Tracking moving wideband sound sources is one of the most challenging issues in the acoustic array signal processing which is based on the direction of arrival(DOA) estimation. Compressive sensing(CS) is a recent theo...Tracking moving wideband sound sources is one of the most challenging issues in the acoustic array signal processing which is based on the direction of arrival(DOA) estimation. Compressive sensing(CS) is a recent theory exploring the signal sparsity representation, which has been proved to be superior for the DOA estimation. However, the spatial aliasing and the offset at endfire are the main obstacles for CS applied in the wideband DOA estimation. We propose a particle filter based compressive sensing method for tracking moving wideband sound sources. First, the initial DOA estimates are obtained by wideband CS algorithms. Then, the real sources are approximated by a set of particles with different weights assigned. The kernel density estimator is used as the likelihood function of particle filter. We present the results for both uniform and random linear array. Simulation results show that the spatial aliasing is disappeared and the offset at endfire is reduced. We show that the proposed method can achieve satisfactory tracking performance regardless of using uniform or random linear array.展开更多
In order to improve the performance of the probability hypothesis density(PHD) algorithm based particle filter(PF) in terms of number estimation and states extraction of multiple targets, a new probability hypothesis ...In order to improve the performance of the probability hypothesis density(PHD) algorithm based particle filter(PF) in terms of number estimation and states extraction of multiple targets, a new probability hypothesis density filter algorithm based on marginalized particle and kernel density estimation is proposed, which utilizes the idea of marginalized particle filter to enhance the estimating performance of the PHD. The state variables are decomposed into linear and non-linear parts. The particle filter is adopted to predict and estimate the nonlinear states of multi-target after dimensionality reduction, while the Kalman filter is applied to estimate the linear parts under linear Gaussian condition. Embedding the information of the linear states into the estimated nonlinear states helps to reduce the estimating variance and improve the accuracy of target number estimation. The meanshift kernel density estimation, being of the inherent nature of searching peak value via an adaptive gradient ascent iteration, is introduced to cluster particles and extract target states, which is independent of the target number and can converge to the local peak position of the PHD distribution while avoiding the errors due to the inaccuracy in modeling and parameters estimation. Experiments show that the proposed algorithm can obtain higher tracking accuracy when using fewer sampling particles and is of lower computational complexity compared with the PF-PHD.展开更多
The choice of the particle's distribution model and the consistency of the result are very important for FastSLAM.The improved auxiliary variable model with FastSLAM,and Stirling Interpolation which is used to app...The choice of the particle's distribution model and the consistency of the result are very important for FastSLAM.The improved auxiliary variable model with FastSLAM,and Stirling Interpolation which is used to approximate the nonlinear functions are provided.This approach improves the precision of the approximation for the nonlinear functions,conquers the drawback of the FastSLAM1.0 by using a model ignoring the measurement data,enhances the estimation consistency of the robot pose,and reduces the degradation speed of the particle in FastSLAM algorithm.Simulation results demonstrate the excellence of the proposed algorithm and give the noise parameter influence on the proposed algorithm.展开更多
Flow against pipeline leakage and the pipe network sudden burst pipe to pipeline leakage flow for the application objects,network congestion avoidance strategy is designed in pipeline leak monitoring.Based on the prop...Flow against pipeline leakage and the pipe network sudden burst pipe to pipeline leakage flow for the application objects,network congestion avoidance strategy is designed in pipeline leak monitoring.Based on the property of Markov chain for network data,a new estimator with particle filter is proposed for congestion control in this paper.The proposed scheme can predict the traffic patterns by the decision-making model.To compare with previous scheme based on fuzzy neural network,the proposed scheme can adaptively adjust the network rate in real-time and reduce the cell loss rate,so that it can avoid the traffic congestion.Simulation results show that network congestion avoidance strategy with particle filter can improve the bandwidth utilization,Transmission Control Protocol (TCP) friendliness and reduce the packet drop rate in Pipeline Flux Leak Monitoring networks.展开更多
