A flight path planning control system and method for a drone and a drone

Abstract

The application discloses a kind of unmanned plane flight path planning control system, method and unmanned plane, including voice input module, path generation module, acquisition data module;Path generation module is connected with voice input module and acquisition data module respectively;Path generation module is used when receiving user demand data, generates execution instruction, and sends execution instruction and user demand data to acquisition data module, generates unmanned plane flight path according to user demand data, acquisition data module sent unmanned plane movement data and target map characteristic value;Acquisition data module is used when receiving execution instruction and user demand data, obtains unmanned plane movement data, and calculates target map characteristic value according to unmanned plane movement data, and unmanned plane movement data and target map characteristic value are transmitted to path generation module;Solve the technical problem that the flight path planning control system of existing unmanned plane leads to the complexity of path planning is higher.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicles (UAVs), and in particular to a flight path planning control system and method for an UAV, and the UAV. Background Art

[0002] An unmanned aerial vehicle (UAV), or drone, is an unmanned aircraft controlled by a wireless remote control device and an autonomous program, or fully or intermittently by a computer. In the civilian sector, UAVs, due to their flexibility, portability, affordability, and ease of operation compared to manned aircraft, have found widespread application in various fields, including agriculture, urban management, scientific research, environmental protection, and public security.

[0003] In the practical application of drones, drone path planning is a very important topic. Drone path planning mainly refers to minimizing the number of drones used in a given scenario, reaching the target location in a shorter time, and improving flight efficiency.

[0004] Existing flight path planning and control systems for drones mostly use the Dijkstra algorithm and the A* algorithm. These methods can obtain the shortest path from the initial point to the destination, but during the implementation process, complete environmental information must be known in advance, resulting in a high complexity in path planning. Summary of the Invention

[0005] The present invention provides a flight path planning control system and method for an unmanned aerial vehicle (UAV), and the UAV, for solving the technical problem that the existing flight path planning control system for UAVs causes high complexity in path planning.

[0006] A first aspect of the present invention provides a flight path planning and control system for an unmanned aerial vehicle, the system comprising a voice input module, a path generation module, and a data collection module;

[0007] The path generation module is connected to the voice input module and the data collection module respectively;

[0008] The voice input module is used to obtain a user voice signal, and generate user demand data based on the user voice signal and transmit the generated user demand data to the path generation module;

[0009] The path generation module is configured to generate an execution instruction upon receiving the user demand data, and send the execution instruction and the user demand data to the data collection module, and generate a UAV flight path based on the user demand data, the UAV motion data sent by the data collection module, and the target map feature value;

[0010] The data collection module is configured to, when the execution instruction and the user demand data are received, acquire the unmanned aerial vehicle motion data, calculate the target map feature value according to the unmanned aerial vehicle motion data, and transmit the unmanned aerial vehicle motion data and the target map feature value to the path generation module.

[0011] Optionally, the user demand data includes voice demand features and text demand features; and the voice input module is specifically configured to:

[0012] acquire a user voice signal, pre-process the user voice signal, and output a target voice digital signal;

[0013] extract a signal feature of the target voice digital signal by using a preset voice feature extraction technology, and output a voice signal feature;

[0014] input the voice signal feature into a preset deep learning model, and output a text sequence;

[0015] convert the text sequence into voice by using a voice conversion technology, and generate voice demand features;

[0016] extract a text feature of the text sequence by using a preset named entity recognition model, generate text demand features, and transmit the text demand features to the path generation module.

[0017] Optionally, the unmanned aerial vehicle motion data includes a target location position, a current time position of the unmanned aerial vehicle, two-dimensional laser scanning data, environment image data, and inertial measurement data; and the data collection module includes a Beidou positioning module and a SLAM module.

[0018] The Beidou positioning module and the SLAM module are both connected to the path generation module.

[0019] The Beidou positioning module is configured to, in response to the execution instruction, acquire the current time position of the unmanned aerial vehicle, generate the target location position according to the text demand features, and transmit the target location position and the current time position of the unmanned aerial vehicle to the path generation module.

[0020] The SLAM module is configured to, in response to the execution instruction, acquire the two-dimensional laser scanning data, the environment image data, and the inertial measurement data, calculate a target map feature value according to the two-dimensional laser scanning data, the environment image data, and the inertial measurement data, and transmit the target map feature value to the path generation module.

[0021] Optionally, the SLAM module is specifically configured to:

[0022] in response to the execution instruction, acquire the two-dimensional laser scanning data, the environment image data, and the inertial measurement data.

[0023] detecting, by using a preset corner point detection algorithm, a plurality of pixel points in the environment image data, and outputting a corner pixel value of each of the pixel points;

[0024] respectively comparing each of the corner pixel values with a preset pixel threshold value;

[0025] taking, as a target pixel point, any pixel point corresponding to a corner pixel value greater than the preset pixel threshold value;

[0026] calculating, by using a preset image processing function group, a feature descriptor corresponding to each of the target pixel points according to a gradient amplitude of each of the target pixel points;

[0027] constructing a feature descriptor matrix according to each of the feature descriptors;

[0028] calculating, by using a preset extended Kalman filter function group, a Kalman gain matrix and a state vector prediction value according to the feature descriptor matrix, the two-dimensional laser scanning data and the inertial measurement data;

[0029] calculating a target map feature value according to the state vector prediction value, the Kalman gain matrix, the feature descriptor matrix, the two-dimensional laser scanning data and the inertial measurement data;

[0030] calculating a Boolean value according to the target map feature value, and determining whether the Boolean value is zero;

[0031] if the Boolean value is not zero, transmitting the target map feature value to the path generation module.

[0032] Optionally, the text requirement feature includes a user constraint condition, an initial state of the unmanned aerial vehicle and a target state of the unmanned aerial vehicle; and the path generation module is specifically configured to:

[0033] performing consistency comparison on the current time position of the unmanned aerial vehicle and the target location position;

[0034] if the comparison is inconsistent, constructing an unmanned aerial vehicle field by using the user constraint condition;

[0035] updating the user constraint condition based on a preset state update function to determine a target user constraint condition;

[0036] constructing an unmanned aerial vehicle problem by using the target user constraint condition, the initial state of the unmanned aerial vehicle, the target state of the unmanned aerial vehicle, the target map feature value, the target location position and the current time position of the unmanned aerial vehicle;

[0037] generating an unmanned aerial vehicle flight path according to the unmanned aerial vehicle problem and the unmanned aerial vehicle field.

