Single Neuron Controller Research Articles

Analysis and control design for input‐series output‐parallel multi‐channel inductive power transfer system

AbstractTo realize high‐power inductive power transfer (IPT) for fast charging of electric vehicles (EVs), an input‐series output‐parallel (ISOP) multi‐channel IPT system is analysed in this paper, and an output control strategy based on single neuron controller is proposed to improve the stability of the system. Firstly, the steady‐state operating conditions of ISOP‐IPT system is analysed based on different compensation networks which shows that the constant‐voltage‐output compensation networks are more suitable for the proposed circuit structure. Then, to improve the voltage equalization between channels, an open‐loop control method combining parameter design and phase‐shift control is proposed against different coil misalignments. At the same time, based on the single neuron controller, the system output closed‐loop control is realized without the need of accurate system modelling. Finally, a three‐channel ISOP‐IPT experimental system was constructed, which operated stably at 2.9 kW with a power efficiency of 93.38%. The system closed‐loop control with control time of 22 ms is realized. The voltage equalization control offset is less than 1%.

Analysis of the Control Characteristics of the Electro-Hydraulic Vibration System Based on the Single-Neuron Control Algorithm

This paper proposes an electro-hydraulic vibration control system based on the single-neuron PID algorithm, which improves the operating frequency of the electro-hydraulic fatigue testing machine and the control accuracy of the load force. Through mathematical modeling of the electro-hydraulic vibration system (EVS), a MATLAB/Simulink simulation, and experimental testing, this study systematically analyzes the output waveform of the EVS as well as the closed-loop situation of load force amplitude and offset under the action of the single-neuron PID algorithm. The results show that: the EVS with a 2D vibration valve as the core, which can control the movement of the spool in the two-degrees-of-freedom direction, can realize the output of an approximate sinusoidal load force waveform from 0 to 800 Hz. The system controlled by the single-neuron PID algorithm is less complex to operate than the traditional PID algorithm. It also has a short rise time for the output load force amplitude curve and a maximum control error of only 1.2%. Furthermore, it exhibits a rapid closed-loop response to the load force offset. The range variability of the load force is measured to be 1.43%. A new scheme for the design of EVS is provided in this study, which broadens the application range of electro-hydraulic fatigue testing machines.

Single-neuron adaptive pitch-depth control for a lift principle AUV with experimental verification

This paper addresses the problem of pitch and depth control for a novel deep-sea, high-speed, and heavy-load autonomous underwater vehicle based on the lift principle (LP-AUV). One of the characteristics of an LP-AUV is the use of downward lift to overcome varying residual buoyancy and achieve rapid diving. The pitch angle determines the action direction of the lift force, thus affecting the safety and stability of diving and navigation. Specifically, the main requirement for the controller is accuracy and no overshoot in either the depth or pitch outputs of the system. To offer accurate and reliable control ability, with various speeds and residual buoyancies, unmatched disturbances and the action of lift taken into consideration, a single-neuron controller with self-learning and self-adaptive capabilities is proposed. Based on the Hebb learning rule, a supervised learning rule that can learn not only from the principles of neural networks but also from tuning experience is designed. Hence, the weights of the single-neuron controller can be constantly adjusted online to improve the control performance of the system more rapidly. Simulation results show the robustness and effectiveness of the proposed controller toward external disturbances. Finally, a series of lake tests using the LP-AUV are conducted to validate the proposed method.

Operant conditioning reveals task-specific responses of single neurons in a brain–machine interface

Objective. Volitional modulation of single cortical neurons holds great potential for the implementation of brain–machine interfaces (BMIs) because it can induce a rapid acquisition of arbitrary associations between machines and neural activity. It can also be used as a framework to study the limits of single-neuron control in BMIs. Approach. We tested the control of a one-dimensional actuator in two BMI tasks which differed only in the neural contingency that determined when a reward was dispensed. A thresholded activity task, commonly implemented in single-neuron BMI control, consisted of reaching or exceeding a neuron activity level, while the second task consisted of reaching and maintaining a narrow neuron activity level (i.e. windowed activity task). Main findings. Single neurons in layer V of the motor cortex of rats improved performance during both the thresholded activity and windowed activity BMI tasks. However, correct performance during the windowed activity task was accompanied by activation of neighboring neurons, not in direct control of the BMI. In contrast, only neurons in direct control of the BMI were active at the time of reward during the thresholded activity task. Significance. These results suggest that thresholded activity single-neuron BMI implementations are more appropriate compared to windowed activity BMI tasks to capitalize on the adaptability of cortical circuits to acquire novel arbitrary skills.

