Madhusree Kundu

Classification and authentication of unknown water samples using machine learning algorithms

This paper proposes the development of water sample classification and authentication, in real life which is based on machine learning algorithms. The proposed techniques used experimental measurements from a pulse voltametry method which is based on an electronic tongue (E-tongue) instrumentation system with silver and platinum electrodes. E-tongue include arrays of solid state ion sensors, transducers even of different types, data collectors and data analysis tools, all oriented to the classification of liquid samples and authentication of unknown liquid samples. The time series signal and the corresponding raw data represent the measurement from a multi-sensor system. The E-tongue system, implemented in a laboratory environment for 6 numbers of different ISI (Bureau of Indian standard) certified water samples (Aquafina, Bisleri, Kingfisher, Oasis, Dolphin, and McDowell) was the data source for developing two types of machine learning algorithms like classification and regression. A water data set consisting of 6 numbers of sample classes containing 4402 numbers of features were considered. A PCA (principal component analysis) based classification and authentication tool was developed in this study as the machine learning component of the E-tongue system. A proposed partial least squares (PLS) based classifier, which was dedicated as well; to authenticate a specific category of water sample evolved out as an integral part of the E-tongue instrumentation system. The developed PCA and PLS based E-tongue system emancipated an overall encouraging authentication percentage accuracy with their excellent performances for the aforesaid categories of water samples.

Dissolution kinetics of nickel laterite ore using different secondary metabolic acids

The dissolution kinetics of nickel laterite ore in aqueous acid solutions of three metabolic acids, i.e., citric acid, oxalic acid and acetic acid were investigated in a batch reactor individually. It was determined that experimental data comply with a shrinking core model. The diffusion coefficients for citric acid, oxalic acid and acetic acid were found to be 1.99×10 -9 cm 2 /s, 2.59×10 -8 cm 2 /s and 1.92×10 -10 cm 2 /s respectively. The

The e‐tongue‐based classification and authentication of mineral water samples using cross‐correlation‐based PCA and Sammon's nonlinear mapping

The present paper elaborates on the design of classifiers based on cross‐correlation‐based principal component analysis (PCA) and Sammons nonlinear mapping (NLM) using current signals obtained from electronic tongue (e‐tongue) with commercial mineral water samples available in the Indian market. The pulse‐voltammetric method is used to capture the electroanalytical/electrochemical characteristics of the sampled mineral waters by considering a real model for the liquid–electrode interface in a given e‐tongue apparatus. Then the cross‐correlation coefficients between the output and input signals are determined. Both PCA and Sammons NLM create a subspace from high‐dimensional mineral water data by considering the principal eigenvectors and minimising the stress function, respectively. The proposed cross‐correlation‐based PCA and Sammons classifiers establish the highest separation distance among the investigated water brands and carries out the authentication of more than one unknown sample of the same brand with a certain degree of variability with respect to their sources. Copyright

Numerical solution of multi-order fractional differential equations using generalized triangular function operational matrices

Present article proposes numerical technique for the solution of linear and nonlinear multi-order fractional differential equations.The proposed method is based on newly computed generalized triangular function operational matrices for Riemann-Liouville fractional order integral.Theoretical error analysis is performed to estimate the upper bound of absolute error between the exact Riemann-Liouville fractional order integral and its approximation in the triangular functions domain. Most fractional differential equations do not have closed form solutions. Development of effective numerical techniques has been an interesting research topic for decades. In this context, this paper proposes a numerical technique, for solving linear and nonlinear multi-order fractional differential equations, based on newly computed generalized triangular function operational matrices for Riemann-Liouville fractional order integral. The orthogonal triangular functions are evolved from a simple dissection of piecewise constant orthogonal block pulse functions. Theoretical error analysis is performed to estimate the upper bound of absolute error between the exact Riemann-Liouville fractional order integral and its approximation in the triangular functions domain. Numerical examples are considered for investigating the applicability and effectiveness of proposed technique to solve multi-order fractional differential equations. The results encourage the use of orthogonal TFs for analysis of real processes exhibiting fractional dynamics.

