line |
stmt |
bran |
cond |
sub |
pod |
time |
code |
1
|
|
|
|
|
|
|
package AI::PSO; |
2
|
|
|
|
|
|
|
|
3
|
1
|
|
|
1
|
|
24139
|
use strict; |
|
1
|
|
|
|
|
2
|
|
|
1
|
|
|
|
|
42
|
|
4
|
1
|
|
|
1
|
|
6
|
use warnings; |
|
1
|
|
|
|
|
2
|
|
|
1
|
|
|
|
|
30
|
|
5
|
1
|
|
|
1
|
|
912
|
use Math::Random; |
|
1
|
|
|
|
|
7356
|
|
|
1
|
|
|
|
|
112
|
|
6
|
1
|
|
|
1
|
|
830
|
use Callback; |
|
1
|
|
|
|
|
1534
|
|
|
1
|
|
|
|
|
2552
|
|
7
|
|
|
|
|
|
|
|
8
|
|
|
|
|
|
|
require Exporter; |
9
|
|
|
|
|
|
|
|
10
|
|
|
|
|
|
|
our @ISA = qw(Exporter); |
11
|
|
|
|
|
|
|
|
12
|
|
|
|
|
|
|
our @EXPORT = qw( |
13
|
|
|
|
|
|
|
pso_set_params |
14
|
|
|
|
|
|
|
pso_register_fitness_function |
15
|
|
|
|
|
|
|
pso_optimize |
16
|
|
|
|
|
|
|
pso_get_solution_array |
17
|
|
|
|
|
|
|
); |
18
|
|
|
|
|
|
|
|
19
|
|
|
|
|
|
|
our $VERSION = '0.86'; |
20
|
|
|
|
|
|
|
|
21
|
|
|
|
|
|
|
|
22
|
|
|
|
|
|
|
######################## BEGIN MODULE CODE ################################# |
23
|
|
|
|
|
|
|
|
24
|
|
|
|
|
|
|
#---------- BEGIN GLOBAL PARAMETERS ------------ |
25
|
|
|
|
|
|
|
|
26
|
|
|
|
|
|
|
#-#-# search parameters #-#-# |
27
|
|
|
|
|
|
|
my $numParticles = 'null'; # This is the number of particles that actually search the problem hyperspace |
28
|
|
|
|
|
|
|
my $numNeighbors = 'null'; # This is the number of neighboring particles that each particle shares information with |
29
|
|
|
|
|
|
|
# which must obviously be less than the number of particles and greater than 0. |
30
|
|
|
|
|
|
|
# TODO: write code to preconstruct different topologies. Such as fully connected, ring, star etc. |
31
|
|
|
|
|
|
|
# Currently, neighbors are chosen by a simple hash function. |
32
|
|
|
|
|
|
|
# It would be fun (no theoretical benefit that I know of) to play with different topologies. |
33
|
|
|
|
|
|
|
my $maxIterations = 'null'; # This is the maximum number of optimization iterations before exiting if the fitness goal is never reached. |
34
|
|
|
|
|
|
|
my $exitFitness = 'null'; # this is the exit criteria. It must be a value between 0 and 1. |
35
|
|
|
|
|
|
|
my $dimensions = 'null'; # this is the number of variables the user is optimizing |
36
|
|
|
|
|
|
|
|
37
|
|
|
|
|
|
|
|
38
|
|
|
|
|
|
|
#-#-# pso position parameters #-#-# |
39
|
|
|
|
|
|
|
my $deltaMin = 'null'; # This is the minimum scalar position change value when searching |
40
|
|
|
|
|
|
|
my $deltaMax = 'null'; # This is the maximum scalar position change value when searching |
41
|
|
|
|
|
|
|
|
42
|
|
|
|
|
|
|
#-#-# my 'how much do I trust myself verses my neighbors' parameters #-#-# |
43
|
|
|
|
|
|
|
my $meWeight = 'null'; # 'individuality' weighting constant (higher weight (than group) means trust individual more, neighbors less) |
44
|
|
|
|
|
|
|
my $meMin = 'null'; # 'individuality' minimum random weight (this should really be between 0, 1) |
45
|
|
|
|
|
|
|
my $meMax = 'null'; # 'individuality' maximum random weight (this should really be between 0, 1) |
46
|
|
|
|
|
|
|
my $themWeight = 'null'; # 'social' weighting constant (higher weight (than individual) means trust group more, self less) |
47
|
|
|
|
|
|
|
