longitudinal e2e mpc

pull/1210/head
Harald Schafer 5 years ago
parent d02f55ae0b
commit 34f2c0da75
  1. 1
      SConstruct
  2. 9
      release/files_common
  3. 77
      selfdrive/controls/lib/long_mpc_model.py
  4. 2
      selfdrive/controls/lib/longitudinal_mpc_model/.gitignore
  5. 31
      selfdrive/controls/lib/longitudinal_mpc_model/SConscript
  6. 0
      selfdrive/controls/lib/longitudinal_mpc_model/__init__.py
  7. 99
      selfdrive/controls/lib/longitudinal_mpc_model/generator.cpp
  8. 212
      selfdrive/controls/lib/longitudinal_mpc_model/lib_mpc_export/acado_auxiliary_functions.c
  9. 138
      selfdrive/controls/lib/longitudinal_mpc_model/lib_mpc_export/acado_auxiliary_functions.h
  10. 342
      selfdrive/controls/lib/longitudinal_mpc_model/lib_mpc_export/acado_common.h
  11. 231
      selfdrive/controls/lib/longitudinal_mpc_model/lib_mpc_export/acado_integrator.c
  12. 70
      selfdrive/controls/lib/longitudinal_mpc_model/lib_mpc_export/acado_qpoases_interface.cpp
  13. 65
      selfdrive/controls/lib/longitudinal_mpc_model/lib_mpc_export/acado_qpoases_interface.hpp
  14. 4944
      selfdrive/controls/lib/longitudinal_mpc_model/lib_mpc_export/acado_solver.c
  15. 32
      selfdrive/controls/lib/longitudinal_mpc_model/libmpc_py.py
  16. 141
      selfdrive/controls/lib/longitudinal_mpc_model/longitudinal_mpc.c
  17. 75
      selfdrive/debug/mpc/longitudinal_mpc_model.py

@ -215,6 +215,7 @@ if arch != "Darwin":
SConscript(['selfdrive/controls/lib/cluster/SConscript']) SConscript(['selfdrive/controls/lib/cluster/SConscript'])
SConscript(['selfdrive/controls/lib/lateral_mpc/SConscript']) SConscript(['selfdrive/controls/lib/lateral_mpc/SConscript'])
SConscript(['selfdrive/controls/lib/longitudinal_mpc/SConscript']) SConscript(['selfdrive/controls/lib/longitudinal_mpc/SConscript'])
SConscript(['selfdrive/controls/lib/longitudinal_mpc_model/SConscript'])
SConscript(['selfdrive/boardd/SConscript']) SConscript(['selfdrive/boardd/SConscript'])
SConscript(['selfdrive/proclogd/SConscript']) SConscript(['selfdrive/proclogd/SConscript'])

@ -214,6 +214,7 @@ selfdrive/controls/lib/vehicle_model.py
selfdrive/controls/lib/speed_smoother.py selfdrive/controls/lib/speed_smoother.py
selfdrive/controls/lib/fcw.py selfdrive/controls/lib/fcw.py
selfdrive/controls/lib/long_mpc.py selfdrive/controls/lib/long_mpc.py
selfdrive/controls/lib/long_mpc_model.py
selfdrive/controls/lib/gps_helpers.py selfdrive/controls/lib/gps_helpers.py
selfdrive/controls/lib/cluster/* selfdrive/controls/lib/cluster/*
@ -234,6 +235,14 @@ selfdrive/controls/lib/longitudinal_mpc/generator.cpp
selfdrive/controls/lib/longitudinal_mpc/libmpc_py.py selfdrive/controls/lib/longitudinal_mpc/libmpc_py.py
selfdrive/controls/lib/longitudinal_mpc/longitudinal_mpc.c selfdrive/controls/lib/longitudinal_mpc/longitudinal_mpc.c
selfdrive/controls/lib/longitudinal_mpc_model/lib_mpc_export/*
selfdrive/controls/lib/longitudinal_mpc_model/.gitignore
selfdrive/controls/lib/longitudinal_mpc_model/SConscript
selfdrive/controls/lib/longitudinal_mpc_model/__init__.py
selfdrive/controls/lib/longitudinal_mpc_model/generator.cpp
selfdrive/controls/lib/longitudinal_mpc_model/libmpc_py.py
selfdrive/controls/lib/longitudinal_mpc_model/longitudinal_mpc.c
selfdrive/locationd/__init__.py selfdrive/locationd/__init__.py
selfdrive/locationd/.gitignore selfdrive/locationd/.gitignore
selfdrive/locationd/SConscript selfdrive/locationd/SConscript

@ -0,0 +1,77 @@
import numpy as np
import math
from selfdrive.swaglog import cloudlog
from common.realtime import sec_since_boot
from selfdrive.controls.lib.longitudinal_mpc_model import libmpc_py
class LongitudinalMpcModel():
def __init__(self):
self.setup_mpc()
self.v_mpc = 0.0
self.v_mpc_future = 0.0
self.a_mpc = 0.0
self.last_cloudlog_t = 0.0
self.ts = list(range(10))
self.valid = False
def setup_mpc(self, v_ego=0.0):
self.libmpc = libmpc_py.libmpc
self.libmpc.init(1.0, 1.0, 1.0, 1.0, 1.0)
self.libmpc.init_with_simulation(v_ego)
self.mpc_solution = libmpc_py.ffi.new("log_t *")
self.cur_state = libmpc_py.ffi.new("state_t *")
self.cur_state[0].x_ego = 0
self.cur_state[0].v_ego = 0
self.cur_state[0].a_ego = 0
def set_cur_state(self, v, a):
self.cur_state[0].x_ego = 0.0
self.cur_state[0].v_ego = v
self.cur_state[0].a_ego = a
def update(self, CS, model):
v_ego = CS.vEgo
longitudinal = model.longitudinal
if len(longitudinal.distances) == 0:
self.valid = False
return
x_poly = list(map(float, np.polyfit(self.ts, longitudinal.distances, 3)))
v_poly = list(map(float, np.polyfit(self.ts, longitudinal.speeds, 3)))
a_poly = list(map(float, np.polyfit(self.ts, longitudinal.accelerations, 3)))
# Calculate mpc
self.libmpc.run_mpc(self.cur_state, self.mpc_solution, x_poly, v_poly, a_poly)
# Get solution. MPC timestep is 0.2 s, so interpolation to 0.05 s is needed
self.v_mpc = self.mpc_solution[0].v_ego[1]
self.a_mpc = self.mpc_solution[0].a_ego[1]
self.v_mpc_future = self.mpc_solution[0].v_ego[10]
self.valid = True
# Reset if NaN or goes through lead car
nans = any(math.isnan(x) for x in self.mpc_solution[0].v_ego)
t = sec_since_boot()
if nans:
if t > self.last_cloudlog_t + 5.0:
self.last_cloudlog_t = t
cloudlog.warning("Longitudinal model mpc reset - backwards")
self.libmpc.init(1.0, 1.0, 1.0, 1.0, 1.0)
self.libmpc.init_with_simulation(v_ego)
self.cur_state[0].v_ego = v_ego
self.cur_state[0].a_ego = 0.0
self.v_mpc = v_ego
self.a_mpc = CS.aEgo
self.valid = False

