A
ACE, see Area control error
Action value, 155
AGC, see Automatic generation control
AI, see Artificial intelligence
ANN, see Artificial neural networks
Approximate inference, 184
Area control error (ACE), 40
Bayesian network, 175
closed-loop system, 197
control action and, 232
dynamic controller, 28
FNN controller, 112
fuzzy logic controller, 40
genetic algorithm, 50
limits, 79
performance index, 236
PI controller, 225
PID controller, 43
power fluctuation and, 186
restriction, 234
restructured power systems, 79
RL-based controllers, 153
signal, 159
step load disturbances, 255
system frequency and, 21
Artificial intelligence (AI), 149, 175
Artificial neural networks (ANNs), 46, 95, 152, see also Neural-network-based AGC design
Automatic generation control (AGC), 11–36
Bayesian-network, see Bayesian-network-based AGC approach
blackouts, 17
communication between control units, 13
control area, 24
dynamic controller, 19
dynamic performance analysis, 26
energy management system, 11
frequency control mechanism, 20
frequency response model, 24–31
droop characteristic, 25
generation-load model, 27
generation rate constraint, 29–30
participation factor, 29
speed governor dead-band, 30
spinning reserve, 29
fundamental changes, 8
fuzzy logic, see Fuzzy logic, AGC systems and
generation rate constraint, 30
genetic algorithm, see Genetic algorithm, application of in AGC synthesis
human machine interface, 12
hydraulic amplifier, 19
intelligent, see Intelligent AGC
load-frequency control, 21
local area network, 12
multiagent systems, see Multiagent systems, AGC design using
neural network, see Neural-network-based AGC design
power system frequency control, 15–23
proportional-integral controller, 19
remote site control center, 14
remote terminal units, 11
renewable energy sources, see Renewable energy sources,; AGC systems concerning
restructured power systems, see Restructured power systems, AGC in
secondary control, 16
security control, supervisory control and data acquisition, 11
shared load shedding, 23
speed changer, 19
speed regulation, 25
summary, 35
tertiary control concept, 17
three-control area power system example, 31–35
turbine-governor dynamics, 34
under-frequency load shedding, 16
wide area network, 15
Automatic voltage regulators (AVRs), 57
AVRs, see Automatic voltage regulators
B
Battery
energy storage (BES), 38
redox flow, 59
Bayesian network (BN), 257
Bayesian-network-based AGC approach, 173–206
acyclic graph, 177
algorithm classes, 177
real-time laboratory experiment, 200–203
thirty-nine-bus test system, 195–200
approximate inference, 184
area control error, 186
artificial intelligence, 175
Bayesian networks toolbox, 194
Bayesian statistical techniques, 176
binary variables, 181
blackout phenomenon, 179
chain rule, 180
conditional probability table, 177
conditioned probability distribution, 178
digital signal processing board, 200
directed acyclic graph, 177–178
exact inference, 184
Gibbs sampling, 184
implementation methodology, 193–195
incomplete data sets, 176
joint probability distribution, 177
load-generation imbalance, 191
long-term control problem, 200
Markov assumptions, 180
Markov chain Monte Carlo, 184
Metropolis–Hastings algorithm, 184
modeling variables, 188
Monte Carlo sampling, 184
network inference, 176
observed evidence, 183
occurrence of disturbance, 192
graphical model example, 179–182
graphical models and representation, 177–179
learning, 184
performed laboratory experiment, 201
pitch angle controls, 197
posterior probability, 190
probabilistic inference, 194
probabilistic knowledge, 175
probability tables, 197
probability theory, 178
proposed intelligent control scheme, 187–192
estimation of amount of load change, 190–192
reinforcement learning, 195
renewable energy technology, 173
root node, 179
severe test scenario, 202
summary, 204
system frequency deviation, 192
system inertia, 191
system load, 183
tie-line power deviations, 185, 193
tie-line power flows, 179
variable speed wind turbines, 200
frequency control and wind turbines, 185–186
