
TuA11 Regular Session, Pikake 2 
Add to My Program 
Energy Systems I 


Chair: Garulli, Andrea  Univ. di Siena 
CoChair: Bentsman, Joseph  Univ. of Illinois at UrbanaChampaign 

10:0010:20, Paper TuA11.1  Add to My Program 
Control of Energy Systems As Distributed Parameter Systems with Software Support by Virtual Software Environments (I) 
Hulko, Gabriel  Slovak Univ. of Tech. 
RohalIlkiv, Boris  Slovak Univ. of Tech. in Bratislava 
Noga, Pavol  Slovak Univ. of Tech. 
Lipar, Slavomir  Slovak Univ. of Tech. in Bratislava 
Keywords: Energy systems, Distributed parameter systems, Adaptive control
Abstract: Thanks to developments in information technology, socalled virtual software environments offer wide possibilities for the estimation of timespace dynamical characteristics of energy systems as well as for modeling, control and design of distributed parameter systems. Based on these advances we present a novel approach to control of energy systems as lumpedinput and distributedparameteroutput systems. A coalburning fluidized bed furnace temperature field adaptive predictive control ensuring optimal conditions for the desulphurization process is presented as a demonstration of the proposed methodology.


10:2010:40, Paper TuA11.2  Add to My Program 
EfficiencyRisk Tradeoffs in Dynamic Oligopoly Markets – with Application to Electricity Markets 
Huang, Qingqing  Massachusetts Inst. of Tech. 
Roozbehani, Mardavij  Massachusetts Inst. of Tech. 
Dahleh, Munther A.  Massachusetts Inst. of Tech. 
Keywords: Energy systems, Emerging control applications, Stochastic systems
Abstract: In this paper, we examine in an abstract framework, how a tradeoff between efficiency and risk arises in different dynamic oligopolistic market architectures. We consider a market in which there is a monopolistic resource provider and agents that enter and exit the market following a random process. Selfinterested and fully rational agents dynamically update their resource consumption decisions over a finite time horizon, under the constraint that the total resource consumption requirements are met before each individual’s deadline. We then compare the statistics of the stationary aggregate demand processes induced by the noncooperative and cooperative load scheduling schemes. We show that although the noncooperative load scheduling scheme leads to an efficiency loss  widely known as the “price of anarchy”  the stationary distribution of the corresponding aggregate demand process has a smaller tail. This tail, which corresponds to rare and undesirable demand spikes, is important in many applications of interest. On the other hand, when the agents can cooperate with each other in optimizing their total cost, a higher market efficiency is achieved at the cost of a higher probability of demand spikes. We thus posit that the origins of endogenous risk in such systems may lie in the market architecture, which is an inherent characteristic of the system.


10:4011:00, Paper TuA11.3  Add to My Program 
Electric Load Forecasting in the Presence of Active Demand 
Paoletti, Simone  Univ. di Siena 
Garulli, Andrea  Univ. di Siena 
Vicino, Antonio  Univ. di Siena 
Keywords: Energy systems, Filtering, System identification
Abstract: Active Demand (AD) is a new concept in smart grids developed within the EU project ADDRESS. It refers to the active participation of households and small commercial consumers in energy systems by means of the flexibility they can offer. Upon receiving realtime price/volume signals, consumers may find convenient to change their load profiles in return of a monetary reward. In this way, they can contribute to the provision of services to the different participants in the electricity system. Since AD causes modifications of the typical consumers' behaviour, classical load forecasting tools not considering AD signals as inputs are expected to give inaccurate results when applied to load time series including AD effects. In this paper, we study this problem by comparing the prediction performances of several linear models of the load exploiting or not AD signals as inputs. The comparison shows that enhanced prediction results can be obtained by suitably combining the use of AD inputs and the extraction of seasonal characteristics. This is demonstrated by applying the considered approaches to simulated AD effects added to real measurements, representing the aggregated load of about 60 consumers from an Italian LV network.


11:0011:20, Paper TuA11.4  Add to My Program 
On Energy Delivery to DelayAverse Flexible Loads: Optimal Algorithm, Consumer Value and Network Level Impacts 
Kefayati, Mahdi  The Univ. of Texas at Austin 
Baldick, Ross  Univ. of Texas, Austin 
Keywords: Energy systems, Stochastic optimal control, Electrical power systems
Abstract: In many cases, demand for electricity can be viewed as demand for energy over a time horizon and not instantaneous demand for power (i.e., energy rate). Demand for energy translates to more flexibility in energy delivery. Assuming availability of proper information, which is promised by smart grids, this flexibility can be utilized to reduce costs for the consumers alongside providing other benefits. In this paper, we propose a stochastic model for delayaverse flexible demands subject to realtime stochastic spot prices. Based on this model, we obtain the optimal consumption policy and discuss its computational efficiency under different assumptions. Using this optimal scheme we quantify the value of time flexibility (i.e., delay tolerance) in terms of the reduction in the expected cost of satisfying consumer demand as well as the costdelay tradeoff. Finally, through simulations, we analyze the collective behavior of such opportunistic loads and their effects on the power system. Beyond obtaining computationally efficient algorithms for optimal behavior of delayaverse flexible loads, our model provides insights into the value of time flexibility and exhibits costdelay tradeoffs. Furthermore, we show that opportunism on the demand side in realtime pricing environments can result in undesired effects in terms of the aggregate power profile of the loads.