@ -0,0 +1,31 @@
Import('env', 'arch')
cpp_path = [
"#phonelibs/acado/include",
"#phonelibs/acado/include/acado",
"#phonelibs/qpoases/INCLUDE",
"#phonelibs/qpoases/INCLUDE/EXTRAS",
"#phonelibs/qpoases/SRC/",
"#phonelibs/qpoases",
"lib_mpc_export"
]
mpc_files = [
"longitudinal_mpc.c",
Glob("lib_mpc_export/*.c"),
Glob("lib_mpc_export/*.cpp"),
]
interface_dir = Dir('lib_mpc_export')
SConscript(['#phonelibs/qpoases/SConscript'], variant_dir='lib_qp', exports=['interface_dir'])
env.SharedLibrary('mpc', mpc_files, LIBS=['m', 'qpoases'], LIBPATH=['lib_qp'], CPPPATH=cpp_path)
# if arch != "aarch64":
# acado_libs = [File("#phonelibs/acado/x64/lib/libacado_toolkit.a"),
# File("#phonelibs/acado/x64/lib/libacado_casadi.a"),
# File("#phonelibs/acado/x64/lib/libacado_csparse.a")]
# env.Program('generator', 'generator.cpp', LIBS=acado_libs, CPPPATH=cpp_path)

@ -0,0 +1,99 @@
#include <acado_code_generation.hpp>
const int controlHorizon = 50;
using namespace std;
int main( )
{
USING_NAMESPACE_ACADO
DifferentialEquation f;
DifferentialState x_ego, v_ego, a_ego, t;
OnlineData x_poly_r0, x_poly_r1, x_poly_r2, x_poly_r3;
OnlineData v_poly_r0, v_poly_r1, v_poly_r2, v_poly_r3;
OnlineData a_poly_r0, a_poly_r1, a_poly_r2, a_poly_r3;
Control j_ego;
// Equations of motion
f << dot(x_ego) == v_ego;
f << dot(v_ego) == a_ego;
f << dot(a_ego) == j_ego;
f << dot(t) == 1;
auto poly_x = x_poly_r0*(t*t*t) + x_poly_r1*(t*t) + x_poly_r2*t + x_poly_r3;
auto poly_v = v_poly_r0*(t*t*t) + v_poly_r1*(t*t) + v_poly_r2*t + v_poly_r3;
auto poly_a = a_poly_r0*(t*t*t) + a_poly_r1*(t*t) + a_poly_r2*t + a_poly_r3;
// Running cost
Function h;
h << x_ego - poly_x;
h << v_ego - poly_v;
h << a_ego - poly_a;
h << a_ego * (0.1 * v_ego + 1.0);
h << j_ego * (0.1 * v_ego + 1.0);
// Weights are defined in mpc.
BMatrix Q(5,5); Q.setAll(true);
// Terminal cost
Function hN;
hN << x_ego - poly_x;
hN << v_ego - poly_v;
hN << a_ego - poly_a;
hN << a_ego * (0.1 * v_ego + 1.0);
// Weights are defined in mpc.
BMatrix QN(4,4); QN.setAll(true);
// Non uniform time grid
// First 5 timesteps are 0.2, after that it's 0.6
DMatrix numSteps(20, 1);
for (int i = 0; i < 5; i++){
numSteps(i) = 1;
}
for (int i = 5; i < 20; i++){
numSteps(i) = 3;
}
// Setup Optimal Control Problem
const double tStart = 0.0;
const double tEnd = 10.0;
OCP ocp( tStart, tEnd, numSteps);
ocp.subjectTo(f);
ocp.minimizeLSQ(Q, h);
ocp.minimizeLSQEndTerm(QN, hN);
//ocp.subjectTo( 0.0 <= v_ego);
ocp.setNOD(12);
OCPexport mpc(ocp);
mpc.set( HESSIAN_APPROXIMATION, GAUSS_NEWTON );
mpc.set( DISCRETIZATION_TYPE, MULTIPLE_SHOOTING );
mpc.set( INTEGRATOR_TYPE, INT_RK4 );
mpc.set( NUM_INTEGRATOR_STEPS, controlHorizon);
mpc.set( MAX_NUM_QP_ITERATIONS, 500);
mpc.set( CG_USE_VARIABLE_WEIGHTING_MATRIX, YES);
mpc.set( SPARSE_QP_SOLUTION, CONDENSING );
mpc.set( QP_SOLVER, QP_QPOASES );
mpc.set( HOTSTART_QP, YES );
mpc.set( GENERATE_TEST_FILE, NO);
mpc.set( GENERATE_MAKE_FILE, NO );
mpc.set( GENERATE_MATLAB_INTERFACE, NO );
mpc.set( GENERATE_SIMULINK_INTERFACE, NO );
if (mpc.exportCode( "lib_mpc_export" ) != SUCCESSFUL_RETURN)
exit( EXIT_FAILURE );
mpc.printDimensionsQP( );
return EXIT_SUCCESS;
}