generalized ACE signal, 186–187
wind power generation, 195
wind turbine generators, 173, 185
BDI agents, see Belief-desire-intention agents
Belief-desire-intention (BDI) agents, 151
BES, see Battery energy storage
BN, see Bayesian network
BNT, see Bayesian network toolbox
C
Case study, see Intelligent power system operation and control
Centralized AGC market, 81
Chain rule, 180
Cognitive control systems, 99
Conditional probability table (CPT), 177
Conditioned probability distribution (CPD), 178
Controller, dynamic, supplementary feedback to, 28
CPD, see Conditioned probability distribution
CPT, see Conditional probability table
Crossover operator, 243
D
DAG, see Directed acyclic graph
Deliberative control, 152
Denormalization factor, 44
DFIG, see Double-fed induction generator
Diesel unit, 269
Digital signal processing (DSP) board, 200
Directed acyclic graph (DAG), 177–178
Dispersed power sources, isolated systems with, see Frequency regulation, isolated systems with dispersed power sources
DOF controller, see Dynamic output feedback controller
Double-fed induction generator (DFIG), 135, 209
DSP board, see Digital signal processing board
Dynamic output feedback (DOF) controller, 253
E
ECS, see Energy capacitor system
EMS, see Energy management system
Energy capacitor system (ECS), 51, 58, 263
Energy capacitor system, frequency regulation using, 229–240
calculated performance index, 236
control signal, restriction of, 231
current stored energy, 231
evaluation of frequency regulation performance, 236–239
excessive control action, 239
frequency regulation performance, 238
load change, 234
load-frequency control, 234
MATLAB/Simulink environment, 229
monitored area control error, 230
performance index, 236
proposed control scheme, 230–233
prevention of excessive control action (type III), 232
restriction of control action (type I), 231
restriction of control action (type II), 232
stored energy level, 235
study system, 233
summary, 239
target stored energy, 231
tie-line power change, 230
Energy management system (EMS), 79
communication, 13
restructured power systems, 79
Energy storage systems (ESSs), 55
ESSs, see Energy storage systems
Exact inference, 184
Exchanges market, 81
F
Federal Energy Regulatory Commission (FERC), 79, 82
FERC, see Federal Energy Regulatory Commission
FFC, see Flat frequency control
Fitness function, 241
Flat frequency control (FFC), 264
FLC, see Fuzzy logic controller
Flexible neural networks (FNNs), 95, 104–107
learning process, 105
momentum terms, 119
supplementary frequency controllers, 120
Flexible sigmoid functions (FSFs), 104
FNNs, see Flexible neural networks
Frequency regulation, energy capacitor system, 229–240
evaluation of frequency regulation performance, 236–239
proposed control scheme, 230–233
prevention of excessive control action (type III), 232
restriction of control action (type I), 231
restriction of control action (type II), 232
study system, 233
summary, 239
Frequency regulation, isolated systems with dispersed power sources, 263–277
automatic generation control, 263
control agents, 266
diesel unit, 269
energy capacitor systems, 263
experimental laboratory system, 267
external commercial power source, 268
feedback control system, 264, 265
flat frequency control, 264
load change scenarios, 276
monitoring agents, 266
multiagent-based AGC system, 264–266
conventional AGC on diesel unit, 264
coordinated AGC on ECS and diesel unit, 264–266
proportional-integral control loop, 264
start-up process, 268
summary, 276
supervisor agent, 266
FSFs, see Flexible sigmoid functions
Fuzzy logic, AGC systems and, 207–228
area control error, 211
centroid method, 224
double-fed induction generator, 209
fuzzy logic control block, 207
fuzzy rule base, 223
mamdani type inference system, 222
MATLAB/Simulink environment, 208, 225
membership functions, 222
nuclear units, 208
participation factor, 213
polar-information-based fuzzy logic AGC, 211–220
control of regulation margin, 217–220
polar-information-based fuzzy logic control, 211–215
