drone_optimization.cpp File Reference

Drone arm center of mass optimization using PSO with high-precision verification. More...

#include <iostream>
#include <vector>
#include <cmath>
#include <memory>
#include <iomanip>
#include <cstdint>
#include <omp.h>
#include <filesystem>
#include <random>
#include <limits>
#include <functional>
#include "../montecarlo/rng/rng_global.hpp"
#include "../montecarlo/rng/rng_factory.hpp"
#include "../montecarlo/geometry.hpp"
#include "../montecarlo/integrators/MCintegrator.hpp"
#include "../montecarlo/domains/hyperrectangle.hpp"
#include "../montecarlo/domains/hypersphere.hpp"
#include "../montecarlo/domains/hypercylinder.hpp"
#include "../montecarlo/domains/polytope.hpp"
#include "../montecarlo/optimizers/PSO.hpp"
#include "../montecarlo/utils/plotter.hpp"
#include <fstream>
#include <stdexcept>

Include dependency graph for drone_optimization.cpp:

Go to the source code of this file.

Classes
class	DroneArmDomain

Functions
template<int dim>
std::vector< mc::geom::Point< dim > >	read_points_from_file_drone (const std::string &filename)
template<int dim>
void	read_normals_and_offsets_from_file_drone (const std::string &filename, std::vector< std::array< double, dim > > &normals, std::vector< double > &offsets)
uint32_t	hash_params (const std::vector< double > &params)
double	objective_function (const std::vector< double > &params)
int	main (int argc, char *argv[])

Variables
constexpr size_t	DIM = 3
	Dimensionality of the problem space (3D geometry)

Detailed Description

Drone arm center of mass optimization using PSO with high-precision verification.

This application demonstrates the use of Particle Swarm Optimization (PSO) combined with Monte Carlo integration to solve a geometric optimization problem: finding the optimal spherical hole placement in a drone arm to achieve a target center of mass position.

Problem Setup:

• Complex 3D domain: rectangular arm + cylindrical motor + optional polytope cabin
• Optimization variables: hole position [x,y,z] and radius r
• Objective: minimize distance between current CM and target (1.0, 0.0, 0.0)

Implementation Highlights:

• Deterministic RNG seeding for reproducible results across thread counts
• Ghost-hole penalty to prevent infeasible solutions
• Fast PSO with 20k samples per evaluation
• High-precision verification with 1M correlated samples
• Hybrid ground truth combining MC baseline and analytic hole subtraction

Performance:

• OpenMP parallelization for multi-core PSO
• Thread-safe objective function with parameter-based seeding
• Typical convergence: <0.1% error in ~50 iterations

Outputs:

• Optimal hole configuration printed to console
• Geometry export to drone_frames/drone_domain.txt
• Auto-generated gnuplot visualization script

See also

PSO, MontecarloIntegrator, DroneArmDomain

Definition in file drone_optimization.cpp.

Function Documentation

hash_params()

uint32_t hash_params

(

const std::vector< double > &

params

)

Definition at line 326 of file drone_optimization.cpp.

main()

int main	(	int	argc,
		char *	argv[]
	)

Definition at line 404 of file drone_optimization.cpp.

Here is the call graph for this function:

objective_function()

double objective_function

(

const std::vector< double > &

params

)

Definition at line 335 of file drone_optimization.cpp.

Here is the call graph for this function:

read_normals_and_offsets_from_file_drone()

template<int dim>

void read_normals_and_offsets_from_file_drone	(	const std::string &	filename,
		std::vector< std::array< double, dim > > &	normals,
		std::vector< double > &	offsets
	)

Definition at line 108 of file drone_optimization.cpp.

read_points_from_file_drone()

template<int dim>

std::vector< mc::geom::Point< dim > > read_points_from_file_drone

(

const std::string &

filename

)

Definition at line 67 of file drone_optimization.cpp.

Variable Documentation

DIM

constexpr size_t DIM = 3

constexpr

Dimensionality of the problem space (3D geometry)

Definition at line 61 of file drone_optimization.cpp.