@ -0,0 +1,212 @@
/*
* This file was auto-generated using the ACADO Toolkit.
*
* While ACADO Toolkit is free software released under the terms of
* the GNU Lesser General Public License (LGPL), the generated code
* as such remains the property of the user who used ACADO Toolkit
* to generate this code. In particular, user dependent data of the code
* do not inherit the GNU LGPL license. On the other hand, parts of the
* generated code that are a direct copy of source code from the
* ACADO Toolkit or the software tools it is based on, remain, as derived
* work, automatically covered by the LGPL license.
*
* ACADO Toolkit is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
*
*/
#include "acado_auxiliary_functions.h"
#include <stdio.h>
real_t* acado_getVariablesX( )
{
return acadoVariables.x;
}
real_t* acado_getVariablesU( )
{
return acadoVariables.u;
}
#if ACADO_NY > 0
real_t* acado_getVariablesY( )
{
return acadoVariables.y;
}
#endif
#if ACADO_NYN > 0
real_t* acado_getVariablesYN( )
{
return acadoVariables.yN;
}
#endif
real_t* acado_getVariablesX0( )
{
#if ACADO_INITIAL_VALUE_FIXED
return acadoVariables.x0;
#else
return 0;
#endif
}
/** Print differential variables. */
void acado_printDifferentialVariables( )
{
int i, j;
printf("\nDifferential variables:\n[\n");
for (i = 0; i < ACADO_N + 1; ++i)
{
for (j = 0; j < ACADO_NX; ++j)
printf("\t%e", acadoVariables.x[i * ACADO_NX + j]);
printf("\n");
}
printf("]\n\n");
}
/** Print control variables. */
void acado_printControlVariables( )
{
int i, j;
printf("\nControl variables:\n[\n");
for (i = 0; i < ACADO_N; ++i)
{
for (j = 0; j < ACADO_NU; ++j)
printf("\t%e", acadoVariables.u[i * ACADO_NU + j]);
printf("\n");
}
printf("]\n\n");
}
/** Print ACADO code generation notice. */
void acado_printHeader( )
{
printf(
"\nACADO Toolkit -- A Toolkit for Automatic Control and Dynamic Optimization.\n"
"Copyright (C) 2008-2015 by Boris Houska, Hans Joachim Ferreau,\n"
"Milan Vukov and Rien Quirynen, KU Leuven.\n"
);
printf(
"Developed within the Optimization in Engineering Center (OPTEC) under\n"
"supervision of Moritz Diehl. All rights reserved.\n\n"
"ACADO Toolkit is distributed under the terms of the GNU Lesser\n"
"General Public License 3 in the hope that it will be useful,\n"
"but WITHOUT ANY WARRANTY; without even the implied warranty of\n"
"MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the\n"
"GNU Lesser General Public License for more details.\n\n"
);
}
#if !(defined _DSPACE)
#if (defined _WIN32 || defined _WIN64) && !(defined __MINGW32__ || defined __MINGW64__)
void acado_tic( acado_timer* t )
{
QueryPerformanceFrequency(&t->freq);
QueryPerformanceCounter(&t->tic);
}
real_t acado_toc( acado_timer* t )
{
QueryPerformanceCounter(&t->toc);
return ((t->toc.QuadPart - t->tic.QuadPart) / (real_t)t->freq.QuadPart);
}
#elif (defined __APPLE__)
void acado_tic( acado_timer* t )
{
/* read current clock cycles */
t->tic = mach_absolute_time();
}
real_t acado_toc( acado_timer* t )
{
uint64_t duration; /* elapsed time in clock cycles*/
t->toc = mach_absolute_time();
duration = t->toc - t->tic;
/*conversion from clock cycles to nanoseconds*/
mach_timebase_info(&(t->tinfo));
duration *= t->tinfo.numer;
duration /= t->tinfo.denom;
return (real_t)duration / 1e9;
}
#else
#if __STDC_VERSION__ >= 199901L
/* C99 mode */
/* read current time */
void acado_tic( acado_timer* t )
{
gettimeofday(&t->tic, 0);
}
/* return time passed since last call to tic on this timer */
real_t acado_toc( acado_timer* t )
{
struct timeval temp;
gettimeofday(&t->toc, 0);
if ((t->toc.tv_usec - t->tic.tv_usec) < 0)
{
temp.tv_sec = t->toc.tv_sec - t->tic.tv_sec - 1;
temp.tv_usec = 1000000 + t->toc.tv_usec - t->tic.tv_usec;
}
else
{
temp.tv_sec = t->toc.tv_sec - t->tic.tv_sec;
temp.tv_usec = t->toc.tv_usec - t->tic.tv_usec;
}
return (real_t)temp.tv_sec + (real_t)temp.tv_usec / 1e6;
}
#else
/* ANSI */
/* read current time */
void acado_tic( acado_timer* t )
{
clock_gettime(CLOCK_MONOTONIC, &t->tic);
}
/* return time passed since last call to tic on this timer */
real_t acado_toc( acado_timer* t )
{
struct timespec temp;
clock_gettime(CLOCK_MONOTONIC, &t->toc);
if ((t->toc.tv_nsec - t->tic.tv_nsec) < 0)
{
temp.tv_sec = t->toc.tv_sec - t->tic.tv_sec - 1;
temp.tv_nsec = 1000000000+t->toc.tv_nsec - t->tic.tv_nsec;
}
else
{
temp.tv_sec = t->toc.tv_sec - t->tic.tv_sec;
temp.tv_nsec = t->toc.tv_nsec - t->tic.tv_nsec;
}
return (real_t)temp.tv_sec + (real_t)temp.tv_nsec / 1e9;
}
#endif /* __STDC_VERSION__ >= 199901L */
#endif /* (defined _WIN32 || _WIN64) */
#endif

@ -0,0 +1,138 @@
/*
* This file was auto-generated using the ACADO Toolkit.
*
* While ACADO Toolkit is free software released under the terms of
* the GNU Lesser General Public License (LGPL), the generated code
* as such remains the property of the user who used ACADO Toolkit
* to generate this code. In particular, user dependent data of the code
* do not inherit the GNU LGPL license. On the other hand, parts of the
* generated code that are a direct copy of source code from the
* ACADO Toolkit or the software tools it is based on, remain, as derived
* work, automatically covered by the LGPL license.
*
* ACADO Toolkit is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
*
*/
#ifndef ACADO_AUXILIARY_FUNCTIONS_H
#define ACADO_AUXILIARY_FUNCTIONS_H
#include "acado_common.h"
#ifndef __MATLAB__
#ifdef __cplusplus
extern "C"
{
#endif /* __cplusplus */
#endif /* __MATLAB__ */
/** Get pointer to the matrix with differential variables. */
real_t* acado_getVariablesX( );
/** Get pointer to the matrix with control variables. */
real_t* acado_getVariablesU( );
#if ACADO_NY > 0
/** Get pointer to the matrix with references/measurements. */
real_t* acado_getVariablesY( );
#endif
#if ACADO_NYN > 0
/** Get pointer to the vector with references/measurement on the last node. */
real_t* acado_getVariablesYN( );
#endif
/** Get pointer to the current state feedback vector. Only applicable for NMPC. */
real_t* acado_getVariablesX0( );
/** Print differential variables. */
void acado_printDifferentialVariables( );
/** Print control variables. */
void acado_printControlVariables( );
/** Print ACADO code generation notice. */
void acado_printHeader( );
/*
* A huge thanks goes to Alexander Domahidi from ETHZ, Switzerland, for
* providing us with the following timing routines.
*/
#if !(defined _DSPACE)
#if (defined _WIN32 || defined _WIN64) && !(defined __MINGW32__ || defined __MINGW64__)
/* Use Windows QueryPerformanceCounter for timing. */
#include <Windows.h>
/** A structure for keeping internal timer data. */
typedef struct acado_timer_
{
LARGE_INTEGER tic;
LARGE_INTEGER toc;
LARGE_INTEGER freq;
} acado_timer;
#elif (defined __APPLE__)
#include "unistd.h"
#include <mach/mach_time.h>
/** A structure for keeping internal timer data. */
typedef struct acado_timer_
{
uint64_t tic;
uint64_t toc;
mach_timebase_info_data_t tinfo;
} acado_timer;
#else
/* Use POSIX clock_gettime() for timing on non-Windows machines. */
#include <time.h>
#if __STDC_VERSION__ >= 199901L
/* C99 mode of operation. */
#include <sys/stat.h>
#include <sys/time.h>
typedef struct acado_timer_
{
struct timeval tic;
struct timeval toc;
} acado_timer;
#else
/* ANSI C */
/** A structure for keeping internal timer data. */
typedef struct acado_timer_
{
struct timespec tic;
struct timespec toc;
} acado_timer;
#endif /* __STDC_VERSION__ >= 199901L */
#endif /* (defined _WIN32 || defined _WIN64) */
/** A function for measurement of the current time. */
void acado_tic( acado_timer* t );
/** A function which returns the elapsed time. */
real_t acado_toc( acado_timer* t );
#endif
#ifndef __MATLAB__
#ifdef __cplusplus
} /* extern "C" */
#endif /* __cplusplus */
#endif /* __MATLAB__ */
#endif /* ACADO_AUXILIARY_FUNCTIONS_H */

