Unmanned aerial vehicle path planning method based on improved leapfrog algorithm

Abstract

The invention relates to the technical field of unmanned aerial vehicle path planning methods, in particular to an unmanned aerial vehicle path planning method based on an improved leapfrog algorithm. According to the method, dynamic weight adjustment information is obtained by carrying out dynamic adaptation matching analysis on flight stage information and environment threat information, a flight stage weight mapping matrix is established, and fitness function weights are dynamically adjusted according to core targets in take-off, cruise and landing stages; weight correction is carried out in combination with environmental data such as obstacle grid density and a meteorological threat coefficient, so that the algorithm focuses on core constraints in the current stage, and target conflicts caused by global unified weight are avoided; according to the method, environmental threats are sensed in real time by means of multi-sensor fusion, when abnormal states such as over-high obstacle density and strong wind are detected, an algorithm is triggered to be suddenly stopped and restarted, a population is reconstructed according to the current position, threat avoidance weight is enhanced, the defect that an existing mechanism can be updated only after the current iteration period is completed is overcome, and the method is suitable for large-scale popularization and application. And the response time to an abnormal environment is shortened.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicle path planning methods, and specifically to an unmanned aerial vehicle path planning method based on an improved shuffled frog leaping algorithm. Background Art

[0002] When the current unmanned aerial vehicle path planning method based on the improved shuffled frog leaping algorithm conducts path planning for different flight stages, the fitness function weight of the predetermined algorithm is usually a globally unified value. Due to the differences in environmental sensitivity, constraint priorities, and dynamic requirements required for different flight stages, it cannot be dynamically adjusted when facing different flight stages. Moreover, when the unmanned aerial vehicle encounters an abnormal environment, the algorithm needs to quickly adjust the path to avoid risks. However, the existing mechanism needs to complete the current iteration cycle before it can be updated, resulting in response delays and path rigidity. Summary of the Invention

[0003] The present invention provides an unmanned aerial vehicle path planning method based on an improved shuffled frog leaping algorithm to solve the above technical problems.

[0004] The first aspect of the present invention provides an unmanned aerial vehicle path planning method based on an improved shuffled frog leaping algorithm, including the following steps:

[0005] Step 1: The central control module of the unmanned aerial vehicle retrieves the stage data set in the database and inputs the stage data set into the multi-stage environment analysis model to obtain flight stage information and environmental threat information.

[0006] As a further improvement of the present invention, for the multi-stage environment analysis model, its specific implementation steps are:

[0007] The stage data set includes environmental data and unmanned aerial vehicle state data; obtain the minimum safe flight height and constraint height given by the pre-designed, denoted as H and H C , and obtain the flight height, speed data, and attitude information of the unmanned aerial vehicle according to the unmanned aerial vehicle state data;

[0008] Denote the flight height of the unmanned aerial vehicle as h. When 0 < h < H, generate the initial selected flight stages of the unmanned aerial vehicle as the take-off stage and the landing stage. Obtain the vertical speed according to the speed data of the unmanned aerial vehicle, and analyze the take-off and landing of the unmanned aerial vehicle based on the vertical speed of the unmanned aerial vehicle and the height change of the unmanned aerial vehicle. When the vertical speed of the unmanned aerial vehicle is positive and the flight height approaches H, and according to the attitude information of the unmanned aerial vehicle, when the attitude angle of the unmanned aerial vehicle is greater than the preset angle threshold, generate the corresponding flight stage as the take-off stage; when the vertical speed of the unmanned aerial vehicle is negative and the flight height approaches 0, and according to the attitude information of the unmanned aerial vehicle, when the attitude angle of the unmanned aerial vehicle approaches 0, generate the corresponding flight stage as the landing stage;

[0009] When H < h < H CWhen the initial flight phase of the drone is the cruise phase, the vertical speed and flight speed are obtained based on the speed data of the drone. When the vertical speed is equal to zero or the drone ascends and descends at a preset stable speed, and the horizontal speed is within the preset stable speed range, a speed stability signal is generated; when the attitude angle is within the preset horizontal flight inclination range according to the attitude information of the drone, an attitude stability signal is generated; when the speed stability signal and the attitude stability signal of the drone are detected, the corresponding flight phase is the cruise phase; the environmental data corresponding to each flight phase is obtained and analyzed to obtain environmental threat information.

