Impact load monitoring method and system for unmanned aerial vehicle collision avoidance

Abstract

The invention relates to the technical field of unmanned aerial vehicle impact load monitoring, in particular to an impact load monitoring method and system for unmanned aerial vehicle collision avoidance, and the method comprises the steps: obtaining the impact load of an unmanned aerial vehicle in each group of experiments and the data of three flight elements of the unmanned aerial vehicle through an intelligent sensor; calculating the gradient influence degree and the impact load synchronization consistency of each flight element; analyzing the deviation condition of the correlation vector of each flight element and the correlation total vector to obtain the contribution degree of each flight element to the impact load so as to calculate the self-adaptive fitting weight of each flight element; and in combination with the data of each flight element and the unmanned aerial vehicle impact load data, fitting to obtain a fitting function between the unmanned aerial vehicle impact load and the three flight elements so as to obtain the maximum impact load, and obtaining the optimal value of the unmanned aerial vehicle airbag parameter by using an optimization algorithm. The impact load of the unmanned aerial vehicle can be effectively reduced, and the safety and effectiveness of the unmanned aerial vehicle during collision are guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicle impact load monitoring, and in particular to an unmanned aerial vehicle impact load monitoring method and system for collision avoidance. Background Art

[0002] In recent years, with the widespread application of drones in various industries, due to the limitations of drone performance and the professionalism of operators, especially when drones are conducting inspections, facing the influence of complex external environments, the active obstacle avoidance function of the drone system is limited or fails, and drones often collide with structures passively, causing damage to the drone and making it unusable. From the perspective of passive collision avoidance, equipping drones with safety protection devices, such as protective covers and drone airbags, has become a focus of attention in the field of passive collision avoidance of drones.

[0003] In the process of developing UAV anti-collision airbags, it is necessary to determine the optimal design parameters of the airbag based on the actual impact load of the UAV. Therefore, the impact load detection for UAV anti-collision is crucial. Since the state of the UAV when it fails in the air is complex and changeable, the measurement of the impact load is difficult. In traditional measurements, the impact load is mainly related to the flight altitude, ambient wind speed and failure flight speed of the UAV. The coupling relationship between the impact load and the three flight elements (flight altitude, ambient wind speed and failure flight speed) is obtained by multiple simulation measurements and polynomial fitting. However, in the actual environment, the coupling relationship between the three flight elements and the impact load is nonlinear, and the degree of influence of the three flight elements on the impact load is different. Therefore, in the traditional method, each flight element is given an equal fitting weight, resulting in a deviation between the fitted impact load and the actual impact load, which affects the effectiveness of the airbag design. Summary of the invention

[0004] In order to solve the technical problem that the fixed threshold value in the above-mentioned revolving door algorithm affects the accuracy of monitoring aquaculture nutrition and animal health data, the purpose of the present invention is to provide an impact load monitoring method and system for UAV collision avoidance. The technical solution adopted by the method is as follows:

[0005] The impact load of the UAV in each group of experiments and the data of the three flight elements of the UAV are obtained through intelligent sensors;

[0006] Based on the data differences of each flight element in all experiments and the gradual change of the impact load of the UAV, the gradual change influence of each flight element is determined. Combined with the difference between the data trend distribution of each flight element and the trend distribution of the impact load of the UAV, the synchronous consistency of the impact load of each flight element is obtained.

[0007] According to the data of each flight element and the impact load data of each UAV, the correlation vector between each flight element and the impact load and the total correlation vector between the three flight elements and the impact load are obtained, the deviation between the correlation vector of each flight element and the total correlation vector is analyzed, and the contribution of each flight element to the impact load is obtained; the adaptive fitting weight of each flight element is calculated in combination with the contribution of the three flight elements to the impact load;

[0008] Based on the adaptive fitting weights of each flight element and combined with the data of each flight element and the impact load data of the UAV, the fitting function between the impact load of the UAV and the three flight elements is fitted to obtain the maximum impact load, and the optimal value of the UAV airbag parameters is obtained using the optimization algorithm.

[0009] Preferably, the flight elements include flight altitude, ambient wind speed and failure flight speed.

