Network Controller Research Articles

Research and Design of Hybrid Optimized Backpropagation (BP) Neural Network PID Algorithm for Integrated Water and Fertilizer Precision Fertilization Control System for Field Crops

China’s field crops such as cotton, wheat, and tomato have been produced on a large scale, but their cultivation process still adopts more traditional manual fertilization methods, which makes the use of chemical fertilizers in China high and causes waste of fertilizer resources and ecological environmental damage. To address the above problems, a hybrid optimization of genetic algorithms and particle swarm optimization (GA–PSO) is used to optimize the initial weights of the backpropagation (BP) neural network, and a hybrid optimization-based BP neural network PID controller is designed to realize the accurate control of fertilizer flow in the integrated water and fertilizer precision fertilization control system for field crops. At the same time, the STM32 microcontroller-based precision fertilizer application control system for integrated water and fertilizer application of large field crops was developed and the performance of the controller was verified experimentally. The results show that the controller has an average maximum overshoot of 5.1% and an average adjustment time of 68.99 s, which is better than the PID and PID control algorithms based on BP neural network (BP–PID) controllers; among them, the hybrid optimization of PID control algorithm based on BP neural network by particle swarm optimization and genetic algorithm(GA–PSO–BP–PID) controller has the best-integrated control performance when the fertilizer application flow rate is 0.6m3/h.

Total harmonic distortion reduction based energy harvesting using grid-based three phase system and integral-derivative

Distributed generation (DG) and solar photovoltaic (PV) systems are just two of the many places multilevel inverters have found a home. The total harmonic distortion (THD) of an inverter's output voltage or current is widely regarded as a useful quality indicator. In order to lessen THD while enhancing the performance of the network, the authors of this research presented a unified controller.An integrated controller consisting of PID (Proportional-Integral-Derivative),Fuzzy,and Artificial Neural Network (ANN) controllers is used in the 31-level multilevel inverter to reduce THD in the PV system. The fundamental frequency of the suggested integrated controller is 50 Hz,and the simulation results show that the THD is reduced to 0.94 percent. The proposed controller's performance shown a considerable decrease in 5th order harmonics for the Cascaded H-bridge Multilevel Inverter,as well as an average percentage reduction in line voltage total harmonic distortion of 17.04 compared to the existing approach.

Design of a BP neural network PID controller for an air suspension system by considering the stiffness of rubber bellows

A rubber bellows has certain creep characteristics, and its role has often been ignored in previous air suspension research. To describe the motion state of a vehicle more accurately during driving, this paper uses amplitude correlation and frequency correlation to describe the mechanical characteristics of the rubber bellows in air spring mechanical characteristics research, establishes an air suspension simulation, and explores the role of the rubber bellows in the suspension. In addition, an adaptive neural network controller is designed to simulate and analyze the Body Acceleration (BA), Sprung Mass Displacement (SMD), Unsprung Mass Displacement (UMD) and Dynamic Wheel Load (DWL) of the suspension. Simulation results show that the rubber bellows has a certain impact on suspension SMD and has a great impact on the performance of the active suspension. The root mean square (RMS) can be increased by 4.55 % but little impact is found on other performance indicators. In addition, compared with the conventional PID control strategy, the adaptive neural network control strategy designed in this paper has better control performance, reducing the RMS of BA and SMD by 27.58 % and 16.67 %, respectively, and improving the control effect by about 4.48 % and 4.17 %, respectively compared with PID. Thus, it can better maintain the stability of the car during driving.

Modeling of a neural network-based system for identification and control of technical object parameters

The article analyzes the effectiveness of a neural network control system for main-taining the pH level in the feedwater of a steam boiler. An intelligent control system im-plements the principle of reverse error propagation through a neural emulator. The sub-system model of steam boiler water tube blowing was used as the research object. The neural network controller and neural emulator were trained on a control system model with a PID controller using the expert correction methodology of tuning coefficients: proportionality, integration constant, and differentiation based on the analysis of tran-sient process quality indicators. The analysis of the transient processes obtained from simulation modeling allows us to state that the trained neural network control system successfully compensates for disturbances over a wide range of changes in the object's parameter values via control channels and disturbances (simulating changes in steam load).

