Intelligent Analysis and Control System for UAV Flight Risks Based on Data Analysis

Abstract

The present invention relates to the technical field of UAV risk analysis, and particularly to an intelligent analysis and control system for UAV flight risks based on data analysis. By identifying obstacles in the flight space, static signals and dynamic signals are obtained. According to the types of signals in the dynamic signals, flight areas are set, and the flight areas include a safe area, a warning area, and a dangerous area. According to the flight areas, the parts of the preset trajectory that are not within the safe area are marked as risk trajectories, and the risk coefficients in the risk trajectories are calculated. The risk coefficients are calculated, and according to the calculation results, corresponding safe flight instructions and obstacle avoidance instructions are obtained, so that the UAV adjusts the corresponding flight trajectory according to different types of obstacles during flight, thereby reducing flight risks and reducing the flight loss of the UAV.

Description

Technical Field

[0001] The present invention relates to the technical field of UAV risk analysis, and particularly to an intelligent analysis and control system for UAV flight risks based on data analysis. Background Art

[0002] An unmanned aerial vehicle, abbreviated as "UAV", is an unpiloted aircraft controlled by a radio remote control device and a self - contained program control device.

[0003] The prior art CN117539269A discloses a UAV multi - aircraft collaborative operation supervision system, which includes a ground control terminal and multiple UAVs; the UAVs establish two - way communication with the ground control terminal, and the communication command contains a serial number for identifying different UAVs; the UAVs fly and operate according to a preset task route, and at the same time, send flight information to the ground control terminal in real time; the ground control terminal receives and uses the flight information to perform flight risk analysis on all the UAVs, and remotely avoids risks for the UAVs with flight risks. The ground control terminal repeats the flight risk analysis and remotely avoids risks for the risk UAVs to achieve multi - aircraft collaborative operation supervision.

[0004] However, during the flight of a UAV, the obstacles encountered by the UAV are random. When the UAV only avoids obstacles based on its flight trajectory, since the flight trajectory of the obstacle in the next time period is uncertain. For example, when the obstacle is a floating balloon, its trajectory will not change in a short time and will float along the predicted trajectory. However, when the obstacle is an object such as a bird or an insect, its flight trajectory will have a sudden deviation, which will increase the flight risk value of the UAV and cause losses to the UAV. Summary of the Invention

[0005] The purpose of the present invention is to solve the problems in the background art, and to propose an intelligent analysis and control system for UAV flight risks based on data analysis.

[0006] To achieve the above purpose, the present invention adopts the following technical solutions:

[0007] An intelligent analysis and control system for UAV flight risks based on data analysis includes a speed detection terminal, a space detection terminal, a foreign object analysis terminal, a flight analysis terminal, and an action execution terminal;

[0008] The speed detection terminal is used to detect the flight speed of the UAV in real time and transmit it to the foreign object analysis terminal;

[0009] The space detection terminal is used to detect obstacles in the flight space of the drone. Through the sensors carried on the drone, it identifies obstacles in the flight space. When the sensors identify an obstacle in the spatial information of the flight space, it captures the obstacle image and generates a static signal and a dynamic signal corresponding to the obstacle image. At the same time, it sets an intermittent time, marks the obstacle as an obstacle target, and conducts tracking to obtain tracking information. Among them, the dynamic signal includes a conscious signal and an unconscious signal. Then, the space detection terminal transmits the static signal, the dynamic signal, and the tracking information to the foreign object analysis terminal;

[0010] The foreign object analysis terminal is used to receive the static signal, the dynamic signal, and the tracking information. Then, it first marks the position where the drone detects the obstacle as the origin and establishes a spatial coordinate system. When it receives the static signal, it annotates the static object in the spatial coordinate system according to the obstacle image and the radar signal, and sets a safe area, a warning area, and a danger area. When it receives the dynamic signal, it first fits the trajectory of the obstacle, then judges the dynamic signal, and sets the corresponding safe area, warning area, and danger area according to the conscious signal and the unconscious signal respectively. Then, the foreign object analysis terminal transmits the safe area, the warning area, and the danger area as the flight area to the flight analysis terminal at the same time;

