A fuel cell life prediction method based on hydrogen-powered drones

Abstract

The present invention relates to the field of battery life testing technology, and more specifically, to a fuel cell life prediction method for hydrogen-powered drones. The method first acquires and segments monitoring data. Then, by analyzing the flight monitoring data from the perspectives of the drone's flight and fuel cell output, it extracts the power demand, output capacity coefficient, and output attenuation coefficient. This data is then integrated with historical influences to obtain a corrected attenuation coefficient. Finally, the life attenuation rate is calculated based on the overall power demand. This method addresses the issue of inaccurate fuel cell life predictions caused by flight state interference, thereby improving the safety of hydrogen-powered drones during flight.

Description

Technical Field

[0001] The present invention relates to the technical field of battery life testing, and in particular to a fuel cell life prediction method based on a hydrogen-powered unmanned aerial vehicle. Background Art

[0002] Hydrogen fuel cells (HFCs) offer advantages such as high energy conversion efficiency and zero pollution emissions, making them ideal for use in lightweight aircraft such as drones. Hydrogen drones require a continuous and stable energy supply to avoid mission interruptions caused by sudden failures, necessitating lifespan predictions for their fuel cells.

[0003] Existing technology predicts the lifespan of a fuel cell by analyzing its output during flight. However, during the operation of a drone, different flight modes will have different impacts on the working state of the fuel cell, especially under high load or frequent starts and stops, making the lifespan prediction results inaccurate. Summary of the Invention

[0004] In order to solve the technical problem that the flight of hydrogen-powered UAVs affects the accuracy of battery life prediction, the purpose of the present invention is to provide a fuel cell life prediction method based on hydrogen-powered UAVs. The technical solutions adopted are as follows:

[0005] Acquire monitoring data from the most recent flight mission of the hydrogen-powered UAV and segment it into segments with a preset time domain length; the monitoring data includes flight altitude and acceleration, and output power and voltage of the fuel cell;

[0006] Obtaining the electrical energy demand for each flight period based on the change in altitude and the acceleration during each period; clustering and classifying the electrical energy demand; obtaining the output capacity coefficient for each period based on the fluctuation of the output power and voltage during each period; and obtaining the output attenuation coefficient for each corresponding period based on the overall characteristics of each type of electrical energy demand and the fluctuation of the output capacity coefficient;

[0007] A preset number of adjacent historical time periods in each time period are recorded as impact time periods; based on the distribution of the power demand in the impact time period of each time period in a category, the power demand and the output attenuation coefficient are combined to obtain a corrected attenuation coefficient; based on the corrected attenuation coefficient of all types of time periods and the overall power demand, the life decay rate is obtained.

[0008] Furthermore, the method for obtaining the output attenuation coefficient includes:

[0009] Obtaining a fitting curve of each type of output capacity coefficient;

[0010] According to the overall slope and intercept of the fitting curve and in combination with the overall electric energy demand, the output attenuation coefficient corresponding to each time period is obtained.

[0011] Furthermore, the method for obtaining the fitting curve includes:

[0012] In each type of the output capacity coefficients, they are sorted according to the time sequence of the corresponding time periods, with the horizontal axis representing the time period and the vertical axis representing the output capacity coefficients, and a fitting curve is obtained by the least square method.

[0013] Furthermore, the method for obtaining the modified attenuation coefficient includes:

[0014] Obtaining an impact factor for each time period according to a time interval corresponding to a mean value of the electric energy demand in the impact time period and a range of the electric energy demand;

[0015] The electric energy demand and the affected factors in each time period within a type of time period are integrated with the output attenuation coefficient to obtain a modified attenuation coefficient.

[0016] Furthermore, the method for obtaining the life decay rate includes:

[0017] The modified attenuation coefficient of each time period is used as the numerator, the overall electric energy demand is used as the denominator, the fractional ratio is used as the life attenuation sub-parameter, and the life attenuation sub-parameters of all types of time periods are integrated to obtain the life attenuation rate.

[0018] Furthermore, the method for obtaining the output capacity coefficient includes:

[0019] The output capacity coefficient of each time period is obtained according to the variance of the output power and the range of the voltage in each time period.

