It is known that long-time integration an effective method for improving the detection performance ofweak and maneuvering radar targets.However, with increase of time, problems across range unit (ARU) Doppler (ADU) may severely limit high-speed targets. The generalized Radon-Fourier transform (GRFT) can optimally deal above problems, but its implementation needs ergodic multidimensional search. To reduce complexity, a novel hybrid approach proposed in this paper to speed up realization GRFT....

The Radon-Fourier transform (RFT) with a long coherent integration time has recently been proposed for detecting moving target an across range cell (ARC) effect. However, without effective clutter suppression, will also be integrated via the RFT, which may affect weak detection. Based on maximal signal-to-clutter-plus-noise ratio (SCNR) criteria, novel adaptive RFT (ARFT) is in this paper to effectively detect "low-observable" background. ARFT can combine and suppression by introducing...

Long-period pulse components in near-field ground motions have been known to cause significant damage flexible structures. In this paper, an empirical mode decomposition (EMD) approach is used identify and separate the dominant pulse, high-frequency, long-period motions. The pulses synthesized by analytical model recently proposed authors are compared extracted from recorded earthquake records using EMD, responses of linear nonlinear structures subjected these analyzed. results demonstrate...

In this paper various semi-active control strategies and linear passive dampers are applied to the benchmark building model evaluate their effectiveness in reducing response quantities of subject prescribed earthquakes. These include: continuous pulse friction (SACPF) controller, discontinuous (SADPF) viscous (LPV) damper, smooth boundary layer (SBLSAF) resetting stiffness damper (RSASD) visco-elastic (VEF) damper. Both SACPF SADPF controllers designed by using augmented structure-pulse...

In this paper, a fall detection method is proposed by employing deep learning and multi-sensors fusion. Continuous wave radar optical cameras are used simultaneously to capture human action information. Based on the abstraction of both microwave characteristics captured information, multiple convolutional neural network (CNN) realize information training recognition. Due fusion multi-sensor overall performance system can be improved remarkably. Detailed experiments given validate method.

Long-time integration is an effective method to improve target detection performance in a noisy background. However, when detecting high-speed and maneuvering targets by long-time integration, it easy encounter the across range unit Doppler effects, which deteriorate of algorithms with low computational complexity, such as moving hybrid (HI). The generalized Radon-Fourier transform (GRFT) detector has proven have best performance. GRFT high burden for ergodic search multidimensional motion...

In this paper, the performance of different active controllers, including an with pulse filter (APF) controller, in which controller is designed using a ground motion recently proposed by authors, has been investigated for benchmark base-isolated building model subject to various near-field motions. The utilizes period, damping factor and shape from recorded motions represent dominant kinematic characteristics actual pulses. Comparison control approaches, such as linear quadratic gaussian...

Health and usage monitoring as a technique for online test, diagnosis or prognosis of structures systems has evolved key technology future critical systems. The technology, often referred to HUMS is usually based around sensors that must be more reliable than the system structure they are monitoring. This paper proposes fault tolerant sensor architecture demonstrates feasibility realising this through design dual mode humidity/pressure MEMS with an integrated temperature function. simple...

It is known that accurate motion parameter estimation makes significant benefits to radar target tracking, imaging and recognition. However, maneuvering targets usually cause across-range-unit across-Doppler-unit effects, which make it difficult accurately estimate the parameters of target. The generalized Radon-Fourier transform (GRFT) has been proposed vector targets, but only noise scenario considered. In this paper, normalized adaptive GRFT (NAGRFT) first for estimation, proven be...

Translational compensation is a crucial step in inverse synthetic aperture radar (ISAR) imaging. However, nonparametric methods may collapse under the condition of low signal-to-noise ratio (SNR), and high-order parametric usually have large computational load. In this article, two cascaded integration are proposed to solve problems above based on generalized Radon-Fourier transform (GRFT). The first method models translational motion as polynomial utilizes subaperture GRFT (SAGRFT), which...

For modern synthetic aperture radar (SAR), it has much more urgent demands on ground moving target indication (GMTI), which includes not only the point targets like cars, truck or tanks but also distributed river ocean surfaces. Among existing GMTI methods, displaced phase center antenna (DPCA) can effectively cancel strong clutter and been widely used. However, its detection performance is closely related to target's signal-to-clutter ratio (SCR) as well radial velocity, cannot detect weak...

