Пример #1
0
def _ds_exp(i):
    _exp = [(p3_atm.AtmosphereCalm(), '/tmp/pat_glider_ds_nw.npz'),
            (p3_atm.AtmosphereCstWind([1, 0, 0]), '/tmp/pat_glider_ds_wc100.npz'),
            (p3_atm.AtmosphereRidge(), '/tmp/pat_glider_ds_wr.npz'),
            (p3_atm.AtmosphereShearX(wind1=5.0, wind2=0.0, xlayer=60.0, zlayer=40.0), '/tmp/pat_glider_ds_ws50.npz'),
            (p3_atm.AtmosphereVgradient(w0=-2, w1=5, h0=20, h1=60), '/tmp/pat_glider_ds_wvg.npz'),
            (p3_atm.AtmosphereVgradient(w0=-7.5, w1=7.5, h0=20, h1=60), '/tmp/pat_glider_ds_wvg2.npz')]
    atm, save_filename = _exp[i]
    return atm, save_filename
Пример #2
0
def test0():
    atm = p3_atm.AtmosphereCstWind([0., 0., 0.])
    dm_ae = p1_fw_dyn.DynamicModel()
    U = [0.0, 0.0, 0, 0, 0]
    print('U {}'.format(U))
    pos_ned = [0, 0, 0]
    #vel_aero = [10, np.deg2rad(4), 0]
    vel_aero = [10, 0., -0.6]
    eulers = np.deg2rad([0, 0, 45])
    #eulers = [1.77592771, 0.99355044, 3.13596614]
    eulers = [1.23060692, 0.14490205, -2.96400653]
    #eulers = [0, 0, 0]
    rvel_body = [0, 0, 0]
    rvel_body = [0.1, 0, 0]
    Xae = np.concatenate((pos_ned, vel_aero, eulers, rvel_body))
    print('Xae {}'.format(Xae))
    Xae_dot = dm_ae.dyn_new(Xae, t=0, U=U, atm=atm)
    print('Xae_dot     {}'.format(Xae_dot))
    Xae_dot_old = dm_ae.dyn_old(Xae, t=0, U=U, atm=atm)
    print('Xae_dot_old {}'.format(Xae_dot_old))
Пример #3
0
def test2():
    atm = p3_atm.AtmosphereCstWind([0., 0., 0.])
    #atm = p3_atm.AtmosphereSinetWind([0., 0., -0.5])
    #atm = p3_atm.AtmosphereSinedWind([0., 0., -0.5])

    #dm = t1dyn.get_default_dm()
    dm = p1_fw_dyn.DynamicModel()
    #dm = p1_fw_dyn.DynamicModel_ee()

    time = np.arange(0, 30, 0.01)
    Xe, Ue = dm.trim({'h': 0, 'va': 10, 'gamma': 0}, report=True)
    X0, X_act0 = np.array(Xe), np.array(Ue),
    U = Ue * np.ones((len(time), dm.input_nb()))
    time, X, X_act = t1dyn.run_simulation(dm, time, X0, X_act0, U, atm=atm)
    #dm.plot_trajectory_as_ae(time, X, X_act, window_title='aero/euler', atm=atm)
    dm.plot_trajectory_as_ee(time,
                             X,
                             X_act,
                             window_title='euclidian/euler',
                             atm=atm)
    plt.show()
Пример #4
0
def main(dt=0.005):
    param_filename = p3_u.pat_ressource('data/vehicles/cularis.xml')
    dm = p1_fw_dyn.DynamicModel(param_filename)
    trim_args = {'h': 30, 'va': 17, 'gamma': 0}
    if 0:
        atm = p3_atm.AtmosphereCalm()
        save_filename = '/tmp/pat_glider_ds_nw.npz'
    if 0:
        atm = p3_atm.AtmosphereCstWind([1, 0, 0])
        save_filename = '/tmp/pat_glider_ds_wc100.npz'
    if 0:
        atm = p3_atm.AtmosphereRidge()
        save_filename = '/tmp/pat_glider_ds_wr.npz'
    if 0:
        atm = p3_atm.AtmosphereShearX(wind1=5.0,
                                      wind2=0.0,
                                      xlayer=60.0,
                                      zlayer=40.0)
        save_filename = '/tmp/pat_glider_ds_ws50.npz'
    if 1:
        atm = p3_atm.AtmosphereVgradient()
        save_filename = '/tmp/pat_glider_ds_wvg.npz'

