Seminar Course 392N ● Spring2011

Lecture 3
Intelligent Energy Systems:
Control and Monitoring
Basics
Dimitry Gorinevsky

ee392n - Spring 2011 Intelligent Energy Systems 1

Stanford University
 Traditional Grid
• Worlds Largest Machine!
– 3300 utilities
– 15,000 generators, 14,000
TX substations
– 211,000 mi of HV lines
(>230kV)
• A variety of interacting
control systems

ee392n - Spring 2011 Intelligent Energy Systems 2 2

Stanford University
 Smart Energy Grid
Intelligent Energy Network
Source IPS

energy
subnet
Load IPS
Intelligent
Power Switch

Generation
Transmission
Distribution
Load
Conventional Electric Grid
Conventional Internet
ee392n - Spring 2011 Intelligent Energy Systems 3
Stanford University
 Intelligent Energy Applications
Tablet Smart
phone

Communications
Internet
Computer
Energy Application
Presentation Layer

Application Logic
Business Logic
(Intelligent Functions)

Database
ee392n - Spring 2011 Intelligent Energy Systems 4
Stanford University
 Control Function
• Control function in a systems perspective

Physical system
Measurement Control Control
System Logic Handles
Sensors Actuators

Plant

ee392n - Spring 2011 Intelligent Energy Systems 5

Stanford University
 Analysis of Control Function
• Control analysis perspective
• Goal: verification of control logic
– Simulation of the closed-loop behavior
– Theoretical analysis
Control
Logic

Measurement System Control Handle

Model Model Model

ee392n - Spring 2011 Intelligent Energy Systems 6

Stanford University
 Key Control Methods
• Control Methods
– Design patterns
– Analysis templates
• P (proportional) control
• I (integral) control
• Switching control
• Optimization
• Cascaded control design

ee392n - Spring 2011 Intelligent Energy Systems 7

Stanford University
 Generation Frequency Control
• Example control command

Controller
sensor measurements

Turbine /Generator

disturbance
Load
ee392n - Spring 2011 Intelligent Energy Systems 8
Stanford University
 Generation Frequency Control
• Simplified classic grid frequency control model
– Dynamics and Control of Electric Power Systems, G. Andersson, ETH Zurich, 2010
http://www.eeh.ee.ethz.ch/en/eeh/education/courses/viewcourse/227-0528-00l.html

Pe  Pload

Swing equation: I  Pm  Pe

  x
x  u  d Pm / I  u
 Pe / I  d
ee392n - Spring 2011 Intelligent Energy Systems 9
Stanford University
 P-control
• P (proportional) feedback control
u  kP x x  u  d
• Closed –loop dynamics x+0
x   k p x  d frequency
droop
k pt 1 k pt Step response
x  x0 e  d (1  e )
kp 1
0.8
u
• Steady state error 0.6
0.4 x(t)
0.2

xs  d s / k p 0 0 2 4

frequency droop

ee392n - Spring 2011 Intelligent Energy Systems 10

Stanford University
 AGC Control Example
• AGC = Automated Generation Control
• AGC frequency control
generation command AGC

frequency measurement

disturbance

ee392n - Spring 2011 Intelligent Energy Systems

Load 11
Stanford University
 AGC Frequency Control
• Frequency control model
x  g  u  c  l,
– x is frequency error
– cl is frequency droop for load l
– u is the generation command
• Control logic
u   k I x
– I (integral) feedback control
• This is simplified analysis

ee392n - Spring 2011 Intelligent Energy Systems 12

Stanford University
 P and I control
• P control of an integrator
d
u  k P x
x  bu  d
 b x

-k p

• I control of a gain system. The same feedback loop

cl
u   k I x
g x
x  g  u  c  l,
-k I 
ee392n - Spring 2011 Intelligent Energy Systems 13
Stanford University
 Cascade (Nested) Loops
• Inner loop has faster time scale than outer loop
• In the outer loop time scale, consider the inner loop as
a gain system that follows its setpoint input
outer loop
setpoint inner loop
(command) Outer Loop setpoint
- Control
Inner Loop
- Control

output Plant
inner loop
outer loop

ee392n - Spring 2011 Intelligent Energy Systems 14

Stanford University
 Switching (On-Off) Control
• State machine model
– Hides the continuous-time dynamics
– Continuous-time conditions for switching
• Simulation analysis
– Stateflow by Mathworks

x  69
off on
x  70 furnace
passive
setpoint cooling heating
x  71

ee392n - Spring 2011 Intelligent Energy Systems 15

Stanford University
 Optimization-based Control
• Is used in many energy applications, e.g., EMS
• Typically, LP or QP problem is solved
– Embedded logic: at each step get new data and compute
new solution
Optimization Problem
Formulation