my $themMin = 'null'; # 'social' minimum random weight (this should really be between 0, 1) |
48
|
|
|
|
|
|
|
my $themMax = 'null'; # 'social' maximum random weight (this should really be between 0, 1) |
49
|
|
|
|
|
|
|
|
50
|
|
|
|
|
|
|
my $psoRandomRange = 'null'; # PSO::.86 new variable to support original unmodified algorithm |
51
|
|
|
|
|
|
|
my $useModifiedAlgorithm = 'null'; |
52
|
|
|
|
|
|
|
|
53
|
|
|
|
|
|
|
#-#-# user/debug parameters #-#-# |
54
|
|
|
|
|
|
|
my $verbose = 0; # This one defaults for obvious reasons... |
55
|
|
|
|
|
|
|
|
56
|
|
|
|
|
|
|
#NOTE: $meWeight and $themWeight should really add up to a constant value. |
57
|
|
|
|
|
|
|
# Swarm Intelligence defines a 'pso random range' constant and then computes two random numbers |
58
|
|
|
|
|
|
|
# within this range by first getting a random number and then subtracting it from the range. |
59
|
|
|
|
|
|
|
# e.g. |
60
|
|
|
|
|
|
|
# $randomRange = 4.0 |
61
|
|
|
|
|
|
|
# $meWeight = random(0, $randomRange); |
62
|
|
|
|
|
|
|
# $themWeight = $randomRange - $meWeight. |
63
|
|
|
|
|
|
|
# |
64
|
|
|
|
|
|
|
# |
65
|
|
|
|
|
|
|
|
66
|
|
|
|
|
|
|
#---------- END GLOBAL PARAMETERS ------------ |
67
|
|
|
|
|
|
|
|
68
|
|
|
|
|
|
|
#---------- BEGIN GLOBAL DATA STRUCTURES -------- |
69
|
|
|
|
|
|
|
# |
70
|
|
|
|
|
|
|
# a particle is a hash of arrays of positions and velocities: |
71
|
|
|
|
|
|
|
# |
72
|
|
|
|
|
|
|
# The position of a particle in the problem hyperspace is defined by the values in the position array... |
73
|
|
|
|
|
|
|
# You can think of each array value as being a dimension, |
74
|
|
|
|
|
|
|
# so in N-dimensional hyperspace, the size of the position vector is N |
75
|
|
|
|
|
|
|
# |
76
|
|
|
|
|
|
|
# A particle updates its position according the Euler integration equation for physical motion: |
77
|
|
|
|
|
|
|
# Xi(t) = Xi(t-1) + Vi(t) |
78
|
|
|
|
|
|
|
# The velocity portion of this contains the stochastic elements of PSO and is defined as: |
79
|
|
|
|
|
|
|
# Vi(t) = Vi(t-1) + P1*[pi - Xi(t-1)] + P2*[pg - Xi(t-1)] |
80
|
|
|
|
|
|
|
# where P1 and P2 add are two random values who's sum adds up to the PSO random range (4.0) |
81
|
|
|
|
|
|
|
# and pi is the individual's best location |
82
|
|
|
|
|
|
|
# and pg is the global (or neighborhoods) best position |
83
|
|
|
|
|
|
|
# |
84
|
|
|
|
|
|
|
# The velocity vector is obviously updated before the position vector... |
85
|
|
|
|
|
|
|
# |
86
|
|
|
|
|
|
|
# |
87
|
|
|
|
|
|
|
my @particles = (); |
88
|
|
|
|
|
|
|
my $user_fitness_function; |
89
|
|
|
|
|
|
|
my @solution = (); |
90
|
|
|
|
|
|
|
#---------- END GLOBAL DATA STRUCTURES -------- |
91
|
|
|
|
|
|
|
|
92
|
|
|
|
|
|
|
|
93
|
|
|
|
|
|
|
#---------- BEGIN EXPORTED SUBROUTINES ---------- |
94
|
|
|
|
|
|
|
|
95
|
|
|
|
|
|
|
# |
96
|
|
|
|
|
|
|
# pso_set_params |
97
|
|
|
|
|
|
|
# - sets the global module parameters from the hash passed in |
98
|
|
|
|
|
|
|
# |
99
|
|
|
|
|
|
|
sub pso_set_params(%) { |
100
|
2
|
|
|
2
|
1
|
310
|
my (%params) = %{$_[0]}; |
|
2
|
|
|
|
|
33
|
|
101
|
2
|
|
|
|
|
7
|
my $retval = 0; |
102
|
|
|
|
|
|
|
|
103
|
|
|
|
|
|
|
#no strict 'refs'; |
104
|
|
|
|
|
|
|
#foreach my $key (keys(%params)) { |
