CN106843282B - M100 development platform-based area complete search and obstacle avoidance system and method - Google Patents

M100 development platform-based area complete search and obstacle avoidance system and method Download PDF

Info

Publication number
CN106843282B
CN106843282B CN201710162023.6A CN201710162023A CN106843282B CN 106843282 B CN106843282 B CN 106843282B CN 201710162023 A CN201710162023 A CN 201710162023A CN 106843282 B CN106843282 B CN 106843282B
Authority
CN
China
Prior art keywords
obstacle
module
obstacle avoidance
mode
cruise
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Expired - Fee Related
Application number
CN201710162023.6A
Other languages
Chinese (zh)
Other versions
CN106843282A (en
Inventor
牛钰茜
孔德优
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Southeast University
Original Assignee
Southeast University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Southeast University filed Critical Southeast University
Priority to CN201710162023.6A priority Critical patent/CN106843282B/en
Publication of CN106843282A publication Critical patent/CN106843282A/en
Application granted granted Critical
Publication of CN106843282B publication Critical patent/CN106843282B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • GPHYSICS
    • G05CONTROLLING; REGULATING
    • G05DSYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D1/00Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
    • G05D1/12Target-seeking control

Landscapes

  • Engineering & Computer Science (AREA)
  • Aviation & Aerospace Engineering (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Automation & Control Theory (AREA)
  • Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)

Abstract

The invention relates to a regional complete search and obstacle avoidance system based on an M100 development platform, which comprises: the system comprises an APP module, a cruise module and an obstacle avoidance module, wherein the APP module is responsible for sending take-off and landing commands and displaying pictures in real time, the cruise module comprises a GPS calibration module, a return base line module and a steering module, the obstacle avoidance module comprises a judgment module and a decision module, the judgment module is responsible for judging whether an obstacle exists in the direction, and the decision module is responsible for obtaining which step the current obstacle avoidance is performed. According to the technical scheme, the obstacle can be avoided while completely autonomous and comprehensive searching is carried out in a specific area, the accuracy and the efficiency of detecting the obstacle are improved by combining the data of the binocular vision system and the ultrasonic data, and the success rate of automatic obstacle avoidance of the unmanned aerial vehicle and the efficiency of area searching are further improved.

