Pub Date : 2019-07-01DOI: 10.1109/IBSSC47189.2019.8973035
A. Jacob, Vedanta Pawar, Vinay Vishwakarma, Anand D. Mane
Video streaming requirement has increased exponentially and video currently consumes 75% of the internet traffic. Due to which video streaming and storage is a huge challenge for service providers. Image and video compression algorithms rely on codecs which are encoders and decoders that lack adaptability. Due to the advent and advances in Deep Learning these issues can be solved. This paper proposes a method for video compression using neural networks that outperforms the H.264/AVC video coding standard as measured using Multi-Scale - Structural Similarity Index (MS-SSIM).The neural network model proposed is a multi-layer architecture consisting of two parts i) Encoder and ii) Decoder. The training of the two parts of the model happens together and during test time the encoder and decoder are separated to be used as just like any another compression encoding and decoding modules. The entire model’s purpose was to try and capitalize on the temporal and spatial dependencies between frames of a video.
{"title":"Deep Learning Approach to Video Compression","authors":"A. Jacob, Vedanta Pawar, Vinay Vishwakarma, Anand D. Mane","doi":"10.1109/IBSSC47189.2019.8973035","DOIUrl":"https://doi.org/10.1109/IBSSC47189.2019.8973035","url":null,"abstract":"Video streaming requirement has increased exponentially and video currently consumes 75% of the internet traffic. Due to which video streaming and storage is a huge challenge for service providers. Image and video compression algorithms rely on codecs which are encoders and decoders that lack adaptability. Due to the advent and advances in Deep Learning these issues can be solved. This paper proposes a method for video compression using neural networks that outperforms the H.264/AVC video coding standard as measured using Multi-Scale - Structural Similarity Index (MS-SSIM).The neural network model proposed is a multi-layer architecture consisting of two parts i) Encoder and ii) Decoder. The training of the two parts of the model happens together and during test time the encoder and decoder are separated to be used as just like any another compression encoding and decoding modules. The entire model’s purpose was to try and capitalize on the temporal and spatial dependencies between frames of a video.","PeriodicalId":148941,"journal":{"name":"2019 IEEE Bombay Section Signature Conference (IBSSC)","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127769871","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 : 2019-07-01DOI: 10.1109/IBSSC47189.2019.8973053
Dharni Shah, Sanaya Shah, V. Sharma, Prof. Vijaya Kamble
Micro-expressions (MEs) are involuntary, subtle expressions which can reveal concealed emotions that people don’t want to show. However, analyzing such rapid facial micro-expressions is very challenging due to their short duration and low intensity. Here, we are emphasizing on macro & micro-expressions recognition on diverse Indian faces and emotions. There are biases in the result due to lack of diversity in the available datasets i.e there are only one or two types of facial features, skin tones, etc. included in the dataset. This leads to misleading results and do not recognize varied real time input. The given macro-expression & micro-expression datasets are cleaned and pre-processed. Pre-processing includes noise removal, cropping and conversion of images to grayscale followed by segmentation. The action units in a large macro-expression dataset is tested and designed to map the data with various macro-expressions followed by training the weights on the provided dataset of micro-expressions using transfer learning. The model is then trained using deep Convolutional neural layers obtaining validation accuracy of 76.9% for macro-expressions and accuracy 71% for micro-expressions respectively which is better than other techniques using CNN.