ART1 Network: Application in Wine Classification

Unsupervised neural network (NN) based on Adaptive Resonance Theory (ART1) was successfully implemented as an alternative to statistical classifier in order to discriminate among the 178 samples of wine possessing 13 numbers of feature variables. A pattern recognition tool, principal component analysis (PCA) was applied to reduce the dimensionality of the feature variables by 5; out of which the first 2 numbers of principal components captured over 55.4 % of the variance of the dataset of wine. Supervised non- hierarchical K-means clustering was used to designate the classes available among the wine samples, hence discrimination. Supervised hierarchical clustering technique was also applied for discrimination with a mention of their classification level in the produced dendograms. After the discrimination made by hierarchical as well as non- hierarchical clustering, the ART1 classifier was designed.

Decoupling Multivariable Processes using Partial Least Squares for Decentralized Control

Multivariable control systems suffer very much from unwanted interactions among control loops. Change in setpoint of one variable may cause other variables to deviate from their respective steady states because of couplings between unpaired variables. Due to unreliability problems, conventional decouplers are not appropriate for higher order processes. This paper proposes Partial Least Squares (PLS), multivariate statistical process control technique (MVSPC), based decoupling strategy to attain satisfactory performance and consistent product quality in spite of disturbances. The proposed scheme was applied on conventional and heat integrated distillation processes. The results have shown the reliability and robustness of Partial Least Squares based decouplers over conventional decouplers. General Terms Process Identification & control, Statistical Process Control

Identification and Control of Distillation Process using Partial Least Squares based Artificial Neural Network

Partial least squares technique has been in use for identification of the dynamics & control for multivariable distillation process. Discrete input-output time series data ) ( Y X were generated by exciting non-linear process models with pseudo random binary signals. Signal to noise ratio was set to 10 by adding white noise to the data. The ARX models as well FIR models in combination with least squares technique were used to build up dynamic inner relations among the scores of the time series data ) ( Y X , which logically built up the framework for PLS based process controllers. In this work, process dynamics was also identified in latent subspace using neural networks. The inverse dynamics of the latent variable based NN process acted as inverse neural controller (DINN). Distillation process without any decoupler could be controlled by a series of NN-SISO controllers General Terms Process Identification & control, Statistical Process Control

Monitoring of Drum-boiler Process Using Statistical Techniques

Present study addresses the monitoring of drum boiler process. Methodologies; based on clustering time series data and moving window based pattern matching have been proposed for the detection of fault in the chosen process. Design databases were created for the process by simulating the developed process model. A modified k-means clustering algorithm using similarity measure as a convergence criterion has been adopted for discriminating among time series data pertaining to various operating conditions. The proposed distance and PCA based combined similarity along with the moving window approach were used to discriminate among the normal operating conditions as well as detection of faults for the processes taken up. measures. The model has been derived using first principles and is characterized by few physical parameters. The parameters used in the model are those from a Swedish Power Plant. The values of all the parameters used were adapted from the work of K. J. Astrom & R. D. Bell. Sixteen (16) numbers of datasets belonging to four operating conditions including an abnormal operating condition were generated to evaluate the performance of the proposed techniques.

Design of Multivariable Neural Controllers Using a Classical Approach

In the present study, the neural network (NN) based multivariable controllers were designed as a series of single input-single output (SISO) controllers or multi variable SISO (MVSISO) controllers utilizing the classical decoupled process models. Multilayer feed forward networks (FFNN) were used as direct inverse neural network (DINN) controllers, which used the inverse dynamics of the decoupled process. To address the disturbance rejection problems, the IMC based neural control architecture was proposed with suitable choice of filter and disturbance transfer function. Multi input - multi output (MIMO) non-linear processes like interacting tank systems, temperature and level control of a mixing tank with hot and cold input streams & a (2×2) distillation process were considered as case studies for that purpose. Simplified as well as ideally decoupled process as well as disturbance transfer functions was used for neural controller design. DINN/IMC based NN controllers performed effectively well in comparison to conventional P/ PI/IMC based PI controllers for set-point tracking & regulator problems.

Control of Yeast Fermentation Bioreactor in Subspace

PCA based temperature controller was used to control Ethanol concentration produced in Yeast fermentation process. The controller was designed at a specific operating point and its disturbance rejection performances were studied. Substrate inlet temperature proved to be the most significant disturbance input from the analysis of open loop responses. -statistic (SPE) of process measurements confirmed that in the face of disturbances and noise the process could be held to the specific operating condition using the controller designed in subspace. General Terms Statistical Process Control