Robotic airship mission path-following control based on ANN and human operator’s skill

A yawing controller based on artificial neural networks (ANN) and human operator's skill is presented for robotic airship mission path tracking. Firstly, consideration of the path tracking errors from the point of view of operators is presented. Then, a data acquisition system is designed to collect flight data under manual control. Thirdly, The processed flight data are used to train and validate a multilayer feedforward ANN offline. Lastly, the trained ANN is reconstructed in the flight control system for yawing control. The experimental results indicate that this solution is valid and the ANN controller is robust even with wind disturbance.

Attempts are being made to enhance the drive performance by intelligent control using fuzzy logic (FL) and neural network techniques. One of the frequently discussed applications of artificial intelligence in control is the replacement of a standard proportional plus integral (PI) speed controller with an FL or artificial neural network (ANN) speed controller. Regardless of all the work, it appears that a thorough comparison of the drive behavior under PI, FL, and ANN speed control is necessary. This article attempts to compare PI, fuzzy, and ANN controllers that are implemented in an embedded system for closed-loop speed control of DC drive fed by a buck-type DC–DC power converter. The PI controller is designed based on the small signal modeling of the system. The PI-like fuzzy controller structure is considered for comparison. Two ANN controllers are designed. One controller uses training data obtained from the simulation of a fuzzy controller and the other uses training data from the simulation of a PI controller. The performance of the controllers is studied for a variety of operating conditions, such as step change in speed command and step change in load torque. The parameters selected for the comparison are the steady-state error and the rise time of the response. It is shown that ANN speed controllers provide a superior speed response in terms of rise time and the steady-state error compared to PI and FL controllers. This advantage arises from the fact that the neural network has the property of generalization and the control surface of the neural controller is smooth. The designed neural network controller is simple, with three neurons only, and so it is best suited for embedded system implementation. It is also found that the ANN controller trained with the training data from a PI controller has a better response compared to the ANN controller trained with data from a fuzzy controller.

Inverter's controller directly determines the grid-integration quality of power conversion systems. To address complex operating conditions, e.g. grid voltage distortion, parameter variations and weak grid scenarios, this paper proposes a lightweight artificial neural network (ANN) controller with strong adaptability. In implementation, the ANN controller generates control signals while simultaneously optimizing its weights in real time using the gradient descent algorithm. Distinctively, the weight gradients are directly calculated using the loss function and weights, which compared to the error backpropagation method, significantly reduces computational complexity, and therefore achieves the computational lightweight feature. In addition, the ANN phase-locked loop (ANN-PLL) is constructed to provide phase alignment for current reference while enabling fully ANNbased inverter control architecture. In principle, the proposed ANN controller relies neither on an offline training dataset nor on the system model, and achieves adaptive weights adjustment in real time with minimal computational effort. Through physical experiments, the proposed lightweight ANN controller was compared with the sliding mode controller and the model-based ANN controller, verifying its superior performance under complex operating conditions, such as grid voltage distortion, input voltage variation, current reference variation, filter parameter variation, and extremely weak grid (short circuit ratio=1.09).

This paper includes an application of layered artificial neural network controller to study the load-frequency control problem in a power system. The control scheme guarantees that steady state error of frequencies and inadvertent interchange of tie-lines are maintained in a given tolerance limitation. The proposed control has been designed for a three-area interconnected power system such that two areas include steam turbines and the other area includes a hydro turbine. Only one artificial neural network (ANN) controller, which controls the inputs of each area in the power system together, is considered. In the study, a back propagation-through-time algorithm is used as a neural network learning rule. The performance of the power system is simulated by using a conventional integral controller and ANN controller, separately. By comparing the results for both cases, the performance of an ANN controller is better than conventional controllers.