trunk line power control, 215–217
power system stabilizer, 209
product method, 224
PSO-based fuzzy logic AGC, 220–226
AGC design methodology, 222–224
particle swarm optimization, 220–221
PSO algorithm for setting of membership functions, 224
control areas with subareas, 207–208
thirty-nine-bus power system, 208–211
summary, 227
switching surface, 211
tie-line power deviation, 213
tie-line power regulation, 227
wind turbines, 227
Fuzzy logic controller (FLC), 40
application, 8
parameter adjustment, 42
prototype, 7
stabilization control, 7
structure possibilities, 39
G
Generation participation matrix (GPM), 89
Generation rate constraint (GRC), 30, 244
AGC simulator, 208
automatic generation control, 30
Bayesian networks, 197
control scheme, 49
dynamic frequency response model, 244
genetic algorithm, 244
system nonlinearities, 37
Genetic algorithm, application of in AGC synthesis, 241–262
Bayesian network toolbox, 259
bilinear matrix inequalities, 254
closed-loop response, 255
condition then action, 243
conventional controllers, optimal tuning of, 244–247
application to AGC design, 249–252
multiobjective optimization, 248–249
double vector, 250
dynamic output feedback controller, 253
flowchart, 242
generation rate constraint, 244
genotypes, 249
governor set point, 258
learning algorithm, 243
mutation, 242
noisy cost functions, 259
one-point crossover, 243
control systems, 243
parameter learning, 257
Pareto-optimal solutions, 248
reinforcement learning, 257
roulette-wheel selection method, 244
speed governor dead-band, 246
static output feedback, 252
summary, 259
tracking robust performance index, 252–255
mixed H2/H∞ SOF design, 253–254
training data matrix, 258
Gibbs sampling, 184
Governor dead-band, 30
GPM, see Generation participation matrix
GRC, see Generation rate constraint
Gridwise, 61
H
High voltage direct current (HVDC) 38
HMI, see Human machine interface
Human machine interface (HMI), 12
HVDC, see High voltage direct current
Hydraulic amplifier, 19
I
IEDs, see Intelligent electronic devices
Independent power producers (IPPs), 95
Inertia constant, 132
Inference engine
definition, 40
if-then rules, 42
Intelligent AGC, 37–75, see also Automatic generation control
adaptive fuzzy controllers, 42
automatic voltage regulators, 57
backpropagation, 47
battery energy storage, 38
combined and other intelligent techniques in AGC, 51–54
defuzzification, 42
denormalization factor, 44
deregulated environment, AGC in, 54–55
distributed generation, 60
dropping system, 62
energy capacitor system, 51, 58
energy storage systems, 55
fault ride-through capability of generation resources, 57
fuel rack position control, 54
fuzzification, 42
generator governor and excitation systems, 57
genetic-algorithm-based AGC, 47–50
if-then rules, 42
inference engine, 40
definition, 40
if-then rules, 42
information technology, 65
interconnection procedures, 65
Kharitonov’s theorem, 38
knowledge management, 39
knowledge rule base, 42
learning algorithm, 44
least mean squares procedure, 47
linear matrix inequalities, 38
lookup table, 42
Lyapunov stability theory, 38
membership functions, 42
microgrids
grid codes, 65
neuro-fuzzy and neural-networks-based AGC, 44–47
normalization, 42
parametric uncertainty, 38
particle swarm optimization, 53
pole placement technique, 38
power system stabilizers, 57
proportional-integral-derivative controller, 40
protective relaying and special protection schemes, 57
Q-parameterization, 38
quantitative feedback theory, 38
redox flow batteries, 59
renewable energy options, AGC and, 55–60
new technical challenges, 57–58
present status and future prediction, 56–57
renewable energy sources, 38, 39, 55
Riccati equation approaches, 38
coordination between regulation powers of DGs/RESs and conventional generators, 64
development of effective intelligent control schemes for contribution of DGs/RESs, 64
improvement of computing techniques and measurement technologies, 64–65
improvement of modeling and analysis tools, 63
revision of existing standards, 65–66
