Back Propagation Neural Network Control Research Articles

Testing Method for Intelligent Loading of Mining Emulsion Pump Based on Digital Relief Valve and BP Neural Network Control Algorithm

The mining emulsion pump is mainly used on a fully mechanized coal mining face, but it is rarely used on other occasions, so research on its loading test method is relatively limited. This paper proposes the application of a digital relief valve to the emulsion pump loading test. In addition, the small number of plungers in the emulsion pump will lead to large flow pulsation and pressure pulsation, and the nominal flow of different types of emulsion pumps varies greatly. These factors lead to the deficiency of a traditional PID control algorithm in control accuracy and efficiency. In order to improve control accuracy and efficiency, firstly, the influence of the flow rate of the tested pump and extension of the linear stepping motor shaft on the working pressure is studied. A backpropagation (BP) artificial neural network (ANN) model is used to fit a functional relationship between the three parameters. The flow rate of the tested pump and target pressure were provided as inputs to predict the extension of the linear stepping motor shaft, thereby realizing the remote intelligent control of the system pressure. Next, a BP ANN model is constructed, and its reliability is verified; the BP neural network algorithm and proportional-integral-derivative (PID) algorithm are compared through simulation. The simulation results show that the BP neural network algorithm has high control accuracy and small overshoot. Finally, two pumps with different flows are tested in a self-developed digital relief valve and test platform. The test results show that the proposed loading test method is intelligent and efficient, and it has high accuracy.

Power quality improvement by PV integrated UPQC using multi-level inverter with resilient back propagation neural network approach

To increase the life and efficiency of power electronics equipment in a utility distribution system, the power quality improvement is essential part. In this work, to improve power quality by using Robust Resilient Back Propagation Neural Network (RBPNN) scheme for a Photovoltaic (PV)-Integrated Unified Power Quality Conditioner (UPQC) with cascaded multi-level inverter configurations are described. Among the proposed methods, there is no need to use a transformer and filter when multilevel UPQC is applied, and it is one of the great advantages. The proposed UQPC offers a PV array composition with a power converter connected to a DC-link capacitor that can compensate for voltage sag, swell, voltage interruption, harmonics and reactive power. The Robust Resilient Back Propagation Neural Network controller is generate gating pulses to the UPQC. The reference currents and voltages for the controller are estimated using Synchronous Reference Frame (SRF) theory. The proposed cascaded multi-level inverter-based UPQC is designed using Matlab/Simulink Software. The simulation results confirm that the proposed method gives good results compared with existing Adaptive neural Fuzzy Inference System (ANFIS) and fuzzy logic methods. A real-time hardware system is also established to validate the simulation results. The effectiveness of the proposed system RBPN-UPQC approach is compared for both simulation and experimental results gives better low THD level 1.22%.

On an enhanced back propagation neural network control of vehicle semi-active suspension with a magnetorheological damper

To improve the ride quality of a vehicle, an enhanced vibration control method is presented for semi-active suspension (SAS) with magnetorheological (MR) damper by combining back propagation neural network (BPNN) and particle swarm optimization (PSO). Based on the test data of MR damper, a non-parametric model of MR damper using adaptive neuro-fuzzy inference system (ANFIS) is first established, and based on that, a dynamics model of the SAS system is derived. Next, a BPNN controller is designed to fulfill the effective control of the current in MR damper. Meanwhile, the improved PSO with adaptive weight and dynamic acceleration constant is introduced to optimize the weights and thresholds of the BPNN controller, which can avoid the designed BPNN falling into the local optimum and then improve the convergence rate of the designed controller. Besides, the stability of the developed controller is analyzed via Lyapunov stability theory. Different from the existing models and methods, the established model can well describe the dynamics behaviors of the actual MR damper, and the proposed control method has better adaptability, convergence speed and precision. Finally, a simulative investigation is performed to validate the effectiveness and feasibility of the proposed controller, compared to existing BP-PID control and the passive suspension, the vehicle acceleration of SAS with this proposed controller is respectively improved by 10% and 30%.

