A New Method Investigation for Robotic Inverted Pendulum Movement and Control

Journal Journal Europeen des Systemes Automatises

DOI 10.18280/jesa.560616

Issue 56

Volume 6

The integration of sensor fusion techniques has played a crucial role in the advancement of the contemporary robot applications era. This is primarily due to the fact that numerous robot applications heavily rely on the amalgamation of data from multiple sensors, which capture information from the immediate surroundings. Examples of such applications include interactive virtual reality games, navigation systems, and fitness trackers. This study presents the implementation of a sensor fusion technique utilizing low-cost sensors, specifically a gyro sensor and a tilt sensor, to precisely estimate the balancing angle of an inverted pendulum robot system. Hence, this research study presents an innovative approach to address the issue of gyro drift, which has a detrimental impact on the precision of orientation calculations in the indirect Kalman filter-based sensor fusion. The emergence of kinematics and dynamics is observed in this particular process. The frequency responses of two distinct sensors, namely a gyro and a tilt sensor, have been subjected to analysis. A technique of sensor fusion has been utilized to counterbalance the inherent drift of the gyroscope sensor. The integration of a low-cost gyro sensor and a tilt sensor through the process of filtering facilitates the determination of the pendulum angle with efficacy, thereby eliminating the necessity of deploying more expensive sensors.

Read Publication

A New Method to Road Traffic Monitoring Using Artificial Systems

Journal AIP Conference Proceedings

DOI 10.1063/5.0167689

Issue 2839

Volume 1

Unmanned aerial vehicles (UAVs) are used in this research to demonstrate a method to traffic monitoring based on automated UAV situation management. Analysis of current techniques of on-board automated detection of abnormal traffic conditions and emergency using artificial vision systems (AVS) is part of the investigation, which also involves preliminary categorization of these events, including the allocation of crises and disasters. A variety of UAV controls are discussed in light of the current scenario. Based on Neyman-Pearson criteria and Bayes, the traffic scenario recognition approach presented in the study is used to identify traffic conditions. In addition, the study examines the current methods for detecting moving and stationary vehicles . as well, Image segmentation and machine learning technologies, such as Deep Learning, are used to recognize cars in this paper. Traffic scenarios may be described using solutions to vehicle tracking and velocity detection difficulties.

Read Publication

Neural Network-Directed Detection and Localization of Faults in Railway Track Circuits: An Application of Dempster-Shafer Theory

Journal Ingénierie des Systèmes d’Information

DOI 10.18280/isi.280306

Issue 28

Volume 3

Track circuits, integral to the security infrastructure of railway traffic systems, govern the operational status of train lines. The swift detection and rectification of faults within these circuits is critical to preserve the integrity and functionality of rail networks. In this study, an innovative approach, leveraging neural networks in tandem with Dempster-Shafer theory, is proposed for detecting and localizing faults in track circuits. The complexities of fault detection are deconstructed into more manageable, capacitor-specific pattern recognition challenges, with the resolutions amalgamated via Dempster-Shafer theory. Simulations demonstrate the efficacy of this method, yielding a detection accuracy exceeding 98% and a localization accuracy surpassing 93%. This marks a significant improvement over contemporary reference techniques, thereby setting a new benchmark in the domain of track circuit fault detection and localization.

Read Publication

Design of End-to-End Speech Recognition System by Used Hybrid CNN-LSTM

Journal AIP Conference Proceedings

DOI 10.1063/5.0167688

Issue 2839

Volume 1

In the study of speaker verification, end-to-end Speech Recognition systems based on (DNNs) have received a lot of attention. The study of audio signal processing in the disciplines of voice recognition and auto-tagging for music However, in light of the fact that we are not aware of any systems that use raw audio streams from end to end. Speaker verification has not looked into these issues is used as an example of a full method for verifying speakers from end to end has been suggested. It accepts unprocessed audio I/P and produces a verified audio O/P result. Preprocessing and an integrated speaker are both included in this setup. The focus was on feature extraction models. The idea that has been floated a pre-emphasis layer and a stride convolution were used together. The first two hidden levels should have a layer for pre-processing. In addition, convolutional models are used to extract speaker features. It has been hypothesized that the long-term short-term memory and layer incorporated in the end-to-end Speech Recognition system that is being suggested.

