Seshadhri Srinivasan

Optimal operation of a district heating power plant with thermal energy storage

This paper presents an optimal control strategy for a district heating power plant with thermal energy storage. The main goal of the control strategy is to reduce the operation costs of the power plant, by scheduling the boilers, the operation of the thermal energy storage and the curtailment on the loads. The problem is stated as a constrained optimization in the form of a Mixed Integer Linear Program (MILP), embedded on an Model Predictive Control (MPC) framework. Particular attention is paid to modeling of boilers operating constraints, including the outlet water flow temperature, to the energy exchanged with the thermal energy storage and to the operating modes of the power plant layout, including the constraints related to the supply water temperature needed from the network. The results are performed using the data and the layout of the power plant located in the city of Ylivieska, in Finland. The cost analysis performed shows the advantages of using the predictive control strategy.

Verifying Response Times in Networked Automation Systems Using Jitter Bounds

Networked Automation Systems (NAS) have to meet stringent response time during operation. Verifying response time of automation is an important step during design phase before deployment. Timing discrepancies due to hardware, software and communication components of NAS affect the response time. This investigation uses model templates for verifying the response time in NAS. First, jitter bounds model the timing fluctuations of NAS components. These jitter bounds are the inputs to model templates that are formal models of timing fluctuations. The model templates are atomic action patterns composed of three composition operators-sequential, alternative, and parallel and embedded in time wrapper that specifies clock driven activation conditions. Model templates in conjunction with formal model of technical process offer an easier way to verify the response time. The investigation demonstrates the proposed verification method using an industrial steam boiler with typical NAS components in plant floor.

Modelling time-varying delays in networked automation systems with heterogeneous networks using machine learning techniques

Time-varying delays affect the performance and reliability of networked automation systems (NAS). Recent trend to use wired and wireless networks within NAS induces network delays that vary depending on many factors such as loading, sharing, length of the channel, protocol, and so on. As these factors are inherently time-varying, developing analytical models capturing the effect of all these parameters is complex. This investigation presents a methodology that combines experiments with machine learning techniques to model time-varying delays in networked automation systems integrated with heterogeneous networks. Experiments are conducted on NAS by varying the factors that influence delays and time stamping obtained using Wireshark are used to compute the delay. The data collected on the factors influencing the delays and the corresponding delay values are used to model the delays. In data-mining techniques, the accuracy of the estimates varies with the number of computing elements in the hidden layer and selecting them using trial-and-error approach is cumbersome. The minimum resource allocation network (MRAN) over comes the short-coming as it decides the number of computing elements (neurons) in the hidden layer using error thresholds and pruning strategy. The data collected from the experiment is the input training set to the MRAN. Once trained, the MRAN model gives a functional representation relating the factors affecting delays and the estimated delay for a given network condition. During testing, MRAN estimates are validated using error measurements. Results show that the MRAN delay model can capture delays with good accuracy and can be used a tool to assist design decisions on engineering automation systems with heterogeneous networks. The proposed model gives a framework to model time-varying delays as a function of factors influencing them and can be modified to include any number of parameters. This is a significant benefit against existing models in literature that capture the delays only for particular conditions.

A receding horizon approach for the power flow management with renewable energy and energ storage systems

Optimal operation of energy grids integrated with renewable energy sources (RES) and energy storage systems (ESS) is challenging due to intermittencies in generation and dynamics of the storage. This investigation presents a multi-period alternate current optimal power flow (ACOPF) algorithm for reducing the grid operating cost. The proposed approach uses a receding horizon strategy that solves a non-linear optimization problem during each period by using forecasts and storage dynamics. As a result, the proposed algorithm can optimize the grid operations considering the intermittent renewable generation and storage dynamics. The multi-period ACOPF is illustrated on a Norwegian distribution network with 85 buses and our results illustrate the suitability of the multi-period approach to optimize the grid operations with RES and ESS.

Estimating Random Delays in Modbus Network Using Experiments and General Linear Regression Neural Networks with Genetic Algorithm Smoothing

Time-varying delays adversely affect the performance of networked control systems (NCS) and in the worst case can destabilize the entire system. Therefore, modeling network delays are important for designing NCS. However, modeling time-varying delays are challenging because of their dependence on multiple parameters, such as length, contention, connected devices, protocol employed, and channel loading. Further, these multiple parameters are inherently random and delays vary in a nonlinear fashion with respect to time. This makes estimating random delays challenging. This investigation presents a methodology to model delays in NCS using experiments and general regression neural network (GRNN) due to their ability to capture nonlinear relationship. To compute the optimal smoothing parameter that computes the best estimates, genetic algorithm is used. The objective of the genetic algorithm is to compute the optimal smoothing parameter that minimizes the mean absolute percentage error (MAPE). Our results illustrate that the resulting GRNN is able to predict the delays with less than 3 % error. The proposed delay model gives a framework to design compensation schemes for NCS subjected to time-varying delays.

