Neural Network Controller for Attitude Control of Quadrotor

This paper proposes the neural network (NN) predictive controller that combines the advantages of NN and predictive control for the automatic voltage regulator (AVR). The NN predictive controller is suggested as a new intelligence controller rather than the conventional controllers for the AVR. This is the first application of the NN predictive controller for AVR. There are five parameters of the NN predictive controller which need a proper tuning to get a good performance by using the NN predictive controller. In recent papers, the parameters of NN predictive controller are typically set by trial and error or by the designer’s expertise. The imperialist competitive algorithm (ICA) is introduced in this paper as a new artificial intelligence technique instead of the trial-and-error or the designer’s expertise methods to get the optimal parameters of NN predictive controller in order to overcome the deviations of the voltage. The performance of the designed NN predictive controller based on the ICA is compared with the designed NN predictive controller based on the genetic algorithm and the conventional proportional–integral–derivative controller based on Ziegler–Nichols technique. The comparison emphasizes the superiority of the suggested NN predictive controller based on the ICA.

The precise position control for electro-hydraulic servo system (EHSS) is a key technical problem for modern industry control. This paper presents a model reference adaptive control (MRAC) strategy combining fractional order calculus theory and neural network to solve the high nonlinearity and parameter uncertainties in the EHSS. The reference model in the MRAC strategy is a fractional second-order model and the controller with adaptive law is designed by a Wiener recurrent neural network (WRNN). Simulations indicate the validity and superiority of the proposed fractional MRAC strategy.

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.

The article explores modern neural network architectures for the automatic detection and recognition of marine surface objects and obstacles of given classes throughout the full image area, applicable for execution in real or near real time on an optoelectronic vision system to au-tomate and improve the safety of civil marine navigation. A formal statement of the problem of automatic detection of objects on images is given. The state-of-the-art algorithms for detecting objects in images based on use of artificial convolutional neural networks were reviewed, their comparison was made and a reasonable choice was made in favor of the most efficient neuralnetwork architecture in terms of computational complexity to recognition accuracy. The subject area is studied, as well as publicly available databases of surface objects suitable for use in the training of algorithms using artificial neural networks. The article concluded that there is insuffi-cient labeled data for training neural network algorithms, as a result of which the authors inde-pendently collected research images and video sequences, prepared and labeled the collected data containing surface marine objects and other obstacles that represent a navigation hazard for ships. Based on the selected neural network architecture, a new neural network algorithm for automatic full-frame detection and recognition of surface objects was developed, and an artificial neural network was trained using the prepared database of images of typical objects. The resulting algorithm was tested by the authors on a validation data set, the quality of its work was estimated using various metrics, and the algorithm’s performance was measured. Conclusions are made about the necessity to expand the collected database of images of typical marine objects, further steps are proposed to improve the accuracy of the developed software and algorithmic complex and its implementation to be used in a marine optoelectronic machine vision system for automa-tion and improving the safety of civil navigation.

Partial-state feedback multivariable MRAC and reduced-order designs

This paper investigates the use of neural networks (NNs) in conventional model reference adaptive control (MRAC) to control a nonlinear magnetic levitation system. In the conventional MRAC scheme, the controller is designed to realize plant output convergence to a reference model output based on a plant which is linear. This scheme is effectively for controlling linear plants with unknown parameters. However, using MRAC to control the nonlinear magnetic levitation system in real time is a difficult control problem. In this paper, we incorporate a NN in MRAC to overcome the problem. The control input is given by the sum of the output of the adaptive controller and the output of the NN. The NN is used to compensate the nonlinearity of the plant that is not taken into consideration in the conventional MRAC. From experiment results, it has been shown that the plant output can converge to the reference model output after using NN in MRAC.

Airway pressure control of respiratory mechanics is crucial for ventilated patients. However, respiratory mechanics are nonlinear and are varying considerably amongst individual patients. In addition nonlinear interactions of a number of controls such as PEEP, Vt,fl02 - just to name a few - complicate the situation for the respiratory therapist on intensive care units. In this paper an alternative approach to traditional control methods - a neural network predictive controller - is proposed to establish an adaptive control of the airway pressure according to the individual features of a mechanically ventilated patient. The experimental results show that this approach can make the ventilator output follow the command pressure accurately as well as successfully identify the individual nonlinear respiratory mechanics of severely sick patients.Control of Respiratory Mechanics with Artificial Neural Networks Hui Zhu College of Mechanical & Electric Eng.

