Real-time Recurrent Learning Algorithm Research Articles

Nonlinear autoregressive spline neural filter and its application

A novel recurrent model called nonlinear autoregressive spline neural filter (NARSNF) is introduced in this paper. Due to its unique spline nonlinear output autoregressive structure, NARSNF exhibits better fitting ability than IIR-spline adaptive filter (IIR-SAF) in identifying nonlinear dynamic systems with output autoregression. Correspondingly, the real-time recurrent learning algorithm based on NARSNF is derived and the Lyapunov stability analysis of learning rates is also provided. Moreover, the NARSNF and bilinear NARSNF (BNARSNF) are proposed to handle compensation of acoustic feedback for nonlinear active noise control (NANC). At last, simulation results in the context of nonlinear dynamic system identification and NANC verify the effectiveness of the proposed models.

A recurrent neural network for variable admittance control in human–robot cooperation: simultaneously and online adjustment of the virtual damping and Inertia parameters

In this manuscript, a recurrent neural network is proposed for variable admittance control in human–robot cooperation tasks. The virtual damping and the virtual inertia of the designed robot’s admittance controller are adjusted online and simultaneously. A Jordan recurrent neural network is designed and trained for this purpose. The network is indirectly trained using the real-time recurrent learning algorithm and based on the velocity error between the reference velocity of the minimum jerk trajectory model and the actual velocity of the robot. The performance of the proposed variable admittance controller is presented in terms of the human required effort, the task completion time, the achieved accuracy at the target, and the oscillations during the movement. Its generalization ability is evaluated experimentally by conducting cooperative tasks along numerous straight-line segments using the KUKA LWR robot and by ten subjects. Finally, a comparison with previous developed variable admittance controllers, where only the variable damping or only the virtual inertia is adjusted, is presented.

Recurrent Neural Network Based MPPT Control of Grid Connected DFIG for Wind Turbine

This paper presents a new maximum-power-point-tracking (MPPT) controller in wind power generation using artificial neural networks (ANN) in order for making the wind turbine function in optimum working point and get high efficiency of wind energy conversion at different conditions. The algorithm uses fully connected recurrent neural network and is trained online using real-time recurrent learning (RTRL) algorithm in order to avoid the oscillation problem in wind-turbine generation systems. It generates control command for speed of the rotor side converter using optimal algorithm to enable the control system in order to track the maximum power point. The rotor speed and wind-turbine torque are the inputs of the networks, and the command signal for the rotor speed of wind turbine is the output. Simulation results verify the performance of the proposed algorithm.

Complex-Valued Pipelined Chebyshev Functional Link Recurrent Neural Network for Joint Compensation of Wideband Transmitter Distortions and Impairments

In this article, a novel digital predistortion (DPD) model based on complex-valued pipelined Chebyshev functional link recurrent neural network (CPCFLRNN) for joint compensation of wideband transmitter distortions and impairments is proposed. The functional link neural network (FLNN) model has attracted much attention from scholars, and many improved models using this structure, such as Chebyshev FLNN, have been applied in the DPD of power amplifiers (PAs). However, these existing neural network models cannot deal with complex-valued input signals simultaneously, and the real-valued model structure will introduce cumbersome training algorithm and result in a long training time. The pipelined recurrent neural network (PRNN) has been successfully applied to nonlinear signal prediction because of its excellent ability for dealing with nonlinear nonstationary signals. Therefore, the PRNN model containing Chebyshev structure is extended to complex domain for the first time to obtain the CPCFLRNN model for DPD application. Considering the strong correlation of in-phase and quadrature phase (I/Q) components of the transmitter signal, the real time recurrent learning (RTRL) algorithm based on fully complex activation function is selected and extended to complex domain to obtain the complex-valued RTRL (CRTRL) algorithm for CPCFLRNN model training. A GaN PA was employed to verify the effectiveness of the proposed models. And the input signal is a 30MHz LTE signals which consists of I/Q imbalance and dc offsets. The experimental results show that the proposed CPCFLRNN model have more accurate modeling effect and better linearization performance compared with the conventional DPD models.