update/definition of new grid codes, 65
updating of deregulation policies, 66
use of advanced communication and information technology, 65
self-tuning fuzzy controller, 42
simulated annealing, 39
sliding mode control, 49
solar energy, 56
structured singular value theory, 38
superconductivity magnetic energy storage, 38
system nonlinearities, 27
tie-line control, 62
transmission system operator, 54, 66
variable renewable power, 58
variable structure control, 49
vertically integrated utility, 54
virtual power plants, 62
Ziegler–Nichols tuning rules, 47
Intelligent electronic devices (IEDs), 11, 14
Intelligent power system operation and control (Japan case study), 1–9
application of intelligent methods to power systems, 2–3
application to power system control and restoration, 6–8
fault diagnosis, 6
application to power system planning, 3–6
expansion planning of distribution systems, 3–4
maintenance scheduling, 6
unit commitment, 5
automatic generation control, fundamental changes, 8
distribution systems, expansion planning of, 3
estimation error, daily power demand, 5
future implementations, 8
fuzzy logic controllers, 7
hybrid unit, scheduling types, 6
intelligent system applications
areas of, 2
objectives, 3
problems for future extension, 8
intelligent techniques, applied, 2
Japanese electric utilities, 1, 2, 9
mathematical models, 1
power demand, 5
power system stabilizer, 7
summary, 9
thermal generators, 5
weekly load forecasting, 4
Intelligrid, 61
IPPs, see Independent power producers
K
Kharitonov’s theorem, 38
Knowledge
acquisition, 6
-based expert systems, 1
Bayesian network, 173, 177, 193
domain, 151
experimental, 101
expert, 184
inference and, 182
insufficient, 37
management, 39
probabilistic, 175
rule base, 42
L
LAN, see Local area network
Learning
algorithm, step size of, 157
ANN, 96
automata algorithm, 157
Bayesian networks, 184
flexible neural networks, 105
genetic algorithm, 243, 255–259
gradient descent algorithm, 47
neural networks, 44
offline, 100
Q-, 155
reinforcement, 152
adaptive critical control, 102
genetic algorithm, 257
stochastic multistage decision problem, 53
LFC, see Load-frequency control
Linear matrix inequalities (LMIs), 38, 55
LMIs, see Linear matrix inequalities
Load-frequency control (LFC), 21, 83
automatic generation control, 21
bilateral AGC scheme, 107
deregulated environment, 54
restructured power systems, 83
simulation results, 234
Local area network (LAN), 12
Lyapunov stability theory, 38
M
Market clearing price (MCP), 85
Markov chain Monte Carlo (MCMC), 184
Markov decision process (MDP), 154
MATLAB/Simulink environment, 208, 225, 229
MCMC, see Markov chain Monte Carlo
MCP, see Market clearing price
MDP, see Markov decision process
Metropolis–Hastings algorithm, 184
Model predictive controller (MPC), 101
Monte Carlo sampling, 184
MPC, see Model predictive controller
Multiagent systems, AGC design using, 149–171
action value, 155
agent for β estimation, 165–169
artificial intelligence, 149
artificial neural networks, 152
autonomy, 150
Bayesian networks, 152
belief-desire-intention agents, 151
Bellman’s equation, 155
closed-loop system, 164
control layer, 153
deliberative control, 152
dimensionality, 149
environment, 150
exploration policy, 158
if-then rules, 151
individual fitness, 163
inference, 151
information structure constraints, 149
input variable, 158
intelligent inference, 151
interfacing layer, 153
layered agent structures, 151
learning automata algorithm, 157
learning speed, 162
Markov decision process, 154
message handling layer, 151
multiagent reinforcement-learning-based AGC, 153–161
application to thirty-nine-bus test system, 158–161
area control agent, 156
multiagent reinforcement learning, 154–155
nonlinearity, 149
problem domain, 151
processing/modeling layer, 153
proportional-integral controllers, 164
pursuit algorithm, 157
Q-function, 155
reactive control, 152
reasoning, 151
recursive least squares algorithm, 168
reinforcement learning, 152
reward matrix, 158
sliding manifold, 152
step size of learning algorithm, 157