[0010] Preferably, the calculation formula for the progressive influence of each flight element is:

[0011] In the formula, B represents the influence of the current flight factor, N represents the number of data, ΔF i Indicates the absolute value of the difference between the i-th and i+1-th data values ​​after the data of the current flight element is sorted, ΔQ i It represents the absolute value of the difference between the i-th and i+1-th impact loads after the UAV impact loads are sorted, and α is a constant to avoid the denominator being zero. The sorting is to arrange the data values ​​of the current flight elements and the UAV impact loads in all experiments in ascending order.

[0012] Preferably, the calculation formula for the synchronous consistency of the impact loads of the flight elements is: Where, C represents the synchronous consistency of the impact load of the current flight elements, ρ x and ρ Q They represent the trend strength of the current flight elements and the impact load of the UAV respectively, where the trend strength is calculated by using the time series decomposition algorithm after sorting.

[0013] Preferably, the process of obtaining the correlation vector between each flight element and the impact load and the total correlation vector between the three flight elements and the impact load is:

[0014] Arrange the data of each flight element in ascending order to obtain the data sequence of each flight element, and arrange the impact load of the UAV in ascending order to form the impact load sequence of the UAV;

[0015] Each correlation vector between each flight element and the impact load is composed of the elements at the same position in the data sequence of each flight element and the impact load sequence of the UAV; each total correlation vector between the three flight elements and the impact load is composed of the elements at the same position in the data sequence of all flight elements and the impact load sequence of the UAV.

[0016] Preferably, the calculation formula for the contribution of each flight element to the impact load is:

[0017] In the formula, D represents the contribution of the current flight element to the impact load, C represents the synchronization consistency of the impact load of the current flight element, N represents the number of data, θ(y i ,y i+1 ) represents the angle between the correlation vectors of the i-th and i+1-th current flight elements and the impact load, θ(z i ,z i+1 ) represents the angle between the total vector associated with the i-th and i+1-th three flight elements and the impact load, and α is a constant to avoid the denominator being zero.

[0018] Preferably, the calculation formula for the adaptive fitting weight of each flight element is: In the formula, w h represents the adaptive fitting weight of the flight altitude, D h , D v and D f They respectively represent the contribution of flight altitude to impact load, the contribution of ambient wind speed to impact load and the contribution of time-dependent flight speed to impact load.

[0019] Preferably, obtaining the fitting function between the impact load of the UAV and the three flight elements includes:

[0020] The data of the impact loads of each UAV and each flight element in all groups of experiments are combined with the adaptive fitting weights of each flight element, and a polynomial fitting algorithm is used to fit the fitting function between the UAV impact load and the three flight elements, where the independent variables in the fitting function are the three flight elements and the dependent variable is the UAV impact load.

[0021] Preferably, obtaining the optimal value of the drone airbag parameter includes:

[0022] According to the relationship between the maximum impact load and the design parameters of the UAV airbag, the response fitting function of the maximum impact load of the UAV is obtained, and the optimal values ​​of each airbag parameter of the UAV are obtained using the genetic optimization algorithm.

[0023] On the other hand, an embodiment of the present application provides an impact load monitoring system for unmanned aerial vehicle collision avoidance, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor implements the steps of any one of the above methods when executing the computer program.

[0024] The present invention has the following beneficial effects:

[0025] In the present invention, the experimental data of the flight parameters and the impact load of the unmanned aerial vehicle in the experimental group are obtained by using an intelligent sensor through a uniform experiment. The synchronization consistency of the current flight element is obtained based on the variation and distribution trend of the single flight element and the impact load, and the influence of the current flight element on the impact load is measured. Based on the deviation of the correlation vector of the single flight element and the correlation total vector of the three flight elements, the contribution of the single flight element is obtained, which reflects the contribution of the current flight element to the impact load under the joint action of the three flight elements, thereby obtaining the fitting weight of the flight element. The data and the fitting weight are brought into the fitting algorithm, and finally the fitting relationship between the flight parameters and the impact load is obtained, and the impact load of the unmanned aerial vehicle that is close to accuracy is obtained. Based on the impact load of the unmanned aerial vehicle, the airbag parameters in the anti-collision airbag for reducing the load of the unmanned aerial vehicle are obtained by using finite element analysis and genetic algorithm. Since the fitting weight of the single flight element is obtained by comprehensively considering the influence degree of the single flight element and the correlation of multiple flight elements in the present invention, the impact load of the unmanned aerial vehicle obtained in the end is closer to the actual impact load, so that the airbag parameters finally designed can effectively reduce the impact load of the unmanned aerial vehicle and ensure the safety and effectiveness of the unmanned aerial vehicle during collision. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0027] Figure 1 A flow chart of an impact load monitoring method for UAV collision avoidance provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0028] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the impact load monitoring method and system for UAV collision avoidance proposed by the present invention, in combination with the accompanying drawings and preferred embodiments, and its specific implementation, structure, features and effects are described as follows. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0029] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0030] The specific scheme of the impact load monitoring method for UAV collision avoidance provided by the present invention is described in detail below with reference to the accompanying drawings.

[0031] See also Figure 1 , which shows an impact load monitoring method and system flow chart for UAV collision avoidance provided by an embodiment of the present invention, the method comprising the following steps:

[0032] Step 100: Obtain the impact load of the UAV in each group of experiments and the data of the three flight elements of the UAV through the intelligent sensor.

[0033] In order to better design the anti-collision airbag parameters of the drone, it is necessary to strictly measure the impact load between the drone and the ground when it fails. In this embodiment, the range information of the drone flight parameters is obtained through the drone's flight manual. The flight altitude refers to the altitude range in which the drone can fly, the ambient wind speed refers to the wind speed range that the drone can tolerate, and the flight speed when it fails refers to the flight speed range of the drone. In this embodiment, the drone flight elements include the flight altitude, the ambient wind speed, and the failure flight speed. In this embodiment, the flight altitude h is selected to be in the range of 0 to 200m, the ambient wind speed v is in the range of 0 to 12m / s, and the failure flight speed f is in the range of 0 to 20m / s.

[0034] In order to reduce the cost of the experiment, this embodiment adopts a uniform experiment to reduce the number of experiments, and uses a model of the same size and mass as the drone during the test, and uses smart sensors to obtain the flight parameters of the drone when it fails, where the smart sensors include an altitude sensor, a wind speed sensor, a speed sensor, and a collision impact sensor. In this embodiment, the uniform experiment group is set to 200 groups, and each group of experimental information includes the impact load of the drone and the data of each flight element.

[0035] In order to reduce the influence of different parameter information sizes and units, the 200 sets of experimental data obtained from the experiment are normalized to obtain normalized parameter information. It should be noted that there are many methods for normalization operations, and implementers can choose them in actual application scenarios. In this embodiment, the Z-Score normalization method is used to normalize the data of each parameter. The specific normalization process is a prior art and will not be repeated in this embodiment.

[0036] Step 200: Based on the data differences of each flight element in all experiments and the gradual changes of the impact load of the UAV, the gradual change influence of each flight element is determined, and the synchronous consistency of the impact load of each flight element is obtained by combining the difference between the data trend distribution of each flight element and the trend distribution of the impact load of the UAV.

[0037] When a drone fails in the air, the impact load between the drone and the ground is closely related to the flight information of the drone, that is, the impact load is related to the flight altitude, environmental wind speed and failure flight speed. The flight altitude determines the normal falling time of the drone after failure, which is often positively correlated with the impact load; the environmental wind speed affects the air resistance of the drone when it falls, but it will affect the falling posture of the drone and indirectly affect the size of the impact load; the failure flight speed determines the initial kinetic energy of the drone. The greater the initial kinetic energy, the greater the impact load at the time of impact. The three flight elements work together to affect the impact load of the drone, and the relationship between the three is complex, and the degree of influence on the impact load is different. Therefore, further analysis is required based on the monitoring data obtained by the intelligent sensor of the experimental group.