A hierarchical wireless sensor networks(WSN) was proposed to estimate the plume source location.Such WSN can be of tremendous help to emergency personnel trying to protect people from terrorist attacks or responding t...A hierarchical wireless sensor networks(WSN) was proposed to estimate the plume source location.Such WSN can be of tremendous help to emergency personnel trying to protect people from terrorist attacks or responding to an accident.The entire surveillant field is divided into several small sub-regions.In each sub-region,the localization algorithm based on the improved particle filter(IPF) was performed to estimate the location.Some improved methods such as weighted centroid,residual resampling were introduced to the IPF algorithm to increase the localization performance.This distributed estimation method eliminates many drawbacks inherent with the traditional centralized optimization method.Simulation results show that localization algorithm is efficient for estimating the plume source location.展开更多
In this paper, an evolutionary recursive Bayesian estimation algorithm is presented, which incorporates the latest observation with a new proposal distribution, and the posterior state density is represented by a Gaus...In this paper, an evolutionary recursive Bayesian estimation algorithm is presented, which incorporates the latest observation with a new proposal distribution, and the posterior state density is represented by a Gaussian mixture model that is recovered from the weighted particle set of the measurement update step by means of a weighted expectation-maximization algorithm. This step replaces the resampling stage needed by most particle filters and relieves the effect caused by sample impoverishment. A nonlinear tracking problem shows that this new approach outperforms other related particle filters.展开更多
PM10 samples of 62 Quartz filters and 75 PTFE filters from 28 Jul. 1999 to 10 Jan. 2000 in London were collected and analyzed. The difference between Quartz filter and PTFE filter in monitoring PM10 and anion ion conc...PM10 samples of 62 Quartz filters and 75 PTFE filters from 28 Jul. 1999 to 10 Jan. 2000 in London were collected and analyzed. The difference between Quartz filter and PTFE filter in monitoring PM10 and anion ion concentration has been studied. The mean PM10, SO42- concentrations of Quartz filters were higher than those of PTFT filters, which had statistically significant （P〈0.05）. The mean PM 10 concentration of Quartz filter was almost 1.5 times of that of PTFE filter. However, there were no statistically significant among CI, NO3-, PO43 comparison of the two kinds filters （P〉0.05）. We should be careful when selecting filter to do research about PM10 and anion ion concentration.展开更多
基金supported by the NFSC Grants 51375385 and 51675425Natural Science Basic Research Plan in Shaanxi Province of China Grants 2016JZ013
文摘Tracking moving wideband sound sources is one of the most challenging issues in the acoustic array signal processing which is based on the direction of arrival(DOA) estimation. Compressive sensing(CS) is a recent theory exploring the signal sparsity representation, which has been proved to be superior for the DOA estimation. However, the spatial aliasing and the offset at endfire are the main obstacles for CS applied in the wideband DOA estimation. We propose a particle filter based compressive sensing method for tracking moving wideband sound sources. First, the initial DOA estimates are obtained by wideband CS algorithms. Then, the real sources are approximated by a set of particles with different weights assigned. The kernel density estimator is used as the likelihood function of particle filter. We present the results for both uniform and random linear array. Simulation results show that the spatial aliasing is disappeared and the offset at endfire is reduced. We show that the proposed method can achieve satisfactory tracking performance regardless of using uniform or random linear array.
基金Project(61101185) supported by the National Natural Science Foundation of ChinaProject(2011AA1221) supported by the National High Technology Research and Development Program of China
文摘In order to improve the performance of the probability hypothesis density(PHD) algorithm based particle filter(PF) in terms of number estimation and states extraction of multiple targets, a new probability hypothesis density filter algorithm based on marginalized particle and kernel density estimation is proposed, which utilizes the idea of marginalized particle filter to enhance the estimating performance of the PHD. The state variables are decomposed into linear and non-linear parts. The particle filter is adopted to predict and estimate the nonlinear states of multi-target after dimensionality reduction, while the Kalman filter is applied to estimate the linear parts under linear Gaussian condition. Embedding the information of the linear states into the estimated nonlinear states helps to reduce the estimating variance and improve the accuracy of target number estimation. The meanshift kernel density estimation, being of the inherent nature of searching peak value via an adaptive gradient ascent iteration, is introduced to cluster particles and extract target states, which is independent of the target number and can converge to the local peak position of the PHD distribution while avoiding the errors due to the inaccuracy in modeling and parameters estimation. Experiments show that the proposed algorithm can obtain higher tracking accuracy when using fewer sampling particles and is of lower computational complexity compared with the PF-PHD.