@ -0,0 +1,342 @@
/*
* This file was auto-generated using the ACADO Toolkit.
*
* While ACADO Toolkit is free software released under the terms of
* the GNU Lesser General Public License (LGPL), the generated code
* as such remains the property of the user who used ACADO Toolkit
* to generate this code. In particular, user dependent data of the code
* do not inherit the GNU LGPL license. On the other hand, parts of the
* generated code that are a direct copy of source code from the
* ACADO Toolkit or the software tools it is based on, remain, as derived
* work, automatically covered by the LGPL license.
*
* ACADO Toolkit is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
*
*/
#ifndef ACADO_COMMON_H
#define ACADO_COMMON_H
#include <math.h>
#include <string.h>
#ifndef __MATLAB__
#ifdef __cplusplus
extern "C"
{
#endif /* __cplusplus */
#endif /* __MATLAB__ */
/** \defgroup ACADO ACADO CGT generated module. */
/** @{ */
/** qpOASES QP solver indicator. */
#define ACADO_QPOASES 0
#define ACADO_QPOASES3 1
/** FORCES QP solver indicator.*/
#define ACADO_FORCES 2
/** qpDUNES QP solver indicator.*/
#define ACADO_QPDUNES 3
/** HPMPC QP solver indicator. */
#define ACADO_HPMPC 4
#define ACADO_GENERIC 5
/** Indicator for determining the QP solver used by the ACADO solver code. */
#define ACADO_QP_SOLVER ACADO_QPOASES
#include "acado_qpoases_interface.hpp"
/*
* Common definitions
*/
/** User defined block based condensing. */
#define ACADO_BLOCK_CONDENSING 0
/** Compute covariance matrix of the last state estimate. */
#define ACADO_COMPUTE_COVARIANCE_MATRIX 0
/** Flag indicating whether constraint values are hard-coded or not. */
#define ACADO_HARDCODED_CONSTRAINT_VALUES 1
/** Indicator for fixed initial state. */
#define ACADO_INITIAL_STATE_FIXED 1
/** Number of control/estimation intervals. */
#define ACADO_N 20
/** Number of online data values. */
#define ACADO_NOD 12
/** Number of path constraints. */
#define ACADO_NPAC 0
/** Number of control variables. */
#define ACADO_NU 1
/** Number of differential variables. */
#define ACADO_NX 4
/** Number of algebraic variables. */
#define ACADO_NXA 0
/** Number of differential derivative variables. */
#define ACADO_NXD 0
/** Number of references/measurements per node on the first N nodes. */
#define ACADO_NY 5
/** Number of references/measurements on the last (N + 1)st node. */
#define ACADO_NYN 4
/** Total number of QP optimization variables. */
#define ACADO_QP_NV 24
/** Number of Runge-Kutta stages per integration step. */
#define ACADO_RK_NSTAGES 4
/** Providing interface for arrival cost. */
#define ACADO_USE_ARRIVAL_COST 0
/** Indicator for usage of non-hard-coded linear terms in the objective. */
#define ACADO_USE_LINEAR_TERMS 0
/** Indicator for type of fixed weighting matrices. */
#define ACADO_WEIGHTING_MATRICES_TYPE 2
/*
* Globally used structure definitions
*/
/** The structure containing the user data.
*
* Via this structure the user "communicates" with the solver code.
*/
typedef struct ACADOvariables_
{
int dummy;
/** Matrix of size: 21 x 4 (row major format)
*
* Matrix containing 21 differential variable vectors.
*/
real_t x[ 84 ];
/** Column vector of size: 20
*
* Matrix containing 20 control variable vectors.
*/
real_t u[ 20 ];
/** Matrix of size: 21 x 12 (row major format)
*
* Matrix containing 21 online data vectors.
*/
real_t od[ 252 ];
/** Column vector of size: 100
*
* Matrix containing 20 reference/measurement vectors of size 5 for first 20 nodes.
*/
real_t y[ 100 ];
/** Column vector of size: 4
*
* Reference/measurement vector for the 21. node.
*/
real_t yN[ 4 ];
/** Matrix of size: 100 x 5 (row major format) */
real_t W[ 500 ];
/** Matrix of size: 4 x 4 (row major format) */
real_t WN[ 16 ];
/** Column vector of size: 4
*
* Current state feedback vector.
*/
real_t x0[ 4 ];
} ACADOvariables;
/** Private workspace used by the auto-generated code.
*
* Data members of this structure are private to the solver.
* In other words, the user code should not modify values of this
* structure.
*/
typedef struct ACADOworkspace_
{
real_t rk_ttt;
/** Row vector of size: 37 */
real_t rk_xxx[ 37 ];
/** Matrix of size: 4 x 24 (row major format) */
real_t rk_kkk[ 96 ];
/** Row vector of size: 37 */
real_t state[ 37 ];
/** Column vector of size: 80 */
real_t d[ 80 ];
/** Column vector of size: 100 */
real_t Dy[ 100 ];
/** Column vector of size: 4 */
real_t DyN[ 4 ];
/** Matrix of size: 80 x 4 (row major format) */
real_t evGx[ 320 ];
/** Column vector of size: 80 */
real_t evGu[ 80 ];
/** Row vector of size: 17 */
real_t objValueIn[ 17 ];
/** Row vector of size: 30 */
real_t objValueOut[ 30 ];
/** Matrix of size: 80 x 4 (row major format) */
real_t Q1[ 320 ];
/** Matrix of size: 80 x 5 (row major format) */
real_t Q2[ 400 ];
/** Column vector of size: 20 */
real_t R1[ 20 ];
/** Matrix of size: 20 x 5 (row major format) */
real_t R2[ 100 ];
/** Column vector of size: 80 */
real_t S1[ 80 ];
/** Matrix of size: 4 x 4 (row major format) */
real_t QN1[ 16 ];
/** Matrix of size: 4 x 4 (row major format) */
real_t QN2[ 16 ];
/** Column vector of size: 4 */
real_t Dx0[ 4 ];
/** Matrix of size: 4 x 4 (row major format) */
real_t T[ 16 ];
/** Column vector of size: 840 */
real_t E[ 840 ];
/** Column vector of size: 840 */
real_t QE[ 840 ];
/** Matrix of size: 80 x 4 (row major format) */
real_t QGx[ 320 ];
/** Column vector of size: 80 */
real_t Qd[ 80 ];
/** Column vector of size: 84 */
real_t QDy[ 84 ];
/** Matrix of size: 20 x 4 (row major format) */
real_t H10[ 80 ];
/** Matrix of size: 24 x 24 (row major format) */
real_t H[ 576 ];
/** Column vector of size: 24 */
real_t g[ 24 ];
/** Column vector of size: 24 */
real_t lb[ 24 ];
/** Column vector of size: 24 */
real_t ub[ 24 ];
/** Column vector of size: 24 */
real_t x[ 24 ];
/** Column vector of size: 24 */
real_t y[ 24 ];
} ACADOworkspace;
/*
* Forward function declarations.
*/
/** Performs the integration and sensitivity propagation for one shooting interval.
*
* \param rk_eta Working array to pass the input values and return the results.
* \param resetIntegrator The internal memory of the integrator can be reset.
* \param rk_index Number of the shooting interval.
*
* \return Status code of the integrator.
*/
int acado_integrate( real_t* const rk_eta, int resetIntegrator, int rk_index );
/** Export of an ACADO symbolic function.
*
* \param in Input to the exported function.
* \param out Output of the exported function.
*/
void acado_rhs_forw(const real_t* in, real_t* out);
/** Preparation step of the RTI scheme.
*
* \return Status of the integration module. =0: OK, otherwise the error code.
*/
int acado_preparationStep( );
/** Feedback/estimation step of the RTI scheme.
*
* \return Status code of the qpOASES QP solver.
*/
int acado_feedbackStep( );
/** Solver initialization. Must be called once before any other function call.
*
* \return =0: OK, otherwise an error code of a QP solver.
*/
int acado_initializeSolver( );
/** Initialize shooting nodes by a forward simulation starting from the first node.
*/
void acado_initializeNodesByForwardSimulation( );
/** Shift differential variables vector by one interval.
*
* \param strategy Shifting strategy: 1. Initialize node 21 with xEnd. 2. Initialize node 21 by forward simulation.
* \param xEnd Value for the x vector on the last node. If =0 the old value is used.
* \param uEnd Value for the u vector on the second to last node. If =0 the old value is used.
*/
void acado_shiftStates( int strategy, real_t* const xEnd, real_t* const uEnd );
/** Shift controls vector by one interval.
*
* \param uEnd Value for the u vector on the second to last node. If =0 the old value is used.
*/
void acado_shiftControls( real_t* const uEnd );
/** Get the KKT tolerance of the current iterate.
*
* \return The KKT tolerance value.
*/
real_t acado_getKKT( );
/** Calculate the objective value.
*
* \return Value of the objective function.
*/
real_t acado_getObjective( );
/*
* Extern declarations.
*/
extern ACADOworkspace acadoWorkspace;
extern ACADOvariables acadoVariables;
/** @} */
#ifndef __MATLAB__
#ifdef __cplusplus
} /* extern "C" */
#endif /* __cplusplus */
#endif /* __MATLAB__ */
#endif /* ACADO_COMMON_H */