{"title":"Emotion Recognition using Micro-expressions","authors":"Dharni Shah, Sanaya Shah, V. Sharma, Prof. Vijaya Kamble","doi":"10.1109/IBSSC47189.2019.8973053","DOIUrl":"https://doi.org/10.1109/IBSSC47189.2019.8973053","url":null,"abstract":"Micro-expressions (MEs) are involuntary, subtle expressions which can reveal concealed emotions that people don’t want to show. However, analyzing such rapid facial micro-expressions is very challenging due to their short duration and low intensity. Here, we are emphasizing on macro & micro-expressions recognition on diverse Indian faces and emotions. There are biases in the result due to lack of diversity in the available datasets i.e there are only one or two types of facial features, skin tones, etc. included in the dataset. This leads to misleading results and do not recognize varied real time input. The given macro-expression & micro-expression datasets are cleaned and pre-processed. Pre-processing includes noise removal, cropping and conversion of images to grayscale followed by segmentation. The action units in a large macro-expression dataset is tested and designed to map the data with various macro-expressions followed by training the weights on the provided dataset of micro-expressions using transfer learning. The model is then trained using deep Convolutional neural layers obtaining validation accuracy of 76.9% for macro-expressions and accuracy 71% for micro-expressions respectively which is better than other techniques using CNN.","PeriodicalId":148941,"journal":{"name":"2019 IEEE Bombay Section Signature Conference (IBSSC)","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124876963","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 : 2019-07-01DOI: 10.1109/IBSSC47189.2019.8973087
Z. Islam, Vikas Singh, N. Verma
The analysis of medical images and to find meaningful patterns in it is a cumbersome task, even with the use of techniques of Computer Vision when the dataset is very large. In such a situation deep learning is a handy tool, because of its ability to learn and extract meaningful patterns and features from the images. The use of multiple modalities of training data to train system has been in practice for conventional machine learning algorithms. Here, in this paper, we are going to present a Deep Learning based architecture for extraction of features from large training set of medical images. The deep learning model is tested against conventional techniques by performing Multimodal fusion, Shared Learning and Cross Learning on it. It was found out that Deep Learning model performs superior than the conventional techniques in multimodal fusion and shared learning settings.
{"title":"Feature learning using Stacked Autoencoder for Multimodal Fusion, Shared and Cross Learning on Medical Images","authors":"Z. Islam, Vikas Singh, N. Verma","doi":"10.1109/IBSSC47189.2019.8973087","DOIUrl":"https://doi.org/10.1109/IBSSC47189.2019.8973087","url":null,"abstract":"The analysis of medical images and to find meaningful patterns in it is a cumbersome task, even with the use of techniques of Computer Vision when the dataset is very large. In such a situation deep learning is a handy tool, because of its ability to learn and extract meaningful patterns and features from the images. The use of multiple modalities of training data to train system has been in practice for conventional machine learning algorithms. Here, in this paper, we are going to present a Deep Learning based architecture for extraction of features from large training set of medical images. The deep learning model is tested against conventional techniques by performing Multimodal fusion, Shared Learning and Cross Learning on it. It was found out that Deep Learning model performs superior than the conventional techniques in multimodal fusion and shared learning settings.","PeriodicalId":148941,"journal":{"name":"2019 IEEE Bombay Section Signature Conference (IBSSC)","volume":"149 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132593213","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 : 2019-07-01DOI: 10.1109/IBSSC47189.2019.8973090
Mandar Bhalerao, Shlok Gujar, Aditya A. Bhave, Anant V. Nimkar
Visual Question Answering (VQA) is a technique by which humans can ask simple questions about an image and get answers. This technique can be extended on video clips to answer simple questions about the things happening in the video. The system will take a video and a natural language question as an input, and it will output a natural language answer. It is a multi-discipline research problem by nature. In this work, we limit our work to answering binary questions, i.e. questions having only yes or no as their answers. It could be further designed to answer complex questions.
