Gaussian Sum PHD Filtering Algorithm for Nonlinear Non-Gaussian Models

A new multi-target filtering algorithm, termed as the Gaussian sum probability hypothesis density （GSPHD） filter, is proposed for nonlinear non-Gaussian tracking models. Provided that the initial prior intensity of the states is Gaussian or can be identified as a Gaussian sum, the analytical results of the algorithm show that the posterior intensity at any subsequent time step remains a Gaussian sum under the assumption that the state noise, the measurement noise, target spawn intensity, new target birth intensity, target survival probability, and detection probability are all Gaussian sums. The analysis also shows that the existing Gaussian mixture probability hypothesis density （GMPHD） filter, which is unsuitable for handling the non-Gaussian noise cases, is no more than a special case of the proposed algorithm, which fills the shortage of incapability of treating non-Gaussian noise. The multi-target tracking simulation results verify the effectiveness of the proposed GSPHD.

Quantum stochastic filters for nonlinear time-domain filtering of communication signals

Principles and performances of quantum stochastic filters are studied for nonlinear time-domain filtering of communication signals. Filtering is realized by combining neural networks with the nonlinear Schroedinger equation and the time-variant probability density function of signals is estimated by solution of the equation. It is shown that obviously different performances can be achieved by the control of weight coefficients of potential fields. Based on this characteristic, a novel filtering algorithm is proposed, and utilizing this algorithm, the nonlinear waveform distortion of output signals and the denoising capability of the filters can be compromised. This will make the application of quantum stochastic filters be greatly extended, such as in applying the filters to the processing of communication signals. The predominant performance of quantum stochastic filters is shown by simulation results.

Eigenvalue spectrum analysis for temporal signals of Kerr optical frequency combs based on nonlinear Fourier transform

Based on the nonlinear Schr?dinger equation(NLSE) with damping, detuning, and driving terms describing the evolution of signals in a Kerr microresonator, we apply periodic nonlinear Fourier transform(NFT) to the study of signals during the generation of the Kerr optical frequency combs(OFCs). We find that the signals in different states, including the Turing pattern, the chaos, the single soliton state, and the multi-solitons state, can be distinguished according to different distributions of the eigenvalue spectrum. Specially, the eigenvalue spectrum of the single soliton pulse is composed of a pair of conjugate symmetric discrete eigenvalues and the quasi-continuous eigenvalue spectrum with eye-like structure.Moreover, we have successfully demonstrated that the number of discrete eigenvalue pairs in the eigenvalue spectrum corresponds to the number of solitons formed in a round-trip time inside the Kerr microresonator. This work shows that some characteristics of the time-domain signal can be well reflected in the nonlinear domain.

Maximum Correntropy Kalman Filtering for Non-Gaussian Systems With State Saturations and Stochastic Nonlinearities

This paper tackles the maximum correntropy Kalman filtering problem for discrete time-varying non-Gaussian systems subject to state saturations and stochastic nonlinearities. The stochastic nonlinearities, which take the form of statemultiplicative noises, are introduced in systems to describe the phenomenon of nonlinear disturbances. To resist non-Gaussian noises, we consider a new performance index called maximum correntropy criterion(MCC) which describes the similarity between two stochastic variables. To enhance the “robustness” of the kernel parameter selection on the resultant filtering performance, the Cauchy kernel function is adopted to calculate the corresponding correntropy. The goal of this paper is to design a Kalman-type filter for the underlying systems via maximizing the correntropy between the system state and its estimate. By taking advantage of an upper bound on the one-step prediction error covariance, a modified MCC-based performance index is constructed. Subsequently, with the assistance of a fixed-point theorem, the filter gain is obtained by maximizing the proposed cost function. In addition, a sufficient condition is deduced to ensure the uniqueness of the fixed point. Finally, the validity of the filtering method is tested by simulating a numerical example.

Transfer characteristics of dynamic biochemical signals in non-reversing pulsatile flows in a shallow Y-shaped microfluidic channel: signal filtering and nonlinear amplitude-frequency modulation

The transports of the dynamic biochemical signals in the non-reversing pulsatile flows in the mixing microchannel of a Y-shaped microfluidic device are ana- lyzed. The results show that the mixing micro-channel acts as a low-pass filter, and the biochemical signals are nonlinearly modulated by the pulsatile flows, which depend on the biochemical signal frequency, the flow signal frequency, and the biochemical signal transporting distance. It is concluded that, the transfer characteristics of the dynamic biochemical signals, which are transported in the time-varying flows, should be carefully considered for better loading biochemical signals on the cells cultured on the bottom of the microfluidic channel.

