Search

In this paper, an Improved Collaborative Filtering Integrated with Diverse Optimization Strategies (ICFIDOS) is proposed to improve the performance of recommendation system. First, data sparsity problem of collaborative filtering was relieved via Slope One algorithm. Then, Hybrid Shilling Attacks Detection Framework is utilized to help recommendation system to defend the traditional and modern Shilling Attacks, and purify the polluted data. After that, cold start problem of initial trusted degree was resolved via combination with the influence of new users and existed users. Furthermore, similarity computation of Item Preference is improved. Finally, every step of our ICFIDOS algorithm was clarified carefully. In experiment, the efficiency of ICFIDOS is validated and has better performance compared to classical UBCF and K_TUBCF algorithm.

Energy storage power station is an effective regulatory resource applied in many ways. It will play a more and more important role in the future power system. And with the decline of battery cost and the improvement of technical level. Its application and installation will increase greatly in the future. Considering that the market keeps pace with the times. This paper focuses on the research on the transaction variety and transaction mode of energy storage power stations participating in the electricity market. This will help to understand the current practical application of energy storage power stations, and effectively put forward the future development direction.

Promoting clean, low-carbon, safe and efficient use of energy in power grid operation is the direction of energy development. Speeding up the development of new energy, green environmental protection and other industries is the guarantee of sustainable development. With the extensive access of large-scale renewable energy to the power grid, the randomness, intermittence and volatility of its power generation will bring severe challenges to the safe operation of the power grid. Therefore, in order to increase the proportion of new energy with randomness and volatility in energy use, the adjustability of power grid resources is very important. With the goal of improving the adjustable resources and regulation ability of the power grid, this paper expounds how to guide and adjust the resource cost and improve the regulation ability of the power grid in the electricity market environment

Power grid regulation resources is an important content to be considered in the construction of smart grid. With the construction of power market becoming more and more perfect, power system regulation resources play a more and more important role through the market mechanism. However, considering the different operation characteristics and load structure of each provincial power grid (including regional power grid), the role of regulating resources is different. At the same time, the market mechanism in each region is not completely the same. Through a comprehensive investigation of the construction of regulation resources market in various regional and provincial power grids, this paper gives an analysis of the current situation of regulation resources in the market.

Numerous dams have been in service for more than thirty years and are gradually aging. Many of them are defective and risky dams. To detect defects on the dam surface, obtain intuitive and reliable information, and evaluate the safety of the dams in real-time, the "Yulong" Dam Maintenance Submersible (DMS) has been designed and built. This operating platform is not limited by the season and can inspect deep-water dams while obtaining High-definition video data of suspicious leakage points and other defects. By carrying technicians on-site to identify defects, grasp the underwater situation of the dam in real-time, and obtain timely and comprehensive information on the dam surface. The DMS is a complicated system that puts harsh requirements on the control system. The engineering application results show that the control system that we designed and implemented is stable, reliable, and highly accurate, and can effectively complete underwater inspection and operation tasks, thereby ensuring the safe operation of the deepwater dams.

With the increasing of large-scale grid-connected wind power, solar energy and high voltage direct current transmission (HVDC), the power supply structure of power systems has undergone tremendous changes, which leading to the insufficient of rapid frequency modulation capability of the power supply. On the other hand, the poor voltage regulation characteristics of capacitor-based reactive power compensation equipment and the lack of dynamic reactive power support capability are a huge challenge for the power grid. As an important way of energy storage, battery energy storage (BES) has the advantages of flexible power and energy allocation according to the needs of different applications. By optimizing the dynamic characteristics of the BES power station, it can effectively assist the power grid in rapid frequency modulation and voltage regulation. In this paper, an additional frequency control strategy is proposed for the coupling characteristics of active and reactive power in power systems. By controlling the voltage, the BES power station can still participate in the inertial response and primary frequency modulation of the grid under the condition of low state-of-charge (SOC), which improves the regulation ability and the flexibility of operation of the grid. the results of simulation analysis on Matlab/Simulink verify the effectiveness of theoretical analysis.

For realizing the robust heading control of the Dam Maintenance Submersible (DMS) with parameter uncertainty and external disturbance, the controller (BSMC-ESO) is proposed based on the Extended State Observer (ESO) and backstepping sliding mode control algorithm. The ESO is applied to estimate the total disturbance and a new sliding mode control method with the backstepping technique is adopted to compute the control output. The stability and convergence of the whole system are proved by the Lyapunov stability theory and verified by simulation experiments. Simulation results show that the BSMC-ESO has high tracking performance and strong robustness against parameter uncertainty and external disturbance.