[0038] In this embodiment, the flight altitude h, ambient wind speed v and failure flight speed f of the three flight elements of the UAV are used as independent variables. The three flight elements neutralize the impact load acting on the UAV, and the impact load is used as the dependent variable. In order to analyze the influence of the three flight elements on the impact load, any of the flight elements is taken as an example for detailed description to analyze the synchronization consistency of the impact load of each flight element. The specific process is as follows:

[0039] A single flight element is a continuous space within the value range. Through uniform experiments, 200 sets of experimental data are sampled. Therefore, when the value of a single flight element changes, if the change of the flight element is larger and the corresponding change of the impact load is also larger, it indicates that the flight element has a greater impact on the impact load.

[0040] Therefore, for the current flight element, the values ​​in the 200 sets of experimental data of the current flight element are arranged in ascending order. Based on the above analysis, the variation influence degree B is obtained by measuring the variation relationship between the flight element and the impact load. In this embodiment, the variation influence degree of each flight element is determined based on the data difference of each flight element in all experiments and the variation of the impact load of the drone. The specific calculation formula is:

[0041] In the formula, B represents the influence of the current flight factor, N represents the number of data, ΔF i Indicates the absolute value of the difference between the i-th and i+1-th data values ​​after the data of the current flight element is sorted, ΔQ i It represents the absolute value of the difference between the i-th and i+1-th impact loads after the impact loads of the UAV are sorted, wherein the sorting is to arrange the data values ​​of the current flight element and the impact load of the UAV in all experiments in ascending order, α is a constant to avoid the denominator being zero, and in this embodiment, α=0.01 is set. It should be noted that the current flight element is any one of the three flight elements {h, v, f}.

[0042] Repeating the above process can obtain the progressive influence of each flight element. According to the above process, it can be understood that if the data values ​​of two adjacent groups of the current flight element change slightly, it can cause a large change in the impact load, that is, If it is much greater than 1, the higher the impact of the flight element on the impact load, the larger the value of the progressive impact degree is. On the contrary, if the impact of the flight element on the impact load is smaller, the value of the progressive impact degree is It is much smaller than 1, that is, the value of the progressive influence is smaller.

[0043] The influence of the gradual change is mainly measured by the gradual change of the values ​​between the current flight elements and the impact load of the UAV. At the same time, the influence of the flight elements is closely related to the overall trend of the impact load data, that is, if the trend change of the current flight elements is more consistent with the trend of the impact load, the higher the degree of connection between the two. In order to measure the trend of the two, the two sequences obtained by arranging the current flight element data and the UAV impact load data in ascending order are respectively decomposed by STL (Seasonal-Trend Decomposition using LOESS). According to the trend items of each sequence, the trend strength calculation formula is used to obtain the trend distribution of the current flight element data and the UAV impact load.

[0044] Based on the above analysis, according to the progressive influence of each flight element, combined with the difference between the data trend distribution of each flight element and the trend distribution of the impact load of the UAV, the synchronous consistency of the impact load of each flight element is obtained. The corresponding calculation formula in this embodiment is:

[0045] In the formula, C represents the synchronous consistency of the impact load of the current flight element, B represents the progressive influence of the current flight element, and ρ x and ρ Q They respectively represent the trend intensity of the current flight elements and the impact load of the UAV.

[0046] Therefore, if the trend of the current flight element and the UAV impact load in the overall data distribution is more consistent, the trend strength ρ of the current flight element is obtained. x The trend strength ρ of the impact load of UAV Q The closer the value is, the greater the value of the impact load synchronization consistency C of the current flight element is, indicating that the change of the current flight element has a greater impact on the impact load of the UAV.

[0047] Step 300: Based on the data of each flight element and the impact load data of the UAV, obtain the correlation vector between each flight element and the impact load and the total correlation vector between the three flight elements and the impact load, analyze the deviation between the correlation vector of each flight element and the total correlation vector, and obtain the contribution of each flight element to the impact load; calculate the adaptive fitting weight of each flight element based on the contribution of the three flight elements to the impact load.

