Neural Network-based Control Research Articles

Compensation of electrical current drift in human-robot collision.

Human-robot collaborative systems are being increasingly adopted in manufacturing environments due to their application flexibility, adaptability, and cost-effectiveness. The majority of robotic systems use electrical current sensors to measure joint torque in industrial robot arms and limit the robot's impact in the event of an unanticipated acceleration/deceleration, such as in the event of a collision with a human operator. However, these electrical current sensors are known to experience sensor drift, which results in measurement inaccuracy that can result in improper joint-torque or end-effector force readings. This paper provides a method to compensate for electrical current drift using a neural network-based controller to control robot velocity. To evaluate the compensation method, an experimental setup was developed where a robot joint collided with the biofidelic test device that mimics the deformation response of the human forearm while simultaneously measuring deformation and contact force using an embedded soft force sensor. The proposed method was shown to compensate for electrical current drift and therefore reduce resulting contact forces between the robot and the biofidelic test device. Hence, this research provides a method to quantify the behavior of electrical current sensor drift on human-robot collision and presents a data-driven methodology for compensation.

Neural network-based control for RRP-based networked systems under DoS attacks with power interval

This paper is concerned with the optimal control problem for networked systems with the round-robin protocol (RRP) under the denial-of-service (DoS) attacker with power interval. In the literature, the control design is studied for the linear networked system subject to the DoS attacks with a known constant power or a known constant probability of data-packet dropouts. In this paper, the objective is to control the unknown nonlinear system in the communication network subject to the DoS attacks with a power interval and a time-varying probability of data-packet dropouts. The effects of the DoS attacks with power interval on a networked system under RRP are modeled accurately. A neural network (NN) -based observer is designed for the nonlinear system under the DoS attacks with power interval, and the relationship between the DoS attacker power interval and state estimation error is obtained. The NN actor–critic policy for the optimal control of the RRP-based networked system under the DoS attacks with power interval is found with adaptive dynamical programming, and stability of the resulting control system is analyzed. The proposed control method is demonstrated by a networked uninterruptible power supply system.

A new control strategy for harmonic reduction in photovoltaic inverters inspired by the autonomous nervous system

Abstract This paper proposes a new inverter control strategy whose main purpose is to reduce the current harmonic distortion resulting from unnecessary control actions without sacrificing the system’s dynamic response. The brain’s capabilities to learn and react to stress are mimicked to generate control actions based on emotional cues. The model is based on the brain emotional learning based intelligent controller, to which an autonomous nervous system was added. The modified controller aims at separating the strategy during transient states from the one during steady states. The proposed method was compared to the PI controller, the PR controller, and a neural network-based controller on Matlab Simulink. It shows major improvements in terms of harmonic distortion and a complete removal of the inter-harmonics. It provides a good dynamic response in transient states and an immunity to irrelevant signal variations during the steady state, which results in an improvement in the harmonic production.

Optimization algorithm of an artificial neural network-based controller and simulation method for animated virtual idol characters

Optimization algorithm of an artificial neural network-based controller and simulation method for animated virtual idol characters

Adaptive neural network-based trajectory tracking outer loop control for a quadrotor

This manuscript introduces a novel adaptive neural network-based controller for trajectory tracking of quadrotors. This controller is conceived as an outer loop controller that interacts with an inner loop controller in a two-loop configuration. The inner loop in this two-loop configuration is assumed to be inaccessible and unmodifiable, which is a realistic hypothesis in the operation of commercial quadrotors. Under this situation, the proposed controller computes appropriate kinematic input commands for the inner loop to achieve trajectory tracking. One remarkable feature of the proposed algorithm is its robustness against parametric uncertainties from the inner loop. An exhaustive error convergence analysis is provided, thus guaranteeing the convergence of the trajectory tracking error. Experimental results and a comparison using other control schemes demonstrate the competitiveness of the proposed scheme, being the latter the best among the tested adaptive neural network-based schemes.

