A demand-oriented UAV parameter self-setting method and system

Abstract

The present invention is applicable to the technical field of drone parameter setting, and in particular relates to a demand-oriented drone parameter self-setting method and system. The method comprises: obtaining the source and destination points of the drone flight, dividing them into segments, and establishing a correspondence between the segments and the requirements; initializing particles, randomly generating initial data for the drone flight, and embedding the segments and initial data into the particles; using a selected objective function, calculating the fitness of the particles in each segment, and sorting the fitness values. By calculating the fitness of the particles, the present invention can determine the quality of the flight path, thereby selecting the most appropriate flight path. By creating a decision tree, the target parameters can be quickly determined according to the requirements, greatly enhancing the adaptability of the requirements. The setting process does not require human intervention, optimizes resource utilization, and improves the efficiency of the drone's mission execution.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicle (UAV) parameter setting, and in particular to a demand-oriented UAV parameter self-setting method and system. Background Art

[0002] Drone requirements include the mission requirements of drones when performing different tasks, such as aerial photography, agricultural spraying, and search and rescue. Each mission may also have different requirements, such as range, accuracy, and stability. In addition, they also include the navigation requirements, communication requirements, and energy requirements of drones during the execution of a single mission. In the specific application scenarios of navigation requirements, users need to adjust various performance indicators in real time, such as flight time, payload capacity, flight altitude, and speed, to meet the flight requirements of the drone. This customized setting is closely related to the real-time location of the drone. This setting method will not only limit the scale of tasks, but also lead to increased labor costs.

[0003] Particle swarms are often used to search for potential paths in flight space due to their excellent global search capabilities, while decision trees can perform feature selection well and help determine the importance of features in solving specific problems.

[0004] Therefore, “how to combine particle swarm and decision tree algorithms to self-set UAV parameters based on demand” is the technical problem that the present invention needs to solve. Summary of the Invention

[0005] The purpose of the present invention is to provide a demand-oriented UAV parameter self-setting method and system to solve the problem of "how to combine particle swarm and decision tree algorithms to self-set UAV parameters based on demand" raised in the above background technology.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] A demand-oriented method for self-setting parameters of an unmanned aerial vehicle, the method comprising:

[0008] Obtain the source and destination points of the drone flight, divide them into segments, and establish a correspondence between the segments and the requirements;

[0009] Initialize particles, randomly generate initial data of the drone's flight, and embed the segments and initial data into the particles. Use the selected objective function to calculate the fitness value of the particles in each segment, sort the fitness values, and select a preset number of particles from the top of the sorted results as leaf nodes.

[0010] Iteratively crossing the leaf nodes, importing the iteration results into the created branch nodes, inserting attribute labels, determining target data in the branch nodes, integrating the leaf nodes and branch nodes, constructing a decision tree, and embedding a root node integrating a flight parameter calculation model into the decision tree;

[0011] The requirements are input into a decision tree, target data are obtained through traversal, the target data are converted into target parameters using the root node, and the target parameters are output from the decision tree.

[0012] Furthermore, the steps of obtaining the source and destination points of the UAV flight, dividing them into segments, and establishing a correspondence between the segments and the requirements include:

[0013] Collecting the segmented environmental data and superimposing the environmental data into the corresponding relationship, wherein the environmental data at least includes: weather data and obstacle data;

[0014] Based on a preset communication link, the environmental data is adjusted in real time.

[0015] Furthermore, the steps of initializing particles, randomly generating initial data of the drone flight, and embedding the segmented and initial data into particles include:

[0016] defining a particle structure and determining an initial state of the particle, wherein the initial state includes position, velocity, and properties;

[0017] A mapping relationship between the initial data and the segments is configured, and the particles are embedded based on the mapping relationship.

