Pub Date : 2022-07-08DOI: 10.1109/CONECCT55679.2022.9865713
Alireza Keyanfar, Liyana Meh, Reihaneh Rabbani
Housing is not a static artifact and contributes to the physical and psychological state of the residents. The impact of the housing effect depends on the residents' quality of life and satisfaction. A review of housing journals indicates that optimal housing is not accessible for most low-income households due to the lack of financial support for tenants due to the landlords' absolute power, dwellers with physical or mental disabilities, and elderlies. Therefore, preferably, they look for a home that is as satisfying as possible. If unavoidable, a housing relocation might bring the housing situation more in line with their preferences. This paper proposes ways to obtain adaptive housing responsive to users' preferences using intelligent technologies to optimize the residents' quality of life. The work shows that implementing sensors for managing a healthy environment, smart carpet as an interface between the environment and the ventilation system, facial emotion recognition, intelligent weather detection, smart curtains, mobile apps, and adjusting interior lighting and color-changing may be valuable to offer a comfortable, user-friendly home environment and facilitates the house choosing for the user by considering their personalities and behavior patterns.
{"title":"Application of intelligence in solving architectural problems in the field of housing","authors":"Alireza Keyanfar, Liyana Meh, Reihaneh Rabbani","doi":"10.1109/CONECCT55679.2022.9865713","DOIUrl":"https://doi.org/10.1109/CONECCT55679.2022.9865713","url":null,"abstract":"Housing is not a static artifact and contributes to the physical and psychological state of the residents. The impact of the housing effect depends on the residents' quality of life and satisfaction. A review of housing journals indicates that optimal housing is not accessible for most low-income households due to the lack of financial support for tenants due to the landlords' absolute power, dwellers with physical or mental disabilities, and elderlies. Therefore, preferably, they look for a home that is as satisfying as possible. If unavoidable, a housing relocation might bring the housing situation more in line with their preferences. This paper proposes ways to obtain adaptive housing responsive to users' preferences using intelligent technologies to optimize the residents' quality of life. The work shows that implementing sensors for managing a healthy environment, smart carpet as an interface between the environment and the ventilation system, facial emotion recognition, intelligent weather detection, smart curtains, mobile apps, and adjusting interior lighting and color-changing may be valuable to offer a comfortable, user-friendly home environment and facilitates the house choosing for the user by considering their personalities and behavior patterns.","PeriodicalId":380005,"journal":{"name":"2022 IEEE International Conference on Electronics, Computing and Communication Technologies (CONECCT)","volume":"30 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125627682","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-07-08DOI: 10.1109/CONECCT55679.2022.9865720
G. Thiagarajan, Deepan Vetrivel, Sanjeev Gurugopinath
Matrix decomposition methods such as the Cholesky and the QR decomposition arise in several applications in signal processing for multiple-input, multiple-output (MIMO) communication systems. The computational complexity of regular Cholesky and QR solvers is known to be $mathcal{O}left( {{N^3}} right)$. To reduce this, several recursive algorithms at both column- and block-levels have been proposed in the literature. In this paper, we utilize one such recursive structure in Cholesky and QR decompositions for matrices with entries from the field of complex numbers, which results in a level of complexity reduction. The use of the considered techniques is discussed in the context of a MIMO decoder. In particular, the utility of proposed methods is illustrated in a MIMO successive interference cancellation based detector. Simulation results are provided to substantiate the performance of a detector under two different antenna and receiver configurations.