[0011] The flight analysis terminal is used to annotate the preset trajectory of the drone and the obstacle trajectory in the spatial coordinate system. At the same time, according to the flight area, it marks the part of the preset trajectory that is not within the safe area as a risk trajectory, calculates the risk coefficient in the risk trajectory, and obtains a safe flight instruction and an obstacle avoidance instruction respectively according to the calculation result. Then, the flight analysis terminal transmits the safe flight instruction and the obstacle avoidance instruction to the action execution terminal;

[0012] The action execution terminal is used to fly according to the received specified corresponding flight trajectory.

[0013] As a further solution of the present invention, the specific acquisition methods of the static signal and the dynamic signal are as follows:

[0014] When the drone is flying in the air, it senses the spatial information in the flight space through the carried sensors;

[0015] When an obstacle is identified in the spatial information, an image of the obstacle is collected to generate an obstacle image. The obstacle image is matched with the obstacle information to obtain the state of the obstacle in the spatial information. If the obstacle state is a static state, a static signal is generated. If the obstacle is in a moving state, the obstacle is further identified, and a conscious signal and an unconscious signal are generated according to the identification results. The conscious signal is the signal generated when the obstacle is a living body, and the unconscious signal is the signal generated when the obstacle is a non-living body. Then, the conscious signal and the unconscious signal are jointly marked as dynamic signals, and the dynamic signals and the static signals are used as obstacle signals and transmitted to the foreign object analysis terminal.

[0016] As a further solution of the present invention, the method for obtaining tracking information is as follows:

[0017] Set an interval time, and the interval time is a threshold. When an obstacle signal is detected, the drone emits a detection signal into the flight space at intervals of the interval time, and records the emission time, emission position, and emission angle. The detection signal includes an image signal and a radar signal. At the same time, when the signal is reflected back, the reception time of the reflected signal and the signal reception angle are recorded. Then, the emission time, emission position, and emission angle of the drone at intervals of the intermittent time, and the reception time of the reflected signal and the signal reception angle are used as tracking signals.

[0018] As a further solution of the present invention, the analysis method for the foreign object analysis terminal to receive a static signal is as follows:

[0019] S1: When the drone detects an obstacle, taking the position where the drone detects the obstacle as the origin and the flight direction of the drone as the positive direction, a spatial coordinate system is established;

[0020] S2: Detect the obstacle signal. If the obstacle signal is a static signal, extract the image signal and the radar signal in the detection signal. First, obtain the contour of the static object according to the image signal, and then, from the transmission and reception times of the radar signal, use the formula to obtain the static distance value D1. The static distance value D1 is the straight-line distance between the drone and the static object at the time point when the drone receives the radar signal. t1 is the time difference between the transmission and reception of the radar signal, θ2 represents the angle at which the drone receives the radar signal, θ1 represents the angle at which the drone emits the radar signal, and VFs is the flight speed of the drone;

[0021] S3: According to the preset trajectory of the drone, and in combination with Ds1 = t1 × VF, obtain the flight distance Ds1 of the drone within the time t1. According to the position of the drone at this time, based on the contour of the static object, the static distance value, and the angle of reception of the radar signal, mark the static object in the spatial coordinate system;

[0022] S4: Mark the position of the static object in the spatial coordinate system as the dangerous area, set the area within a distance of X1 around the static object as the warning area, and mark the area outside the warning area as the safe area, where X1 is a threshold value.

[0023] As a further solution of the present invention, the analysis method for the foreign object analysis terminal to receive the dynamic signal is as follows:

[0024] When the obstacle signal is a dynamic signal, extract the tracking information and process it according to steps S2 and S3. Mark the positions of the UAV and the obstacle target at each time interval in the spatial coordinate system. At this time, there are multiple position coordinates Dzi(x i , y i , z i ) in the spatial coordinate system, where i = 1, 2,..., I, indicating that the UAV has detected the obstacle target I times;

[0025] When the continuous acquisition times are greater than or equal to a1, fit the position coordinates Dzi in the spatial coordinate system using a mathematical model to obtain the obstacle trajectory, and the obstacle trajectory is the flight trajectory of the obstacle target;

[0026] Use to obtain the obstacle moving speed Vi, I = i + 1, and tg represents the time interval;

[0027] Use to obtain the acceleration adj of the obstacle target moving, where j = 1, 2,..., J, and then take the average value of the acceleration to obtain the average acceleration adv.