[0020] Furthermore, the method for obtaining the electric energy demand includes:

[0021] The power demand for flight in each period is obtained based on the overall acceleration value of the flight process in each period and the height difference between the first and last moments.

[0022] Furthermore, the preset number is 5.

[0023] Furthermore, the preset time domain length is 1 second.

[0024] Furthermore, the clustering algorithm used in the cluster classification is the DBSCAN clustering algorithm.

[0025] The present invention has the following beneficial effects:

[0026] The present invention first obtains monitoring data and segments it to provide a data basis; further obtains the power demand according to the change in altitude and acceleration in each time period, and performs cluster classification based on this to characterize the degree of power demand during flight, and provide a basis for subsequent analysis of battery life attenuation from the perspective of UAV flight; further obtains the output capacity coefficient according to the fluctuation of output power and voltage in each time period to characterize the output capacity of the battery, and provide a basis for subsequent analysis of life attenuation from the perspective of fuel cell output; further obtains the corresponding output attenuation coefficient according to the overall characteristics of each type of power demand, combined with the fluctuation of the output capacity coefficient, and preliminarily characterizes the output attenuation degree of the fuel cell; further obtains the corrected attenuation coefficient according to the distribution of power demand of the influencing time period of each time period in a type of time period, combined with the power demand and the output attenuation coefficient, eliminates the influence of energy demand at the front of the time sequence, and improves the accuracy of characterizing the degree of battery output attenuation; finally, obtains the life attenuation rate according to the corrected attenuation coefficient and the overall power demand of all types of time periods. The present invention extracts power demand, output capacity coefficient and output attenuation coefficient by analyzing flight monitoring data, integrates historical influences to obtain a corrected attenuation coefficient, and finally calculates the life attenuation rate based on the overall power demand, thereby solving the problem of inaccurate fuel cell life prediction caused by flight status interference. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. 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 any creative work.

[0028] Figure 1 A flowchart of a method for predicting fuel cell lifespan for a hydrogen-powered UAV provided by one embodiment of the present invention;

[0029] Figure 2 A schematic diagram of a fitting curve of an output capacity coefficient provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0030] To further illustrate the technical means and effectiveness of the present invention in achieving its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, describes in detail the specific implementation, structure, features, and effectiveness of a fuel cell life prediction method for a hydrogen-powered drone. In the following description, references to different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.

[0031] 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.

[0032] The following describes in detail a specific scheme of a fuel cell life prediction method based on a hydrogen-powered UAV provided by the present invention in conjunction with the accompanying drawings.

[0033] See also Figure 1 , which shows a flow chart of a fuel cell life prediction method based on a hydrogen-powered UAV provided by one embodiment of the present invention, specifically including:

[0034] Step S1: Acquire monitoring data from the most recent flight mission of the hydrogen-powered UAV and segment it into segments according to a preset time domain length; the monitoring data includes the flight altitude and acceleration, and the output power and voltage of the fuel cell.

[0035] In one embodiment of the present invention, monitoring data from the most recent flight mission of a hydrogen-powered UAV is first obtained to provide a data basis. Considering that a flight mission of a hydrogen-powered UAV lasts a long time, the data is segmented into preset time domain lengths to facilitate a more detailed analysis of the flight mode of the hydrogen-powered UAV and an accurate assessment of the fuel cell life.

[0036] A flexible printed circuit board (FPCB) is embedded between the fuel cell stacks, housing an integrated thin-film voltage sensor (±0.5mV accuracy) to collect the fuel cell output voltage during operation. A Hall-effect sensor (DC-10kHz bandwidth) is used for current detection, installed on the power output bus to collect real-time output current. Position and acceleration sensors are also installed on the drone to provide real-time information on the hydrogen drone's altitude and acceleration during flight. The method of obtaining power through current and voltage is already well-established and will not be elaborated on here.

[0037] The acquisition frequency of all data is set to 50 Hz, and the preset time domain length is 1 second, that is, each flight period is 1 second. In other embodiments of the present invention, the implementer can set other data acquisition frequencies and preset time domain lengths.

[0038] Step S2: Obtain the power demand for flight in each time period based on the change in altitude and acceleration in each time period; cluster and classify the power demand; obtain the output capacity coefficient for each time period based on the fluctuation of output power and voltage in each time period; obtain the output attenuation coefficient for each type of time period based on the overall characteristics of each type of power demand and the fluctuation of the output capacity coefficient.