Single-Neuron Adaptive Hysteresis Compensation of Piezoelectric Actuator Based on Hebb Learning Rules.

This paper presents an adaptive hysteresis compensation approach for a piezoelectric actuator (PEA) using single-neuron adaptive control. For a given desired trajectory, the control input to the PEA is dynamically adjusted by the error between the actual and desired trajectories using Hebb learning rules. A single neuron with self-learning and self-adaptive capabilities is a non-linear processing unit, which is ideal for time-variant systems. Based on the single-neuron control, the compensation of the PEA’s hysteresis can be regarded as a process of transmitting biological neuron information. Through the error information between the actual and desired trajectories, the control input is adjusted via the weight adjustment method of neuron learning. In addition, this paper also integrates the combination of Hebb learning rules and supervised learning as teacher signals, which can quickly respond to control signals. The weights of the single-neuron controller can be constantly adjusted online to improve the control performance of the system. Experimental results show that the proposed single-neuron adaptive hysteresis compensation method can track continuous and discontinuous trajectories well. The single-neuron adaptive controller has better adaptive and self-learning performance against the rate-dependence of the PEA’s hysteresis.

Generalized Correntropy Predictive Control for Waste Heat Recovery Systems Based on Organic Rankine Cycle

Organic Rankine cycle (ORC) is one of the most promising technologies to recover energy from low temperature waste heat. The heat sources usually experience fluctuations in temperature and mass flow rate, which makes it difficult to obtain satisfactory control performances of ORC systems. In this paper, a single neuron adaptive multi-step predictive control scheme is developed for an ORC based waste heat recovery (WHR) system with non-Gaussian disturbances. Since the non-Gaussian disturbances existed in WHR system follow heavy-tailed distribution, generalized correntropy is adopted as the performance index to characterize the system uncertainties. The weights of the single neuron controller are updated by optimizing the performance function. The whole implementation procedures of the proposed control strategy for the WHR are presented. As a contrast, the performances of the minimum error entropy (MEE) based controller and the mean square error (MSE) based controller are also tested. The proposed control scheme is confirmed to be more effective through simulations.

A Novel Adaptive Neuro-Control Approach for Permanent Magnet Synchronous Motor Speed Control

A speed controller for permanent magnet synchronous motors (PMSMs) under the field oriented control (FOC) method is discussed in this paper. First, a novel adaptive neuro-control approach, single artificial neuron goal representation heuristic dynamic programming (SAN-GrHDP) for speed regulation of PMSMs, is presented. For both current loops, PI controllers are adopted, respectively. Compared with the conventional single artificial neuron (SAN) control strategy, the proposed approach assumes an unknown mathematic model of the PMSM and guides the selection value of parameter K online. Besides, the proposed design can develop an internal reinforcement learning signal to guide the dynamic optimal control of the PMSM in the process. Finally, nonlinear optimal control simulations and experiments on the speed regulation of a PMSM are implemented in Matlab2016a and TMS320F28335, a 32-bit floating-point digital signal processor (DSP), respectively. To achieve a comparative study, the conventional SAN and SAN-GrHDP approaches are set up under identical conditions and parameters. Simulation and experiment results verify that the proposed controller can improve the speed control performance of PMSMs.

Single Neuron Stochastic Predictive PID Control Algorithm for Nonlinear and Non-Gaussian Systems Using the Survival Information Potential Criterion

This paper presents a novel stochastic predictive tracking control strategy for nonlinear and non-Gaussian stochastic systems based on the single neuron controller structure in the framework of information theory. Firstly, in order to characterize the randomness of the control system, survival information potential (SIP), instead of entropy, is adopted to formulate the performance index, which is not shift-invariant, i.e., its value varies with the change of the distribution location. Then, the optimal weights of the single neuron controller can be obtained by minimizing the presented SIP based predictive control criterion. Furthermore, mean-square convergence of the proposed control algorithm is also analyzed from the energy conservation perspective. Finally, a numerical example is given to show the effectiveness of the proposed method.