Photovoltaic systems play a pivotal role in renewable energy initiatives. To enhance the efficiency of solar panels amid changing environmental conditions, effective Maximum Power Point Tracking (MPPT) is essential. This study introduces an innovative control approach based on an Artificial Neural Network (ANN) controller tailored for photovoltaic systems. The aim is to elevate the precision and adaptability of MPPT, thereby improving solar energy harvesting. This research integrated an ANN controller into a photovoltaic system in order dynamically optimize the operating point of solar panels in response to environmental changes. The performance of the ANN controller was compared with traditional MPPT approaches using simulation in Simulink/Matlab. The results of the simulation showed that the ANN controller performed better than the traditional MPPT techniques, highlighting the effectiveness of this method for dynamically changing solar panel performance. The ANN particularly demonstrates higher precision and adaptability when environmental conditions vary. The strategy consistently achieves and maintains the maximum power point, enhancing overall energy harvesting efficiency. The integration of an ANN controller marks a significant advance in solar energy control. The study highlights the superiority of the ANN controller through rigorous simulations, demonstrating increased accuracy and adaptability. This approach not only proves effective, but also has the potential to outperform other MPPT strategies in terms of stability and responsiveness

In the field of autonomous robotics, enhancing the navigation system of robots is a crucial aspect that directly impacts their performance. This study presents a novel approach to addressing this challenge with an Artificial Neural Network (ANN) model. The research focuses on improving the navigation capabilities of a differential drive robot using the line-following method for route tracking and the dead reckoning technique for localization. It investigates a differential drive robot model controlled with a PID controller and derives the transfer function of the PID model. Through simulations, it becomes apparent that the PID model exhibits a continuous overshoot in its response, which negatively affects the behaviour of the robot's wheels. Ordinarily, continuous manual tuning will be required to correctly tune the PID controller to a value where the overshoots will be negligible, and this could be onerous. To overcome this limitation, an ANN controller is proposed, leveraging the learning capabilities of the neural network. Data from the PID controller transfer function is utilized to train the ANN model, enabling it to understand patterns and relationships. The ANN controller is then substituted in place of the PID controller in the simulation. The results showcase a remarkable 13.1% improvement in the robot's wheel response, highlighting the transformative potential of this approach for revolutionizing autonomous robot navigation in industrial applications. By using the transfer function of the PID model to train an ANN model, this study offers a powerful framework for enhancing the navigation performance of a differential drive autonomous robot and shows performance improvements in control, flexibility, and adaptation to changing conditions. These discoveries have significant ramifications for the industry and will pave the way for intelligent and effective autonomous robot navigation systems. The research provides a comprehensive understanding of the challenges associated with the differential drive robot model controlled with a PID controller and offers a robust approach to how this can be alleviated. The significance of this study lies in its ability to address the continuous overshoot issue observed in the PID controller's response by training an ANN controller with data from the PID controller. The proposed approach minimizes overshoot and improves the robot's wheel response, ultimately enhancing its navigation capabilities. Overall, this study demonstrates the potential of an ANN model to revolutionize autonomous robot navigation in industrial applications. The notable improvement achieved in the robot's wheel response validates the effectiveness of this approach. Future research can further optimize this integrated approach in real-world scenarios, leading to intelligent and efficient autonomous robot navigation systems across diverse industrial settings.

BACKGROUND: Currently, artificial intelligence in the form of artificial neural networks is being actively implemented in a number of areas of our lives, including medicine. In particular, in otorhinolaryngology, artificial neural networks are used to analyze images obtained during endoscopic examinations of patients (e.g., videolaryngoscopy) [1–3]. The interpretation of laryngoscopic images often presents significant difficulties for practicing physicians, which reduces the frequency of detection of precancerous laryngeal diseases and contributes to the increase in the number of patients with stage III–IV laryngeal cancer [4, 5]. This underscores the significance of prompt performance and accurate interpretation of the findings of endoscopic examinations of patients with laryngeal disorders. Artificial neural networks can be employed to analyze the results of videolaryngoscopy, furnishing the physician with supplementary information that can enhance diagnostic accuracy and diminish the probability of error [6, 7]. AIM: The study aims to develop and train an artificial neural network for recognizing characteristic features of laryngeal neoplasms and variants of laryngeal normality. MATERIALS AND METHODS: The study was conducted under the grant of the Moscow Center for Innovative Technologies in Healthcare (grant No. 2112-1/22) entitled “Using Neural Networks (Artificial Intelligence Algorithms) for Control and Improving the Quality of Diagnosis and Treatment of Diseases of Laryngeal and Ear Structures through Digital Technologies”.The following methods were used during the course of the study: data collection for the creation of a photobank (dataset) of medical images obtained during videolaryngoscopy; data partitioning for the formation of datasets for individual nosologies and groups of diseases; the method of consilium; analysis of the accuracy of recognition and classification of digital endoscopic images; and training of classification neural networks. Consequently, a dataset comprising 1,471 laryngeal images in digital formats (JPEG, BMP) was assembled, labelled, and uploaded for the purpose of training the artificial neural network. Of the total number of images, 410 were classified as pertaining to laryngeal formation, while 1061 were classified as variants of normality. Subsequently, the neural network was trained and tested to identify the signs of normal and laryngeal masses. RESULTS: The results of the testing of the artificial neural network indicated the formation of an inaccuracy matrix, the calculation of the value of recognition accuracy, the calculation of the quality indicators of the model performance, and the construction of the ROC curve. The developed and trained artificial neural network demonstrated an accuracy of 86% in recognizing the signs of laryngeal masses and norms. CONCLUSIONS: This study demonstrates that a trained artificial neural network can successfully distinguish between signs of normal and laryngeal masses in endoscopic photographs. With further training of the neural network and achievement of high accuracy, this technology can be used in clinical practice as an assistant in the interpretation of laryngoscopic images and early diagnosis of laryngeal masses. It can also be employed to control and improve the quality of diagnosis and treatment of diseases of the throat, nose, and ears by primary care physicians.