An Adaptive Fuzzy Neural Network Based On Progressive Gaussian Approximate Filter with Variable Step Size

The nonlinear filtering problem is a hot spot in robot navigation research. As an advanced method that can effectively improve the robustness and accuracy of the system, the progressive Gaussian approximate filter with variable step size (PGAFVS) still has some shortcomings, how to resolve the nonlinear filtering problem in the application of tightly coupled integration under the premise of the prior uncertainty and further promote robustness high measurement accuracy, which becomes the purpose of this paper. This paper formulates the processing of trajectory tracking measurement noise problem as a Kalman filtering procedure and the measurement noise covariance matrix in controller, is jointly estimated based on the progressive Gaussian approximate filter (PGAF), after that, PGAFVS can be deduced. Then we proposed an adaptive fuzzy and backpropagation neural network controller based on PGAFVS (AFNPGA-VS) that can improve the application of tightly coupled integration under the premise of the prior uncertainty and further promote robustness high measurement accuracy. The simulation results show that the proposed algorithm outperforms the state-of-the-art methods.

DESIGN OF DYNAMIC NONLINEAR INTELLIGENT CONTROL SYSTEM FOR SIGNAL PROCESSING

Based on back propagation neural network (BPNN), we design a dynamic nonlinear intelligent control system for signal processing and analysis in this work. Performances of the traditional PID and BPNN control systems is compared in a case of heat power engineering system design. In the design, the training times, regulation time of the BPNN control system are evaluated when [Formula: see text] and [Formula: see text], respectively. Furthermore, the regulation time, oscillation time and amplitude of the traditional PID and BPNN control systems are compared in the dead zone. Results of the comparison show that the BPNN based dynamic nonlinear intelligent control system demonstrates superior performances over the traditional PID control system. The research results prove the advantage of BPNN in the design of nonlinear control systems and provide a scientific and effective reference for the follow-up research works.

Improved BP-NN Controller of PMSM for Speed Regulation

We have studied the speed regulation of the permanent magnet synchronous motor (PMSM) servo system in this paper. To optimize the PMSM servo system's speed-control performance with disturbances, a non-linear speed-control technique using a back-propagation neural network (BP-NN) algorithm for the controller design of the PMSM speed loop is introduced. To solve the slow convergence speed and easy to fall into the local minimum problem of BP-NN, we develope an improved BP-NN control algorithm by limiting the range of neural network outputs of the proportional coefficient Kp, integral coefficient Ki of the controller, and add adaptive gain factor β, that is the internal gain correction ratio. Compared with the conventional PI control method, our improved BP-NN control algorithm makes the settling time faster without static error, overshoot or oscillation. Simulation comparisons have been made for our improved BP-NN control method and the conventional PI control method to verify the proposed method's effectiveness.

A Back Propagation Neural Network with Double Learning Rate for PID Controller in Phase-Shifted Full-Bridge Soft-switching Power Supply

This paper mainly focuses on the control strategy for phase-shifting full-bridge soft switching electrolytic silver power supply based on Zero Voltage Switching (ZVS) soft switching technology. Taking into consideration the low performance of traditional PID control for phase-shifting full-bridge soft-switching, this paper introduce a PID improved by Back Propagation (BP) neural network with one single learning rate which is used to calculate weights from the input layer to the hidden layer and weights from the hidden layer to the output layer. After testing, it is found that setting independent learning rate for calculation of weights from the input layer to the hidden layer and weights from the hidden layer to the output layer which will not have an adverse effect on the design of the controller. Instead, the learning rate can be set according to the respective characteristics of the weights between the two layers, which is called double learning rate BP neural network PID. The simulation results indicate that compared with the single learning rate BP neural network PID control, the double learning rate BP neural network control has higher response speed, less over-shoot, short time to enter the steady state and strong immunity.

Neural Network-Based Closed-Loop Deep Brain Stimulation for Modulation of Pathological Oscillation in Parkinson’s Disease

Aiming at the problem that the Proportional-Integral-Derivative (PID) control strategy needs to readjust controller parameters for different Parkinson's disease (PD) states. This work proposes an improved control strategy that considers an artificial neural network control scheme. A backpropagation neural network (BPNN) controller is designed to solve the above problem and further to improve the performance of the closed-loop control strategy. The training data set of the BPNN controller is obtained by controlling eight different PD states (PD a - PD h ) by the PID controller and the BPNN controller is trained by the training data set to obtain a set of optimal weights. By modulating other different PD states (e.g. PD1 - PD3), the effectiveness of the PID-structure controller and BPNN controller are compared. We find that the BPNN controller can modulate different PD states without changing the controller parameters and reduce energy expenditure by 58.26%. This work is helpful for the design of more effective closed-loop deep brain stimulation (DBS) systems for clinical applications and provides a framework for the further development of closed-loop DBS.