    #ref_traj = p3_traj3d.BankedCircleRefTraj(c=[100, 0, -40], r=60, slope=np.deg2rad(10))
    #ctl = p3_guid.GuidanceDS(dm, ref_traj, trim_args, dt, lookahead_dist=15., max_phi=np.deg2rad(60))
    ctl_logger = p3_guid.GuidancePurePursuitLogger()
    time, Xae, U = ctl_logger.load(save_filename)

    # aliasing state variables
    _pos = Xae[:, p3_fr.SixDOFAeroEuler.sv_slice_pos]
    _alt = -Xae[:, p3_fr.SixDOFAeroEuler.sv_z]
    _va = Xae[:, p3_fr.SixDOFAeroEuler.sv_va]
    _va2 = _va**2
    _alpha = Xae[:, p3_fr.SixDOFAeroEuler.sv_alpha]
    _eul = Xae[:, p3_fr.SixDOFAeroEuler.sv_slice_eul]
    _phi, _theta, _psi = [_eul[:, _i] for _i in range(3)]
    _gamma = _theta - _alpha
    # aliasing input variables
    _throttle = U[:, dm.iv_dth()]

    # compute state variable in euclidian/euler format
    Xee = np.array([dm.to_six_dof_euclidian_euler(_X, atm, 0.) for _X in Xae])
    _vi = Xee[:, p3_fr.SixDOFEuclidianEuler.sv_slice_vel]
    groundspeed_3d = np.linalg.norm(_vi, axis=1)  # I hate this variable name

    #ctl_logger.plot_chronograms(time, X, U, ctl, atm2)

    # Compute Energy
    e_kin_air = 0.5 * dm.P.m * _va2
    alt_0 = 30.
    e_pot = dm.P.m * 9.81 * (_alt - alt_0)
    e_tot_air = e_kin_air + e_pot

    # Compute wind in body frame
    #df.wx.iloc[i] = -V_a * cos(theta-np.deg2rad(alpha)) + V_g
    #df.wz.iloc[i] = V_a * sin(theta-np.deg2rad(alpha)) + V_z
    #Vg is gps ground speed, in X-Y
    #and Vz is climb speed
    # def estimate_inplane_wind(df_in):
    #     df = df_in.copy() # Be careful, this is important !!!
    #     for i in range(df.index.shape[0]):
    #         V_a   = df.airspeed.iloc[i]
    #         theta = df.theta.iloc[i]
    #         #alpha = df.alpha.iloc[i]
    #         alpha = alpha_func(df.index[i])
    #         V_g = df.vel.iloc[i]
    #         #         V_g = df.vel_3d.iloc[i]
    #         V_z = -df.climb.iloc[i] #
    #         df.wx.iloc[i] = -V_a * cos(theta-np.deg2rad(alpha)) + V_g
    #         df.wz.iloc[i] = V_a * sin(theta-np.deg2rad(alpha)) + V_z # going up is negative for wz
    #         df.alpha[i] = alpha
    #     return df
    Wned = np.array([atm.get_wind_ned(_p, _t) for _p, _t in zip(_pos, time)])
    Wb = np.array(
        [p3_fr.vel_world_to_body_eul(_v, _e) for _v, _e in zip(Wned, _eul)])
    Wx, Wy, Wz = [Wb[:, _i] for _i in range(3)]
    dt = time[1] - time[0]
    dWx, dWz = np.gradient(Wx, dt), np.gradient(Wz, dt)
    # compute power
    rho = 1.225
    AR = 11.84
    e = 0.85
    # I don't have accel, let's compute it
    Az = np.gradient(_vi[:, 2], dt)
    Lift = -Az * dm.P.m
    #Lift = -Az * mass
    CL = Lift / (0.5 * rho * _va2 * dm.P.Sref)
    CD = dm.P.CD0 + CL**2 / (np.pi * AR * e)
    D = -0.5 * rho * _va2 * dm.P.Sref * CD
    P_drag = _va * D
    P_dwx = -dWx * (-_va * np.sign(_gamma) * np.cos(_gamma))
    P_dwz = dWz * (_va * np.sin(_gamma))

