Adaptive Neural Network Approach Research Articles

An adaptive neural network approach for operator functional state prediction using psychophysiological data

In highly automated human-machine systems, human operator functional state (OFS) prediction is an important ap- proach to prevent accidents caused by operator fatigue, high mental workload, over anxiety, etc. In this paper, psychophysio- logical indices, i.e. heart rate, heart rate variability, task load index and engagement index recorded from operators who execute process control tasks are selected for OFS prediction. An adaptive differential evolution based neural network (ACADE-NN) is investigated. The behavior of ant colony foraging is introduced to self-adapt the control parameters of DE along with the mutation strategy at different evolution phases. The performance of ACADE is verified in the benchmark function tests. The designed ACADE-NN prediction model is used for estimation of the operator functional state. The empirical results illustrate that the proposed adaptive model is effective for most of the operators. The model outperforms the compared modeling methods and yields good generalization comparatively. It can describe the relationship between psychophysiological variables and OFS. It's applicable to assess the operator functional state in safety-critical applications.

An adaptive neural network approach for ship roll stabilization via fin control

Combining the adaptive backstepping technique with neural network, an adaptive neural-network-based fin control design method is proposed for the ship roll stabilization. A barrier Lyapunov function is employed to address the problem of the output constraints, and an adaption mechanism is to solve the problem of unknown parameters. The ship roll stabilization is achieved via a fin control system, where the disturbance is estimated and compensated to improve the robustness of the controller. The stability analysis shows that the proposed approach guarantees all signals in the closed-loop system to be semi-globally, uniformly and ultimately bounded (SGUUB). The simulation results verify the performance and effectiveness of the proposed approach.

Adaptive neural networks control for camera stabilization with active suspension system

The camera always suffers from image instability on the moving vehicle due to unintentional vibrations caused by road roughness. This article presents an adaptive neural network approach mixed with linear quadratic regulator control for a quarter-car active suspension system to stabilize the image captured area of the camera. An active suspension system provides extra force through the actuator which allows it to suppress vertical vibration of sprung mass. First, to deal with the road disturbance and the system uncertainties, radial basis function neural network is proposed to construct the map between the state error and the compensation component, which can correct the optimal state-feedback control law. The weights matrix of radial basis function neural network is adaptively tuned online. Then, the closed-loop stability and asymptotic convergence performance is guaranteed by Lyapunov analysis. Finally, the simulation results demonstrate that the proposed controller effectively suppresses the vibration of the camera and enhances the stabilization of the entire camera, where different excitations are considered to validate the system performance.

Adaptive Neural Network Approach for a Class of Uncertain Non-affine Nonlinear Systems

Adaptive Neural Network Approach for a Class of Uncertain Non-affine Nonlinear Systems

Adaptive Neural Network Approach for a Class of Uncertain Non-affine Nonlinear Systems

The paper proposes a new output feedback adaptive tracking control scheme using neural network for a class of uncertain non-affine nonlinear systems that only the system output variables can be measured. The scheme adopts low-pass filter to transform non-affine nonlinear systems into affine in the pseudo-input dynamics. No state observer is employed and few adapting parameters to be tuned and Lipschiz assumption, SPR condition are not required. Only the output error is used in control laws and weights update laws which make the system construct simple. Boundedness for the output tracking error and all states in the closed-loop system are guaranteed, and simulation results have verified the effectiveness of the proposed approach.

Coal Mine Gas Prediction Model Based on Particle Swarm Optimization Algorithm

For traditional methods for coal mine gas emission prediction accuracy is not high, an adaptive mutation particle swarm optimization neural network approach is introduced. The algorithm increases the mutation operation in iterative process, and adaptive adjusts mutation probability of the size, in order to enhance the ability to jump out of the local optima. The simulation results show that the method can be better predicted coal mine gas, has a certain practicality.