summary, 169
symbolic representation, 151
training set, 157
uncertainty, 149
uniform probability distribution, 157
using GA to determine actions and states, 161–164
application to three-control area power system, 163–164
finding individual’s fitness and variation ranges, 162–163
value function, 154
N
NERC, see North American Electric
Reliability Council Network inference, 176
Neural-network-based AGC design, 95–122
ANN-based control systems, 97–104
ANNs in control systems, 100–104
learning and adaptation, 99–100
area control error, 112
artificial FNNs, 96
artificial neural networks, 95
backpropagation learning algorithm, 99
Bellman equation, 100
bilateral AGC scheme and modeling, 107–109
cognitive control systems, 99
direct analytic computation, 99
droop characteristic, 109
error function procedure, 105
error gradient, 106
experimental knowledge, 101
feedback control applications, 100
feedback systems, components, 99
feedforward networks, 101
flexible neural network, 95, 104–107
flexible sigmoid functions, 104
independent power producers, 95
intelligent control design, 117
intelligent systems, 99
learning algorithm 107
model predictive controller, 101
multiextremum optimization problems, 104
neoadaptive systems, 99
neuron, mathematical model of, 98
offline learning, 100
open-loop ANN, 104
perceptron, 98
plant identification, structures for, 102
proportional-integral controller, 96
recurrent networks, 101
recursive update techniques, 99
root mean squares error, 111
sampling time, 116
sigmoid function parameters, 105
sigmoid unit function, 110
simulation scenarios, system response, 116
speed governor, 109
stabilizing coefficient, 107
supplementary control unit, 96
system output error, 112
tracking errors, 113
uniform random number, 115
Normalization, 42
North American Electric Reliability Council (NERC), 21, 29, 77, 143
Nuclear plants, 30
Nuclear units
areas of operation, 233
power generation from, 208
O
Oak Ridge National Laboratory (ORNL), 79
Offline learning, 100
One-point crossover, 243
ORNL, see Oak Ridge National Laboratory
P
Pareto-optimal solutions, 248
Particle swarm optimization (PSO), 53
Perceptron, 98
PI controller, see Proportional-integral controller
PID controller, see Proportional-integral-derivative controller
Pole placement technique, 38
Pool market, 81
Power system stabilizer (PSS), 7, 57, 133
Probabilistic inference, 194
Proportional-integral (PI) controller, 31, 164
fuzzy-based, 43
multiagent systems, 159
neural networks, 96
parameters., 250
supplementary control, 19
Proportional-integral-derivative (PID) controller, 40
PSO, see Particle swarm optimization
PSS, see Power system stabilizer
Pursuit algorithm, 157
Q
Q-function, 155
Q-parameterization, 38
Quantitative feedback theory, 38
R
Reactive control, 152
Recursive least squares (RLS) algorithm, 168
Redox flow (RF) batteries, 59
Reinforcement learning, 152
adaptive critical control, 102
genetic algorithm, 257
stochastic multistage decision problem, 53
Remote terminal units (RTUs), 11
Renewable energy sources (RESs), 8, 38
Bayesian networks, 173
deregulated environment, 95
frequency regulation in interconnected networks, 263
intelligent AGC, 38
restructured power systems, 79
Renewable energy sources, AGC systems concerning, 123–148
ACE signal definition, 126
aggregation techniques, 144
central generating units, 145
double-fed induction generators, 135
drooping characteristic, 124, 130
emergency frequency control and
frequency drop, 136
frequency response analysis, 128–131
generator response, 134
governor-turbine model, 124
grid codes, 144
induction generator, 139
inertia constant, 132
intelligent system infrastructures, 145
interconnection procedures, 142
key issues and new perspectives, 142–146
further research needs, 144–146
need for revision of performance standards, 142–144
power system inertia, lack of, 137
power system stabilizer, 133
proportional-integral controller, 131
RES power variation, 123
thirty-nine-bus test system, 133–137
solar isolation, 131
standards redesign, 143