[0038] Optionally, the system further comprises a voice output module, a display output module, and an inspection module and a flight module connected in sequence;

[0039] The inspection module is connected to the voice output module, the display output module and the path generation module respectively;

[0040] The voice output module is communicatively connected with the voice input module;

[0041] The voice output module is used to receive the voice demand characteristics sent by the voice input module and play the voice according to the voice demand characteristics;

[0042] The display output module is used to receive the UAV flight path sent by the inspection module and display the UAV flight path;

[0043] The checking module is configured to, upon receiving the UAV flight path sent by the path generating module, determine whether the UAV flight path meets the preset flight conditions; if the UAV flight path meets the preset flight conditions, generate a flight instruction, and transmit the flight instruction and the UAV flight path to the flight module;

[0044] The flight module is used to respond to the flight instruction and control the drone to fly according to the drone flight path.

[0045] A second aspect of the present invention provides a method for planning a flight path of an unmanned aerial vehicle, comprising:

[0046] When receiving a user voice signal, collecting drone motion data, and calculating target map feature values ​​based on the drone motion data;

[0047] generating user demand data according to the user voice signal;

[0048] A drone flight path is generated according to the user demand data, the drone motion data, and the target map feature value.

[0049] A third aspect of the present invention provides a drone, comprising a drone body and a propeller pair;

[0050] The propeller pairs are symmetrically arranged on both sides of the drone fuselage;

[0051] The top of the drone body is provided with a voice input button, a user operation screen and a map display screen;

[0052] A SLAM sensor is provided at the bottom of the drone fuselage;

[0053] The unmanned aerial vehicle body is internally provided with a control mainboard, which is connected with the voice input button, the user operation screen, the map display screen, the propeller pair and the SLAM sensor respectively.

[0054] The control mainboard is loaded with the flight path planning control system of the unmanned aerial vehicle as described above.

[0055] The fourth aspect of the present application provides a computer device, comprising a memory and a processor, the memory stores a computer program, and the computer program is executed by the processor to make the processor execute the steps of the flight path planning method of the unmanned aerial vehicle as described above.

[0056] The fifth aspect of the present application provides a computer readable storage medium, which stores a computer program, and the computer program is executed to realize the flight path planning method of the unmanned aerial vehicle as described above.

[0057] From the above technical solutions, the present application has the following advantages:

[0058] The first aspect of the above technical solution of the present application provides a flight path planning control system of an unmanned aerial vehicle, which comprises a voice input module, a path generation module and a data acquisition module; the path generation module is connected with the voice input module and the data acquisition module; the voice input module is used to obtain user voice signals and generate user demand data according to the user voice signals, and then transmit the user demand data to the path generation module; the path generation module is used to generate execution instructions when receiving the user demand data, and send the execution instructions and the user demand data to the data acquisition module, and generate the flight path of the unmanned aerial vehicle according to the user demand data, the unmanned aerial vehicle motion data sent by the data acquisition module and the target map feature value; the data acquisition module is used to obtain the unmanned aerial vehicle motion data when receiving the execution instructions and the user demand data, calculate the target map feature value according to the unmanned aerial vehicle motion data, and transmit the unmanned aerial vehicle motion data and the target map feature value to the path generation module; based on the above scheme, the process of generating the flight path of the unmanned aerial vehicle by the path generation module according to the user demand data, the unmanned aerial vehicle motion data sent by the data acquisition module and the target map feature value does not need to know the complete environmental information in advance, and can reduce the complexity of path planning.

[0059] The second aspect of the above technical solution of the present application provides a flight path planning method of an unmanned aerial vehicle, which acquires unmanned aerial vehicle motion data when receiving user voice signals, and calculates target map feature values according to the unmanned aerial vehicle motion data; generates user demand data according to the user voice signals; generates the flight path of the unmanned aerial vehicle according to the user demand data, the unmanned aerial vehicle motion data and the target map feature values; the above process does not need to know the complete environmental information in advance, and can reduce the complexity of path planning.

[0060] The third aspect of the above technical solution of the present application provides a UAV, comprising a pair of propellers symmetrically arranged on both sides of a UAV body; a voice input button, a user operation screen and a map display screen are arranged on the top of the UAV body; a SLAM sensor is arranged on the bottom of the UAV body; a control mainboard is built in the UAV body, and the control mainboard is connected with the voice input button, the user operation screen, the map display screen, the pair of propellers and the SLAM sensor respectively; the control mainboard is loaded with a flight path planning control system of the UAV; based on the above solution, the UAV generates a flight path of the UAV based on the flight path planning control system loaded on the control mainboard, according to user voice signals and UAV motion data obtained through the voice input button and the SLAM sensor, and the process does not need to know complete environmental information in advance, and can reduce the complexity of path planning. BRIEF DESCRIPTION OF DRAWINGS

[0061] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or the prior art description will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0062] Figure 1 A structural schematic diagram of a flight path planning control system of a UAV provided for the first embodiment of the present application;

[0063] Figure 2 A flowchart of calculating target map feature values of a SLAM module provided for the first embodiment of the present application;

[0064] Figure 3 A work flowchart of the flight path planning control system of the UAV provided for the first embodiment of the present application;

[0065] Figure 4 A step flowchart of a flight path planning method of a UAV provided for the second embodiment of the present application;

[0066] Figure 5 A structural schematic diagram of a UAV provided for the third embodiment of the present application;

[0067] Among them, the meanings of the marks of the drawings are as follows:

[0068] 1, propeller; 2, map display screen; 3, UAV body; 4, voice input button; 5, SLAM sensor; 6, user operation screen. DETAILED DESCRIPTION

[0069] The embodiment of the present application provides a flight path planning control system, method and unmanned aerial vehicle for solving the technical problem of high complexity of path planning caused by the existing flight path planning control system for unmanned aerial vehicles.

[0070] In order to make the application purpose, features and advantages of the present application more obvious and easy to understand, the technical solutions in the embodiments of the present application will be described clearly and completely below in combination with the drawings in the embodiments of the present application. Obviously, the following described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the protection scope of the present application.

[0071] Please refer to Figure 1 , Figure 1 The structure diagram of the flight path planning control system of the unmanned aerial vehicle provided in the first embodiment of the present application is shown.

[0072] The flight path planning control system of the unmanned aerial vehicle provided in the present application comprises a voice input module, a path generation module and a data acquisition module.

[0073] The voice input module is used for obtaining a user voice signal and generating user demand data according to the user voice signal and then transmitting the user demand data to the path generation module.

[0074] It should be noted that the voice input module is mainly responsible for receiving the voice signal of the user and converting it into a text instruction. The voice input module receives the voice signal of the user by using a microphone device to ensure the clarity and integrity of the voice signal.

[0075] In the present embodiment, the voice input module converts the voice signal into a text instruction by using a voice recognition technology and processes noise and dialect factors to improve the recognition accuracy. The instruction analysis module analyzes the converted text instruction and extracts the flight demand of the user, i.e. the voice demand features and the text demand features, which both include the flight time, the flight destination, the required avoidance area and other information.

[0076] The path generation module is used for generating an execution instruction when receiving the user demand data and sending the execution instruction and the user demand data to the data acquisition module, and generating the flight path of the unmanned aerial vehicle according to the user demand data, the unmanned aerial vehicle motion data sent by the data acquisition module and the target map feature value.

