COMBINATION OF DISTRIBUTED KALMAN FILTER AND BP NEURAL NETWORK FOR ESG BIAS MODEL IDENTIFICATION

By combining the distributed Kalman filter （DKF） with the back propagation neural network （BPNN）,a novel method is proposed to identify the bias of electrostatic suspended gyroscope （ESG）. Firstly,the data sets of multi-measurements of the same ESG in different noise environments are ＂mapped＂ into a sensor network,and DKF with embedded consensus filters is then used to preprocess the data sets. After transforming the preprocessed results into the trained input and the desired output of neural network,BPNN with the learning rate and the momentum term is further utilized to identify the ESG bias. As demonstrated in the experiment,the proposed approach is effective for the model identification of the ESG bias.

Self-Triggered Consensus Filtering over Asynchronous Communication Sensor Networks

In this paper,a self-triggered consensus filtering is developed for a class of discrete-time distributed filtering systems.Different from existing event-triggered filtering,the self-triggered one does not require to continuously judge the trigger condition at each sampling instant and can save computational burden while achieving good state estimation.The triggering policy is presented for pre-computing the next execution time for measurements according to the filter’s own data and the latest released data of its neighbors at the current time.However,a challenging problem is that data will be asynchronously transmitted within the filtering network because each node self-triggers independently.Therefore,a co-design of the self-triggered policy and asynchronous distributed filter is developed to ensure consensus of the state estimates.Finally,a numerical example is given to illustrate the effectiveness of the consensus filtering approach.

Outlier-resistant distributed fusion filtering for nonlinear discrete-time singular systems under a dynamic event-triggered scheme

This paper investigates the problem of outlier-resistant distributed fusion filtering(DFF)for a class of multi-sensor nonlinear singular systems(MSNSSs)under a dynamic event-triggered scheme(DETS).To relieve the effect of measurement outliers in data transmission,a self-adaptive saturation function is used.Moreover,to further reduce the energy consumption of each sensor node and improve the efficiency of resource utilization,a DETS is adopted to regulate the frequency of data transmission.For the addressed MSNSSs,our purpose is to construct the local outlier-resistant filter under the effects of the measurement outliers and the DETS;the local upper bound(UB)on the filtering error covariance(FEC)is derived by solving the difference equations and minimized by designing proper filter gains.Furthermore,according to the local filters and their UBs,a DFF algorithm is presented in terms of the inverse covariance intersection fusion rule.As such,the proposed DFF algorithm has the advantages of reducing the frequency of data transmission and the impact of measurement outliers,thereby improving the estimation performance.Moreover,the uniform boundedness of the filtering error is discussed and a corresponding sufficient condition is presented.Finally,the validity of the developed algorithm is checked using a simulation example.

FPGA Implementation of Deep Leaning Model for Video Analytics

In recent years,deep neural networks have become a fascinating and influential research subject,and they play a critical role in video processing and analytics.Since,video analytics are predominantly hardware centric,exploration of implementing the deep neural networks in the hardware needs its brighter light of research.However,the computational complexity and resource constraints of deep neural networks are increasing exponentially by time.Convolutional neural networks are one of the most popular deep learning architecture especially for image classification and video analytics.But these algorithms need an efficient implement strategy for incorporating more real time computations in terms of handling the videos in the hardware.Field programmable Gate arrays(FPGA)is thought to be more advantageous in implementing the convolutional neural networks when compared to Graphics Processing Unit(GPU)in terms of energy efficient and low computational complexity.But still,an intelligent architecture is required for implementing the CNN in FPGA for processing the videos.This paper introduces a modern high-performance,energy-efficient Bat Pruned Ensembled Convolutional networks(BPEC-CNN)for processing the video in the hardware.The system integrates the Bat Evolutionary Pruned layers for CNN and implements the new shared Distributed Filtering Structures(DFS)for handing the filter layers in CNN with pipelined data-path in FPGA.In addition,the proposed system adopts the hardware-software co-design methodology for an energy efficiency and less computational complexity.The extensive experimentations are carried out using CASIA video datasets with ARTIX-7 FPGA boards(number)and various algorithms centric parameters such as accuracy,sensitivity,specificity and architecture centric parameters such as the power,area and throughput are analyzed.These results are then compared with the existing pruned CNN architectures such as CNN-Prunner in which the proposed architecture has been shown 25%better performance than the existing architectures.