    #matplotlib.rcParams['text.usetex'] = True
    plt.rcParams["font.family"] = "Times New Roman"
    plt.rcParams["font.size"] = 11
    fig = plt.figure(figsize=(10, 14.5))
    axes = fig.subplots(6, 1, sharex=True)
    fig.subplots_adjust(top=0.975,
                        bottom=0.035,
                        left=0.125,
                        right=0.97,
                        hspace=0.165,
                        wspace=0.205)
    ax = axes[0]  #fig.add_subplot(611)
    ax.set_title('Energy in Air-Path Frame')
    ax.plot(time, e_tot_air, label='$E_{Total}$')
    ax.plot(time, e_kin_air, label='$E_{Kinetic}$')
    ax.plot(time, e_pot, label='$E_{Potential}$')
    ax.grid()
    ax.legend()
    ax.set_ylabel('Energy [J]')

    ax = axes[1]  #fig.add_subplot(612)
    ax.plot(time, _alt - alt_0, label='Height AGL')
    ax.plot(time, _va, label='$V_a$')
    ax.plot(time, groundspeed_3d, label='$V_{i}$')
    ax.grid()
    ax.set_ylabel('Height AGL [$m$] \nSpeed [$m/s$]')
    ax.legend()

    ax = axes[2]  #fig.add_subplot(613)
    ax.plot(time, np.rad2deg(_gamma), label='$\gamma$ [deg]')
    ax.plot(time, Wx, label='$W_x$')
    ax.plot(time, dWx, label='$\dot{W_x}$')
    ax.plot(time, Wz, label='$W_z$')
    ax.plot(time, -_vi[:, 2], label='$Vi_z$')
    ax.grid()
    ax.legend()
    ax.set_ylabel(
        'Flight Path ($\gamma$) [deg] \n  Wind Speed [$m/s$] \n  Gradient [$m/s^2$]'
    )

    ax = axes[3]  #fig.add_subplot(614)
    ax.plot(time,
            P_drag,
            color='red',
            label='$P_D$  [W], $\sum{P_D}$ = %0.2f [Ws]' %
            (np.nansum(P_drag) / 100))
    ax.plot(time,
            P_dwx,
            color='black',
            alpha=0.6,
            label='$P_{\dot{W}_X}$ [W], $\sum{P_{\dot{W}_X}}$ = %0.2f [Ws]' %
            (np.nansum(P_dwx) / 100))
    ax.fill_between(time,
                    0,
                    P_dwx,
                    where=(P_dwx >= 0),
                    alpha=0.50,
                    color='green',
                    interpolate=True)
    ax.fill_between(time,
                    0,
                    P_dwx,
                    where=(P_dwx < 0),
                    alpha=0.50,
                    color='red',
                    interpolate=True)
    ax.grid()
    ax.legend()
    ax.set_ylabel('Power ($P_{\dot{W}_X}\, & \, P_D$)  [W]')

    ax = axes[4]  #fig.add_subplot(615)
    ax.plot(time,
            P_dwz,
            color='black',
            alpha=0.3,
            label='$P_{\dot{W_Z}}$, $\sum{P_{\dot{W_Z}}}$ = %0.2f [Ws]' %
            (np.nansum(P_dwz) / 100))
    ax.fill_between(time,
                    0,
                    P_dwz,
                    where=(P_dwz >= 0),
                    alpha=0.50,
                    color='green',
                    interpolate=True)  #, label='P$_{\dot{w}}$');
    ax.fill_between(time,
                    0,
                    P_dwz,
                    where=(P_dwz < 0),
                    alpha=0.50,
                    color='red',
                    interpolate=True)  #, label='P$_{\dot{w}}$');
    ax.grid()
    plt.legend()
    ax.set_ylabel('Power ($P_{\dot{W}_Z}$)  [W]')
    ax.set_xticklabels([])

    ax = axes[5]  #fig.add_subplot(616);
    ax.plot(time,
            _throttle,
            label='Throttle,  Averaged = %0.2f' % (np.nanmean(_throttle)))
    #ax.fill_between(time,0,electrical_power, where=(_throttle >= throttle_limit), alpha=0.20, color='grey', interpolate=True)
    #ax.plot(time,electrical_power, label='$P_{Elec.}$ [W] , $\sum{P_{Elec.}}$ = %0.2f [Ws]' % (np.nansum(electrical_power[throttle>=throttle_limit])/100.* propulsion_eff) ); #np.nanmean(electrical_power),
    ax.grid()
    plt.legend()
    ax.set_ylabel('Throttle [\%] \n Elec. Power ($P_{Elec.}$) [W]')
    ax.set_xlabel('Time [s]')

    #ax.set_xlim([20, 30])
    plt.savefig('/tmp/traj_murat_ds.png', dpi=120, bbox_inches='tight')
Пример #5
0
def main(h=0):
    atm = p3_atm.AtmosphereCstWind([0., 0., 0.])
    dm = p1_fw_dyn.DynamicModel_ee()
    #fit_va(atm, dm)
    #fit_phi(atm, dm)
    fit_va_and_phi(atm, dm, compute=True)