105
|
|
|
|
|
|
|
# $$key = $params{$key}; |
106
|
|
|
|
|
|
|
#} |
107
|
|
|
|
|
|
|
#use strict 'refs'; |
108
|
|
|
|
|
|
|
|
109
|
2
|
50
|
|
|
|
10
|
$numParticles = defined($params{numParticles}) ? $params{numParticles} : 'null'; |
110
|
2
|
50
|
|
|
|
5
|
$numNeighbors = defined($params{numNeighbors}) ? $params{numNeighbors} : 'null'; |
111
|
2
|
50
|
|
|
|
7
|
$maxIterations = defined($params{maxIterations}) ? $params{maxIterations} : 'null'; |
112
|
2
|
50
|
|
|
|
6
|
$dimensions = defined($params{dimensions}) ? $params{dimensions} : 'null'; |
113
|
2
|
50
|
|
|
|
7
|
$exitFitness = defined($params{exitFitness}) ? $params{exitFitness} : 'null'; |
114
|
2
|
50
|
|
|
|
9
|
$deltaMin = defined($params{deltaMin}) ? $params{deltaMin} : 'null'; |
115
|
2
|
50
|
|
|
|
9
|
$deltaMax = defined($params{deltaMax}) ? $params{deltaMax} : 'null'; |
116
|
2
|
50
|
|
|
|
7
|
$meWeight = defined($params{meWeight}) ? $params{meWeight} : 'null'; |
117
|
2
|
50
|
|
|
|
6
|
$meMin = defined($params{meMin}) ? $params{meMin} : 'null'; |
118
|
2
|
50
|
|
|
|
5
|
$meMax = defined($params{meMax}) ? $params{meMax} : 'null'; |
119
|
2
|
50
|
|
|
|
7
|
$themWeight = defined($params{themWeight}) ? $params{themWeight} : 'null'; |
120
|
2
|
50
|
|
|
|
7
|
$themMin = defined($params{themMin}) ? $params{themMin} : 'null'; |
121
|
2
|
50
|
|
|
|
7
|
$themMax = defined($params{themMax}) ? $params{themMax} : 'null'; |
122
|
|
|
|
|
|
|
|
123
|
2
|
100
|
|
|
|
20
|
$psoRandomRange = defined($params{psoRandomRange}) ? $params{psoRandomRange} : 'null'; |
124
|
|
|
|
|
|
|
|
125
|
2
|
50
|
|
|
|
7
|
$verbose = defined($params{verbose}) ? $params{verbose} : $verbose; |
126
|
|
|
|
|
|
|
|
127
|
2
|
|
|
|
|
4
|
my $param_string; |
128
|
2
|
100
|
|
|
|
12
|
if($psoRandomRange =~ m/null/) { |
129
|
1
|
|
|
|
|
27
|
$param_string = "$numParticles:$numNeighbors:$maxIterations:$dimensions:$exitFitness:$deltaMin:$deltaMax:$meWeight:$meMin:$meMax:$themWeight:$themMin:$themMax"; |
130
|
|
|
|
|
|
|
} else { |
131
|
1
|
|
|
|
|
11
|
$param_string = "$numParticles:$numNeighbors:$maxIterations:$dimensions:$exitFitness:$deltaMin:$deltaMax:$psoRandomRange"; |
132
|
|
|
|
|
|
|
} |
133
|
|
|
|
|
|
|
|
134
|
2
|
50
|
|
|
|
10
|
$retval = 1 if($param_string =~ m/null/); |
135
|
|
|
|
|
|
|
|
136
|
2
|
|
|
|
|
18
|
return $retval; |
137
|
|
|
|
|
|
|
} |
138
|
|
|
|
|
|
|
|
139
|
|
|
|
|
|
|
|
140
|
|
|
|
|
|
|
# |
141
|
|
|
|
|
|
|
# pso_register_fitness_function |
142
|
|
|
|
|
|
|
# - sets the user-defined callback fitness function |
143
|
|
|
|
|
|
|
# |
144
|
|
|
|
|
|
|
sub pso_register_fitness_function($) { |
145
|
2
|
|
|
2
|
1
|
5
|
my ($func) = @_; |
146
|
2
|
|
|
|
|
4
|
$user_fitness_function = new Callback(\&{"main::$func"}); |
|
2
|
|
|
|
|
20
|
|
147
|
2
|
|
|
|
|
78
|
return 0; |
148
|
|
|
|
|
|
|
} |
149
|
|
|
|
|
|
|
|
150
|
|
|
|
|
|
|
|
151
|
|
|
|
|
|
|
# |
152
|
|
|
|
|
|
|
# pso_optimize |
153
|
|
|
|
|
|
|
# - runs the particle swarm optimization algorithm |
154
|
|
|
|
|
|
|
# |
155
|
|
|
|
|
|
|
sub pso_optimize() { |
156
|
2
|
|
|
2
|
1
|
7
|
&init(); |
157
|
2
|
|
|
|
|
28
|
return &swarm(); |
158
|
|
|
|
|
|
|
} |
159
|
|
|
|
|
|
|
|
160