Read Publication

Enhancement and modification of automatic speaker verification by utilizing hidden Markov model

Journal Indonesian Journal of Electrical Engineering and Computer Science

DOI 10.11591/ijeecs.v27.i3.pp1397-1403

Issue 27

Volume 3

The purpose of this study is to discuss the design and implementation of autonomous surface vehicle (ASV) systems. There’s a lot riding on the advancement and improvement of ASV applications, especially given the benefits they provide over other biometric approaches. Modern speaker recognition systems rely on statistical models like hidden Markov model (HMM), support vector machine (SVM), artificial neural networks (ANN), generalized method of moments (GMM), and combined models to identify speakers. Using a French dataset, this study investigates the effectiveness of prompted text speaker verification. At a context-free, single mixed mono phony level, this study has been constructing a continuous speech system based on HMM. After that, suitable voice data is used to build the client and world models. In order to verify speakers, the text-dependent speaker verification system uses sentence HMM that have been concatenated for the key text. Normalized log-likelihood is determined from client model forced by Viterbi algorithm and world model, in the verification step as the difference between the log- likelihood. At long last, a method for figuring out the verification results is revealed.

Read Publication

Enhancement of motor speed identification using artificial neural networks

Journal Indonesian Journal of Electrical Engineering and Computer Science

DOI 10.11591/ijeecs.v27.i3.pp1388-1396

Issue 27

Volume 3

In this study have been utilized a modified version of ant colony optimization to improve the thresholds of neural networks and weights by including the rank- weight approach. Furthermore, this technique easily overcome the drawbacks speed up convergence into the minimum while training the back propagation neural network. The improved ant colony optimization-back propagation neural. not only has the capacity to map extensively, but it also enhances operating efficiency noticeably, according to the simulation findings. The simulation results revealed that the speed sensor replaced with the ant colony optimization rw-optimized back propagation neural network - speed identification and motor’s speed determined using this approach the result is satisfactory.

Read Publication

Reinforcement learning for speech recognition using recurrent neural networks

Journal 2022 2nd Asian Conference on Innovation in Technology, ASIANCON 2022

DOI 10.1109/ASIANCON55314.2022.9908930

Read Publication

Foretelling Diabetic Disease Using a Machine Learning Algorithms

Journal 2022 International Conference for Advancement in Technology, ICONAT 2022

DOI 10.1109/ICONAT53423.2022.9726070

Continuous monitoring and adjustment of insulin dosages are necessary for diabetics in order to maintain diabetics levels as near to normal levels possible. long-term and short-term complications might result from blood glucose levels that are out of the usual range. If a person's blood glucose levels were predicted automatically, they would be able to take preventative measures before they had a problem. Here, in this study provide a strategy that leverages a general To generate features for a (SVM) model trained on specific patient data sets, we used a physiological model of blood glucose dynamics. Almost a quarter of hypoglycemia incidents may be predicted 30 minutes in advance using a novel algorithm that beats diabetes specialists. There are now only 42 percent false alarms, but the vast majority of them occur in near-hypoglycemia regions, so patients who react to these hypoglycemic warnings would not be harmed by action.

Read Publication

Design of automatic speech recognition in noisy environments enhancement and modification

DOI 10.21533/pen.v10i1.2575

Issue 10

Volume 1

Recurrent neural networks (RNN) and feed-forward multi-layer perceptron’s have been proposed for determining the absence and presence of speech in continuous voice signals when there is a variety of background noise levels present. The Aurora2 and Aurora3 were used to conduct detailed performance evaluations on vocal activity detection. When a Recurrent neural network feeds on automatic speech recognition particular features and acoustic features, the best outcomes can be achieved, according to this study. Aurora2 and the French, Romanian and Norway portions of the Aurora3 corpus is also proposed for detailed studies of ASR. When noise presence probability is utilized to change for encoding speech, phone subsequent probabilities are employed; the WER is reduced by 10.3 percent.