[0048] Through step 200, the synchronous consistency of the impact load of each flight element on the impact load of the UAV can be obtained, and the influence of the current flight element on the impact load can be measured. However, since the data of the experimental group is obtained by sampling the continuous value range of the three flight elements through uniform experiments, the sampling of the same flight element is not distributed at equal intervals. The impact load of each set of data is the result of the combined effect of the three flight elements, and the synchronous consistency is only considered from the perspective of a single flight element, and the influence of the three flight elements is not measured from a comprehensive perspective, which may cause deviations in the measurement of importance, and thus further analysis is required in combination with the data of the experimental group.

[0049] In a uniform experiment, when the value of a single flight element changes, the values ​​of other flight elements also change. Therefore, it is necessary to measure the contribution of a single flight element to the impact load of the UAV according to the degree of change of the flight element. Therefore, in this embodiment, the data of each flight element is arranged in ascending order to obtain a data sequence of each flight element, and the impact load of the UAV is arranged in ascending order to form a UAV impact load sequence. Taking a single flight element as an example, this embodiment takes the flight altitude as an example, and arranges the data values ​​of the flight altitude in all experiments in ascending order to form a data sequence of the flight altitude. Then, the data sequence of the flight altitude and the elements at the corresponding positions in the UAV impact load sequence form various correlation vectors between the flight altitude and the impact load, such as y. i=(h i ,Q i ), y i represents the correlation vector between the ith flight altitude and the impact load, h i , Q i are the i-th element (flight altitude) in the data sequence of flight altitude and the i-th element (drone impact load) in the drone impact load sequence. Further, the data sequences of the three flight elements are obtained, and the data at each position in the three data sequences are combined with the drone impact load at the same position in the drone impact load sequence to construct the total correlation vectors between the three flight elements and the impact load, for example, z i =(h i ,v i ,f i ,Q i ), z i represents the total correlation vector between the ith three flight elements and the impact load, v i 、f i They are the i-th element in the data sequence of ambient wind speed and the data sequence of failure flight speed respectively.

[0050] If the angle between the correlation vector of a single flight element of two connected experimental groups is closer to the angle between the correlation vector of the three flight elements, it means that the flight altitude contributes more to the change of the impact load between the two experimental groups. That is, if θ(y i ,y i+1 ) and θ(z i ,z i+1 ) is closer, the higher the contribution of flight altitude to impact load.

[0051] Therefore, in this embodiment, the contribution of each flight element to the impact load is obtained by analyzing the deviation between the correlation vector of each flight element and the total correlation vector. The specific calculation formula is:

[0052] In the formula, D represents the contribution of the current flight element to the impact load, C represents the synchronization consistency of the impact load of the current flight element, N represents the number of data, θ(y i ,y i+1 ) represents the angle between the correlation vectors of the i-th and i+1-th current flight elements and the impact load, θ(z i ,z i+1 ) represents the angle between the total vector associated with the i-th and i+1-th three flight elements and the impact load, and α is a constant to avoid the denominator being zero.

[0053] For each flight element, if the angle between the correlation vector of a single flight element and the total correlation vector of the three flight elements in each adjacent experimental group is closer, it means that the other two flight elements have a smaller impact on the impact load of the UAV, reflecting that the current flight element, that is, the high flight speed, has a greater impact on the impact load.

[0054] The same method as above is used to obtain the contribution of flight altitude to impact load D h , the contribution of ambient wind speed to impact load D v The contribution of flight speed to impact load D f Furthermore, the contribution of a single flight element to the impact load is analyzed in proportion to the contribution and value of the three flight elements to the impact load, so as to obtain the adaptive fitting weight of a single flight element. Taking the flight altitude as an example, the adaptive fitting weight expression is:

[0055] In the formula, w h represents the adaptive fitting weight of the flight altitude, D h , D v and D f They respectively represent the contribution of flight altitude to impact load, the contribution of ambient wind speed to impact load and the contribution of time-dependent flight speed to impact load.

[0056] If the final flight altitude contributes more to the impact load and accounts for a higher proportion of the three flight elements, the adaptive fitting weight of the flight altitude will be larger, indicating that the flight altitude is more important in data fitting.