Design of Metaheuristic Optimization with Deep-Learning-Assisted Solar-Operated On-Board Smart Charging Station for Mass Transport Passenger Vehicle

Electric vehicles (EVs) have become popular in reducing the negative impact of ICE automobiles on the environment. EVs have been predicted to be an important mode of mass transit around the globe in recent years. Several charging stations in island and remote areas are dependent on off-grid power sources and renewable energy. Solar energy is used in the daytime as it is based on several environmental components. The creation of efficient power trackers is necessary for solar arrays to produce power at their peak efficiency. To deliver energy during emergencies and store it in case there is an excess, energy storage systems are required. It has long been known that reliable battery management technology is essential for maintaining precise battery charge levels and avoiding overcharging. This study suggests an ideal deep-learning-assisted solar-operated off-board smart charging station (ODL-SOOSCS) design method as a result. The development of on-board smart charging for mass transit EVs is the main goal of the ODL-SOOSCS technique that is being described. In the ODL-SOOSCS approach described here, a perovskite solar film serves as the generating module, and the energy it generates is stored in a module with a hybrid ultracapacitor and a lithium-ion battery. Broad bridge converters and solar panels are incorporated into the deep belief network (DBN) controller, which doubles as an EV charging station. An oppositional bird swarm optimization (OBSO) algorithm is used as a hyperparameter optimizer to improve the performance of the DBN model. Moreover, an MPPT device is exploited for monitoring and providing maximal output of the solar panel if the power sources are PV arrays. The proposed system combines the power of metaheuristic optimization algorithms with deep learning techniques to create an efficient and smart charging station for mass transport passenger vehicles. This integration of two powerful technologies is a novel approach toward solving the complex problem of charging electric vehicles in mass transportation systems. The experimental validation of the ODL-SOOSCS technique is tested on distinct converter topologies. A widespread experimental analysis assures the promising performance of the ODL-SOOSCS method over other current methodologies.

An advanced bionic knee joint mechanism with neural network controller.

In this article, a tensegrity-based knee mechanism is studied for developing a high-efficiency rehabilitation knee exoskeleton. Moreover, the kinematics and dynamics models of the knee mechanism are explored for bringing about further improvement in controller design. In addition, to estimate the performance of the bionic knee joint, based on the limit function of knee patella, the limit position functionality of the bionic knee joint is developed for enhancing the bionic property. Furthermore, to eliminate the noise item and other disturbances that are constantly generated in the rehabilitation process, a noise-tolerant zeroing neural network (NTZNN) algorithm is utilized to establish the controller. This indicates that the controller shows an anti-noise performance; hence, it is quite unique from other bionic knee mechanism controllers. Eventually, the anti-noise performance and the calculation of the precision of the NTZNN controller are verified through several simulation and contrast results.

Towards a real-time IoT: Approaches for incoming packet processing in cyber–physical systems

Embedded real-time devices for monitoring, controlling, and collaboration purposes in cyber–physical systems are now commonly equipped with IP networking capabilities. However, the reception and processing of IP packets generates workloads in unpredictable frequencies as networks are outside of a developer’s control and difficult to anticipate, especially when networks are connected to the internet. As of now, embedded network controllers and IP stacks are not designed for real-time capabilities, even when used in real-time environments and operating systems.Our work focuses on real-time aware packet reception from open network connections, without a real-time networking infrastructure. This article presents two experimentally evaluated modifications to the IP processing subsystem and embedded network interface controllers of constrained IoT devices. The first, our software approach, introduces early packet classification and priority-aware processing in the network driver. In our experiments this allowed the network subsystem to remain active at a seven-fold increase in network traffic load before disabling the receive interrupts as a last resort. The second, our hardware approach, makes changes to the network interface controller, applying interrupt moderation based on real-time priorities to minimize the number of network-generated interrupts. Furthermore, this article provides an outlook on how the software and hardware approaches can be combined in a co-designed packet receive architecture.