Read Publication

Enhancement automatic speech recognition by deep neural networks

Journal Periodicals of Engineering and Natural Sciences

DOI 10.21533/pen.v9i4.2450

Issue 9

Volume 4

The performance of speech recognition tasks utilizing systems based on deep learning has improved dramatically in recent years by utilizing different deep designs and learning methodologies. A popular way to boosting the number of training data is called Data Augmentation (DA), and research shows that using DA is effective in teaching neural network models how to make invariant predictions. furthermore, EM approaches have piqued machine-learning researchers' attention as a means of improving classifier performance. In this study, have been presented a unique deep neural network speech recognition that employs both EM and DA approaches to improve the system's prediction accuracy. firstly, reveal an approach based on vocal tract length disturbance that already exists and then propose a Feature perturbation is an alternative Data Augmentation approach. in order to make amendment training data sets. This is followed by an integration of the posterior probabilities obtained from several DNN acoustic models trained on diverse datasets.

Read Publication

Design of rotary inverted pendulum swinging-up and stabilizing

Journal Periodicals of Engineering and Natural Sciences

DOI 10.21533/pen.v9i4.2453

Issue 9

Volume 4

The mechanical design of inverted pendulum systems may be very diverse. This study makes use of mathematics to explain the expanded unstable Rotary planer inverted pendulum arrangement (Rotary -inverted system). Pin connections are used in planer-inverted pendulums to attach the pendulum to the rotating-Rotary actuation base. This design of a Rotary -inverted pendulum is explored in the development of underactuated robotic systems because it best simulates the balance of a broomstick in the hand by treating the elbow and shoulder as revolute joints. It is necessary to utilize the Lagrangian equation of motion when creating the dynamical equation for the Rotary -inverted pendulum. MATLAB Simulink is used to simulate a nonlinear computational model of the system so as to test the accuracy of the mathematical model.

Read Publication

Classification of acoustic data using the ff neural network and random forest method

Journal 2021 International Conference on Smart Generation Computing, Communication and Networking, SMART GENCON 2021

DOI 10.1109/SMARTGENCON51891.2021.9645847

According to their acoustic analysis, speaker identification systems are intended to identify the speaker or group of speakers. Many methods are used to conduct acoustic analysis on a speech signal, with time and frequency domain analysis being the most common. The MFCC and Fundamental Frequency techniques are used in this article to extract acoustic information from speech samples. Two distinct methods, Random-forest and Feed Forward Neural Network, are used to classify the findings. The combination of the FFNN classifier with the acoustic model yielded a recognition accuracy of 91.4 percent. This paper makes use of the CMU ARCTIC Database.

Read Publication

Design Speech Recognition Systems in the nosily Environment by Utilizing intelligent Devices

Journal 2021 2nd International Conference on Smart Technologies in Computing, Electrical and Electronics, ICSTCEE 2021 - Proceedings

DOI 10.1109/ICSTCEE54422.2021.9708576

Many individuals have always found the ability to recognize human speech interesting because of the diversity of applications in virtually every industry. Improvements in human voice/speech recognition capacity and quality have been made possible via advancements in science and technology, particularly when using equipment known as a terminal. Speech recognition enables devices to alter speech data in a manner that is understandable, and this means that information has been completely identified and comprehended. The main aim of recognizing human voice is to be able to tailor information (for humans) for device use. The main goal of voice recognition systems is to allow the device to interact with the user and provide new possibilities. The growing use of intelligent terminals, as well as their substantial scientific and technological potential, calls into question the capabilities of human voice recognition in the workplace. The aim of this article is to demonstrate human speech recognition capabilities utilizing intelligent terminal devices. Also, this will primarily be investigated in the transportation (vehicle) context, where the advantages and disadvantages of these devices and concepts will be assessed.

Read Publication

Separately excited DC motor speed using ANN neural network

Journal AIP Conference Proceedings

DOI 10.1063/5.0068893

Issue 2404

Volume 1

Separately Excited Direct Current Motors (SEDCM) is termed by high efficiency in electrical traction manufacturing. SEDCMs are utilized by high power applications such as aircrafts and ships traction. Speed control is more likely required to maintain the performance of the motor in different load scenarios. Conventional methods of speed control such as proportional integral controller (PIC) are reported good performance in error tackling but it may consume longer time and computational cost. In this paper, computational speed controller is proposed which uses neural network to predict speed and then to produce the reference voltage in order to update the armature terminal voltage. In this study have been assumed a three layers neural network to implement the control on speed of motor and the model is outperformed over the conventional controller models.