A Hybrid Model for Real-Time Probabilistic Flood Forecasting Using Elman Neural Network with Heterogeneity of Error Distributions

Traditional static neural networks often fail to describe dynamic flood processes, while recurrent neural networks can reflect this dynamic feature of flooding. In this paper, a real-time framework for probabilistic flood forecasting using an Elman neural network is presented. Based on this framework, flood forecasting models with different lead times are developed and trained by a real-time recurrent learning algorithm for forecasting the inflow of the Xianghongdian reservoir of the Huai River in East China. The performances of these models are evaluated. The forecasting model having a 3 h lead time meets the precision requirements and is chosen as the deterministic flood forecasting model. Compared with the multilayer perceptron having a 3 h lead time, the relative error of flood volume is 5.28% less, and the coefficient of efficiency is 0.105 greater. We further analyze the error characteristics of the selected model and derive the discharge probability density function based on the heterogeneity of error distributions. The forecasted discharge intervals with different confidence levels, the expected values, and the median values are obtained. The results show that the average relative errors of flood volume and peak discharge obtained by the median value forecasting are −1.66% and 5.69% respectively, and the coefficient of efficiency is 0.784. The performance of the median value forecasting was slightly better than that of the deterministic forecasting, and considerably better than that of the expected value forecasting. This study demonstrates that the proposed model has high practicability and can provide decision support for flood control.

Training Performance of Recurrent Neural Network using RTRL and BPTT for Gamelan Onset Detection

Gamelan is one of Indonesia’s traditional musical instruments. Signal variations in gamelan music are caused by differences in play style and the process of making gamelan. Gamelan music analysis usually using supervised learning method like Recurrent Neural Network (RNN). This paper will compare the performance of Simple Recurrent Neural Network training process using a gradient-based algorithm Backpropagation Through Time (BPTT) and Real-Time Recurrent Learning (RTRL) algorithm. The performance of the algorithm during training process was necessary to be evaluated, in order to know which algorithm has better performance and faster process to approach convergences on the training method of the recurrent neural network. The performance results of the algorithm training process will be compared and evaluated by the means of a Normalized Negative Log-likelihood (NNL). BPTT resulted better and faster forming convergence in terms of the number of epoch parameter with NNL 0.0121 In terms of the value of learning rate, BPTT perform better at learning rate 0.1 with NNL 0.0174 and RTRL performs better at learning rate 0.4 with NNL 0.0382.

Hand Recognition Obtained by Simulation of Hand Regard.

Eye-hand coordination of an infant is observed during the early months of their development. Hand regard, which is an example of this coordination, occurs at about 2 months. It is considered that after experiencing hand regard, an infant may recognize their own hands. However, it is unknown how an infant recognizes their hands through hand regard. Accordingly, the process by which an infant recognizes their hands and distinguishes between their hands and other objects was simulated. A simple neural network was trained with a modified real-time recurrent learning (RTRL) algorithm to deal with time-varying input and output during hand regard. The simulation results show that information about recognition of the modeled hands of an infant is stored in cell assemblies, which were self-organized. Cell assemblies appear during the phase of U-shaped developments of hand regard, and the configuration of the cell assemblies changes with each U-shaped development. Furthermore, movements like general movements (GMs) appear during the phase of U-shaped developments of hand regard.

Study of Closed-Loop Model Reference Adaptive Control of Smart MicroGrid with QNU and Recurrent Learning

An adaptive quadratic polynomial neural unit (QNU) controller for optimization of a conventional Smart Microgrid control loop is studied and proposed. The parameters associated with the studied grid plants are considered to be known in this study, with the fact that the load is unknown and time-variant. A sample-by-sample real-time recurrent learning algorithm of an additional QNU controller is derived, with its performance tested and discussed as a result of this paper.