Design and verification of Cyber-Physical Systems using TrueTime, evolutionary optimization and UPPAAL

Abstract Timing imperfections in Cyber-Physical Systems (CPS) components affect their performance and reliability. This investigation presents a methodology to design and verify CPS using multi-objective evolutionary optimization, model checking and supporting software tools. The time-varying delays in CPS are modeled as constant delays plus jitter. It is shown that the CPS design problem is a trade-off between performance and jitter margin. To make good trade-offs, this investigation models the design problem as a constrained multi-objective optimization problem. The design algorithm results in a non-linear and multi-objective optimization problem that is generally difficult to solve. To overcome this computation barrier, this investigation uses the evolutionary algorithm Non-dominated Sorting Genetic Algorithm II (NSGA II). Solution of the optimization problem computes the proportional integral and derivative controller gains that simultaneously maximize the jitter margin while improving the system performance. Implementing the CPS controller in embedded platform requires the selection and validation of the process schedules as well as network protocols. To validate the CPS design, the computed controller gains are simulated in TrueTime to verify the performance for a given network protocol and processor scheduling policy. Three scheduling policies are considered due to their appropriateness for use with temporal imperfections: fixed priority, earliest deadline first and deadline monotonic. Finally, to verify the CPS for timing guarantees, a timed-automata model is used that defines the timing interfaces among the components. The formal model is then used to verify CPS response time and properties such as safety using computation tree logic in UPPAAL model checker. The proposed CPS design and verification approach is illustrated on an industrial mine pump example. Our results demonstrate that the proposed approach can be used to design and validate CPS for performance and verify timing guarantees. The proposed method provides a systematic design and verification approach that can be used for deployments of CPS in industries.

Stochastic Wheel-Slip Compensation Based Robot Localization and Mapping

Wheel slip compensation is vital for building accurate and reliable dead reckoning based robot localization and mapping algorithms. This investigation presents stochastic slip compensa ...

Verification of design contracts for cyber-physical system design using evolutionary optimization

Cyber-Physical Systems (CPS) have enabled plethora of applications and are seen as future solutions to build large-scale infrastructure. Controller meeting design requirements and performance objectives are required for building CPS. This is non-trivial due to the interplay and temporal constraints among the physical, computing, communication, and control entities within CPS. This paper presents design contracts (an agreement) between software, network and control engineers to build CPS controllers. Jitter bounds are used to synthesize the design contract that are verified by solving multiobjective optimization problem (MOP) to specify temporal parameter (jitter bound) and also guarantee performance. To solve the MOP, NSGA-II is used (Nondominated sorting genetic algorithm) due to its ability to generate non-dominated solutions; a desirable feature for design involving tradeoff. The synthesis and verification procedure is illustrated using experiments with simple example.

Optimization of energy exchanges in utility grids with applications to residential, industrial and tertiary cases

This paper focuses on an efficient Energy Management System (EMS) developed within the European project e-Gotham. Relying on previous results on modeling and controlling microgrids operations, we propose an optimal control strategy to manage the operations of an utility grid. The goal of the proposed strategy is to minimize the operations, maintenance and generations costs balancing a time-varying power demand. The overall problem is stated as a Mixed Integer Linear Programming (MILP) problem. A Model Predictive Control (MPC) technique is used to compensate the system uncertainties. As case of study a residential, an industrial and a tertiary pilot are considered to test the proposed control strategy.

Model checking response times in Networked Automation Systems using jitter bounds

HighlightsProposes methodology and workflow to verify Networked Automation Systems (NAS) timing performance during design, rather after deployment; thereby, eliminating the need for costly hardware upgrades and design modifications.A component based modelling approach for NAS capturing the timing performance, specifications and behaviour.The proposed method can be used study timing properties and guarantees in NAS.The proposed workflow can be used to benchmark different automation solutions and propose alternatives.The workflow is illustrated using experiments and industrial deployment example. Response time (RT) of Networked Automation Systems (NAS) is affected by timing imperfections induced due to the network, computing and hardware components. Guaranteeing RT in the presence of such timing imperfections is essential for building dependable NAS, and to avoid costly upgrades after deployment in industries.This investigation proposes a methodology and work-flow that combines modelling, simulation, verification, experiments, and software tools to verify the RT of the NAS during the design, rather than after deployment. The RT evaluation work-flow has three phases: model building, modelling and verification. During the model building phase component reaction times are specified and their timing performance is measured by combining experiments with simulation. During the modelling phase, component based mathematical models that capture the network architecture and inter-connection are proposed. Composition of the component models gives the NAS model required for studying the RT performance on system level. Finally, in the verification step, the NAS formal models are abstracted as UPPAAL timed automata with their timing interfaces. To model timing interfaces, the action patterns, and their timing wrapper are proposed. The formal model of high level of abstraction is used to verify the total response time of the NAS where the reactions to be verified are specified using a subset of timed computation tree logic (TCTL) in UPPAAL model checker. The proposed approach is illustrated on an industrial steam boiler deployment.