Description

M100 development platform-based area complete search and obstacle avoidance system and method
Technical Field
The invention relates to an omnibearing automatic obstacle avoidance and cruising method based on M100 of Dajiang company, and belongs to the technical field of unmanned aerial vehicle cruising and obstacle avoidance.
Background
Along with the application of unmanned aerial vehicle is more and more extensive, the environment that unmanned aerial vehicle used is more and more complicated, to unmanned aerial vehicle perception environment, avoids the requirement of barrier more and more high. At present, obstacle avoidance methods of unmanned aerial vehicles mainly include TOF technologies, visual obstacle avoidance, ultrasonic obstacle avoidance, infrared obstacle avoidance, radar obstacle avoidance and the like, and one or two methods are often adopted for the methods. The ToF technology is easily affected by light pollution and dust and smoke, the target identification and elimination capacity is weak, the radar obstacle avoidance is based on medium and low frequency, the radar is large in size, and the radar is difficult to install and use on the unmanned aerial vehicle; in addition, most of systems for avoiding obstacles do not realize real autonomous flight, the whole process from take-off to landing of the unmanned aerial vehicle cannot realize full automation, and only how to avoid the obstacles is considered. The existing unmanned aerial vehicle obstacle avoidance system does not consider how to combine with cruising in a specific area, if only avoiding obstacles, the searched areas have large overlap, and complete search is probably not achieved.
M100 is a flight platform which is developed by the company of great Xinjiang, stable, reliable, powerful and flexible to expand, and can be developed for the second time. Wherein, M100 has been provided with 5 directional binocular vision systems and ultrasonic systems, can acquire depth map data and ultrasonic data. Based on the foundation of the platform and the problems existing in the existing unmanned aerial vehicle obstacle avoidance, the applicant develops an omnibearing automatic obstacle avoidance and cruising method based on the M100 of the Dajiang company.
Disclosure of Invention
The invention provides a region complete search and obstacle avoidance system and method based on an M100 development platform aiming at the technical problems in the prior art, the technical scheme utilizes a mobile phone APP to control the mobile phone APP to automatically take off and land, and image data are transmitted back in real time; different types of obstacles are better identified by combining ultrasonic data and depth map data; adjusting the angle of the holder under different conditions to realize complete regional scanning; the method achieves real autonomous flight and complete area search, and better combines cruising and obstacle avoidance.
In order to achieve the above object, a technical solution of the present invention is as follows, a system for complete area search and obstacle avoidance based on an M100 development platform, wherein the system includes: the system comprises an APP module, a cruise module and an obstacle avoidance module, wherein the cruise module is used as a main thread, the APP module and the obstacle avoidance module are used as sub-threads, the following variables are used for information transmission in the cruise module and the obstacle avoidance module, and the variables controlled by the cruise module have control rights (when the control rights are in a cruise mode, the obstacle avoidance mode thread cannot run, otherwise, the cruise mode waits; the variables controlled by the obstacle avoidance module have the control right, the flying height, the flying speed and whether the variables are safe or not when returning to the base line. The system comprises an APP module, a base line returning module and a steering module, wherein the APP module is responsible for sending takeoff and landing commands and displaying pictures in real time, the cruise module comprises a GPS calibration module, the base line returning module and the steering module, the GPS calibration module is responsible for GPS coordinate acquisition and calibration, the base line returning module is responsible for safely returning to the base line after obstacle avoidance is finished, the steering module is responsible for arriving at a turning position and stably adjusting the direction and calculating the next target point, the obstacle avoidance module comprises a judgment module and a decision module, the judgment module is responsible for judging whether an obstacle exists in the direction or not, and the decision module is responsible for obtaining which step the current obstacle avoidance. Due to the limitation of bandwidth, the unmanned aerial vehicle can only transmit depth map data in two directions in the binocular vision system, and the method specifically comprises the following steps: all be equipped with image sensor and ultrasonic sensor in unmanned aerial vehicle's top, the place ahead, left, right side, below, the ultrasonic data of 5 directions all can obtain, but the depth map data can only transmit two directions, here we adopt image sensor No. 1 to transmit the image data in the place ahead all the time, then according to different situation image sensor No. 2 transmission left, right side or the depth map data of top.
The utility model discloses a cruise control system, including APP, M100, LAN, and at this moment, M100 takes off, and the APP sends and keeps away barrier mode one, and the data of APP promptly sends for M100 through the LAN, and at this moment, M100 takes off, and the control authority is keeping away barrier mode one, and the height-adjusting back gets into cruise mode, and the control authority specifically includes at cruise mode promptly: this system divide into two threads, and the module of cruising is as the main thread, keeps away the barrier module and manages as the sub-thread, and the control right is managed by the module of cruising, and the module that possess the control right can change unmanned aerial vehicle's flight state, and the barrier mode is kept away in the originated acquiescence of unmanned aerial vehicle entering, and after the adjustment height, the control right is regained to the mode of cruising. The system can completely and automatically search in a specific area and simultaneously avoid the obstacle, the accuracy and the efficiency of detecting the obstacle are improved by combining the data of the binocular vision system and the ultrasonic data, and the success rate of automatically avoiding the obstacle of the unmanned aerial vehicle and the efficiency of area search are improved.
A method for an M100 development platform-based area complete search and obstacle avoidance system comprises the following steps: 1) the method comprises the steps that an APP end sends a takeoff command, M100 enters an obstacle avoidance mode I, namely a control right is in an obstacle avoidance mode, and after the height is adjusted, the APP end enters a cruise mode, namely the control right is in the cruise mode; in the cruise mode, the M100 flies according to a track planned by a cruise rule, the cradle head swings at a certain angle, and a shot picture is transmitted to the APP end in real time; 2) the cruising speed varies with the result of the calculation; the image sensor 1 is always started, and the direction is the front; judging whether an obstacle exists in front or not by combining the depth map data and the ultrasonic data; if so, applying a control right in the obstacle avoidance mode, and handing the control right to the obstacle avoidance mode in the cruise mode under the condition of non-turning to enter the obstacle avoidance mode I; 3) the right side of the image sensor No. 2 is opened, and whether an obstacle exists on the right side is judged; if yes, the right side is unsafe, M100 turns left until no obstacle exists in the front, and enters a left turn back state; if not, the right side is safe, and M100 turns right until no obstacle exists in the front; 4) entering a state of right turn back; in the state of left (right) turning, the tripod head swings at 90 degrees between the direction of the image sensor and the ground, the No. 2 right (left) side of the image sensor is started, and whether an obstacle exists in the front or not is judged preferentially; if yes, M100 turns left (right) until no obstacle exists in front; if not, judging whether the right (left) party is safe; if not, M100 continues to advance; if yes, turning M100 right (left), and judging whether the direction is the cruising direction; if not, M100 continues to turn right (left) until the direction is the cruising direction; if yes, M100 is in a forward state and is deviated to the left (right), and whether an obstacle exists on the right (left) side or not is judged; if yes, M100 advances until no obstacle exists in the right (left) direction, and then translates to the right (left); if not, translating to the right (left) until returning to the base line, and returning the control right to the cruise mode; 5) when the electric quantity is smaller than a certain threshold value or the user clicks on the APP to return to the navigation, the M100 enters an obstacle avoidance mode II; opening the upper part of the image sensor No. 2, and judging whether an obstacle exists above the image sensor No. 2; if yes, the M100 flies according to the obstacle avoidance mode I until no obstacle exists above; if not, the M100 rises to a certain height, and flies back to the starting point directly, and the task is finished.
