Parameters

The following is a list of parameters expected by tmplanner and their types. By default, these are loaded from parameters/tmplanner.yaml, unless stated otherwise.

Map parameters

• width: environment width [m] (double)
• height: environment height [m] (double)
• map_resolution_x: map resolution in the x-direction (width) [m/cell] (double)
• map_resolution_y: map resolution in the y-direction (height) [m/cell] (double)
• frame_id: name of the map reference frame (string)
• T_W_MAP: transformation from the world to the map reference frame (4x4 matrix, see parameters/transform_world_map.yaml)

tmplanner_discrete

• upper_threshold, lower_threshold: upper/lower occupancy probability thresholds (doubles, range: 0.0 - 1.0)

tmplanner_continuous

• cov_fun: covariance function for the Gaussian Process (string, "matern3")
• log_hyperparams: hyperparameter logarithms for the covariance function (for "matern3", [length scale, signal variance squared] - vector of doubles)

Planning parameters

• control_poses: number of control poses/points for a finite-horizon polynomial plan (int, range: 1 - 10)
• initial_pos_x, initial_pos_y, initial_pos_z: position of first measurement [m] (doubles)
• maximum_height: maximum altitude to fly [m] (double)
• minimum_height: minimum altitude to fly [m] (double)
• reference_speed, reference_acceleration, reference_yaw: references for polynomial path optimization
• lattice_min_height_points: number of lattice points at the lowest altitude level for the 3-D grid search (int)
• lattice_height_increment: difference between successive altitude levels on the lattice [m] (double)
• time_budget: total allocated time budget for planning [s] (double)

tmplanner_continuous

• use_threshold: whether to use a lower threshold for adaptive planning. Only map regions above this threshold are used to compute information gain (bool)
• lower_threshold: value of the lower threshold as described above (double)

Sensor parameters

• measurement_frequency: frequency at which measurements are taken [Hz] (double)
• fov_angle_x, fov_angle_y: camera field of view angles [deg] (double)
• T_IMU_CAM: extrinsic camera calibration; transformation from the IMU to the camera reference frame (4x4 matrix, see parameters/camera_calib.yaml)
• intrinsics_matrix: intrinsic camera calibration (vector of doubles)

tmplanner_discrete

• saturation_height: maximum altitude for receiving valid sensor measurements [m] (double)
• true_positive_coeffs: polynomial coefficients for the true positive classification curve - P(occ|occ) - in descending order (vector of doubles)
• false_negative_coeffs: polynomial coefficients for the false negative classification curve - P(occ|free) - in descending order (vector of doubles)

tmplanner_continuous

• coefficient_A, coefficient_B: coefficients for the exponential altitude-dependent sensor variance model - var = A * (1 - e^(-B * altitude))

Optimization parameters

• optimization_method: routine to use for optimizing the initial trajectory from the 3-D grid search (string, "cmaes"/"none")
• cmaes_step_size, cmaes_maximum_fevals, cmaes_offsprings: parameters for the CMA-ES, as described here (doubles)

'tmplanner_discrete`

• optimization_objective: informative objective for the 3D grid search (string, "information"/"classification")