Neural Network-Based Control for Wind Turbine System with PMSG: A Case Study for Hafar Al-Batin

Neural Network-Based Control for Wind Turbine System with PMSG: A Case Study for Hafar Al-Batin

Vibration and Trajectory Tracking Control of Engineering Mechanical Arm Based on Neural Network.

We offer a neural network-based control method to control the vibration of the engineering mechanical arm and the trajectory in order to solve the problem of large errors in tracking the path when the engineering mechanical arm is unstable and under the influence of the outside world. A mechanical arm network is used to perform tasks related to learning the unknown dynamic properties of a engineering mechanical arms keyboard without the need for prior learning. Given the dynamic equations of the engineering mechanical arm, the dynamic properties of the mechanical arm were studied using a positive feedback network. The adaptive neural network management system was developed, and the stability and integrity of the closed-loop system were proved by Lyapunov's function. Engineering mechanical arm motion trajectory control errors were modeled and validated in the Matlab/Simulink environment. The simulation results show that the management of the adaptive neural network is able to better control the desired path of the engineering mechanical arm in the presence of external interference, and the fluctuation range of input torque is small. The PID control has a large error in the expected trajectory tracking of the engineering mechanical arm, the fluctuation range of the input torque is as high as 20, and the jitter phenomenon is more serious. The use of detailed comparisons and adaptive neural network monitoring can perform well in manipulating the trajectory of the engineering mechanical arm. The engineering mechanical arm uses an adaptive neural network control method, in which the control precision of engineering mechanical arm motion trajectory can be improved and the out-of-control phenomenon of mechanical arm motion can be reduced.

Appointed-time prescribed performance control for 6-DOF spacecraft rendezvous and docking operations under input saturation

This study investigates the adaptive appointed-time control problem of chaser spacecraft rendezvous and docking with unknown target spacecraft information, parametric uncertainties, external disturbances, input saturation and performance constraints. Two kinds of novel prescribed performance function are first proposed to relax the limitation of the initial values of state signals and the contradiction between input saturation and performance constraints. Using dynamic surface control theory and a novel compensation system, an adaptive anti-saturation appointed-time prescribed performance controller and an adaptive anti-saturation appointed-time neural network-based prescribed performance controller are presented. An auxiliary variable is used to compensate for input saturation. Considering the strong approximation of the radial basis function neural network, an appointed-time neural network-based control scheme is proposed to compensate for external disturbances and parametric uncertainties. All closed-loop states converge to a neighborhood of zero in appointed-time, and the performance constraints of the relative position as well as the relative attitude are satisfied. A number of simulation scenarios are considered to verify the effectiveness of the proposed control strategies.

Agilicious: Open-source and open-hardware agile quadrotor for vision-based flight.

Autonomous, agile quadrotor flight raises fundamental challenges for robotics research in terms of perception, planning, learning, and control. A versatile and standardized platform is needed to accelerate research and let practitioners focus on the core problems. To this end, we present Agilicious, a codesigned hardware and software framework tailored to autonomous, agile quadrotor flight. It is completely open source and open hardware and supports both model-based and neural network-based controllers. Also, it provides high thrust-to-weight and torque-to-inertia ratios for agility, onboard vision sensors, graphics processing unit (GPU)-accelerated compute hardware for real-time perception and neural network inference, a real-time flight controller, and a versatile software stack. In contrast to existing frameworks, Agilicious offers a unique combination of flexible software stack and high-performance hardware. We compare Agilicious with prior works and demonstrate it on different agile tasks, using both model-based and neural network-based controllers. Our demonstrators include trajectory tracking at up to 5g and 70 kilometers per hour in a motion capture system, and vision-based acrobatic flight and obstacle avoidance in both structured and unstructured environments using solely onboard perception. Last, we demonstrate its use for hardware-in-the-loop simulation in virtual reality environments. Because of its versatility, we believe that Agilicious supports the next generation of scientific and industrial quadrotor research.