[0018] Furthermore, the steps of calculating the fitness values ​​of particles in each segment using the selected objective function, sorting the fitness values, and selecting a preset number of particles from the top of the sorted results as leaf nodes include:

[0019] Integrate path length cost, threat cost and force efficiency to generate fitness function;

[0020] The path length cost is calculated by using the Euclidean distance formula to calculate the sum of the Euclidean distances of the particles in the segments;

[0021] The Euclidean distance formula is:

[0022]

[0023] Where (x i ,y i ,z i ) is the coordinate of point i, (x i+1 ,y i+1 ,z i+1) is the coordinate of i+1, which is input in advance by the staff;

[0024] The threat cost is the sum of the threat costs of each point;

[0025]

[0026] Where K is the total number of threats, k is the k-th threat, and j(k) is the total cost from each point on the path to the k-th threat.

[0027] j(k)=(c1×r+c2×H)d×v;

[0028] c1 and c2 are adjustment parameters, r is the radius of the threat, H is the threat level of the threat, and the threat level is specified in advance by the staff; d is the distance to the threat, and v is the speed;

[0029] Use the preset force efficiency formula to calculate the force efficiency of the drone;

[0030] The force efficiency formula is:

[0031]

[0032] Where T is the flight time, m is the mass of the drone, q is the battery capacity of the drone, and U is the voltage. Try for loss;

[0033] The fitness function is determined as the objective function, particles are decoded, and leaf nodes are constructed using the data obtained after decoding.

[0034] Furthermore, the steps of iterating and crossing the leaf nodes, importing the iteration results into the created branch nodes, inserting attribute labels, determining target data in the branch nodes, integrating the leaf nodes and branch nodes, constructing a decision tree, and embedding a root node integrating a flight parameter calculation model into the decision tree include:

[0035] Performing single-point crossover on the particles in the leaf node to generate new particles, and updating the leaf node with the new particles;

[0036] In the leaf node, continue to iterate the new particle, and determine the target data after the iteration is completed;

[0037] Determine the goal and framework of the decision tree, classify the branch nodes into the framework, and hang the leaf nodes down to the branch nodes to construct the decision tree;

[0038] Utilize the attribute tags to build a query link between the segment and target data.

[0039] Furthermore, the step of inputting the requirements into a decision tree, traversing to obtain target data, converting the target data into target parameters using the root node, and outputting the target parameters from the decision tree includes:

[0040] Input the requirement into the branch node of the decision tree, cross the corresponding relationship and query link, and locate the target data corresponding to the requirement;

[0041] The flight parameter calculation model in the root node is used to convert the target data into target parameters, obtain the control authority of the UAV, and use the target parameters to set the UAV.

[0042] Furthermore, the method further comprises:

[0043] The design data of the UAV is obtained, and the risk level of the environmental data is determined. If the risk level is greater than a preset threshold, the target parameters are monitored using a preset monitoring mechanism, and based on the monitoring results, an alarm message is generated and sent to the control terminal of the UAV.

[0044] Furthermore, the system includes:

[0045] A segmentation module is used to obtain the source and destination points of the UAV flight, divide them into segments, and establish a corresponding relationship between the segments and the requirements;

[0046] A calculation module is used to initialize particles, randomly generate initial data of the UAV flight, embed the segments and initial data into the particles, calculate the fitness value of the particles in each segment using a selected objective function, sort the fitness values, and select a preset number of particles from the top of the sorted results as leaf nodes;

[0047] a construction module, configured to iteratively cross the leaf nodes, import the iteration results into the created branch nodes, insert attribute labels, determine target data in the branch nodes, integrate the leaf nodes and branch nodes, construct a decision tree, and embed a root node integrating a flight parameter calculation model into the decision tree;

[0048] The output module is used to input the requirements into the decision tree, traverse to obtain target data, use the root node to convert the target data into target parameters, and output the target parameters from the decision tree.

[0049] Furthermore, the segmentation module includes:

[0050] a superposition unit, configured to collect the segmented environmental data and superimpose the environmental data into the corresponding relationship, wherein the environmental data includes at least weather data and obstacle data;

[0051] The adjustment unit is used to adjust the environmental data in real time according to a preset communication link.