Measured Embedded Optimizer Control

Data Solver Variables

Sensors Plant Actuators

ee392n - Spring 2011 Intelligent Energy Systems 16
Stanford University
 Cascade (Hierarchical) Control
• Hierarchical decomposition
– Cascade loop design
– Time scale separation

ee392n - Spring 2011 Intelligent Energy Systems 17

Stanford University
 Hierarchical Control Examples
• Frequency control
– I (AGC)  P (Generator)
• ADR – Automated Demand Response
– Optimization  Switching
• Energy flow control in EMS
– Optimization  PI
• Building control:
– PI  Switching
– Optimization

ee392n - Spring 2011 Intelligent Energy Systems 18

Stanford University
 Power Generation Time Scales
• Power generation and distribution
Power Supply
• Energy supply side Scheduling

1/10 1 10 100 1000 Time (s)

http://www.eeh.ee.ethz.ch/en/eeh/education/courses/viewcourse/227-0528-00l.html
ee392n - Spring 2011 Intelligent Energy Systems 19
Stanford University
 Power Demand Time Scales
• Power consumption
– DR, Homes, Buildings, Plants
• Demand side
Enterprise Demand
Scheduling
Building HVAC

Home Thermostat

Demand Response

100 1,000 10,000 Time (s)

ee392n - Spring 2011 Intelligent Energy Systems 20
Stanford University
 Research Topics: Control
• Potential topics for the term paper.
• Distribution system control and optimization
– Voltage and frequency stability
– Distributed control for Distributed Generation
– Distribution Management System: energy optimization, DR

ee392n - Spring 2011 Intelligent Energy Systems 21

Stanford University
 Monitoring & Decision Support
• Open-loop functions
- Data presentation to a user

Physical system
Measurement Monitoring Data
System & Decision Presentation
Sensors Support

Physical
system

ee392n - Spring 2011 Intelligent Energy Systems 22

Stanford University
 Monitoring Goals
• Situational awareness
– Anomaly detection
– State estimation
• Health management
– Fault isolation
– Condition based maintenances

ee392n - Spring 2011 Intelligent Energy Systems 23

Stanford University
 Condition Based Maintenance
• CBM+ Initiative

ee392n - Spring 2011 Intelligent Energy Systems 24

Stanford University
 SPC: Shewhart Control Chart
• W.Shewhart, Bell Labs, 1924
• Statistical Process Control (SPC)
• UCL = mean + 3·
• LCL = mean - 3·
Upper
Control
quality variable

Limit
Walter Shewhart
mean (1891-1967)

Lower
Control
Limit
3 6 9 12 15
sample
ee392n - Spring 2011 Intelligent Energy Systems 25
Stanford University
 Multivariable SPC
• Two correlated univariate processes
y1(t) and y2(t)  y1 
y 
cov(y1,y2) = Q, Q-1= LTL  y2 
• Uncorrelated linear combinations
 1 
z(t) = L·[y(t)-]  
 2 
z   y    Q 1  y    ~  22
2 T

• Declare fault (anomaly) if

 y   T
Q 1  y     c 2
ee392n - Spring 2011 Intelligent Energy Systems 26
Stanford University
 Multivariate SPC - Hotelling's T2
• Empirical parameter estimates
1 n
ˆ   y (t )  E  X 
n t 1
n
1
Qˆ   ( y (t )   )( y T (t )   T )  cov  y   
n t 1
Harold Hotelling
(1895-1973)
• Hotelling's T2 statistics is
T 2   y (t )    Qˆ 1  y (t )   
T

• T2 can be trended as a univariate SPC variable

ee392n - Spring 2011 Intelligent Energy Systems 27

Stanford University
 Advanced Monitoring Methods
• Estimation is dual to control
– SPC is a counterpart of switching control
• Predictive estimation – forecasting, prognostics
– Feedback update of estimates (P feedback  EWMA)
• Cascaded design
– Hierarchy of monitoring loops at different time scales
• Optimization-based methods
– Optimal estimation

ee392n - Spring 2011 Intelligent Energy Systems 28

Stanford University
 Research Topics: Monitoring
• Potential topics for the term paper.
• Asset monitoring
– Transformers
• Electric power circuit state monitoring
– Using phasor measurements
– Next chart

ee392n - Spring 2011 Intelligent Energy Systems 29

Stanford University
 Electric Power Circuit Monitoring
0  Ax  Bf  w
y  Cx  Df  v
model
Optimization Problem
State estimate
Electric • Fault isolation
Power System
Measurements:
• Currents
• Voltages
• Breakers, relays
ACC, 2009
ee392n - Spring 2011 Intelligent Energy Systems 30
Stanford University
 End of Lecture 3

ee392n - Spring 2011 Intelligent Energy Systems 31

Stanford University