Adaptive Flutter Suppression for a Fighter Wing via Recurrent Neural Networks over a Wide Transonic Range

The paper presents a digital adaptive controller of recurrent neural networks for the active flutter suppression of a wing structure over a wide transonic range. The basic idea behind the controller is as follows. At first, the parameters of recurrent neural networks, such as the number of neurons and the learning rate, are initially determined so as to suppress the flutter under a specific flight condition in the transonic regime. Then, the controller automatically adjusts itself for a new flight condition by updating the synaptic weights of networks online via the real-time recurrent learning algorithm. Hence, the controller is able to suppress the aeroelastic instability of the wing structure over a range of flight conditions in the transonic regime. To demonstrate the effectiveness and robustness of the controller, the aeroservoelastic model of a typical fighter wing with a tip missile was established and a single-input/single-output controller was synthesized. Numerical simulations of the open/closed-loop aeroservoelastic simulations were made to demonstrate the efficacy of the adaptive controller with respect to the change of flight parameters in the transonic regime.

Neural network decoupling technique and its application to a powered wheelchair system.

This paper proposes a neural network decoupling technique for an uncertain multivariable system. Based on a linear diagonalization technique, a reference model is designed using nominal parameters to provide training signals for a neural network decoupler. A neural network model is designed to learn the dynamics of the uncertain multivariable system in order to avoid required calculations of the plant Jacobian. To avoid overfitting problem, both neural networks are trained by the Lavenberg-Marquardt with Bayesian regulation algorithm that uses a real-time recurrent learning algorithm to obtain gradient information. Three experimental results in the powered wheelchair control application confirm that the proposed technique effectively minimises the coupling effects caused by input-output interactions even under the condition of system uncertainties.

Learning long-term dependencies in segmented-memory recurrent neural networks with backpropagation of error

In general, recurrent neural networks have difficulties in learning long-term dependencies. The segmented-memory recurrent neural network (SMRNN) architecture together with the extended real-time recurrent learning (eRTRL) algorithm was proposed to circumvent this problem. Due to its computational complexity eRTRL becomes impractical with increasing network size. Therefore, we introduce the less complex extended backpropagation through time (eBPTT) for SMRNN together with a layer-local unsupervised pre-training procedure. A comparison on the information latching problem showed that eRTRL is better able to handle the latching of information over longer periods of time, even though eBPTT guaranteed a better generalisation when training was successful. Further, pre-training significantly improved the ability to learn long-term dependencies with eBPTT. Therefore, the proposed eBPTT algorithm is suited for tasks that require big networks where eRTRL is impractical. The pre-training procedure itself is independent of the supervised learning algorithm and can improve learning in SMRNN in general.

Improved Generalization in Recurrent Neural Networks Using the Tangent Plane Algorithm

The tangent plane algorithm for real time recurrent learning (TPA-RTRL) is an effective online training method for fully recurrent neural networks. TPA-RTRL uses the method of approaching tangent planes to accelerate the learning processes. Compared to the original gradient descent real time recurrent learning algorithm (GD-RTRL) it is very fast and avoids problems like local minima of the search space. However, the TPA-RTRL algorithm actively encourages the formation of large weight values that can be harmful to generalization. This paper presents a new TPA-RTRL variant that encourages small weight values to decay to zero by using a weight elimination procedure built into the geometry of the algorithm. Experimental results show that the new algorithm gives good generalization over a range of network sizes whilst retaining the fast convergence speed of the TPA-RTRL algorithm.