This paper studies the influence of synchronous condensers under different initial reactive power states on system transient voltage stability in received power grids with ultra-high voltage direct current (UHVDC) systems. It is helpful for the synchronous condensers incorporating the system static voltage regulation. The voltage stability problem of an actual received power grid with an UHVDC system is introduced. And then, the voltage stability under different initial reactive power states of synchronous condensers and grid voltage control strategies are presented. The scenarios that the synchronous condensers can take part in the system static voltage regulation are summarized. At last, the static voltage regulation model of synchronous condensers is established with an objective function and sets of constraints. By using the actual power grid of UHVDC System, it is verified that the synchronous condenser can participate in the static voltage regulation in this scenarios, and the transient voltage stability of the system will not get worse.

Stroke is the fifth leading cause of death for Americans. Due to critical consequences and cost, an efficient stroke system of care (i.e., acute treatment, post-stroke acute rehabilitation) is necessary. Early determination of hospital discharge disposition is important for stroke management, and can make an immense impact in optimizing acute treatment and planning post-acute rehabilitation with desired outcomes. With the rise of sophisticated machine learning models, many researchers have gained momentum in using such models in their studies. However, due to the lack of explanations that models provide, the integrity of the prediction result is often challenged. Our goal is to predict post-stroke hospital discharge disposition using machine learning models to increase the prediction capability while providing explanations for the results using an interpretation method, such as Local Interpretable Model-agnoistic Explanations (LIME). Our results demonstrate the effectiveness of LIME in providing interpretability to machine learning models and suggests further exploration in performance improvement.

The study of static voltage stability is actually based on asymptotic stability, which is affected by the operation state of the power grid. In this paper, the voltage stability margin index is deduced, and the stability margin of the existing power grid under different actual working conditions is analyzed based on the proposed index. the leading factors and the worst system conditions are determined by practical experience. The correctness of the proposed method is verified by the simulation analysis of the actual provincial power grid.

STATCOM is a typical reactive power regulating equipment, which is more and more widely used in provincial power grid. In this paper, ZX and SD stations will invest in STATCOM with rated capacity of 120 Mvar as an example. Including the ontology characteristics of STATCOM and its impact on the power grid after it is put into operation. The analysis results can be used to evaluate the influence of STATCOM on the voltage security and stability of provincial power grid in practical application.

The screening and ranking of expected accidents is often the first step in the safety and stability prevention and control strategy. The analysis time of the screening process is long and the calculation is large. In this paper, a second-order screening method for anticipatory accidents is proposed. in essence, all the expected events are screened quickly and the target set is obtained. then, according to the characteristics of each event in the target set, some events are selected for margin analysis. The proposed method can quickly obtain the prevention area. The simulation results show that the method is correct.

Because the direct background light generated by the sun is strong, its strong background light power is much higher than the LED’s power, and the signal is completely submerged in the background noise, causing communication interruption. To solve the above problems, this paper establishes a visible light communication system between LED traffic lights and vehicles under strong background light, and introduces the power inversion algorithm into the system to suppress the influence of strong background light, then analyzes and compares the bit error rate of the systems under different signal-to-interference ratio. The simulation results show that the power inversion using the 3-element receiving array can generate a null of about -70dB in the direction of strong background light, so that the noise value of the output is significantly reduced, the signal gain remains unchanged. When the signal interference ratio of system is -15, the BER is 10-4, which can supply the need of communication.

The increasing penetration of wind power generation displaces a major portion of conventional generating plant reducing system inertia and deteriorating system frequency dynamic characteristics. In this sense, the induction machines (IMs) including induction motor loads and induction generators, which occupy a fair proportion in power consumption and generation, should be viewed with some concern in frequency response. Therefore, this paper firstly estimates IMs available inertia. It is shown that the IMs available inertia is affected by both rotor inertia and slip frequency regulation and presents a time-varied property. Besides, the frequency response model of a power system constituted of a SG, an aggregate Type 3 WTs and an IM load is established. The effect of IMs available inertia on frequency dynamics is analyzed based on this system. Time-domain simulation results of the system verify the analyses.