static compensator, 139
step load disturbance, 136
summary, 146
system dynamical model, Jacobian, 133
system inertia, 132
tie-line power changes, 126
turbine-governor parameters, 124
under-frequency load shedding, 138
updated AGC frequency response model, 124–128
wind generation, output, 141
wind turbines
compatibility, 143
generators, 135
pitch control, 127
wind velocity, 131
RESs, see Renewable energy sources
Restructured power systems, AGC in, 77–94
AGC configurations and frameworks, 79–84
AGC response and updated model, 86–92
AGC model and bilateral contracts, 89–91
AGC system and market operator, 86–89
need for intelligent AGC markets, 91–92
area control error, 79
availability price, 85
bilateral contracts, 88
capacity components, 85
centralized AGC market, 81
control area in new environment, 77–79
control block, 83
decentralized AGC market, 81
energy dispatch, 86
energy management system, 79
energy markets, 92
exchanges market, 81
fixed allowance, 85
frequency control, 79
generation participation matrix, 89
hierarchical scheme, 84
independent-system operators, 78
load following, 79
load-frequency control, 83
market clearing price, 85
market failure, 86
opportunity cost components, 85
participation factors, 87
payment of opportunity cost, 86
pluralistic scheme, 84
pool market, 81
price-based AGC market, 85
price for kinetic energy, 85
renewable energy sources, 79
scheduling, 79
self-procurement, 88
spot market, 88
summary, 92
synchronous area, 83
tendering, 88
transmission system operators, 83
utilization components, 85
utilization frequency payment, 86
utilization payment, 86
RF batteries, see Redox flow batteries
Riccati equation approaches, 38
RLS algorithm, see Recursive least squares algorithm
Root node, 179
Roulette-wheel selection method, 244
RTUs, see Remote terminal units
S
SA, see Simulated annealing
SCADA, see Security control, supervisory control and data acquisition
Security control, supervisory control and data acquisition (SCADA), 11
energy management system, 11
fast communication, 145
major elements, 12
power system control via, 64
renewable energy sources, 143
standards, 15
Self-tuning fuzzy controller, 42
SFPs, see Sigmoid function parameters
Sigmoid function parameters (SFPs), 105, 116
Simulated annealing (SA), 39
Sliding manifold, 152
Sliding mode control (SMC), 49
SmartGrids, 61
SMC, see Sliding mode control
SMES, see Superconductivity magnetic energy storage
SOF, see Static output feedback
Speed changer, 19
Speed governor dead-band, 30
Spinning reserve, 29
Static output feedback (SOF), 252
Structured singular value theory, 38
Superconductivity magnetic energy storage (SMES), 38
T
Target stored energy, 231
Tertiary control concept, 17
Thirty-nine-bus test system, 195–200
Tie-line power deviations, 185, 193
Transmission system operators (TSOs), 54, 83
Trunk line power control, 215–217
TSOs, see Transmission system operators
Turbine-governor dynamics, 34
U
UCTE, see Union for the Coordination of Transmission of Electricity
UFLS, see Under-frequency load shedding
Under-frequency load shedding (UFLS), 16
automatic generation control, 16
delays, 59
disconnected loads, 18
frequency decline and, 139
load-generation imbalance and, 16
L-step, 24
strategy guideline, 22
Uniform random number (URN), 115, 119, 250
Union for the Coordination of Transmission of Electricity (UCTE), 17, 22, 142
URN, see Uniform random number
V
Variable speed wind turbines (VSWTs)
wind farm, 200
Variable structure control (VSC), 49
Vertically integrated utility, 54
Virtual power plants (VPPs), 62
VPPs, see Virtual power plants
VSC, see Variable structure control
VSWTs, see Variable speed wind turbines
W
WAN, see Wide area network
WF, see Wind farm
Wide area network (WAN), 15
Wind farm (WF), 200
automatic generation control system for, 59
thirty-nine-bus test system, 133, 195
variable speed wind turbines, 200
Wind power
dynamic influence of, 59
geographical spreading of, 60
Wind turbine generators (WTGs), 185
WTGs, see Wind turbine generators
Z
Ziegler–Nichols tuning rules, 47