|
|
|
|
|
|
|
# |
161
|
|
|
|
|
|
|
# pso_get_solution_array |
162
|
|
|
|
|
|
|
# - returns the array of parameters corresponding to the best solution so far |
163
|
|
|
|
|
|
|
sub pso_get_solution_array() { |
164
|
2
|
|
|
2
|
1
|
9
|
return @solution; |
165
|
|
|
|
|
|
|
} |
166
|
|
|
|
|
|
|
|
167
|
|
|
|
|
|
|
|
168
|
|
|
|
|
|
|
#---------- END EXPORTED SUBROUTINES ---------- |
169
|
|
|
|
|
|
|
|
170
|
|
|
|
|
|
|
|
171
|
|
|
|
|
|
|
|
172
|
|
|
|
|
|
|
#--------- BEGIN INTERNAL SUBROUTINES ----------- |
173
|
|
|
|
|
|
|
|
174
|
|
|
|
|
|
|
# |
175
|
|
|
|
|
|
|
# init |
176
|
|
|
|
|
|
|
# - initializes global variables |
177
|
|
|
|
|
|
|
# - initializes particle data structures |
178
|
|
|
|
|
|
|
# |
179
|
|
|
|
|
|
|
sub init() { |
180
|
2
|
100
|
|
2
|
0
|
11
|
if($psoRandomRange =~ m/null/) { |
181
|
1
|
|
|
|
|
10
|
$useModifiedAlgorithm = 1; |
182
|
|
|
|
|
|
|
} else { |
183
|
1
|
|
|
|
|
2
|
$useModifiedAlgorithm = 0; |
184
|
|
|
|
|
|
|
} |
185
|
2
|
|
|
|
|
9
|
&initialize_particles(); |
186
|
|
|
|
|
|
|
} |
187
|
|
|
|
|
|
|
|
188
|
|
|
|
|
|
|
# |
189
|
|
|
|
|
|
|
# initialize_particles |
190
|
|
|
|
|
|
|
# - sets up internal data structures |
191
|
|
|
|
|
|
|
# - initializes particle positions and velocities with an element of randomness |
192
|
|
|
|
|
|
|
# |
193
|
|
|
|
|
|
|
sub initialize_particles() { |
194
|
2
|
|
|
2
|
0
|
7
|
for(my $p = 0; $p < $numParticles; $p++) { |
195
|
8
|
|
|
|
|
153
|
$particles[$p] = {}; # each particle is a hash of arrays with the array sizes being the dimensionality of the problem space |
196
|
8
|
|
|
|
|
32
|
$particles[$p]{nextPos} = []; # nextPos is the array of positions to move to on the next positional update |
197
|
8
|
|
|
|
|
21
|
$particles[$p]{bestPos} = []; # bestPos is the position of that has yielded the best fitness for this particle (it gets updated when a better fitness is found) |
198
|
8
|
|
|
|
|
16
|
$particles[$p]{currPos} = []; # currPos is the current position of this particle in the problem space |
199
|
8
|
|
|
|
|
15
|
$particles[$p]{velocity} = []; # velocity ... come on ... |
200
|
|
|
|
|
|
|
|
201
|
8
|
|
|
|
|
21
|
for(my $d = 0; $d < $dimensions; $d++) { |
202
|
32
|
|
|
|
|
266
|
$particles[$p]{nextPos}[$d] = &random($deltaMin, $deltaMax); |
203
|
32
|
|
|
|
|
314
|
$particles[$p]{currPos}[$d] = &random($deltaMin, $deltaMax); |
204
|
32
|
|
|
|
|
298
|
$particles[$p]{bestPos}[$d] = &random($deltaMin, $deltaMax); |
205
|
32
|
|
|
|
|
296
|
$particles[$p]{velocity}[$d] = &random($deltaMin, $deltaMax); |
206
|
|
|
|
|
|
|
} |
207
|
|
|
|
|
|
|
} |
208
|
|
|
|
|
|
|
} |
209
|
|
|
|
|
|
|
|
210
|
|
|
|
|
|
|
|
211
|
|
|
|
|
|
|
|
212
|
|
|
|
|
|
|
# |
213
|
|
|
|
|
|
|
# initialize_neighbors |
214
|
|
|
|
|
|
|
# NOTE: I made this a separate subroutine so that different topologies of neighbors can be created and used instead of this. |
215
|
|
|
|
|
|
|
# NOTE: This subroutine is currently not used because we access neighbors by index to the particle array rather than storing their references |
216
|
|
|
|
|
|
|
# |
217
|
|
|
|
|
|
|
# - adds a neighbor array to the particle hash data structure |
218
|
|
|
|
|
|
|