{"title":"Visual Question Answering Using Video Clips","authors":"Mandar Bhalerao, Shlok Gujar, Aditya A. Bhave, Anant V. Nimkar","doi":"10.1109/IBSSC47189.2019.8973090","DOIUrl":"https://doi.org/10.1109/IBSSC47189.2019.8973090","url":null,"abstract":"Visual Question Answering (VQA) is a technique by which humans can ask simple questions about an image and get answers. This technique can be extended on video clips to answer simple questions about the things happening in the video. The system will take a video and a natural language question as an input, and it will output a natural language answer. It is a multi-discipline research problem by nature. In this work, we limit our work to answering binary questions, i.e. questions having only yes or no as their answers. It could be further designed to answer complex questions.","PeriodicalId":148941,"journal":{"name":"2019 IEEE Bombay Section Signature Conference (IBSSC)","volume":"23 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132979686","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 : 2019-07-01DOI: 10.1109/IBSSC47189.2019.8973021
Deepa Abin, M. Solanki, Neha Waghchaure, Snehal Shivthare, Rosilin Augustine
The prominent factor affecting the quality of metals are the various kind of defects present on their surfaces. Identifying these defects and taking remedial measures to overcome the defects is of paramount importance to maintain quality. Manual inspection of defects is a tedious process and may sometimes be inaccurate. The objective of this paper is to study various classification techniques and their performance in identifying rust from the metal surfaces. Auto color correlogram has been used on the images for feature extraction. We have evaluated the performance of 13 different classification techniques and they have been compared on the basis of their accuracy and error rates. Accuracy in the range of 95% - 97% was obtained by classification techniques like Bagging, LogitBoost and ensemble method such as Random Forest, whereas J48 gave the least error rate.
{"title":"Machine Learning approach for Defect Identification in Machinery parts","authors":"Deepa Abin, M. Solanki, Neha Waghchaure, Snehal Shivthare, Rosilin Augustine","doi":"10.1109/IBSSC47189.2019.8973021","DOIUrl":"https://doi.org/10.1109/IBSSC47189.2019.8973021","url":null,"abstract":"The prominent factor affecting the quality of metals are the various kind of defects present on their surfaces. Identifying these defects and taking remedial measures to overcome the defects is of paramount importance to maintain quality. Manual inspection of defects is a tedious process and may sometimes be inaccurate. The objective of this paper is to study various classification techniques and their performance in identifying rust from the metal surfaces. Auto color correlogram has been used on the images for feature extraction. We have evaluated the performance of 13 different classification techniques and they have been compared on the basis of their accuracy and error rates. Accuracy in the range of 95% - 97% was obtained by classification techniques like Bagging, LogitBoost and ensemble method such as Random Forest, whereas J48 gave the least error rate.","PeriodicalId":148941,"journal":{"name":"2019 IEEE Bombay Section Signature Conference (IBSSC)","volume":"98 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134581155","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 : 2019-07-01DOI: 10.1109/IBSSC47189.2019.8973103
Trupti Chandak, Chaitanya Ghorpade, Sanyam Shukla
Malicious activities can harm the security of the system. These activities must be avoided. Network traffic data can be monitored and analyzed by using intrusion detection system. Different data mining classification techniques are used to detect network attacks. Dimensionality reduction performs key role in the Intrusion Detection System, since detecting anomalies is time-consuming. Recently a lot of work has been done in feature selection. But, most of the authors have modified the KDD99 test dataset. Modification of training dataset is valid but modifying test dataset is against the machine learning ethics. This work comprises some of the recently proposed feature selection algorithm such as Information gain, Gain Ratio and Correlation-based feature selection with the objective of determining the reduced feature set. The performance is evaluated using a combination of any two feature selection technique. This study proposes a new heuristic based feature selection algorithm using naive Bayes classifier to detect the important reduced feature set. The results are evaluated on c4.5 decision tree classifier and the results are compared with the existing works. The evaluated results show that the proposed reduced feature set gives the effective and efficient performance.