[0052] Furthermore, the calculation module includes:

[0053] A definition unit, used to define a particle structure and determine an initial state of the particle, wherein the initial state includes position, velocity, and attributes;

[0054] a mapping unit, configured to configure a mapping relationship between the initial data and the segments, and embed the particles based on the mapping relationship;

[0055] A generation unit is used to integrate path length cost, threat cost and force efficiency to generate a fitness function;

[0056] The decoding unit is used to determine the fitness function as the objective function, decode the particles, and construct leaf nodes using the data obtained after decoding.

[0057] Compared with the prior art, the present invention has the following beneficial effects:

[0058] 1. By obtaining segments, the needs can be subdivided, resource utilization can be better optimized, and the efficiency of task execution can be improved. By utilizing particles, the flight path can be optimized and the flight parameters can be adapted. By calculating the fitness value of the particles, the quality of the flight path can be judged, so as to select the most appropriate flight path. By creating a decision tree, the target parameters can be quickly determined according to the needs, which greatly enhances the adaptability of the needs. The setting process does not require human intervention, which improves the efficiency of the UAV's task execution.

[0059] 2. By determining environmental data, the data source of the UAV can be expanded, flight performance can be optimized, and flight safety can be ensured. By determining design data and risk levels, real-time monitoring of the UAV can be achieved, further improving the stability of the UAV flight. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention.

[0061] Figure 1 A flowchart of a demand-oriented UAV parameter self-setting method provided by an embodiment of the present invention;

[0062] Figure 2 A flowchart of the first sub-process of the demand-oriented UAV parameter self-setting method provided by an embodiment of the present invention;

[0063] Figure 3A second sub-flow chart of the demand-oriented UAV parameter self-setting method provided by an embodiment of the present invention;

[0064] Figure 4 A third sub-flow chart of the demand-oriented UAV parameter self-setting method provided by an embodiment of the present invention;

[0065] Figure 5 A fourth sub-flow chart of the demand-oriented UAV parameter self-setting method provided by an embodiment of the present invention;

[0066] Figure 6 A block diagram of the demand-oriented UAV parameter self-setting system provided by an embodiment of the present invention;

[0067] Figure 7 A block diagram of the components of the segmentation module in the demand-oriented UAV parameter self-setting system provided by an embodiment of the present invention;

[0068] Figure 8 A block diagram of the components of the computing module in the demand-oriented UAV parameter self-setting system provided by an embodiment of the present invention;

[0069] Figure 9 A block diagram of the components of the building blocks in the demand-oriented UAV parameter self-setting system provided by an embodiment of the present invention;

[0070] Figure 10 A block diagram of the output module in the demand-oriented drone parameter self-setting system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0071] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0072] In Example 1, Figure 1 The implementation process of the demand-oriented drone parameter self-setting method provided by an embodiment of the present invention is shown and described in detail below:

[0073] S100: Obtain the source point and destination point of the UAV flight, divide them into segments, and establish a corresponding relationship between the segments and requirements.

[0074] According to the UAV's mission planning, determine the source and destination points of the UAV's flight. The source point is the starting point of the UAV, and the destination point can be the end point of the UAV or the execution point of the mission. Divide an area according to the source and destination points, and segment the area. The number of segments is not limited, but it must be at least two. Establish a corresponding relationship between the segments and the needs.

[0075] For example, after determining segment A and segment B, the requirements for each segment are determined. The requirements may be to maintain a certain altitude in segment A in order to perform the flight mission, or other requirements. The requirements for each segment are derived from the flight mission.

[0076] S200: Initialize particles, randomly generate initial data of the drone's flight, and embed the segments and initial data into particles. Use the selected objective function to calculate the fitness value of particles in each segment, sort the fitness values, and select a preset number of particles from the top of the sorting results as leaf nodes.

[0077] Initialize the particles, where each particle represents a flight path, and randomly generate flight data. Use the flight data as initial data, then embed the segmentation and initial data into the particles, and use the particles to simulate the flight of the UAV. After the simulation is completed, use the objective function to calculate the fitness of each particle. The higher the fitness, the better the particle is and the closer it is to the correct solution. Use the fitness from high to low to sort the particles. From the sorting results, select a preset number of particles as leaf nodes. The preset number is not limited, and the leaf nodes are constructed using particles.