[0077] It should be noted that the path generation module is mainly responsible for planning path generation. Through the voice input of the user and the initial position and environmental state of the unmanned aerial vehicle determined by the SLAM module (Simultaneous Localization and Mapping), a numerical planning problem is generated and a numerical planning solver is used to generate a planning path.

[0078] In the embodiment, the application uses ROS (Robot Operating System) as an underlying communication and coordination platform to realize efficient communication and data sharing between modules. Through the sensor driver and data processing node in ROS, obstacle information, path information and dynamic changes in the environment are obtained and processed. The system obtains environmental data in real time through sensors, constructs a dynamic environment model, and uses SLAM technology and the Beidou satellite positioning system to realize dynamic updating of the environment and self-positioning. The application uses numerical planning technology to realize real-time optimal path planning of the unmanned aerial vehicle. The initial state and geographical position of the unmanned aerial vehicle are obtained through the SLAM technology, and a numerical planning task is constructed according to the overall goal of the task and the resource constraints. The planner is called to plan and solve the task to generate a planning path, and the planning path is updated in real time according to the environmental perception module.

[0079] The data acquisition module is configured to, when receiving the execution instruction and the user demand data, acquire the unmanned aerial vehicle motion data, and calculate target map feature values according to the unmanned aerial vehicle motion data, and transmit the unmanned aerial vehicle motion data and the target map feature values to the path generation module.

[0080] As a further improvement, the user demand data includes voice demand features and text demand features; the voice input module is specifically configured to:

[0081] acquire a user voice signal, pre-process the user voice signal, and output a target voice digital signal; extract signal features of the target voice digital signal by using a preset voice feature extraction technology, and output voice signal features; input the voice signal features into a preset deep learning model as an input, and output a text sequence; convert the text sequence into a voice by using a voice conversion technology, and generate voice demand features; extract text features of the text sequence by using a preset named entity recognition model, and generate text demand features and transmit the text demand features to the path generation module.

[0082] It should be noted that the voice input module performs multiple steps such as audio acquisition, preprocessing, feature extraction, model recognition, text processing and demand extraction; specifically, the voice signal is first acquired by a microphone and pre-processed, the pre-processing process includes removing noise and invalid parts, the signal amplitude is scaled to a specific range through normalization operation, and noise is eliminated through a low-pass filter, and the pre-processing process can be represented as:

[0083]

[0084] Where x[n] is a continuous analog signal, representing the value of the nth sampling point; x a is the user's voice signal; n is the number of sampling points; T s is the sampling period; f s is the sampling rate, which means the number of samples per second; x q [n] is a discrete digital signal, which represents the digital signal of the nth sampling point; Q(·) is a quantization function, which represents the conversion of a continuous analog signal into a discrete digital signal; x norm [n] is the normalized digital signal of the nth sampling point; max(·) is the maximum function, which is used to extract the maximum value in the sampling process; y[n] is the target speech digital signal; α is the filter coefficient, which takes a value between 0 and 1 and represents the contribution ratio of the current input signal and the output signal at the previous moment to the current output signal; y[n-1] is the target speech digital signal at the historical moment, which represents the target speech digital signal at the previous moment.

[0085] Furthermore, features are extracted from the audio signal (target speech digital signal) using a preset speech feature extraction technology. The preset speech feature extraction technology is Mel Frequency Cepstral Coefficients (MFCC). The process can be expressed as:

[0086] C(n)=MFCC(y[n]);

[0087] Where C(n) is the speech signal feature, which represents the MFCC coefficient of the nth sampling point; MFCC(·) is the Mel-frequency cepstral coefficient, a feature extraction method; and y[n] is the target speech digital signal.

[0088] Furthermore, the speech signal features are used as the input of the preset deep learning model, and a text sequence is output. That is, the extracted features C(n) are analyzed using the preset deep learning model, and the corresponding text sequence V is output. This process can be expressed as: V = LSTM(C(n)), where V is the text sequence, C(n) is the speech signal feature, and LSTM(·) is the preset deep learning model.

[0089] Further, the text sequence is converted into speech by using a speech conversion technology to generate speech demand features, wherein the speech conversion technology is a TTS technology (Text To Speech), and the extracted speech demand features Speech are fed back to the user by using a speech output module; a pre-set named entity recognition model is used to extract text features from the text sequence to generate text demand features and transmit them to the path generation module, wherein the pre-set named entity recognition model is a named entity recognition model (NER, Named Entity Recognition); the above process can be represented as:

[0090]

[0091] wherein V is the text sequence; Speech is the speech demand feature; {(w i ,t i )} is the text demand feature, which is composed of a plurality of words w i and a plurality of entity types t i ; NER(·) is the pre-set named entity recognition model; TTS(·) is the speech conversion technology.

[0092] As a further improvement, the UAV motion data includes target location position, UAV current time position, two-dimensional laser scanning data, environment image data and inertial measurement data; the data acquisition module includes a Beidou positioning module and a SLAM module;

[0093] The Beidou positioning module and the SLAM module are connected with the path generation module;

[0094] The Beidou positioning module is used to obtain the UAV current time position in response to the execution instruction, and generate the target location position according to the text demand features, and transmit the target location position and the UAV current time position to the path generation module;

[0095] The SLAM module is used to obtain the two-dimensional laser scanning data, the environment image data and the inertial measurement data in response to the execution instruction, and calculate the target map feature value according to the two-dimensional laser scanning data, the environment image data and the inertial measurement data, and transmit the target map feature value to the path generation module.

[0096] It should be noted that the Beidou positioning module obtains the UAV current time position in response to the execution instruction, and generates the target location position according to the flight destination in the text demand features, and transmits the target location position and the UAV current time position to the path generation module.

[0097] In this embodiment, the SLAM module is used to determine the drone's position and environmental status, helping it navigate during flight using simultaneous localization and mapping (SLAM) technology. Sensors such as lidar and cameras are used to acquire environmental data, building and updating an environmental map in real time. SLAM algorithms (such as particle filters and Kalman filters) are used to achieve high-precision positioning and navigation.

[0098] As a further improvement, the SLAM module is specifically used to:

[0099] In response to the execution instruction, the two-dimensional laser scanning data, environmental image data and inertial measurement data are obtained; a preset corner detection algorithm is used to detect corners of multiple pixels in the environmental image data, and the corner pixel value of each pixel is output; each corner pixel value is compared with the preset pixel threshold; any pixel corresponding to a corner pixel value greater than the preset pixel threshold is used as a target pixel; a preset image processing function group is used to calculate the feature descriptor corresponding to each target pixel according to the gradient amplitude of each target pixel; a feature descriptor matrix is ​​constructed based on each feature descriptor; a preset extended Kalman filter function group is used to calculate the Kalman gain matrix and the state vector prediction value based on the feature descriptor matrix, the two-dimensional laser scanning data and the inertial measurement data; the target map eigenvalue is calculated based on the state vector prediction value, the Kalman gain matrix, the feature descriptor matrix, the two-dimensional laser scanning data and the inertial measurement data; a Boolean value is calculated based on the target map eigenvalue, and it is determined whether the Boolean value is zero; if the Boolean value is not zero, the target map eigenvalue is transmitted to the path generation module.