[0057] Step 400: Based on the adaptive fitting weights of each flight element, and in combination with the data of each flight element and the impact load data of the UAV, a fitting function between the impact load of the UAV and the three flight elements is fitted to obtain the maximum impact load, and the optimal value of the UAV airbag parameters is obtained using an optimization algorithm.

[0058] Based on the process of step 300, the adaptive fitting weight of each flight element can be obtained, that is, the adaptive fitting weight w of the flight altitude h , the adaptive fitting weight w of the ambient wind speed v and the adaptive fitting weight w of the failure flight speed f The data of 200 sets of uniform experiments and the adaptive fitting weights are then used in the polynomial fitting algorithm, where the polynomial order is set to 3, the convergence tolerance is 5%, and the maximum number of iterations is 200. The fitting function of the impact load of the UAV and the three flight elements is obtained, which is represented by Q = O (h, v, f), and the maximum impact load of the UAV is obtained, which is recorded as A max .

[0059] By performing finite element model analysis on the drone, the safe overload value of the drone is obtained, that is, under the safe overload value, the collision damage of the drone will not cause the normal operation of the drone. The method for obtaining the safe overload value of the drone is a prior art and will not be described in detail in this embodiment. The implementer can obtain it by himself in the actual application scenario.

[0060] In the design of drone airbags, the parameters that need to be designed are: airbag exhaust hole area x1, airbag initial inflation pressure x2, airbag exhaust threshold x3, and the maximum impact load A of the drone is obtained by the second-order response surface method. max The response surface fitting function is as follows: In order to further optimize the corresponding response surface model parameters and find a set of airbag design parameter values ​​that minimize the impact load of the UAV collision, this embodiment uses a genetic algorithm to optimize the airbag parameters, wherein the initial population is randomly generated, the population size is set to 150, a roulette wheel is used for selection, a single point crossover is used for crossover operation, the crossover coefficient is 0.82, random mutation is used for mutation operation, the coefficient of variation is 0.1, and the maximum number of iterations is set to 600. After iteration, the optimal parameter values ​​of each airbag parameter of the UAV are obtained. The genetic optimization algorithm belongs to the known technology, and the specific steps are not repeated here.

[0061] It should be noted that the response surface fitting function of the maximum impact load of the drone airbag and the optimization process of the airbag parameters are existing well-known technologies and are not elaborated in detail in this embodiment. In actual application, the implementer can select other optimization algorithms for optimization and set them by the implementer.

[0062] In this embodiment, the optimal parameter values ​​corresponding to the optimization by the genetic algorithm are that the exhaust hole area x1 of the drone airbag is 742.31 mm2, the initial inflation pressure x2 is 134.68 Kpa, and the exhaust threshold x3 is 270 Kpa. All the obtained optimized parameters are input into the finite element software for calculation, and the impact load value borne by the drone at this time is 11.9 g, which is much smaller than the safe overload value of 53.4 g of the drone, and is also smaller than the initial value of 27.6 g before optimization. Therefore, the above process of this embodiment can achieve the expected optimization effect. The specific results are shown in Table 1.

[0063] Table 1

[0064]

[0065]

[0066] Based on the same inventive concept as the above method, an embodiment of the present application also provides an impact load monitoring system for UAV collision avoidance, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, the steps of any one of the above-mentioned impact load monitoring methods for UAV collision avoidance are implemented.

[0067] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0068] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. A method for monitoring impact load for UAV collision avoidance, characterized in that: The method comprises the following steps: The impact load of the UAV in each group of experiments and the data of the three flight elements of the UAV are obtained through intelligent sensors; Based on the data differences of each flight element in all experiments and the gradual change of the impact load of the UAV, the gradual change influence of each flight element is determined. Combined with the difference between the data trend distribution of each flight element and the trend distribution of the impact load of the UAV, the synchronous consistency of the impact load of each flight element is obtained. According to the data of each flight element and the impact load data of each UAV, the correlation vector between each flight element and the impact load and the total correlation vector between the three flight elements and the impact load are obtained, the deviation between the correlation vector of each flight element and the total correlation vector is analyzed, and the contribution of each flight element to the impact load is obtained; the adaptive fitting weight of each flight element is calculated in combination with the contribution of the three flight elements to the impact load; Based on the adaptive fitting weights of each flight element and combined with the data of each flight element and the impact load data of the UAV, the fitting function between the impact load of the UAV and the three flight elements is fitted to obtain the maximum impact load, and the optimal value of the UAV airbag parameters is obtained using the optimization algorithm.