The retrieval of the nonlinear random waves from the non-Gaussian characteristics of the sea surface elevation distribution

In this paper, without recourse to the nonlinear dynamical equations of the waves, the nonlinear random waves are retrieved from the non-Gaussian characteristic of the sea surface elevation distribution. The question of coincidence of the nonlinear wave profile, spectrum and its distributions of maximum (or minimum) values of the sea surface elevation with results derived from some existing nonlinear theories is expounded under the narrow-band spectrum condition. Taking the shoaling sea wave as an example, the nonlinear random wave process and its spectrum in shallow water are retrieved from both the non-Gaussian characteristics of the sea surface elevation distribution in shallow water and the normal sea waves in deep water and compared with the values actually measured. Results show that they can coincide with the actually measured values quite well, thus, this can confirm that the method proposed in this paper is feasible.

Nonlinear signal transduction network with multistate

Signal transduction is an important and basic mechanism to cell life activities.The stochastic state transition of receptor induces the release of signaling molecular,which triggers the state transition of other receptors.It constructs a nonlinear sigaling network,and leads to robust switchlike properties which are critical to biological function.Network architectures and state transitions of receptor affect the performance of this biological network.In this work,we perform a study of nonlinear signaling on biological polymorphic network by analyzing network dynamics of the Ca^(2+)-induced Ca^(2+)release(CICR)mechanism,where fast and slow processes are involved and the receptor has four conformational states.Three types of networks,Erdos–R´enyi(ER)network,Watts–Strogatz(WS)network,and BaraB´asi–Albert(BA)network,are considered with different parameters.The dynamics of the biological networks exhibit different patterns at different time scales.At short time scale,the second open state is essential to reproduce the quasi-bistable regime,which emerges at a critical strength of connection for all three states involved in the fast processes and disappears at another critical point.The pattern at short time scale is not sensitive to the network architecture.At long time scale,only monostable regime is observed,and difference of network architectures affects the results more seriously.Our finding identifies features of nonlinear signaling networks with multistate that may underlie their biological function.

Estimation Performance for the Cubature Particle Filter under Nonlinear/Non-Gaussian Environments

This paper evaluates the state estimation performance for processing nonlinear/non-Gaussian systems using the cubature particle lter(CPF),which is an estimation algorithm that combines the cubature Kalman lter(CKF)and the particle lter(PF).The CPF is essentially a realization of PF where the third-degree cubature rule based on numerical integration method is adopted to approximate the proposal distribution.It is benecial where the CKF is used to generate the importance density function in the PF framework for effectively resolving the nonlinear/non-Gaussian problems.Based on the spherical-radial transformation to generate an even number of equally weighted cubature points,the CKF uses cubature points with the same weights through the spherical-radial integration rule and employs an analytical probability density function(pdf)to capture the mean and covariance of the posterior distribution using the total probability theorem and subsequently uses the measurement to update with Bayes’rule.It is capable of acquiring a maximum a posteriori probability estimate of the nonlinear system,and thus the importance density function can be used to approximate the true posterior density distribution.In Bayesian ltering,the nonlinear lter performs well when all conditional densities are assumed Gaussian.When applied to the nonlinear/non-Gaussian distribution systems,the CPF algorithm can remarkably improve the estimation accuracy as compared to the other particle lterbased approaches,such as the extended particle lter(EPF),and unscented particle lter(UPF),and also the Kalman lter(KF)-type approaches,such as the extended Kalman lter(EKF),unscented Kalman lter(UKF)and CKF.Two illustrative examples are presented showing that the CPF achieves better performance as compared to the other approaches.

Detecting the Nonlinearity of Fish Acoustic Signals

This paper discusses the nonlinearity of fish acoustic signals by using the surrogate data method. We compare the difference of three test statistics - time-irreversibility Trey, correlation dimension D2 and auto mutual information function I between the original data and the surrogate data. We come to the conclusion that there exists nonlinearity in the fish acoustic signals and there exist deterministic nonlinear components; therefore nonlinear dynamic theory can be used to analyze fish acoustic signals.