{"title":"Recursive Matrix Decomposition Methods and Applications in Wireless Communication","authors":"G. Thiagarajan, Deepan Vetrivel, Sanjeev Gurugopinath","doi":"10.1109/CONECCT55679.2022.9865720","DOIUrl":"https://doi.org/10.1109/CONECCT55679.2022.9865720","url":null,"abstract":"Matrix decomposition methods such as the Cholesky and the QR decomposition arise in several applications in signal processing for multiple-input, multiple-output (MIMO) communication systems. The computational complexity of regular Cholesky and QR solvers is known to be $mathcal{O}left( {{N^3}} right)$. To reduce this, several recursive algorithms at both column- and block-levels have been proposed in the literature. In this paper, we utilize one such recursive structure in Cholesky and QR decompositions for matrices with entries from the field of complex numbers, which results in a level of complexity reduction. The use of the considered techniques is discussed in the context of a MIMO decoder. In particular, the utility of proposed methods is illustrated in a MIMO successive interference cancellation based detector. Simulation results are provided to substantiate the performance of a detector under two different antenna and receiver configurations.","PeriodicalId":380005,"journal":{"name":"2022 IEEE International Conference on Electronics, Computing and Communication Technologies (CONECCT)","volume":"62 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134261127","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-07-08DOI: 10.1109/CONECCT55679.2022.9865709
A. Chandra, G. Singh, V. Pant
Due to the integration of distributed generation (DG), fault level of the microgrid changes significantly. The fault current behaves dynamically and inhibits the activities of conventional protection algorithms. To address this serious issue for protection, a simple and fast protection algorithm, based on Empirical Mode Decomposition (EMD) and Hilbert Transform (HT) is proposed in this paper for the protection of grid-connected 14 BUS looped microgrid system. This scheme estimates the energy difference of current signals retrieved from both ends of the feeder. Further, the fault detection signal is generated from the Hilbert spectral energy difference. Nonetheless, for the comparative analysis of the performance of proposed EMD based fault detection technique, Variational Mode Decomposition (VMD) is also performed. Further high resistance fault cases are also examined to verify effectiveness of this proposed scheme. As this technique is essentially reckoning on the energy difference of current signals, it does not suffer from the difficulties associated with dynamic current behaviour of a microgrid. This proposed system is simulated in PSCAD simulation software and the programming for signal analysis is performed in MATLAB.
{"title":"Fault Detection of Grid-Connected Looped Microgrid based on Estimated Energy Difference Signal","authors":"A. Chandra, G. Singh, V. Pant","doi":"10.1109/CONECCT55679.2022.9865709","DOIUrl":"https://doi.org/10.1109/CONECCT55679.2022.9865709","url":null,"abstract":"Due to the integration of distributed generation (DG), fault level of the microgrid changes significantly. The fault current behaves dynamically and inhibits the activities of conventional protection algorithms. To address this serious issue for protection, a simple and fast protection algorithm, based on Empirical Mode Decomposition (EMD) and Hilbert Transform (HT) is proposed in this paper for the protection of grid-connected 14 BUS looped microgrid system. This scheme estimates the energy difference of current signals retrieved from both ends of the feeder. Further, the fault detection signal is generated from the Hilbert spectral energy difference. Nonetheless, for the comparative analysis of the performance of proposed EMD based fault detection technique, Variational Mode Decomposition (VMD) is also performed. Further high resistance fault cases are also examined to verify effectiveness of this proposed scheme. As this technique is essentially reckoning on the energy difference of current signals, it does not suffer from the difficulties associated with dynamic current behaviour of a microgrid. This proposed system is simulated in PSCAD simulation software and the programming for signal analysis is performed in MATLAB.","PeriodicalId":380005,"journal":{"name":"2022 IEEE International Conference on Electronics, Computing and Communication Technologies (CONECCT)","volume":"23 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133364544","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-07-08DOI: 10.1109/CONECCT55679.2022.9865113
Dhruti Ranjan Gaan, M. Kumar, Dinakarn E, S. S
Stepper motors are widely used in precise positioning applications because of their ease of use, open loop control, almost constant torque output throughout the operating range and simplified control circuitry. A lot of improvement has been done in drive schemes of these stepper motors; implementation of microstepping to improve positional accuracy and to reduce disturbance torque in compared to full step; Pulse Width Modulation (PWM) mode of current control to counter back Electro-Motive-Force (EMF) voltage loss etc. In microstepping mode, stepper motor is driven with quantized sine and cosine currents. Due to these varying current generations, signal integrity issues within its own circuitry as well as to the neighboring circuits come to forefront. This paper describes PWM mode of microstepping drive using Field Programmable Gate Array (FPGA), N-MOSFET (Metal-Oxide- Semiconductor-Field-Effect-Transistor) & MOSFET drivers, signal integrity issues observed during motor operations & high current switching and best practices to tackle these issues. Logic signals are generated in FPGA. Microstepping current amplitude values are stored in PROM which are being accessed by FPGA and written to DAC for voltage reference. This paper presents the design, layout implementation, analysis, signal integrity issues, mitigation techniques and experimental results of multiple motor drives.