A review on neural networks with random weights

One of the most used artificial intelligence techniques for maximum power point tracking is artificial neural networks. In order to achieve successful results in maximum power point tracking, the training process of artificial neural networks is important. Metaheuristic algorithms are used extensively in the literature for neural network training. An important group of metaheuristic algorithms is swarm-intelligent-based optimization algorithms. In this study, feed-forward neural network training is carried out for maximum power point tracking by using 13 swarm-intelligent-based optimization algorithms. These algorithms are artificial bee colony, butterfly optimization, cuckoo search, chicken swarm optimization, dragonfly algorithm, firefly algorithm, grasshopper optimization algorithm, krill herd algorithm, particle swarm optimization, salp swarm algorithm, selfish herd optimizer, tunicate swarm algorithm, and tuna swarm optimization. Mean squared error is used as the error metric, and the performances of the algorithms in different network structures are evaluated. Considering the results, a success ranking score is obtained for each algorithm. The three most successful algorithms in both training and testing processes are the firefly algorithm, selfish herd optimizer, and grasshopper optimization algorithm, respectively. The training error values obtained with these algorithms are 4.5 × 10−4, 1.6 × 10−3, and 2.3 × 10−3, respectively. The test error values are 4.6 × 10−4, 1.6 × 10−3, and 2.4 × 10−3, respectively. With these algorithms, effective results have been achieved in a low number of evaluations. In addition to these three algorithms, other algorithms have also achieved mostly acceptable results. This shows that the related algorithms are generally successful ANFIS training algorithms for maximum power point tracking.

The present paper investigates an advanced control system of the wind energy conversion using an Artificial Neural Network (ANN) controller, which is developed in two modes: the offline mode, required for testing different sets of neural network parameters to find the optimal neural network controller (structure, activation function, and training algorithm), and the online mode where the optimal ANN controller is used in the five-phase PMSG based wind turbine system. The required data to generate the ANN model is obtained from a conventional PI controller. The controller inputs are the error between reference and actual speed (e), and the error change (de = e(k)-e(k-1)). As for the ANN output, it is the decreasing or increasing electromagnetic torque (dTe*). The studied neural network controller has been tested and validated using different wind speed values. Results and analysis are presented, and the contribution has been demonstrated.

Artificial neural networks (ANNs) are used to solve many problems, such as modeling, identification, prediction, and classification. The success of ANN is directly related to the training process. Meta-heuristic algorithms are used extensively for ANN training. Within the scope of this study, a feed-forward artificial neural network (FFNN) is trained using the marine predators algorithm (MPA), one of the current meta-heuristic algorithms. Namely, this study is aimed to evaluate the performance of MPA in ANN training in detail. Identification/modeling of nonlinear systems is chosen as the problem. Six nonlinear systems are used in the applications. Some of them are static, and some are dynamic. Mean squared error (MSE) is utilized as the error metric. Effective training and testing results were obtained using MPA. The best mean error values obtained for six nonlinear systems are 2.3 × 10−4, 1.8 × 10−3, 1.0 × 10−4, 1.0 × 10−4, 1.2 × 10−5, and 2.5 × 10−4. The performance of MPA is compared with 16 meta-heuristic algorithms. The results have shown that the performance of MPA is better than other algorithms in ANN training for the identification of nonlinear systems.