This paper studies the influence of operation parameters on the coupling relationship of each region by perturbation method based on the study of source-network-load characteristics. The influence of any running parameter on the coupling relationship between nodes or regions can be obtained according to the method described in this paper. The influence of electrical parameters on source-network-load characteristic can be further decoupled by analyzing the influence of electrical parameters on the proposed coupling characteristic index. The practical operation system of a province proves the correctness and validity of the proposed method

When the high-speed hydraulic cylinder drives the parts to the stroke terminal, it is often accompanied by strong hydraulic shock. Cushioning is an important aspect in hydraulic cylinder performance, which can greatly reduce the vibration, noise and temperature rise of oil caused by hydraulic shock. In this paper, we presented a variable annular gap cushioning for hydraulic cylinder. Firstly, the working principle of the variable annular gap cushioning in hydraulic cylinder is provided. Secondly, the mathematical model and geometric model of the hydraulic cylinder are established, and the dynamic CFD simulation of the hydraulic cylinder is conducted. Based on the pressure and velocity characteristics of the hydraulic cylinder under different cushioning parameters, the optimal structural parameters of the variable annular gap cushioning can be determined. In addition, the simulation results show that the pressure peak value and pressure fluctuation obviously decrease under the optimal structural parameters of cushioning. The proposed cushion structure and numerical results have scientific and engineering value to promote the innovative development of hydraulic cylinder.

A particle swarm optimization (TVPSO) algorithm with time varying parameters is proposed to improve the performance of particle swarm optimization (PSO) algorithm by two improvements. Aiming at the fact general PSO algorithms have the disadvantages of falling into local optima caused by linearly decreased inertia weight. TVPSO uses the related properties of the trigonometric function to improve the dynamic changes of inertia weight along With Time. The inertia weight maintains a large value in the initial stage, and decreases gradually and reaches a small value at the end. Thus, the global search capability and convergence performance were improved; In order to cope with changes in inertia weight, learning factors also change with time. TVPSO and the other latest particle swarm optimization algorithms are tested on 10 functions at the same time. Experimental results show that TPSO has faster search speed and stronger global search capabilities.

A modified differential evolution (MDE) algorithm is developed to solve the economic load dispatch (ELD) problems without or with valve-point effects (VPE). MDE mainly has two improvements compared to the original differential evolution (DE) algorithm. The whole evolution process is divided into the early and late phases. The early phase uses two mutation strategies, and the late phase only uses a mutation strategy. The operation is suitable for any evolution phase. The disturbance with a small probability is added to avoid falling into the local optima in the mutation operation. The control parameters are adaptively adjusted to improve the convergence performance of MDE. Additionally, a repair method is used to handle equality constraints, which enables MDE to find feasible solutions rapidly. Five ELD cases are selected to testify the effectiveness of the proposed method. Results show that MDE performs better than the other methods for the ELD problems.

With the increase in global temperature, the rise in electricity demand has been accelerated. Air conditioning accounts for most of the residential power demand. In order to improve energy efficiency and manage the A/C power demand, smart meter data can be exploited to extract intelligent information about A/C power usage. In this study, we will demonstrate deep learning-based non-intrusive load monitoring algorithms (NILM) to disaggregate the A/C power consumption from the total power consumption measured by smart meter. Deep learning can provide end-to-end solutions to deal with large-scale data and is widely used to address real world challenging problems. We will use Pecan Street smart meter data to show the performances of the proposed A/C operating cycle classification and A/C power consumption estimation. Our preliminary results will manifest the effectiveness/potential of deep learning-based algorithms and suggest further investigation for performance improvement/model generalization.

In the United States, the residential and commercial sectors are responsible for almost half of total energy consumed. Additionally, air conditioning accounts for the majority of energy consumed in residential buildings. In order to reduce the energy consumption of the building, we identify the behavior of A/C energy consumption by estimating the future A/C load. In this paper, we will explore deep learning-based algorithms to estimate A/C energy consumption using the smart meter data and outdoor temperature data. We will utilize data provided by Pecan Street to validate the performance of proposed task. Our results will demonstrate the strength of deep learning algorithms in solving load forecasting problems, and help in further investigations.

Stroke is the fifth leading cause of death and a major cause of long-term disability in the United States. Because of its life-threatening consequences, the stroke system of care including acute treatment and post-acute rehabilitation is crucial. In order to support the stroke system of care in the age of big data, predictive analytics can be applied to forecast what will happen in the future. For example, we can predict post-stroke discharge disposition at an acute stroke admission, which will facilitate optimizing acute treatment and planning post-acute rehabilitation with the desired outcomes. In this preliminary study, we will explore deep learning for post-stroke discharge disposition prediction and evaluate prediction performance using hospital discharge data provided by Tennessee Department of Health. Deep learning is a powerful tool to deal with large-scale data and provides an end-to-end solution to multi-class classification including dimensionality reduction. Our preliminary results will demonstrate the effectiveness of deep learning and suggest the further exploration for performance improvement and clinical adoption.