Distributed Spatiotemporal Neural Network for Nonlinear Dynamic Transmitter Modeling and Adaptive Digital Predistortion

This paper presents an adaptive neural network (NN) approach for the behavioral modeling of wireless transmitters exhibiting dynamic nonlinearities that are mainly caused by the power amplifier (PA). The proposed distributed spatiotemporal NN mimics the functionality of the mammal cerebellum, which is capable of very fast learning and contains features of interpolation. PAs' memory effects are modeled by using linear affine projection on a local function generated by preceding signal inputs. The applicability of the proposed model is validated in the frequency and time domains for forward and reverse modeling using a highly nonlinear Doherty amplifier and a class AB PA driven by wideband code division multiple access and WiMAX signals. The modeling performance is compared with existing techniques to establish it as a successful model that requires a relatively less demanding processing speed and memory requirement during the identification procedure. This model was found to be effective for adaptive applications such as baseband predistortion-based linearization of wireless transmitters.

Harmonics estimation using KF-Adaline algorithm and elimination with hybrid active power filter in distorted power system signals

Harmonics estimation and its elimination for a signal distorted with additive noise is an interdisciplinary area of interest for many researchers. This paper presents Kalman filter – adaptive linear neural network (KF-Adaline) approach for harmonics estimation and hybrid active power filter (HAPF) with modified PWM control technique for its elimination in distorted power system signals. In the estimation process, the weight of the Adaline is updated using Kalman filter algorithm. Harmonics components are estimated from the updated weights of the Adaline. In order to mitigate these harmonics, HAPF with modified PWM control is proposed. The modified PWM control technique is based on comparing simultaneously a triangular high frequency carrier signal with a slow varying regulation signal and it is opposite. A laboratory prototype for HAPF with modified PWM technique is built for harmonics elimination in distorted power system signals. Simulation and experimental results are presented to verify the good behav...

Adaptive neural control for strict-feedback stochastic nonlinear systems with time-delay

The problem of robust stabilization is investigated for strict-feedback stochastic nonlinear time-delay systems via adaptive neural network approach. Neural networks are used to model the unknown packaged functions, then the adaptive neural control law is constructed by a novel Lyapunov–Krasovskii functional and backstepping. It is shown that all the variables in the closed-loop system are semi-globally stochastic bounded, and the state variables converge into a small neighborhood in the sense of probability.

Advances in adaptive nonlinear manifolds and dimensionality reduction

Recent decades have witnessed a much increased demand for advanced, effective and efficient methods and tools for analyzing, understanding and dealing with data of increasingly complex, high dimensionality and large volume. Whether it is in biology, neuroscience, modern medicine and social sciences or in engineering and computer vision, data are being sampled, collected and cumulated in an unprecedented speed. It is no longer a trivial task to analyze huge amounts of high dimensional data. A systematic, automated way of interpreting data and representing them has become a great challenge facing almost all fields and research in this emerging area has flourished. Several lines of research have embarked on this timely challenge and tremendous progresses and advances have been made recently. Traditional and linear methods are being extended or enhanced in order to meet the new challenges. This paper elaborates on these recent advances and discusses various state-of-the-art algorithms proposed from statistics, geometry and adaptive neural networks. The developments mainly follow three lines: multidimensional scaling, eigen-decomposition as well as principal manifolds. Neural approaches and adaptive or incremental methods are also reviewed. In the first line, traditional multidimensional scaling (MDS) has been extended not only to be more adaptive such as neural scale, curvilinear component analysis (CCA) and visualization induced self-organizing map (ViSOM) for online learning, but also to be more local scaling such as Isomap for enhanced flexibility for nonlinear data sets. The second line extends linear principal component analysis (PCA) and has attracted a huge amount of interest and enjoyed flourishing advances with methods like kernel PCA (KPCA), locally linear embedding (LLE) and Laplacian eigenmap. The advantage is obvious: a nonlinear problem is transformed into a linear one and a unique solution can then be sought. The third line starts with the nonlinear principal curve and surface and links up with adaptive neural network approaches such as self-organizing map (SOM) and ViSOM. Many of these frameworks have been further improved and enhanced for incremental learning and mapping function generalization. This paper discusses these recent advances and their connections. Their application issues and implementation matters will also be briefly enlightened and commented on.