A New Processing Method for the Nonlinear Signals Produced by Electromagnetic Flowmeters in Conditions of Pipe Partial Filling

When a pipe is partially filled with a given working liquid,the relationship between the electromotive force(EMF)measured by the sensor(flowmeter)and the average velocity is nonlinear and non-monotonic.This relationship varies with the inclination of the pipe,the fluid density,the pipe wall friction coefficient,and other factors.Therefore,existing measurement methods cannot meet the accuracy requirements of many industrial applications.In this study,a new processing method is proposed by which the flow rate can be measured with an ordinary electromagnetic flowmeter even if the pipe is only partially filled.First,a B-spline curve fitting method is applied to a limited set of measurements.Second,matrix inversion required in the B-spline curve method is optimized in order to reduce the number of needed computations.Dedicated experimental tests prove that the proposed method can effectively measure the average flow velocity of the fluid.When the fluid level of the pipeline is between 50%and 100%,the relative error is less than 3.5%.

Broad-Band All-Optical Wavelength Conversion of Differential Phase-Shift Keyed Signal Using an SOA-Based Nonlinear Polarization Switch

Broad-band all-optical wavelength conversion of differential phase-shift keyed （DPSK） signal is experimentally demonstrated. This scheme is composed of a one-bit delay interferometer demodulation stage followed by a semiconductor optical amplifier （SOA） based nonlinear polarization switch. A wavelength converter for the 10 G b/s DPSK signal is presented, which has a wide wavelength range of more than 30 nm. The converted signals experience small power penalties less than 1.4 dB compared with the original signal, at a bit error rate of 10-9. Additionally, the optical spectra, the measured waveforms and the open eye diagrams of the converted signals show a high quality wavelength conversion performance.

Boundedness in a chemotaxis system with nonlinear signal production

This work deals with the zero-Neumann boundary problem to a fully parabolic chemotaxis system with a nonlinear signal production function f（s） fulfilling 0 〈 f（s） 〈 Ksα for all s 〉 0, where K and α are positive parameters. It is shown that whenever 0 〈 α 〈 2/n （where n denotes the spatial dimension） and under suitable assumptions on the initial data, this problem admits a unique global classical solution that is uniformly-in-time bounded in any spatial dimension. The proof is based on some a priori estimate techniques.

Nonlinear Liquid Crystal Optical Signal Amplifiers

The principle, structure and system of nonlinear liquid crystal optical signal amplifiers are described. Experimental results are theoretically analysed for optical signal amplifers. It shows that this type of the optical signal amplifiers are comprised of liquid crystal light sensitive medium which can receive a modulated signal optic wave and a pump wave, and can be applied to optical transmission systems.

Quaternion Approach to Solve Coupled Nonlinear Schrdinger Equation and Crosstalk of Quarter-Phase-Shift-Key Signals in Polarization Multiplexing Systems

The quaternion approach to solve the coupled nonlinear Schrodinger equations （CNSEs） in fibers is proposed, converting the CNSEs to a single variable equation by using a conception of eigen-quaternion of coupled quater- nion. The crosstalk of quarter-phase-shift-key signals caused by fiber nonlinearity in polarization multiplexing systems with 100 Cbps bit-rate is investigated and simulated. The results demonstrate that the crosstalk is like a rotated ghosting of input constellation. For the 50 km conventional fiber link, when the total power is less than 4roW, the crosstalk effect can be neglected; when the power is larger than 20roW, the crosstalk is very obvious. In addition, the crosstalk can not be detected according to the output eye diagram and state of polarization in Poincare sphere in the trunk fiber, making it difficult for the monitoring of optical trunk link.

All-Optical Sampling Using Nonlinear Polarization Rotation in a Single Semiconductor Optical Amplifier

We propose a novel all-optical sampling method using nonlinear polarization rotation in a semiconductor optical amplifier. A rate-equation model capable of describing the all-optical sampling mechanism is presented in this paper. Based on this model, we investigate the optimized operating parameters of the proposed system by simulating the output intensity of the probe light as functions of the input polarization angle, the phase induced by the polarization controller, and the ori- entation of the polarization beam splitter. The simulated results show that we can obtain a good linear slope and a large linear dynamic range,which is suitable for all-optical sampling. The operating power of the pump light can be less than lmW. The presented all-optical sampling method can potentially operate at a sampling rate up to hundreds GS/s and needs low optical power.