Guided modes may exist on periodic structures above the lightline, and they are called bound states in the continuum (BICs). For BICs on a 1D periodic array of circular cylinders, we show that the BICs disappear and become resonant modes when the cylinders are slightly distored so that the reflection symmetries are broken.

Sensitivity analysis is important for evaluating the performance of practical devices which are affected by fabrication errors. It is also very useful in optimal designs for various structures and devices. Based on the Dirichlet-to-Neumann map method, which is an efficient numerical method for modelling photonic crystal devices, we perform a sensitivity analysis for photonic crystal waveguide-cavity systems.

Falls are common and dangerous for the elderly or individuals with decreased independence or functional limitations. Fall recognition is extremely important for fallers, healthcare providers, and society. Immediate fall recognition triggers emergency services and potentially decreases individuals time with injury without care. Acute post-fall intervention works to mitigate life threatening fall consequences, decrease fall risk through rehabilitation, and improve quality of life. Extended from our research on real-time fall risk estimation with the functional reach test and Timed Up and Go test built in mStroke, a real-time and automatic mobile health system for post-stroke recovery and rehabilitation, our investigation here is expanded to include fall recognition by taking advantage of wearable technologies and machine learning algorithms. Up to three wearable sensors are employed to acquire raw motion data related to activities of daily living or falls. Feature selection and classification on the basis of machine learning algorithms are explored for fall recognition. The fall recognition performances are presented to justify their accuracy and reliability. Meanwhile, the effects of sensor placement/location and the feature number on the recognition performance are also discussed in this paper.

One Man Atmospheric System (OMAS), is an articulated anthropomorphic single person submersible which consists of the Suit, Tether Management Systems (TMS), Launch and Recovery Systems (LARS) and Surface Control Systems (SCS). The submersible is presently capable of allowing a pilot to operate at the worksite. OMAS maintains the internal pressure at one atmosphere. There is no need for decompression schedules and it increases the bottom times. Meanwhile, the pilot is put directly at the worksite where they can observe dexterously and control tools directly. So OMAS are most effective. The images in operation which are transferred synchronously to the surface can be used to monitor and guide the operation. Power Distribution & Control System is the most important system for OMAS. It is charged with power distribution of Suit, TMS and SCS, controlling navigating and working for OMAS. It also provide the channel for exchanging of information, the display and alaming for OMAS. According to physical location, control system is divided into three parts: surface control, TMS control and suit control. And these three parts functionally response their own control requirements. Firstly, this paper introduce the general situation of OMAS. Secondly, the design for power distribution & control system of OMAS is provided. Then, debugging for the system, situation of pool trial and sea trial are given, controlling system is improved through pool trial and sea trial. At last, summarize the design for power distribution & control system. Finally, at the end of the paper, and with the benefit of many years experience of underwater technology, some advice is presented.

The quantization method of coupling degree is proposed and used to identify the critical nodes in the power system. This method combines the change of injection current on the basis of network analysis technique. The degree of coupling between the various parts of the power grid after the disturbance can be analysed by this method which overcomes shortcomings of ignoring the different operation states when the network technology is applied to the power grid. The proposed index can accurately reflect the influence degree and has a very clear physical meaning. The proposed method is more in line with the actual situation of the power system when compared with the existing complex network model by the application of the IEEE standard sample system. The method is also simulated on the actual central province Power Grid in China, and result shows that the concept of coupling degree is easy to calculate and can real-time track status changes.

This paper will design a general, flexible, and efficient framework for data acquisition, data compression, and data reconstruction in advanced metering infrastructure (AMI). Compressed distributed sensing will be utilized to acquire load data from smart meters and transmit them to the central control unit. Different sparse binary measurement matrices will be exploited for different time instances when data acquisitions are performed. Each sparse binary measurement matrix corresponds to one data gathering scheme using compressed distributed sensing. This paper proposes to perform joint reconstruction of the two-dimensional load profile at the central control unit. Both spatial and temporal correlations will be explicitly employed to facilitate data reconstruction with high accuracy and fidelity. Meanwhile, the desirable data compression ratio can be achieved.