As an improvement of the present invention, in the step 1), in the cruise mode, the step of flying the M100 according to a predefined track, swinging the pan-tilt according to a certain angle, and transmitting a shot picture to the APP end in real time includes: before determining the flight track, firstly acquiring GPS coordinates of four vertexes of a cruise area, wherein the flight track adopts a snake-shaped roundabout mode, and the terminal point of the line and the place to be turned are calculated when the unmanned aerial vehicle walks a section of broken line, so that the calculation amount can be greatly reduced; in the cruise mode, the swinging angle is closely related to the flying height of the unmanned aerial vehicle and the step length of the cruise track, the fixed holder swings 45 degrees left and right, the flying height is 2m, the step length can be calculated, the complete search of the area is achieved with the least time and the least electric quantity, and the efficiency is improved.
As an improvement of the present invention, in step 2), the cruise speed varies with the result of calculation as follows, and the cruise speed has three stages: the ultrasonic data return distance comprises a fast speed, a slow speed I and a slow speed II (in actual test, 3.5m/s, 2m/s and 0.7m/s are respectively adopted), and two thresholds are provided: a threshold value for entering an early warning state and a threshold value for entering an obstacle avoidance mode (in actual tests, 5m and 2m are respectively adopted, and the threshold value for entering the early warning state is larger than the threshold value for entering the obstacle avoidance mode). If no obstacle is detected in the advancing direction and the distance of returning ultrasonic data is greater than the threshold value for entering the early warning state, the vehicle can advance rapidly; if no obstacle is detected in the forward direction, the ultrasonic data are smaller than the threshold value for entering the early warning state and larger than the threshold value for entering the obstacle avoidance mode, the unmanned aerial vehicle enters the early warning state by advancing at a slow speed, namely the obstacle possibly appears in the front; if the front detected obstacle or the ultrasonic data is smaller than the threshold value for entering the obstacle avoidance mode, the unmanned aerial vehicle enters the obstacle avoidance mode at a slow speed II; after obstacle avoidance is finished, determining the cruising speed through a new round of judgment;
in the step 2), the step of judging whether the front part has the obstacle is as follows: firstly, acquiring most original depth map data and ultrasonic data, wherein the ultrasonic data is the distance from an unmanned aerial vehicle to a nearest obstacle, and the depth map data is an unprocessed single-channel image; and then carrying out corrosion, expansion and binarization on the obtained depth map, wherein the binarization process comprises the following steps: the method comprises the steps of converting black and white parts without obstacles into black and converting gray parts with obstacles into white to obtain a binarized image, displaying the obstacles in the image as white at the moment, then calculating the areas of all continuous white areas, and considering that the obstacles exist in the direction when the area of at least one continuous white area is larger than a certain threshold value. In addition, in an actual test, some rows of pixels below the front image are found to be the ground, and the pixels are judged to be obstacles due to the influence of illumination, so that only depth map data above the image is processed to judge whether the obstacles exist, the calculation amount is reduced, and the efficiency is improved; the ultrasonic data is acquired at a high frequency, the ultrasonic data can change quickly along with the flight state of the unmanned aerial vehicle, the result is unstable, for example, a thin obstacle is arranged in front, and two adjacent ultrasonic data can be completely different, so that a segmented sampling method is adopted, every 10 data are in one group, only one value is taken in the group, and when the new data are smaller than the current minimum distance, the current minimum distance is updated, so that the judgment error caused by the obstacle shape can be avoided, and the safety is improved; if the current minimum distance is greater than a certain threshold, it is also considered that there is an obstacle in that direction.
As an improvement of the present invention, in the step 3), the step of entering the state of "after right turn" is as follows, M100 observes whether there is an obstacle in front, and when a danger in front is found, the default right side of the image sensor No. 2 is turned on, and at this time, observes whether the right side is safe, turns right if safe, and turns left if unsafe. There is a special case here that unmanned aerial vehicle flies into a narrower space, all has the circumstances of barrier about, and unmanned aerial vehicle can turn left, nevertheless because unmanned aerial vehicle's the state of turning right and turning left all belongs to the original place action, even also can observe earlier whether the place ahead has the barrier after turning left, if have the barrier, then can continue turning left, and unmanned aerial vehicle will withdraw from this narrower space this moment.
As an improvement of the present invention, in the step 4), the step of preferentially determining whether there is an obstacle in front includes: in the "left (right) turn back" state, since there is already an obstacle on the left, in order to make the real-time transmitted image more meaningful, the pan-tilt transmits the image about the obstacle to the APP end, the pan-tilt swing direction is as shown in fig. 3; in the "left (right) turn back" state, each time it is judged whether there is an obstacle in front, it is judged preferentially whether there is an obstacle in the direction of movement, and then the data of the other image sensor is judged.
As an improvement of the present invention, in the step 4), if yes, M100 is in the "forward" state and is biased to the left (right), and it is determined whether there is an obstacle on the right (left) side; if yes, M100 advances until no obstacle exists in the right (left) direction, and then translates to the right (left); if not, translating to the right (left) until returning to the base line, and returning the control right to the cruise mode comprises the following steps: after the direction is adjusted to the cruising direction, what is needed to do next is to translate back to the cruising line, and during the translation, attention is paid to observe whether the direction of translation has an obstacle or not, if so, the direction of translation advances until the obstacle is crossed, and then the translation is continued.
As an improvement of the invention, in the step 5), the step of entering the obstacle avoidance mode two by the M100 is as follows, the cruise of the M100 development platform and the obstacle avoidance mode of the obstacle avoidance method of the invention have two types, namely the obstacle avoidance mode one and the obstacle avoidance mode two, and the difference between the two modes is that the obstacle avoidance mode two is the obstacle avoidance mode adopted when the unmanned aerial vehicle returns, when the obstacle avoidance mode two is in state, the image sensor 2 is opened above, when there is no obstacle above, the unmanned aerial vehicle rises to a certain height and then returns to the navigation directly, otherwise, the unmanned aerial vehicle flies in the obstacle avoidance mode one.
Compared with the prior art, the unmanned aerial vehicle has the advantages that 1) the technical scheme realizes real autonomous flight and complete area search, cruise and obstacle avoidance are better combined, after the user clicks a 'take-off' button at the APP end, no person can fly according to a preset cruise track, if the user encounters an obstacle in the process, no person can avoid the obstacle, the user returns to a cruise base line after obstacle avoidance is finished and keeps the original cruise direction, the unmanned aerial vehicle continuously flies according to the cruise track, when the electric quantity is lower than a certain threshold value or the user clicks a 'return flight' button at the APP end, the unmanned aerial vehicle ascends to a certain height under the condition that no obstacle exists above the unmanned aerial vehicle, the unmanned aerial vehicle returns to a flight point, and if the obstacle exists above the unmanned aerial vehicle, the unmanned aerial vehicle firstly flies to the area without the obstacle above the unmanned aerial vehicle and then ascends. The cradle head can adjust the swinging direction along with different modes, shot pictures are transmitted to the APP end in real time, only a user clicks a button on the APP in the whole process, and then the unmanned aerial vehicle can perform completely autonomous area search and obstacle avoidance; 2) in the aspect of avoiding obstacles, ultrasonic data and image sensor data are combined meaningfully, so that the method can adapt to more complex environments and improve safety and reliability; 3) in the aspect of area search, the angle of the holder is adjusted according to different conditions, more comprehensive information is obtained, and the comprehensive search of the area can be achieved through path planning; 4) in the aspect of independently flying, only need the user to click the button of APP end, can realize whole flight process to return safely, combine through data and the ultrasonic data to binocular vision system, improve the precision and the efficiency of detecting the barrier, and then improve the automatic success rate of keeping away the barrier of unmanned aerial vehicle and the efficiency of area search.
Drawings
FIG. 1 is a schematic diagram of a cruise track of the area complete search and obstacle avoidance system and method based on an M100 development platform according to the present invention;
FIG. 2 is a schematic view of the swing direction of the area complete search and obstacle avoidance system and method based on the M100 development platform according to the present invention;
