-
Notifications
You must be signed in to change notification settings - Fork 4
/
Copy pathfunc_train_val.m
235 lines (156 loc) · 5.26 KB
/
func_train_val.m
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
function [rmse_l, rmse_c, rmse_m, rmse_i, t] = func_train_val(n, train_t, reset_t, noise_level, bias)
dt=0.01;
input_infor={'xy', 'qdt'};
dim_in=length(input_infor) * 4;
dim_out=2;
m1=1;m2=1;
l1=0.5;l2=0.5;
lc1=0.25;lc2=0.25;
I1=0.03;I2=0.03;
properties=[m1, m2, l1, l2, lc1, lc2, I1, I2];
val_t=500;
% prepare for the reservoir computing, the hyperparameters are given by the
% Bayesian optimization.
hyperpara_set = [0.756250, 0.756250, 0.843750, -3.125, 106.71875, bias];
eig_rho = hyperpara_set(1);
W_in_a = hyperpara_set(2);
alpha = hyperpara_set(3);
beta = 10^hyperpara_set(4);
k = round( hyperpara_set(5)/200*n);
kb = hyperpara_set(6);
W_in = W_in_a*(2*rand(n,dim_in)-1);
res_net=sprandsym(n,k/n);
eig_D=eigs(res_net,1);
res_net=(eig_rho/(abs(eig_D))).*res_net;
res_net=full(res_net);
section_len=round(reset_t/dt);
washup_length=round(1002/dt);
train_length=round(train_t/dt)+round(5/dt); % 50000
val_length=round(val_t/dt); % 300
time_length = train_length + 2 * val_length + 3 * washup_length + 100;
res_infor=struct('W_in', W_in, 'res_net', res_net, 'alpha', alpha, 'kb', kb, 'beta', beta, 'n', n);
time_infor=struct('section_len', section_len, 'washup_length', washup_length, ...
'train_length', train_length, 'val_length', val_length, 'time_length', time_length);
% generate training and validation data
[xy, q, qdt, q2dt, tau] = robot_data_generator(time_infor, noise_level, dt, properties);
xy=xy(washup_length:end, :);
q=q(washup_length:end, :);
qdt=qdt(washup_length:end, :);
q2dt=q2dt(washup_length:end, :);
tau=tau(washup_length:end, :);
data_reservoir = struct('xy', xy, 'q', q, 'qdt', qdt, 'q2dt', q2dt, 'tau', tau);
clearvars q qdt q2dt tau
tic;
[Wout, r_end] = func_reservoir_train(data_reservoir, time_infor, input_infor, res_infor, dim_in, dim_out);
t = toc;
clearvars data_reservoir
% after training, we test on four trajectories: lorenz, circle, mg17 and
% eight symbol.
% lorenz
disturbance = 0.00;
measurement_noise = 0.00;
plot_movie = 0;
time_infor.val_length=150000;
traj_type = 'lorenz';
bridge_type = 'cubic';
save_rend=0;
idx=1;
if exist('traj_frequency','var') == 0
traj_frequency = 75;
end
blur.blur = 0;
failure.type = 'none';
if save_rend == 0
start_info.q=0;
start_info.qdt=0;
start_info.q2dt=0;
start_info.tau=0;
end
[control_infor, output_infor, time_infor, r_end] = func_reservoir_validate(traj_type,...
bridge_type, time_infor, input_infor, res_infor, start_info, properties, dim_in, dim_out, ...
Wout, r_end, dt, plot_movie, save_rend,failure,blur, traj_frequency);
% update
data_pred=output_infor.data_pred;
data_control=control_infor.data_control;
rmse_start_time = round(time_infor.val_length * 3/5);
rmse_end_time = time_infor.val_length - 100;
rmse_l = func_rmse(data_pred, data_control, rmse_start_time, rmse_end_time);
% circle
disturbance = 0.00;
measurement_noise = 0.00;
plot_movie = 0;
time_infor.val_length=150000;
traj_type = 'circle';
bridge_type = 'cubic';
save_rend=0;
idx=1;
blur.blur = 0;
failure.type = 'none';
if save_rend == 0
start_info.q=0;
start_info.qdt=0;
start_info.q2dt=0;
start_info.tau=0;
end
[control_infor, output_infor, time_infor, r_end] = func_reservoir_validate(traj_type,...
bridge_type, time_infor, input_infor, res_infor, start_info, properties, dim_in, dim_out, ...
Wout, r_end, dt, plot_movie, save_rend,failure,blur, traj_frequency);
% update
data_pred=output_infor.data_pred;
data_control=control_infor.data_control;
rmse_start_time = round(time_infor.val_length * 3/5);
rmse_end_time = time_infor.val_length - 100;
rmse_c = func_rmse(data_pred, data_control, rmse_start_time, rmse_end_time);
% mg17
disturbance = 0.00;
measurement_noise = 0.00;
plot_movie = 0;
time_infor.val_length=150000;
traj_type = 'mg17';
bridge_type = 'cubic';
save_rend=0;
idx=1;
blur.blur = 0;
failure.type = 'none';
if save_rend == 0
start_info.q=0;
start_info.qdt=0;
start_info.q2dt=0;
start_info.tau=0;
end
[control_infor, output_infor, time_infor, r_end] = func_reservoir_validate(traj_type,...
bridge_type, time_infor, input_infor, res_infor, start_info, properties, dim_in, dim_out, ...
Wout, r_end, dt, plot_movie, save_rend,failure,blur, traj_frequency);
% update
data_pred=output_infor.data_pred;
data_control=control_infor.data_control;
rmse_start_time = round(time_infor.val_length * 3/5);
rmse_end_time = time_infor.val_length - 100;
rmse_m = func_rmse(data_pred, data_control, rmse_start_time, rmse_end_time);
% infty
disturbance = 0.00;
measurement_noise = 0.00;
plot_movie = 0;
time_infor.val_length=150000;
traj_type = 'infty';
bridge_type = 'cubic';
save_rend=0;
idx=1;
blur.blur = 0;
failure.type = 'none';
if save_rend == 0
start_info.q=0;
start_info.qdt=0;
start_info.q2dt=0;
start_info.tau=0;
end
[control_infor, output_infor, time_infor, r_end] = func_reservoir_validate(traj_type,...
bridge_type, time_infor, input_infor, res_infor, start_info, properties, dim_in, dim_out, ...
Wout, r_end, dt, plot_movie, save_rend,failure,blur,traj_frequency);
% update
data_pred=output_infor.data_pred;
data_control=control_infor.data_control;
rmse_start_time = round(time_infor.val_length * 3/5);
rmse_end_time = time_infor.val_length - 100;
rmse_i = func_rmse(data_pred, data_control, rmse_start_time, rmse_end_time);
end