[0010] As a further improvement of the present invention, the environmental data is analyzed, and the specific analysis method is as follows:

[0011] The static environment and the dynamic environment are obtained from the environmental data. The geometric parameters of each static obstacle and no-fly zone are obtained according to the static environment, and the three-dimensional spatial modeling of each static obstacle and no-fly zone is obtained through the geometric parameters. The three-dimensional spatial modeling of the static obstacles and no-fly zone is marked as obstacle grids through a three-dimensional rasterization algorithm. Based on the search step of the drone, the obstacle grids corresponding to each static obstacle are obtained, and the total number of obstacle grids is counted. The ratio of the total number of obstacle grids to the total number of search grids within the search step range is calculated to obtain the obstacle grid density, and the obstacle grid density is used as the corresponding static threat coefficient.

[0012] The meteorological conditions and dynamic obstacle information are obtained from the dynamic environment; the wind speed condition, precipitation condition, visibility condition, and temperature condition are obtained according to the meteorological conditions. The threat analysis of each meteorological condition is carried out according to a preset meteorological threat analysis unit. The meteorological threat analysis unit includes the safety meteorological threshold and the meteorological upper limit threshold preset for each meteorological condition, which are respectively denoted as where q represents the meteorological condition; the part of each meteorological condition that exceeds the corresponding safety meteorological threshold is divided into multiple meteorological threat intervals where n, n + 1, n + 2, n + 3 are all constants; based on the multiple meteorological threat intervals corresponding to each meteorological adjustment, the corresponding meteorological threat coefficients are set for each meteorological threat interval corresponding to each meteorological condition The threat analysis model corresponding to the meteorological condition is Each meteorological condition is input into the corresponding threat analysis model to obtain the corresponding meteorological threat coefficient.

[0013] Obtain the forward direction of the drone, mark the area within the preset angle range of the forward direction of the drone as a high-risk area, and set a high-risk coefficient. Denote the area other than the high-risk area as the normal area. According to the dynamic obstacle information, obtain the obstacle distances between each dynamic obstacle and the drone. Obtain the risk warning boundary pre-designed for the drone. When a dynamic obstacle enters the risk warning boundary, mark the corresponding dynamic obstacle as a threatening obstacle. Based on the dynamic obstacle information, obtain the obstacle space volume and obstacle distance of each threatening obstacle;

[0014] Perform normalization processing on the obstacle space volume and obstacle distance and take their numerical values. Through Obtain the dynamic risk value; multiply the dynamic risk value of the threatening obstacle corresponding to the high-risk area by the high-risk coefficient to obtain the dynamic threat coefficient; mark the dynamic risk value corresponding to the normal area as the dynamic threat coefficient.

[0015] Aggregate the meteorological threat coefficients and dynamic threat coefficients corresponding to the dynamic environment to obtain the total dynamic threat information. Aggregate the total dynamic threat information and the static threat coefficient corresponding to the static environment to obtain the environmental threat information.

[0016] Step 2: The central control module of the drone performs dynamic adaptive matching analysis on the flight phase information and the environmental threat information to obtain dynamic weight adjustment information.

[0017] As a further improvement of the present invention, perform dynamic adaptive matching analysis on the flight phase information and the environmental threat information. The specific analysis method is as follows:

[0018] Based on the central control module, obtain multiple drone paths for the currently preset analysis time period through the improved shuffled frog leaping algorithm. Identify the corresponding path lengths for each drone path. According to the environmental threat information, obtain the static threat coefficient and the total dynamic threat information. Obtain the basic weight vectors for each flight phase pre-designed, and the basic weight vectors include path length weight, static threat weight, and dynamic threat weight.

[0019] Based on the static threat coefficient and the total dynamic threat information, obtain the obstacle grid density, meteorological threat coefficient, and dynamic threat coefficient corresponding to each drone path. Calculate and sum the meteorological threat coefficients corresponding to each meteorological condition to obtain the total meteorological threat coefficient. Use the multivariate Bessel function weighted sum model to process the obstacle grid density, the total meteorological threat coefficient, and the dynamic threat coefficient and output the total threat value. When the total threat value of a drone path is greater than the preset threat threshold, mark the corresponding drone path as a high-threat path.

[0020] Identify the flight phases of the UAV. When the flight phase information of the UAV corresponds to the take-off phase and the landing phase, the corresponding dynamic weight adjustment information is to reduce the local search step size corresponding to the improved shuffled frog leaping algorithm and reduce the path length weight, and obtain the high-threat occupancy ratio of the high-threat paths in the total number of UAV paths. When the high-threat occupancy ratio is greater than the preset upper limit ratio value, increase the static threat weight and the dynamic threat weight on the basis of the basic weight vector.