[0028] As a further solution of the present invention, the method for obtaining the safe area, warning area, and dangerous area under the dynamic signal includes:

[0029] Judge the dynamic signal. When the dynamic signal is an unintentional signal, according to the obstacle trajectory, obtain the regional direction of the preset trajectory on the obstacle trajectory. Take any point on the obstacle trajectory as the starting point, and set the first safety distance d1 and the second safety distance d2 respectively. Mark the area within the first safety distance d1 in the regional direction as the dangerous area, mark the area between the first safety distance d1 and the second safety distance d2 as the warning area, and mark the area outside the second safety distance d2 as the safe area;

[0030] When the dynamic signal is an intentional signal, taking any point on the obstacle trajectory as the center point, set a first sphere and a second sphere with the first safety distance d1 and the second safety distance d2 as the radii respectively, where the radius of the first sphere is d1 and the radius of the second sphere is d2. Mark the area within the first sphere as the dangerous area, mark the area between the first sphere and the second sphere as the warning area, and mark the area outside the second sphere as the safe area. d1 < d2, and both d1 and d2 are thresholds.

[0031] As a further solution of the present invention, the method for obtaining the safe flight instruction and the obstacle avoidance instruction is as follows:

[0032] Mark the preset trajectory, the obstacle trajectory and its flight area in the space coordinate system respectively, and judge whether the entire preset trajectory of the UAV is within the safe area. If it is within the safe area, a safe flight instruction will be generated;

[0033] If there is a part of the preset trajectory that is not within the safe area, extract this part of the preset trajectory and mark it as the risk trajectory. At the same time, divide the distance of the risk trajectory by the flight speed to obtain the risk time tz;

[0034] Adopt to obtain the risk coefficient Fx, where Zx is the order value corresponding to the obstacle target, dz is the minimum distance between the risk trajectory and the obstacle target, adv is the average acceleration, c1, c2 and c3 are fixed coefficients respectively, λ1, λ2, λ3 and λ4 are proportionality coefficients respectively, Qn is the risk coefficient of the risk trajectory in the flight area, n takes values of 1 and 2, representing the warning area and the dangerous area respectively, and Q1 < Q2;

[0035] Compare the risk coefficient Fx with the risk threshold Fy. When the risk coefficient Fx is less than the risk threshold, a safe flight instruction is generated. On the contrary, when the risk coefficient is greater than or equal to the risk threshold, an arbitrary point is taken in the space coordinate system at this time, the trajectory after taking the arbitrary point is fitted, and the fitted curve is marked as the virtual trajectory. At this time, the virtual trajectory is processed again according to the above steps until the risk coefficient Fx is less than the risk threshold, and then this virtual trajectory is marked as the obstacle avoidance instruction.

[0036] As a further solution of the present invention, it further includes a flight information acquisition end and an information storage end;

[0037] The flight information acquisition end is used to collect the flight information of the UAV and transmit it to the foreign object analysis end. The flight information includes the preset trajectory;

[0038] An information storage terminal is used to store the risk information existing during the flight of the drone and transmit it to the spatial information detection terminal. The risk information includes the obstacle information encountered during the flight of the drone and the parameters set for the obstacle information. The obstacle information refers to the obstacles and the states of the obstacles, and the set parameters refer to the order values set according to the states of the obstacles.