[0039] During the operation of hydrogen-powered drones, changes in flight modes (such as climbing, cruising, and hovering) will cause changes in the load on the fuel cell, resulting in varying degrees of ripples in the fuel cell output power. Therefore, it is necessary to analyze the power output status of the fuel cell according to the different flight states of the hydrogen-powered drone;

[0040] Considering that the change in altitude and acceleration within a time period is an important reflection of the flight status, the power demand for each period of flight is obtained based on the change in altitude and acceleration in each period of flight, which characterizes the degree of power demand during flight, provides a basis for the subsequent classification of flight periods, and provides a basis for the subsequent analysis of battery life attenuation from the perspective of UAV flight.

[0041] Preferably, in one embodiment of the present invention, the higher the overall acceleration value within a time period, the greater the change in flight speed, which reflects a greater change in flight state and a greater power demand; the greater the difference in altitude between the beginning and end of a time period, the greater the change in altitude and the greater the power demand;

[0042] Based on this, the power demand for flight in each period is obtained according to the overall acceleration value of the flight process in each period and the height difference between the first and last moments.

[0043] As an example, the absolute value of the difference in altitude between the first and last moments of a time period is used as the altitude difference value, and the average of the absolute values ​​of all accelerations in each time period is multiplied by the altitude difference value. The product is used as the independent variable and normalized and mapped through the th(x) function. The mapping result is used as the power demand for flight in each time period.

[0044] Among them, the higher the energy demand of the flight process, the greater the pressure on the fuel cell of the hydrogen drone, and the more likely it is to cause the fuel cell life to be lost.

[0045] In other embodiments of the present invention, linear normalization can also be used to obtain the power demand. Specifically, the data to be normalized is normalized within the corresponding data dimension. The normalization used in the embodiments of the present invention can adopt this method, which is a technical means well known to those skilled in the art.

[0046] In another embodiment of the present invention, it is also taken into consideration that the higher the flight speed of the UAV, the greater the power consumption and the greater the demand, so the monitoring data also includes the flight speed, and the average value of the speed within the time period, the average value of the absolute value of the acceleration and the height difference value are fused by multiplication or weighted summation, and the fusion result is linearly normalized to obtain the power demand.

[0047] The sum of the absolute values ​​of the height differences at all adjacent moments within a time period can also be counted as the height change during the time period, replacing the height difference value to obtain the power demand.

[0048] Considering that the power demand of flights at different times is different and has different impacts on fuel cells, in order to analyze the loss of fuel cells caused by UAV flight, the power demand is clustered and classified, and data subsets with similar working conditions are screened out to improve the analysis accuracy.

[0049] In one embodiment of the present invention, the clustering algorithm used for cluster classification is the DBSCAN (Density-Based Spatial Clustering of Applications with Noise) clustering algorithm, which is already an existing technology and will not be described in detail.

[0050] During the operation of a hydrogen-powered drone, the fuel cell's output power fluctuates due to the fuel cell's lifespan degradation or load changes during operation. Therefore, it is necessary to obtain the fuel cell output during operation to predict the fuel cell's lifespan. Therefore, based on the fluctuations in output power and voltage during each period, the output capacity coefficient for each period is obtained to characterize the battery's output capacity, providing a basis for subsequent analysis of lifespan degradation from the perspective of fuel cell output.

[0051] Preferably, in one embodiment of the present invention, considering that the flight mode within a time period is relatively consistent, the larger the variance of the output power, the more unstable the output power, and the larger the range of voltage, the more unstable the battery output. Therefore, the output capacity coefficient of each time period is obtained based on the variance of the output power and the range of voltage in each time period.

[0052] As an example, the product of the output power variance and the voltage range within a period is used as the independent variable, and negative correlation mapping is performed through the exp(-x) function. The mapping result is used as the output capacity coefficient of the corresponding period, and x represents the independent variable.

[0053] The fluctuation of output power is expressed in the form of variance, and the fluctuation of voltage is expressed in the form of range. The output capacity coefficient of each time period is obtained by integrating the two.