[0100] Please note that Figure 2 The steps performed by the SLAM module include sensor data acquisition, front-end feature extraction and matching, back-end processing to update positioning and map, loop detection and map reconstruction; Among them, the process of sensor data acquisition is to respond to execution instructions and obtain two-dimensional laser scanning data, environmental image data and inertial measurement data. The SLAM module obtains environmental data through a variety of sensors, including lidar, camera and inertial measurement unit (IMU); the lidar provides the two-dimensional laser scanning data L at the current time t t , represents the distance information of the surrounding environment, and can accurately measure the distance between the drone and objects in the environment. The camera captures the environmental image data I at the current time t t , records the visual information of the environment, and is often used for feature point extraction and matching. The IMU provides the inertial measurement data U at the current time t t , including acceleration and angular velocity information, which is used to estimate the motion state and attitude changes of the drone.

[0101] Furthermore, the front-end feature extraction and matching process is to extract useful features from sensor data for use in subsequent map construction and positioning. For environmental image data, the Harris Corner Detection algorithm is used to detect key points. That is, a preset corner detection algorithm is used to detect corners of multiple pixels in the environmental image data and output the corner pixel value of each pixel. The calculation process of the corner pixel value can be expressed as:

[0102]

[0103] Where R is the pixel value of the corner point; M is the second-order matrix of the pixel point; det(M) is the determinant of the matrix; trace(M) is the trace of the matrix; α is an empirical constant with a value of 0.04-0.06; I x is the gradient of the pixel in the horizontal direction; I y is the vertical gradient of the pixel.

[0104] Furthermore, the pixel value of each corner point is compared with the preset pixel threshold; the pixel point corresponding to each corner point pixel value greater than the preset pixel threshold is taken as the target pixel point, and the preset image processing function group is used to calculate the feature descriptor corresponding to each target pixel point according to the gradient amplitude of each target pixel point; the present invention uses SIFT descriptor (Scale-Invariant Feature Transform), that is, feature descriptor to represent image features, and after detecting the map key point (target pixel point) through Harris corner detection algorithm, calculates the gradient amplitude of the target pixel point, the projection angle q (x, y, z) of the xy plane, and the angle of the gradient direction relative to the z axis The preset pixel threshold can be set as needed, and the present invention is not limited thereto. The preset image processing function can be expressed as follows:

[0105]

[0106] Among them, M(x,y,z) is the gradient amplitude; G x (x, y, z) is the gradient in the horizontal direction; G y (x, y, z) is the gradient in the vertical direction; G z (x,y,z) is the gradient in the vertical direction; q(x,y,z) is the projection angle on the xy plane; is the angle of the gradient direction relative to the z-axis; SIFT(·) is the SIFT descriptor; d x,y,z is the feature descriptor.

[0107] Furthermore, the back-end processing process for updating positioning and maps is as follows: using an extended Kalman filter (EKF) to estimate the position and attitude of the UAV, that is, using a preset extended Kalman filter function group based on the feature descriptor matrix, two-dimensional laser scanning data and inertial measurement data to calculate the Kalman gain matrix and state vector prediction value; wherein, the preset extended Kalman filter function group can be expressed as:

[0108]

[0109] in, is the predicted value of the state vector at the current time t, where the state vector x is composed of the drone posture feature x r Combined with the map feature m, the drone posture feature x r Mainly includes v x ,v y ,v z , w represents the linear velocity and angular velocity of the drone in the horizontal, vertical, and vertical directions respectively. The map feature m mainly includes the three-dimensional coordinates and descriptors of static objects and feature points in the environment; x t-1 is the state vector at time t-1, indicating the state vector at the previous moment; U t is the inertial measurement data at the current time t; P t - is the predicted value of the covariance matrix at the current time t; F t is the Jacobian matrix of the state transfer matrix; P t-1 is the covariance matrix at time t-1; is the transpose of the Jacobian matrix of the state transfer matrix; Q t is the process noise covariance matrix; K t is the Kalman gain matrix at the current moment; is the transpose of the Jacobian matrix of the observation function h(·) to the state; H t is the Jacobian matrix of the observation function h(·) to the state; R t is the covariance matrix of observation noise; SIFT t is the feature descriptor matrix at the current moment; is the horizontal coordinate of the i-th target pixel at the current moment; is the ordinate of the i-th target pixel at the current moment; is the vertical coordinate of the i-th target pixel at the current moment; is the feature descriptor of the i-th target pixel at the current moment; is the predicted value of the state vector at time t-1; K t-1 is the Kalman gain matrix at time t-1; z t-1is the observation data at time t-1, which includes two-dimensional laser scanning data and inertial measurement data; h(·) is an observation function, which describes the mapping relationship from state to observation; SIFT t-1 is the feature descriptor matrix at time t-1; I is an identity matrix; K t-1 is the Kalman gain matrix at time t-1; H t-1 is the Jacobian matrix of the observation function h(·) with respect to the state at time t-1; is the covariance matrix prediction value at time t-1; f(·) is a state transition function.

[0110] It is worth mentioning that the initial value of the covariance matrix and the initial value of the state vector can be set as needed, and the application is not limited thereto.

[0111] Further, according to the state vector prediction value, the Kalman gain matrix, the feature descriptor matrix, the two-dimensional laser scanning data and the inertial measurement data, the target map feature value is calculated and transmitted to the path generation module; this process can be represented as:

[0112]

[0113] Wherein, m t is the target map feature value at the current time t; is the map feature prediction value in the state vector prediction value at the current time t; K t is the Kalman gain matrix at the current time; z t is the observation data at the current time t, which includes two-dimensional laser scanning data and inertial measurement data; is the state vector prediction value at the current time t; SIFT t is the feature descriptor matrix at the current time.

[0114] Further, the process of loop detection is: detecting whether the unmanned aerial vehicle has returned to the previously visited position during the movement, thereby improving the consistency and accuracy of the map, calculating a Boolean value according to the target map feature value, and judging whether the Boolean value is zero; the process of map reconstruction is: if the Boolean value is not zero, the target map feature value is transmitted to the path generation module; if the Boolean value is zero, the subsequent update step is stopped, and the unmanned aerial vehicle continues to fly, and when the next time t+1 is reached, the unmanned aerial vehicle movement data is reacquired, and the unmanned aerial vehicle flight path is updated using the reacquired unmanned aerial vehicle movement data, until the unmanned aerial vehicle reaches the target location; wherein the calculation process of the Boolean value is:

[0115] Loop_closure=detect_loop(m t-1 ,m t );

[0116] Wherein, Loop closure is a Boolean value; detect loop (·) is a loop detection method, which can specifically use feature matching, bag-of-words model, image sequence matching, etc.; m t-1 is a target map feature value at time t-1; m t is a target map feature value at current time t.