Read Publication

mahmood Dynamic spectrum sharing is the best way to modify spectrum resources

Journal 2021 Asian Conference on Innovation in Technology, ASIANCON 2021

DOI 10.1109/ASIANCON51346.2021.9544912

As a result, spectrum agreement could not fulfill the increasing technological needs. But with cognitive radio, spectrum use and resource substitution have come together. A distressing number of people are now relying on technology such as radio to meet their daily needs, since technologies in all industries, especially those that employ licensed band, are moving toward greater availability. To combat the radio spectrum congestion, users may be allocated to different bands to balance out the frequency spectrum. To improve radio coverage, we have suggested a model in this study. This work is emphasizing the performance constraints that affect spectrum sharing. Various spectrum-sharing approaches are researched for time delay and throughput.

Read Publication

Optimization the accuracy of ffnn based speaker recognition system using pso algorithm

Journal International Journal on Communications Antenna and Propagation

DOI 10.15866/irecap.v11i4.19883

Issue 11

Volume 4

Speaker recognition systems use a model that learns a speaker's speech by inputting an audio recording and processing it. Time-varying signal, with frequencies that continuously change, is identified as a speech signal. There are many uncertain attributes to speech; thus traditional speech recognition techniques such as using zero crossings and the Fourier Transform are not up to the task. It aims to be accomplished with the aim of helping two causes. The first part is designed to address speaker identification technology that is resistant to noise. While most prior solutions have relied on changing mel frequency cepstrum coefficients, with a Fundamental frequency feature coefficient, this proposal integrates both of these modifications with a new cepstrum component. In order to construct the feature matrix, the system is fed with two-hundred and fifty speech imprints that are used to apply features extraction techniques. The matrix is used to teach the algorithm about features, and each one is then evaluated using incomplete data (thirty percent of total data in features matrix). Speaker recognition models with improved accuracy are developed by studying the algorithms invasively. These variables (metrics) are generated for each algorithm and applied to the algorithm for recognition accuracy and the time required to achieve that accuracy. When tested against previous research, the findings show that the Feed Forward Neural Network-based Particle Swarm Optimization method has been better. This model can accurately identify 96% of the input with less processing time. According to the findings, optimization utilizing advanced particle swarm optimization (a.k.a. Particle Swarm Optimization) is most likely responsible for the higher accuracy seen in speaker identification.

Read Publication

Design of multiband slot patch antennas for modern wireless applications

Journal International Journal on Communications Antenna and Propagation

DOI 10.15866/irecap.v10i5.19071

Issue 10

Volume 5

In the past few years, microstrip antenna have been used in many applications because of their thin planar profile that could be easily integrated into various surfaces of aircraft, rockets and other customer-beneficial products. They are easily integrated into the same board and they allow the addition of active devices. These antennas are used in different areas like satellite navigation, mobile communication, automobiles, internet services and radars. In this study, antennas of the Coplanar (CPW-Fed) system have been used. Multi-band microstrip antennas have been designed and simulated with the help of the HFSS program about the return loss (S11). The goal of this study is to obtain multiband for working on several frequencies. Microstrip patch slot antennas using Coplanar (CPW-Fed) and working on different frequencies have been designed.

Read Publication

The Impact of Relay Node Deployment in Vehicle Ad Hoc Network: Reachability Enhancement Approach

Journal 2019 Global Conference for Advancement in Technology, GCAT 2019

DOI 10.1109/GCAT47503.2019.8978445

Speaker VANET or vehicular ad hoc network is intelligent means of transportation which is essentially ensuring the safety norms. This technology can be established in any road such as highway or inter-city urban roads; it is basically a set of mobile vehicles connected with static base station. VANET is adopted ad hoc nature where dynamic vehicles are capable to communicate with each other due to less infrastructure network strategy as no clear administration tasks are functional. However, VANET is reported efficient in low and medium speeds i.e. 10 through 40 km/h. This paper aims to enhance network reachability in random traffic conditions.