[0039] Compared with the existing technologies, the advantages of the present invention are as follows:

[0040] In the present invention, the obstacles that the drone will encounter during flight are classified. When the drone detects an obstacle during flight, it first identifies the obstacle and generates a static signal and a dynamic signal according to the identification result. When the dynamic signal is detected, it first analyzes the movement trajectory of the obstacle target to obtain the obstacle trajectory, and respectively sets a safe area, a warning area, and a danger area according to the obstacle trajectory. The part of the preset trajectory that is not within the safe area is marked as a risk trajectory, and the risk coefficient in the risk trajectory is calculated. According to the calculation result, a safe flight instruction and an obstacle avoidance instruction are respectively obtained, so that the drone adjusts the corresponding flight trajectory according to different types of obstacles during flight, thereby reducing the flight risk and reducing the flight loss of the drone. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 It is a schematic structural diagram of the system of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0042] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments.

[0043] Referring to Figure 1 , the intelligent analysis and control system for the flight risk of the drone based on data analysis includes a flight information acquisition terminal, an information storage terminal, a speed detection terminal, a spatial detection terminal, a foreign object analysis terminal, a flight analysis terminal, and an action execution terminal;

[0044] The flight information acquisition terminal is used to acquire the flight information of the drone, where the flight information includes a preset trajectory, and the preset trajectory is the preset flight route when the drone flies. Then, the flight information acquisition terminal transmits the flight information to the foreign object analysis terminal;

[0045] The information storage terminal is used to store the risk information existing during the flight of the drone. Specifically, the risk information includes the obstacle information encountered during the flight of the drone and the parameters set for the obstacle information. The obstacle information refers to the obstacles and their states, and the set parameters refer to the order values set according to the obstacle states. For example, during the flight of the drone, the encountered obstacles are divided into static objects and dynamic objects. Static objects refer to objects that are fixed and immovable, such as buildings, etc. Dynamic objects refer to objects that move freely in space. Among them, dynamic objects include living things and non-living things. Living things refer to objects with vital signs, including various birds or flying insects, etc. Non-living things refer to objects without vital signs, including plastic bags or floating balloons, etc. Here, it should be further explained that objects controlled by humans are regarded as living things. After that, the information storage terminal is unidirectionally electrically connected to the space detection terminal;

[0046] The speed detection terminal is used to detect the real-time flight speed of the drone and mark the real-time detected flight speed as the flight speed per hour. After that, the speed detection terminal transmits the flight speed per hour to the foreign object analysis terminal. It should be further explained that during the flight of the drone, there are multiple modes set inside the system, and the flight speeds in each mode are different. However, due to reasons such as weather or its own factors, there will be a certain difference between the actual flight speed and the flight speed in the set mode. Therefore, during the flight of the drone, it is necessary to detect the actual flight speed to ensure the accuracy of data collection and analysis, and thus reduce the risk rate of the drone flight;

[0047] The space detection terminal is used to detect the obstacles in the flight space of the drone. Specifically, the method for detecting the obstacles in the flight space of the drone is as follows:

[0048] When the drone is flying in the air, it senses the space information in the flight space through the sensors carried on the drone. Among them, the sensors include cameras and lidar, etc. In this embodiment, the sensors are set to carry both a camera and a lidar;

[0049] If an obstacle is identified in the space information, first collect the image of the obstacle to generate an obstacle image, and match the obstacle image with the obstacle information in the risk information to obtain the state of the obstacle in the space information. If the obstacle state is a static state, generate a static signal. If the obstacle is in a moving state, further identify the obstacle and generate a conscious signal and an unconscious signal according to the identification result. Specifically, when the obstacle is a living thing, a conscious signal is generated at this time. When the obstacle is a non-living thing, an unconscious signal is generated at this time. At the same time, both the conscious signal and the unconscious signal are jointly marked as dynamic signals;

[0050] After that, the space detection end fuses the static signal, the awareness signal, and the unintentional signal into an obstacle signal. At the same time, when the space detection end generates an obstacle signal, it targets the detected obstacle and marks it as an obstacle target. Then, the space detection end conducts target tracking on the obstacle target to obtain tracking information. The specific method for obtaining the tracking information is as follows:

[0051] There is an interval time set in the space detection end. The interval time is a threshold, and its specific value is set by those skilled in the art. When an obstacle signal is detected, the drone emits detection signals into the flight space at intervals of the interval time and records the emission time, emission position, and emission angle. The detection signals include image signals and radar signals. At the same time, when the signals are reflected back, the time of receiving the reflected signals and the angles of receiving the signals are recorded. Then, the space detection end transmits the obstacle signal and the tracking information to the foreign object analysis end;

[0052] Based on the received obstacle signal and tracking information, the foreign object analysis end judges the movement trajectory of the foreign object. The specific method for judging the movement trajectory is as follows:

[0053] S1: When the drone detects an obstacle, taking the position where the drone detects the obstacle as the origin, and taking the flight direction of the drone as the positive direction, a space coordinate system is established. At the same time, the flight speed of the drone at this time is extracted and marked as VFs;

[0054] S2: Then the obstacle signal is extracted. If the obstacle signal is a static signal, the image signal and the radar signal in the detection signal are respectively extracted. The contour of the static object is obtained according to the image signal. Then, according to the transmission and reception times of the radar signal, the formula is used to obtain the static distance value D1. The static distance value D1 is the straight-line distance between the drone and the static object at the time point when the drone receives the radar signal. t1 is the time difference between the transmission and reception of the radar signal, θ2 represents the angle at which the drone receives the radar signal, and θ1 represents the angle at which the drone emits the radar signal;

[0055] S3: Then, in the space coordinate system, first, according to the preset trajectory of the drone and combined with Ds1 = t1 × VF, the flight distance Ds1 of the drone within t1 time is obtained, and the position of the drone at this time is marked in the space coordinate system. Then, according to the contour of the static object, the static distance value, and the angle of receiving the radar signal, the static object is marked in the space coordinate system;

[0056] S4: At this time, mark the position of the static object in the space coordinate system as the dangerous area, and at the same time set the area within a distance of X1 around the static object as the warning area. Finally, mark the area outside the warning area as the safe area. X1 is a threshold value, and the specific value is set by those skilled in the art. It should be further noted that there is no intersection between the dangerous area, the warning area, and the safe area;

[0057] S5: When the obstacle signal is a dynamic signal, extract the emission time and emission position of the detection signal in the tracking information, as well as the signal reception time, reception position, and signal reception angle of the detection signal. The signal here is the detection signal. Then, according to the above steps S2 and S3, mark the positions of the UAV and the obstacle target at each time interval in the space coordinate system. At this time, there are multiple position coordinates Dzi(x i , y i , z i ) in the space coordinate system, where i = 1, 2,..., I, indicating that the UAV has detected the obstacle target I times;

[0058] S6: When the continuous acquisition times are greater than or equal to a1, that is, I ≥ a1, at this time, fit the position coordinates Dzi in the space coordinate system using a mathematical model to obtain the curve and its algebraic expression respectively. The curve is the flight trajectory of the obstacle target, and mark the flight trajectory of the obstacle target as the obstacle trajectory. In this embodiment, the mathematical model is fitted using Matlab;

[0059] S7: At the same time, use to obtain the obstacle moving speed Vi between two adjacent position coordinates. It should be noted that here I = i + 1, and tg represents the time interval;

[0060] S8: Use to obtain the acceleration adj of the obstacle target moving, where j = 1, 2,..., J. Then take the average value of the acceleration, that is, use to obtain the average acceleration adv;

[0061] S9: Then, identify the specific signals in the dynamic signal. When the dynamic signal is an unintentional signal, at this time, according to the obstacle trajectory, obtain the regional direction of the preset trajectory in the obstacle trajectory area. At the same time, starting from any point on the obstacle trajectory, set the first safety distance d1 and the second safety distance d2 respectively. Mark the area within the first safety distance d1 in the regional direction as the dangerous area, mark the area between the first safety distance d1 and the second safety distance d2 as the warning area, and mark the area outside the second safety distance d2 as the safe area. It should be further noted here that the space area outside the regional direction also belongs to the safe area;