# - sets the neighbor based on the default neighbor hash function |
219
|
|
|
|
|
|
|
# |
220
|
|
|
|
|
|
|
sub initialize_neighbors() { |
221
|
0
|
|
|
0
|
0
|
0
|
for(my $p = 0; $p < $numParticles; $p++) { |
222
|
0
|
|
|
|
|
0
|
for(my $n = 0; $n < $numNeighbors; $n++) { |
223
|
0
|
|
|
|
|
0
|
$particles[$p]{neighbor}[$n] = $particles[&get_index_of_neighbor($p, $n)]; |
224
|
|
|
|
|
|
|
} |
225
|
|
|
|
|
|
|
} |
226
|
|
|
|
|
|
|
} |
227
|
|
|
|
|
|
|
|
228
|
|
|
|
|
|
|
|
229
|
|
|
|
|
|
|
sub dump_particle($) { |
230
|
2
|
|
|
2
|
0
|
8
|
$| = 1; |
231
|
2
|
|
|
|
|
4
|
my ($index) = @_; |
232
|
2
|
|
|
|
|
55
|
print STDERR "[particle $index]\n"; |
233
|
2
|
|
|
|
|
3
|
print STDERR "\t[bestPos] ==> " . &compute_fitness(@{$particles[$index]{bestPos}}) . "\n"; |
|
2
|
|
|
|
|
9
|
|
234
|
2
|
|
|
|
|
5
|
foreach my $pos (@{$particles[$index]{bestPos}}) { |
|
2
|
|
|
|
|
6
|
|
235
|
8
|
|
|
|
|
54
|
print STDERR "\t\t$pos\n"; |
236
|
|
|
|
|
|
|
} |
237
|
2
|
|
|
|
|
5
|
print STDERR "\t[currPos] ==> " . &compute_fitness(@{$particles[$index]{currPos}}) . "\n"; |
|
2
|
|
|
|
|
7
|
|
238
|
2
|
|
|
|
|
5
|
foreach my $pos (@{$particles[$index]{currPos}}) { |
|
2
|
|
|
|
|
6
|
|
239
|
8
|
|
|
|
|
55
|
print STDERR "\t\t$pos\n"; |
240
|
|
|
|
|
|
|
} |
241
|
2
|
|
|
|
|
6
|
print STDERR "\t[nextPos] ==> " . &compute_fitness(@{$particles[$index]{nextPos}}) . "\n"; |
|
2
|
|
|
|
|
5
|
|
242
|
2
|
|
|
|
|
5
|
foreach my $pos (@{$particles[$index]{nextPos}}) { |
|
2
|
|
|
|
|
5
|
|
243
|
8
|
|
|
|
|
63
|
print STDERR "\t\t$pos\n"; |
244
|
|
|
|
|
|
|
} |
245
|
2
|
|
|
|
|
10
|
print STDERR "\t[velocity]\n"; |
246
|
2
|
|
|
|
|
5
|
foreach my $pos (@{$particles[$index]{velocity}}) { |
|
2
|
|
|
|
|
5
|
|
247
|
8
|
|
|
|
|
48
|
print STDERR "\t\t$pos\n"; |
248
|
|
|
|
|
|
|
} |
249
|
|
|
|
|
|
|
} |
250
|
|
|
|
|
|
|
|
251
|
|
|
|
|
|
|
# |
252
|
|
|
|
|
|
|
# swarm |
253
|
|
|
|
|
|
|
# - runs the particle swarm algorithm |
254
|
|
|
|
|
|
|
# |
255
|
|
|
|
|
|
|
sub swarm() { |
256
|
2
|
|
|
2
|
0
|
8
|
for(my $iter = 0; $iter < $maxIterations; $iter++) { |
257
|
25
|
|
|
|
|
59
|
for(my $p = 0; $p < $numParticles; $p++) { |
258
|
|
|
|
|
|
|
|
259
|
|
|
|
|
|
|
## update position |
260
|
100
|
|
|
|
|
198
|
for(my $d = 0; $d < $dimensions; $d++) { |
261
|
400
|
|
|
|
|
985
|
$particles[$p]{currPos}[$d] = $particles[$p]{nextPos}[$d]; |
262
|
|
|
|
|
|
|
} |
263
|
|
|
|
|
|
|
|
264
|
|
|
|
|
|
|
## test _current_ fitness of position |
265
|
100
|
|
|
|
|
116
|
my $fitness = &compute_fitness(@{$particles[$p]{currPos}}); |
|
100
|
|
|
|
|
228
|
|
266
|
|
|
|
|
|
|
# if this position in hyperspace is the best so far... |
267
|
100
|
100
|
|
|
|
146
|
if($fitness > &compute_fitness(@{$particles[$p]{bestPos}})) { |
|
100
|
|
|
|
|
326
|
|
268
|
|
|
|
|
|
|
# for each dimension, set the best position as the current position |
269
|
16
|
|
|
|
|
37
|
for(my $d2 = 0; $d2 < $dimensions; $d2++) { |
270
|
64
|
|
|
|
|
166
|
$particles[$p]{bestPos}[$d2] = $particles[$p]{currPos}[$d2]; |
271
|
|
|
|
|
|
|
} |
272
|
|
|
|
|
|
|
} |
273
|
|
|
|
|
|
|
|
274
|
|
|
|
|
|
|
## check for exit criteria |
275
|
100
|
100
|
|
|
|
194
|
if($fitness >= $exitFitness) { |