[0078] S300: Iterate and cross the leaf nodes, import the iteration results into the created branch nodes, insert attribute labels, determine target data in the branch nodes, integrate the leaf nodes and branch nodes, build a decision tree, and embed a root node that integrates a flight parameter calculation model into the decision tree.

[0079] The particles in the leaf nodes are iteratively crossed with each other. The advantage of this is that the best particles can be obtained, and branch nodes can be created at the same time. After the iterative crossover is completed, the data in the obtained particles are transferred to the branch nodes, and attribute labels are inserted into the branch nodes. The attribute labels include the segments and requirements corresponding to the particles, etc., and the target data is determined in the root node. The leaf nodes and branch nodes are integrated to build a decision tree. The root node is then inserted into the decision tree, and the flight parameter calculation model is integrated into the root node. It should be noted that there is more than one particle in the leaf node, and similarly, there is more than one particle in the branch node, so the objective function is also needed in the branch node to calculate the fitness of the particles in order to find the optimal solution.

[0080] S400: Input the demand into a decision tree, traverse to obtain target data, use the root node to convert the target data into target parameters, and output the target parameters from the decision tree.

[0081] After the decision tree is created, the requirements are input into the decision tree. By traversing the decision tree, the corresponding particles can be obtained. The particles are used to determine the flight path between segments. The data in the particles are then used to determine the target data of the drone. The target data is transferred to the root node, converted into target parameters, and output.

[0082] In Example 2, Figure 2 The implementation process of the demand-oriented drone parameter self-setting method provided by an embodiment of the present invention is shown and described in detail below:

[0083] S101: Collect the segmented environmental data and superimpose the environmental data into the corresponding relationship, wherein the environmental data at least includes weather data and obstacle data.

[0084] Collect environmental data between segments, where the environmental data can be weather data or obstacle data.

[0085] S102: Based on a preset communication link, the environmental data is adjusted in real time.

[0086] Build a communication link between the drone and the source of environmental data. The source of environmental data here can be a data supplier or other sources. Use the communication link to adjust the environmental data to ensure the timeliness of the environmental data.

[0087] In Example 3, Figure 3 The implementation process of the demand-oriented drone parameter self-setting method provided by an embodiment of the present invention is shown. The steps of initializing particles, randomly generating initial data of drone flight, and embedding the segmentation and initial data into particles are described in detail below:

[0088] S201: Define the particle structure and determine the initial state of the particle, where the initial state includes position, velocity, and attributes.

[0089] Determine the particle structure and the initial state of the particle, where the initial state includes the specific position, initial velocity and other relevant properties of the particle.

[0090] S202: configuring a mapping relationship between the initial data and segments, and embedding the particles based on the mapping relationship.

[0091] The correspondence between the initial data and the segments is determined and the correspondence is embedded into the particles.

[0092] In Example 4, Figure 3The implementation process of the demand-oriented drone parameter self-setting method provided by an embodiment of the present invention is shown. The following details the steps of using the selected objective function to calculate the fitness value of particles in each segment, sorting the fitness values, and selecting a preset number of particles from the top of the sorted results as leaf nodes.

[0093] S203: Integrate path length cost, threat cost and force efficiency to generate fitness function;

[0094] The path length cost is calculated by using the Euclidean distance formula to calculate the sum of the Euclidean distances of the particles in the segments;

[0095] The Euclidean distance formula is:

[0096]

[0097] Where (x i ,y i ,z i ) is the coordinate of point i, (x i+1 ,y i+1 ,z i+1 ) is the coordinate of i+1, which is input in advance by the staff;

[0098] The threat cost is the sum of the threat costs of each point;

[0099]

[0100] Where K is the total number of threats, k is the k-th threat, and j(k) is the total cost from each point on the path to the k-th threat.