[0021] When the flight phase information of the UAV corresponds to the cruise phase, the corresponding dynamic weight adjustment information is to increase the local search step size and increase the path length weight. Similarly, when the high-threat occupancy ratio is greater than the preset upper limit ratio value, increase the static threat weight and the dynamic threat weight on the basis of the basic weight vector.

[0022] As a further improvement of the present invention, the specific processing process of the multivariate Bessel function weighted sum model is as follows:

[0023] Normalize the obstacle grid density, the total meteorological threat coefficient, and the dynamic threat coefficient and take their values, and denote them as sg, dw; use the multivariate Bessel function weighted sum formula Calculate the total threat value wf corresponding to each UAV path k ; where, k represents the kth UAV path; β i represents the weight coefficients corresponding to the three variables of the obstacle grid density, the total meteorological threat coefficient, and the dynamic threat coefficient; sg max 、 dw max respectively represent the maximum values corresponding to the three variables; J v represents the first kind of Bessel function, and the order ν>0 ensures monotonic increase in the r k value range. For example, when ν = 1, J1(r) first increases and then decreases when r>0, but r can be limited k within its monotonic increasing interval.

[0024] Step 3: The UAV central control module extracts the total threat value corresponding to the travel path, and performs real-time environment determination and analysis based on the total threat value to obtain algorithm dynamic control information.

[0025] As a further improvement of the present invention, the specific method for performing real-time environment determination and analysis based on the total threat value is as follows: Analyze the total threat value to obtain the three variables of the obstacle grid density, the total meteorological threat coefficient, and the dynamic threat coefficient corresponding to the current UAV path, and perform real-time monitoring on the three variables, and set threat red line values for the three variables respectively. When the obstacle grid density is greater than the corresponding threat red line value, generate a red line signal. Similarly, obtain the red line signals of the total meteorological threat coefficient and the dynamic threat coefficient;

[0026] Monitor the red line signal through the built-in dynamic control warning mechanism, and generate an algorithm dynamic control signal for algorithm restart when the red line signal appears.

[0027] Step 4: The UAV central control module mobilizes the improved shuffled frog leaping algorithm and adjusts the information path planning based on the dynamic weight; receives the algorithm control signal and performs emergency disposal execution on the improved shuffled frog leaping algorithm.

[0028] As a further improvement of the present invention, the emergency disposal execution includes the following steps: interruption trigger, status saving, population reconstruction, weight resetting, and rapid replanning;

[0029] Interruption trigger: When the dynamic control warning mechanism generates an algorithm restart signal, the UAV central control module sends an interruption signal to the improved shuffled frog leaping algorithm;

[0030] Status saving: After receiving the interruption signal, the improved shuffled frog leaping algorithm immediately terminates the current iteration cycle and stores the current global optimal path solution and the corresponding fitness function parameters;

[0031] Population reconstruction: Taking the current position of the UAV as a new starting point, generate a new initial population;

[0032] Weight resetting: Temporarily strengthen the weight corresponding to the red line signal to improve the avoidance priority of the corresponding threat;

[0033] Rapid replanning: Execute the improved shuffled frog leaping algorithm within the preset number of iterations, and output the emergency obstacle avoidance path

[0034] Step 5: The UAV central control module performs dynamic constraint verification on the UAV path to generate path feasibility verification information.

[0035] In the technical solution provided by the present invention, compared with the prior art, the beneficial effects are:

[0036] 1. By establishing a weight mapping matrix for the flight phase, the present invention dynamically adjusts the fitness function weight according to the core objectives of the takeoff, cruise, and landing phases, and combines environmental data such as obstacle grid density and meteorological threat coefficient for weight correction, enabling the algorithm to focus on the core constraints of the current phase and avoiding target conflicts caused by a globally unified weight.

[0037] 2. The present invention uses multi-sensor fusion to real-time sense environmental threats. When detecting abnormal states such as too high obstacle density and strong wind, it triggers the algorithm to stop and restart urgently, reconstructs the population with the current position, and strengthens the threat avoidance weight, avoiding the drawback that the existing mechanism needs to complete the current iteration cycle to update, and shortening the response time to the abnormal environment. Description of the Drawings

[0038] To more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings required for the description of the embodiments. The following drawings are not deliberately drawn to scale in actual size, and the focus is on showing the gist of the present application.