Neuro-Sliding Control for Underwater ROV's Subject to Unknown Disturbances.

Proposed in this paper is a model-free and chattering-free second order sliding mode control (2nd-SMC) in combination with a backpropagation neural network (BP-NN) control scheme for underwater vehicles to deal with external disturbances (i.e., ocean currents) and parameter variations caused, for instance, by the continuous interchange of tools. The compound controller, here called the neuro-sliding control (NSC), takes advantage of the 2nd-SMC robustness and fast response to drive the position tracking error to zero. Simultaneously, the BP-NN contributes with its capability to estimate and to compensate online the hydrodynamic variations of the vehicle. When a change in the vehicle’s hydrodynamics occurs, the 2nd-SMC may no longer be able to compensate for the variations since its feedback gains are tuned for a different condition; thus, in order to preserve the desired performance, it is necessary to re-tune the feedback gains, which a cumbersome and time consuming task. To solve this, a viable choice is to implement a BP-NN control scheme along with the 2nd-SMC that adds or removes energy from the system according to the current condition it is in, in order to keep, or even improve, its performance. The effectiveness of the proposed compound controller was supported by experiments carried out on a mini-ROV.

SOM-Based Direct Inverse Trajectory Control System for Double-Propeller Boat Maneuvers

This paper presents the development of neural-network-based control system using self-organizing-maps (SOMs) for the maneuvers of a double-propeller boat. The performance characteristics of the developed SOM controller system are compared with a widely-used supervised learning mechanism, the backpropagation neural network (BPNN) controller. The experimental results show that the proposed unsupervised SOM controller can control the boat model with very low error, although most artificial neural network (ANN)-based controllers are usually designed using supervised learning approaches. The important characteristic of the proposed SOM controller system is that it utilizes a mapping principle instead of an error calculation such as that in the BPNN controller system; consequently, the proposed SOM controller system is not very sensitive to non-ideal training data, which produces a low control error for the generated elliptical trajectory data. It is also confirmed in these experiments that when more mapping neurons are utilized in the SOM controller, a lower control error is achieved. It is expected that in a real implementation, the SOM controller could provide more robust control than the BPNN controller in handling small disturbances such as light winds and small waves.

A New Approach to Modeling and Controlling a Pneumatic Muscle Actuator-Driven Setup Using Back Propagation Neural Networks

Pneumatic muscle actuators (PMAs) own excellent compliance and a high power-to-weight ratio and have been widely used in bionic robots and rehabilitated robots. However, the high nonlinear characteristics of PMAs due to inherent construction and pneumatic driving principle bring great challenges in applications acquired accurately modeling and controlling. To tackle the tricky problem, a single PMA mass setup is constructed, and a back propagation neural network (BPNN) is employed to identify the dynamics of the setup. An offline model is built up using sampled data, and online modifications are performed to further improve the quality of the model. An adaptive controller based on BPNN is designed using gradient descent information of the built-up model. Experiments of identifying the PMA setup using BPNN and position tracking by adaptive BPNN controller are performed, and results demonstrate the good capacity in accurate controlling of the PMA setup.

Development of a human head and neck muscle activation control model based on BPNN

With the development of vehicle active safety technology and its combination with passive safety technologies, there is a need for an updated tool to assess the performance of vehicle safety systems in both the in-crash and pre-crash phases. Consequently, developing human body models with active muscles is critical to achieve improved bio-fidelity simulation results to evaluate the effects of safety systems. In the present study, an innovative human active muscle control model based on a back-propagation neural network (BPNN) controller was developed, and the active muscle response was modelled using a feedback control strategy with the BPNN algorithm implemented in a human head and neck Sim Mechanics model. The BPNN controller was used to dynamically control the muscle activation level, the most complicated and essential factor in active muscle force calculation based on the Hill equation. Then, the active muscle control model was evaluated by simulating the human response in volunteer sled tests. The results validated the proposed active muscle control method as successful and easily realized. Achieving muscle active control would further improve the bio-fidelity of the human body model.

