Graduate Exam Abstract
Yi Han
Ph.D. Final
October 29, 2014, 11:30 am - 1:30 pm
Mechanical Engineering Conference Room
Microgrid Optimization, Modelling and Control
Abstract: Abstract
Microgrid has drawn more and more attention because of the enormous benefits it can bring to traditional power system
and local communities. Microgrids are essentially modern, small-scale (electrical) power distribution systems. They can
afford benefits, such as enhancing system reliability, reducing capital investment and carbon footprint, and
diversifying energy sources. Microgrids contain several generators, whose size may range from several tens of kilowatts
to a few megawatts. They are different from traditional centralized electricity networks, which transmit vast amounts of
electrical energy across long distances at very high voltages. However, they are similar to utility scale power
distribution grids, which generate, transmit and regulate electricity to the consumer locally.
To improve the efficiency of microgrids, and to reduce fossil fuel usage and pollution, renewable energy sources are
integrated with traditional microgrids. Renewable energy sources include photovoltaic power, hydro power, geothermal
power and wind power. These are clean and abundantly available energy sources. However, because the output power of
renewable energy is strongly correlated with weather conditions, their outputs are not consistent over time. Hence,
renewable generation systems significantly impact microgrid stability, and can cause large frequency and voltage
deviations in a microgrid. In addition to the load and renewable energy fluctuations, system modelling uncertainties and
unmodelled system dynamics can have a large impact when it comes to model based controller design.
Overall, we propose to tackle these issues:
Microgrid Efficiency Improvement:
Besides increasing renewable energy content within the microgrid, system overall efficiency can be further enhanced by
optimally dispatch resources. This means optimally dispatching load to generation units within a microgrid for fuel
usage minimization. Further, if a historic load profile as well as generators’ fuel consumption curves is known, we are
able to find the best generators combination for fuel usage minimization. In this optimization problem, real-world
constraints are put into consideration as well.
Natural Gas Engine Modelling and Control:
Natural gas engines have become more and more popular because of the lower cost of natural gas and their lower
environment impact. However, the applications of natural gas engine based generation units are limited, since the longer
transport delay between fuel injection and torque makes the transient response unsuitable for grid frequency control.
As load is dispatched to generators that have different fuel usage curves, generators are running at dissimilar
operation points. They can even be turned on and off depending on the load profile and other circumstances. To handle
fast load variations, maximize renewable energy content, allow for rapid adjustment for generators operating points, and
keep Microgrid frequency deviation within the allowed range in islanded mode, advanced control approaches are needed.
Here robust control and L1 adaptive control are implemented and compared with classical control techniques. For
controller design and validation purposes, control orientated mathematical models for internal combustion engine,
storage system and renewable generation system are developed and validated. The voltage control of a synchronized
generator is realized through the closed-loop current control of the exciter. Within a microgird, internal combustion
engines are used as primary movers to rotate alternator. Hence, the internal combustion engine control is mainly focused
on grid frequency control by controlling the engine crank shaft speed.
Storage System Modelling and Control for Transient Performances Improvement:
In the proposed setup, microgrid transient performances are further improved by adding a storage system (e.g., battery).
Since the high cost of storage system, its capacity is limited. It only reacts on high frequency load fluctuation
(caused by renewable energy) or load transients with constrained power delivery capability. Its state of charge and
output power are monitored and controlled accordingly. The load variations with slower dynamics and bigger amplitude are
taken care of by internal combustion engines in the configuration. From control point of view, we would like to use
Multi-Input-Multi-Output (MIMO) control to replace multiple individual Single-Input-Single-Output (SISO) control loop in
the system. By doing so, system internal connections between different inputs and outputs are taken cared. In other
word, the controller can access more information from various inputs, and at the same time it can maneuver the multiple
actuators at the outputs. With tighter emission regulations, multiple SISO loop setup cannot longer provide satisfactory
performances, more advanced MIMO control techniques are needed.
Through this research, the methodology of dealing with systems consisting of subsystems that have inherently distinct
properties for improving overall system performances are developed. A slower system with bigger capacity is mostly
cheaper in terms of per unit deliverables. A faster system with small capacity is more expensive to deliver the same
amount of output. In a microgrid, the capital cost of per kW output power of internal combustion engine is much cheaper
than per kW output capital cost of a battery. To improve the entire system performance without significant costs
increment, we would like to separate the wheat and chaff of both systems and bring the advantages of both systems
together. In microgrids, the battery system and internal combustion engine system work collectively. We maximize the
usage of battery and avoiding any saturation or drain. At the same time we encourage the internal combustion engines to
take the big trend of load fluctuations. The developed methodology can be applied to solve problems in various
disciplinaries.
Adviser: Dr. Peter Young
Co-Adviser: N/A
Non-ECE Member: Dr. Chunk Anderson, Computer Science
Member 3: Dr. Edwin Chong, Electrical and Computer Engineering
Addional Members: Dr. Ali Pezeshki, Electrical and Computer Engineering
Publications:
1. Han, Yi and Peter Young. "Modelling of Lean Burn Natural Gas Engine." manuscript is ready for submission.
2. Han, Yi, Peter Young, and Daniel Zimmerle. "Optimal Generator Selection for Fuel Usage Minimization in Microgrids." Submitted to Elsevier on Electric Power Systems Research.
3. Han, Yi, Peter Young, Abhishek Jain and Daniel Zimmerle. "Robust Control for Microgrid Frequency Deviation Reduction with Attached Storage System." IEEE Transactions on Smart Grid DOI: 10.1109/TSG.2014.2320984.
4. Han, Yi, Peter Young, and Daniel Zimmerle. "Microgrid generation units optimum dispatch for fuel consumption minimization." Journal of Ambient Intelligence and Humanized Computing, 4, no. 6 (2013): 685-701.
5. Han, Yi, Jain, Abhishek Jain, Peter Young and Daniel ZImmerle. "Robust Control of Microgrid Frequency with Attached Storage System." In Decision and Control Conference (CDC), 2013 52nd IEEE Conference on, IEEE, 2013
6. Han, Yi, Peter Young, and Daniel Zimmerle. "Constrained optimum generator dispatch for fuel consumption minimization." In Power and Energy Society General Meeting (PES), 2013 IEEE, pp. 1-5. IEEE, 2013.
7. Han, Yi, Peter Young, and Daniel Zimmerle. "Optimal selection of generators in a microgrid for fuel usage minimization." In Power and Energy Society General Meeting (PES), 2013 IEEE, pp. 1-5. IEEE, 2013.
8. Han, Yi, Peter Young, and Daniel Zimmerle. "Optimum generation units dispatch for fuel consumption minimization." In Decision and Control and European Control Conference (CDC-ECC), 2011 50th IEEE Conference on, pp. 7206-7211. IEEE, 2011.
9. Han, Yi, R. Tzoneva, and S. Behardien. "MATLAB, LabVIEW and FPGA linear control of an inverted pendulum." In AFRICON 2007, pp. 1-7. IEEE, 2007.
Program of Study:
ECE 514
ECE 520
ECE 565
ECE 611
ECE 612
ECE 652
ECE 799
MATH 560