2. The impact load monitoring method for UAV collision avoidance according to claim 1 is characterized in that: The flight elements include flight altitude, ambient wind speed and failure flight speed.

3. The impact load monitoring method for UAV collision avoidance according to claim 1 is characterized in that: The calculation formula of the progressive influence of each flight element is: In the formula, B represents the influence of the current flight factor, N represents the number of data, ΔF i Indicates the absolute value of the difference between the i-th and i+1-th data values ​​after the data of the current flight element is sorted, ΔQ i It represents the absolute value of the difference between the i-th and i+1-th impact loads after the UAV impact loads are sorted, and α is a constant to avoid the denominator being zero. The sorting is to arrange the data values ​​of the current flight elements and the UAV impact loads in all experiments in ascending order.

4. The impact load monitoring method for UAV collision avoidance according to claim 3 is characterized in that: The calculation formula for the synchronous consistency of the impact loads of the flight elements is: Where, C represents the synchronous consistency of the impact load of the current flight elements, ρ x and ρ Q They represent the trend strength of the current flight elements and the impact load of the UAV respectively, where the trend strength is calculated by using the time series decomposition algorithm after sorting.

5. The impact load monitoring method for UAV collision avoidance according to claim 1 is characterized in that: The process of obtaining the correlation vector between each flight element and the impact load and the total correlation vector between the three flight elements and the impact load is as follows: Arrange the data of each flight element in ascending order to obtain the data sequence of each flight element, and arrange the impact load of the UAV in ascending order to form the impact load sequence of the UAV; Each correlation vector between each flight element and the impact load is composed of the elements at the same position in the data sequence of each flight element and the impact load sequence of the UAV; each total correlation vector between the three flight elements and the impact load is composed of the elements at the same position in the data sequence of all flight elements and the impact load sequence of the UAV.

6. The impact load monitoring method for UAV collision avoidance according to claim 1 is characterized in that: The calculation formula for the contribution of each flight element to the impact load is: In the formula, D represents the contribution of the current flight element to the impact load, C represents the synchronization consistency of the impact load of the current flight element, N represents the number of data, θ(y i ,y i+1 ) represents the angle between the correlation vectors of the i-th and i+1-th current flight elements and the impact load, θ(z i ,z i+1 ) represents the angle between the total vector associated with the i-th and i+1-th three flight elements and the impact load, and α is a constant to avoid the denominator being zero.

7. The impact load monitoring method for UAV collision avoidance according to claim 1 is characterized in that: The calculation formula of the adaptive fitting weight of each flight element is: In the formula, w h represents the adaptive fitting weight of the flight altitude, D h , D v and D f They respectively represent the contribution of flight altitude to impact load, the contribution of ambient wind speed to impact load and the contribution of time-dependent flight speed to impact load.

8. The impact load monitoring method for UAV collision avoidance according to claim 1 is characterized in that: The acquisition of the fitting function between the impact load of the UAV and the three flight elements includes: The data of the impact loads of each UAV and each flight element in all groups of experiments are combined with the adaptive fitting weights of each flight element, and a polynomial fitting algorithm is used to fit the fitting function between the UAV impact load and the three flight elements, where the independent variables in the fitting function are the three flight elements, and the dependent variable is the UAV impact load.

9. The impact load monitoring method for UAV collision avoidance according to claim 1 is characterized in that: The acquisition of the optimal value of the drone airbag parameter includes: According to the relationship between the maximum impact load and the design parameters of the UAV airbag, the response fitting function of the maximum impact load of the UAV is obtained, and the optimal values ​​of each airbag parameter of the UAV are obtained using the genetic optimization algorithm.

10. An impact load monitoring system for unmanned aerial vehicle collision avoidance, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 9 are implemented.

Images