The limit cycle oscillation of divergent instability control based on classical flutter of blade section

Numerical simulation of a novel fuzzy control and back propagation neural network (BPNN) control for divergent instability based on classical flutter of 5-DOF wind turbine blade section driven by pitch adjustment has been investigated. The work is dedicated to solving destructive flap/lag/twist divergent instability from classical flutter, which might occur during the gust wind action, and might cause fracture failure of the blade itself and tower body. In order to investigate the optimal control method, the parameters of blade section are specially designed so as to simulate the actual situation, which lead to absolutely divergent motions (ADM) under gust wind load. The control of ADM often leads to limit cycle oscillation (LCO), the larger amplitude of which is likely to cause fracture failure of tower body. A novel fuzzy control method with adjustable quantization gain and BPNN control strategy are investigated in order to effectively eliminate LCO (leading to direct convergence of the system) or reduce the amplitude of LCO. The obvious effects of fuzzy control and BPNN control are illustrated by numerical comparisons of vibration suppression from nonlinear time response, amplitude of LCO and frequency spectrum analysis. An experimental platform is built based on hardware-in-the-loop simulation by way of PLC-OPC technology in order to test the real-time performance of the control algorithm. The feasibility of the control algorithm is demonstrated by the experimental results displayed by touch-screen hardware.

Research on the optimization mechanism of loading path in hydroforming process

In this paper, the optimization mechanism of loading path in hydroforming process is researched. Firstly, the geometric model of an X-shaped tube is established by using 3D drawing software UG; the DYNAFORM software is used to simulate the forming performance of the X tube under different loading paths. The backward displacement is taken as the main factor of the loading path, and the loading path is presented in the form of three factor graphs; it can show the relationship of the main factors of the loading path visually, accurately, and precisely, which are axial feed, internal pressure, and back displacement. Secondly, the orthogonal test method is used to select the optimal loading path, and the back propagation (BP) neural network based on genetic algorithm is used to optimize the loading path of X tubes. Through synthetic consideration of the interrelation of the minimum wall thickness, the maximum wall thickness, the height of branch, and the contact area between branch tube and back punch, the average performance index function is established in the BP neural network control algorithm to optimize the learning efficiency and shorten the calculation time. Finally, verified by experiment, the optimization method of loading path for X tube hydroforming process could control the precision error of results between the simulation and the experiment within 5%, and has high accuracy and good feasibility.

Tracking Control of an Ultrasonic Linear Motor Actuated Stage Using a Sliding-mode Controller with Friction Compensation

The purpose of this paper is to investigate the nonlinear friction and compensation for a piezoelectric ceramic ultrasonic motor (USM). Although the architecture of the USM is different from the general electric-mechanical motor, the mathematic model for the USM motor can use the same friction model to formulate the friction phenomenon. To establish the feedforward controller, the system identification for the USM is needed to study to design the model-based controller. To obtain the optimal system parameters of the USM, PSO and CSS algorithms are studied to identify the system parameters for the nonlinear friction model. For the controller design, a non-model based controller, using back-propagation neural network controller to perform tracking tasks, and the model-based controller, which consists of the feed-forward controller based on the system identification and the sliding-mode control, are discussed in this paper. Finally, the two real-time tracking tasks are used to validate the proposed method.

Tracking Control of an Ultrasonic Linear Motor Actuated Stage Using a Sliding-mode Controller with Friction Compensation

The purpose of this paper is to investigate the nonlinear friction and compensation for a piezoelectric ceramic ultrasonic motor (USM). Although the architecture of the USM is different from the general electric-mechanical motor, the mathematic model for the USM motor can use the same friction model to formulate the friction phenomenon. To establish the feedforward controller, the system identification for the USM is needed to study to design the model-based controller. To obtain the optimal system parameters of the USM, PSO and CSS algorithms are studied to identify the system parameters for the nonlinear friction model. For the controller design, a non-model based controller, using back-propagation neural network controller to perform tracking tasks, and the model-based controller, which consists of the feed-forward controller based on the system identification and the sliding-mode control, are discussed in this paper. Finally, the two real-time tracking tasks are used to validate the proposed method.