[0039] Figure 1 It is the flowchart of the method of the present invention;

[0040] Figure 2 It is the execution step diagram of the emergency response of the present invention. Detailed implementation manners

[0041] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying 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 the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0042] For ease of understanding, the following describes the specific process of the embodiments of the present invention. Please refer to Figure 1-2 In an embodiment of the present invention, an embodiment of a UAV path planning method based on an improved shuffled frog leaping algorithm includes:

[0043] Step 1: Multi-stage environment modeling: The UAV central control module retrieves the stage data set in the database and inputs the stage data set into the multi-stage environment analysis model to obtain flight stage information and environmental threat information. Based on multi-dimensional judgments of altitude, speed, and attitude, single-index misjudgment is avoided, and the flight stage is accurately distinguished, providing a real-time stage label for subsequent weight adjustment and search strategy adaptation.

[0044] The multi-stage environment analysis model, its specific execution steps are:

[0045] The stage data set includes environmental data and UAV state data; the pre-designed minimum safe flight altitude and constraint altitude are obtained, denoted as H and H C , and the flight altitude, speed data, and attitude information of the UAV are obtained according to the UAV state data;

[0046] Record the flight altitude of the drone as h. When 0 < h < H, the preliminary flight phases of the drone are the takeoff phase and the landing phase. Obtain the vertical velocity based on the velocity data of the drone. Analyze the takeoff and landing of the drone based on the vertical velocity of the drone and the change in the altitude of the drone. When the vertical velocity of the drone is positive and the flight altitude approaches H, and when the attitude angle of the drone obtained according to the attitude information of the drone is greater than the preset angle threshold, the corresponding flight phase generated is the takeoff phase; when the vertical velocity of the drone is negative and the flight altitude approaches 0, and when the attitude angle of the drone obtained according to the attitude information of the drone approaches 0, the corresponding flight phase generated is the landing phase;

[0047] When H < h < H C At this time, the preliminary flight phase of the drone is generated as the cruise phase. Obtain the vertical velocity and the flight velocity based on the velocity data of the drone. Generate a velocity stability signal when the vertical velocity is equal to zero or rises and falls at a preset stable velocity, and the horizontal velocity is within the preset stable velocity range; generate an attitude stability signal when the attitude angle obtained according to the attitude information of the drone is within the preset horizontal flight inclination range; when the velocity stability signal and the attitude stability signal of the drone are detected, the corresponding flight phase generated is the cruise phase; Obtain the environmental data corresponding to each flight phase and analyze the environmental data to obtain environmental threat information.

[0048] Furthermore, analyze the environmental data, and the specific analysis method is as follows:

[0049] Obtain the static environment and the dynamic environment based on the environmental data. Obtain the geometric parameters of each static obstacle and no-fly zone according to the static environment. Obtain the three-dimensional spatial modeling of each static obstacle and no-fly zone through the geometric parameters. Mark the three-dimensional spatial modeling of the static obstacles and no-fly zone as obstacle grids through the three-dimensional rasterization algorithm. Obtain the obstacle grids corresponding to each static obstacle based on the search step of the drone, and statistically obtain the total amount of obstacle grids. Calculate the ratio of the total amount of obstacle grids to the total amount of search grids within the search step range to obtain the obstacle grid density, and use the obstacle grid density as the corresponding static threat coefficient.

[0050] Obtain the meteorological conditions and dynamic obstacle information according to the dynamic environment; obtain the wind speed condition, precipitation condition, visibility condition, and temperature condition according to the meteorological conditions. Perform threat analysis on each meteorological condition according to the preset meteorological threat analysis unit. The meteorological threat analysis unit includes the safety meteorological threshold and the meteorological upper limit threshold preset for each meteorological condition, which are respectively recorded as Among them, q represents the meteorological condition; Divide the part of each meteorological condition that exceeds the corresponding safety meteorological threshold into multiple meteorological threat intervals Among them, n, n + 1, n + 2, and n + 3 are all constants; based on the multiple meteorological threat intervals corresponding to each meteorological adjustment, corresponding meteorological threat coefficients are set for each meteorological threat interval corresponding to each meteorological condition The threat analysis model corresponding to the meteorological condition is Each meteorological condition is respectively input into the corresponding threat analysis model to obtain the corresponding meteorological threat coefficient;

[0051] For example, for the precipitation intensity pi corresponding to the precipitation condition, it is input into the threat analysis model When the precipitation intensity is greater than the corresponding precipitation meteorological upper threshold, the corresponding precipitation meteorological threat coefficient output is 1.