276
|
|
|
|
|
|
|
#...write solution |
277
|
2
|
|
|
|
|
22
|
print "Y:$iter:$p:$fitness\n"; |
278
|
2
|
|
|
|
|
6
|
&save_solution(@{$particles[$p]{bestPos}}); |
|
2
|
|
|
|
|
8
|
|
279
|
2
|
|
|
|
|
7
|
&dump_particle($p); |
280
|
2
|
|
|
|
|
36
|
return 0; |
281
|
|
|
|
|
|
|
} else { |
282
|
98
|
50
|
|
|
|
177
|
if($verbose == 1) { |
283
|
98
|
|
|
|
|
1427
|
print "N:$iter:$p:$fitness\n" |
284
|
|
|
|
|
|
|
} |
285
|
98
|
50
|
|
|
|
341
|
if($verbose == 2) { |
286
|
0
|
|
|
|
|
0
|
&dump_particle($p); |
287
|
|
|
|
|
|
|
} |
288
|
|
|
|
|
|
|
} |
289
|
|
|
|
|
|
|
} |
290
|
|
|
|
|
|
|
|
291
|
|
|
|
|
|
|
## at this point we've updated our position, but haven't reached the end of the search |
292
|
|
|
|
|
|
|
## so we turn to our neighbors for help. |
293
|
|
|
|
|
|
|
## (we see if they are doing any better than we are, |
294
|
|
|
|
|
|
|
## and if so, we try to fly over closer to their position) |
295
|
|
|
|
|
|
|
|
296
|
23
|
|
|
|
|
54
|
for(my $p = 0; $p < $numParticles; $p++) { |
297
|
92
|
|
|
|
|
145
|
my $n = &get_index_of_best_fit_neighbor($p); |
298
|
92
|
|
|
|
|
137
|
my @meDelta = (); # array of self position updates |
299
|
92
|
|
|
|
|
100
|
my @themDelta = (); # array of neighbor position updates |
300
|
92
|
|
|
|
|
196
|
for(my $d = 0; $d < $dimensions; $d++) { |
301
|
368
|
100
|
|
|
|
532
|
if($useModifiedAlgorithm) { # this if shold be moved out much further, but i'm working on code refactoring first |
302
|
272
|
|
|
|
|
418
|
my $meFactor = $meWeight * &random($meMin, $meMax); |
303
|
272
|
|
|
|
|
2556
|
my $themFactor = $themWeight * &random($themMin, $themMax); |
304
|
272
|
|
|
|
|
2597
|
$meDelta[$d] = $particles[$p]{bestPos}[$d] - $particles[$p]{currPos}[$d]; |
305
|
272
|
|
|
|
|
548
|
$themDelta[$d] = $particles[$n]{bestPos}[$d] - $particles[$p]{currPos}[$d]; |
306
|
272
|
|
|
|
|
501
|
my $delta = ($meFactor * $meDelta[$d]) + ($themFactor * $themDelta[$d]); |
307
|
272
|
|
|
|
|
374
|
$delta += $particles[$p]{velocity}[$d]; |
308
|
|
|
|
|
|
|
|
309
|
|
|
|
|
|
|
# do the PSO position and velocity updates |
310
|
272
|
|
|
|
|
398
|
$particles[$p]{velocity}[$d] = &clamp_velocity($delta); |
311
|
272
|
|
|
|
|
1101
|
$particles[$p]{nextPos}[$d] = $particles[$p]{currPos}[$d] + $particles[$p]{velocity}[$d]; |
312
|
|
|
|
|
|
|
} else { |
313
|
96
|
|
|
|
|
161
|
my $rho1 = &random(0, $psoRandomRange); |
314
|
96
|
|
|
|
|
941
|
my $rho2 = $psoRandomRange - $rho1; |
315
|
96
|
|
|
|
|
218
|
$meDelta[$d] = $particles[$p]{bestPos}[$d] - $particles[$p]{currPos}[$d]; |
316
|
96
|
|
|
|
|
193
|
$themDelta[$d] = $particles[$n]{bestPos}[$d] - $particles[$p]{currPos}[$d]; |
317
|
96
|
|
|
|
|
174
|
my $delta = ($rho1 * $meDelta[$d]) + ($rho2 * $themDelta[$d]); |
318
|
96
|
|
|
|
|
138
|
$delta += $particles[$p]{velocity}[$d]; |
319
|
|
|
|
|
|
|
|
320
|
|
|
|
|
|
|
# do the PSO position and velocity updates |
321
|
96
|
|
|
|
|
166
|
$particles[$p]{velocity}[$d] = &clamp_velocity($delta); |
322
|
96
|
|
|
|
|
425
|
$particles[$p]{nextPos}[$d] = $particles[$p]{currPos}[$d] + $particles[$p]{velocity}[$d]; |
323
|
|
|
|
|
|
|
} |
324
|
|
|
|
|
|
|
} |
325
|
|
|
|
|
|
|
} |
326
|
|
|