The optimal detector of the moderately fluctuating model has been shown to be a weighted noncoherent integrator sampled pulse returns. Nevertheless, calculation weights in needs dynamic inversion covariance matrix, which results huge computation complexity. In this paper, general description hybrid integration (HID) for target is given achieve good balance between detection performance and HID, time divided into several subapertures with coherent implemented within each subaperture, among...

In the area of computational intelligence, recommender systems play important roles in both commercial and social activities. Among its implement methods, matrix factorization is one most popular approach with various merits. However, traditional methods either roughly aggregate it into final task or ignore users/items similarity, which may easily introduce cause over-fitting under-fitting. To tackle this problem, paper, we propose a similarity based for by considering global local...

In this paper, effectiveness of various semi-active friction control strategies in protecting the base-isolated benchmark building model with pendulum systems as isolators and subject to prescribed near-field ground motions has been investigated. These include discontinuous pulse filter (SADPF) controller, continuous (SACPF) controller visco-elastic (VEF) controller. addition, linear passive viscous (LPV) damper is also considered for comparison. Both SACPF SADPF controllers are designed by...

This paper presents a technique that enables online testing of sensors through the superposition test stimulus onto measurand. Perturbations due to surrounding environment can very often introduce fluctuations in output is major concern for this type sensor testing. proposes novel approach tackle problem by encoding using pseudo-random sequence. The trade-off between level rejection perturbation and overall time studied. In case MEMS accelerometer considered paper, it demonstrated more than...

A novel on-line monitoring technique for a range of MEMS and integrated sensor systems is presented based on the injection test stimuli into bias structure transducer functions. The "Bias Superposition" utilises both signal extraction techniques to achieve an indication structural integrity interface. has been successfully applied thick film conductance sensor.

A centralized multisensor unscented joint probabilistic data association algorithm, CMSUJPDA, is proposed for the multitarget tracking problem of nonlinear system in clutter. In UKF used propagation state distribution at first. Then measurements to track implemented according method JPDA. Based on this, CMSUJPDA algorithm derived by use idea sequential MSJPDA. Due higher order accuracy than EKF, probabilities and estimates our are not affected linearization error. Hence compared with...

In the over-the-horizon radar (OTHR), coherent integration time can approach to tens of seconds. such a long time, effects across range unit and Doppler uncooperative targets may become obvious, which will cause SNR loss for traditional moving target detection methods like MTD. this paper second-order generalized Radon-Fourier transform (GRFT) is introduced OTHR highly maneuvering target, accumulate energy along all possible trajectories after phase compensation obtain improvement...

Detection and motion estimation of weak, high-speed maneuvering targets are urgently needed in modern radar, long-time integration is an effective method to improve the performance. This paper studies parameter based on hybrid a long time. For uniformly accelerated targets, this derives closed-form expressions resolutions range, velocity acceleration integration, which different from existing results coherent methods. work helpful analyze performance as well speed up implementation...

Health and usage monitoring of structures active systems has received a wealth attention in recent years. Microsystems are seen as core technology for the realisation these monitors they offer potential multi-sensor integration with electronics, wireless connectivity future self-powering capability. This article focuses on solutions aircraft wiring where on-line detection degradation incipient failure would deliver improved safety enhanced maintenance efficiency. Several approaches presented...

Long-time integration is an effective method to improve the signal-to-noise ratio (SNR) and detection performance of highly maneuvering targets. The existing methods for long-time mainly focus on detecting targets with high-order motion under fixed model, losing sight possibility that may have abrupt change at unknown instant. Therefore, this paper proposes a novel target model (ACM) based generalized Radon-Fourier transform (GRFT), which could coherently integrate echoes before after ACM...

Fault-tolerance is an important design requirement in critical sensor applications. It conventionally achieved by using redundant components, which increases system's size, cost and complexity, that are often sacrificed or comprised due to associated limitations. This paper proposes a novel solution achieve fault-tolerance at the system level instead of usual approach targets component level. The architecture consists multi-functional sensors used replace conventional single mode sensors,...