[0117] As a further improvement, the text requirement features further include user constraints, initial state of the UAV and target state of the UAV; the path generation module is specifically configured to:

[0118] Consistency comparison is performed on the current time position of the UAV and the target location position; if the comparison is inconsistent, the user constraints are used to construct the UAV domain; the user constraints are updated based on the preset state update function to determine the target user constraints; the target user constraints, the initial state of the UAV, the target state of the UAV, the target map feature value, the target location position and the current time position of the UAV are used to construct the UAV problem; and the UAV flight path is generated according to the UAV problem and the UAV domain.

[0119] It should be noted that, if the consistency comparison of the current time position of the UAV and the target location position is consistent, it indicates that the UAV has reached the target location position, and there is no need to update the flight path, and the operation is stopped.

[0120] It should be noted that, if the comparison is inconsistent, the user constraints are used to construct the UAV domain; that is, according to the predicates B, objects O, actions A, action costs c involved in the pre-set UAV flight problem and the user constraints C in the text requirement features that need to be met by the UAV during flight, the corresponding domain domain is constructed, which can be represented as:

[0121] domain = DOMAIM (B, O, A, c, C) ;

[0122] Wherein, B is a predicate; O is an object; A is an action; c is an action cost; domain is a UAV domain; DOMAIM (·) is a domain file construction method, which can specifically use a language large model to help generate.

[0123] Further, the initial problem file mainly contains the initial state S0 of the UAV in the text requirement features, the target state S g(UAV initial state and UAV target state) and the specific values ​​of the flight constraints specified by the user, including the UAV's maximum flight time CT, the UAV's maximum flight distance CD, and the UAV's maximum flight altitude CH, etc. The UAV's initial position IP, target position GP, ​​and real-time position P during flight are determined by the Beidou satellite positioning module through the Beidou satellite positioning system, which can be expressed as:

[0124]

[0125] Among them, T1 is the time coordinate; Y1 is the satellite's spatial vertical coordinate; X1 is the satellite's spatial horizontal coordinate; Z1 is the satellite's spatial vertical coordinate; T is the drone's time coordinate; X is the drone's spatial horizontal coordinate; Y is the drone's spatial vertical coordinate; Z is the drone's spatial vertical coordinate.

[0126] Furthermore, the initial problem file can be expressed as:

[0127] problem_init=PROBLEM(S0,S g ,CT,CD,CH,IP,GP,m t );

[0128] Among them, problem_init is the initial problem file; PROBLEM(·) is the problem generation method, which can be generated by using a large language model; S0 is the initial state of the drone; S g is the target state; CT is the maximum flight time of the UAV; CD is the maximum flight distance of the UAV; CH is the maximum flight altitude of the UAV; IP is the initial position of the UAV; GP is the target position; m t The target map feature value at the current time t.

[0129] Furthermore, during the flight, according to the target map feature value m provided by the SLAM module t , the real-time position of the UAV provided by the Beidou satellite and the updated numerical constraints are used to dynamically update the problem file, that is, the user constraints are updated based on the preset state update function to determine the target user constraints; the target user constraints, the initial state of the UAV, the target state of the UAV, the target map feature value, the target location and the current position of the UAV are used to construct the UAV problem; among them, the user constraints include the maximum flight time of the UAV, the maximum flight distance of the UAV and the maximum flight altitude of the UAV, and the target user constraints include the target maximum flight time of the UAV, the target maximum flight distance of the UAV and the target maximum flight altitude of the UAV; the process can be expressed as:

[0130]

[0131] wherein CT' is the target maximum flight time of the UAV; CD' is the target maximum flight distance of the UAV; CH' is the target maximum flight height of the UAV; Condition_update(·) is a condition updating method, mainly for numerical deduction of the numerical conditions such as CT, CD, CH according to the current status of the UAV; PROBLEM(·) is a problem generating method, which can specifically use a language large model to help generate; problem t is the UAV problem at the current time.

[0132] Further, according to the description of the domain file and the problem file (i.e. the UAV problem and the UAV domain), a numerical planner LNM-PLAN is called to obtain the optimal planning path under the current state. The UAV flies according to the planning path, wherein the process can be represented as:

[0133] Plan t =LNM_PALN(domain,problem t );

[0134] wherein Plan t is the UAV flight path at the current time; domain is the UAV domain; problem t is the UAV problem at the current time; LNM_PALN(·) is a numerical planner.

[0135] It is worth mentioning that the UAV flies according to the UAV flight path at the current time, and when the next time t+1 is reached, the UAV motion data is reacquired, and the UAV flight path is updated according to the reacquired UAV motion data, until the real-time position and the target location position are consistent, and the planning of the path is stopped.

[0136] As a further improvement, the system further comprises a voice output module, a display output module, and a checking module and a flight module connected in sequence;

[0137] The checking module is connected with the voice output module, the display output module and the path generation module respectively;

[0138] The voice output module is in communication connection with the voice input module;

[0139] The voice output module is configured to receive the voice demand feature sent by the voice input module and play voice according to the voice demand feature; the display output module is configured to receive the UAV flight path sent by the checking module and display the UAV flight path; the checking module is configured to, when receiving the UAV flight path sent by the path generation module, judge whether the UAV flight path meets the preset flight condition, if the UAV flight path meets the preset flight condition, generate a flight instruction and transmit the flight instruction and the UAV flight path to the flight module; the flight module is configured to control the UAV to fly according to the UAV flight path in response to the flight instruction.

[0140] It should be noted that the voice output module is mainly responsible for providing feedback and confirmation to the user, covering the confirmation of the flight destination and demand, the success of the generated path planning, and the flight restrictions of the current location, etc. Through voice feedback, the module can confirm the user's flight destination and demand. By using natural language processing technology, the voice feedback is ensured to be clear and accurate. In terms of path planning, the module will inform the user whether the path is successfully generated, and if not, the reasons and possible solutions will be provided. At the same time, the module will also prompt the user of the flight restrictions of the current location, such as no-fly zones and weather conditions, etc.

[0141] Further, the display output module is mainly responsible for displaying the state information of the UAV and the flight path generated after the specified flight demand. The state information display displays the real-time state information of the UAV, including battery level, flight speed and height, etc., providing a clear and easy-to-understand graphical interface. The path display displays the generated flight path after the user specifies the flight demand, providing detailed information of the path, such as the places to pass through and the estimated flight time, etc.