Intelligent Beam Steering for Wireless Communication Using Programmable Metasurfaces

Reconfigurable Intelligent Surfaces (RIS) are well established as a promising solution to the blockage problem in millimeter-wave (mm-wave) and terahertz (THz) communications, envisioned to serve demanding networking applications, such as 6G and vehicular. HyperSurfaces (HSF) is a revolutionary enabling technology for RIS, complementing Software Defined Metasurfaces (SDM) with an embedded network of controllers to enhance intelligence and autonomous operation in wireless networks. In this work, we consider feedback-based autonomous reconfiguration of the HSF controller states to establish a reliable communication channel between a transmitter and a receiver via programmable reflection on the HSF when Line-of-sight (LoS) between them is absent. The problem is to regulate the angle of reflection on the metasurface such that the power at the receiver is maximized. Extremum Seeking Control (ESC) is employed with the control signals generated mapped into appropriate metasurface coding signals which are communicated to the controllers via the embedded controller network (CN). This information dissemination process incurs delays which can compromise the stability of the feedback system and are thus accounted for in the performance evaluation. Extensive simulation results demonstrate the effectiveness of the proposed method to maximize the power at the receiver within a reasonable time even when the latter is mobile. The spatiotemporal nature of the traffic for different sampling periods is also characterized.

Разработка нейросетевого регулятора для дизеля

The paper considers issues of using a neural network in the thermal engine controller in order to improve the quality of its control and setting conveniences. The study object was a speed controller for the D500 promising locomotive diesel engine. Tasks for design and development of a neural network controller were formulated, and the network input signals were determined. To adjust the neural network, the reinforcement learning method was introduced, where it interacted with the diesel engine computer model in a closed system. The criterion in setting up the network was the accuracy of the control program execution. A system of rewards was assigned, according to which the network was learning. Based on the results of studying the neural network controller influence on the quality of the control system operation, the network minimum possible composition for solving the problem presented was determined. Study results are presented in the form of graphs of rewards alteration during the learning process for various options of the neural network controller structure, as well as the control system simulated transient processes over the entire range of the diesel engine speed.

Simulation model of ANN and PID controller for TCP/AQM wireless networks by using MATLAB/Simulink

<p>The wireless network transmission control protocol/active queue management (TCP/AQM) is a network that was chosen as a topic for research a basis is laid to simulate the proposed network and to conduct the simulation under certain conditions. To manage the queue control protocol (TCP/AQM) was chosen. The solution for many modern systems depends on placing additional units called controllers, which work to improve the performance of the work of the systems. The current simulation system can be described according to the test cases that were conducted, where four test cases were identified with sequential steps. In this work there are two control methods by simulation and mathematical model of wireless network TCP/AQM with proportional, integral and derivative (PID) controller and neural network. Simulation is conducted for cases in order to determine the performance of each case through comparison according to appropriate criteria to determine the best. The first case is a wireless communication network system with with a traditional controller PID type. The second is a wireless communication network system, a large neural network controller. Simulations were conducted to choose the best methods among those suggested for nonlinear systems and to enhance and achieve the possibility of adopting MATLAB to perform the required simulation.</p>

Discovering Closed-Loop Failures of Vision-Based Controllers via Reachability Analysis

