Page 321 - Neural Network Modeling and Identification of Dynamical Systems

312 INDEX

neural network, 103, 110 Engine thrust aircraft, 216 Hidden layer neurons, 46, 60, 61,
object, 24, 106 Engine thrust control, 134, 152 132, 133, 176
performance, 145, 147, 149 Error Hidden layers, 43, 46, 108, 117,
problem, 5, 114, 124 covariance, 67, 68 118, 121, 125, 224
quality, 13, 117, 154, 158 covariance estimation, 67, 68 Hidden layers outputs, 101
scheme, 106, 141, 153 function, 52–59, 62, 63, 65, 67, Hybrid ANN models, 101
signal, 80, 101, 116, 131, 154, 155, 156, 178, 183, 187, Hypersonic aircraft, 135, 145
171, 183, 193, 194, 201, 189–192, 195
203, 212, 219 derivatives, 53, 59, 60, 64, 178 I
signals, 80 Hessian, 182, 191 Identification problem, 4, 5, 140,
surfaces, 149, 159, 208, 213 landscape, 187 159, 208, 213, 216, 217,
synthesis, 19, 94 value, 182 225
theory, 102, 112 gradient, 66 Incremental learning, 52, 212
variables, 17, 72, 73, 75, 77, 78, signal, 131 Incremental learning process, 72
80, 82, 96, 101, 103, 105, values, 150, 152–154 Indirect adaptive control, 25, 26
108, 173, 202, 210, 217 EXogeneous inputs, 46, 96, 98 Indirect approach, 73, 80, 103–106
vector, 17, 18, 115, 156 Extended Kalman filter (EKF), 67 Inputs, 39, 41–43, 48, 50, 51, 59, 61,
Controllability, 11, 139, 208 62, 66, 70, 85, 95–98, 101,
Controllable F 107, 224
influences, 18 ANN, 108
Feedforward network, 39–41, 43,
Controllable system, 12, 13, 15 control, 63, 85
65, 96, 97, 99, 101
Controller, 24–27, 102, 104, Feedforward neural network, 41, network, 43, 45, 60–62
116–119, 124, 140, 144, 43, 96, 99, 100, 117, 166, neurons, 45, 50
147–149, 156, 159 189, 192 Instantaneous error function, 64,
adjusting, 104, 106, 107 Flight, 12, 15, 17, 18, 25, 105, 109, 65, 178
network, 140 112, 149, 153, 157, 158, Intelligent control, 102
PI, 139 201, 204, 206, 217 International Standard
Controlling, 6, 9, 25, 93, 117, 160 aircraft, 10 Atmosphere (ISA), 218
Correcting controller (CC), 75 Interneurons, 69, 70
altitude, 160, 210, 218
Interpolation error, 185
conditions, 102, 160
D Intersubnets, 70
Deflection angle, 18, 78, 105, 108, control, 132 Inverse Hessian, 55, 56
mode, 112, 139, 158
109, 134, 157, 200, 201, path angle, 18, 217 Inverse Hessian approximation,
208, 209, 217, 225 regimes, 12, 21, 118, 139, 153, 55, 56
Derivatives 157, 158, 160, 200 Inverse Hessian matrix, 194
error function, 53, 59, 60, 64, 178
Designing control laws Functional basis (FB), 35 K
for control systems, 112 Kalman filter (KF), 68, 133
for multimode objects, 5 G
Discrete time, 19, 20, 85, 171 Gain Scheduling (GS), 25 L
instants, 63, 182 Gradient descent (GD), 53 Layered ANN model, 39, 132
state space, 170 Layered feedforward networks,
Disturbed motion, 86, 115–117, H 42, 101
160, 208 Hessian, 55, 57, 59–61, 63, 65, 178, Layered Feedforward Neural
180, 181, 183, 190, 191 Network (LFNN), 45,
E computation, 55 53, 58, 60, 64–66, 166,
Elevator actuator, 134, 150, 157 error function, 182, 191 168, 175, 176, 181
ENC optimality criterion, 121, 123 matrix, 57, 62 Learning
Engine thrust, 134, 153, 212, 218 nonsingular, 54 algorithm, 67, 77