[0101] j(k)=(c1×r+c2×H)d×v;

[0102] c1 and c2 are adjustment parameters, r is the radius of the threat, H is the threat level of the threat, and the threat level is specified in advance by the staff; d is the distance to the threat, and v is the speed;

[0103] Use the preset force efficiency formula to calculate the force efficiency of the drone;

[0104] The force efficiency formula is:

[0105]

[0106] Where T is the flight time, m is the mass of the drone, q is the battery capacity of the drone, and U is the voltage. Try for the loss.

[0107] The fitness function is generated using the path length cost, threat cost and force efficiency. The Euclidean distance formula is used to calculate the path length cost, and the threat cost formula is used to calculate the threat cost on the path length. Finally, the force efficiency of the drone is calculated. The path length cost, threat cost and force efficiency are integrated to determine the fitness function of the drone. The purpose of using the fitness function is to find the optimal solution for particles in the segment.

[0108] S204: Determine the fitness function as the objective function, decode particles, and construct leaf nodes using the data obtained after decoding.

[0109] The fitness function is used as the objective function, the fitness value of the decoded particle is calculated using the objective function, and the leaf node is constructed using the fitness value.

[0110] In Example 5, Figure 4 The implementation process of the demand-oriented drone parameter self-setting method provided by an embodiment of the present invention is shown. The following details the steps of iteratively crossing the leaf nodes, importing the iteration results into the created branch nodes, inserting attribute labels, determining target data in the branch nodes, integrating the leaf nodes and branch nodes, constructing a decision tree, and embedding a root node integrating a flight parameter calculation model into the decision tree.

[0111] S301: Performing single-point crossover on particles in the leaf node to generate new particles, and using the new particles to update the leaf node.

[0112] New particles are obtained by performing single-point crossover on particles, where the number of new particles is at least one, and the new particles are used to cover the particles in the leaf nodes.

[0113] S302: In the leaf node, continue to iterate the new particle, and determine target data after the iteration is completed.

[0114] Continue to iterate the new particles in the leaf node, and set the termination condition in the leaf node. When the iteration result meets the termination condition, the iteration stops. In each iteration process, new particles will be generated and the data will be updated. The data in the particles after the iteration is completed will be determined as the target data.

[0115] S303: Determine the goal and framework of the decision tree, classify the branch nodes into the framework, and hang the leaf nodes to the branch nodes to construct the decision tree.

[0116] According to the goal, determine the framework of the decision tree, integrate the branch nodes into the framework, and then hang the leaf nodes down to the branch nodes to complete the construction of the decision tree.

[0117] S304: Using the attribute tags, a query link between the segment and the target data is established.

[0118] The attribute tags include the segments and requirements corresponding to the particles, and query links are built at the same time. In practical applications, the requirements and query links can be used to quickly locate the branch nodes, and the data in the branch nodes can be used to control the flight of the drone.

[0119] In Example 5, Figure 5 The implementation process of the demand-oriented drone parameter self-setting method provided by an embodiment of the present invention is shown. The following details the steps of inputting the demand into the decision tree, traversing to obtain the corresponding target data, converting the target data into target parameters using the root node, and outputting the target parameters from the decision tree.

[0120] S401: Input the demand into a branch node in a decision tree, cross the corresponding relationship and query link, and locate the target data corresponding to the demand.

[0121] The requirements are input into the decision tree, and the target data is determined by utilizing the corresponding relationship and query link. The target data is the optimal flight action of the drone. By flying according to this flight action, the safety and stability of the flight are optimized, and the energy consumption is minimized.

[0122] S402: Using the flight parameter calculation model in the root node, convert the target data into target parameters, obtain the control authority of the UAV, and use the target parameters to set the UAV.

[0123] A flight parameter calculation model is embedded in the root node, and a data set is selected from the existing database to train the flight parameter calculation model. The flight parameter calculation model is used to convert target data into target parameters, and the target parameters are used to set the parameters of the UAV.