{"title":"Effective Analysis of Feature Selection Algorithms for Network based Intrusion Detection System","authors":"Trupti Chandak, Chaitanya Ghorpade, Sanyam Shukla","doi":"10.1109/IBSSC47189.2019.8973103","DOIUrl":"https://doi.org/10.1109/IBSSC47189.2019.8973103","url":null,"abstract":"Malicious activities can harm the security of the system. These activities must be avoided. Network traffic data can be monitored and analyzed by using intrusion detection system. Different data mining classification techniques are used to detect network attacks. Dimensionality reduction performs key role in the Intrusion Detection System, since detecting anomalies is time-consuming. Recently a lot of work has been done in feature selection. But, most of the authors have modified the KDD99 test dataset. Modification of training dataset is valid but modifying test dataset is against the machine learning ethics. This work comprises some of the recently proposed feature selection algorithm such as Information gain, Gain Ratio and Correlation-based feature selection with the objective of determining the reduced feature set. The performance is evaluated using a combination of any two feature selection technique. This study proposes a new heuristic based feature selection algorithm using naive Bayes classifier to detect the important reduced feature set. The results are evaluated on c4.5 decision tree classifier and the results are compared with the existing works. The evaluated results show that the proposed reduced feature set gives the effective and efficient performance.","PeriodicalId":148941,"journal":{"name":"2019 IEEE Bombay Section Signature Conference (IBSSC)","volume":"344 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134050778","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 : 2019-07-01DOI: 10.1109/IBSSC47189.2019.8973218
M. Phatak, Manasi S. Patwardhan, Meenakshi S. Arya
Aesthetics is defined by the properties of arts and beauty, thus making it a very subjective domain. In our day to day lives, with the increase of multimedia requirements, the aesthetic appeal of images and videos has gained much importance in varied fields like advertising, film making, User-Interface design, social networking etc. Visual attributes greatly affect the aesthetic sense of the viewers. In this paper, to start with, we dive into the details of low level, middle level and high level image attributes that contribute towards the aesthetic appeal of images. Videos share their attributes with images except for the presence of motion in a video. Next, we proceed towards the handcrafted and deep learning techniques for assessing image and video attributes for their aesthetic appeal. Motion is an important but seldom explored visual attribute that affects video aesthetic appeal. Typically, slow motion creates an impact and appreciation amongst the viewers as they absorb the contents of the video better in comparison to faster motion in the video. Surveys conducted showcased the human inclination towards slowly paced videos in comparison to the fast-paced ones. We have experimented with the deep learning framework for detecting motion in nature based videos. Deep learning achieves an impressive performance in comparison to the handcrafted methods, thus reinforcing current trust in the deep learning frameworks for multimedia analysis.
{"title":"Deep Learning for motion based video aesthetics","authors":"M. Phatak, Manasi S. Patwardhan, Meenakshi S. Arya","doi":"10.1109/IBSSC47189.2019.8973218","DOIUrl":"https://doi.org/10.1109/IBSSC47189.2019.8973218","url":null,"abstract":"Aesthetics is defined by the properties of arts and beauty, thus making it a very subjective domain. In our day to day lives, with the increase of multimedia requirements, the aesthetic appeal of images and videos has gained much importance in varied fields like advertising, film making, User-Interface design, social networking etc. Visual attributes greatly affect the aesthetic sense of the viewers. In this paper, to start with, we dive into the details of low level, middle level and high level image attributes that contribute towards the aesthetic appeal of images. Videos share their attributes with images except for the presence of motion in a video. Next, we proceed towards the handcrafted and deep learning techniques for assessing image and video attributes for their aesthetic appeal. Motion is an important but seldom explored visual attribute that affects video aesthetic appeal. Typically, slow motion creates an impact and appreciation amongst the viewers as they absorb the contents of the video better in comparison to faster motion in the video. Surveys conducted showcased the human inclination towards slowly paced videos in comparison to the fast-paced ones. We have experimented with the deep learning framework for detecting motion in nature based videos. Deep learning achieves an impressive performance in comparison to the handcrafted methods, thus reinforcing current trust in the deep learning frameworks for multimedia analysis.","PeriodicalId":148941,"journal":{"name":"2019 IEEE Bombay Section Signature Conference (IBSSC)","volume":"90 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133749164","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 : 2019-07-01DOI: 10.1109/IBSSC47189.2019.8973098
Shraddha Suratkar, F. Kazi, R. Gaikwad, Akshay Shete, Raj Kabra, Shantanu Khirsagar
Intrusion Detection systems are used for detecting attacks on a system. The host-based intrusion detection system (HIDS) detect the ongoing attacks on a Host system. HIDS model is proposed using System Call Analysis consisting of two modules, an Anomaly Detection module and a Multi-HMM module for state prediction. Anomaly Detection module uses Long Short-term memory (LSTM) architecture, a special type of Recurrent Neural Network, for detection of anomalies in system call traces. It models the normal behaviour of the system using system call patterns which enables it to detect even ‘Zero-day’ attacks. The State prediction module is based on Multiple Hidden Markov Model (Multi-HMM), in which each HMM model a known attack. It takes a sequence of system calls as input and predicts next ‘N’ most probable system calls during the attack. After performing a number of experiments, results show that the model has high recognition rate and low false alarm rate.