@ -0,0 +1,231 @@
/*
* This file was auto-generated using the ACADO Toolkit.
*
* While ACADO Toolkit is free software released under the terms of
* the GNU Lesser General Public License (LGPL), the generated code
* as such remains the property of the user who used ACADO Toolkit
* to generate this code. In particular, user dependent data of the code
* do not inherit the GNU LGPL license. On the other hand, parts of the
* generated code that are a direct copy of source code from the
* ACADO Toolkit or the software tools it is based on, remain, as derived
* work, automatically covered by the LGPL license.
*
* ACADO Toolkit is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
*
*/
#include "acado_common.h"
void acado_rhs_forw(const real_t* in, real_t* out)
{
const real_t* xd = in;
const real_t* u = in + 24;
/* Compute outputs: */
out[0] = xd[1];
out[1] = xd[2];
out[2] = u[0];
out[3] = (real_t)(1.0000000000000000e+00);
out[4] = xd[8];
out[5] = xd[9];
out[6] = xd[10];
out[7] = xd[11];
out[8] = xd[12];
out[9] = xd[13];
out[10] = xd[14];
out[11] = xd[15];
out[12] = (real_t)(0.0000000000000000e+00);
out[13] = (real_t)(0.0000000000000000e+00);
out[14] = (real_t)(0.0000000000000000e+00);
out[15] = (real_t)(0.0000000000000000e+00);
out[16] = (real_t)(0.0000000000000000e+00);
out[17] = (real_t)(0.0000000000000000e+00);
out[18] = (real_t)(0.0000000000000000e+00);
out[19] = (real_t)(0.0000000000000000e+00);
out[20] = xd[21];
out[21] = xd[22];
out[22] = (real_t)(1.0000000000000000e+00);
out[23] = (real_t)(0.0000000000000000e+00);
}
/* Fixed step size:0.2 */
int acado_integrate( real_t* const rk_eta, int resetIntegrator, int rk_index )
{
int error;
int run1;
int numSteps[20] = {1, 1, 1, 1, 1, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3};
int numInts = numSteps[rk_index];
acadoWorkspace.rk_ttt = 0.0000000000000000e+00;
rk_eta[4] = 1.0000000000000000e+00;
rk_eta[5] = 0.0000000000000000e+00;
rk_eta[6] = 0.0000000000000000e+00;
rk_eta[7] = 0.0000000000000000e+00;
rk_eta[8] = 0.0000000000000000e+00;
rk_eta[9] = 1.0000000000000000e+00;
rk_eta[10] = 0.0000000000000000e+00;
rk_eta[11] = 0.0000000000000000e+00;
rk_eta[12] = 0.0000000000000000e+00;
rk_eta[13] = 0.0000000000000000e+00;
rk_eta[14] = 1.0000000000000000e+00;
rk_eta[15] = 0.0000000000000000e+00;
rk_eta[16] = 0.0000000000000000e+00;
rk_eta[17] = 0.0000000000000000e+00;
rk_eta[18] = 0.0000000000000000e+00;
rk_eta[19] = 1.0000000000000000e+00;
rk_eta[20] = 0.0000000000000000e+00;
rk_eta[21] = 0.0000000000000000e+00;
rk_eta[22] = 0.0000000000000000e+00;
rk_eta[23] = 0.0000000000000000e+00;
acadoWorkspace.rk_xxx[24] = rk_eta[24];
acadoWorkspace.rk_xxx[25] = rk_eta[25];
acadoWorkspace.rk_xxx[26] = rk_eta[26];
acadoWorkspace.rk_xxx[27] = rk_eta[27];
acadoWorkspace.rk_xxx[28] = rk_eta[28];
acadoWorkspace.rk_xxx[29] = rk_eta[29];
acadoWorkspace.rk_xxx[30] = rk_eta[30];
acadoWorkspace.rk_xxx[31] = rk_eta[31];
acadoWorkspace.rk_xxx[32] = rk_eta[32];
acadoWorkspace.rk_xxx[33] = rk_eta[33];
acadoWorkspace.rk_xxx[34] = rk_eta[34];
acadoWorkspace.rk_xxx[35] = rk_eta[35];
acadoWorkspace.rk_xxx[36] = rk_eta[36];
for (run1 = 0; run1 < 1; ++run1)
{
for(run1 = 0; run1 < numInts; run1++ ) {
acadoWorkspace.rk_xxx[0] = + rk_eta[0];
acadoWorkspace.rk_xxx[1] = + rk_eta[1];
acadoWorkspace.rk_xxx[2] = + rk_eta[2];
acadoWorkspace.rk_xxx[3] = + rk_eta[3];
acadoWorkspace.rk_xxx[4] = + rk_eta[4];
acadoWorkspace.rk_xxx[5] = + rk_eta[5];
acadoWorkspace.rk_xxx[6] = + rk_eta[6];
acadoWorkspace.rk_xxx[7] = + rk_eta[7];
acadoWorkspace.rk_xxx[8] = + rk_eta[8];
acadoWorkspace.rk_xxx[9] = + rk_eta[9];
acadoWorkspace.rk_xxx[10] = + rk_eta[10];
acadoWorkspace.rk_xxx[11] = + rk_eta[11];
acadoWorkspace.rk_xxx[12] = + rk_eta[12];
acadoWorkspace.rk_xxx[13] = + rk_eta[13];
acadoWorkspace.rk_xxx[14] = + rk_eta[14];
acadoWorkspace.rk_xxx[15] = + rk_eta[15];
acadoWorkspace.rk_xxx[16] = + rk_eta[16];
acadoWorkspace.rk_xxx[17] = + rk_eta[17];
acadoWorkspace.rk_xxx[18] = + rk_eta[18];
acadoWorkspace.rk_xxx[19] = + rk_eta[19];
acadoWorkspace.rk_xxx[20] = + rk_eta[20];
acadoWorkspace.rk_xxx[21] = + rk_eta[21];
acadoWorkspace.rk_xxx[22] = + rk_eta[22];
acadoWorkspace.rk_xxx[23] = + rk_eta[23];
acado_rhs_forw( acadoWorkspace.rk_xxx, acadoWorkspace.rk_kkk );
acadoWorkspace.rk_xxx[0] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[0] + rk_eta[0];
acadoWorkspace.rk_xxx[1] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[1] + rk_eta[1];
acadoWorkspace.rk_xxx[2] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[2] + rk_eta[2];
acadoWorkspace.rk_xxx[3] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[3] + rk_eta[3];
acadoWorkspace.rk_xxx[4] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[4] + rk_eta[4];
acadoWorkspace.rk_xxx[5] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[5] + rk_eta[5];
acadoWorkspace.rk_xxx[6] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[6] + rk_eta[6];
acadoWorkspace.rk_xxx[7] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[7] + rk_eta[7];
acadoWorkspace.rk_xxx[8] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[8] + rk_eta[8];
acadoWorkspace.rk_xxx[9] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[9] + rk_eta[9];
acadoWorkspace.rk_xxx[10] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[10] + rk_eta[10];
acadoWorkspace.rk_xxx[11] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[11] + rk_eta[11];
acadoWorkspace.rk_xxx[12] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[12] + rk_eta[12];
acadoWorkspace.rk_xxx[13] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[13] + rk_eta[13];
acadoWorkspace.rk_xxx[14] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[14] + rk_eta[14];
acadoWorkspace.rk_xxx[15] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[15] + rk_eta[15];
acadoWorkspace.rk_xxx[16] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[16] + rk_eta[16];
acadoWorkspace.rk_xxx[17] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[17] + rk_eta[17];
acadoWorkspace.rk_xxx[18] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[18] + rk_eta[18];
acadoWorkspace.rk_xxx[19] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[19] + rk_eta[19];
acadoWorkspace.rk_xxx[20] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[20] + rk_eta[20];
acadoWorkspace.rk_xxx[21] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[21] + rk_eta[21];
acadoWorkspace.rk_xxx[22] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[22] + rk_eta[22];
acadoWorkspace.rk_xxx[23] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[23] + rk_eta[23];
acado_rhs_forw( acadoWorkspace.rk_xxx, &(acadoWorkspace.rk_kkk[ 24 ]) );
acadoWorkspace.rk_xxx[0] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[24] + rk_eta[0];
acadoWorkspace.rk_xxx[1] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[25] + rk_eta[1];
acadoWorkspace.rk_xxx[2] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[26] + rk_eta[2];
acadoWorkspace.rk_xxx[3] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[27] + rk_eta[3];
acadoWorkspace.rk_xxx[4] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[28] + rk_eta[4];
acadoWorkspace.rk_xxx[5] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[29] + rk_eta[5];
acadoWorkspace.rk_xxx[6] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[30] + rk_eta[6];
acadoWorkspace.rk_xxx[7] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[31] + rk_eta[7];
acadoWorkspace.rk_xxx[8] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[32] + rk_eta[8];
acadoWorkspace.rk_xxx[9] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[33] + rk_eta[9];
acadoWorkspace.rk_xxx[10] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[34] + rk_eta[10];
acadoWorkspace.rk_xxx[11] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[35] + rk_eta[11];
acadoWorkspace.rk_xxx[12] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[36] + rk_eta[12];
acadoWorkspace.rk_xxx[13] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[37] + rk_eta[13];
acadoWorkspace.rk_xxx[14] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[38] + rk_eta[14];
acadoWorkspace.rk_xxx[15] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[39] + rk_eta[15];
acadoWorkspace.rk_xxx[16] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[40] + rk_eta[16];
acadoWorkspace.rk_xxx[17] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[41] + rk_eta[17];
acadoWorkspace.rk_xxx[18] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[42] + rk_eta[18];
acadoWorkspace.rk_xxx[19] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[43] + rk_eta[19];
acadoWorkspace.rk_xxx[20] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[44] + rk_eta[20];
acadoWorkspace.rk_xxx[21] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[45] + rk_eta[21];
acadoWorkspace.rk_xxx[22] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[46] + rk_eta[22];
acadoWorkspace.rk_xxx[23] = + (real_t)9.9999999999999964e-02*acadoWorkspace.rk_kkk[47] + rk_eta[23];
acado_rhs_forw( acadoWorkspace.rk_xxx, &(acadoWorkspace.rk_kkk[ 48 ]) );
acadoWorkspace.rk_xxx[0] = + (real_t)1.9999999999999993e-01*acadoWorkspace.rk_kkk[48] + rk_eta[0];
acadoWorkspace.rk_xxx[1] = + (real_t)1.9999999999999993e-01*acadoWorkspace.rk_kkk[49] + rk_eta[1];
acadoWorkspace.rk_xxx[2] = + (real_t)1.9999999999999993e-01*acadoWorkspace.rk_kkk[50] + rk_eta[2];
acadoWorkspace.rk_xxx[3] = + (real_t)1.9999999999999993e-01*acadoWorkspace.rk_kkk[51] + rk_eta[3];
acadoWorkspace.rk_xxx[4] = + (real_t)1.9999999999999993e-01*acadoWorkspace.rk_kkk[52] + rk_eta[4];
acadoWorkspace.rk_xxx[5] = + (real_t)1.9999999999999993e-01*acadoWorkspace.rk_kkk[53] + rk_eta[5];
acadoWorkspace.rk_xxx[6] = + (real_t)1.9999999999999993e-01*acadoWorkspace.rk_kkk[54] + rk_eta[6];
acadoWorkspace.rk_xxx[7] = + (real_t)1.9999999999999993e-01*acadoWorkspace.rk_kkk[55] + rk_eta[7];
acadoWorkspace.rk_xxx[8] = + (real_t)1.9999999999999993e-01*acadoWorkspace.rk_kkk[56] + rk_eta[8];
acadoWorkspace.rk_xxx[9] = + (real_t)1.9999999999999993e-01*acadoWorkspace.rk_kkk[57] + rk_eta[9];
acadoWorkspace.rk_xxx[10] = + (real_t)1.9999999999999993e-01*acadoWorkspace.rk_kkk[58] + rk_eta[10];
acadoWorkspace.rk_xxx[11] = + (real_t)1.9999999999999993e-01*acadoWorkspace.rk_kkk[59] + rk_eta[11];
acadoWorkspace.rk_xxx[12] = + (real_t)1.9999999999999993e-01*acadoWorkspace.rk_kkk[60] + rk_eta[12];
acadoWorkspace.rk_xxx[13] = + (real_t)1.9999999999999993e-01*acadoWorkspace.rk_kkk[61] + rk_eta[13];
acadoWorkspace.rk_xxx[14] = + (real_t)1.9999999999999993e-01*acadoWorkspace.rk_kkk[62] + rk_eta[14];
acadoWorkspace.rk_xxx[15] = + (real_t)1.9999999999999993e-01*acadoWorkspace.rk_kkk[63] + rk_eta[15];
acadoWorkspace.rk_xxx[16] = + (real_t)1.9999999999999993e-01*acadoWorkspace.rk_kkk[64] + rk_eta[16];
acadoWorkspace.rk_xxx[17] = + (real_t)1.9999999999999993e-01*acadoWorkspace.rk_kkk[65] + rk_eta[17];
acadoWorkspace.rk_xxx[18] = + (real_t)1.9999999999999993e-01*acadoWorkspace.rk_kkk[66] + rk_eta[18];
acadoWorkspace.rk_xxx[19] = + (real_t)1.9999999999999993e-01*acadoWorkspace.rk_kkk[67] + rk_eta[19];
acadoWorkspace.rk_xxx[20] = + (real_t)1.9999999999999993e-01*acadoWorkspace.rk_kkk[68] + rk_eta[20];
acadoWorkspace.rk_xxx[21] = + (real_t)1.9999999999999993e-01*acadoWorkspace.rk_kkk[69] + rk_eta[21];
acadoWorkspace.rk_xxx[22] = + (real_t)1.9999999999999993e-01*acadoWorkspace.rk_kkk[70] + rk_eta[22];
acadoWorkspace.rk_xxx[23] = + (real_t)1.9999999999999993e-01*acadoWorkspace.rk_kkk[71] + rk_eta[23];
acado_rhs_forw( acadoWorkspace.rk_xxx, &(acadoWorkspace.rk_kkk[ 72 ]) );
rk_eta[0] += + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[0] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[24] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[48] + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[72];
rk_eta[1] += + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[1] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[25] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[49] + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[73];
rk_eta[2] += + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[2] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[26] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[50] + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[74];
rk_eta[3] += + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[3] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[27] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[51] + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[75];
rk_eta[4] += + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[4] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[28] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[52] + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[76];
rk_eta[5] += + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[5] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[29] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[53] + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[77];
rk_eta[6] += + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[6] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[30] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[54] + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[78];
rk_eta[7] += + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[7] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[31] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[55] + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[79];
rk_eta[8] += + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[8] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[32] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[56] + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[80];
rk_eta[9] += + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[9] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[33] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[57] + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[81];
rk_eta[10] += + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[10] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[34] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[58] + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[82];
rk_eta[11] += + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[11] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[35] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[59] + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[83];
rk_eta[12] += + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[12] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[36] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[60] + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[84];
rk_eta[13] += + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[13] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[37] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[61] + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[85];
rk_eta[14] += + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[14] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[38] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[62] + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[86];
rk_eta[15] += + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[15] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[39] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[63] + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[87];
rk_eta[16] += + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[16] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[40] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[64] + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[88];
rk_eta[17] += + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[17] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[41] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[65] + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[89];
rk_eta[18] += + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[18] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[42] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[66] + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[90];
rk_eta[19] += + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[19] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[43] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[67] + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[91];
rk_eta[20] += + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[20] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[44] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[68] + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[92];
rk_eta[21] += + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[21] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[45] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[69] + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[93];
rk_eta[22] += + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[22] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[46] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[70] + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[94];
rk_eta[23] += + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[23] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[47] + (real_t)6.6666666666666638e-02*acadoWorkspace.rk_kkk[71] + (real_t)3.3333333333333319e-02*acadoWorkspace.rk_kkk[95];
acadoWorkspace.rk_ttt += 1.0000000000000000e+00;
}
}
error = 0;
return error;
}