Robot SLAM with Ad hoc wireless network adapted to search and rescue environments

An innovative multi-robot simultaneous localization and mapping(SLAM)is proposed based on a mobile Ad hoc local wireless sensor network(Ad-WSN).Multiple followed-robots equipped with the wireless link RS232/485module act as mobile nodes,with various on-board sensors,Tp-link wireless local area network cards,and Tp-link wireless routers.The master robot with embedded industrial PC and a complete robot control system autonomously performs the SLAM task by exchanging information with multiple followed-robots by using this self-organizing mobile wireless network.The PC on the remote console can monitor multi-robot SLAM on-site and provide direct motion control of the robots.This mobile Ad-WSN complements an environment devoid of usual GPS signals for the robots performing SLAM task in search and rescue environments.In post-disaster areas,the network is usually absent or variable and the site scene is cluttered with obstacles.To adapt to such harsh situations,the proposed self-organizing mobile Ad-WSN enables robots to complete the SLAM process while improving the performances of object of interest identification and exploration area coverage.The information of localization and mapping can communicate freely among multiple robots and remote PC control center via this mobile Ad-WSN.Therefore,the autonomous master robot runs SLAM algorithms while exchanging information with multiple followed-robots and with the remote PC control center via this local WSN environment.Simulations and experiments validate the improved performances of the exploration area coverage,object marked,and loop closure,which are adapted to search and rescue post-disaster cluttered environments.

Distributed Resilient Fusion Filtering for Nonlinear Systems with Random Sensor Delays:Optimized Algorithm Design and Boundedness Analysis

This paper is concerned with the distributed resilient fusion filtering(DRFF)problem for a class of time-varying multi-sensor nonlinear stochastic systems(MNSSs)with random sensor delays(RSDs).The phenomenon of the RSDs is modeled by a set of random variables with certain statistical features.In addition,the nonlinear function is handled via Taylor expansion in order to deal with the nonlinear fusion filtering problem.The aim of the addressed issue is to propose a DRFF scheme for MNSSs such that,for both RSDs and estimator gain perturbations,certain upper bounds of estimation error covariance(EEC)are given and locally minimized at every sample time.In the light of the obtained local filters,a new DRFF algorithm is developed via the matrix-weighted fusion method.Furthermore,a sufficient condition is presented,which can guarantee that the local upper bound of the EEC is bounded.Finally,a numerical example is provided,which can show the usefulness of the developed DRFF approach.

Convergence analysis of distributed Kalman filtering for relative sensing networks

We study the distributed Kalman filtering problem in relative sensing networks with rigorous analysis.The relative sensing network is modeled by an undirected graph while nodes in this network are running homogeneous dynamical models. The sufficient and necessary condition for the observability of the whole system is given with detailed proof. By local information and measurement communication, we design a novel distributed suboptimal estimator based on the Kalman filtering technique for comparison with a centralized optimal estimator. We present sufficient conditions for its convergence with respect to the topology of the network and the numerical solutions of n linear matrix inequality（LMI） equations combining system parameters. Finally, we perform several numerical simulations to verify the effectiveness of the given algorithms.

Distributed Kalman filter for UWB/INS integrated pedestrian localization under colored measurement noise

Colored Measurement Noise(CMN)has a great impact on the accuracy of human localization in indoor environments with Inertial Navigation System(INS)integrated with Ultra Wide Band(UWB).To mitigate its influence,a distributed Kalman Filter(dKF)is developed for Gauss-Markov CMN with switching Colouredness Factor Matrix(CFM).In the proposed scheme,a data fusion filter employs the difference between the INS-and UWB-based distance measurements.The main filter produces a final optimal estimate of the human position by fusing the estimates from local filters.The effect of CMN is overcome by using measurement differencing of noisy observations.The tests show that the proposed dKF developed for CMN with CFM can reduce the localization error compared to the original dKF,and thus effectively improve the localization accuracy.