Reinforced recurrent neural networks for multi-step-ahead flood forecasts

Summary Considering true values cannot be available at every time step in an online learning algorithm for multi-step-ahead (MSA) forecasts, a MSA reinforced real-time recurrent learning algorithm for recurrent neural networks (R-RTRL NN) is proposed. The main merit of the proposed method is to repeatedly adjust model parameters with the current information including the latest observed values and model’s outputs to enhance the reliability and the forecast accuracy of the proposed method. The sequential formulation of the R-RTRL NN is derived. To demonstrate its reliability and effectiveness, the proposed R-RTRL NN is implemented to make 2-, 4- and 6-step-ahead forecasts in a famous benchmark chaotic time series and a reservoir flood inflow series in North Taiwan. For comparison purpose, three comparative neural networks (two dynamic and one static neural networks) were performed. Numerical and experimental results indicate that the R-RTRL NN not only achieves superior performance to comparative networks but significantly improves the precision of MSA forecasts for both chaotic time series and reservoir inflow case during typhoon events with effective mitigation in the time-lag problem.

Biological modeling of complex chemotaxis behaviors for C. elegans under speed regulation—a dynamic neural networks approach

In this paper, the modeling of several complex chemotaxis behaviors of C. elegans is explored, which include food attraction, toxin avoidance, and locomotion speed regulation. We first model the chemotaxis behaviors of food attraction and toxin avoidance separately. Then, an integrated chemotaxis behavioral model is proposed, which performs the two chemotaxis behaviors simultaneously. The novelty and the uniqueness of the proposed chemotaxis behavioral models are characterized by several attributes. First, all the chemotaxis behavioral model sare on biological basis, namely, the proposed chemotaxis behavior models are constructed by extracting the neural wire diagram from sensory neurons to motor neurons, where sensory neurons are specific for chemotaxis behaviors. Second, the chemotaxis behavioral models are able to perform turning and speed regulation. Third, chemotaxis behaviors are characterized by a set of switching logic functions that decide the orientation and speed. All models are implemented using dynamic neural networks (DNN) and trained using the real time recurrent learning (RTRL) algorithm. By incorporating a speed regulation mechanism, C. elegans can stop spontaneously when approaching food source or leaving away from toxin. The testing results and the comparison with experiment results verify that the proposed chemotaxis behavioral models can well mimic the chemotaxis behaviors of C. elegans in different environments.

Reinforced two-step-ahead weight adjustment technique for online training of recurrent neural networks.

A reliable forecast of future events possesses great value. The main purpose of this paper is to propose an innovative learning technique for reinforcing the accuracy of two-step-ahead (2SA) forecasts. The real-time recurrent learning (RTRL) algorithm for recurrent neural networks (RNNs) can effectively model the dynamics of complex processes and has been used successfully in one-step-ahead forecasts for various time series. A reinforced RTRL algorithm for 2SA forecasts using RNNs is proposed in this paper, and its performance is investigated by two famous benchmark time series and a streamflow during flood events in Taiwan. Results demonstrate that the proposed reinforced 2SA RTRL algorithm for RNNs can adequately forecast the benchmark (theoretical) time series, significantly improve the accuracy of flood forecasts, and effectively reduce time-lag effects.

Learning in fully recurrent neural networks by approaching tangent planes to constraint surfaces

In this paper we present a new variant of the online real time recurrent learning algorithm proposed by Williams and Zipser (1989). Whilst the original algorithm utilises gradient information to guide the search towards the minimum training error, it is very slow in most applications and often gets stuck in local minima of the search space. It is also sensitive to the choice of learning rate and requires careful tuning. The new variant adjusts weights by moving to the tangent planes to constraint surfaces. It is simple to implement and requires no parameters to be set manually. Experimental results show that this new algorithm gives significantly faster convergence whilst avoiding problems like local minima.