{"title":"Signal Integrity Issues in FPGA based multi-motor microstepping Drives","authors":"Dhruti Ranjan Gaan, M. Kumar, Dinakarn E, S. S","doi":"10.1109/CONECCT55679.2022.9865113","DOIUrl":"https://doi.org/10.1109/CONECCT55679.2022.9865113","url":null,"abstract":"Stepper motors are widely used in precise positioning applications because of their ease of use, open loop control, almost constant torque output throughout the operating range and simplified control circuitry. A lot of improvement has been done in drive schemes of these stepper motors; implementation of microstepping to improve positional accuracy and to reduce disturbance torque in compared to full step; Pulse Width Modulation (PWM) mode of current control to counter back Electro-Motive-Force (EMF) voltage loss etc. In microstepping mode, stepper motor is driven with quantized sine and cosine currents. Due to these varying current generations, signal integrity issues within its own circuitry as well as to the neighboring circuits come to forefront. This paper describes PWM mode of microstepping drive using Field Programmable Gate Array (FPGA), N-MOSFET (Metal-Oxide- Semiconductor-Field-Effect-Transistor) & MOSFET drivers, signal integrity issues observed during motor operations & high current switching and best practices to tackle these issues. Logic signals are generated in FPGA. Microstepping current amplitude values are stored in PROM which are being accessed by FPGA and written to DAC for voltage reference. This paper presents the design, layout implementation, analysis, signal integrity issues, mitigation techniques and experimental results of multiple motor drives.","PeriodicalId":380005,"journal":{"name":"2022 IEEE International Conference on Electronics, Computing and Communication Technologies (CONECCT)","volume":"79 ","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"113991687","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-07-08DOI: 10.1109/CONECCT55679.2022.9865745
V. Advaith, Anushka Shivkumar, B. S. Sowmya Lakshmi
Part-of-speech (POS) tagging is one of the vital Natural Language Processing (NLP) tasks that entails categorising words in a text (corpus) in accordance with a specific part of the speech, based on the word’s context. POS tagging for Indian Languages is not widely explored. Kannada is extremely inflectional and contains one of the most complex and richest collections of linguistic traits. Hence, developing a POS tagger for a resource-poor language such as Kannada is difficult The morphological complexity of Hindi becomes a challenge despite there having been numerous attempts of building a POS tagger for the language. The proposed work deals with the development of a POS tagger for both Kannada and Hindi by employing Machine Learning (ML) and Deep Learning (DL) algorithms. The results obtained are based on experiments conducted on a corpus consisting of around 3 lakh unique words for Kannada and Hindi combined. The 17 POS tags have been taken from the BIS tag set.