Machine learning driven image-based controllers allow robotic systems to take intelligent actions based on the visual feedback from their environment. Understanding when these controllers might lead to system safety violations is important for their integration in safety-critical applications and engineering corrective safety measures for the system. Existing methods leverage simulation-based testing (or falsification) to find the failures of vision-based controllers, i.e., the visual inputs that lead to closed-loop safety violations. However, these techniques do not scale well to the scenarios involving high-dimensional and complex visual inputs, such as RGB images. In this work, we cast the problem of finding closed-loop vision failures as a Hamilton-Jacobi (HJ) reachability problem. Our approach blends simulation-based analysis with HJ reachability methods to compute an approximation of the backward reachable tube (BRT) of the system, i.e., the set of unsafe states for the system under vision-based controllers. Utilizing the BRT, we can tractably and systematically find the system states and corresponding visual inputs that lead to closed-loop failures. These visual inputs can be subsequently analyzed to find the input characteristics that might have caused the failure. Besides its scalability to high-dimensional visual inputs, an explicit computation of BRT allows the proposed approach to capture non-trivial system failures that are difficult to expose via random simulations. We demonstrate our framework on two case studies involving an RGB image-based neural network controller for (a) autonomous indoor navigation, and (b) autonomous aircraft taxiing.

Bridgeless isolated sepic PFC for EV battery charging using ANN controller

The power quality of an Electric Vehicle (EV) battery charger is improved in this work using a bridgeless isolated single ended primary inductance converter (SEPIC). Owing to the excessive amount of conduction losses associated with the conventional Diode Bridge Rectifier (DBR) based Power Factor Correction (PFC) circuit; a Bridgeless circuit model is preferred in this work. The safe functioning of the PFC circuit is further ensured by selecting an isolated topology over a non-isolated converter topology. With the use of an artificial neural network (ANN) controller, the converter's operation is improved in terms of rising time, overshoot, steady state error and settling time. Distortions in the input current are curtailed with the aid of Hysteresis Current Controller (HCC). The entire work of the proposed converter in terms of design equations and several operating modes is detailed elaborately. The input current displays functioning with a unity power factor for the full charging period. The proposed Bridgeless Isolated SEPIC PFC's efficacy in improving the power quality of the EV charger system is determined using simulation results from MATLAB and hardware implementation utilising a DSP30F411 controller.

Energy Management For Renewable Hybrid System Based On Artificial Neural Networks (ANN)

Abstract: It has become necessary in recent years to use renewable energy sources, such as wind and solar energy, to meet the load demand. Because of their endless existence and environmentally friendly nature, they are useful for power generation. The management of energy in hybrid wind-solar power sources is the topic of this article. Using a common current source interface multiple-input DC-DC converter, the photovoltaic (PV) array, wind turbine, and battery storage are all connected. The power control between intermittent renewable energy generation, energy storage, and the grid is maintained by the artificial neural network (ANN) controller. For wind turbines to produce the most electricity, variable speed control can be used. For solar systems, the Maximum Power Point Tracking (MPPT) algorithm is used. By keeping the common DC voltage constant, the grid interface inverter sends the energy drawn from the PV array and wind turbine into the grid. Using MATLAB Simulink, the complete control strategy is subjected to a simulation study. The results of the simulation show the photovoltaic/wind hybrid system's dynamic behavior and control effectiveness

King of the castle: organisational influences on authority gradients between network controllers and other team members

The rail system relies on the effective coordination of multiple disciplines and teams situated within an operational hierarchy to meet a single operational objective—the safe and timely movement of rail traffic. Power and status dispersals across these teams and the various roles within them impact interaction and communication. This study drew on the perceptions of network controllers, to identify organisational factors influencing power imbalances that generate authority gradients between network controllers and other team members. Network controllers (N = 55) across eight Australasian organisations engaged in interviews using the Scenario Invention Task Technique to explore perceptions of risk. Thematic analysis revealed relationships between teams were affected by: (1) the accountability mechanisms adopted by organisations; (2) the way power was vested in roles; and (3) the status attached to roles. This insight into organisational power hierarchies and the generation of authority gradients provides opportunities for understanding teamwork error. Practitioner summary: Communication is impaired by authority gradients across teams in rail and is a contributing factor in incidents occurring on the network. This paper explores the organisational influences on power hierarchies across teams from the perspective of the network controller, pointing to an adversarial culture, resulting in tribalism impeding team interactions.