[0142] Further, the checking module is mainly responsible for checking the feasibility of the flight path generated by the path generation module, and dynamically adjusting the path according to the weather and regions passed through in the flight path, to ensure that no unforeseen obstacles or dangers are encountered during the actual flight; the flight module is mainly responsible for executing the detailed flight path checked by the checking module. And when the UAV is flying, the flight state of the UAV is monitored in real time to detect possible abnormal situations, through connection with the path generation module, the relevant data (inertial measurement data, environmental image data, etc.) are reacquired and timely adjusted to ensure the smooth completion of the flight task.

[0143] It is worth mentioning that the preset flight condition is the latest published no-fly zone obtained by the checking module through the Internet of Things from the relevant website, and the checking module checks whether the region passed through by the path contains the no-fly zone; through the planning path checking result, that is, the judgment result, the planning path is timely modified, and the no-fly zone is added as a constraint condition to the domain file, to ensure that the next generated planning path will not pass through the no-fly zone again.

[0144] Exemplarily, please refer to Figure 3 , according to the voice input module receives the user's voice instruction (voice signal), get user flight demand (voice demand characteristics and text demand characteristics), and through the Beidou satellite module obtains the real-time positioning position of the unmanned aerial vehicle, and the target location position to be reached, according to the SLAM module obtains the unmanned aerial vehicle motion data, according to the unmanned aerial vehicle motion data output target map characteristic value, path generation module judges whether the current real-time position is consistent with the target location position, if consistent, it indicates that the destination is reached, if not consistent, it indicates that the destination is not reached, combined with the current state of the unmanned aerial vehicle in the user flight demand, target map characteristic value and other data, output the current time of the unmanned aerial vehicle planning path, through the inspection module to judge whether the unmanned aerial vehicle planning path is executable, if executable, then through the flight module controls the unmanned aerial vehicle to fly according to the path, if not executable, then correct the planning path.

[0145] As a comparison of technical effects, reference can be made in combination with the prior art. The interaction and control of the existing unmanned aerial vehicle equipment: through the Internet of Things technology and the unmanned aerial vehicle, the equipment state is obtained in real time. Through the display screen, microphone, loudspeaker and other devices, human-computer interaction is carried out. Path planning and optimization: using numerical planning method, the optimal path is planned in complex three-dimensional flight environment to minimize flight time, energy consumption or cost. Dynamic environment adaptability: using the real-time data transmission and updating mechanism provided by ROS, combined with numerical planning method, the unmanned aerial vehicle can perceive and adapt to dynamic environmental changes in real time, such as weather changes, flight obstacles, etc. Path safety and conflict avoidance: combined with the perception and control module provided by ROS, the path safety control and conflict avoidance of the unmanned aerial vehicle are realized to ensure that it complies with the flight rules and safely flies during flight.

[0146] At the same time, the existing path planning algorithm has slow response speed in dynamic environment, and it is difficult to perceive and avoid sudden obstacles in real time, resulting in insufficient safety of unmanned aerial vehicle flight. Multi-objective optimization difficulty: many path planning methods are difficult to consider multiple optimization objectives (such as shortest path, lowest energy consumption, least time, etc.) at the same time, usually only single objective optimization can be carried out, which cannot meet the comprehensive demand in actual application. Poor adaptability to dynamic environment: the unmanned aerial vehicle may encounter dynamic changes in the environment (such as moving obstacles, weather changes, etc.) during flight, and the existing technology is difficult to update and adjust the flight path in real time, with poor adaptability. Low system integration: the unmanned aerial vehicle path planning system usually lacks unified architecture and interface standard, the cooperative work efficiency between each functional module is low, and efficient system integration is difficult to realize. Low computing efficiency and high resource consumption: the traditional path planning algorithm has high computational complexity, especially when dealing with large-scale data and complex environment, the computing efficiency is low, and it is difficult to realize real-time path planning under the condition of limited computing resources.

[0147] To solve the above problems, the application provides a flight path planning control system for a UAV, which fully utilizes ROS as a core platform to realize the control of the UAV, acquisition and processing of sensor data, and other functions. Through the modular structure and rich function packages of ROS, the system achieves high scalability and flexibility. Each module of the UAV, such as navigation, remote control, and sensor data processing, can be uniformly managed and coordinated through ROS, making the development, debugging, and maintenance of the system more convenient and efficient. The SLAM technology is used to realize dynamic updating of the environment and self-positioning. The numerical planning technology is used to efficiently, safely, and optimally plan the flight path of the UAV according to different tasks and constraints. In a complex environment, the UAV can realize autonomous flight by reasonably analyzing and optimizing the path to achieve the optimal goal of task requirements and path planning.

[0148] In summary, the application uses ROS as a core platform, and the system has good scalability and flexibility, can easily extend new function modules or integrate new sensors, and is suitable for different application scenarios and needs. The system integrates the SLAM module to achieve efficient dynamic environment updating and self-positioning. The advanced speech recognition technology can achieve high-accuracy and fast-response speech recognition, and has demand extraction and multi-round dialogue functions, making user interaction more intelligent and convenient. The advanced numerical planning technology can efficiently, safely, and optimally plan the flight path of the UAV according to different tasks and constraints, and compared with traditional planning methods, it can obtain a more optimized flight path, has real-time path planning and execution capabilities, can respond to environmental changes and flight task requirements in a timely manner, and has good robustness to cope with complex environments and abnormal situations, ensuring the safe flight of the UAV. The application uses efficient numerical planning algorithms and the real-time communication mechanism provided by ROS to realize real-time path planning and obstacle avoidance of the UAV in a dynamic environment. The numerical planning method considers factors such as flight time, energy consumption, and cost to plan the optimal flight path to meet different application needs. The real-time data transmission and updating mechanism of ROS enables the UAV to perceive and adapt to environmental changes in real time and adjust the flight path in a timely manner. The modular system architecture based on ROS ensures seamless communication and cooperation between function modules, improves the overall performance and scalability of the system, and optimizes the computational complexity of the path planning algorithm to realize efficient real-time path planning through parallel computing technology and special hardware acceleration.

[0149] It is worth mentioning that, in addition to using numerical programming algorithm, the application can also consider integrating deep learning technology to use neural network to learn and plan the flight path of the unmanned aerial vehicle. Deep learning has strong representation learning ability and can learn flight environment and task requirements from a large amount of data to realize more intelligent and adaptive path planning. In combination with augmented reality technology, an unmanned aerial vehicle navigation system can be designed to superimpose real-time camera images and virtual maps to provide a more intuitive and intuitive navigation interface. The flight path planning system of the unmanned aerial vehicle can also be used for environmental monitoring and inspection tasks, including air quality monitoring, vegetation coverage investigation, building structure detection, etc. Through automatic path planning and execution, large-scale monitoring tasks can be efficiently completed to provide data support and decision reference for environmental protection and urban management.