{"title":"Parts of Speech Tagging for Kannada and Hindi Languages using ML and DL models","authors":"V. Advaith, Anushka Shivkumar, B. S. Sowmya Lakshmi","doi":"10.1109/CONECCT55679.2022.9865745","DOIUrl":"https://doi.org/10.1109/CONECCT55679.2022.9865745","url":null,"abstract":"Part-of-speech (POS) tagging is one of the vital Natural Language Processing (NLP) tasks that entails categorising words in a text (corpus) in accordance with a specific part of the speech, based on the word’s context. POS tagging for Indian Languages is not widely explored. Kannada is extremely inflectional and contains one of the most complex and richest collections of linguistic traits. Hence, developing a POS tagger for a resource-poor language such as Kannada is difficult The morphological complexity of Hindi becomes a challenge despite there having been numerous attempts of building a POS tagger for the language. The proposed work deals with the development of a POS tagger for both Kannada and Hindi by employing Machine Learning (ML) and Deep Learning (DL) algorithms. The results obtained are based on experiments conducted on a corpus consisting of around 3 lakh unique words for Kannada and Hindi combined. The 17 POS tags have been taken from the BIS tag set.","PeriodicalId":380005,"journal":{"name":"2022 IEEE International Conference on Electronics, Computing and Communication Technologies (CONECCT)","volume":"146 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123341228","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
There’s been an increase in attempts to control the surge in pollution levels due to extensive exploitation of conventional fossil fuels. These efforts have fueled research for alternative green solutions. Lithium-ion batteries are immensely beneficial for energy storage system. They are extremely advantageous in the automobile industry, particularly as a source to power Electric Vehicles (EVs). Lithium-ion batteries are also vital for powering consumer electronics. The State of Charge (SOC) measurement is used to calculate the remaining usage time of batteries, is one of the most pertinent metric. The goal of current research has been to develop accurate State of Charge (SOC) prediction algorithms. All existing methods require significant amount of superior-quality curated dataset. However, battery researchers have minimal access to commercial battery datasets and therefore must rely on open-access public datasets that lack the required heterogeneity to generate generalised SOC algorithms. To resolve this issue of lack of data, we introduce a Sample Convolution and Interaction Networks (SCINet) to produce resilient synthetic battery data. The code implementation can be found on: https://github.com/vinayakrajurs/Sample-Convolution-Interaction-Syntheic-Data
{"title":"Synthetic Data Generation using Resilient Sample Convolution and Interactive Learning Approach","authors":"Vinayak Raj Urs, Vageesh Maiya, Janamejaya Channegowda, Chaitanya Lingaraj","doi":"10.1109/CONECCT55679.2022.9865732","DOIUrl":"https://doi.org/10.1109/CONECCT55679.2022.9865732","url":null,"abstract":"There’s been an increase in attempts to control the surge in pollution levels due to extensive exploitation of conventional fossil fuels. These efforts have fueled research for alternative green solutions. Lithium-ion batteries are immensely beneficial for energy storage system. They are extremely advantageous in the automobile industry, particularly as a source to power Electric Vehicles (EVs). Lithium-ion batteries are also vital for powering consumer electronics. The State of Charge (SOC) measurement is used to calculate the remaining usage time of batteries, is one of the most pertinent metric. The goal of current research has been to develop accurate State of Charge (SOC) prediction algorithms. All existing methods require significant amount of superior-quality curated dataset. However, battery researchers have minimal access to commercial battery datasets and therefore must rely on open-access public datasets that lack the required heterogeneity to generate generalised SOC algorithms. To resolve this issue of lack of data, we introduce a Sample Convolution and Interaction Networks (SCINet) to produce resilient synthetic battery data. The code implementation can be found on: https://github.com/vinayakrajurs/Sample-Convolution-Interaction-Syntheic-Data","PeriodicalId":380005,"journal":{"name":"2022 IEEE International Conference on Electronics, Computing and Communication Technologies (CONECCT)","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125154233","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-07-08DOI: 10.1109/CONECCT55679.2022.9865761
Jamal Pasha, S. Karpagavalli
Autonomous companion robots have shown to be particularly beneficial for gathering information in areas where people are restricted. Controlling them is often a challenge ng feat due to the environment's ambiguity and the nonlinear dynamics of the grounds. Despite the fact that a variety of controller designs are feasible, and some are documented in the literature, it is unknown which designs are best suited for a certain context. In this paper, we attempted to design a robot that can be adapted for usage in any environment by making only skeleton alterations, we designed the controller with integrating Neural network nodes with Q-learning algorithm to regulate movement of the robot using LIDAR samples. With military applications in consideration, we implemented encryptors to send and receive data, and we distributed all dumps to the controller to ensure that we only needed to be connected when delivering data to the owner. As all this requires high processing speed and storage, we recommend using ESP32-S2 for its high clock speed.