Switching robust neural network control of perishable inventory with fixed shelf life products under time-varying uncertain demand

Managing the inventories of perishable items becomes a more challenging task when the demand for such items is uncertain and has intermittent, lumpy or erratic character. In this study practical problem of perishable inventory products, and control is presented and solved. This is an extended version of the research presented in [Cholodowicz E, Orlowski P (2022) Robust control of perishable inventory with uncertain lead time using neural networks and genetic algorithm. In: International Conference on Computational Science. Springer, pp 46–59]. The objective of this study is to find order quantities for the perishable inventory system, that are optimal in terms of stock level, fill rate and outdating. To find the solution, a new approach is developed. The proposed approach is a combination of an artificial neural network, control theory and hybrid optimization. The proposed controller consists of robust neural network controllers which are tuned for selected perturbation bound. Weights switching operation is performed based on the demand uncertainty estimation. The validity of the results is tested on intermittent, lumpy, smooth and erratic demand data series coming from the food industry. One of the key findings of this study is that applying the proposed Switching Robust Neural Network controller provides a significant decrease in cost function in comparison to Robust Neural Network controllers. It is found that the biggest decrease is for smooth demand and the smallest one for the lumpy demand pattern. What distinguishes our work from the current state of the art is the fact that we do not limit that demand is deterministic and constant, we propose a method which calculates the optimal order quantities to minimize costs under different patterns of demand which is uncertain.

Solving the multicommodity flow problem using an evolutionary routing algorithm in a computer network environment.

The continued increase in Internet traffic requires that routing algorithms make the best use of all available network resources. Most of the current deployed networks are not doing so due to their use of single path routing algorithms. In this work we propose the use of a multipath capable routing algorithm using Evolutionary Algorithms (EAs) that take into account all the traffic going over the network and the link capacities by leveraging the information available at the Software Defined Network (SDN) controller. The designed routing algorithm uses Per-Packet multipath routing to make the best use of the network's resources. Per-Packet multipath is known to have adverse affects when used with TCP, so we propose modifications to the Multipath TCP (MPTCP) protocol to overcome this. Network simulations are performed on a real world network model with 41 nodes and 60 bidirectional links. Results for the EA routing solution with the modified MPTCP protocol show a 29% increase in the total network Goodput, and a more than 50% average reduction in a flow's end-to-end delay, when compared to OSPF and standard TCP under the same network topology and flow request conditions.

Optimal Design and Performance Investigation of Artificial Neural Network Controller for Solar- and Battery-Connected Unified Power Quality Conditioner

Nowadays, integration of renewable sources into the local distribution system and the nonlinear behavior of advanced power electronic equipment have made a large impact on the power quality (PQ). The unified power quality conditioner (UPQC) is a multifunctional FACTS device, which is a combination of both shunt active filter and series active filters via a common DC link. Presently, the artificial intelligence is playing a vital role in the development of the intelligent control methods. Traditional training methods of artificial neural network (ANN) like back propagation and Levenberg-Marquardt may get stuck in local optimal solution which leads to the invention of ANN trained optimally by metaheuristic algorithms. This paper develops a firefly algorithm-trained ANN (FF-ANNC) controller for the shunt active filter and proportional integral controller (PI-C) for the series active filter of the UPQC integrated with the solar energy system and battery energy storage via boost converter (B-C) and buck boost converters (B-B-C). The main aim of the proposed FF-ANNC is to reduce the mean square error (MSE) thereby achieving the constant DC link capacitor voltage (DLCV) during load and irradiation variations, reduction of imperfections in current waveforms, improvement in power factor (PF), and mitigation of sag, swell, disturbances, and unbalances in the grid voltage. The working of developed FF-ANNC was tested on five test studies with different types of loads and source voltage balancing/unbalancing conditions. However, to demonstrate supremacy of the suggested FF-ANNC, a comparative study with the training methods like genetic algorithm (GA) and ant colony optimization (AC-O) and also with other methods that exist in literature like PI-C, fuzzy logic controller (FL-C), and artificial neuro fuzzy interface system (ANFI-S) was conducted. The proposed method reduces the total harmonic distortion to 2.39%, 2.32%, 2.27%, 2.45%, and 2.66% which are lower than the existing methods that are available in literature. The FF-ANNC shows an excellent performance in reducing voltage fluctuations and total harmonic distortion (THD) successfully and thereby improving PF.

