Skip to content

Embed URL

HTTPS clone URL

Subversion checkout URL

You can clone with
or
.
Download ZIP
/*
Monte Carlo localization simulator by Daniel L. Lu
Outputs SVG images pertaining to a one-dimensional robot attempting to localize
in a circular corridor with three doors.
Copyright (c) 2013, Daniel L. Lu
All rights reserved.
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are met:
1. Redistributions of source code must retain the above copyright notice, this
list of conditions and the following disclaimer.
2. Redistributions in binary form must reproduce the above copyright notice,
this list of conditions and the following disclaimer in the documentation
and/or other materials provided with the distribution.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND
ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR
ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
(INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/
#include <iostream>
#include <cstdlib>
#include <cmath>
#include <ctime>
#include <vector>
#include <fstream>
using namespace std;
static const long double
DITHER = 10,
MOVEMENT_D_NOISE = 0.05, // pixels noise proportional to distance moved
MOVEMENT_P_NOISE = 5, // pixels noise independent of distance moved
SENSITIVITY = 500, // pixels width of gaussian for door
KIDNAP_PROBABILITY = 0.1;
static const int
L = 1000, // map size
M = 500, // number of particles
doors[] = {300, 450, 700}, // position of doors
D = 3; // number of doors
struct particle {
long double z;
long double w;
};
class robot {
public:
robot() {
srand(3);
z = 100;
for(int m=0; m<M; m++) {
particle x;
x.z = L*(rand()/(long double)RAND_MAX);
x.w = 1;
belief.push_back(x);
}
prints = 0;
t = 0;
}
void mcl(bool sensor, long double command) {
vector <particle> belief_;
svg(false, belief);
long double cumsum = 0, *integral = new long double[belief.size()], trial;
int m = 0, i, di;
z += command;
for(m=0; m<M; m++) {
motion_update(command, &(belief[m]));
}
svg(false, belief);
m = 0;
for(m=0; m<M; m++) {
/* this loop can be merged with the previous loop in normal implementations but here they are separate
because we need to output the SVG in between. */
particle x = belief.back();
sensor_update(sensor, &x);
cumsum += x.w;
integral[m] = cumsum;
belief.pop_back();
belief_.push_back(x);
}
svg(true, belief_);
for(m=0; m<M; m++) {
trial = cumsum*(rand()/(long double)RAND_MAX);
/* this part is much better done using binary search but I'm too lazy to write the 2 extra lines for that*/
for(i=0; i<M; i++) {
if(integral[i]>trial) break;
}
belief.push_back(belief_[i]);
belief[m].z = wrap(belief[m].z + noise()*DITHER);
}
svg(false, belief);
belief_.clear();
t++;
prints = 0;
delete [] integral;
}
void svg(bool showweights, vector<particle>bel) {
char filename[200];
sprintf(filename, "mcl_t_%d_%d.svg", t, prints++);
fstream fout(filename, fstream::out);
fout << "<svg xmlns=\"http://www.w3.org/2000/svg\" version=\"1.1\" width=\"" << L << "px\" height=\"160px\">" << endl;
long double maxweight = 0;
fout << "<rect x=\"0\" y=\"0\" width=\"" << L << "\" height=\"100\" style=\"fill:#eeeeee;stroke-width:0\"/>" << endl;
fout << "<rect x=\"0\" y=\"100\" width=\"" << L << "\" height=\"60\" style=\"fill:#ffffff;stroke-width:0\"/>" << endl;
for(int m=0; m<M; m++) {
if(bel[m].w > maxweight) maxweight = bel[m].w;
fout << "<line x1=\"" << bel[m].z << "\" y1=\"100\" x2=\""
<< bel[m].z << "\" y2=\"160\" style=\"stroke:#000000;opacity:0.2;stroke-width:1\"/>" << endl;
}
if(showweights) {
for(int m=0; m<M; m++) {
fout << "<line x1=\"" << bel[m].z << "\" y1=\"" << 160 - 60*bel[m].w/maxweight << "\" x2=\""
<< bel[m].z << "\" y2=\"160\" style=\"stroke:#ff0000;stroke-width:1\"/>" << endl;
}
}
fout << "<line x1=\"0\" y1=\"100\" x2=\"" << L << "\" y2=\"100\" style=\"stroke:#cccccc;stroke-width:1\"/>" << endl;
/* draw the doors */
for(int d=0; d<D; d++) {
fout << "<g transform=\"translate(" << doors[d] << ",0)\">" << endl;
fout << " <rect x=\"" << -34 << "\" y=\"10\" width=\"68\" height=\"90\" style=\"fill:#aa7060;stroke:#995544;stroke-width:2\"/>" << endl;
fout << " <circle cx=\"" << 16 << "\" cy=\"55\" r=\"6\" style=\"fill:#333333;stroke:#111111;stroke-width:2\"/>" << endl;
fout << "</g>" << endl;
}
/* draw the robot itself */
fout << "<g transform=\"translate(" << z << ",0)\">" << endl;
fout << " <rect x=\"" << -31 << "\" y=\"20\" width=\"62\" height=\"70\" rx=\"16\" style=\"fill:#5588ff;stroke:#3355cc;stroke-width:2\"/>" << endl;
fout << " <circle cx=\"" << -16 << "\" cy=\"90\" r=\"10\" style=\"fill:#555555;stroke:#333333;stroke-width:2\"/>" << endl; // left wheel
fout << " <circle cx=\"" << 16 << "\" cy=\"90\" r=\"10\" style=\"fill:#555555;stroke:#333333;stroke-width:2\"/>" << endl; // right wheel
fout << " <circle cx=\"" << -12 << "\" cy=\"32\" r=\"5\" style=\"fill:#333333;stroke:#111111;stroke-width:2\"/>" << endl; // left eye
fout << " <circle cx=\"" << 12 << "\" cy=\"32\" r=\"5\" style=\"fill:#333333;stroke:#111111;stroke-width:2\"/>" << endl; // right eye
fout << " <path d=\"M-7,43 h14 a7,-7 0 0,1 -14,0 z\" style=\"fill:#ee0000;stroke:#aa0000;stroke-width:2\"/>" << endl; // mouth
fout << "</g>" << endl;
fout << "</svg>" << endl;
}
private:
void motion_update(long double command, particle *x) {
x->z = wrap(x->z + command + command*noise()*MOVEMENT_D_NOISE + noise()*MOVEMENT_P_NOISE);
}
void sensor_update(bool sensor, particle *x) {
if(sensor) {
x->w = KIDNAP_PROBABILITY;
for(int d=0; d<D; d++) {
x->w += exp(-pow(x->z - doors[d], 2)/SENSITIVITY);
}
} else {
x->w = 1 + KIDNAP_PROBABILITY;
for(int d=0; d<D; d++) {
x->w -= exp(-pow(x->z - doors[d], 2)/SENSITIVITY);
}
}
}
inline long double noise() {
return (rand()/(long double)RAND_MAX)-(rand()/(long double)RAND_MAX);
}
inline long double wrap(long double z) {
return z - L*floor(z/L);
}
long double z;
int prints, t;
vector <particle> belief;
};
int main() {
robot A;
A.mcl(true, 200);
A.mcl(false, 250);
A.mcl(true, -100);
}
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Something went wrong with that request. Please try again.