{"title":"Spydobot-AI Based autonomous spider like robot for spying","authors":"Jamal Pasha, S. Karpagavalli","doi":"10.1109/CONECCT55679.2022.9865761","DOIUrl":"https://doi.org/10.1109/CONECCT55679.2022.9865761","url":null,"abstract":"Autonomous companion robots have shown to be particularly beneficial for gathering information in areas where people are restricted. Controlling them is often a challenge ng feat due to the environment's ambiguity and the nonlinear dynamics of the grounds. Despite the fact that a variety of controller designs are feasible, and some are documented in the literature, it is unknown which designs are best suited for a certain context. In this paper, we attempted to design a robot that can be adapted for usage in any environment by making only skeleton alterations, we designed the controller with integrating Neural network nodes with Q-learning algorithm to regulate movement of the robot using LIDAR samples. With military applications in consideration, we implemented encryptors to send and receive data, and we distributed all dumps to the controller to ensure that we only needed to be connected when delivering data to the owner. As all this requires high processing speed and storage, we recommend using ESP32-S2 for its high clock speed.","PeriodicalId":380005,"journal":{"name":"2022 IEEE International Conference on Electronics, Computing and Communication Technologies (CONECCT)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130886207","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Conventional squint mode ω-k algorithm requires higher memory and processing time to generate synthetic aperture radar (SAR) images in the case of a large range to scene center, high squint, and large swath. The high memory and processing time requirements make the algorithm the least preferred solution to implement in real-time SAR image generation on multi-core processor hardware. This paper presents various real-time implementation problems associated with conventional squint mode ω-k algorithm implementation. This paper presents a squint ω-k algorithm with a modified time-domain reference function generation approach to mitigate the multi-core processor’s processing time and memory requirement problems for real-time SAR image generation. The proposed method is implemented on real-time hardware and validated on real SAR data for 1-meter and 3-meter resolutions.
{"title":"Squint SAR Algorithm for Real-Time SAR Imaging","authors":"Peeyush Sahay, Vaidya DhavalKumar B., Kadali Lokesh Kiran","doi":"10.1109/CONECCT55679.2022.9865811","DOIUrl":"https://doi.org/10.1109/CONECCT55679.2022.9865811","url":null,"abstract":"Conventional squint mode ω-k algorithm requires higher memory and processing time to generate synthetic aperture radar (SAR) images in the case of a large range to scene center, high squint, and large swath. The high memory and processing time requirements make the algorithm the least preferred solution to implement in real-time SAR image generation on multi-core processor hardware. This paper presents various real-time implementation problems associated with conventional squint mode ω-k algorithm implementation. This paper presents a squint ω-k algorithm with a modified time-domain reference function generation approach to mitigate the multi-core processor’s processing time and memory requirement problems for real-time SAR image generation. The proposed method is implemented on real-time hardware and validated on real SAR data for 1-meter and 3-meter resolutions.","PeriodicalId":380005,"journal":{"name":"2022 IEEE International Conference on Electronics, Computing and Communication Technologies (CONECCT)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130048907","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-07-08DOI: 10.1109/CONECCT55679.2022.9865817
Arunabha Majumder, Debadrata Sarkar, Sagnik Chakraborty, Abhijit Singh, S. Roy, Aman Arora
The pneumatic artificial muscle (PAM) is considered one of the most preferred actuators in a variety of robotic and industrial applications. However, due to their inherent nonlinearities and hysteretic properties, they are difficult to model and the controller’s design becomes more sophisticated. The position control problem of a PAM having different regions of operations at various axial loads is considered in this paper. A neural network-based gain scheduled proportional-integral-derivative (PID-NN) control scheme has been synthesized and compared to the classical linear PID controllers. The PID gains for different operating regions at different loads are determined using Zeigler Nichols sustained oscillation method. These sets of PID gains are then used to determine the neural network (NN) model that schedules them based on the region of operations and axial loads. To validate the efficacy of the proposed control scheme with regards to different step inputs and a sinusoidal input reference tracking performance, experimental studies are conducted, and comparisons have been made with the PID controller. The experimental results for position control confirm the efficacy of the proposed control strategy.