Inhibitory single neuron control of seizures and epileptic traveling waves in humans

Inhibitory neuronal activity is critical for the normal functioning of the brain, but is thought to go awry during neurological disorders such as epilepsy. Animal models have suggested both decreased and increased inhibition as possible initiators of epileptic activity, but it is not known if, or how, human inhibitory neurons shape seizures. Here, using large-scale recordings of neocortical single neurons in patients with secondarily generalized tonic-clonic seizures, we show that fast-spiking (FS) inhibitory activity first increases as a seizure spreads across the neocortex, impeding and altering the spatial flow of fast epileptic traveling waves. Unexpectedly, however, FS cells cease firing less than half-way through a seizure. We use biophysically-realistic computational models to show that this cessation is due to FS cells entering depolarization block as a result of extracellular potassium accumulation during the seizure and not because they are inhibited by other inhibitory subtypes. Strikingly, this absence of FS inhibitory activity is accompanied by dramatic increases in local seizure amplitude along with unobstructed traveling waves and is seen during all secondarily generalized seizures examined, independent of etiology or focus. FS cessation also leads to prominent spike-and-wave events, suggesting that FS cell dynamics control the transition between the tonic and clonic phases of these seizures. Thus, it may be possible to curtail human seizures by preventing inhibitory neurons from entering potassium-dependent depolarization block, a novel and potentially powerful therapeutic avenue in treating multiple kinds of epilepsies.

Single Neuron PID Control of a Turbofan Engine with Inlet Pressure Distortion

In this paper, a preliminary exploration and research about the adaptive control law has been carried on for the turbofan engine with inlet pressure distortion. First, a real-time dynamic component-based model which contains the inlet distortion effects was established. Then, we select a single neuron PID controller with adaptive parameters to complete the dynamic acceleration process. The results illustrate that the single neuron control algorithm with gain adaptive has strong robustness, self-learning and anti-jamming capability and the performance of a turbofan engine can be effectively improved by using the turbine expansion ratio adaptive control law.

Improved single neuron controller for multivariable stochastic systems with non-Gaussianities and unmodeled dynamics

In this paper, a new adaptive control approach is presented for multivariate nonlinear non-Gaussian systems with unknown models. A more general and systematic statistical measure, called (h,ϕ)-entropy, is adopted here to characterize the uncertainty of the considered systems. By using the “sliding window” technique, the non-parameter estimate of the (h,ϕ)-entropy is formulated. Then, the improved neuron based controllers are developed for multivariate nonlinear non-Gaussian systems by minimizing the entropies of the tracking errors in closed loops. The condition to guarantee the strictly decreasing entropy of tracking error is presented. Moreover, the convergence in the mean-square sense has been analyzed for all the weights in the neural controllers. Finally, the comparative simulation results are presented to show that the performance of the proposed algorithm is superior to that of PID control strategy.

Statistical Information Based Single Neuron Adaptive Control for Non-Gaussian Stochastic Systems

Based on information theory, the single neuron adaptive control problem for stochastic systems with non-Gaussian noises is investigated in this paper. Here, the statistic information of the output within a receding window rather than the output value is used for the tracking problem. Firstly, the single neuron controller structure, which has the ability of self-learning and self-adaptation, is established. Then, an improved performance criterion is given to train the weights of the single neuron. Furthermore, the mean-square convergent condition of the proposed control algorithm is formulated. Finally, comparative simulation results are presented to show that the proposed algorithm is superior to the PID controller. The contributions of this work are twofold: (1) the optimal control algorithm is formulated in the data-driven framework, which needn’t the precise system model that is usually difficult to obtain; (2) the control problem of non-Gaussian systems can be effectively dealt with by the simple single neuron controller under improved minimum entropy criterion.

PID Controll of Single Neuron for Switched Reluctance Motors Based on RBF Neural Network

This paper presents a novel approach of single neuron PID control for switched reluctance motors based on RBF neural network on-line identification. The method is adjusted to the nonlinearity of switched reluctance motors, and use the single neurons capable of self-learning and self-adaption to form the single neuron adaptive controller of switched reluctance motors. It not only has simple structure and strong robustness, but but can adapt to environmental changes. Also we construct a RBF network system to identify the system online, and to build its online reference model, using a single neuron controller to achieve self-learning of its parameters, in order to achieve their online adjustment, and to obtain better control effect.

Immune Single Neuron Control of Level in Spherical Vessel

To solve the problem of nonlinear characteristics of spherical vessels, an immune single neuron PID controller (ISNPID) is proposed in this paper. The ISNPID controller combines the immune principle with single neuron PID, it can improve the learning rate and reduce the adjusting time of single neuron PID controller. The interim process of the spherical vessel system could also be optimized. Good control performance of nonlinear system could be realized according to the on-line self-adjusting gain by the ISNPID controller. Simulation result shows that ISNPID controller has strong adaptability and robustness. Efficiency and applicability of the ISNPID controller could be proved in practical application.

