Nirav Bhatt

Deconstructing principal component analysis using a data reconciliation perspective

Abstract Data reconciliation (DR) and principal component analysis (PCA) are two popular data analysis techniques in process industries. Data reconciliation is used to obtain accurate and consistent estimates of variables and parameters from erroneous measurements. PCA is primarily used as a method for reducing the dimensionality of high dimensional data and as a preprocessing technique for denoising measurements. These techniques have been developed and deployed independently of each other. The primary purpose of this article is to elucidate the close relationship between these two seemingly disparate techniques. This leads to a unified framework for applying PCA and DR. Further, we show how the two techniques can be deployed together in a collaborative and consistent manner to process data. The framework has been extended to deal with partially measured systems and to incorporate partial knowledge available about the process model.

Identification and Multivariable Gain-Scheduling Control for Cloud Computing Systems

This paper presents the dynamic modeling and performance control of a Web server hosted on a private cloud. The cloud hosting Web server is a variable capacity system with two control inputs: 1) the number of virtual machines (VMs), which is indicative of the capacity of the cloud, and 2) the admission control used for regulating workload. As the workload and the hosting conditions change frequently, the linear parameter-varying (LPV) framework is well suited to derive the model. For the hosted Web server, we obtain an multiple input multiple output (MIMO) LPV model with performance metrics such as the response time and the throughput, which is then converted to polytopic LPV form using tensor product transformation. Finally, we design a gain scheduled linear quadratic regulator controller for performance guarantees with optimal cost of VMs. The identification, validation, and control experiments are demonstrated on the open source Eucalyptus cloud platform. The HTTP requests are generated using customized synthetic workload generator tool.

A novel approach for phase identification in smart grids using Graph Theory and Principal Component Analysis

Consumers with low demand, like households, are generally supplied single-phase power by connecting their service mains to one of the phases of a distribution transformer. The distribution companies face the problem of keeping a record of consumer connectivity to a phase due to uninformed changes that happen. The exact phase connectivity information is important for the efficient operation and control of distribution system. We propose a new data driven approach to the problem based on Principal Component Analysis (PCA) and its Graph Theoretic interpretations, using energy measurements in equally timed short intervals, generated from smart meters. We propose an algorithm for inferring phase connectivity from noisy measurements. The algorithm is demonstrated using simulated data for phase connectivities in distribution networks.

An LPV approach to performance modeling of a web server on a private cloud

This paper presents dynamic modeling of a webserver hosted on a private cloud using grey box identification technique. In contrast to the results in literature, we model the web-server as a linear parameter varying (LPV) state space system valid around several well defined operating regions. We programmatically generate synthetic HTTP load based on open source workload tool called httperf to obtain response time as performance metric. Finally, we validate the models on test data by conducting experiments in the actual Eucalyptus cloud environment.

Incremental Kinetic Identification based on Experimental data From Steady-state Plug Flow Reactors

Abstract This work develops an incremental model identification approach for analyzing concentrations from non-isothermal steady-state plug flow reactors (SPFR) with tubular geometry. A model of non-isothermal SPFRs consists of the material and energy balance equations of SPFR in form of a set of differential equations. Since SPFRs often are used for studying gas phase reactions, the pressure drop equation will also be included. Two scenarios of reactor operations for collecting concentration measurements are distinguished: (S1) Concentrations of species are measured along the length of reactors, and (S2) concentrations of species in the outlet stream are measured for the given inlet concentration condition at various volumetric flowrates. For Scenarios S1 and S2, the incremental identification is extended to identify reaction kinetics from experimental data. Scenario S1 is corroborated with a simulated example of pyrolysis of dimethylformamide in an isothermal tubular reactor.

Identifying Topology of Low Voltage Distribution Networks Based on Smart Meter Data

In a power distribution network, the network topology information is essential for an efficient operation. This network connectivity information is often not available at the low voltage (LV) level due to uninformed changes that happen from time to time. In this paper, we propose a novel data-driven approach to identify the underlying network topology for LV distribution networks including the load phase connectivity from time series of energy measurements. The proposed method involves the application of principal component analysis and its graph-theoretic interpretation to infer the steady state network topology from smart meter energy measurements. The method is demonstrated through simulation on randomly generated networks and also on IEEE recognized Roy Billinton distribution test system.