[0150] In the embodiment of the application, the flight path planning control system of the unmanned aerial vehicle is provided, which comprises a voice input module, a path generation module and a data acquisition module. The path generation module is connected with the voice input module and the data acquisition module. The voice input module is used to obtain user voice signals and generate user demand data according to the user voice signals and then transmit the user demand data to the path generation module. The path generation module is used to generate execution instructions when receiving the user demand data, and send the execution instructions and the user demand data to the data acquisition module, generate the flight path of the unmanned aerial vehicle according to the user demand data, the unmanned aerial vehicle motion data sent by the data acquisition module and the target map feature value. The data acquisition module is used to obtain the unmanned aerial vehicle motion data when receiving the execution instructions and the user demand data, calculate the target map feature value according to the unmanned aerial vehicle motion data, and transmit the unmanned aerial vehicle motion data and the target map feature value to the path generation module. Based on the above scheme, the process of generating the flight path of the unmanned aerial vehicle by the path generation module according to the user demand data, the unmanned aerial vehicle motion data sent by the data acquisition module and the target map feature value does not need to know the complete environmental information in advance, and the complexity of path planning can be reduced.

[0151] Please refer to Figure 4 , Figure 4 The flight path planning method of the unmanned aerial vehicle provided in the second embodiment of the application is shown in the step flowchart.

[0152] The flight path planning method of the unmanned aerial vehicle provided in the application comprises:

[0153] Step 401, when receiving the user voice signals, collect the unmanned aerial vehicle motion data and calculate the target map feature value according to the unmanned aerial vehicle motion data.

[0154] In the embodiment, when receiving the user voice signals, the unmanned aerial vehicle motion data is collected, and the target map feature value is calculated according to the unmanned aerial vehicle motion data.

[0155] Step 402, generating user demand data according to the user voice signal.

[0156] In this embodiment, the user demand data is generated according to the user voice signal.

[0157] Step 403, generating the flight path of the unmanned aerial vehicle according to the user demand data, the unmanned aerial vehicle motion data and the target map feature value.

[0158] In this embodiment, the flight path of the unmanned aerial vehicle is generated according to the user demand data, the unmanned aerial vehicle motion data and the target map feature value.

[0159] In the embodiment of the application, the flight path planning method of the unmanned aerial vehicle is provided, when the user voice signal is received, the unmanned aerial vehicle motion data is collected, and the target map feature value is calculated according to the unmanned aerial vehicle motion data; the user demand data is generated according to the user voice signal; the flight path of the unmanned aerial vehicle is generated according to the user demand data, the unmanned aerial vehicle motion data and the target map feature value; the above process does not need to know the complete environment information in advance, and the complexity of the path planning can be reduced.

[0160] Please refer to Figure 5 , Figure 5 The structure diagram of the unmanned aerial vehicle provided in the third embodiment of the application.

[0161] The unmanned aerial vehicle provided by the application comprises an unmanned aerial vehicle body 3 and a pair of propellers.

[0162] The pair of propellers is symmetrically arranged on both sides of the unmanned aerial vehicle body 3.

[0163] The top of the unmanned aerial vehicle body 3 is provided with a voice input button 4, a user operation screen 6 and a map display screen 2.

[0164] The bottom of the unmanned aerial vehicle body 3 is provided with a SLAM sensor 5.

[0165] The unmanned aerial vehicle body 3 is internally provided with a control mainboard, and the control mainboard is connected with the voice input button 4, the user operation screen 6, the map display screen 2, the pair of propellers and the SLAM sensor 5 respectively.

[0166] The control mainboard is loaded with the flight path planning control system of the unmanned aerial vehicle as described in the above first embodiment.

[0167] The pair of propellers comprises two propellers 1.

[0168] It should be noted that the unmanned aerial vehicle body 3 is made of light and strong materials such as carbon fiber, aluminum alloy, plastic, etc. to provide the necessary strength and lightness; the unmanned aerial vehicle body 3 also has a built-in battery and a control mainboard connected to the control motor, which is a brushless motor (BLDC) that provides the necessary thrust for the unmanned aerial vehicle flight; the propeller 1 converts the rotation of the motor into thrust, and propellers of different sizes and shapes are suitable for different flight requirements; the battery provides power for the motor and other electronic devices, usually a lithium polymer (LiPo) battery, with capacity and voltage selected according to specific needs; the unmanned aerial vehicle body 3 also has a built-in flight controller (FCU, File Control Unit) connected to the control mainboard, an attitude sensor, and an electronic speed controller (ESC, Electronic Stability Control); the flight controller (FCU) is responsible for processing various sensor data, executing control algorithms, and outputting control signals to adjust the speed and direction of the motor; the attitude sensor includes a gyroscope and an accelerometer to monitor the attitude and motion state of the unmanned aerial vehicle in real time; the electronic speed controller (ESC) can adjust the speed and direction of the motor and receive instructions from the flight controller to control the motor; SLAM sensor: including laser radar (LiDAR), vision sensor (such as depth camera) and inertial measurement unit (IMU). For simultaneous localization and mapping, to help the unmanned aerial vehicle navigate and locate outdoors.

[0169] Further, the unmanned aerial vehicle body 3 is also provided with a microphone and a loudspeaker, by pressing the voice input button 4, the microphone can recognize voice and receive user voice instructions; the loudspeaker provides voice feedback to the user about the processing status of the unmanned aerial vehicle. The map display screen 2 is used to display the map and display the current position of the unmanned aerial vehicle on the map, and the user can also input the required flight requirements by hand through the user operation screen 6.

[0170] In the embodiment of the present application, the unmanned aerial vehicle provided by the present application comprises a pair of propellers symmetrically arranged on both sides of the unmanned aerial vehicle body; a voice input button, a user operation screen and a map display screen are arranged on the top of the unmanned aerial vehicle body; a SLAM sensor is arranged on the bottom of the unmanned aerial vehicle body; a control mainboard is built-in the unmanned aerial vehicle body, and the control mainboard is connected with the voice input button, the user operation screen, the map display screen, the pair of propellers and the SLAM sensor respectively; the control mainboard is loaded with a flight path planning control system of the unmanned aerial vehicle; based on the above scheme, the unmanned aerial vehicle generates the flight path of the unmanned aerial vehicle based on the flight path planning control system of the unmanned aerial vehicle loaded on the control mainboard, according to the user voice signal and the unmanned aerial vehicle motion data obtained through the voice input button and the SLAM sensor, without needing to know the complete environmental information in advance, which can reduce the complexity of path planning.

[0171] The embodiment of the present application further provides a computer device, comprising a memory and a processor, the memory stores a computer program; the computer program is executed by the processor, so that the processor executes the steps of the flight path planning method of the unmanned aerial vehicle in the above embodiment two.

[0172] The embodiment of the present application further provides a computer readable storage medium, which stores a computer program / instruction, and the computer program / instruction is executed by a processor to realize the steps of the flight path planning method of the unmanned aerial vehicle in the above embodiment two.