FIG. 3 is a schematic view of the swing direction of the area complete search and obstacle avoidance system and method based on the M100 development platform according to the present invention;
FIG. 4 is a flowchart of an obstacle avoidance mode of the M100 development platform based area complete search and obstacle avoidance system and method of the present invention;
FIG. 5 is a second flowchart of an obstacle avoidance mode of the system and method for area complete search and obstacle avoidance based on M100 development platform of the present invention;
FIG. 6 is a system diagram of the system and method for area complete search and obstacle avoidance according to the present invention based on M100 development platform;
fig. 7 is an exemplary schematic diagram of the system and method for area complete search and obstacle avoidance based on the M100 development platform according to the present invention.
Fig. 8 is a rectangular area schematic diagram of cruising based on the M100 development platform area complete search and obstacle avoidance system and method of the present invention.
Fig. 9 is a schematic diagram of calculating an offset of two known GPS values of the M100 development platform-based area complete search and obstacle avoidance system and method of the present invention.
Fig. 10 is a schematic diagram of four GPS values of a modified rectangular area represented by the system and method for area complete search and obstacle avoidance based on the M100 development platform of the present invention.
Fig. 11 is a schematic diagram of a method for calculating each one-way end point based on the M100 development platform area complete search and obstacle avoidance system and method of the present invention.
The specific implementation mode is as follows:
for the purpose of enhancing an understanding of the present invention, the present embodiment will be described in detail below with reference to the accompanying drawings.
Example 1: referring to fig. 6, a system for area complete search and obstacle avoidance based on an M100 development platform includes: the system comprises an APP module, a cruise module and an obstacle avoidance module, wherein the cruise module is used as a main thread, the APP module and the obstacle avoidance module are used as sub-threads, the following variables are used for information transmission in the cruise module and the obstacle avoidance module, and the variables controlled by the cruise module have control rights (when the control rights are in a cruise mode, the obstacle avoidance mode thread cannot run, otherwise, the cruise mode waits; the variables controlled by the obstacle avoidance module have the control right, the flying height, the flying speed and whether the variables are safe or not when returning to the base line. The system comprises an APP module, a base line returning module and a steering module, wherein the APP module is responsible for sending takeoff and landing commands and displaying pictures in real time, the cruise module comprises a GPS calibration module, the base line returning module and the steering module, the GPS calibration module is responsible for GPS coordinate acquisition and calibration, the base line returning module is responsible for safely returning to the base line after obstacle avoidance is finished, the steering module is responsible for arriving at a turning position and stably adjusting the direction and calculating the next target point, the obstacle avoidance module comprises a judgment module and a decision module, the judgment module is responsible for judging whether an obstacle exists in the direction or not, and the decision module is responsible for obtaining which step the current obstacle avoidance. Due to the limitation of bandwidth, the unmanned aerial vehicle can only transmit depth map data in two directions in the binocular vision system, and the method specifically comprises the following steps: all be equipped with image sensor and ultrasonic sensor in unmanned aerial vehicle's top, the place ahead, left, right side, below, the ultrasonic data of 5 directions all can obtain, but the depth map data can only transmit two directions, here we adopt image sensor No. 1 to transmit the image data in the place ahead all the time, then according to different situation image sensor No. 2 transmission left, right side or the depth map data of top. The utility model discloses a cruise control system, including APP, M100, LAN, and at this moment, M100 takes off, and the APP sends and keeps away barrier mode one, and the data of APP promptly sends for M100 through the LAN, and at this moment, M100 takes off, and the control authority is keeping away barrier mode one, and the height-adjusting back gets into cruise mode, and the control authority specifically includes at cruise mode promptly: this system divide into two threads, and the module of cruising is as the main thread, keeps away the barrier module and manages as the sub-thread, and the control right is managed by the module of cruising, and the module that possess the control right can change unmanned aerial vehicle's flight state, and the barrier mode is kept away in the originated acquiescence of unmanned aerial vehicle entering, and after the adjustment height, the control right is regained to the mode of cruising. The system can completely and automatically search in a specific area and simultaneously avoid the obstacle, the accuracy and the efficiency of detecting the obstacle are improved by combining the data of the binocular vision system and the ultrasonic data, and the success rate of automatically avoiding the obstacle of the unmanned aerial vehicle and the efficiency of area search are improved.
Example 2: a method of a region complete search and obstacle avoidance system based on an M100 development platform comprises the following steps: 1) the method comprises the steps that an APP end sends a takeoff command, M100 enters an obstacle avoidance mode I, namely a control right is in an obstacle avoidance mode, and after the height is adjusted, the APP end enters a cruise mode, namely the control right is in the cruise mode; in the cruise mode, the M100 flies according to a track planned by a cruise rule, the cradle head swings at a certain angle, and a shot picture is transmitted to the APP end in real time; 2) the cruising speed varies with the result of the calculation; the image sensor 1 is always started, and the direction is the front; judging whether an obstacle exists in front or not by combining the depth map data and the ultrasonic data; if so, applying a control right in the obstacle avoidance mode, and handing the control right to the obstacle avoidance mode in the cruise mode under the condition of non-turning to enter the obstacle avoidance mode I; 3) the right side of the image sensor No. 2 is opened, and whether an obstacle exists on the right side is judged; if yes, the right side is unsafe, M100 turns left until no obstacle exists in the front, and enters a left turn back state; if not, the right side is safe, and M100 turns right until no obstacle exists in the front; 4) entering a state of right turn back; in the state of left (right) turning, the tripod head swings at 90 degrees between the direction of the image sensor and the ground, the No. 2 right (left) side of the image sensor is started, and whether an obstacle exists in the front or not is judged preferentially; if yes, M100 turns left (right) until no obstacle exists in front; if not, judging whether the right (left) party is safe; if not, M100 continues to advance; if yes, turning M100 right (left), and judging whether the direction is the cruising direction; if not, M100 continues to turn right (left) until the direction is the cruising direction; if yes, M100 is in a forward state and is deviated to the left (right), and whether an obstacle exists on the right (left) side or not is judged; if yes, M100 advances until no obstacle exists in the right (left) direction, and then translates to the right (left); if not, translating to the right (left) until returning to the base line, and returning the control right to the cruise mode; 5) when the electric quantity is smaller than a certain threshold value or the user clicks on the APP to return to the navigation, the M100 enters an obstacle avoidance mode II; opening the upper part of the image sensor No. 2, and judging whether an obstacle exists above the image sensor No. 2; if yes, the M100 flies according to the obstacle avoidance mode I until no obstacle exists above; if not, the M100 rises to a certain height, and flies back to the starting point directly, and the task is finished.
In step 1), in the cruise mode, M100 flies according to the track stipulated in advance, and the cloud platform swings according to certain angle to the picture that will shoot transmits the step that the APP end in real time includes: before determining the flight path, firstly acquiring the GPS coordinates of four vertexes of a cruise area, wherein the flight path adopts a snake-shaped roundabout mode, as shown in FIG. 1c, when an unmanned aerial vehicle walks a section of broken line, the terminal point of the line and the place to be turned are calculated, so that the calculation amount can be greatly reduced; in the cruise mode, the swing angle of the cloud deck is as shown in fig. 2, the swing angle is closely related to the flying height of the unmanned aerial vehicle and the step length of the cruise track, the cloud deck is fixed to swing left and right at 45 degrees and the flying height is 2m, so that the step length can be calculated, the complete search of the area is achieved with the least time and the least amount of electricity, and the efficiency is improved.