On distributed Kalman filter based state estimation algorithm over a bearings-only sensor network

This paper studies the distributed state estimation problem for a class of discrete-time linear time-varying systems over a bearingsonly sensor network.A novel fusion estimation algorithm of the distance between the target and each sensor is constructed with the mean square error matrix of corresponding estimation being timely provided.Then,the refined estimation of distance is presented by minimizing the mean square error matrix.Furthermore,the distributed Kalman filter based state estimation algorithm is proposed based on the refined distance estimation.It is rigorously proven that the proposed method has the consistency and stability.Finally,numerical simulation results show the effectiveness of our methods.

Distributed adaptive Kalman filter based on variational Bayesian technique

In this paper, distributed Kalman filter design is studied for linear dynamics with unknown measurement noise variance, which modeled by Wishart distribution. To solve the problem in a multi-agent network, a distributed adaptive Kalman filter is proposed with the help of variational Bayesian, where the posterior distribution of joint state and noise variance is approximated by a free-form distribution. The con vergence of the proposed algorithm is proved in two main steps: n oise statistics is estimated, where each age nt only use its local information in variational Bayesian expectation (VB-E) step, and state is estimated by a consensus algorithm in variational Bayesian maximum (VB-M) step. Finally, a distributed target tracking problem is investigated with simulations for illustration.

Event-based Control and Filtering of Networked Systems:A Survey

In recent years, theoretical and practical research on event-based communication strategies has gained considerable research attention due primarily to their irreplaceable superiority in resource-constrained systems（especially networked systems）. For networked systems, event-based transmission scheme is capable of improving the efficiency in resource utilization and prolonging the lifetime of the network components compared with the widely adopted periodic transmission scheme. As such, it would be interesting to 1） examining how the event-triggering mechanisms affect the control or filtering performance for networked systems, and 2） developing some suitable approaches for the controller and filter design problems. In this paper, a bibliographical review is presented on event-based control and filtering problems for various networked systems. First, the event-driven communication scheme is introduced in detail according to its engineering background, characteristic, and representative research frameworks. Then, different event-based control and filtering（or state estimation） problems are categorized and then discussed. Finally, we conclude the paper by outlining future research challenges for event-based networked systems.

Freshness constraints of an age of information based event-triggered Kalman consensus filter algorithm over a wireless sensor network

This paper presents the design of a new event-triggered Kalman consensus filter(ET-KCF)algorithm for use over a wireless sensor network(WSN).This algorithm is based on information freshness,which is calculated as the age of information(Aol)of the sampled data.The proposed algorithm integrates the traditional event-triggered mechanism,information freshness calculation method,and Kalman consensus filter(KCF)algorithm to estimate the concentrations of pollutants in the aircraft more efficiently.The proposed method also considers the influence of data packet loss and the aircraft's loss of communication path over the WSN,and presents an Aol-freshness-based threshold selection method for the ET-KCF algorithm,which compares the packet Aol to the minimum average Aol of the system.This method can obviously reduce the energy consumption because the transmission of expired information is reduced.Finally,the convergence of the algorithm is proved using the Lyapunov stability theory and matrix theory.Simulation results show that this algorithm has better fault tolerance compared to the existing KCF and lower power consumption than other ET-KCFs.

Enhancement of filtration efficiency by electrical charges on nebulized particles

There have been few investigations of effects of electrical charge, carried by lab-generated particles, on filtration efficiency testing. Here, we measured the elementary charge on particles and the fraction of particles carrying that charge with a combined electrometer, differential mobility analyzer, and scanning mobility particle sizer. A typical solid NaCI aerosol and liquid diethylhexyl sebacate （DEHS） aerosol were generated with Collison and Laskin nebulizers, respectively. Our experimental results showed that NaCI aerosols carried more charge after aerosol generation. The average net elementary charge per particle was approximately 0.07. The NaCI aerosol was overall positively charged but contained a mixture of neutral and charged particles. Individual particles could carry at most four elementary charges. According to constant theorem, we speculated that original NaC1 aerosol contained 17% neutral, 45% positive-, and 38% negative-charged particles in the diameter range from 30 to 300nm. A Kr-85 neutralizer was used to decrease the charge on the NaCI particles. Our results indicated that the DEHS aerosol was electrically neutral. The effects of electric charge on particle collection by electret and electroneutral high efficiency particulate air （HEPA） filters were analyzed. Theoretical calculations suggested that charges on original NaCI aerosol particles enhanced the filtration efficiency of HEPA filters,