A neural network-based control chart for monitoring and interpreting autocorrelated multivariate processes using layer-wise relevance propagation

Recent advances in sensing and information technology are enabling multivariate sensory data from industrial equipment to be collected at a high sampling frequency. The resulting data streams often exhibit strong autocorrelation. In this study, we propose a neural network-based residual control chart to monitor the autocorrelated multivariate processes for real-time detection of abnormal equipment performance. Furthermore, we propose an interpretation method, based on layer-wise relevance propagation (LRP), to identify the responsible variable when an out-of-control condition is detected by the control chart. We compare the proposed techniques with several existing methods. Numerical studies and a real-data application demonstrate that the proposed method has good monitoring performance and can provide effective interpretation results. The proposed methods can be applied broadly for system condition monitoring and change detection in a variety of industrial systems applications.

Bioinspired Central Pattern Generator and T-S Fuzzy Neural Network-Based Control of a Robotic Manta for Depth and Heading Tracking

Aiming at the difficult problem of motion control of robotic manta with pectoral fin flexible deformation, this paper proposes a control scheme that combines the bioinspired Central Pattern Generator (CPG) and T-S Fuzzy neural network (NN)-based control. An improved CPG drive network is presented for the multi-stage fin structure of the robotic manta. Considering the unknown dynamics and the external environmental disturbances, a sensor-based classic T-S Fuzzy NN controller is designed for heading and depth control. Finally, a pool test demonstrates the effectiveness and robustness of the proposed controller: the robotic manta can track the depth and heading with an error of ±6 cm and ±6°, satisfying accuracy requirements.

Learning-based autonomous vascular guidewire navigation without human demonstration in the venous system of a porcine liver

PurposeThe navigation of endovascular guidewires is a dexterous task where physicians and patients can benefit from automation. Machine learning-based controllers are promising to help master this task. However, human-generated training data are scarce and resource-intensive to generate. We investigate if a neural network-based controller trained without human-generated data can learn human-like behaviors.MethodsWe trained and evaluated a neural network-based controller via deep reinforcement learning in a finite element simulation to navigate the venous system of a porcine liver without human-generated data. The behavior is compared to manual expert navigation, and real-world transferability is evaluated.ResultsThe controller achieves a success rate of 100% in simulation. The controller applies a wiggling behavior, where the guidewire tip is continuously rotated alternately clockwise and counterclockwise like the human expert applies. In the ex vivo porcine liver, the success rate drops to 30%, because either the wrong branch is probed, or the guidewire becomes entangled.ConclusionIn this work, we prove that a learning-based controller is capable of learning human-like guidewire navigation behavior without human-generated data, therefore, mitigating the requirement to produce resource-intensive human-generated training data. Limitations are the restriction to one vessel geometry, the neglected safeness of navigation, and the reduced transferability to the real world.

Research on Finite Element Structure of Vehicle Suspension Control Arm Based on Neural Network Sensing Control

To improve the adaptive control of the neural network under the influence of vehicle suspension control, the neural network control method is proposed. The specific content of the method analyzes the nonlinear properties of vehicle suspensions, proposes neural network-based adaptive control strategies, and develops neural network-based nonlinear algorithms and neural identifiers. Genetic algorithms perform predictive control of rear suspension through a compensation network. The experimental results show that the model structure is order n = m = 2 , the AN1 network node is 4-6-1, the AN2 network node is 5-4-1, the AN3 network node is 6-4-1, and the learning correction rate is α = 0.90 . In the actual simulation calculation, the number of nodes in the hidden layer of the network is increased, and the minimum number of nodes is chosen to determine the structure of the network, since the control effect obtained is not fundamentally changed. The suspension, which is controlled by the neural network’s adaptive control, has a vibration-reducing effect and is more effective by increasing the control of the rear suspension. The neural network has been shown to be able to effectively control the vehicle’s control arm.