[0052] Obtain the forward direction of the UAV, mark the area within the preset angle range of the forward direction of the UAV as the high-risk area, and set a high-risk coefficient. Mark the area outside the high-risk area as the normal area. According to the dynamic obstacle information, obtain the obstacle distance between each dynamic obstacle and the UAV. Obtain the risk warning boundary pre-designed by the UAV. When a dynamic obstacle enters the risk warning boundary, mark the corresponding dynamic obstacle as a threat obstacle. Based on the dynamic obstacle information, obtain the obstacle space volume and obstacle distance of each threat obstacle;

[0053] Perform normalization processing on the obstacle space volume and obstacle distance and take their values. Through Obtain the dynamic risk value; multiply the dynamic risk value of the threat obstacle corresponding to the high-risk area by the high-risk coefficient to obtain the dynamic threat coefficient; mark the dynamic risk value corresponding to the normal area as the dynamic threat coefficient.

[0054] Collect the meteorological threat coefficients and dynamic threat coefficients corresponding to the dynamic environment to obtain the total dynamic threat information, and collect the total dynamic threat information and the static threat coefficient corresponding to the static environment to obtain the environmental threat information.

[0055] Step 2: Dynamic fitness matching: The UAV central control module performs dynamic adaptation matching analysis on the flight phase information and environmental threat information to obtain dynamic weight adjustment information. Through the adaptation of the dynamic weight and the search step size, improve the planning efficiency of each phase. Prioritize avoiding near-field obstacles during the takeoff phase, and balance the distance and threat during the cruise phase to reduce ineffective searches

[0056] Perform dynamic adaptation matching analysis on the flight phase information and environmental threat information. The specific analysis method is as follows:

[0057] Based on the central control module, obtain multiple UAV paths for the currently preset analysis time period obtained by improving the shuffled frog leaping algorithm. Identify the corresponding path lengths for each UAV path. Obtain the static threat coefficient and the total dynamic threat information according to the environmental threat information. Obtain the basic weight vectors for each flight phase given in advance. The basic weight vectors include path length weight, static threat weight, and dynamic threat weight.

[0058] Based on the static threat coefficient and the total dynamic threat information, obtain the obstacle grid density, meteorological threat coefficient, and dynamic threat coefficient corresponding to each UAV path. Calculate and sum the meteorological threat coefficients corresponding to each meteorological condition to obtain the total meteorological threat coefficient. Use the multivariate Bessel function weighted sum model to process the obstacle grid density, the total meteorological threat coefficient, and the dynamic threat coefficient and output the threat total value. When the threat total value of the UAV path is greater than the preset threat threshold, mark the corresponding UAV path as a high-threat path.

[0059] Identify the flight phase of the UAV. When the flight phase information of the UAV corresponds to the takeoff phase and the landing phase, the corresponding dynamic weight adjustment information is to reduce the local search step size corresponding to the improved shuffled frog leaping algorithm and reduce the path length weight. Obtain the high-threat ratio of the high-threat paths accounting for the total number of UAV paths. When the high-threat ratio is greater than the preset upper limit ratio, increase the static threat weight and the dynamic threat weight on the basis of the basic weight vector.

[0060] When the flight phase information of the UAV corresponds to the cruise phase, the corresponding dynamic weight adjustment information is to expand the local search step size and increase the path length weight. Similarly, when the high-threat ratio is greater than the preset upper limit ratio, increase the static threat weight and the dynamic threat weight on the basis of the basic weight vector.

[0061] The specific processing process of the multivariate Bessel function weighted sum model is as follows:

[0062] Normalize the obstacle grid density, the total meteorological threat coefficient, and the dynamic threat coefficient and take their values, and denote them as sg, dw; Use the multivariate Bessel function weighted sum formula Calculate the threat total value wf corresponding to each UAV path k ; where, k represents the kth UAV path; β i represents the weight coefficients corresponding to the three variables of the obstacle grid density, the total meteorological threat coefficient, and the dynamic threat coefficient; sg max 、 dw max respectively represent the maximum values corresponding to the three variables; J v represents the first kind of Bessel function, and the order ν>0 ensures that at rk It is monotonically increasing within the value range. For example, when ν = 1, J1(r) first increases and then decreases when r > 0, but r can be restricted k within its monotonically increasing interval.

