double_pendulum/examples/ideal/acados_mpc.py at main · based-robotics/double_pendulum · GitHub

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
import os
from datetime import datetime
import numpy as np
import yaml

from double_pendulum.model.symbolic_plant import SymbolicDoublePendulum
from double_pendulum.model.model_parameters import model_parameters
from double_pendulum.simulation.simulation import Simulator

from double_pendulum.utils.plotting import plot_timeseries
from double_pendulum.utils.csv_trajectory import save_trajectory, load_trajectory

from double_pendulum.controller.acados_mpc.acados_mpc import AcadosMpc

actuated_joint = 0
mass = [0.608, 0.630]
length = [0.3, 0.2]
com = [0.275, 0.166]
damping = [0.081, 0.0]
gravity = 9.81
inertia = [0.09, 0.09]
actuated_joint=0    #0 = pendubot, 1 = acrobot
gravity=9.81
cfric = [0.093, 0.186]
Fmax = 6    #torque limit
bend_the_rules=True

torque_limit = [0.0, 0.0]
torque_limit[actuated_joint] = Fmax
if bend_the_rules:
    torque_limit[np.argmin(torque_limit)] = 0.5

# controller parameters
N_horizon=20
prediction_horizon=0.5
Nlp_max_iter=100000000

if actuated_joint == 1: #acrobot
    Q_mat = 2*np.diag([1000, 1000, 100, 100])
    Qf_mat = 2*np.diag([100000, 100000, 100, 100])
    R_mat = 2*np.diag([0.0001, 0.0001])

if actuated_joint == 0: #pendubot
    Q_mat = 2*np.diag([1000, 1000, 100, 100])
    Qf_mat = 2*np.diag([100000, 100000, 100, 100])
    R_mat = 2*np.diag([0.0001, 0.0001])

vmax = 30 #rad/s
vf = 0 # terminal velcoity constraint

mpc_cycle_dt = 0.005 # inner loop
outer_cycle_dt = 0.005 # outer loop

max_solve_time = mpc_cycle_dt * 20

bend_the_rules = True
friction_compensation_mpc = True
friction_compensation_added = False
solver_type = "SQP_RTI" #Select here [SQP, SQP_RTI, or DDP]

# simulation parameter
dt = outer_cycle_dt
t_final = 20  # 5.985
start = np.array([0,0,0,0])
goal = np.array([np.pi, 0, 0.0, 0.0])

# simulation plant
plant = SymbolicDoublePendulum(
    mass,
    length ,
    com ,
    damping ,
    gravity,
    cfric ,
    inertia ,
    torque_limit=torque_limit
)

sim = Simulator(plant=plant)

if not friction_compensation_mpc:
    cfric = np.zeros_like(cfric)

controller = AcadosMpc(
    mass,
    length ,
    com ,
    damping ,
    cfric ,
    gravity,
    inertia ,
    torque_limit
)

controller.set_start(start)
controller.set_goal(goal)
controller.set_parameters(
    N_horizon=N_horizon,
    prediction_horizon=prediction_horizon,
    Nlp_max_iter=Nlp_max_iter,
    max_solve_time=max_solve_time,
    solver_type=solver_type,
    wrap_angle=False,
    fallback_on_solver_fail=True,
    pd_KP=0.0,
    pd_KD=0.0,
    pd_KI=0,
    cheating_on_inactive_joint=bend_the_rules,
    mpc_cycle_dt=mpc_cycle_dt,
    outer_cycle_dt=outer_cycle_dt,
    pd_tracking=False,
    warm_start=True
)

controller.set_velocity_constraints(v_max=vmax, v_final=vf)
controller.set_cost_parameters(Q_mat=Q_mat, Qf_mat=Qf_mat, R_mat=R_mat)
if friction_compensation_added:
    controller.set_friction_compensation(coulomb_fric=cfric)
controller.init()

T, X, U = sim.simulate_and_animate(t0=0.0, x0=start,
                                   tf=t_final, dt=dt, controller=controller,
                                   integrator="runge_kutta",
                                   plot_forecast=True,
                                   scale=0.5,
                                   save_video=False)

plot_timeseries(T, X, U, None,
               plot_energy=False,
               pos_y_lines=[0.0, np.pi],
               tau_y_lines=[-torque_limit[0], torque_limit[0],torque_limit[1],torque_limit[1]],
               scale=0.5)