Artificial neural network based identification of process dynamics and neural network controller design for continuous distillation column

In this study, an equation based nonlinear autoregressive with exogenous input neural network model for a multiple input and output process was proposed to enhance the accuracy of the model, a control system’s speed, and robustness. A continuous lab-scale binary distillation column was considered for a case study, and an experiment was conducted. The forward and inverse neural network models have been developed and incorporated into a multivariable internal model control strategy to control the distillate and bottom compositions of the column simultaneously. These models demonstrated excellent performances, as supported by lower mean square error values of 2.73E-05 and 1.26E-04 for the forward and inverse models. Then, the proposed control method was applied to control both set point changes and disturbance changes, and in each case, it is contrasted with the traditional proportional-integral-derivative (PID) technique. The integrated absolute and integrated square errors for the proposed control were 0.1603 and 2.32E-03, respectively, while for PID control, these performance indexes were 0.7822 and 8.67E-03, respectively, for set-point tracking in top composition. Similarly, for bottom composition in tracking set points and rejecting disturbance, these performance indexes were also very less for the neural network control scheme.

A Neural Network Controller Design for the Mecanum Wheel Mobile Robot

Advanced controllers are an excellent choice for the trajectory tracking problem of Wheeled Mobile Robots (WMRs). However, these controllers pose a challenge to the hardware structure of WMRs due to the controller's complex structure and the large number of calculations needed. In that context, designing a controller with a simple structure and a small number of computations but good real-time performance is necessary in order to improve the tracking accuracy for the WMRs without requiring high hardware architecture. In this work, a neural network controller with a simple structure for the trajectory-tracking of a Mecanum-Wheel Mobile robot (MWMR) based on a reference controller is proposed. A two-layer feedforward neural network is designed as a tracking controller for the robot. The neural network is trained with a sample input-output data set so that the error between the neural network output and the reference control signal of the supervisory controller is minimal. The neural network parameters are trained to update over time. The simulation results verified the effectiveness of the neural network controller, whose parameters are tuned adaptively to ensure a fast convergence to the desired Bézier trajectory.

EEG-Based Emotion Recognition via Neural Architecture Search

With the flourishing development of deep learning (DL) and the convolution neural network (CNN), electroencephalogram-based (EEG) emotion recognition is occupying an increasingly crucial part in the field of brain-computer interface (BCI). However, currently employed architectures have mostly been designed manually by human experts, which is a time-consuming and labor-intensive process. In this paper, we proposed a novel neural architecture search (NAS) framework based on reinforcement learning (RL) for EEG-based emotion recognition, which can automatically design network architectures. The proposed NAS mainly contains three parts: search strategy, search space, and evaluation strategy. During the search process, a recurrent network (RNN) controller is used to select the optimal network structure in the search space. We trained the controller with RL to maximize the expected reward of the generated models on a validation set and force parameter sharing among the models. We evaluated the performance of NAS on the DEAP and DREAMER dataset. On the DEAP dataset, the average accuracies reached 97.94%, 97.74%, and 97.82% on arousal, valence, and dominance respectively. On the DREAMER dataset, average accuracies reached 96.62%, 96.29% and 96.61% on arousal, valence, and dominance, respectively. The experimental results demonstrated that the proposed NAS outperforms the state-of-the-art CNN-based methods.