[0063] For example, let a certain path variable: sg = 80, the maximum value of obstacle density sg max = 100, The maximum value of the total meteorological threat coefficient dw = 70, the maximum value of the dynamic threat coefficient dw max = 100; After standardization: r k = 0.82 + 0.62 + 0.72 = 1.49 ≈ 1.22; Take β sg = 0.4, β dw = 0.3, ν = 1, look up the Bessel function J1(1.22) ≈ 0.4;

[0064] wf k = 0.4×0.491 + 0.3×0.491 + 0.3×0.491 = 0.491.

[0065] Step 3. Real-time environmental perception and determination: The UAV central control module extracts the total threat value corresponding to the travel path, and performs real-time environmental determination and analysis based on the total threat value to obtain algorithm dynamic control information.

[0066] The specific method for performing real-time environmental determination and analysis based on the total threat value is: Analyze the total threat value to obtain three variables corresponding to the current UAV path, namely obstacle grid density, total meteorological threat coefficient, and dynamic threat coefficient. Monitor the three variables in real time, and set threat red line values for the three variables respectively. When the obstacle grid density is greater than the corresponding threat red line value, a red line signal is generated. Similarly, red line signals for the total meteorological threat coefficient and the dynamic threat coefficient are obtained;

[0067] Monitor the red line signals through the built-in dynamic control warning mechanism. When a red line signal appears, generate an algorithm dynamic control signal for algorithm restart.

[0068] For example, set the red line value of obstacle grid density = 0.8, the red line value of total meteorological threat coefficient = 0.9, and the red line value of dynamic threat coefficient = 0.8. When the UAV enters the obstacle-dense area, the obstacle grid density = 0.85, triggering a red line signal. The dynamic control warning mechanism generates an algorithm restart signal, pre-warns of abnormal environments, and provides real-time signals for emergency handling

[0069] Step 4. Dynamic control of the improved shuffled frog leaping algorithm: The UAV central control module activates the improved shuffled frog leaping algorithm and adaptively adjusts the population initialization, subgroup division, and local search of the shuffled frog leaping algorithm based on the dynamic weight adjustment information; receives the algorithm control signal to perform emergency handling on the improved shuffled frog leaping algorithm.

[0070] The emergency handling execution includes the following steps:

[0071] Interrupt trigger: When the dynamic control warning mechanism generates an algorithm restart signal, the UAV central control module sends an interrupt signal to the improved shuffled frog leaping algorithm;

[0072] Status saving: After receiving the interrupt signal, the improved shuffled frog leaping algorithm immediately terminates the current iteration cycle and stores the current global optimal path solution and the corresponding fitness function parameters;

[0073] Population reconstruction: Taking the current position of the UAV as a new starting point, generate a new initial population;

[0074] Weight reset: Temporarily strengthen the weight corresponding to the red line signal to improve the avoidance priority of the corresponding threat;

[0075] Fast replanning: Execute the improved shuffled frog leaping algorithm within the preset number of iterations and output an emergency obstacle avoidance path.

[0076] Step 5. Path feasibility verification: The UAV central control module performs power constraint verification on the UAV path to generate path feasibility verification information.

[0077] As mentioned above, the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A method for path planning of an unmanned aerial vehicle based on an improved shuffled frog leaping algorithm, characterized in that, Including: The UAV central control module retrieves the stage data set in the database, and inputs the stage data set into the multi-stage environment analysis model to obtain flight stage information and environmental threat information; The UAV central control module performs dynamic adaptive matching analysis on the flight stage information and environmental threat information to obtain dynamic weight adjustment information; The UAV central control module extracts the total threat value corresponding to the travel path, and performs real-time environment determination analysis based on the total threat value to obtain algorithm dynamic control information; The UAV central control module mobilizes the improved shuffled frog leaping algorithm and performs path planning based on the dynamic weight adjustment information; receives the algorithm control signal and performs emergency disposal execution on the improved shuffled frog leaping algorithm; Perform power constraint verification on the UAV path to generate path feasibility verification information.