11:2011:40, Paper TuA11.5  Add to My Program 
Wavelet Multiresolution Model Based Generalized Predictive Control for Hybrid CombustionGasification Chemical Looping Process (I) 
Zhang, Shu  Univ. of Illinois at UrbanaChampaign 
Bentsman, Joseph  Univ. of Illinois at UrbanaChampaign 
Lou, Xinsheng  ALSTOM Power, Inc. 
Neuschaefer, Carl  Alstom Power Inc 
Keywords: Energy systems, NL system identification, Adaptive control
Abstract: Chemical looping (CL) process is a novel technology that separates oxygen from nitrogen to facilitate carbon dioxide capture in the design of clean coal power plants. The process, based on the multiphase gassolid flow, has an extremely challenging nonlinear multiscale dynamics with jumps, rendering traditional robust control techniques, such as switching Hinfinity design, difficult to apply and marginally successful. In an effort to model and control such a complex system, we present a generalized predictive control (GPC) scheme based on multiresolution wavelet model structure that characterizes well the nonlinear dynamics of single loop gas/solid flow. The NARX model, nonlinear in the wavelet basis, but linear in parameters, is used for the online chemical looping process identification. The control inputs and wavelet model parameters are calculated by optimizing the cost function using a gradient descent method. The convergence of the proposed GPC scheme is derived using Lyapunov function. Experimental results are provided to demonstrate the effectiveness of the proposed control strategy.


11:4012:00, Paper TuA11.6  Add to My Program 
Optimal Active Control of a Wave Energy Converter (I) 
Abraham, Edo  Imperial Coll. London 
Kerrigan, Eric C.  Imperial Coll. London 
Keywords: Energy systems, Optimal control, Optimization algorithms
Abstract: This paper investigates optimal active control schemes applied to a point absorber wave energy converter within a receding horizon fashion. A variational formulation of the power maximization problem is adapted to solve the optimal control problem. The optimal control method is shown to be of a bangbang type for a power takeoff mechanism that incorporates both linear dampers and active control elements. We also consider a direct transcription of the optimal control problem as a general nonlinear program. A variation of the projected gradient optimization scheme is formulated and shown to be feasible and computationally inexpensive compared to a standard NLP solver. Since the system model is bilinear and the cost function is nonconvex quadratic, the resulting optimization problem is not a convex quadratic program. Results will be compared with an optimal command latching method to demonstrate the improvement in absorbed power. Time domain simulations are generated under irregular sea conditions.


12:0012:20, Paper TuA11.7  Add to My Program 
Power Optimization for Photovoltaic MicroConverters Using Multivariable NewtonBased ExtremumSeeking 
Ghaffari, Azad  Joint Doctoral Programs between San DiegoStateUniversityandUnive 
Krstic, Miroslav  Univ. of California, San Diego 
Seshagiri, Sridhar  San Diego State Univ. 
Keywords: Energy systems, Optimization algorithms, Adaptive control
Abstract: Extremumseeking (ES) is a realtime optimization technique that has been applied to maximum power point tracking (MPPT) design for photovoltaic (PV) microconverter systems, where each PV module is coupled with its own DCDC converter. However, most existing designs are scalar, i.e., employ one ES MPPT loop around each converter, and all current designs, whether scalar or mutivariable, are gradientbased. The convergence rate of gradientbased designs depends on the Hessian, which in turn is dependent on environmental conditions such as irradiance and temperature. Consequently, when applied to large PV arrays, the variability in environmental conditions and/or PV module degradation result in nonuniform transients in the convergence to the maximum power point (MPP). Using a multivariable gradientbased ES algorithm for the entire system instead of a scalar one for each PV module, while decreasing the sensitivity to the Hessian, does not eliminate this dependence. We present a recently developed Newtonbased ES algorithm that simultaneously employs estimates of the gradient and Hessian in the peak power tracking. The convergence rate of such a design to the MPP is independent of the Hessian, with tunable transient performance that is independent of environmental conditions. We present simulation results that show the effectiveness of the proposed algorithm in comparison to multivariable gradientbased ES.