[0124] In Example 6, different from Example 1, in this embodiment of the present invention, the method further includes:

[0125] The design data of the UAV is obtained, and the risk level of the environmental data is determined. If the risk level is greater than a preset threshold, the target parameters are monitored using a preset monitoring mechanism, and based on the monitoring results, an alarm message is generated and sent to the control terminal of the UAV.

[0126] Obtain the design data of the drone from the drone manufacturer, and determine the risk level of the drone flight based on the environmental data. If the risk level is greater than the preset threshold, it proves that the flight environment is relatively harsh and it is necessary to pay special attention to the flight parameters. It should be noted that the preset threshold is determined based on the design data; use the preset monitoring mechanism to monitor the target parameters. The preset monitoring mechanism can be to compare the design data and the target parameters. If the target parameters are greater than the design data, it means that the drone needs to fly beyond its own capabilities. At this time, an alarm message needs to be generated and sent to the drone's control terminal. If the target parameters are not greater than the design data, continue monitoring.

[0127] Figure 6 The following is a structural block diagram of a demand-oriented UAV parameter self-setting system provided by an embodiment of the present invention. The demand-oriented UAV parameter self-setting system 1 includes:

[0128] The segmentation module 11 is used to obtain the source and destination points of the UAV flight, divide them into segments, and establish a corresponding relationship between the segments and the requirements;

[0129] The calculation module 12 is used to initialize particles, randomly generate initial data of the UAV flight, embed the segmentation and initial data into the particles, calculate the fitness value of the particles in each segment using the selected objective function, sort the fitness values, and select a preset number of particles from the top of the sorted results as leaf nodes;

[0130] A construction module 13 is configured to iteratively intersect the leaf nodes, import the iteration results into the created branch nodes, insert attribute labels, determine target data in the branch nodes, integrate the leaf nodes and branch nodes, construct a decision tree, and embed a root node integrating a flight parameter calculation model into the decision tree;

[0131] The output module 14 is used to input the requirements into the decision tree, traverse to obtain corresponding target data, convert the target data into target parameters using the root node, and output the target parameters from the decision tree.

[0132] Figure 7 The following is a structural block diagram of a demand-oriented UAV parameter self-setting system provided by an embodiment of the present invention. The segmentation module 11 includes:

[0133] A superposition unit 111 is configured to collect the segmented environmental data and superimpose the environmental data onto the corresponding relationship, wherein the environmental data includes at least weather data and obstacle data;

[0134] The adjustment unit 112 is configured to adjust the environmental data in real time according to a preset communication link.

[0135] Figure 8 The following is a structural block diagram of a demand-oriented UAV parameter self-setting system provided by an embodiment of the present invention. The calculation module 12 includes:

[0136] A definition unit 121 is used to define a particle structure and determine an initial state of the particle, wherein the initial state includes position, velocity, and attributes;

[0137] A mapping unit 122 is configured to configure a mapping relationship between the initial data and the segments, and embed the particles based on the mapping relationship;

[0138] A generating unit 123 is used to integrate the path length cost, the threat cost and the force efficiency to generate a fitness function;

[0139] The decoding unit 124 is configured to determine the fitness function as an objective function, decode particles, and construct leaf nodes using the data obtained after decoding.

[0140] Figure 9 The following is a structural block diagram of a demand-oriented UAV parameter self-setting system provided by an embodiment of the present invention. The building block 13 includes:

[0141] A crossover unit 131 is configured to perform single-point crossover on particles in the leaf node to generate new particles, and update the leaf node using the new particles;

[0142] An iterative unit 132 is configured to continue iterating the new particle in the leaf node and determine target data after the iteration is completed;

[0143] The inclusion unit 133 is used to determine the goal and framework of the decision tree, include the branch nodes in the framework, and attach the leaf nodes to the branch nodes to construct the decision tree;

[0144] The building unit 134 is configured to build a query link between the segment and target data using the attribute tag.