The paper introduces an approach to determining the number of people in a room based on data from observations of air parameters using a multilayer neural network. Monitoring of air parameters, the ability to predict their values and manage ventilation systems are important to ensure a healthy and comfortable indoor environment. The purpose of the research is to develop mathematical models and methods of forecasting air parameters based on the results of real-time monitoring. Different approaches to pre-dicting air parameters and the number of people in rooms using mathematical models in the form of equations and artificial neural networks with different architectures and types of training functions are considered. The paper proposes an approach to forecasting with the help of a multilayer neural network, which allows taking into account various factors, the nature and quantitative values of which cannot be taken into account and cannot be added to the model parameters. The CO2 data together with the indoor occupancy data from the meters are used to train the neural network. In the future, a neural network trained on such data can be used to predict the number of people in a room based only on data from air quality analyzers. The issue of choosing the architecture of a multilayer neural network and its pa-rameters for solving the forecasting problem has been investigated. Neural network training is carried out by the method of error back propagation. To evaluate the forecast-ing quality of the model, the metrics of mean square error of forecast and mean absolute error of forecast are used. The Kruskal-Wallis criterion is used to take into account the results of forecasting on several samples. Based on the results of the computational ex-periment, the optimal network architecture is determined. The model successfully repro-duces the nature of the distribution of the predicted parameter, as it captures the com-plex nonlinear dependencies between the various factors of the model. Recommendations are given to improve the performance of the model.

The objective of this study is to compare the predictive ability of Bayesian regularization with Levenberg–Marquardt Artificial Neural Networks. To examine the best architecture of neural networks, the model was tested with one-, two-, three-, four-, and five-neuron architectures, respectively. MATLAB (2011a) was used for analyzing the Bayesian regularization and Levenberg–Marquardt learning algorithms. It is concluded that the Bayesian regularization training algorithm shows better performance than the Levenberg–Marquardt algorithm. The advantage of a Bayesian regularization artificial neural network is its ability to reveal potentially complex relationships, meaning it can be used in quantitative studies to provide a robust model.

Artificial neural networks (ANN) have been widely used in the field of data classification. Normally, training of neural network is applied with the traditional back propagation technique. As, this approach has various drawbacks, training of neural network is done with Particle Swarm Optimization (PSO). PSO has been widely used to solve the diverse kind of optimization problems. Population initialization performs a significant role in meta-heuristic algorithms. This paper describes a new initialization population approach Log Logistic termed as PSOLL-NN to create the initialization of the swarm. The proposed algorithm has been tested for weight optimization of feed forward neural network; and compared with back propagation Algorithm (BPA), standard PSO (PSO-NN), PSO initialized with Halton Sequence (PSOH-NN), Torus sequence (PSOT-NN) and Sobol sequence (PSOS-NN). The experimental results show that the proposed technique performed exceptionally better than the other traditional techniques. Moreover, the outcome of our work presents a foresight that how the proposed initialization technique can be used as an efficient alternative to standard training approaches for the data classification problems.

In this study, an approach for predicting the three-body abrasive wear behavior of unfilled and graphite filled carbon fabric reinforced epoxy composite using two modeling techniques - Taguchi analysis and artificial neural network are presented. A set of experiments were conducted using an orthogonal array based on Taguchi techniques to acquire data in a controlled manner. The results showed that the addition of graphite particulate into carbon epoxy composite led to a decrease in its abrasive wear resistance, and the wear loss increased with an increase in abrading distance and loads. To investigate the effect of control parameters on the wear behavior of the composites, an analysis of variance was performed, and the S/N ratio was calculated. The results found that the normal load had the highest physical as well as statistical influence on the abrasive wear of the composites followed by abrading distance and filler content. To predict the wear properties of composites as a function of testing conditions, 3-[5]1-1 neural network architecture with Levenberg Marquardt (LM) training algorithm was used. By comparing the correlations obtained by Taguchi regression analysis and artificial neural network with the experimental results it was found that the artificial neural network predicts the wear rate better than regression analysis. Therefore, a well-trained artificial neural network system can be very helpful in estimating the weight loss in the complex three-body abrasive wear situation of polymer composites.