2. The method for path planning of an unmanned aerial vehicle based on an improved frog leaping algorithm according to claim 1, wherein, The specific execution steps of the multi-stage environment analysis model are as follows: The phase dataset includes environmental data and UAV status data; obtain the minimum safe flight altitude and the constraint altitude given by the pre-designed, denoted as H and H C , and obtain the flight altitude, speed data, and attitude information of the UAV according to the UAV status data; Record the flight height of the UAV as h. When 0 < h < H, the initial selected flight stages of the UAV are generated as the takeoff stage and the landing stage. Obtain the vertical speed according to the speed data of the UAV, and analyze the takeoff and landing of the UAV based on the vertical speed of the UAV and the height change of the UAV. When the vertical speed of the UAV is positive and the flight height approaches H, and according to the attitude information of the UAV, the attitude angle of the UAV is greater than the preset angle threshold, the corresponding flight stage is generated as the takeoff stage; when the vertical speed of the UAV is negative and the flight height approaches 0, and according to the attitude information of the UAV, the attitude angle of the UAV approaches 0, the corresponding flight stage is generated as the landing stage; When H < h < H C During this period, the initial selected flight phase of the drone is the cruise phase. The vertical speed and flight speed are obtained based on the speed data of the drone. A speed stability signal is generated when the vertical speed is equal to zero or the drone ascends and descends at a preset stable speed, and the horizontal speed is within the preset stable speed range. An attitude stability signal is generated when the attitude angle obtained based on the attitude information of the drone is within the preset horizontal flight inclination range. When the speed stability signal and the attitude stability signal of the drone are detected, the corresponding flight phase is generated as the cruise phase. The environmental data corresponding to each flight phase is obtained and analyzed to obtain environmental threat information.

3. The method for path planning of an unmanned aerial vehicle based on an improved shuffled frog leaping algorithm according to claim 2, wherein The specific analysis method for analyzing the environmental data is as follows: Obtain the static environment and dynamic environment according to the environmental data, obtain the geometric parameters of each static obstacle and no-fly zone according to the static environment, obtain the three-dimensional space modeling of each static obstacle and no-fly zone through the geometric parameters, mark the three-dimensional space modeling of the static obstacle and no-fly zone as obstacle grids through the three-dimensional rasterization algorithm, obtain the obstacle grids corresponding to each static obstacle based on the search step of the UAV, and count the total number of obstacle grids. Calculate the ratio of the total number of obstacle grids to the total number of search grids within the search step range to obtain the obstacle grid density, and use the obstacle grid density as the corresponding static threat coefficient; analyze the dynamic environment to obtain the total dynamic threat information, and combine the total dynamic threat information with the static threat coefficient corresponding to the static environment to obtain the environmental threat information.

4. The method for path planning of an unmanned aerial vehicle based on an improved frog leaping algorithm according to claim 3, wherein, The specific analysis method for analyzing according to the dynamic environment is as follows: Identify the meteorological conditions and dynamic obstacle information in the dynamic environment; obtain the wind speed condition, precipitation condition, visibility condition, and temperature condition according to the meteorological conditions, and perform threat analysis on each meteorological condition according to the preset meteorological threat analysis unit. The meteorological threat analysis unit includes the safety meteorological threshold and meteorological upper limit threshold preset for each meteorological condition, denoted as where q represents the meteorological condition; divide the part of each meteorological condition that exceeds the corresponding safety meteorological threshold into multiple meteorological threat intervals where n, n + 1, n + 2, n + 3 are all constants; based on the multiple meteorological threat intervals corresponding to each meteorological adjustment, set the corresponding meteorological threat coefficients for each meteorological threat interval corresponding to each meteorological condition The threat analysis model corresponding to the meteorological condition is Input each meteorological condition into the corresponding threat analysis model to obtain the corresponding meteorological threat coefficient; Obtain the forward direction of the UAV, mark the area within the preset angle range of the forward direction of the UAV as the high-risk area, and set a high-risk coefficient. Denote the area outside the high-risk area as the normal area. Obtain the obstacle distance between each dynamic obstacle and the UAV according to the dynamic obstacle information. Obtain the risk warning boundary pre-designed by the UAV. When the dynamic obstacle enters the risk warning boundary, mark the corresponding dynamic obstacle as a threat obstacle, and obtain the obstacle space volume and obstacle distance of each threat obstacle based on the dynamic obstacle information; Normalize the obstacle space volume and obstacle distance and take their numerical values. Through obtain the dynamic risk value; calculate the product of the dynamic risk value of the threat obstacle corresponding to the high-risk area and the high-risk coefficient to obtain the dynamic threat coefficient; mark the dynamic risk value corresponding to the normal area as the dynamic threat coefficient; combine the meteorological threat coefficients and the dynamic threat coefficient corresponding to the dynamic environment to obtain the total dynamic threat information.