In the step 2), the cruising speed is changed along with the calculation result as follows, and the cruising speed has three levels: the ultrasonic data return distance comprises a fast speed, a slow speed I and a slow speed II (in actual test, 3.5m/s, 2m/s and 0.7m/s are respectively adopted), and two thresholds are provided: a threshold value for entering an early warning state and a threshold value for entering an obstacle avoidance mode (in actual tests, 5m and 2m are respectively adopted, and the threshold value for entering the early warning state is larger than the threshold value for entering the obstacle avoidance mode). If no obstacle is detected in the advancing direction and the distance of returning ultrasonic data is greater than the threshold value for entering the early warning state, the vehicle can advance rapidly; if no obstacle is detected in the forward direction, the ultrasonic data are smaller than the threshold value for entering the early warning state and larger than the threshold value for entering the obstacle avoidance mode, the unmanned aerial vehicle enters the early warning state by advancing at a slow speed, namely the obstacle possibly appears in the front; if the front detected obstacle or the ultrasonic data is smaller than the threshold value for entering the obstacle avoidance mode, the unmanned aerial vehicle enters the obstacle avoidance mode at a slow speed II; after obstacle avoidance is finished, determining the cruising speed through a new round of judgment;
in the step 2), the step of judging whether the front part has the obstacle is as follows: firstly, acquiring most original depth map data and ultrasonic data, wherein the ultrasonic data is the distance from an unmanned aerial vehicle to a nearest obstacle, and the depth map data is an unprocessed single-channel image; and then carrying out corrosion, expansion and binarization on the obtained depth map, wherein the binarization process comprises the following steps: the method comprises the steps of converting black and white parts without obstacles into black and converting gray parts with obstacles into white to obtain a binarized image, displaying the obstacles in the image as white at the moment, then calculating the areas of all continuous white areas, and considering that the obstacles exist in the direction when the area of at least one continuous white area is larger than a certain threshold value. In addition, in an actual test, some rows of pixels below the front image are found to be the ground, and the pixels are judged to be obstacles due to the influence of illumination, so that only depth map data above the image is processed to judge whether the obstacles exist, the calculation amount is reduced, and the efficiency is improved; the ultrasonic data is acquired at a high frequency, the ultrasonic data can change quickly along with the flight state of the unmanned aerial vehicle, the result is unstable, for example, a thin obstacle is arranged in front, and two adjacent ultrasonic data can be completely different, so that a segmented sampling method is adopted, every 10 data are in one group, only one value is taken in the group, and when the new data are smaller than the current minimum distance, the current minimum distance is updated, so that the judgment error caused by the obstacle shape can be avoided, and the safety is improved; if the current minimum distance is greater than a certain threshold, it is also considered that there is an obstacle in that direction.
In the step 3), the step of entering the state of right turn is as follows, wherein the M100 observes whether an obstacle exists in the front, when the danger in the front is found, the default right side of the image sensor No. 2 is opened, at this time, the right side is observed whether the obstacle is safe, the right turn is performed safely, and the left turn is performed when the obstacle is not safe. There is a special case here that unmanned aerial vehicle flies into a narrower space, all has the circumstances of barrier about, and unmanned aerial vehicle can turn left, nevertheless because unmanned aerial vehicle's the state of turning right and turning left all belongs to the original place action, even also can observe earlier whether the place ahead has the barrier after turning left, if have the barrier, then can continue turning left, and unmanned aerial vehicle will withdraw from this narrower space this moment.
In the step 4), the step of preferentially judging whether the front part has the obstacle is as follows: in the "left (right) turn back" state, since there is already an obstacle on the left, in order to make the real-time transmitted image more meaningful, the pan-tilt transmits the image about the obstacle to the APP end, the pan-tilt swing direction is as shown in fig. 3; in the "left (right) turn back" state, each time it is judged whether there is an obstacle in front, it is judged preferentially whether there is an obstacle in the direction of movement, and then the data of the other image sensor is judged. In the step 4), if yes, M100 is in a "forward" state and is deviated to the left (right), and whether an obstacle exists on the right (left) side is judged; if yes, M100 advances until no obstacle exists in the right (left) direction, and then translates to the right (left); if not, translating to the right (left) until returning to the base line, and returning the control right to the cruise mode comprises the following steps: after the direction is adjusted to the cruising direction, what is needed to do next is to translate back to the cruising line, and during the translation, attention is paid to observe whether the direction of translation has an obstacle or not, if so, the direction of translation advances until the obstacle is crossed, and then the translation is continued.
In the step 5), the step that the M100 enters the obstacle avoidance mode II is as follows, the cruise based on the M100 development platform and the obstacle avoidance mode of the obstacle avoidance method are two, namely the obstacle avoidance mode I and the obstacle avoidance mode II, the difference between the two modes is that the obstacle avoidance mode II is the obstacle avoidance mode adopted when the unmanned aerial vehicle returns, when the obstacle avoidance mode II is in the state, the image sensor 2 is opened above, when no obstacle exists above, the unmanned aerial vehicle rises to a certain height and then returns to the ground directly, otherwise, the unmanned aerial vehicle flies in the obstacle avoidance mode I. Simulation application example: referring to fig. 1-11, a method for an area complete search and obstacle avoidance system based on an M100 development platform is as follows, firstly, an area needing to cruise is determined. Through the existing map software, the actual measurement by an instrument or an airplane, the selection in the map selection area of the APP and other methods, the GPS coordinate values of four corners of a rectangular area (as shown in figure 8) needing cruising can be obtained, because certain errors exist in the GPS values obtained through measurement, a program slightly adjusts the obtained GPS values when cruising starts, so that the cruising area is rectangular, and the process is given in a cruising module. Or one GPS value in four corners of the rectangular cruise area can be obtained through the method, and the program can automatically calculate the GPS coordinate values of the other three points by using the following method through specifying the deviation values of each point of the other three points in the northeast coordinate system and the deviation values of the point in the northeast and east directions.
Suppose the latitude and longitude of the GPS of a known point are x respectively0And y0The GPS of the point to be calculated is x1And y1The point to be calculated is offset from the known point in both the north and east directions by x and y in the northeast coordinate system (see fig. 9). The GPS value of the point to be calculated can be calculated by the equation:
x1=x/6378137+x0
y1=y/(6378137*cos(x1/2+x0/2))+y0
after determining the cruising area and the first point (one of the four corners) to start cruising, the user needs to specify the step length of the drone as in fig. 1c to determine the distance the drone is stepped while cruising. And then, a user clicks a takeoff button at the APP end, the unmanned aerial vehicle firstly flies to a cruise starting point, and then cruise and obstacle avoidance are carried out.
At the beginning of the cruise process, the program analyzes the acquired GPS at the four corner points, as shown in fig. 10. The program will preferably use points 0, 1, 2 for flight path planning, and point 3 as the termination reference point. Since the search area is a rectangular area, line 01 should be parallel and equal to line 23, line 12 should be parallel and equal to line 03, and line 01 should be perpendicular to line 12, so that theoretical point 3 can be obtained from points 0, 1, and 2, if the error from the GPS value of the given point 3 is not large, the GPS value of 3 points is corrected to the value obtained by the theoretical calculation, otherwise, the given point is considered to be not in compliance with the rectangular specification. The cruise algorithm generates an S-shaped flight path in a rectangular area, the path traverses the whole cruise area to realize traversal search, the flight path is generated in real time, a single path is a simple straight line, the single path flies from a single-way starting point to a single-way ending point in a straight line mode, when the single-way ending point is reached, the unmanned aerial vehicle flies to the boundary of a cruise area or flies by a stepping length, the flight direction needs to be adjusted at the moment, then the target GPS coordinate of the next straight path is calculated, the target is set as the single-way ending point of the next path, and the GPS coordinate at the moment is set as the single-way starting point of the next path until the search of the whole area is completed (fig. 1b and 1 c). The target GPS coordinates for each single pass are calculated using point 0 and point 1 as reference points, and the target point must be on line 12 and line 03, so all the target GPS coordinates for the entire cruise area can be calculated from point 0, point 1 and the step length.