Traditional neural network training is usually based on the maximum likelihood to obtain the appropriate network parameters including weights and biases given the training data. However, if the available data are finite and noisy, the maximum likelihood-based network training can cause the neural network after being trained to overfit the noisy data. This problem has been overcome by using the Bayesian inference applied to the neural network training in various applications. The Bayesian inference can allow values of regularization parameters to be found using only the training data. In addition, the Bayesian approach also allows different models (e.g., neural networks with different numbers of hidden units to be compared using only the training data). Neural networks trained with Bayesian inference can be also known as Bayesian neural networks (BNNs). This chapter focuses on BNNs for classification problems with considerations of model complexity of BNNs conveniently handled by a method known as the evidence framework.

Due to its simple structure and robustness, the traditional proportional-integral-derivative (PID) controller is commonly used in the field of industrial automation and process control, but it does not function well with nonlinear systems, time-delayed linear systems and time-varying systems. A new type of PID controller based on artificial neural networks and evolutionary algorithms is presented in this paper. An powerful instrument for a highly nonlinear system is the Artificial Neural Network. The interest in the study of the nonlinear system has increased through the implementation of a high-speed computer system,. In complex systems such as robotics and process control systems, the Neuro Control Algorithm is often applied. Systems of process management is also nonlinear and hard to control consistently.. This paper presents a comprehensive analysis in Which is offline trained by a multilayered feed forward back propagation neural network to act as a process control system controller, That is to say, a temperature control device without prior knowledge of its dynamics. Via the implementation of a range of input vectors to the neural network, the inverse dynamics model is developed. Based on these input vectors, the output of the neural network It is being studied by explicitly configuring it to monitor the operation. In this paper, based on set-point adjustment, impact of disturbances in load and variable dead time, compassion between the PID controller and ANN is conducted. The outcome shows that ANN outperforms the controller of the PID.

This paper aims at improving the excitation control system for the doubly fed induction generator-based wind farm (WF). The proposed decentralized coordinated neural control is consisted of two controllers, i.e., a multiple model predictive controller (MPC) and an artificial neural network (ANN) controller. The MPC is based on the interaction measurement model and designed to guarantee the performance of system-wide optimization around the chosen operating points, while the ANN controller is used to calculate the weights of MPC controllers and trained to guarantee the control performance over the full operating region. The proposed control strategy allows to reduce the computational cost, and at the same time to take advantage of the distributed nature of the power system. Moreover, a simple, generic hybrid power system is used to demonstrate transient stability contributions. The problems of large-scale WF integration and its impact on transient stabilities of power systems are discussed through the eigenvalue analysis and time domain simulation. The results show that system damping and transient stability of hybrid power system are considerably improved.

In this paper, an artificial neural network (ANN) controller based position sensorless control of permanent magnet synchronous motor (PMSM) using Space vector modulation-Direct torque control (SVM-DTC) for variable speed drive has been presented. The SVM-DTC require the initial position of the rotor during the starting of the PMSM drive. The installation of the shaft-mounted position sensor requires additional space, assembly, wiring circuit, and is fragile component. The speed sensor-less control of PMSM enhances the performance of drives in harsh environments and reduces the overall cost of the drive and improve mechanical reliability. The speed estimation requires the knowledge of drive parameters, the model-based speed control technique is suitable for low and medium-speed motor drive applications without knowing the exact parameter of the PMSM drive. The Rotor Flux based Model Reference adaptive system (RF-MRAS) is used for a wide speed operation and estimates rotor angle in dynamic conditions. The presence of an integrator in the voltage model of RF-MRAS affects the low speed performance of the drive, therefore to improve the speed response at low speed, the ANN controller is used to replace the Proportional-Integral (PI) controller, which is employed in the adaptive model of the speed observer. The performance of the control scheme is simulated at variable speed and load conditions with the help of the OPAL-RT 4500 simulation platform.