Online identification of nonlinear multivariable processes using self-generating RBF neural networks

This paper addresses the problem of online model identification for multivariable processes with nonlinear and time-varying dynamic characteristics. For this purpose, two online multivariable identification approaches with self-organizing neural network model structures will be presented. The two adaptive radial basis function (RBF) neural networks are called as the growing and pruning radial basis function (GAP-RBF) and minimal resource allocation network (MRAN). The resulting identification algorithms start without a predefined model structure and the dynamic model is generated autonomously using the sequential input-output data pairs in real-time applications. The extended Kalman filter (EKF) learning algorithm has been extended for both of the adaptive RBF-based neural network approaches to estimate the free parameters of the identified multivariable model. The unscented Kalman filter (UKF) has been proposed as an alternative learning algorithm to enhance the accuracy and robustness of nonlinear multivariable processes in both the GAP-RBF and MRAN based approaches. In addition, this paper intends to study comparatively the general applicability of the particle filter (PF)-based approaches for the case of non-Gaussian noisy environments. For this purpose, the Unscented Particle Filter (UPF) is employed to be used as alternative to the EKF and UKF for online parameter estimation of self-generating RBF neural networks. The performance of the proposed online identification approaches is evaluated on a highly nonlinear time-varying multivariable non-isothermal continuous stirred tank reactor (CSTR) benchmark problem. Simulation results demonstrate the good performances of all identification approaches, especially the GAP-RBF approach incorporated with the UKF and UPF learning algorithms. Copyright © 2010 John Wiley and Sons Asia Pte Ltd and Chinese Automatic Control Society

Fuzzy adaptive neural network approach to path loss prediction in urban areas at GSM-900 band

This paper presents the results of the Adaptive-Network Based Fuzzy Inference System (ANFIS) for the prediction of path loss in a specific urban environment. A new algorithm based ANFIS for tuning the path loss model is introduced in this work. The performance of the path loss model which is obtained from proposed algorithm is compared to the Bertoni-Walfisch model, which is one of the best studied for propagation analysis involving buildings. This comparison is based on the mean square error between predicted and measured values. According to the indicated error criterion, the errors related to the predictions that are obtained from the algorithm are less than the errors that are obtained from the Bertoni-Walfisch Model. In this study, propagation measurements were carried out in the 900 MHz band in the city of Istanbul, Turkey.

Suspended sediment concentration estimation by an adaptive neuro-fuzzy and neural network approaches using hydro-meteorological data

Correct estimation of sediment volume carried by a river is very important for many water resources projects. Conventional sediment rating curves, however, are not able to provide sufficiently accurate results. In this paper, an adaptive neuro-fuzzy approach is proposed to estimate suspended sediment concentration on rivers. The daily rainfall, streamflow and suspended sediment concentration data from Mad River Catchment near Arcata, USA are used as a case study. In the first part of the study, various combinations of current daily rainfall, streamflow and past daily streamflow, suspended sediment data are used as inputs to the neuro-fuzzy computing technique so as to estimate current suspended sediment. In the second part of the study, the potential of neuro-fuzzy technique is compared with those of the three different artificial neural networks (ANN) techniques, namely, the generalized regression neural networks (GRNN), radial basis neural networks (RBNN) and multi-layer perceptron (MLP) and two different sediment rating curves (SRC). The comparison results reveal that the neuro-fuzzy models perform better than the other models in daily suspended sediment concentration estimation for the particular data sets used in this study.

An adaptive PNN-DS approach to classification using multi-sensor information fusion

In this paper, an adaptive neural network approach to classification which combines modified probabilistic neural network and D-S evidence theory (PNN-DS) is proposed. It attempts to deal with the ...

Chaos synchronization of coupled neurons under external electrical stimulation using adaptive H∞ control

In this paper, H∞ control via an adaptive neural network approach is proposed to realize the synchronization of two gap junction coupled chaotic FitzHugh—Nagumo (FHN) neurons under external electrical stimulation. The control scheme is robust to the uncertainties such as unmodelled dynamics, ionic channel noise and external disturbances. Asymptotic chaos synchronization is obtained by proper choice of the control parameters. The simulation results demonstrate the effectiveness of the proposed control method.