[0145] Figure 10 The following is a structural block diagram of a demand-oriented UAV parameter self-setting system provided by an embodiment of the present invention. The output module 14 includes:

[0146] The spanning unit 141 is used to input the demand into the branch node in the decision tree, span the corresponding relationship and query link, and locate the target data corresponding to the demand;

[0147] The setting unit 142 is used to use the flight parameter calculation model in the root node to convert the target data into target parameters, obtain the control authority of the drone, and use the target parameters to set the drone.

[0148] The segmentation module 11 is mainly used to complete step S100, the calculation module 12 is mainly used to complete step S200, the construction module 13 is mainly used to complete step S300, and the output module 14 is mainly used to complete step S400;

[0149] The superposition unit 111 is mainly used to complete step S101, and the adjustment unit 112 is mainly used to complete step S102;

[0150] The definition unit 121 is mainly used to complete step S201, the mapping unit 122 is mainly used to complete step S202, the generation unit 123 is mainly used to complete step S203, and the decoding unit 124 is mainly used to complete step S204;

[0151] The cross unit 131 is mainly used to complete step S301, the iteration unit 132 is mainly used to complete step S302, the classification unit 133 is mainly used to complete step S303, and the construction unit 134 is mainly used to complete step S304;

[0152] The spanning unit 141 is mainly used to complete step S401, and the setting unit 142 is mainly used to complete step S402.

[0153] The technical features of the above-mentioned embodiments can be combined arbitrarily. In order to make the description concise, not all possible combinations of the technical features in the above-mentioned embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0154] The above-described embodiments merely illustrate several implementations of the present invention, and while their descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that a person skilled in the art would be able to make numerous variations and improvements without departing from the spirit of the present invention, all of which fall within the scope of protection of the present invention. Therefore, the scope of protection of the present invention shall be determined by the appended claims.

[0155] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A demand-oriented UAV parameter self-setting method, characterized in that: The method comprises: Obtain the source and destination points of the drone flight, divide them into segments, and establish a correspondence between the segments and the requirements; Initialize particles, randomly generate initial data of the drone's flight, and embed the segments and initial data into the particles. Use the selected objective function to calculate the fitness value of the particles in each segment, sort the fitness values, and select a preset number of particles from the top of the sorted results as leaf nodes. Iteratively crossing the leaf nodes, importing the iteration results into the created branch nodes, inserting attribute labels, determining target data in the branch nodes, integrating the leaf nodes and branch nodes, constructing a decision tree, and embedding a root node integrating a flight parameter calculation model into the decision tree; The requirements are input into the decision tree, target data is obtained by traversing the tree, the target data is converted into target parameters using the root node, and the target parameters are output from the decision tree.

2. The demand-oriented UAV parameter self-setting method according to claim 1, characterized in that: The steps of obtaining the source and destination points of the UAV flight, dividing the points into segments, and establishing a corresponding relationship between the segments and the requirements include: Collecting the segmented environmental data and superimposing the environmental data into the corresponding relationship, wherein the environmental data at least includes: weather data and obstacle data; Based on a preset communication link, the environmental data is adjusted in real time.

3. The demand-oriented UAV parameter self-setting method according to claim 1, characterized in that: The steps of initializing particles, randomly generating initial data of the drone flight, and embedding the segmentation and initial data into particles include: defining a particle structure and determining an initial state of the particle, wherein the initial state includes position, velocity, and properties; A mapping relationship between the initial data and the segments is configured, and the particles are embedded based on the mapping relationship.

4. The demand-oriented UAV parameter self-setting method according to claim 3, characterized in that: The steps of calculating the fitness values ​​of particles in each segment using the selected objective function, sorting the fitness values, and selecting a preset number of particles from the top of the sorted results as leaf nodes include: Integrate path length cost, threat cost and force efficiency to generate fitness function; The path length cost is calculated by using the Euclidean distance formula to calculate the sum of the Euclidean distances of the particles in the segments; The Euclidean distance formula is: Where (x i ,y i ,z i ) is the coordinate of point i, (x i+1 ,y i+1 ,z i+1 ) is the coordinate of i+1, which is pre-entered by the staff; The threat cost is the sum of the threat costs of each point; Where K is the total number of threats, k is the kth threat, and j(k) is the total cost from each point on the path to the kth threat; j(k)=(c1×r+c2×H)d×v; c1 and c2 are adjustment parameters, r is the radius of the threat, H is the threat level of the threat, which is pre-specified by the staff, d is the distance to the threat, and v is the speed; Use the preset force efficiency formula to calculate the force efficiency of the drone; The force efficiency formula is: Where T is the flight time, m is the mass of the drone, q is the battery capacity of the drone, and U is the voltage. Try for loss; The fitness function is determined as the objective function, particles are decoded, and leaf nodes are constructed using the data obtained after decoding.