{"title":"Neural Network-Based Gain Scheduled Position Control of a Pneumatic Artificial Muscle","authors":"Arunabha Majumder, Debadrata Sarkar, Sagnik Chakraborty, Abhijit Singh, S. Roy, Aman Arora","doi":"10.1109/CONECCT55679.2022.9865817","DOIUrl":"https://doi.org/10.1109/CONECCT55679.2022.9865817","url":null,"abstract":"The pneumatic artificial muscle (PAM) is considered one of the most preferred actuators in a variety of robotic and industrial applications. However, due to their inherent nonlinearities and hysteretic properties, they are difficult to model and the controller’s design becomes more sophisticated. The position control problem of a PAM having different regions of operations at various axial loads is considered in this paper. A neural network-based gain scheduled proportional-integral-derivative (PID-NN) control scheme has been synthesized and compared to the classical linear PID controllers. The PID gains for different operating regions at different loads are determined using Zeigler Nichols sustained oscillation method. These sets of PID gains are then used to determine the neural network (NN) model that schedules them based on the region of operations and axial loads. To validate the efficacy of the proposed control scheme with regards to different step inputs and a sinusoidal input reference tracking performance, experimental studies are conducted, and comparisons have been made with the PID controller. The experimental results for position control confirm the efficacy of the proposed control strategy.","PeriodicalId":380005,"journal":{"name":"2022 IEEE International Conference on Electronics, Computing and Communication Technologies (CONECCT)","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127643479","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-07-08DOI: 10.1109/CONECCT55679.2022.9865841
Arjun Thangaraju, Cory E. Merkel
The paper aims to take a deep dive into one of the emerging fields in Deep Learning namely, Adversarial attacks and defenses. We will first see what we mean when we talk of Adversarial examples and learn why they are important? After this, we will explore different types of Adversarial attacks and defenses. Here, we specifically tackle the cases associated with Image Classification. This is done by delving into their respective concepts along with understanding the tools and frameworks required to execute them. The implementation of the FGSM (Fast Gradient Signed Method) attack and the effectiveness of the Adversarial training defense to combat it are discussed. This is done by first analyzing the drop in accuracy from performing the FGSM attack on a MNIST CNN (Convolutional Neural Network) classifier followed by an improvement in the same accuracy metric by defending against the attack using the Adversarial training defense.
{"title":"Exploring Adversarial Attacks and Defenses in Deep Learning","authors":"Arjun Thangaraju, Cory E. Merkel","doi":"10.1109/CONECCT55679.2022.9865841","DOIUrl":"https://doi.org/10.1109/CONECCT55679.2022.9865841","url":null,"abstract":"The paper aims to take a deep dive into one of the emerging fields in Deep Learning namely, Adversarial attacks and defenses. We will first see what we mean when we talk of Adversarial examples and learn why they are important? After this, we will explore different types of Adversarial attacks and defenses. Here, we specifically tackle the cases associated with Image Classification. This is done by delving into their respective concepts along with understanding the tools and frameworks required to execute them. The implementation of the FGSM (Fast Gradient Signed Method) attack and the effectiveness of the Adversarial training defense to combat it are discussed. This is done by first analyzing the drop in accuracy from performing the FGSM attack on a MNIST CNN (Convolutional Neural Network) classifier followed by an improvement in the same accuracy metric by defending against the attack using the Adversarial training defense.","PeriodicalId":380005,"journal":{"name":"2022 IEEE International Conference on Electronics, Computing and Communication Technologies (CONECCT)","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121932319","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：