Diagnosis and rectification of model-process mismatch in chemical reaction systems

In chemical reaction systems, a reliable kinetic model is essential to predict the evolution of concentrations. Often, variations in physicochemical or operational conditions lead to change in a part or whole of the reaction kinetics. This leads to a mismatch between the process model and the process. Consequently, the current model fails to capture the behaviour of the underlying system. Hence, it is important to detect such a change and rectify the model appropriately. In this work, we formulate this problem in a fault diagnosis and identification framework. We propose a framework for solving this problem in the following three steps: (1) Detection of overall model-process mismatch, (2) Isolation of faulty rate model, and (3) Rectification of faulty rate model. The detection step is carried out using a global test which identifies if there is a fault in the system. The isolation of faulty reaction is accomplished using a nodal test statistic. The model-process mismatch in that of the isolated reaction is rectified using a bank of extended Kalman filters. The proposed approach is illustrated using a simulation example of the Acetoacetylation of Pyrrole system.

A novel vehicle model for longitudinal motion analysis

Vehicles with adaptive cruise control (ACC) are an important component of smart and (semi-) autonomous roads and highway systems. These systems are envisaged to handle a large volume of manually driven and autonomous vehicles, ensuring smooth traffic flow. A reliable model capturing the dynamics of a vehicle is essential for developing techniques for model-based adaptive cruise control. The existing literature presents detailed models for individual subsystems are such as engine, brake, wheel etc. However, for analysing longitudinal motion, it is desirable to have a novel model capturing all aspects of the vehicle behaviour. In this work, we propose such a unified model of vehicle powertrain describing the dynamics of different components and interaction between them affecting the longitudinal motion. Lastly, using simulation studies, it is shown that the developed model can emulate the vehicle dynamics in the real-life scenarios such as maneuvering through congestion, stop and go, and sudden braking.

Improving the accuracy of hyaluronic acid molecular weight estimation by conventional size exclusion chromatography

There is an unreasonably high variation in the literature reports on molecular weight of hyaluronic acid (HA) estimated using conventional size exclusion chromatography (SEC). This variation is most likely due to errors in estimation. Working with commercially available HA molecular weight standards, this work examines the extent of error in molecular weight estimation due to two factors: use of non-HA based calibration and concentration of sample injected into the SEC column. We develop a multivariate regression correlation to correct for concentration effect. Our analysis showed that, SEC calibration based on non-HA standards like polyethylene oxide and pullulan led to approximately 2 and 10 times overestimation, respectively, when compared to HA-based calibration. Further, we found that injected sample concentration has an effect on molecular weight estimation. Even at 1g/l injected sample concentration, HA molecular weight standards of 0.7 and 1.64MDa showed appreciable underestimation of 11-24%. The multivariate correlation developed was found to reduce error in estimations at 1g/l to <4%. The correlation was also successfully applied to accurately estimate the molecular weight of HA produced by a recombinant Lactococcus lactis fermentation.

Online Approach for Diagnosis and Rectification of Model–Plant Mismatch in Open Reaction Systems using Incremental Framework

Abstract A reliable dynamic model is essential for model–based control, monitoring, and optimization of reaction systems. Hence, a change in a part or whole of the reaction kinetics of these systems leads to poor performance. In this work, the problem of model–plant mismatch in open reaction system is studied. We propose an online fault diagnosis and rectification framework for solving the problem of model–plant mismatch for open reaction systems. The framework combines the concept of the extents of reaction and flowrate in reaction systems and incremental model identification approach for isolation and rectification of the deficient part of the model. The proposed framework will be demonstrated via a simulation example of the acetoacetylation of pyrrole in a semi–batch reactor for two scenarios: (i) shift in the change of one of the reaction rates, and (ii) change in the inlet flowrate.