@ -0,0 +1,70 @@
/*
* This file was auto-generated using the ACADO Toolkit.
*
* While ACADO Toolkit is free software released under the terms of
* the GNU Lesser General Public License (LGPL), the generated code
* as such remains the property of the user who used ACADO Toolkit
* to generate this code. In particular, user dependent data of the code
* do not inherit the GNU LGPL license. On the other hand, parts of the
* generated code that are a direct copy of source code from the
* ACADO Toolkit or the software tools it is based on, remain, as derived
* work, automatically covered by the LGPL license.
*
* ACADO Toolkit is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
*
*/
extern "C"
{
#include "acado_common.h"
}
#include "INCLUDE/QProblemB.hpp"
#if ACADO_COMPUTE_COVARIANCE_MATRIX == 1
#include "INCLUDE/EXTRAS/SolutionAnalysis.hpp"
#endif /* ACADO_COMPUTE_COVARIANCE_MATRIX */
static int acado_nWSR;
#if ACADO_COMPUTE_COVARIANCE_MATRIX == 1
static SolutionAnalysis acado_sa;
#endif /* ACADO_COMPUTE_COVARIANCE_MATRIX */
int acado_solve( void )
{
acado_nWSR = QPOASES_NWSRMAX;
QProblemB qp( 24 );
returnValue retVal = qp.init(acadoWorkspace.H, acadoWorkspace.g, acadoWorkspace.lb, acadoWorkspace.ub, acado_nWSR, acadoWorkspace.y);
qp.getPrimalSolution( acadoWorkspace.x );
qp.getDualSolution( acadoWorkspace.y );
#if ACADO_COMPUTE_COVARIANCE_MATRIX == 1
if (retVal != SUCCESSFUL_RETURN)
return (int)retVal;
retVal = acado_sa.getHessianInverse( &qp,var );
#endif /* ACADO_COMPUTE_COVARIANCE_MATRIX */
return (int)retVal;
}
int acado_getNWSR( void )
{
return acado_nWSR;
}
const char* acado_getErrorString( int error )
{
return MessageHandling::getErrorString( error );
}