5. A method for path planning of an unmanned aerial vehicle based on an improved frog leaping algorithm according to claim 1, characterized in that The specific analysis method for performing dynamic adaptive matching analysis on the flight stage information and environmental threat information is as follows: Based on the central control module, obtain multiple UAV paths for the currently preset analysis time period obtained by improving the shuffled frog leaping algorithm. Identify the corresponding path lengths for each UAV path. Obtain the static threat coefficient and the total dynamic threat information according to the environmental threat information. Obtain the basic weight vectors for each flight stage designed in advance. The basic weight vectors include path length weight, static threat weight, and dynamic threat weight; Based on the static threat coefficient and the total dynamic threat information, obtain the obstacle grid density, meteorological threat coefficient, and dynamic threat coefficient corresponding to each UAV path. Calculate and sum the meteorological threat coefficients corresponding to each meteorological condition to obtain the total meteorological threat coefficient. Use the multivariate Bessel function weighted sum model to process the obstacle grid density, the total meteorological threat coefficient, and the dynamic threat coefficient and output the total threat value. When the total threat value of the UAV path is greater than the preset threat threshold, mark the corresponding UAV path as a high-threat path; Identify the flight stage of the UAV. When the flight stage information of the UAV corresponds to the takeoff stage and the landing stage, the corresponding dynamic weight adjustment information is to reduce the local search step size corresponding to the improved shuffled frog leaping algorithm and reduce the path length weight. Obtain the high-threat ratio of the high-threat paths accounting for the total number of UAV paths. When the high-threat ratio is greater than the preset upper limit ratio, increase the static threat weight and the dynamic threat weight on the basis of the basic weight vector; When the flight stage information of the UAV corresponds to the cruise stage, the corresponding dynamic weight adjustment information is to expand the local search step size and increase the path length weight. Similarly, when the high-threat ratio is greater than the preset upper limit ratio, increase the static threat weight and the dynamic threat weight on the basis of the basic weight vector.

6. The method for path planning of an unmanned aerial vehicle based on an improved frog leaping algorithm according to claim 5, wherein, The specific processing process of the multivariate Bessel function weighted sum model is as follows: Normalize the obstacle grid density, the total meteorological threat coefficient, and the dynamic threat coefficient, and take their values, which are denoted as sg and dw respectively; use the weighted sum formula of the multivariate Bessel function to calculate the total threat value wf corresponding to each UAV path k ; where k represents the k-th UAV path; β i represents the weight coefficients corresponding to the three variables of the obstacle grid density, the total meteorological threat coefficient, and the dynamic threat coefficient; sg max , dw max represent the maximum values corresponding to the three variables respectively; J v represents the first kind of Bessel function, and the order ν > 0 ensures monotonic increase within the range of r k value.

7. A method for unmanned aerial vehicle path planning based on an improved frog leaping algorithm according to claim 1, characterized in that, The specific method for performing real-time environmental determination and analysis based on the total threat value is as follows: Analyze the total threat value to obtain three variables corresponding to the current UAV path, namely obstacle grid density, total meteorological threat coefficient, and dynamic threat coefficient. Monitor the three variables in real time, and set threat red line values for the three variables respectively. When the obstacle grid density is greater than the corresponding threat red line value, generate a red line signal. Similarly, obtain the red line signals for the total meteorological threat coefficient and the dynamic threat coefficient; Monitor the red line signals through the built-in dynamic control warning mechanism. When a red line signal appears, generate an algorithm dynamic control signal for algorithm restart.

8. A method for path planning of an unmanned aerial vehicle based on an improved frog leaping algorithm according to claim 1, characterized in that The emergency disposal execution includes the following steps: interruption trigger, status saving, population reconstruction, weight reset, and rapid replanning; For the interruption trigger, when the dynamic control warning mechanism generates an algorithm restart signal, the UAV central control module sends an interruption signal to the improved shuffled frog leaping algorithm; For the status saving, after the improved shuffled frog leaping algorithm receives the interruption signal, immediately terminate the current iteration cycle and store the current global optimal path solution and the corresponding fitness function parameters.

9. A method for path planning of an unmanned aerial vehicle based on an improved frog leaping algorithm according to claim 8, characterized in that, For the population reconstruction, take the current position of the UAV as a new starting point and generate a new initial population; For the weight reset, temporarily strengthen the weight corresponding to the red line signal to improve the avoidance priority of the corresponding threat; The fast replanning executes the improved shuffled frog leaping algorithm within the number of iterations given in the pre-design and outputs an emergency obstacle avoidance path.

Images