[0062] When the dynamic signal is an intentional signal, at this time, taking any point on the obstacle trajectory as the center point, a first sphere and a second sphere are set with the first safety distance d1 and the second safety distance d2 as the radii respectively. The radius of the first sphere is d1, and the radius of the second sphere is d2. At this time, the area within the first sphere is marked as the dangerous area, the area between the first sphere and the second sphere is marked as the warning area, and the area outside the second sphere is marked as the safe area. d1 < d2, and both d1 and d2 are thresholds, and their specific values are set by those skilled in the art according to the actual situation;

[0063] After that, the foreign object analysis terminal fuses the safe area, the warning area and the dangerous area into a flight area and transmits it to the flight analysis terminal;

[0064] The flight analysis terminal is used to generate an obstacle avoidance instruction for the UAV according to the flight area of the obstacle target and transmit it to the action execution terminal. The specific method for generating the obstacle avoidance instruction is as follows:

[0065] SS1: Mark the preset trajectory, the obstacle trajectory and its flight area of the UAV in the space coordinate system. Then, first judge whether the preset trajectory of the UAV is entirely within the safe area. If it is within the safe area, a safe flight instruction will be generated;

[0066] SS2: If there is a part of the preset trajectory that is not within the safe area, extract this part of the preset trajectory and mark it as the risk trajectory. Divide the distance of the risk trajectory by the flight speed to obtain the risk time tz;

[0067] SS3: Adopt to obtain the risk coefficient Fx. Zx is the order value corresponding to the obstacle target, dz is the minimum distance between the risk trajectory and the obstacle target, c1, c2 and c3 are fixed coefficients respectively, and their specific values are obtained by technicians through multiple calculations. λ1, λ2, λ3 and λ4 are proportionality coefficients respectively, Qn is the danger coefficient of the risk trajectory in the flight area, n takes values of 1 and 2, representing the warning area and the dangerous area respectively, and Q1 < Q2;

[0068] SS4: Then compare the risk coefficient Fx with the risk threshold Fy. When the risk coefficient Fx is less than the risk threshold, a safe flight instruction is generated. On the contrary, when the risk coefficient is greater than or equal to the risk threshold, at this time, an arbitrary point is taken in the space coordinate system, the trajectory after taking the arbitrary point is fitted, and at the same time, the fitted curve is marked as the virtual trajectory. At this time, the virtual trajectory is processed according to the above steps SS1 to SS3 until the risk coefficient Fx is less than the risk threshold. Then, this virtual trajectory is marked as the obstacle avoidance instruction. The risk threshold is set by those skilled in the art;

[0069] It should be further noted that when arbitrarily selecting points in the space coordinate system here, only the path in the risk trajectory is changed, and the trajectory in the safe area will not change, further reducing the impact of emergencies on the preset trajectory;

[0070] After that, the flight analysis terminal transmits the normal flight instruction and the obstacle avoidance instruction to the action execution terminal;

[0071] Based on the received normal flight instruction and obstacle avoidance instruction, the action execution terminal enables the UAV to fly according to the corresponding trajectory, thereby reducing the risk incidence rate of UAV flight.

[0072] The above is only a preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution and inventive concept of the present invention, making equivalent substitutions or changes, should be covered by the protection scope of the present invention.