@ -0,0 +1,65 @@
/*
* This file was auto-generated using the ACADO Toolkit.
*
* While ACADO Toolkit is free software released under the terms of
* the GNU Lesser General Public License (LGPL), the generated code
* as such remains the property of the user who used ACADO Toolkit
* to generate this code. In particular, user dependent data of the code
* do not inherit the GNU LGPL license. On the other hand, parts of the
* generated code that are a direct copy of source code from the
* ACADO Toolkit or the software tools it is based on, remain, as derived
* work, automatically covered by the LGPL license.
*
* ACADO Toolkit is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
*
*/
#ifndef QPOASES_HEADER
#define QPOASES_HEADER
#ifdef PC_DEBUG
#include <stdio.h>
#endif /* PC_DEBUG */
#include <math.h>
#ifdef __cplusplus
#define EXTERNC extern "C"
#else
#define EXTERNC
#endif
/*
* A set of options for qpOASES
*/
/** Maximum number of optimization variables. */
#define QPOASES_NVMAX 24
/** Maximum number of constraints. */
#define QPOASES_NCMAX 0
/** Maximum number of working set recalculations. */
#define QPOASES_NWSRMAX 500
/** Print level for qpOASES. */
#define QPOASES_PRINTLEVEL PL_NONE
/** The value of EPS */
#define QPOASES_EPS 2.221e-16
/** Internally used floating point type */
typedef double real_t;
/*
* Forward function declarations
*/
/** A function that calls the QP solver */
EXTERNC int acado_solve( void );
/** Get the number of active set changes */
EXTERNC int acado_getNWSR( void );
/** Get the error string. */
const char* acado_getErrorString( int error );
#endif /* QPOASES_HEADER */

@ -0,0 +1,32 @@
import os
from cffi import FFI
from common.ffi_wrapper import suffix
mpc_dir = os.path.join(os.path.dirname(os.path.abspath(__file__)))
libmpc_fn = os.path.join(mpc_dir, "libmpc"+suffix())
ffi = FFI()
ffi.cdef("""
typedef struct {
double x_ego, v_ego, a_ego;
} state_t;
typedef struct {
double x_ego[21];
double v_ego[21];
double a_ego[21];
double t[21];
double j_ego[20];
double cost;
} log_t;
void init(double xCost, double vCost, double aCost, double accelCost, double jerkCost);
void init_with_simulation(double v_ego);
int run_mpc(state_t * x0, log_t * solution, double x_poly[4], double v_poly[4], double a_poly[4]);
""")
libmpc = ffi.dlopen(libmpc_fn)