According to the development of ocean technology and equipment, diving into the deep blue becomes one of aims for international researchers who engage in works about the ocean. Acoustic survey devices(including imaging sonar, positioning sonar, anticollision sonar and Doppler Velocity Log(DVL) etc) are applied widely in deepsea vehicles(including Human Occupied Vehicle(HOV), Remotely Operated Vehicle(ROV), Autonomous Underwater Vehicle(AUV), Glider and Deep-tow etc). The purpose of acoustic survey devices is very explicit, and they are all important for deepsea vehicles when they are working underwater. Acoustic image infront of the deepsea vehicles can be given by imaging sonar, which is the most important acoustic information for deepsea vehicles. Positioning sonar can offer the comparative position between the deepsea vehicle and the mother ship. By combining with the information of the Global Position System(GPS) on the mother ship, we can calculate the absolute position of the deepsea vehicle. The information of all directions is given by anticollision sonar, which can ensure the security of deepsea vehicles. DVL can offer the velocity of the deepsea vehicle, and its information of the altitude can be set as the redundancy of the anticollision sonar which is fixed downwards. Firstly, this paper introduces the application of acoustic survey devices used on deepsea vehicles inland and abroad, including the general situation and development of applications. Secondly, a detailed introduction of acoustic survey devices used on the deepsea vehicles inland(including Jiaolong HOV, Haima ROV, 4500m AUV and OMAS(One Man Atmospheric System) etc) is focused in the paper(take the Jiaolong HOV for example). This section introduces the application of imaging sonar, bathymetric side-scan sonar, positioning sonar and anticollision sonar on the HOV, including the acoustic image and data, collected on the bottom of sea (5000–7000m). Furthermore, the advantages and deficiencies of the application of acoustic survey devices are analyzed. The status quo, problems and requirements of acoustic survey devices applied in deepsea vehicles are also summarized in this section. Finally, at the end of the paper, and with the benefit of many years experience of acoustic survey devices used in deepsea vehicles, some advice is presented about environmental conditions, platform carried and performance index of applying acoustic survey devices.

The scientific cost drive Parameters identification and screening can improve the civil airplane development cost estimation accuracy, and then realize control and management of the development cost effectively. According to the estimation of the conceptual design stage of civil aircraft cost is studied, considering the lack of the information of civil aircraft cost parameters, parameters have been collected exist the problem of Multicollinearity, firstly, through qualitative identification method of civil aircraft cost parameter system and construct a generalized grey relative correlation model and relational clustering model combination method of screening cost drive parameters, algorithm steps are given; and collected to verify the model at home and abroad of 9 different kinds of civil aircraft information, the result shows that the proposed model can effectively screening exit key cost drive parameters. The results show that the prediction method established in this paper has improved significantly on the accuracy, illustrating the effectiveness and scientific of this method.

Researchers are increasingly interested in the secondary use of EHR data to detect specific outcomes and adverse conditions. Our aim is to develop valid, robust, and practical EHR-derived models for identifying postoperative surgical site infections (SSIs). SSIs can be classified into superficial, deep, and organ/space, and are costly with significant morbidity. Compared with administrative/claims data that previous research heavily relied on, our use of EHR data has the potential to allow for the construction of more informative SSI detection models. Unfortunately, secondary use of EHR data can be challenging due to its often incomplete nature -- some specific tests are just ordered to only a subset of patients (e.g., 52% of the surgical patients in our cohort do not have any white blood cell count data within 30 days after the operation). Mostly researchers ignore it by excluding cases or single variables with missing data, or imputing missing values for variables with slight amount of missing data. However, because of the high missingness rate in our data, to simply discard incomplete cases may result in losing important indicators of SSI. In our previous work, we only utilized the complete cases to detect SSIs within 30 days after surgery using the gold standard outcome from a validated national surgical registry -- National Surgical Quality Improvement Project (NSQIP)[1]. In the current study, we sought to explore several popular treatments of missing data. The performance of the models after applying different treatments are compared to that of the reference model based on the complete cases.

This paper will present a general and efficient methodology for data acquisition in Advanced Metering Infrastructure (AMI). Compressed distributed sensing using random walk (CDS(RW)) will be explored to acquire user load data from smart meters. This paper proposes to perform joint reconstruction of 2D user load profile using both spatial and temporal correlations. In this way, high data compression ratio can be achieved. Meanwhile, convex optimization will be the solver for the 2D user load profile reconstruction problem, which can guarantee both convergence and global optimality. Finally, taking power theft classification as a motivated example, this paper will demonstrate the performance will be acceptable even using the reconstructed user load profile for classification.

Photonic crystal heterostructure mode-gap cavities, constructed by connecting two kinds of photonic crystal waveguides with different lattice constants, are important photonic devices. In this paper, we extend the Dirichlet-to-Neumann map method to analyze mode-gap cavities in two-dimensional photonic crystal heterostructures. The efficiency of our method is illustrated by numerical simulations.