Complex-valued pipelined decision feedback recurrent neural network for non-linear channel equalisation

A novel complex-valued non-linear equaliser-based pipelined decision feedback recurrent neural network (CPDFRNN) is proposed in this study for non-linear channel equalisation in wireless communication systems. The CPDFRNN with low computational complexity, a modular structure comprising a number of modules that are interconnected in a chained form, is an extension of the recently proposed real-valued pipelined decision feedback recurrent neural equalisers. Each module is implemented by a small-scale complex-valued decision feedback recurrent neural network (CDFRNN). Moreover, a decision feedback part in each module can overcome the unstable characteristic of the complex-valued recurrent neural network (CRNN). To suit the modularity of the CPDFRNN, an adaptive amplitude complex-valued real-time recurrent learning (CRTRL) algorithm is presented. Simulations demonstrate that the CPDFRNN equaliser using the amplitude CRTRL algorithm with less computational complexity not only eliminates the adverse effects of the nesting architecture, but also provides a superior performance over the CRNN and CDFRNN equalisers for non-linear channels in wireless communication systems.

Genetic Algorithm Based RNN Structure for Rayleigh Fading MIMO Channel Estimation

Multi-Input Multi-Output (MIMO) wireless communication system is an emerging area which offers substantial advantages for achieving high data rate with cost effective design. The spectral efficiency and reliability of MIMO systems greatly depends on the assumption that the transmitter and/or the receiver have perfect knowledge of channel state information (CSI). Hence, it is required to predict the CSI at the receiver and send the updated CSI back to the transmitter. In this paper particularly the Frequency-Flat Rayleigh i.i.d MIMO channel is considered, where as the Rayleigh fading channel provides complex and random coefficients. These complex channel parameters are estimated by using split complex real-time recurrent learning (SCRTRL) and fully complex real-time recurrent learning (FCRTRL) algorithm based recurrent neural network (RNN) structure with complex weights. The RNN with SCRTRL and RNN with FCRTRL algorithm produced a premature convergence which is overcome by the proposed genetic algorithm (GA) based learning process.

Low-Complexity Nonlinear Adaptive Filter Based on a Pipelined Bilinear Recurrent Neural Network

To reduce the computational complexity of the bilinear recurrent neural network (BLRNN), a novel low-complexity nonlinear adaptive filter with a pipelined bilinear recurrent neural network (PBLRNN) is presented in this paper. The PBLRNN, inheriting the modular architectures of the pipelined RNN proposed by Haykin and Li, comprises a number of BLRNN modules that are cascaded in a chained form. Each module is implemented by a small-scale BLRNN with internal dynamics. Since those modules of the PBLRNN can be performed simultaneously in a pipelined parallelism fashion, it would result in a significant improvement of computational efficiency. Moreover, due to nesting module, the performance of the PBLRNN can be further improved. To suit for the modular architectures, a modified adaptive amplitude real-time recurrent learning algorithm is derived on the gradient descent approach. Extensive simulations are carried out to evaluate the performance of the PBLRNN on nonlinear system identification, nonlinear channel equalization, and chaotic time series prediction. Experimental results show that the PBLRNN provides considerably better performance compared to the single BLRNN and RNN models.

Pipelined functional link artificial recurrent neural network with the decision feedback structure for nonlinear channel equalization

This paper presents a computationally efficient nonlinear adaptive filter by a pipelined functional link artificial decision feedback recurrent neural network (PFLADFRNN) for the design of a nonlinear channel equalizer. It aims to reduce computational burden and improve nonlinear processing capabilities of the functional link artificial recurrent neural network (FLANN). The proposed equalizer consists of several simple small-scale functional link artificial decision feedback recurrent neural network (FLADFRNN) modules with less computational complexity. Since it is a module nesting architecture comprising a number of modules that are interconnected in a chained form, its performance can be further improved. Moreover, the equalizer with a decision feedback recurrent structure overcomes the unstableness thanks to its nature of infinite impulse response structure. Finally, the performance of the PFLADFRNN modules is evaluated by a modified real-time recurrent learning algorithm via extensive simulations for different linear and nonlinear channel models in digital communication systems. The comparisons of multilayer perceptron, FLANN and reduced decision feedback FLANN equalizers have clearly indicated the convergence rate, bit error rate, steady-state error and computational complexity, respectively, for nonlinear channel equalization.