From the coordinates of point 0 and point 1, assume the GPS of point 0 to be x0And y0Point 1 is x1And y1If the north east coordinate system is established with the point 0 as the origin, the north and east offsets x and y of the point 1 and the point 0 can be obtained by the foregoing offset calculation method, so that the offset angle θ of the rectangular cruise area in the north east coordinate system is arctan (y/x), the offset angles of the line 12 and the line 03 and the north east coordinate system are (90- θ) degrees, and then the offset of one step length relative to the north east coordinate system can be obtained by the triangle transformation from the step length (set as t):
xt=t*cos(90-θ),yt=t*sin(90-θ),
from the GPS values of point 0 and point 1, all target GPS on lines 12 and 03 can be calculated from the GPS offset calculation described above:
target GPS on line 12:
xd=(x/6378137+x0)+xt
yd=(y/(6378137*cos((x/6378137+x0)/2+x0/2))+y0)+yt
target GPS on-line 03:
xd=i*xt
yd=i*yt
therefore, the S-shaped path of the M100 flying in the search area can be generated, in the algorithm, the M100 can perform self-positioning and flying through a GPS, the cruise algorithm does not generate the whole flying path at one time, the target GPS coordinate one-way ending point target GPS coordinate of the next straight-line path is calculated when the flying direction needs to be adjusted (namely, the vehicle needs to turn), and the GPS coordinate at the moment is set as the starting coordinate of the one-way starting point of the next path until the search of the whole area is completed.
When normally cruising, when M100's sensor detects the place ahead and the barrier appears, cruise and will pause, unmanned aerial vehicle gets into and keeps away the barrier mode, when keeping away the barrier and accomplish the back, cruises and takes over unmanned aerial vehicle, nevertheless after keeping away the barrier, unmanned aerial vehicle has flown away from original air route, before continuing to cruise, need get back to on the original air route. Taking fig. 1d as an example, after obstacle avoidance is finished, the unmanned aerial vehicle deviates from the original route to the right, and at this time, the position of the original route baseline of the aircraft can be determined through the one-way starting point and the one-way ending point. The establishing method of the baseline equation is to establish a northeast coordinate system by taking a one-way starting point as an origin, taking a one-way ending point as a point on a route baseline, solving a linear equation to obtain the baseline equation, and calculating the distance between the current position of the unmanned aerial vehicle and the baseline by using a distance formula between the point and the linear equation. At the moment, the unmanned aerial vehicle enters a state of returning to the baseline, a sensor on the left side of the unmanned aerial vehicle is detected, if an obstacle is detected on the left side, the unmanned aerial vehicle flies forwards, whether the obstacle exists in the front or not is detected, and if the obstacle appears again in the front, the unmanned aerial vehicle enters an obstacle avoidance mode again; if there is no obstacle on the left, the aircraft head is kept to fly horizontally towards the left unchanged until the aircraft head returns to the original route baseline, and if there is an obstacle on the left, the aircraft head continues to fly forwards. And continuing to search the task until the unmanned aerial vehicle rectifies the original air route.
In the obstacle avoidance module, after the APP sends a takeoff command, the M100 enters an obstacle avoidance mode one. The mode is divided into 5 states of 'forward', 'right turn back', 'left turn' and 'left turn back'. The state transition is shown in fig. 4.
At the beginning, the default state is 'forward', the direction of the image sensor 1 is the front, in this case, M100 firstly adjusts the height to 1.5-2M in situ, and the control right is handed to the cruise module after adjusting the height in the obstacle avoidance module when the control right is started. And then, if an obstacle exists in front, the obstacle avoidance module applies for a control right, and the cruise module gives the control right to the obstacle avoidance module under the condition of non-in-situ rotation. At this time, the right side is observed firstly by default (the direction of the image sensor No. 2 is right, the image on the right side is transmitted), and if the right side is safe, the state of right turning is entered; if the right side is unsafe, the system enters a left-turn state. In the "turn right" and "turn left" states, M100 rotates in place until the sensor sees no obstacle in front (which also avoids the situation where M100 hits while there is an obstacle in the left), and then enters the "turn right" and "turn left" states.
In the "right turn" and "left turn" states, where the target of M100 is to fly over an obstacle, and then return to the previous cruising direction, we have found through practical tests that on the basis of "right turn" and "left turn" M100 must turn 10 more to ensure safety. For the' right turn, firstly checking whether the image sensor No. 2 is opened on the left side, then observing the surrounding obstacle condition, if the obstacle exists on the left side, continuing to move forward, in the process of advancing, two cases are divided, wherein the first case is that no obstacle is found in the front until no obstacle is found in the left direction, then M100 turns left, if the direction after turning left is the direction of cruising, the state is changed to 'advancing', marked as turning right, if the direction after turning left is not the direction of cruising, M100 turns left again until the direction is the direction of cruising, the second case is that, in the process of advancing, an obstacle is found in front of the left obstacle-free vehicle, then the M100 enters into the right turn again until no obstacle exists in front, then enters into the right turn, the surrounding obstacle condition is observed, the above-described case one and case two are repeated until the direction of M100 is the direction of cruising (i.e., case one). "after left turn" and "after right turn" are similar and will not be described in detail herein.
At this time, M100 enters the "forward" state. After the M100 direction is correct, it is followed by returning to the cruise baseline. If the state of turning right is changed into the state of going forward, the default is to turn right, and if the state of turning left is changed into the state of going forward, the default is to turn left, and here, the GPS also performs a check. When the direction of the image sensor 2 deviates to the right, the image sensor is opened on the left, the M100 translates to the left, if an obstacle is detected on the left, the M100 stops moving to the left, the image sensor moves forwards until no obstacle exists on the left, then the image sensor continues moving to the left until the image sensor returns to a base line of the cruise, and the control right is handed to the cruise from the obstacle avoidance. The left and right are similar and will not be described again. In the obstacle avoidance mode, after the APP sends the return command, the M100 enters the obstacle avoidance mode two, as shown in fig. 5. The mode is divided into 6 states of 'ascending', 'advancing', 'left turning back', 'right turning' and 'right turning back'. The first obstacle avoidance mode and the second obstacle avoidance mode are different in that the second obstacle avoidance mode is an obstacle avoidance strategy in the process of returning to the navigation, and the M100 needs to ascend to a certain height and directly fly back to the starting point, so that the image sensor 2 is opened above the starting point. If there is an obstacle above, the same as the obstacle avoidance mode, M100 keeps the original height to avoid the obstacle until there is no obstacle above, and then M100 directly rises to fly back to the starting point.
In the cruising and obstacle avoidance process, the angle of the holder is set as follows. When the control right is in the cruise module, the tripod head swings downwards by 90 degrees left and right, as shown in figure 2; when the control right is in the obstacle avoidance module, the rotational direction of the holder is the number 2 of the image sensor and the 90-degree swing below the image sensor, as shown in fig. 3.
Assuming that an obstacle as shown in fig. 7 is encountered during cruising, at this time, M100 is switched from the cruising mode to the obstacle avoidance mode on the premise of a non-turning state, enters a "forward" state, defaults to the right sensor being on, detects that there is no obstacle on the right, enters a "right turn" state, when turning right until there is no obstacle in front, then enters a "right turn back" state, as shown in fig. 7, the front again encounters an obstacle in the "right turn back" state, and thus again enters the "right turn" state, then "right turn back" state, in which the left sensor is on, and after M100 has advanced a distance, it finds that there is no obstacle on the left, thus turning left, and then proceeds until there is no obstacle, and since there is no return to the original cruising direction after this time of turning left, it is necessary to turn left again, at this time, the M100 direction and the cruising direction are identical, the obstacle avoidance mode is ended, and is switched to the cruising mode, m100 advances until no obstacle exists on the left, then translates to the original cruise baseline, and continues to fly to the target point.
It should be noted that the above-mentioned embodiments are not intended to limit the scope of the present invention, and all equivalent modifications and substitutions based on the above-mentioned technical solutions are within the scope of the present invention as defined in the claims.