基金National High-Tech Research and Development Program of China(No.2003AA1Z2130)Science and Technology Project of Zhejiang Province,China(No.2005C11001-02)
文摘The choice of the particle's distribution model and the consistency of the result are very important for FastSLAM.The improved auxiliary variable model with FastSLAM,and Stirling Interpolation which is used to approximate the nonlinear functions are provided.This approach improves the precision of the approximation for the nonlinear functions,conquers the drawback of the FastSLAM1.0 by using a model ignoring the measurement data,enhances the estimation consistency of the robot pose,and reduces the degradation speed of the particle in FastSLAM algorithm.Simulation results demonstrate the excellence of the proposed algorithm and give the noise parameter influence on the proposed algorithm.
基金Xinjiang Production and Construction Corps Reforms Project Courses,China(No.200905)Tarim University Principal Youth Fund, China(No.TDZKQN05002)Tarim University Quality of Higher Education Courses and Research Funding,China(No.TDGJ0914)
文摘Flow against pipeline leakage and the pipe network sudden burst pipe to pipeline leakage flow for the application objects,network congestion avoidance strategy is designed in pipeline leak monitoring.Based on the property of Markov chain for network data,a new estimator with particle filter is proposed for congestion control in this paper.The proposed scheme can predict the traffic patterns by the decision-making model.To compare with previous scheme based on fuzzy neural network,the proposed scheme can adaptively adjust the network rate in real-time and reduce the cell loss rate,so that it can avoid the traffic congestion.Simulation results show that network congestion avoidance strategy with particle filter can improve the bandwidth utilization,Transmission Control Protocol (TCP) friendliness and reduce the packet drop rate in Pipeline Flux Leak Monitoring networks.
基金National High Technology Research and Development Program of China(863Program,No.2004AA412050)
文摘A hierarchical wireless sensor networks(WSN) was proposed to estimate the plume source location.Such WSN can be of tremendous help to emergency personnel trying to protect people from terrorist attacks or responding to an accident.The entire surveillant field is divided into several small sub-regions.In each sub-region,the localization algorithm based on the improved particle filter(IPF) was performed to estimate the location.Some improved methods such as weighted centroid,residual resampling were introduced to the IPF algorithm to increase the localization performance.This distributed estimation method eliminates many drawbacks inherent with the traditional centralized optimization method.Simulation results show that localization algorithm is efficient for estimating the plume source location.
基金Sponsored by the National Security Major Basic Research Project of China(Grant No.973 -61334)
文摘In this paper, an evolutionary recursive Bayesian estimation algorithm is presented, which incorporates the latest observation with a new proposal distribution, and the posterior state density is represented by a Gaussian mixture model that is recovered from the weighted particle set of the measurement update step by means of a weighted expectation-maximization algorithm. This step replaces the resampling stage needed by most particle filters and relieves the effect caused by sample impoverishment. A nonlinear tracking problem shows that this new approach outperforms other related particle filters.
文摘PM10 samples of 62 Quartz filters and 75 PTFE filters from 28 Jul. 1999 to 10 Jan. 2000 in London were collected and analyzed. The difference between Quartz filter and PTFE filter in monitoring PM10 and anion ion concentration has been studied. The mean PM10, SO42- concentrations of Quartz filters were higher than those of PTFT filters, which had statistically significant （P〈0.05）. The mean PM 10 concentration of Quartz filter was almost 1.5 times of that of PTFE filter. However, there were no statistically significant among CI, NO3-, PO43 comparison of the two kinds filters （P〉0.05）. We should be careful when selecting filter to do research about PM10 and anion ion concentration.