[0173] In several embodiments provided in the present application, it should be understood that the disclosed system, device and method can be implemented in other manners. For example, the above described device embodiments are merely schematic, and the division of units is merely a logical function division, and there can be another division manner in actual implementation. For example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed mutual couplings or direct couplings or communication connections between different units, can be indirect couplings or communication connections through some interfaces, devices or units, and can be electrical, mechanical or in other forms.

[0174] The units described as separated components can or can not be physically separated, and the components displayed as units can or can not be physical units, i.e., they can be located in one place, or can be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purposes of the embodiments of the present application.

[0175] The above described and above embodiments are merely used to illustrate the technical solutions of the present application, rather than limit the present application; although the present application is described in detail with reference to the foregoing embodiments, it should be understood by those skilled in the art that the technical solutions recorded in the foregoing embodiments can be modified, or some technical features can be replaced by equivalent replacements; and these modifications or replacements do not make the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A flight path planning and control system for an unmanned aerial vehicle, characterized in that: The system includes a voice input module, a path generation module, and a data collection module; The path generation module is connected to the voice input module and the data collection module respectively; The voice input module is used to obtain a user voice signal, and generate user demand data based on the user voice signal and transmit the generated user demand data to the path generation module; The path generation module is configured to generate an execution instruction upon receiving the user demand data, and send the execution instruction and the user demand data to the data collection module, and generate a UAV flight path based on the user demand data, the UAV motion data sent by the data collection module, and the target map feature value; The data collection module is configured to, upon receiving the execution instruction and the user demand data, obtain the UAV motion data, calculate the target map feature value based on the UAV motion data, and transmit the UAV motion data and the target map feature value to the path generation module; The user demand data includes voice demand features and text demand features; The UAV motion data includes the target location, the current location of the UAV, two-dimensional laser scanning data, environmental image data and inertial measurement data; The text requirement features include user constraints, the initial state of the drone, and the target state of the drone; the path generation module is specifically used to: Performing a consistency comparison between the current position of the drone and the position of the target location; If the comparison is inconsistent, the user constraints are adopted to construct the drone field; Updating the user constraint condition based on a preset state update function to determine the target user constraint condition; Constructing a drone problem using the target user constraint, the drone initial state, the drone target state, the target map feature value, the target location, and the drone current location; A drone flight path is generated based on the drone problem and the drone domain.

2. The UAV flight path planning and control system according to claim 1, characterized in that: The voice input module is specifically used to: Acquire a user voice signal, pre-process the user voice signal, and output a target voice digital signal; Extracting signal features from the target voice digital signal using a preset voice feature extraction technology, and outputting voice signal features; Using the speech signal features as input to a preset deep learning model and outputting a text sequence; Performing voice conversion on the text sequence using voice conversion technology to generate voice demand features; A preset named entity recognition model is used to extract text features from the text sequence, generate text requirement features and transmit them to the path generation module.

3. The UAV flight path planning and control system according to claim 2, characterized in that: The data acquisition module includes a Beidou positioning module and a SLAM module; The Beidou positioning module and the SLAM module are both connected to the path generation module; The Beidou positioning module is used to respond to the execution instruction, obtain the current position of the drone, generate the target location according to the text requirement characteristics, and transmit the target location and the current position of the drone to the path generation module; The SLAM module is used to respond to the execution instruction, obtain two-dimensional laser scanning data, environmental image data and inertial measurement data, and calculate the target map feature value based on the two-dimensional laser scanning data, the environmental image data and the inertial measurement data, and transmit the target map feature value to the path generation module.

4. The UAV flight path planning and control system according to claim 3, characterized in that: The SLAM module is specifically used for: In response to the execution instruction, acquiring two-dimensional laser scanning data, environmental image data and inertial measurement data; Performing corner point detection on a plurality of pixel points in the environmental image data using a preset corner point detection algorithm, and outputting a corner point pixel value of each pixel point; Comparing the pixel value of each corner point with a preset pixel threshold respectively; Taking any pixel point corresponding to a corner pixel value greater than the preset pixel threshold as a target pixel point; Calculating a feature descriptor corresponding to each target pixel point according to the gradient amplitude of each target pixel point using a preset image processing function group; Constructing a feature descriptor matrix according to each of the feature descriptors; Using a preset extended Kalman filter function group to calculate a Kalman gain matrix and a state vector prediction value according to the feature descriptor matrix, the two-dimensional laser scanning data and the inertial measurement data; Calculating target map feature values ​​according to the state vector prediction value, the Kalman gain matrix, the feature descriptor matrix, the two-dimensional laser scanning data, and the inertial measurement data; Calculating a Boolean value according to the target map feature value, and determining whether the Boolean value is zero; If the Boolean value is not zero, the target map feature value is transmitted to the path generation module.

5. The flight path planning and control system of the UAV according to claim 2, characterized in that: The system also includes a voice output module, a display output module, and an inspection module and a flight module connected in sequence; The inspection module is connected to the voice output module, the display output module and the path generation module respectively; The voice output module is communicatively connected with the voice input module; The voice output module is used to receive the voice demand characteristics sent by the voice input module and play the voice according to the voice demand characteristics; The display output module is used to receive the UAV flight path sent by the inspection module and display the UAV flight path; The checking module is configured to, upon receiving the UAV flight path sent by the path generating module, determine whether the UAV flight path meets the preset flight conditions; if the UAV flight path meets the preset flight conditions, generate a flight instruction, and transmit the flight instruction and the UAV flight path to the flight module; The flight module is used to respond to the flight instruction and control the drone to fly according to the drone flight path.

6. A flight path planning method for an unmanned aerial vehicle, applied to the flight path planning control system for an unmanned aerial vehicle according to claim 1, characterized in that: include: When receiving a user voice signal, collecting drone motion data, and calculating target map feature values ​​based on the drone motion data; generating user demand data according to the user voice signal; A drone flight path is generated according to the user demand data, the drone motion data, and the target map feature value.

7. A drone, characterized in that: Includes drone body and propeller pair; The propeller pairs are symmetrically arranged on both sides of the drone fuselage; The top of the drone body is provided with a voice input button, a user operation screen and a map display screen; A SLAM sensor is provided at the bottom of the drone fuselage; The drone body is equipped with a control mainboard, which is respectively connected to the voice input button, the user operation screen, the map display screen, the propeller pair and the SLAM sensor; The control mainboard is equipped with a flight path planning control system for the UAV as described in any one of claims 1-5.

8. A computer device, characterized in that: The method comprises a memory and a processor, wherein a computer program is stored in the memory, and when the computer program is executed by the processor, the processor executes the steps of the flight path planning method for a drone as claimed in claim 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed, the flight path planning method for the UAV as claimed in claim 6 is implemented.

Images