Claims (5)

1. A method for a region complete search and obstacle avoidance system based on an M100 development platform comprises the following steps: the system comprises an APP module, a cruise module and an obstacle avoidance module, wherein the APP module is responsible for sending take-off and landing commands and displaying pictures in real time, the cruise module comprises a GPS calibration module, a return baseline module and a steering module, the GPS calibration module is responsible for GPS coordinate acquisition and calibration, the return baseline module is responsible for safely returning to a baseline after obstacle avoidance is finished, the steering module is responsible for stably adjusting the direction of a turning part and calculating the next target point, the obstacle avoidance module comprises a judgment module and a decision module, the judgment module is responsible for judging whether an obstacle exists in the direction, and the decision module is responsible for obtaining which step the current obstacle avoidance is carried out; the method is characterized by comprising the following steps: step 1) the APP end sends a takeoff command, the M100 enters an obstacle avoidance mode I, namely a control right is in an obstacle avoidance mode, and after the height is adjusted, the M100 enters a cruise mode, namely the control right is in the cruise mode; in the cruise mode, the M100 flies according to a track planned by a cruise rule, the cradle head swings at a certain angle, and a shot picture is transmitted to the APP end in real time; step 2), changing the cruising speed along with the calculated result; the image sensor 1 is always started, and the direction is the front; judging whether an obstacle exists in front or not by combining the depth map data and the ultrasonic data; if so, applying a control right in the obstacle avoidance mode, and handing the control right to the obstacle avoidance mode in the cruise mode under the condition of non-turning to enter the obstacle avoidance mode I; step 3), the right side of the image sensor No. 2 is opened, and whether an obstacle exists on the right side is judged; if yes, the right side is unsafe, M100 turns left until no obstacle exists in the front, and enters a left turn back state; if not, the right side is safe, and M100 turns right until no obstacle exists in the front; step 4), entering a state of right turning; in the state of left/right turning, the tripod head swings at 90 degrees between the direction of the image sensor and the ground, the No. 2 right/left side of the image sensor is opened, and whether an obstacle exists in the front or not is judged preferentially; if yes, M100 turns left/right until no obstacle exists in front; if not, judging whether the right/left side is safe; if not, M100 continues to advance; if so, turning the M100 right/left, and judging whether the direction is the cruising direction; if not, M100 continues to turn right/left until the direction is the cruising direction; if yes, M100 is in a forward state at the moment and is deviated to the left/right, and whether an obstacle exists on the right/left side is judged; if yes, M100 advances until no obstacle exists in the right/left direction, and then translates to the right/left direction; if not, translating to the right/left until returning to the base line, and returning the control right to the cruise mode; step 5), when the electric quantity is smaller than a certain threshold value or the user clicks on the APP to return to the navigation, the M100 enters an obstacle avoidance mode II; opening the upper part of the image sensor No. 2, and judging whether an obstacle exists above the image sensor No. 2; if yes, the M100 flies according to the obstacle avoidance mode I until no obstacle exists above; if not, the M100 rises to a certain height, the M flies back to the starting point directly, and the task is finished;
in step 1), in the cruise mode, M100 flies according to the track stipulated in advance, and the cloud platform swings according to certain angle to the picture that will shoot transmits the step that the APP end in real time includes: before determining the flight track, firstly acquiring GPS coordinates of four vertexes of a cruise area, wherein the flight track adopts a snake-shaped roundabout mode, and when an unmanned aerial vehicle walks a section of broken line, the terminal point of the line and the place to be turned are calculated; in the cruising mode, the swing angle of the holder swings 45 degrees left and right, and the flying height is 2 m;
in the step 2), the cruising speed is changed along with the calculation result as follows, and if no obstacle is detected in the advancing direction and the ultrasonic data is greater than a certain threshold value, the vehicle is advanced rapidly; if the ultrasonic data are smaller than the threshold value, the unmanned aerial vehicle advances at a slow speed and enters an early warning state, namely, an obstacle possibly appears in the front; if an obstacle is detected in front of the unmanned aerial vehicle, the unmanned aerial vehicle enters an obstacle avoidance mode at a slower speed, and whether the ultrasonic threshold of the obstacle is smaller than the ultrasonic threshold of the obstacle entering an early warning state is judged; after obstacle avoidance is finished, determining the cruising speed through a new round of judgment;
in the step 2), the step of judging whether the front part has the obstacle is as follows: firstly, acquiring most original depth map data and ultrasonic data, wherein the ultrasonic data is the distance from an unmanned aerial vehicle to a nearest obstacle, and the depth map data is an unprocessed single-channel image; and then corroding, expanding and binarizing the acquired depth map to obtain a binarized image, displaying the obstacle in the image as white at the moment, then calculating the areas of all continuous white areas, and when at least one continuous white area is larger than a certain threshold value, considering that the obstacle exists in the direction.
2. The method for the area complete search and obstacle avoidance system based on the M100 development platform as claimed in claim 1, wherein in the step 4), the step of entering the state of turning right is as follows, the M100 observes whether there is an obstacle in front, when the danger in front is found, the default right side of the image sensor No. 2 is turned on, at this time, the observation is carried out whether the right side is safe, the right side is turned right if safe, and the left side is turned if unsafe.
3. The method for the area complete search and obstacle avoidance system based on the M100 development platform as claimed in claim 2, wherein in the step 4), the step of preferentially determining whether there is an obstacle in front is as follows: in the "right/left turn back" state, each time whether there is an obstacle in front is judged, it is judged preferentially whether there is an obstacle in the direction of movement, and then the data of the other image sensor is judged.
4. The method for the area complete search and obstacle avoidance system based on the M100 development platform as claimed in claim 3, wherein in the step 4), the step of returning the control right to the cruise mode comprises the following steps: after the direction is adjusted to the cruising direction, what is needed to do next is to translate back to the cruising line, and during the translation, attention is paid to observe whether the direction of translation has an obstacle or not, if so, the direction of translation advances until the obstacle is crossed, and then the translation is continued.
5. The method for the area complete search and obstacle avoidance system based on the M100 development platform as claimed in claim 4, wherein in the step 5), the step of entering the obstacle avoidance mode two by the M100 is as follows, when the obstacle avoidance mode two is in a state, the image sensor No. 2 is opened above, when there is no obstacle above, the unmanned aerial vehicle ascends to a certain height and then directly navigates back, otherwise, the obstacle avoidance mode one is adopted for flying.
CN201710162023.6A 2017-03-17 2017-03-17 M100 development platform-based area complete search and obstacle avoidance system and method Expired - Fee Related CN106843282B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201710162023.6A CN106843282B (en) 2017-03-17 2017-03-17 M100 development platform-based area complete search and obstacle avoidance system and method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201710162023.6A CN106843282B (en) 2017-03-17 2017-03-17 M100 development platform-based area complete search and obstacle avoidance system and method