A Single Neuron with Stochastic Resonance for Noisy Square Pulse Train Signal Transmission

A nonlinear single neuron is demonstrated to exhibit stochastic resonance by theoretical analysis and numerical simulations. This single neuron is used for noisy periodic signal transmission, and significant performance of raising input output SNR gain can be achieved. The research of this paper not only gives a very simple model of neuron with stochastic resonance, but also enlarges the application scope of neuron to the transmission of periodic signals.

The Marginal Rao-Blackwellized Particle Filter for Mixed Linear/Nonlinear State Space Models

In this paper, the marginal Rao-Blackwellized particle filter （MRBPF）, which fuses the Rao-Blackwellized particle filter （RBPF） algorithm and the marginal particle filter （MPF） algorithm, is presented. The state space is divided into linear and non-linear parts, which can be estimated separately by the MPF and the optional Kalman filter. Through simulation in the terrain aided navigation （TAN） domain, it is demonstrated that, compared with the RBPF, the root mean square errors （RMSE） and the error variance of the nonlinear state estimations by the proposed MRBPF are respectively reduced by 29% and 96%, while the unique particle count is increased by 80%. It is also found that the MRBPF has better convergence properties, and analysis has shown that the existing RBPF is nothing more than a special case of the MRBPF.

A New Particle Filter for Nonlinear Tracking Problems

A new particle filter is presented for nonlinear tracking problems. Inpractice, maneuvering target-tracking systems are usually nonlinear and incompletely observed, andthe main difficulty of maneuvering target-tracking problem lies in the fact that the maneuverabilityat every step is of high uncertainties. Here a new smoothing particle filter algorithm is proposed,which combines the particle filter to tackle the non-linear and non-Gaussian peculiarities of theproblem, together with smoothing of the PDF of system modes and thus settles the estimate problem ofthe target maneuverability. The simulation comparison with the auxiliary particle filters showsthat the approach has superiority and yields performance improvements in solving nonlinear trackingproblems.

Effect of Multipoint Heterogeneity on Nonlinear Transformations for Geological Modeling: Porosity-Permeability Relations Revisited

An analysis of statistical expected values for transformations is performed in this study to quantify the effect of heterogeneity on spatial geological modeling and evaluations. Algebraic transformations are frequently applied to data from logging to allow for the modeling of geological properties. Transformations may be powers, products, and exponential operations which are commonly used in well-known relations （e.g., porosity-permeability transforms）. The results of this study show that correct computations must account for residual transformation terms which arise due to lack of independence among heterogeneous geological properties. In the case of an exponential porosity-permeability transform, the values may be positive. This proves that a simple exponential model back-transformed from linear regression underestimates permeability. In the case of transformations involving two or more properties, residual terms may represent the contribution of heterogeneous components which occur when properties vary together, regardless of a pair-wise linear independence. A consequence of power- and product-transform models is that regression equations within those transformations need corrections via residual cumulants. A generalization of this result is that transformations of multivariate spatial attributes require multiple-point random variable relations. This analysis provides practical solutions leading to a methodology for nonlinear modeling using correct back transformations in geology.

Adaptive Learning with Large Variability of Teaching Signals for Neural Networks and Its Application to Motion Control of an Industrial Robot

Recently, various control methods represented by proportional-integral-derivative （PID） control are used for robotic control. To cope with the requirements for high response and precision, advanced feedforward controllers such as gravity compensator, Coriolis/centrifugal force compensator and friction compensators have been built in the controller. Generally, it causes heavy computational load when calculating the compensating value within a short sampling period. In this paper, integrated recurrent neural networks are applied as a feedforward controller for PUMA560 manipulator. The feedforward controller works instead of gravity and Coriolis/centrifugal force compensators. In the learning process of the neural network by using back propagation algorithm, the learning coefficient and gain of sigmoid function are tuned intuitively and empirically according to teaching signals. The tuning is complicated because it is being conducted by trial and error. Especially, when the scale of teaching signal is large, the problem becomes crucial. To cope with the problem which concerns the learning performance, a simple and adaptive learning technique for large scale teaching signals is proposed. The learning techniques and control effectiveness are evaluated through simulations using the dynamic model of PUMA560 manipulator.