Feedforward–Feedback Controller Based on a Trained Quaternion Neural Network Using a Generalised HR Calculus with Application to Trajectory Control of a Three-Link Robot Manipulator

This study derives a learning algorithm for a quaternion neural network using the steepest descent method extended to quaternion numbers. This applies the generalised Hamiltonian–Real calculus to obtain derivatives of a real–valued cost function concerning quaternion variables and designs a feedback–feedforward controller as a control system application using such a network. The quaternion neural network is trained in real-time by introducing a feedback error learning framework to the controller. Thus, the quaternion neural network-based controller functions as an adaptive-type controller. The designed controller is applied to the control problem of a three-link robot manipulator, with the control task of making the robot manipulator’s end effector follow a desired trajectory in the Cartesian space. Computational experiments are conducted to investigate the learning capability and the characteristics of the quaternion neural network used in the controller. The experimental results confirm the feasibility of using the derived learning algorithm based on the generalised Hamiltonian–Real calculus to train the quaternion neural network and the availability of such a network for a control systems application.

Optimization and Power Management of Solar PV-based Integrated Energy System for Distributed Green Hydrogen Production

Photovoltaic-based integrated energy systems act as a possible modern technological solution for clean and affordable green hydrogen production. However, research attempts are required from the scientific community to develop power management, control, optimization algorithms, and new techniques for these integrated energy systems effective and economical operation. The integrated energy system considered in this work consists of a solar photovoltaic, battery, grid and proton exchange membrane (PEM) electrolyzer. PEM electrolyzer with a power rating of 100 kW is modelled as a controlled current sink to interfere with the DC bus directly. This work proposes a power management and control algorithm for the photovoltaic-based integrated energy system considering different parameters like availability of solar photovoltaic power, DC link voltage, battery state of charge (SOC), tariff and availability of grid power. Effective power-sharing among the different power sources increases system reliability and stability. Optimization is performed for optimal sizing and costing to achieve economical operation such that these photovoltaic-based integrated energy systems become affordable and are encouraged for wide use. The photovoltaic system operates at the maximum power point through a neural network-based control strategy; in addition to this, stacking in neural network topologies is also discussed for improving system accuracy in tracking the reference voltage for MPPT operation. The DC link voltage is controlled by interfacing the battery with the DC link via a bidirectional buck-boost power converter. The contribution of the work is to highlight the role of renewable energy sources for continuous green hydrogen production. Green hydrogen will be a critical driving force for the future transportation system, fuel cell-based power generation, and industries utilizing hydrogen in direct or indirect forms-that will help transition towards a carbon-free society. Optimization and time-domain simulation results indicated the economic and technical feasibility of the proposed photovoltaic-based integrated energy system for green hydrogen production with the best optimal configuration producing hydrogen at the cost of $ 4.806/kg and total net present cost of $749904.

Enhanced Oblique Decision Tree Enabled Policy Extraction for Deep Reinforcement Learning in Power System Emergency Control

Deep reinforcement learning (DRL) algorithms have successfully solved many challenging problems in various power system control scenarios. However, their decision-making process is usually regarded as black-boxes. Furthermore, how DRL models interact with human intelligence remains an open problem. Thus, this paper proposes a policy extraction framework to extract a complex DRL model into an explainable policy. This framework includes three parts: 1) DRL training and data generation. We train an agent for a specific control task and generate data, which contains the control policy of the agent. 2) Policy extraction. We propose an information gain rate based weighted oblique decision tree (IGR-WODT) for DRL policy extraction. 3) Policy evaluation. We define three metrics to evaluate the performance of the proposed approach. A case study for the under-voltage load shedding problem shows that the IGR-WODT presents a performance enhancement compared with DRL, weighted oblique decision tree, and univariate decision tree. The proposed policy extraction method could provide an intuitive explanation of the neural network decision-making process to the dispatchers when making final decisions on power grid operation. Also, the resulted rule-based controller could replace the deep neural network-based controller in many field edge devices with limited computing resources, providing comparable performance.