|
|
|
|
|
327
|
|
|
|
|
|
|
} |
328
|
|
|
|
|
|
|
|
329
|
|
|
|
|
|
|
# |
330
|
|
|
|
|
|
|
# at this point we have exceeded the maximum number of iterations, so let's at least print out the best result so far |
331
|
|
|
|
|
|
|
# |
332
|
0
|
|
|
|
|
0
|
print STDERR "MAX ITERATIONS REACHED WITHOUT MEETING EXIT CRITERION...printing best solution\n"; |
333
|
0
|
|
|
|
|
0
|
my $bestFit = -1; |
334
|
0
|
|
|
|
|
0
|
my $bestPartIndex = -1; |
335
|
0
|
|
|
|
|
0
|
for(my $p = 0; $p < $numParticles; $p++) { |
336
|
0
|
|
|
|
|
0
|
my $endFit = &compute_fitness(@{$particles[$p]{bestPos}}); |
|
0
|
|
|
|
|
0
|
|
337
|
0
|
0
|
|
|
|
0
|
if($endFit >= $bestFit) { |
338
|
0
|
|
|
|
|
0
|
$bestFit = $endFit; |
339
|
0
|
|
|
|
|
0
|
$bestPartIndex = $p; |
340
|
|
|
|
|
|
|
} |
341
|
|
|
|
|
|
|
|
342
|
|
|
|
|
|
|
} |
343
|
0
|
|
|
|
|
0
|
&save_solution(@{$particles[$bestPartIndex]{bestPos}}); |
|
0
|
|
|
|
|
0
|
|
344
|
0
|
|
|
|
|
0
|
&dump_particle($bestPartIndex); |
345
|
0
|
|
|
|
|
0
|
return 1; |
346
|
|
|
|
|
|
|
} |
347
|
|
|
|
|
|
|
|
348
|
|
|
|
|
|
|
# |
349
|
|
|
|
|
|
|
# save solution |
350
|
|
|
|
|
|
|
# - simply copies the given array into the global solution array |
351
|
|
|
|
|
|
|
# |
352
|
|
|
|
|
|
|
sub save_solution(@) { |
353
|
2
|
|
|
2
|
0
|
8
|
@solution = @_; |
354
|
|
|
|
|
|
|
} |
355
|
|
|
|
|
|
|
|
356
|
|
|
|
|
|
|
|
357
|
|
|
|
|
|
|
# |
358
|
|
|
|
|
|
|
# compute_fitness |
359
|
|
|
|
|
|
|
# - computes the fitness of a particle by using the user-specified fitness function |
360
|
|
|
|
|
|
|
# |
361
|
|
|
|
|
|
|
# NOTE: I originally had a 'fitness cache' so that particles that stumbled upon the same |
362
|
|
|
|
|
|
|
# position wouldn't have to recalculate their fitness (which is often expensive). |
363
|
|
|
|
|
|
|
# However, this may be undesirable behavior for the user (if you come across the same position |
364
|
|
|
|
|
|
|
# then you may be settling in on a local maxima so you might want to randomize things and |
365
|
|
|
|
|
|
|
# keep searching. For this reason, I'm leaving the cache out. It would be trivial |
366
|
|
|
|
|
|
|
# for users to implement their own cache since they are passed the same array of values. |
367
|
|
|
|
|
|
|
# |
368
|
|
|
|
|
|
|
sub compute_fitness(@) { |
369
|
641
|
|
|
641
|
0
|
1172
|
my (@values) = @_; |
370
|
641
|
|
|
|
|
672
|
my $return_fitness = 0; |
371
|
|
|
|
|
|
|
|
372
|
|
|
|
|
|
|
# no strict 'refs'; |
373
|
|
|
|
|
|
|
# if(defined(&{"main::$user_fitness_function"})) { |
374
|
|
|
|
|
|
|
# $return_fitness = &$user_fitness_function(@values); |
375
|
|
|
|
|
|
|
# } else { |
376
|
|
|
|
|
|
|
# warn "error running user_fitness_function\n"; |
377
|
|
|
|
|
|
|
# exit 1; |
378
|
|
|
|
|
|
|
# } |
379
|
|
|
|
|
|
|
# use strict 'refs'; |
380
|
|
|
|
|
|
|
|
381
|
641
|
|
|
|
|
1640
|
$return_fitness = $user_fitness_function->call(@values); |
382
|
|
|
|
|
|
|
|
383
|
641
|
|
|
|
|
15387
|
return $return_fitness; |
384
|
|
|
|
|
|
|
} |
385
|
|
|
|
|
|
|
|
386
|
|
|
|
|
|
|
|
387
|
|
|
|
|
|
|
# |
388
|
|
|
|
|
|
|
# random |
389
|
|
|
|
|
|
|
# - returns a random number that is between the first and second arguments using the Math::Random module |
390