[0054] As the life of the UAV fuel cell decreases, the power output of the fuel cell may still be poor when the hydrogen-powered UAV is in a flight process with low power demand. Therefore, based on the overall characteristics of each type of power demand and the fluctuation of the output capacity coefficient, the corresponding output attenuation coefficient for each time period is obtained to characterize the output attenuation degree of the fuel cell and provide a basis for the subsequent evaluation of life attenuation.

[0055] Preferably, in one embodiment of the present invention, it is convenient to analyze the fluctuation of the output capacity coefficient and obtain a fitting curve for each type of output capacity coefficient;

[0056] As an example: the methods for obtaining the fitting curve include:

[0057] In each type of output capacity coefficient, the corresponding time periods are sorted according to their chronological order, with the horizontal axis representing the time period and the vertical axis representing the output capacity coefficient. The fitting curve is obtained using the least squares method.

[0058] See also Figure 2 , which shows a schematic diagram of a fitting curve of a type of output capacity coefficient provided by an embodiment of the present invention; Figure 2 The horizontal axis is the serial number of the time period (a type of time period is sorted in chronological order), and the vertical axis is the data axis of the output capacity coefficient.

[0059] Considering that the greater the downward trend of the overall slope of the fitting curve, the smaller the intercept of the vertical axis, the lower the output capacity of the fuel cell, and the faster it decreases, the higher the degree of attenuation; and the lower the overall power demand, the lower the output capacity of the fuel cell, which reflects that when the hydrogen drone is in a flight process with low power demand, the power output of the fuel cell is still poor, and the output attenuation coefficient is larger;

[0060] Therefore, according to the overall slope and intercept of the fitting curve and the overall power demand, the output attenuation coefficient corresponding to each time period is obtained.

[0061] As an example, the linear normalization result of the average slope of each point of the fitting curve in the corresponding data dimension is used as the slope factor, which represents the degree of decrease of the overall slope. The smaller the slope factor, the greater the degree of decrease. The intercept of the fitting curve, that is, the product of the output capacity coefficient of the time period with the smallest time sequence and the slope factor is used as the denominator, the average value of the electric energy demand corresponding to a type of time period is used as the numerator, the fractional ratio is used as the independent variable, and negative correlation mapping is performed through the exp(-x) function. The mapping result is used as the output attenuation coefficient of the corresponding time period.

[0062] Among them, the fluctuation of the output capacity coefficient is shown by analyzing the decline degree and intercept of the fitting curve of the output capacity coefficient.

[0063] It should be noted that the time periods correspond to the electric energy demands, and the time periods are classified while the electric energy demands are clustered and classified. The analysis process for each type of time period is the same, and only any one type is described here. In other embodiments of the present invention, the product of the output capacity coefficient of the time period with the smallest time sequence and the slope factor can be used as the numerator, and the average value of the electric energy demand corresponding to a type of time period can be used as the denominator. The fractional ratio is linearly normalized in the corresponding data dimension, and the result is used as the output attenuation coefficient of the corresponding time period.

[0064] Step S3: Record a preset number of adjacent historical periods in each period as impact periods; obtain a corrected attenuation coefficient based on the distribution of power demand in the impact period of each period in a category, combined with the power demand and the output attenuation coefficient; obtain the life attenuation rate based on the corrected attenuation coefficient of all types of periods and the overall power demand.

[0065] Since the flight mode of the drone will change continuously during the flight, there may be a flight process with high power demand in the front, which will affect the power output capacity of the fuel cell in the flight process with low power demand in the back, thereby causing a deviation in the obtained output attenuation coefficient. Therefore, a preset number of adjacent historical time periods in each time period are recorded as the impact period, and the interference of the impact period on each time period is analyzed.

[0066] Taking into account the distribution of electric energy demand in the influencing period of each time period, it represents the degree of influence of the energy demand in the influencing period on each time period; the electric energy demand of the time period itself represents the susceptibility of itself to being affected, and the output attenuation coefficient provides a correction basis. Therefore, according to the distribution of electric energy demand in the influencing period of each time period in a type of time period, combined with the electric energy demand and the output attenuation coefficient, the corrected attenuation coefficient is obtained to eliminate the influence of the energy demand at the front of the time sequence and improve the accuracy of characterizing the degree of battery output attenuation.