A lot of studies provide strong evidence that traditional predictive regression models face significant challenges in out-of sample predictability tests due to model uncertainty and parameter instability. Recent studies introduce particular strategies that overcome these problems. Support Vector Machine (SVM) is a relatively new learning algorithm that has the desirable characteristics of the control of the decision function, the use of the kernel method, and the sparsity of the solution. In this paper, we present a theoretical and empirical framework to apply the Support Vector Machines strategy to predict the stock market. Firstly, four company-specific and six macroeconomic factors that may influence the stock trend are selected for further stock multivariate analysis. Secondly, Support Vector Machine is used in analyzing the relationship of these factors and predicting the stock performance. Our results suggest that SVM is a powerful predictive tool for stock predictions in the financial market.

One of the major objectives of cognitive radar is to form a dynamic closed feedback loop to adapt the spectrum of transmit waveforms to avoid certain interference. In this paper, we build an automatic closed-loop cognitive radar to support experimental study of the radar system in real-world situations. Convex optimization is applied to jointly design sounding waveforms and the matched filters with spectral power suppressed in arbitrary bands and with low correlation sidelobes as well. Target localization is demonstrated under strong interferences. Experimental results are provided to evaluate the performance of the cognitive radar system.

Spectrum sensing is a fundamental problem in cognitive radio. How to sense the presence of primary user promptly in order to avoid the unexpected interference is a key issue to the system. The motivation of our work is to detect the primary user signal using small size data in short time. In this paper, a quickest detection based approach is proposed for spectrum sensing. This approach employs covariance matrix estimation instead of sample covariance matrix as the first step, then the core idea of sequential detection or quickest detection is borrowed and utilized here to improve the performance of traditional eigenvalue based MME and AGM detectors. The main advantage of the proposed approach is that it requires short data to detect quickly and it works at lower SNR environments than some traditional methods. A performance comparison between the proposed approach and other traditional methods is provided, by the simulation on captured digital TV (DTV) signal. The simulation results show this proposed approach exhibits performance improvement while the threshold keeps robust.

Batch Experiments with the combination use of inhibitors were conducted to identify the mechanism of N2O emission from anoxic-aerobic sequencing batch reactors (SBR) acclimated under different oxygenation levels. Oxygenation has an important effect on the N2O emission and most of the N2O was produced during the aerobic phase, in spite of the oxygenation level. During the anoxic phase, nitrate ammonification was the major producer of N2O while denitrification performed as a sink of N2O. During the aerobic phase, both nitrifier denitrification and coupled nitrification-denitrification were responsible for N2O emission in SBR with low and high oxygenation, while nitrifier denitrification was the major source of N2O emission in SBR with mid oxygenation.

Synthetic aperture radar (SAR) is a form of imaging radar. The defining characteristic of SAR is the usage of relative motion between an antenna and its target region to provide long-term coherent-signal variations that can be exploited to obtain fine spatial resolution for radar image. Due to the improvement of SAR data acquisition technique and the flexibility of SAR sensor deployment, multi-pass SAR imageries can be easily obtained. More information and intelligence can be extracted from SAR imageries. This paper will address one challenging issue for SAR data fusion, i.e. multi-pass SAR change detection. Two methods are proposed in this paper to perform multi-pass SAR change detection. One method is based on robust principal component analysis (PCA) and the other method uses template matching plus thresholding. Both methods explore the local statistics to indicate the change of each pixel in the SAR imageries. The visualization performances illustrate the potential of these two methods for the issue of SAR change detection.

A Cognitive Radio Network (CRN) based Wireless Sensor Network (WSN), as an extension of CRN, is explored for radio frequency (RF) passive target intrusion detection. Compared to a cheap WSN, the CRN based WSN is expected to deliver better results due to its strong communication functions and powerful computing ability. Issues addressed in this paper include experimental architecture, waveform design, and machine learning algorithm for classification. In particular, passive target intrusion is experimentally demonstrated using multiple WARP platforms that serve as the cognitive/sensor nodes. In contrast to traditional localization methods relying on radio propagation properties, the technique used in this research is based on machine learning with measured data, considering complicated multipath environment and high dimensional sensing data collected by the CRN based WSN. Preliminary experimental results are quite encouraging, suggesting that a large-scale CRN based WSN supported by machine learning techniques has promising potential for passive target intrusion detection in harsh RF environments.