5. The demand-oriented UAV parameter self-setting method according to claim 1, characterized in that: The steps of iterating and crossing the leaf nodes, importing the iteration results into the created branch nodes, inserting attribute labels, determining target data in the branch nodes, integrating the leaf nodes and branch nodes, constructing a decision tree, and embedding a root node integrated with a flight parameter calculation model into the decision tree include: Performing single-point crossover on the particles in the leaf node to generate new particles, and updating the leaf node with the new particles; In the leaf node, continue to iterate the new particle, and determine the target data after the iteration is completed; Determine the goal and framework of the decision tree, classify the branch nodes into the framework, and hang the leaf nodes down to the branch nodes to construct the decision tree; Utilize the attribute tags to build a query link between the segment and target data.

6. The demand-oriented UAV parameter self-setting method according to claim 5, characterized in that: The step of inputting the demand into the decision tree, traversing to obtain target data, converting the target data into target parameters using the root node, and outputting the target parameters from the decision tree includes: Input the requirements into the branch nodes in the decision tree, cross the corresponding relationships and query links, and locate the target data; The flight parameter calculation model in the root node is used to convert the target data into target parameters, obtain the control authority of the UAV, and use the target parameters to set the UAV.

7. The demand-oriented UAV parameter self-setting method according to claim 1, characterized in that: The method further comprises: The design data of the UAV is obtained, and the risk level of the environmental data is determined. If the risk level is greater than a preset threshold, the target parameters are monitored using a preset monitoring mechanism, and based on the monitoring results, an alarm message is generated and sent to the control terminal of the UAV.

8. A demand-oriented UAV parameter self-setting system, characterized in that: The system comprises: A segmentation module is used to obtain the source and destination points of the UAV flight, divide them into segments, and establish a corresponding relationship between the segments and the requirements; A calculation module is used to initialize particles, randomly generate initial data of the UAV flight, embed the segments and initial data into the particles, calculate the fitness value of the particles in each segment using a selected objective function, sort the fitness values, and select a preset number of particles from the top of the sorted results as leaf nodes; a construction module, configured to iteratively cross the leaf nodes, import the iteration results into the created branch nodes, insert attribute labels, determine target data in the branch nodes, integrate the leaf nodes and branch nodes, construct a decision tree, and embed a root node integrating a flight parameter calculation model into the decision tree; The output module is used to input the requirements into the decision tree, traverse to obtain target data, use the root node to convert the target data into target parameters, and output the target parameters from the decision tree.

9. The demand-oriented UAV parameter self-setting system according to claim 8, characterized in that: The segmentation module includes: a superposition unit, configured to collect the segmented environmental data and superimpose the environmental data into the corresponding relationship, wherein the environmental data includes at least weather data and obstacle data; The adjustment unit is used to adjust the environmental data in real time according to a preset communication link.

10. The demand-oriented UAV parameter self-setting system according to claim 8, characterized in that: The calculation module includes: A definition unit, used to define a particle structure and determine an initial state of the particle, wherein the initial state includes position, velocity, and attributes; a mapping unit, configured to configure a mapping relationship between the initial data and the segments, and embed the particles based on the mapping relationship; A generation unit is used to integrate path length cost, threat cost and force efficiency to generate a fitness function; The decoding unit is used to determine the fitness function as the objective function, decode the particles, and construct leaf nodes using the data obtained after decoding.

Images