[0067] Preferably, in one embodiment of the present invention, the larger the average value of the power demand in the influencing period, the greater the load pressure on the fuel cell caused by the energy demand in the previous period, and the more likely it is to affect the current period; the larger the time interval corresponding to the range of the power demand in the influencing period, the longer the duration of the high level of power demand, and the greater the impact on the current period; and the smaller the power demand in the current period itself, the more likely it is to be affected by the interference of the previous influencing period, and the output attenuation coefficient of the fuel cell needs to be reduced accordingly;

[0068] Based on this, the influence factor of each time period is obtained according to the time interval corresponding to the mean value of the power demand in the influencing period of each time period and the range of the power demand;

[0069] The power demand and the influencing factors of each time period within a certain time period are integrated with the output attenuation coefficient to obtain the corrected attenuation coefficient.

[0070] As an example, in the impact period of each period, the product of the time interval between the periods corresponding to the maximum and minimum values ​​of the electric energy demand and the mean value of the electric energy demand is used as the independent variable, and negative correlation mapping is performed through the exp(-x) function. The mapping result is used as the impact factor of each period;

[0071] For a type of time period, the average value of the ratio of the power demand in each time period to its affected factor, and the product of the output attenuation coefficient are used as independent variables. After mapping through the th(x) function, the mapping result is used as the corrected attenuation coefficient corresponding to the type of time period.

[0072] In other embodiments of the present invention, the th(x) function may be replaced by a linear normalization method, and normalization may be performed in the corresponding data dimension to obtain a modified attenuation coefficient.

[0073] The actual attenuation of a fuel cell's life depends not only on changes in its output capacity (i.e., the corrected attenuation coefficient), but also on the actual load pressure (i.e., the amount of electrical energy required during flight). Furthermore, different flight modes of hydrogen-powered drones will have different effects on the life of the fuel cell, so the life attenuation rate is ultimately obtained based on the corrected attenuation coefficient and overall electrical energy demand for all time periods.

[0074] Preferably, in one embodiment of the present invention, considering that under normal circumstances the power output attenuation rate of the flight mode with low power demand is low, and when the life of the fuel cell is consumed, the power output attenuation rate of the flight mode with low power demand will be higher, based on this, the corrected attenuation coefficient of each type of time period is used as the numerator, the overall power demand is used as the denominator, and the fractional ratio is used as the life attenuation sub-parameter, and the life attenuation sub-parameters of all types of time periods are integrated to obtain the life attenuation rate.

[0075] As an example, the life decay sub-parameter represents the corrected attenuation coefficient of the battery output per unit of electrical energy demand in each time period. The larger the life decay sub-parameter, the higher the battery decay rate in the flight mode corresponding to the time period. Finally, the average value of the life decay sub-parameters of all time periods is used as the independent variable, and after mapping through the th(x) function, the mapping result is used as the life decay rate.

[0076] Similarly, in other embodiments of the present invention, the th(x) function may be replaced by a linear normalization method, and normalization may be performed in the corresponding data dimension to obtain the life decay rate.

[0077] In one embodiment of the present invention, after obtaining the life decay rate, the method further includes: determining whether the fuel cell has reached the end of its life based on the fuel cell life decay rate after the flight, for example, determining that the fuel cell has reached the end of its life when the life decay rate is higher than 0.6;

[0078] When the fuel cell reaches the end of its life, it should be replaced and properly recycled. Modern hydrogen fuel cells contain a variety of precious metals and other resources. Recycling not only helps reduce environmental pollution, but also helps reuse resources.

[0079] In summary, in response to the technical problem that the flight of hydrogen-powered UAVs affects the accuracy of battery life prediction, the present invention proposes a fuel cell life prediction method based on hydrogen-powered UAVs. The present invention first obtains monitoring data and segments it; further obtains the power demand based on the change in altitude and acceleration in each time period, and performs cluster classification based on this; further obtains the output capacity coefficient based on the fluctuation of output power and voltage in each time period; further obtains the corresponding output attenuation coefficient based on the overall characteristics of each type of power demand, combined with the fluctuation of the output capacity coefficient; further obtains the corrected attenuation coefficient based on the distribution of power demand in the influencing period of each time period within a type of time period, combined with the power demand and the output attenuation coefficient; finally, obtains the life attenuation rate based on the corrected attenuation coefficient of all types of time periods and the overall power demand. The present invention extracts power demand, output capacity coefficient and output attenuation coefficient by analyzing flight monitoring data, obtains the corrected attenuation coefficient by integrating historical influence, and finally calculates the life attenuation rate based on the overall power demand, thereby solving the problem of inaccurate fuel cell life prediction caused by flight state interference.