{"title":"Multi Hidden Markov Models for Improved Anomaly Detection Using System Call Analysis","authors":"Shraddha Suratkar, F. Kazi, R. Gaikwad, Akshay Shete, Raj Kabra, Shantanu Khirsagar","doi":"10.1109/IBSSC47189.2019.8973098","DOIUrl":"https://doi.org/10.1109/IBSSC47189.2019.8973098","url":null,"abstract":"Intrusion Detection systems are used for detecting attacks on a system. The host-based intrusion detection system (HIDS) detect the ongoing attacks on a Host system. HIDS model is proposed using System Call Analysis consisting of two modules, an Anomaly Detection module and a Multi-HMM module for state prediction. Anomaly Detection module uses Long Short-term memory (LSTM) architecture, a special type of Recurrent Neural Network, for detection of anomalies in system call traces. It models the normal behaviour of the system using system call patterns which enables it to detect even ‘Zero-day’ attacks. The State prediction module is based on Multiple Hidden Markov Model (Multi-HMM), in which each HMM model a known attack. It takes a sequence of system calls as input and predicts next ‘N’ most probable system calls during the attack. After performing a number of experiments, results show that the model has high recognition rate and low false alarm rate.","PeriodicalId":148941,"journal":{"name":"2019 IEEE Bombay Section Signature Conference (IBSSC)","volume":"23 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122964069","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 : 2019-07-01DOI: 10.1109/IBSSC47189.2019.8973036
Pushkar Prakash Arya, S. Chakrabarty
Fractional order (FO) controller provide more number of tuning parameters than integer order (IO) counterpart. In this paper, a fractional order (FO) internal model control (IMC) is considered which provides two tuning parameters $(lambda$ and $beta)$ as compared to integer order (IO) counterpart. These two parameters are tuned to get the desired $varphi_{m}$ and $omega_{g}$ for higher order systems. Usually, we use IO approximation of FO terms in the controllers. It is observed that 5-15 order IO approximation provide almost accurate behavior for the FO terms. The problem with such approximation is the resulting higher order controller. In this work, a balanced truncation method is used to get a reduced order controller which retains important properties of higher order controller. The viability of reduced order controller is measured in terms of maximum sensitivity $(M_{s})$ and the frequency at which the system is most sensitive $(omega_{m})$. Further, the results are compared in terms of rise time $(T_{r})$, settling time $(T_{s})$, maximum overshoot (%$M)$, integral square error (ISE), integral absolute error (IAE) and integral of the time weighted absolute error (ITAE). The importance of using reduced order controller is checked using three examples: first, a minimum phase (MP) system, second, a non-minimum phase (NMP) system with right hand plane (RHP) zero and third, a first order plus time delay (FOPTD) system. It is observed that the lower order controller can be used for higher order controller.
分数阶(FO)控制器比整数阶(IO)控制器提供更多的调优参数。本文考虑了分数阶(FO)内模控制(IMC)，与整数阶(IO)内模控制相比，IMC提供了两个调优参数$(lambda$和$beta)$。对这两个参数进行调优，以获得更高阶系统所需的$varphi_{m}$和$omega_{g}$。通常，我们在控制器中使用IO逼近FO项。观察到，5-15阶IO近似为FO项提供了几乎准确的行为。这种近似的问题是产生的高阶控制器。本文采用平衡截断法得到了保留高阶控制器重要特性的降阶控制器。降阶控制器的可行性是根据最大灵敏度$(M_{s})$和系统最敏感的频率$(omega_{m})$来衡量的。进一步比较了上升时间$(T_{r})$、沉降时间$(T_{s})$、最大超调量(%$M)$, integral square error (ISE), integral absolute error (IAE) and integral of the time weighted absolute error (ITAE). The importance of using reduced order controller is checked using three examples: first, a minimum phase (MP) system, second, a non-minimum phase (NMP) system with right hand plane (RHP) zero and third, a first order plus time delay (FOPTD) system. It is observed that the lower order controller can be used for higher order controller.
