Unmanned aerial vehicle and localization method for unmanned aerial vehicle
An unmanned aerial vehicle aerial vehicle and a localization method for an unmanned aerial vehicle aerial vehicle are described. In an embodiment, an unmanned aerial vehicle comprises: a first ultra-wide band node; a second ultra-wide band node; and a localization processor configured to use signals received at the first ultra-wide band node and the second ultra-wide band node from a plurality of anchor nodes located within an environment to estimate a pose state of the unmanned aerial vehicle, wherein the first ultra-wide band node and the second ultra-wide band node are arranged on the unmanned aerial vehicle at positions offset from one another.
1 . An unmanned aerial vehicle comprising:
a first ultra-wide band node;
a second ultra-wide band node; and
a localization processor configured to use signals received at the first ultra-wide band node and the second ultra-wide band node from a plurality of anchor nodes located within an environment to calculate an estimated global pose state of the unmanned aerial vehicle, the estimated global pose state indicating a location and orientation of the unmanned aerial vehicle relative to the plurality of anchor nodes in the environment, wherein the estimated global pose state is calculated using ultra-wide band data comprising distance measurements between ranging ultra-wide band nodes and anchor nodes from the plurality anchor nodes, positions of the anchor nodes in the environment and positions of the ranging ultra-wide band nodes on the unmanned aerial vehicle, the ranging ultra-wide band nodes being the first ultra-wide band node and the second ultra-wide band node,
wherein the first ultra-wide band node and the second ultra-wide band node are arranged on the unmanned aerial vehicle at positions offset from one another.
2 . The unmanned aerial vehicle according to claim 1 , further comprising an inertial measurement unit configured to detect linear acceleration and angular velocity of the unmanned aerial vehicle and wherein the localization processor is further configured to use measurements from the inertial measurement unit to predict the pose state of the unmanned aerial vehicle.
3 . The unmanned aerial vehicle according to claim 2 , wherein the localization processor is further configured to pre-integrate the measurements from the inertial measurement unit to obtain a set of pre-integrated inertial measurement unit measurements and to use the set of pre-integrated inertial movement unit measurements to estimate the pose state of the unmanned aerial vehicle.
4 . The unmanned aerial vehicle according to claim 2 , wherein the localization processor is further configured to compare a predicted distance calculated from the pose state of the unmanned aerial vehicle determined from the inertial measurement measurements with distance predictions determined from measurements by the first ultra-wide band node and the second ultra-wide band node and thereby identify outliners in the measurements by the first ultra-wide band node and the second ultra-wide band node.
5 . The unmanned aerial vehicle according to claim 2 , further comprising an on-board self-localization module configured to generate on-board self-localization data.
6 . The unmanned aerial vehicle according to claim 5 , wherein the on-board self-localization module comprises a lidar sensor and the on-board self-localization data comprises laser scan data.
7 . The unmanned aerial vehicle according to claim 5 , wherein the on-board self-localization module comprises a camera and the on-board self-localization data comprises image data.
8 . The unmanned aerial vehicle according to claim 5 , wherein the localization processor is further configured to estimate the pose state of the unmanned aerial vehicle using measurement data comprising measurements by the first ultra-wide band node and the second ultra-wide band node, the measurements from the inertial measurement unit and the on-board self-localization data by minimizing a cost function between the measurement data and a model predicted pose state.
9 . The unmanned aerial vehicle according to claim 8 , wherein the on-board self-localization module comprises a lidar sensor and the on-board self-localization data comprises laser scan data and the localization processor is further configured to use the on-board self-localization data to extract lidar features from the laser scan data and match the lidar features with a local map to obtain a set of feature-map-matching coefficients, wherein the cost function depends on the set of feature-map-matching coefficients.
10 . The unmanned aerial vehicle according to claim 8 , wherein the on-board self-localization module comprises a camera and the on-board self-localization data comprises image data and the localization processor is further configured to use the on-board self-localization data to extract a set of visual features from the image data, wherein the cost function depends on the set of visual features.
11 . The unmanned aerial vehicle according to claim 8 , wherein the localization processor is further configured to estimate the pose state of the unmanned aerial vehicle using measurement data corresponding to a sliding time window.
12 . A localization method for an unmanned aerial vehicle, the unmanned aerial vehicle comprising: a first ultra-wide band node; and a second ultra-wide band node, wherein the first ultra-wide band node and the second ultra-wide band node are arranged on the unmanned aerial vehicle at positions offset from one another, the method comprising:
receiving signals at the first ultra-wide band node and the second ultra-wide band node from a plurality of anchor nodes located within an environment; and
calculating an estimated global pose state of the unmanned aerial vehicle, the estimated global pose state indicating a location and orientation of the unmanned aerial vehicle relative to the plurality of anchor nodes in the environment based on the signals, wherein the estimated global pose state is calculated using ultra-wide band data comprising distance measurements between ranging ultra-wide band nodes and anchor nodes from the plurality of anchor nodes, positions of the anchor nodes in the environment and positions of the ranging ultra-wide band nodes on the unmanned aerial vehicle, the ranging ultra-wide band nodes being the first ultra-wide band node and the second ultra-wide band node.
13 . The localization method according to claim 12 , wherein the unmanned aerial vehicle further comprises: an inertial measurement unit configured to detect linear acceleration and angular velocity of the unmanned aerial vehicle, and the method further comprises using measurements from the inertial measurement unit to predict the pose state of the unmanned aerial vehicle.
14 . The localization method according to claim 13 , further comprising pre-integrating the measurements from the inertial measurement unit to obtain a set of pre-integrated inertial measurement unit measurements and using the set of pre-integrated inertial measurement unit measurements to estimate the pose state of the unmanned aerial vehicle.
15 . The localization method according to claim 13 , further comprising comparing the predicted distance from the pose prediction of the unmanned aerial vehicle determined from measurements from the inertial measurement unit with the distance measurements by the first ultra-wide band node and the second ultra-wide band node and thereby identify outliners in the measurements by the first ultra-wide band node and the second ultra-wide band node.
16 . The localization method according to claim 13 , further comprising receiving on-board self-localization data generated by on-board self-localization module of the unmanned aerial vehicle.
17 . The localization method according to claim 16 , further comprising estimating the pose state of the unmanned aerial vehicle using measurement data comprising measurements by the first ultra-wide band node and the second ultra-wide band node, the measurements from the inertial measurement unit and the on-board self-localization data by minimizing a cost function between the measurement data and a model predicted pose state.
18 . The localization method according to claim 17 , wherein the on-board self-localization data comprises laser scan data, and the method further comprises extracting lidar features from the laser scan data and matching the lidar features with a local map and thereby obtaining a set of feature-map-matching coefficients, wherein the cost function depends on the set of feature-map-matching coefficients.
19 . The localization method according to claim 17 , wherein the on-board self-localization data comprises image data, and the method further comprises extracting visual features from the image data and thereby obtaining a set of visual features, wherein the cost function depends on the set of visual features.
20 . A non-transitory computer readable medium storing processor executable instructions which when executed on a processor cause the processor to carry out the localization method according to claim 12 .