Design and Implementation of Deep Neural Network-Based Control for Automatic Parking Maneuver Process.

This article focuses on the design, test, and validation of a deep neural network (DNN)-based control scheme capable of predicting optimal motion commands for autonomous ground vehicles (AGVs) during the parking maneuver process. The proposed design utilizes a multilayer structure. In the first layer, a desensitized trajectory optimization method is iteratively performed to establish a set of time-optimal parking trajectories with the consideration of noise-perturbed initial configurations. Subsequently, by using the preplanned optimal parking trajectory data set, several DNNs are trained in order to learn the functional relationship between the system state-control actions in the second layer. To obtain further improvements regarding the DNN performances, a simple yet effective data aggregation approach is designed and applied. These trained DNNs are then utilized as the motion controllers to generate feedback actions in real time. Numerical results were executed to demonstrate the effectiveness and the real-time applicability of using the proposed control scheme to plan and steer the AGV parking maneuver. Experimental results were also provided to justify the algorithm performance in real-world implementations.

Automatic synthesizing multi-robot cooperation strategies based on Brain Storm Robotics

With the increasing task complexity and environmental uncertainty, it is hard to achieve adaptability and robustness by manual design methods for multi-robot cooperation tasks. Automatic synthesis approaches with trial and error mechanisms are getting more and more attention. By encoding the strategies to be designed as “ideas”, the newly proposed Brain Storm Robotics (BSR) framework can obtain the sufficiently good solutions for particular tasks after a series of operations on the ideas. However, the original BSR only shows designing the rule base for a fuzzy controller. This paper proposes an automatic design approach for neural network-based strategies for robotic swarms with the BSR framework to realize cooperative behaviors. Two design cases are studied: one is the direct strategy search for a swarm aggregation behavior; the other is synthesizing a backpropagation neural network-based controller for coordinated formation control, which has both evolution and learning characteristics. The results show that the proposed method can automatically find control strategies with scalability for multi-robot cooperation, which has the potential for further development.

Deterministic learning from neural control for a class of sampled-data nonlinear systems

This study investigates the deterministic learning and control issues for uncertain sampled-data nonlinear systems (SDNSs). The problem of how to acquire/learn knowledge from adaptive control for SDNSs with uncertain affine terms is studied. To be specific, an appropriate neural network-based (NNB) control strategy is first presented to ensure tracking performance. To further realize learning, the exponential stability (ES) of the integrated closed-loop system coupled with the estimation error of the NN weights is considered. As the uncertain affine term prevents learning from occurring, the integrated system is converted into a discrete linear time-varying (DLTV) perturbed system by employing the state conversion technique. Tracking convergence allows the persistent excitation condition (PEC) of the NNs to be established, which guarantees the ES of the integrated DLTV system. Thus, accurate modeling of closed-loop sampled dynamics is obtained. By reutilizing the experiential knowledge obtained, a knowledge-based controller is constructed for high-performance control. Finally, simulations are performed to verify the presented strategy.

SECURITY APPLICATION OF NEURAL NETWORKS UNDER THE INSPECTION OF NONLINEAR DYNAMIC SYSTEMS

Based on the improved BP neural network, this paper establishes an adaptive online controlling model and adopts the model to optimize the controlling accuracies in discrete nonlinear dynamic systems and inverted pendulum systems. To avoid the local minimum problem of the BP neural network’s objective function in the training process, this paper proposes a neural network training method based on the quasi-Newton method (BFGS) optimization algorithm. Compared with other control methods, the neural network-based inverted pendulum control method proposed in this paper has higher control accuracy. Through the control simulation of its power system uses a discrete control method and the control of the inverted pendulum model system, this paper verifies the validity and good control has significantly improved the control method.