|
|
|
|
|
|
|
# |
391
|
|
|
|
|
|
|
sub random($$) { |
392
|
768
|
|
|
768
|
0
|
923
|
my ($min, $max) = @_; |
393
|
768
|
|
|
|
|
1719
|
return random_uniform(1, $min, $max) |
394
|
|
|
|
|
|
|
} |
395
|
|
|
|
|
|
|
|
396
|
|
|
|
|
|
|
|
397
|
|
|
|
|
|
|
# |
398
|
|
|
|
|
|
|
# get_index_of_neighbor |
399
|
|
|
|
|
|
|
# |
400
|
|
|
|
|
|
|
# - returns the index of Nth neighbor of the index for particle P |
401
|
|
|
|
|
|
|
# ==> A neighbor is one of the next K particles following P where K is the neighborhood size. |
402
|
|
|
|
|
|
|
# So, particle 1 has neighbors 2, 3, 4, 5 if K = 4. particle 4 has neighbors 5, 6, 7, 8 |
403
|
|
|
|
|
|
|
# ... |
404
|
|
|
|
|
|
|
# |
405
|
|
|
|
|
|
|
sub get_index_of_neighbor($$) { |
406
|
276
|
|
|
276
|
0
|
315
|
my ($particleIndex, $neighborNum) = @_; |
407
|
|
|
|
|
|
|
# TODO: insert error checking code / defensive programming |
408
|
276
|
|
|
|
|
451
|
return ($particleIndex + $neighborNum) % $numParticles; |
409
|
|
|
|
|
|
|
} |
410
|
|
|
|
|
|
|
|
411
|
|
|
|
|
|
|
|
412
|
|
|
|
|
|
|
# |
413
|
|
|
|
|
|
|
# get_index_of_best_fit_neighbor |
414
|
|
|
|
|
|
|
# - returns the index of the neighbor with the best fitness (when given a particle index)... |
415
|
|
|
|
|
|
|
# |
416
|
|
|
|
|
|
|
sub get_index_of_best_fit_neighbor($) { |
417
|
92
|
|
|
92
|
0
|
101
|
my ($particleIndex) = @_; |
418
|
92
|
|
|
|
|
91
|
my $bestNeighborFitness = 0; |
419
|
92
|
|
|
|
|
95
|
my $bestNeighborIndex = 0; |
420
|
92
|
|
|
|
|
95
|
my $particleNeighborIndex = 0; |
421
|
92
|
|
|
|
|
208
|
for(my $neighbor = 0; $neighbor < $numNeighbors; $neighbor++) { |
422
|
276
|
|
|
|
|
443
|
$particleNeighborIndex = &get_index_of_neighbor($particleIndex, $neighbor); |
423
|
276
|
100
|
|
|
|
303
|
if(&compute_fitness(@{$particles[$particleNeighborIndex]{bestPos}}) > $bestNeighborFitness) { |
|
276
|
|
|
|
|
673
|
|
424
|
159
|
|
|
|
|
161
|
$bestNeighborFitness = &compute_fitness(@{$particles[$particleNeighborIndex]{bestPos}}); |
|
159
|
|
|
|
|
393
|
|
425
|
159
|
|
|
|
|
457
|
$bestNeighborIndex = $particleNeighborIndex; |
426
|
|
|
|
|
|
|
} |
427
|
|
|
|
|
|
|
} |
428
|
|
|
|
|
|
|
# TODO: insert error checking code / defensive programming |
429
|
92
|
|
|
|
|
187
|
return $particleNeighborIndex; |
430
|
|
|
|
|
|
|
} |
431
|
|
|
|
|
|
|
|
432
|
|
|
|
|
|
|
# |
433
|
|
|
|
|
|
|
# clamp_velocity |
434
|
|
|
|
|
|
|
# - restricts the change in velocity to be within a certain range (prevents large jumps in problem hyperspace) |
435
|
|
|
|
|
|
|
# |
436
|
|
|
|
|
|
|
sub clamp_velocity($) { |
437
|
368
|
|
|
368
|
0
|
408
|
my ($dx) = @_; |
438
|
368
|
100
|
|
|
|
866
|
if($dx < $deltaMin) { |
|
|
100
|
|
|
|
|
|
439
|
116
|
|
|
|
|
135
|
$dx = $deltaMin; |
440
|
|
|
|
|
|
|
} elsif($dx > $deltaMax) { |
441
|
32
|
|
|
|
|
36
|
$dx = $deltaMax; |
442
|
|
|
|
|
|
|
} |
443
|
368
|
|
|
|
|
786
|
return $dx; |
444
|
|
|
|
|
|
|
} |
445
|
|
|
|
|
|
|
#--------- END INTERNAL SUBROUTINES ----------- |
446
|
|
|
|
|
|
|
|
447
|
|
|
|
|
|
|
|
448
|
|
|
|
|
|
|
1; |
449
|
|
|
|
|
|
|
######################## END MODULE CODE ################################# |
450
|
|
|
|
|
|
|
__END__ |