mathematical model of uniplanar capacitance sensor. An uniplanar capacitance sensor with specific size was taken as an example in this study, and discrete difference scheme - a numerical analysis method - was utilized to establish the discrete approximate model, which is convenient to calculate and study the mathematical model of partial differential equations established in previous research work. The approximate model obtained was adopted to analyze the working conditions of an uniplanar capacitance sensor and to find out the relationship between the permittivity e of wood and the capacity C of capacitance sensor. This result can be applied to the practice of engineering and the method presented in this study is easily extended to uniplanar capacitance sensors of general size and structure.

This paper diagnoses the emission spectra of the Ar plasmas, which are excited at different pressure from 0.1 to 0.8atm, and Ar-air plasma in open air with a unique microwave plasma generator—APMPS+JET system1. The results demonstrate that the previous excited atom stations die out obviously as air is added; various factors play a part in the transform of atomic spectra lines, such as energy input, and voltage as well as the proportion of air input. With the decrease of pressure, the electron temperature and the intensity of spectrum increases in Ar plasma; the electron temperature is about 1.5∼2eV at 0.8atm and rises to 4∼C5eV at 0.1atm; the low wavelength components strengthen simultaneously.

Aiming to the shortages of fuzzy c-means clustering applied to pattern recognition, an improved method by genetic algorithm is proposed. This method can not only automatically optimizes the classification number, but also search the global optimal solution for the clustering center. The experimental results demonstrate this proposed method is excellent for pattern recognition.

Automatic Fault detection and diagnosis for sensor is necessary, which affects the performance of the control system seriously. The KPCA effectively captures the nonlinear relationship of the process variables, which computes principal component in high-dimensional feature space by means of integral operators and nonlinear kernel functions. The KPCA method was used in diagnosing for four familiar sensor faults. At first it detected fault by Q statistics, at second it identified fault by T2 contribution percent variation. The experiment showed the KPCA method had good performance in fault detection and diagnosis.

Carrier tracking is one of the key technologies of software GPS receiver, whose performance mainly depends on the carrier tracking algorithm. According to a typical structure of carrier tracking loop, this paper introduces the principles of the commonly used devices in carrier tracking, such as four-quadrant frequency discriminator (FQFD) algorithm and the principle of expanded Kalman filter (EKF) algorithm, then analyses the tracking performances of these two algorithms with the input signals by setting the frequency as a constant value, or a step equation, or a simple equation or a quadratic equation of time. Based on these two algorithms, a new hybrid carrier tracking algorithm is designed. The performance of newly hybrid tracking algorithm is simulated and the result demonstrates that it is able to track carriers with different sorts of frequency changing laws, with small tracking errors.

In order to improve the reliability and matching precision of image matching navigation system, a novel filtering and fusion algorithm for sequence image matching is proposed. The essence of the matching results filtering and fusion problems for aircraft matching navigation system is firstly analyzed. Then, a new Windows Median Filter (WMF) is designed to filtering the longitudinal matching position sequence. Furthermore, based on Least Squares Estimate (LSE) method, the filtering and fusion algorithm is presented, the principle and flowchart of the algorithm is given. Simulation experiments and practical examples show the feasibility and availability of the proposed algorithm in eliminating the mismatching points and random matching disturbance for sequence image matching navigation system.