NOVEL STAGING TOOL FOR LOCALIZED PROSTATE CANCER: A PILOT STUDY USING GENETIC ADAPTIVE NEURAL NETWORKS

Water consumption of plants is a key parameter for formulating irrigation system, and the precise prediction play a important role in improving the use efficiency of limited water resources. In this experiment, by using the method of artificial neural network and MATLAB DATA PROCESSING SYSTEM combined with the meteorological data of air temperature, relative air humidity, solar radiation, wind speed, soil water content and dew point temperature as the input variable, the author established the artificial neural network system to forecast the seedling water consumption of P.×euramericana cv."74/76", and through the experiments it has been examined that two neural network system models can be applied in forecasting water consumption of seedlings, and the average relative error of Back Propagation (BP) neural network prediction model was 0.07, the General Regression Neural Network (GRNN) prediction model was 0.05, moreover, the latter had good stability, while that of the former was poor. Therefore, we propose that GRNN model can be used in prediction of seedling water consumption. Furthermore, the maximum relative error of GRNN predication model was 0.106, the minimum relative error was 0.015. The GRNN model is superior to the BP neural network model that the former performs a higher forecasting accuracy with relatively shorter time consumption and faster speed in training.

This paper investigates the depth control of an unmanned underwater remotely operated vehicle (ROV) using neural network predictive control (NNPC). The NNPC is applied to control the depth of the ROV to improve the performances of system response in terms of overshoot. To assess the viability of the method, the system was simulated using MATLAB/Simulink by neural network predictive control toolbox. In this paper also investigates the number of data samples (1000, 5000 and 10,000) to train neural network. The simulation reveals that the NNPC has the better performance in terms of its response, but the execution time will be increased. The comparison between other controller such as conventional PI controller, Linear Quadratic Regulation (LQR) and fuzzy logic controller also covered in this paper where the main advantage of NNPC is the fastest system response on depth control.

This study presents establishment of multiple input prediction model for automotive coil spring fatigue life estimation to shorten automotive suspension design process. Automotive suspension design is a lengthy work where any changes of the design lead to repetition of the entire process. It was hypothesised that the established model could be used to predict the spring design fatigue life without using any strain measurements. To initiate this model establishment, five sets of strain and acceleration measurement across different road conditions were collected and used for validations. To include spring stiffness as a parameter, a quarter car model was generated to obtain the force time histories from spring and vertical vibration of vehicle mass. In addition, artificial road profiles of road classes “A” to “D” were also generated for the quarter car simulation. Through adjusting the spring stiffness in the quarter car model, the spring and vehicle responses were varied. The simulated force time histories were used to predict springs’ fatigue life while acceleration time histories were used to calculate ISO 2631 ride-related vertical vibration. Subsequently, multiple linear regression approach was applied to determine the relationship between vehicle body frequency, ISO 2631 ride-related vertical vibration and spring fatigue life. The obtained regression had shown significance to the spring fatigue life with coefficient of determination of 0.8320. Reciprocally, multiple linear regression models were also used to predict the ISO 2631 ride-related vertical vibration with a coefficient of determination at 0.8810 and mean squared error values below 0.3430. To optimise the prediction results, artificial neural network was trained for the fatigue and vibration prediction purposes. The architectures of the artificial neural network were designed in terms of number of neurons and hidden layers to achieve a higher coefficient of determination of 0.9926 and lower mean squared error of 0.0824. For vibration prediction, the vehicle body frequency and spring fatigue life has shown a significant coefficient of determination to the ISO 2631 weighted vertical vibration, reaching 0.9579 with mean squared error of 0.0004. Based on the experimental strain and acceleration results, the predicted fatigue lives of multiple linear regression models were correlated well with the experimental results with coefficient of determination value of 0.9275. Meanwhile, the maximum difference of vibration prediction to experimental value using multiple linear regression models was only 18%. For artificial neural network predictions, the fatigue lives were mostly distributed within 1:2 or 2:1 life correlation and vibration prediction results were within 12%. For a good prediction, the target correlation value was above 0.80 to demonstrate a good fitted curve and the difference below 20%. The trained artificial neural network has shown outstanding capability in fatigue life or ride-related vertical vibration predictions. In this research, the main novelty was the trained artificial neural network for spring fatigue life or ride-related vertical vibration predictions which serve to reduce some procedures of automotive suspension design. The outcome of this study can be used to provide a new knowledge towards the field of fatigue research as well as vehicle ride dynamics. This research contributes to automotive industries especially in suspension spring design where the analysis of fatigue and ride-related vibration are provided.