Mapping polynomial fitting into feedforward neural networks for modeling nonlinear dynamic systems and beyond

This study presents an explicit demonstration on constructing a multilayer feedforward neural network to approximate polynomials and conduct polynomial fitting. Built on an algebraic analysis of sigmoidal activation functions rather than incremental training, this work reveals the capability of the “universal approximator” by relating the “soft computing tool” to an important class of conventional computing tools widely used in modeling nonlinear dynamic systems and many other scientific computing applications. The authors strive to enable physical interpretations and afford full control when applying the highly adaptive, powerful yet subjective neural network approach. This work is a part of the effort of bridging the gap between the black-box and mechanics-based parametric modeling.

A new adaptive neural network and heuristics hybrid approach for job-shop scheduling

A new adaptive neural network and heuristics hybrid approach for job-shop scheduling is presented. The neural network has the property of adapting its connection weights and biases of neural units while solving the feasible solution. Two heuristics are presented, which can be combined with the neural network. One heuristic is used to accelerate the solving process of the neural network and guarantee its convergence, the other heuristic is used to obtain non-delay schedules from the feasible solutions gained by the neural network. Computer simulations have shown that the proposed hybrid approach is of high speed and efficiency. The strategy for solving practical job-shop scheduling problems is provided. Scope and purpose Job-shop scheduling is usually a strongly NP-complete problem of combinatorial optimization problems and is the most typical one of the production scheduling problems. It is usually very hard to find its optimal solution. Practically researchers turn to search its near-optimal solutions with all kind of heuristic algorithms. The scope of this paper is to present a new hybrid approach in dealing with this job-shop scheduling problem based on adaptive neural network and heuristics.

Constraint satisfaction adaptive neural network and heuristics combined approaches for generalized job-shop scheduling

This paper presents a constraint satisfaction adaptive neural network, together with several heuristics, to solve the generalized job-shop scheduling problem, one of NP-complete constraint satisfaction problems. The proposed neural network can be easily constructed and can adaptively adjust its weights of connections and biases of units based on the sequence and resource constraints of the job-shop scheduling problem during its processing. Several heuristics that can be combined with the neural network are also presented. In the combined approaches, the neural network is used to obtain feasible solutions, the heuristic algorithms are used to improve the performance of the neural network and the quality of the obtained solutions. Simulations have shown that the proposed neural network and its combined approaches are efficient with respect to the quality of solutions and the solving speed.

Constraint satisfaction adaptive neural network and efficient heuristics for job-shop scheduling

An efficient constraint satisfaction based adaptive neural network and heuristics hybrid approach for job-shop scheduling is presented. The adaptive neural network has the property of adaptively adjusting its connection weights and biases of neural units according to the sequence and resource constraints of job-shop scheduling problem while solving feasible solution. Two heuristics are used in the hybrid approach: one is used to accelerate the solving process of neural network and guarantee its convergence, the other is used to obtain non-delay schedule from solved feasible solution by neural network. Computer simulations have shown that the proposed hybrid approach is of high speed and excellent efficiency.

Competition-based neural network for the multiple travelling salesmen problem with minmax objective

In this paper, a new algorithm in competition-based network has been introduced to solve the minmax multiple travelling salesmen problem (MTSP) which needs the maximum distance among all salesmen to be minimum. As in the previous approaches, the generalized 2-opt exchange heuristic algorithms and the elastic net algorithm are reviewed and applied to the minmax MTSP problem solution. Furthermore, a comprehensive empirical study has been provided in order to investigate the performance of the algorithms. The adaptive approach can obtain the superior solution in all instances, compared to the generalized 2-opt exchange heuristic and the elastic net. In additional evaluation, the adaptive algorithm is combined with a simple improvement heuristic and compared with a recently adaptive tabu search. As a result, the adaptive approach can obtain the appropriate solutions 3% in average of the best solution of the adaptive tabu search heuristic accompanied with the higher speed of 31% in average. Scope and purpose Recently, the neural networks have been one of much interest to OR community; in particular to those dealing with the combinatorial optimization problem such as travelling salesman problem (TSP). Nevertheless, the neural approaches are difficult to apply to more complex problems as additional constraints must be satisfactorily considered. In addition, the comparison between the heuristic approaches and the neural approaches has never been reported before. In this paper, an adaptive neural network approach has been proposed for solving the minmax multiple travelling salesmen problem, which is an extension of the TSP, and it is compares with previous heuristics: elastic net, generalized 2-opt exchange heuristic and adaptive tabu search heuristic.