A Differential Evolution Based Adaptive Neural Network Pitch Controller for a Doubly Fed Wind Turbine Generator System

Extraction of maximum energy from wind and transferring it to the grid with high efficiency are challenging problems. To this end, this study proposes a smart pitch controller for a wind turbine-doubly fed induction generator system using a Differential Evolution (DE) based adaptive neural network. The nominal weights for the back-propagation neural network controller are obtained from input-output training data generated by DE optimization method. These weights are then adaptively updated in time domain depending on the variation of the system outputs. The adaptive control strategy has been tested through simulation of complete system dynamics comprising of the turbine-generator system and its various components. It has been observed that the DE based smart pitch controller is able to achieve efficient energy transfer to the grid and at the same time provide a good damping profile. Locally collected wind data was used in the testing phase.

A modified Smith predictive scheme based back-propagation neural network approach for FOPDT processes control

In this paper, a novel control scheme to deal with process uncertainties in the form of disturbance loads and modelling errors, as well as time-varying process parameters is proposed by applying the back-propagation neural network (BPNN) approach. A BPNN predictive controller that replaces the entire Smith predictor structure is initially trained offline. Lyapunov direct method is used to prove that the convergence of this BPNN is guaranteed by selecting a suitable learning rate during the learning process. However, the Smith predictor based BPNN control is an off-line training based algorithm, which is a time consuming method and requires a known process plant input from the controller. A desired control input to the process is difficult to obtain for the training of the network. As a result a group of proper training data (target control inputs and outputs) can hardly be provided. In order to overcome this problem, a BPNN with an on-line training algorithm is introduced for the control of a First Order plus Dead Time (FOPDT) process. The stability analysis is carried out using the Lyapunov criterion to demonstrate the network convergence ability. Simulation results show that this proposed online trained neural Smith predictor based controller provides excellent robustness to process modelling errors and disturbance loads, and high adaptability to time varying processes parameters.

Intelligent controller for managing power flow within standalone hybrid power systems

This study presents a novel adaptive management strategy for power flow in standalone hybrid power systems. The method introduces an on-line energy management by using a hierarchical controller between three energy sources: photovoltaic (PV) panels, battery storage and proton exchange membrane fuel cell. The proposed method includes a feed-forward, back-propagation neural network controller in the first layer, which is added in order to achieve the maximum power point for the different types of PV panels. In the second layer, a fuzzy logic controller has been developed to optimise performance by distributing the power inside the hybrid system and by managing the charge and discharge of the current flow. Finally, and in the third layer, local controllers are presented to regulate the fuel cell/battery set points in order to reach to best performance. Moreover, perturb and observe algorithm with two different controller techniques - the linear proportional-integral (PI) and the non-linear passivity-based controller - are provided for a comparison with the proposed maximum power point tracking controller system. The comparison revealed the robustness of the proposed PV control system for solar irradiance and load resistance changes. Real-time measured parameters and practical load profiles are used as inputs for the developed management system. The proposed model and its control strategy offer a proper tool for optimising the hybrid power system performance, such as the one used in smart-house applications.

Research on the Speed of Diesel Engine Based on Improved BP Neural Network Controller

Marine diesel engine is of characteristics of non-linear and time-invariant, so it is difficult to be controlled with traditional PID controller. An adaptive controller based on back-propagation (BP) neural networks was put forwarded for marine diesel engine speed control system, where two neural networks are proposed to control the position loop and speed loop. The adaptive controller was improved was improved via introducing relative error in target evaluation function of the BP neural network, and obtain sensitivity function of diesel engine output with respect to its input using a differential equation. The controller has self-learning and adaptive capacity. It can also optimize the PID controller parameters online. The controller was experimentally evaluated on rack position actuator of marine diesel engine simulated based on a diesel hardware-in-loop system of dSPACE. Finally, tests on a diesel engine demonstrated that the controller can satisfy the transient and steady demands of speed regulation system.