[0080] It should be noted that the order in which the embodiments of the present invention are described above is for illustrative purposes only and does not necessarily represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or sequential order shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0081] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments.

Claims

1. A fuel cell life prediction method based on a hydrogen-powered drone, characterized in that: The method comprises: Acquire monitoring data from the most recent flight mission of the hydrogen-powered UAV and segment it into segments with a preset time domain length; the monitoring data includes flight altitude and acceleration, and output power and voltage of the fuel cell; Obtaining the electrical energy demand for each flight period based on the change in altitude and the acceleration during each period; clustering and classifying the electrical energy demand; obtaining the output capacity coefficient for each period based on the fluctuation of the output power and voltage during each period; and obtaining the output attenuation coefficient for each corresponding period based on the overall characteristics of each type of electrical energy demand and the fluctuation of the output capacity coefficient; A preset number of adjacent historical periods in each period are recorded as impact periods; based on the distribution of the power demand in the impact period of each period in a category, the power demand and the output attenuation coefficient are combined to obtain a corrected attenuation coefficient; based on the corrected attenuation coefficient and the overall power demand of all types of time periods, the life decay rate is obtained.

2. The method for predicting fuel cell life of a hydrogen-powered UAV according to claim 1, characterized in that: The method for obtaining the output attenuation coefficient includes: Obtaining a fitting curve of each type of output capacity coefficient; According to the overall slope and intercept of the fitting curve and in combination with the overall electric energy demand, the output attenuation coefficient corresponding to each time period is obtained.

3. The method for predicting fuel cell life of a hydrogen-powered UAV according to claim 2, characterized in that: The method for obtaining the fitting curve includes: In each type of the output capacity coefficients, they are sorted according to the time sequence of the corresponding time periods, with the horizontal axis representing the time period and the vertical axis representing the output capacity coefficients, and a fitting curve is obtained by the least square method.

4. The method for predicting fuel cell life of a hydrogen-powered UAV according to claim 1, characterized in that: The method for obtaining the modified attenuation coefficient includes: Obtaining an impact factor for each time period according to a time interval corresponding to a mean value of the electric energy demand in the impact time period and a range of the electric energy demand; The electric energy demand and the affected factors in each time period within a type of time period are integrated with the output attenuation coefficient to obtain a modified attenuation coefficient.

5. The method for predicting fuel cell life of a hydrogen-powered UAV according to claim 1, characterized in that: The method for obtaining the life decay rate includes: The modified attenuation coefficient of each time period is used as the numerator, the overall power demand is used as the denominator, and the fractional ratio is used as the life attenuation sub-parameter. The life attenuation sub-parameters of all types of time periods are integrated to obtain the life attenuation rate.

6. The method for predicting fuel cell life of a hydrogen-powered UAV according to claim 1, characterized in that: The method for obtaining the output capacity coefficient includes: The output capacity coefficient of each time period is obtained according to the variance of the output power and the range of the voltage in each time period.

7. The method for predicting fuel cell life of a hydrogen-powered UAV according to claim 1, characterized in that: The method for obtaining the electric energy demand includes: The power demand for flight in each period is obtained based on the overall acceleration value of the flight process in each period and the height difference between the first and last moments.

8. The method for predicting fuel cell life of a hydrogen-powered UAV according to claim 1, characterized in that: The preset number is 5.

9. The method for predicting fuel cell life of a hydrogen-powered UAV according to claim 1, characterized in that: The preset time domain length is 1 second.

10. The method for predicting fuel cell life of a hydrogen-powered UAV according to claim 1, characterized in that: The clustering algorithm used in the cluster classification is the DBSCAN clustering algorithm.

Images