Read Publication

Design of new multiband slot antennas for Wi-Fi devices

Journal International Journal on Communications Antenna and Propagation

DOI 10.15866/irecap.v9i5.16754

Issue 9

Volume 5

Microstrip antennas cover a wide range of application in different areas, such as internet services, mobile communication, and satellite navigation. Microstrip antenna consists of two metal layers and a substrate. In the two metal layers, the top one is called patch and the bottom layer is called ground. Between the patch and the ground, there is the substrate layer, which has a certain dielectric constant and thickness. The aim of this paper is to present a way of getting multi-band antennas for Wi-Fi devices. 2.5 GHz (2.5-2.69GHZ), 3.5 GHz (3.3-3.8GHz) and 5 GHz (5.25-5.85GHz) have been used for Wi-MAX applications. HFSS software has been used for simulation process. The multiband slot antennas have been simulated and manufactured. Multi-band microstrip antennas have been designed and simulated in terms of the return loss (S11) and the gain. These designs have been fabricated and their return loss has been measured and compared with the simulation results.

Read Publication

Optimization and Integration of RFID Navigation System by Using Different Location Algorithms

Journal International Review of Electrical Engineering

DOI 10.15866/iree.v14i4.16684

Issue 14

Volume 4

The detection of mobile objects locations is vital for many fields of technology such as banking systems, social media, and driving assistance technologies (i.e. Google Maps). This approach is firstly proposed for supporting security and arm forces. Then, location-based services have gained the attention of data scientist so, it has become essential for providing access for particular files of service in accordance with geographical area/countries restrictions. So-to-say, location-based services have become paramount in modern days technologies. These systems are more likely required large knowledge of signal processing, radio frequency and electromagnetic theories. Several estimation algorithms are deployed in this paper for evaluating the location of mobile object over small scale indoor environment. In order to tackle connectivity problems such as No Line of Sight (LoS), energy consumption and time delay; Radio-frequency Identification (RFID) system is used to satisfy the requirements of real-life systems. In this paper, three location optimization algorithms namely Triangulation, Weighted Centroid Localization (WCL) and Bayes Decision Rule (with Kalman filter integration) have been used to evaluate the location of that mobile object. Every algorithm is examined with efforts in order the accuracy of each one to detect the location. However, the performance of all the algorithms is compared; the outcomes of Kalman filter integration with Bayes Decision rule algorithm have been more accurate.

Read Publication

Optimization and Integration of RFID Navigation System by Using Different Location Algorithms

Journal International Review of Electrical Engineering

DOI 10.15866/iree.v14i4.16684

Issue 14

Volume 4

The detection of mobile objects locations is vital for many fields of technology such as banking systems, social media, and driving assistance technologies (i.e. Google Maps). This approach is firstly proposed for supporting security and arm forces. Then, location-based services have gained the attention of data scientist so, it has become essential for providing access for particular files of service in accordance with geographical area/countries restrictions. So-to-say, location-based services have become paramount in modern days technologies. These systems are more likely required large knowledge of signal processing, radio frequency and electromagnetic theories. Several estimation algorithms are deployed in this paper for evaluating the location of mobile object over small scale indoor environment. In order to tackle connectivity problems such as No Line of Sight (LoS), energy consumption and time delay; Radio-frequency Identification (RFID) system is used to satisfy the requirements of real-life systems. In this paper, three location optimization algorithms namely Triangulation, Weighted Centroid Localization (WCL) and Bayes Decision Rule (with Kalman filter integration) have been used to evaluate the location of that mobile object. Every algorithm is examined with efforts in order the accuracy of each one to detect the location. However, the performance of all the algorithms is compared; the outcomes of Kalman filter integration with Bayes Decision rule algorithm have been more accurate.

Read Publication

Radar Target Detection by Using Levenberg-Marquardt Algorithm

Journal Przegląd Elektrotechniczny

DOI 10.15199/48.2023.03.47

Issue 99

Volume 3

Typically, with radar systems, a person is required to assist in the process of detecting a target. As a result of this human factor, radar systems are not completely dependable since their performance differs across operators. In this work, an intelligent radar system for border monitoring is described. Artificial neural networks trained using the Levenberg-Marquardt technique have been used to identify and categorize targets automatically in the radar system under development. Inverse Synthetic Aperture Radar images captured with high resolution by the radar’s detecting module serve as both input and output data for the neural network. The simulation findings show that intelligent radar can identify various targets. Both human operators and a competing radar system were no match for the radar’s efficiency. These findings suggest that in the future, intelligent technologies may be able to take the position of human radar operators in high-risk security environments.