@ -0,0 +1,141 @@
#include "acado_common.h"
#include "acado_auxiliary_functions.h"
#include <stdio.h>
#include <math.h>
#define NX ACADO_NX /* Number of differential state variables. */
#define NXA ACADO_NXA /* Number of algebraic variables. */
#define NU ACADO_NU /* Number of control inputs. */
#define NOD ACADO_NOD /* Number of online data values. */
#define NY ACADO_NY /* Number of measurements/references on nodes 0..N - 1. */
#define NYN ACADO_NYN /* Number of measurements/references on node N. */
#define N ACADO_N /* Number of intervals in the horizon. */
ACADOvariables acadoVariables;
ACADOworkspace acadoWorkspace;
typedef struct {
double x_ego, v_ego, a_ego;
} state_t;
typedef struct {
double x_ego[N+1];
double v_ego[N+1];
double a_ego[N+1];
double t[N+1];
double j_ego[N];
double cost;
} log_t;
void init(double xCost, double vCost, double aCost, double accelCost, double jerkCost){
acado_initializeSolver();
int i;
const int STEP_MULTIPLIER = 3;
/* Initialize the states and controls. */
for (i = 0; i < NX * (N + 1); ++i) acadoVariables.x[ i ] = 0.0;
for (i = 0; i < NU * N; ++i) acadoVariables.u[ i ] = 0.0;
/* Initialize the measurements/reference. */
for (i = 0; i < NY * N; ++i) acadoVariables.y[ i ] = 0.0;
for (i = 0; i < NYN; ++i) acadoVariables.yN[ i ] = 0.0;
/* MPC: initialize the current state feedback. */
for (i = 0; i < NX; ++i) acadoVariables.x0[ i ] = 0.0;
// Set weights
for (i = 0; i < N; i++) {
int f = 1;
if (i > 4){
f = STEP_MULTIPLIER;
}
// Setup diagonal entries
acadoVariables.W[NY*NY*i + (NY+1)*0] = xCost * f;
acadoVariables.W[NY*NY*i + (NY+1)*1] = vCost * f;
acadoVariables.W[NY*NY*i + (NY+1)*2] = aCost * f;
acadoVariables.W[NY*NY*i + (NY+1)*3] = accelCost * f;
acadoVariables.W[NY*NY*i + (NY+1)*4] = jerkCost * f;
}
acadoVariables.WN[(NYN+1)*0] = xCost * STEP_MULTIPLIER;
acadoVariables.WN[(NYN+1)*1] = vCost * STEP_MULTIPLIER;
acadoVariables.WN[(NYN+1)*2] = aCost * STEP_MULTIPLIER;
acadoVariables.WN[(NYN+1)*3] = accelCost * STEP_MULTIPLIER;
}
void init_with_simulation(double v_ego){
int i;
double x_ego = 0.0;
double dt = 0.2;
double t = 0.0;
for (i = 0; i < N + 1; ++i){
if (i > 4){
dt = 0.6;
}
acadoVariables.x[i*NX] = x_ego;
acadoVariables.x[i*NX+1] = v_ego;
acadoVariables.x[i*NX+2] = 0;
acadoVariables.x[i*NX+3] = t;
x_ego += v_ego * dt;
t += dt;
}
for (i = 0; i < NU * N; ++i) acadoVariables.u[ i ] = 0.0;
for (i = 0; i < NY * N; ++i) acadoVariables.y[ i ] = 0.0;
for (i = 0; i < NYN; ++i) acadoVariables.yN[ i ] = 0.0;
}
int run_mpc(state_t * x0, log_t * solution,
double x_poly[4], double v_poly[4], double a_poly[4]){
int i;
for (i = 0; i < N + 1; ++i){
acadoVariables.od[i*NOD+0] = x_poly[0];
acadoVariables.od[i*NOD+1] = x_poly[1];
acadoVariables.od[i*NOD+2] = x_poly[2];
acadoVariables.od[i*NOD+3] = x_poly[3];
acadoVariables.od[i*NOD+4] = v_poly[0];
acadoVariables.od[i*NOD+5] = v_poly[1];
acadoVariables.od[i*NOD+6] = v_poly[2];
acadoVariables.od[i*NOD+7] = v_poly[3];
acadoVariables.od[i*NOD+8] = a_poly[0];
acadoVariables.od[i*NOD+9] = a_poly[1];
acadoVariables.od[i*NOD+10] = a_poly[2];
acadoVariables.od[i*NOD+11] = a_poly[3];
}
acadoVariables.x[0] = acadoVariables.x0[0] = x0->x_ego;
acadoVariables.x[1] = acadoVariables.x0[1] = x0->v_ego;
acadoVariables.x[2] = acadoVariables.x0[2] = x0->a_ego;
acadoVariables.x[3] = acadoVariables.x0[3] = 0;
acado_preparationStep();
acado_feedbackStep();
for (i = 0; i <= N; i++){
solution->x_ego[i] = acadoVariables.x[i*NX];
solution->v_ego[i] = acadoVariables.x[i*NX+1];
solution->a_ego[i] = acadoVariables.x[i*NX+2];
solution->t[i] = acadoVariables.x[i*NX+3];
if (i < N){
solution->j_ego[i] = acadoVariables.u[i];
}
}
solution->cost = acado_getObjective();
// Dont shift states here. Current solution is closer to next timestep than if
// we shift by 0.1 seconds.
return acado_getNWSR();
}

@ -0,0 +1,75 @@
#!/usr/bin/env python3
import numpy as np
import matplotlib.pyplot as plt
from selfdrive.controls.lib.longitudinal_mpc_model import libmpc_py
libmpc = libmpc_py.libmpc
dt = 1
speeds = [6.109375, 5.9765625, 6.6367188, 7.6875, 8.7578125, 9.4375, 10.21875, 11.070312, 11.679688, 12.21875]
accelerations = [0.15405273, 0.39575195, 0.36669922, 0.29248047, 0.27856445, 0.27832031, 0.29736328, 0.22705078, 0.16003418, 0.15185547]
ts = [t * dt for t in range(len(speeds))]
# TODO: Get from actual model packet
x = 0.0
positions = []
for v in speeds:
positions.append(x)
x += v * dt
# Polyfit trajectories
x_poly = list(map(float, np.polyfit(ts, positions, 3)))
v_poly = list(map(float, np.polyfit(ts, speeds, 3)))
a_poly = list(map(float, np.polyfit(ts, accelerations, 3)))
x_poly = libmpc_py.ffi.new("double[4]", x_poly)
v_poly = libmpc_py.ffi.new("double[4]", v_poly)
a_poly = libmpc_py.ffi.new("double[4]", a_poly)
cur_state = libmpc_py.ffi.new("state_t *")
cur_state[0].x_ego = 0
cur_state[0].v_ego = 10
cur_state[0].a_ego = 0
libmpc.init(1.0, 1.0, 1.0, 1.0, 1.0)
mpc_solution = libmpc_py.ffi.new("log_t *")
libmpc.init_with_simulation(cur_state[0].v_ego)
libmpc.run_mpc(cur_state, mpc_solution, x_poly, v_poly, a_poly)
# Converge to solution
for _ in range(10):
libmpc.run_mpc(cur_state, mpc_solution, x_poly, v_poly, a_poly)
ts_sol = list(mpc_solution[0].t)
x_sol = list(mpc_solution[0].x_ego)
v_sol = list(mpc_solution[0].v_ego)
a_sol = list(mpc_solution[0].a_ego)
plt.figure()
plt.subplot(3, 1, 1)
plt.plot(ts, positions, 'k--')
plt.plot(ts_sol, x_sol)
plt.ylabel('Position [m]')
plt.xlabel('Time [s]')
plt.subplot(3, 1, 2)
plt.plot(ts, speeds, 'k--')
plt.plot(ts_sol, v_sol)
plt.xlabel('Time [s]')
plt.ylabel('Speed [m/s]')
plt.subplot(3, 1, 3)
plt.plot(ts, accelerations, 'k--')
plt.plot(ts_sol, a_sol)
plt.xlabel('Time [s]')
plt.ylabel('Acceleration [m/s^2]')
plt.show()
Loading…
Cancel
Save