Claims

1. The intelligent analysis and control system of UAV flight risk based on data analysis is characterized by: It includes speed detection terminal, space detection terminal, foreign object analysis terminal, flight analysis terminal and action execution terminal; The speed detection terminal is used to detect the flight speed of the drone in real time and transmit it to the foreign object analysis terminal; The space detection end is used to detect obstacles in the flight space of the drone. The obstacles in the flight space are identified by the sensors carried by the drone. When the sensor identifies an obstacle in the spatial information of the flight space, it collects an obstacle image, generates static signals and dynamic signals according to the obstacle image, sets an interval time, marks the obstacle as an obstacle target, and tracks it to obtain tracking information, wherein the dynamic signal includes a conscious signal and an unintentional signal. The space detection end then transmits the static signal, dynamic signal and tracking information to the foreign object analysis end; The foreign object analysis end is used to receive static signals, dynamic signals and tracking information, and then first mark the position of the obstacle detected by the drone as the origin and establish a spatial coordinate system. When a static signal is received, the static object is marked in the spatial coordinate system according to the obstacle image and the radar signal, and a safe area, a warning area and a dangerous area are set. When a dynamic signal is received, the trajectory of the obstacle is first fitted, and then the dynamic signal is judged, and the corresponding safe area, warning area and dangerous area are set according to the intentional signal and the unintentional signal. After that, the foreign object analysis end uses the safe area, warning area and dangerous area as the flight area at the same time and transmits them to the flight analysis end. Among them, the setting method of the safe area, warning area and dangerous area in the dynamic signal includes: The dynamic signal is judged. When the dynamic signal is an unintentional signal, the regional direction of the preset trajectory on the obstacle trajectory is obtained according to the obstacle trajectory. Taking any point on the obstacle trajectory as the starting point, the first safety distance d1 and the second safety distance d2 are set respectively. The area within the first safety distance d1 in the regional direction is marked as a dangerous area, the area between the first safety distance d1 and the second safety distance d2 is marked as a warning area, and the area outside the second safety distance d2 is marked as a safe area; If the dynamic signal is an intentional signal, take any point on the obstacle trajectory as the center point, set the first sphere and the second sphere as the radius respectively, where the radius of the first sphere is d1 and the radius of the second sphere is d2, mark the area inside the first sphere as the danger area, mark the area between the first sphere and the second sphere as the warning area, and mark the area outside the second sphere as the safety area, d1<d2, and d1 and d2 are both thresholds; The flight analysis end is used to mark the preset trajectory and obstacle trajectory of the drone in the spatial coordinate system, and at the same time, according to the flight area, mark the part of the preset trajectory that is not in the safe area as a risk trajectory, and calculate the risk coefficient in the risk trajectory. According to the calculation results, the safe flight instruction and the obstacle avoidance instruction are obtained respectively. Then, the flight analysis end transmits the safe flight instruction and the obstacle avoidance instruction to the action execution end, wherein the method for obtaining the safe flight instruction and the obstacle avoidance instruction includes: Mark the preset trajectory, obstacle trajectory and flight area in the spatial coordinate system respectively to determine whether the preset trajectory of the drone is all within the safe area. If it is within the safe area, a safe flight instruction will be generated; If there is a part of the preset trajectory that is not in the safe area, extract this part of the preset trajectory and mark it as a risk trajectory. At the same time, divide the distance of the risk trajectory by the flight speed to get the risk time tz; use The risk coefficient Fx is obtained, where Zx is the order value corresponding to the obstacle target, dz is the minimum distance between the risk trajectory and the obstacle target, adv is the average acceleration, c1, c2 and c3 are fixed coefficients, λ1, λ2, λ3 and λ4 are proportional coefficients, Qn is the risk coefficient of the risk trajectory in the flight area, n takes the values ​​of 1 and 2, representing the warning area and the danger area respectively, and Q1<Q2; The risk factor Fx is compared with the risk threshold Fy. When the risk factor Fx is less than the risk threshold, a safe flight instruction is generated. On the contrary, when the risk factor Fx is greater than or equal to the risk threshold, an arbitrary point is selected in the spatial coordinate system, and the trajectory after the arbitrary point is fitted. At the same time, the fitted curve is marked as a virtual trajectory. At this time, the virtual trajectory is reprocessed according to the above steps until the risk factor Fx is less than the risk threshold, and then the virtual trajectory is marked as an obstacle avoidance instruction; The action execution end is used to fly according to the received designated corresponding flight trajectory.