{"title":"Reduced order controller for FO-IMC with desired phase margin and gain cross-over frequency","authors":"Pushkar Prakash Arya, S. Chakrabarty","doi":"10.1109/IBSSC47189.2019.8973036","DOIUrl":"https://doi.org/10.1109/IBSSC47189.2019.8973036","url":null,"abstract":"Fractional order (FO) controller provide more number of tuning parameters than integer order (IO) counterpart. In this paper, a fractional order (FO) internal model control (IMC) is considered which provides two tuning parameters $(lambda$ and $beta)$ as compared to integer order (IO) counterpart. These two parameters are tuned to get the desired $varphi_{m}$ and $omega_{g}$ for higher order systems. Usually, we use IO approximation of FO terms in the controllers. It is observed that 5-15 order IO approximation provide almost accurate behavior for the FO terms. The problem with such approximation is the resulting higher order controller. In this work, a balanced truncation method is used to get a reduced order controller which retains important properties of higher order controller. The viability of reduced order controller is measured in terms of maximum sensitivity $(M_{s})$ and the frequency at which the system is most sensitive $(omega_{m})$. Further, the results are compared in terms of rise time $(T_{r})$, settling time $(T_{s})$, maximum overshoot (%$M)$, integral square error (ISE), integral absolute error (IAE) and integral of the time weighted absolute error (ITAE). The importance of using reduced order controller is checked using three examples: first, a minimum phase (MP) system, second, a non-minimum phase (NMP) system with right hand plane (RHP) zero and third, a first order plus time delay (FOPTD) system. It is observed that the lower order controller can be used for higher order controller.","PeriodicalId":148941,"journal":{"name":"2019 IEEE Bombay Section Signature Conference (IBSSC)","volume":"86 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124955954","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 : 2019-07-01DOI: 10.1109/IBSSC47189.2019.8973108
Tushar Gawande, R. Deshmukh, Rajendra Patrika, S. Deshmukh
Gas sensor arrays have proven to be an effective and low-cost solution for measurement of pollutants in large area wherein they can be deployed in the form of network mesh. However, while using multiple identical sensor arrays, the inherent sensor array variability associated with these sensor array needs to be tackled prior to its deployment. In the present work MEMS based gas sensors are used to form three identical gas sensor arrays. Out of the three sensor arrays one was identified as master sensor array and other two as slave sensor arrays. Mestimators were used to develop a robust regression model to map the responses of slave onto master using various ethanol concentrations. The results of the same are presented in this paper.
{"title":"Inherent MEMS sensor array variability reduction using robust regression","authors":"Tushar Gawande, R. Deshmukh, Rajendra Patrika, S. Deshmukh","doi":"10.1109/IBSSC47189.2019.8973108","DOIUrl":"https://doi.org/10.1109/IBSSC47189.2019.8973108","url":null,"abstract":"Gas sensor arrays have proven to be an effective and low-cost solution for measurement of pollutants in large area wherein they can be deployed in the form of network mesh. However, while using multiple identical sensor arrays, the inherent sensor array variability associated with these sensor array needs to be tackled prior to its deployment. In the present work MEMS based gas sensors are used to form three identical gas sensor arrays. Out of the three sensor arrays one was identified as master sensor array and other two as slave sensor arrays. Mestimators were used to develop a robust regression model to map the responses of slave onto master using various ethanol concentrations. The results of the same are presented in this paper.","PeriodicalId":148941,"journal":{"name":"2019 IEEE Bombay Section Signature Conference (IBSSC)","volume":"56 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126708077","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：