Optimal Design and Control of CPU Heat Sink Processes

This paper considers the optimal plate fin design and control for central processing unit (CPU) heat sink processes. First, we apply a finite element method to investigate the heat transfer phenomena of a heat sink process. To have a better heat dispersion performance, a real-coded genetic algorithm is then utilized to search for an optimal set of plate-fin shape parameters. The objective function to be minimized is the entropy generation rate which can take simultaneously the two major factors, heat transfer rate and air resistance, into consideration in the design. The present optimization scheme is able to achieve a better design for heat dispersion than existing methods. To attenuate environmental and time-varying disturbances, a direct adaptive control scheme is then developed for the CPU heat sink process. It is based on using a bounded single neuron controller (SNC) along with a parameter tuning algorithm to regulate the temperature of a selected control point. Extensive comparisons of the SNC-based control performance with the on-off control as well as a PI controller show that the proposed scheme provides excellent control performance despite the existence of unexpected process uncertainties.

Research on Anti-Sway Control System for Overhead Crane Based on Single Neuron

During the operation of trolley crane systems, the anti-sway control of the suspended load is always a key problem .Due to the uncertain properties of the trolley crane model, it is difficult for the conventional PID controller to tune its parameters online, so it can not reduce the swing effectively. In this paper, we design a control system based on single neuron controllers which using single neuron controllers to control the swing motion. The computer simulation results show that the single neuron controller has better adaptability and robustness than PID controller in term of different rope length and load.

Direct adaptive process control based on using a single neuron controller: Survey and some new results

This paper surveys the process control techniques for using an adaptive single neuron controller (SNC). In addition to an introduction to a derived parameter tuning algorithm, this paper focuses on how the SNC can be implemented in combination with existing strategies to deal with processes in the presence of diversified dynamics. Following this lead, some alternative SNC control schemes are presented and the related convergence analyses are provided. Besides, some new results on the development of model‐based SNC control systems are established, which enables the SNC to start in a systematic way by the assignment of a pre‐specified performance index. To demonstrate the effectiveness and applicability of the proposed SNC control schemes, examples are provided. Extensive simulation results show that the proposed modelbased SNC control system is a simple, easy‐to‐implement, and promising approach to the direct adaptive control of chemical processes.

Process control of pH neutralization based on adaptive algorithm of universal learning network

This paper presents an adaptive algorithm of universal learning network (ULN) and its application to identify pure time delay of a plant model. Universal learning network can be used in model predictive control for stabilizing a class of nonlinear systems with long time delay. Depending on ULN model with single neuron controller, the control architectures are introduced and applied to pH neutralization process. Simulation results prove the applicability and effectiveness of the ULN model. The general architecture and adaptive learning algorithm give ULN more representing abilities to model and control the nonlinear black box systems with long time delay.

A Neuro-Sliding Mode Controller for STATCOM

The paper presents a new controller design for STATCOM for damping electromechanical oscillations in a power system. The STATCOM is modeled as a reactive current source and the magnitude of this current is obtained by using a sliding mode control strategy. The parameters of the sliding mode controller (SMC) are obtained using linearised model of single-machine infinite-bus power system and a pole placement technique. In order to avoid repeated computation of the sliding mode control gains at different operating conditions, a single neuron controller is used to adapt the parameters of the SMC. Several simulation results are presented to show the marked improvement in transient stability enhancement of the single-machine infinite-bus and multimachine power systems over a wide range of operating conditions.

A radial basis function neural network controller for UPFC

This paper presents the design of radial basis function neural network controllers (RBFNN) for UPFC to improve the transient stability performance of a power system. The RBFNN uses either a single neuron or multi-neuron architecture and the parameters are dynamically adjusted using an error surface derived from active or reactive power/voltage deviations at the UPFC injection bus. The performance of the new single neuron controller is evaluated using both single-machine infinite-bus and three-machine power systems subjected to various transient disturbances. In the case of three-machine 8-bus power system, the performance of the single neuron RBF controller is compared with a BP (backpropagation) algorithm based multi-layered ANN controller. Further it is seen that by using a multi-input multi-neuron RBF controller, instead of a single neuron one, the critical clearing time and damping performance are improved. The new RBFNN controller for UPFC exhibits a superior damping performance in comparison to the existing PI controllers. Its simple architecture reduces the computational burden thereby making it attractive for real-time implementation.