2. The UAV flight risk intelligent analysis and control system based on data analysis according to claim 1 is characterized in that: The specific methods for obtaining static signals and dynamic signals are as follows: When the drone is flying in the air, it senses spatial information in the flying space through the sensors it carries; If an obstacle is identified in the spatial information, the image of the obstacle is collected to generate an obstacle image, and the obstacle image is matched with the obstacle information to obtain the obstacle state in the spatial information. If the obstacle state is static, a static signal is generated. If the obstacle is in motion, the obstacle is further identified, and a conscious signal and an unintentional signal are generated accordingly according to the identification result. The conscious signal is the signal generated when the obstacle is a living thing, and the unintentional signal is the signal generated when the obstacle is a non-living thing. Then, the conscious signal and the unintentional signal are jointly marked as a dynamic signal, and the dynamic signal and the static signal are used as obstacle signals and transmitted to the foreign body analysis end.

3. The UAV flight risk intelligent analysis and control system based on data analysis according to claim 2 is characterized in that: The method to obtain tracking information is: Set the interval time, which is the threshold. When an obstacle signal is detected, the UAV transmits a detection signal to the flight space at intervals, and records the transmission time, transmission position and transmission angle. The detection signal includes image signal and radar signal. At the same time, when the signal is reflected back, the reflected signal reception time and signal reception angle are recorded. After that, the transmission time, transmission position and transmission angle of the UAV at each interval and the reflected signal reception time and signal reception angle are used as tracking signals.

4. The UAV flight risk intelligent analysis and control system based on data analysis according to claim 1 is characterized in that: The analysis method of the static signal received by the foreign body analysis end is: S1: When the drone detects an obstacle, the spatial coordinate system is established with the position where the drone detects the obstacle as the origin and the flight direction of the drone as the positive direction; S2: Detect obstacle signals. If the obstacle signal is a static signal, extract the image signal and radar signal from the detection signal. First, obtain the outline of the static object based on the image signal, and then use the formula based on the sending and receiving time of the radar signal. The static distance value D1 is obtained. The static distance value D1 is the straight-line distance between the drone and the static object at the time when the drone receives the radar signal. t1 is the time difference between sending and receiving the radar signal. θ2 represents the angle at which the drone receives the radar signal. θ1 represents the angle at which the drone transmits the radar signal. VFs is the flight speed of the drone. S3: According to the preset trajectory of the drone, the flight distance Ds1 of the drone in time t1 is obtained in combination with Ds1=t1×VF. According to the position of the drone at this time, the outline of the static object, the static distance value and the angle of radar signal reception, the static object is marked in the spatial coordinate system; S4: Mark the position of the static object in the spatial coordinate system as a dangerous area, set the area within a distance X1 around the static object as a warning area, and mark the area outside the warning area as a safe area, where X1 is the threshold.

5. The UAV flight risk intelligent analysis and control system based on data analysis according to claim 4 is characterized in that: The analysis method of the dynamic signal received by the foreign body analysis end is: If the obstacle signal is a dynamic signal, the tracking information is extracted and processed according to steps S2 and S3. The position of the UAV and the position of the obstacle target in each interval are marked in the spatial coordinate system. At this time, there are multiple position coordinates Dzi(x i ,y i , z i ), i = 1, 2, ..., I, indicating that the UAV has detected the obstacle target I times; When the number of continuous acquisitions is greater than or equal to a1, the position coordinate Dzi in the space coordinate system is fitted using a mathematical model to obtain the obstacle trajectory, which is the flight trajectory of the obstacle target; use Obtain the obstacle moving speed Vi, I = i + 1, tg represents the interval time; use The acceleration adj of the obstacle target is obtained, j=1, 2, ..., J, and then the acceleration is averaged to obtain the acceleration average adv.

6. The UAV flight risk intelligent analysis and control system based on data analysis according to claim 1 is characterized in that: It also includes a flight information collection terminal and an information storage terminal; The flight information collection terminal is used to collect the flight information of the UAV and transmit it to the foreign object analysis terminal. The flight information includes the preset trajectory; The information storage end is used to store the risk information that exists during the flight of the UAV and transmit it to the space information detection end. The risk information includes the obstacle information encountered by the UAV during flight and the parameters set for the obstacle information. The obstacle information refers to the obstacles and the obstacle status, and the set parameters refer to the order values ​​set according to the obstacle status.

Images