Publications (2)

Publication Number Publication Date
CN106843282A CN106843282A (en) 2017-06-13
CN106843282B true CN106843282B (en) 2020-11-03

Family

ID=59143940

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201710162023.6A Expired - Fee Related CN106843282B (en) 2017-03-17 2017-03-17 M100 development platform-based area complete search and obstacle avoidance system and method

Country Status (1)

Country Link
CN (1) CN106843282B (en)

Families Citing this family (18)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107402578A (en) * 2017-06-21 2017-11-28 中国科学院深圳先进技术研究院 Unmanned plane panorama obstacle cognitive method, device, equipment and storage medium
CN109314868B (en) * 2017-10-24 2022-03-22 深圳市大疆创新科技有限公司 Interference source positioning method of base station, unmanned aerial vehicle and computer readable storage medium
EP3508936B1 (en) * 2017-11-02 2020-03-18 Autel Robotics Co., Ltd. Obstacle avoidance method and apparatus, movable object, and computer-readable storage medium
CN108058838A (en) * 2017-12-03 2018-05-22 中国直升机设计研究所 A kind of helicopter collision avoidance system based on binocular distance measurement
KR101886404B1 (en) * 2018-01-11 2018-08-09 박민경 Race system and method of unmanned vehicle
WO2020024101A1 (en) * 2018-07-31 2020-02-06 深圳市大疆创新科技有限公司 Map data source processing method, map updating method, device, and storage medium
CN109407103A (en) * 2018-09-07 2019-03-01 昆明理工大学 A kind of unmanned plane greasy weather obstacle recognition system and its recognition methods
CN111736622B (en) * 2019-03-25 2023-04-07 海鹰航空通用装备有限责任公司 Unmanned aerial vehicle obstacle avoidance method and system based on combination of binocular vision and IMU
CN110147096A (en) * 2019-04-01 2019-08-20 江苏大学 A kind of multimachine control method for coordinating that view-based access control model follows
CN110032216A (en) * 2019-04-11 2019-07-19 江苏首控制造技术有限公司 A kind of unmanned plane rescue method applied to SAR at Sea
CN110530390A (en) * 2019-09-16 2019-12-03 哈尔滨工程大学 A kind of non-particle vehicle path planning method under narrow environment
CN110824127A (en) * 2019-10-23 2020-02-21 军事科学院军事医学研究院环境医学与作业医学研究所 Intelligent unmanned aerial vehicle water quality monitoring cruise operation monitoring and alarm control method
CN112835376B (en) * 2019-11-22 2023-03-10 中国电力科学研究院有限公司 Machine head positioning method and system for unmanned aerial vehicle electricity testing
CN112306064A (en) * 2020-11-04 2021-02-02 河北省机电一体化中试基地 RGV control system and method for binocular vision identification
CN112286230A (en) * 2020-11-13 2021-01-29 国家电网有限公司 Unmanned aerial vehicle visual image algorithm, obstacle avoidance step and information fusion processing system thereof
CN113838313B (en) * 2021-11-29 2022-02-18 中国民用航空总局第二研究所 Obstacle identification method for course beacon channel clearance jitter
CN115617078B (en) * 2022-12-12 2023-07-07 北京理工大学 Unmanned aerial vehicle three-dimensional track rapid planning method based on puffing obstacle
CN115933752B (en) * 2023-02-16 2023-06-06 杭州硬十科技有限公司 Unmanned aerial vehicle dynamic obstacle avoidance prediction control method based on AI

Family Cites Families (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7228232B2 (en) * 2005-01-24 2007-06-05 International Business Machines Corporation Navigating a UAV with obstacle avoidance algorithms
CN104538899A (en) * 2015-01-19 2015-04-22 中兴长天信息技术(北京)有限公司 Wireless-transmission-based unmanned aerial vehicle platform for power line inspection
CN104656665B (en) * 2015-03-06 2017-07-28 云南电网有限责任公司电力科学研究院 A kind of general avoidance module of new unmanned plane and step
CN104820429B (en) * 2015-04-28 2017-05-03 南京航空航天大学 Ultrasonic distance detection-based unmanned aerial vehicle obstacle avoidance system and control method thereof
CN105589472B (en) * 2016-03-03 2018-10-23 深圳市智美达科技股份有限公司 Unmanned equipment avoids the method, apparatus and system of obstacle
CN105785393B (en) * 2016-05-27 2019-01-04 四川桑莱特智能电气设备股份有限公司 A kind of unmanned plane real time imagery obstacle avoidance system and method based on laser radar
CN106168810A (en) * 2016-09-18 2016-11-30 中国空气动力研究与发展中心高速空气动力研究所 A kind of unmanned plane during flying obstacle avoidance system based on RTK and method
CN106324619A (en) * 2016-10-28 2017-01-11 武汉大学 Automatic obstacle avoiding method of substation inspection robot

Also Published As

Publication number Publication date
CN106843282A (en) 2017-06-13

Similar Documents

Publication Publication Date Title
CN106843282B (en) M100 development platform-based area complete search and obstacle avoidance system and method
US10053218B2 (en) System and method for positioning an unmanned aerial vehicle
JP7274674B1 (en) Performing 3D reconstruction with unmanned aerial vehicle
CN109923492B (en) Flight path determination
CN106873627B (en) Multi-rotor unmanned aerial vehicle and method for automatically inspecting power transmission line
US20180204469A1 (en) Unmanned aerial vehicle visual point cloud navigation
CN106950978B (en) Fixed-wing unmanned plane obstacle avoidance system and its barrier-avoiding method and fixed-wing unmanned plane
CN105182992A (en) Unmanned aerial vehicle control method and device
CN108594849A (en) Unmanned aerial vehicle obstacle avoidance method based on millimeter wave radar
US20190113937A1 (en) Measuring device, control device for unmanned aerial vehicle and computer program product for controlling unmanned aerial vehicle
CN105116915B (en) A kind of paragliding flight path control system using multimodal satellite navigation
CN107544531B (en) Line inspection method and device and unmanned aerial vehicle
CN103941750A (en) Device and method for composition based on small quad-rotor unmanned aerial vehicle
EP3850456B1 (en) Control and navigation systems, pose optimisation, mapping, and localisation techniques
CN104298248A (en) Accurate visual positioning and orienting method for rotor wing unmanned aerial vehicle
JP2015006874A (en) Systems and methods for autonomous landing using three dimensional evidence grid
CN111796602A (en) Plant protection unmanned aerial vehicle barrier is surveyed and early warning system
JP6138326B1 (en) MOBILE BODY, MOBILE BODY CONTROL METHOD, PROGRAM FOR CONTROLLING MOBILE BODY, CONTROL SYSTEM, AND INFORMATION PROCESSING DEVICE
CN111665508B (en) Helicopter-mounted terrain following and avoiding visual navigation system and navigation method
CN110850894A (en) Automatic return method and device for unmanned aerial vehicle, unmanned aerial vehicle and storage medium
JP2019016197A (en) Moving entity induction system
Theodore et al. Flight trials of a rotorcraft unmanned aerial vehicle landing autonomously at unprepared sites
Springer et al. Autonomous drone landing with fiducial markers and a gimbal-mounted camera for active tracking
Ax et al. UAV based laser measurement for vegetation control at high-voltage transmission lines
CN112306078B (en) Method and system for automatically avoiding obstacle wires of unmanned aerial vehicle

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20201103

CF01 Termination of patent right due to non-payment of annual fee