Dynamic adjustment method and system for thermal runaway early warning threshold value of hydrogen energy unmanned aerial vehicle

Abstract

The invention discloses a dynamic adjustment method and system for a thermal runaway early warning threshold value of a hydrogen energy unmanned aerial vehicle, and relates to the technical field of data analysis. The system comprises a data acquisition and mapping module, an early warning event classification module, a feature extraction and trend analysis module and a threshold dynamic adjustment module. The data acquisition and mapping module is responsible for acquiring historical monitoring data and establishing a mapping table; the early warning event classification module classifies the early warning events and extracts feedback information; the feature extraction and trend analysis module extracts features from historical monitoring data and performs trend analysis; and the threshold value dynamic adjustment module adjusts the thermal runaway early warning threshold value according to the difference coefficient. The modules work cooperatively, the thermal runaway early warning threshold value can be dynamically adjusted according to historical monitoring data and early warning event records of the hydrogen energy unmanned aerial vehicle, and the safety of the unmanned aerial vehicle and the ability of coping with sudden risks are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data analysis, and in particular to a method and system for dynamically adjusting a thermal runaway warning threshold of a hydrogen-powered unmanned aerial vehicle. Background Art

[0002] With the continuous development of drone technology, hydrogen drones have gradually gained widespread application as a green and environmentally friendly flight platform. However, the risk of thermal runaway of hydrogen drones is still one of the key factors affecting their safety. Thermal runaway refers to the excessive temperature inside the drone due to changes in the internal and external environment, equipment failure or abnormal operation, which in turn causes serious accidents such as combustion and explosion. Therefore, how to timely and accurately monitor and warn of thermal runaway events has become an important research topic to ensure the flight safety of hydrogen drones.

[0003] At present, most existing thermal runaway warning systems rely on statically set thresholds for warning, that is, once a parameter (such as temperature, air pressure, etc.) exceeds the set threshold during flight, the system will trigger a warning. However, the flight state and environmental conditions (such as temperature, humidity, flight altitude, etc.) of hydrogen-powered drones are dynamically changeable, and fixed thresholds are difficult to adapt to changes in different flight states and environments, which can easily lead to false alarms. For example, in certain specific flight states (such as takeoff or high-speed cruising), the temperature of the drone may fluctuate briefly, but this does not mean that there is a risk of thermal runaway, and the static threshold may not be able to accurately identify this situation. Summary of the invention

[0004] The purpose of the present invention is to provide a method and system for dynamically adjusting the thermal runaway warning threshold of a hydrogen-powered UAV to solve the problems raised in the above-mentioned background technology.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0006] A method for dynamically adjusting a thermal runaway warning threshold of a hydrogen-powered UAV comprises the following steps:

[0007] Step S100. Obtain the thermal runaway warning event records and historical monitoring data of the hydrogen-powered UAV in the past period of time, match the corresponding historical monitoring data based on the timestamp of the thermal runaway warning event records; analyze the matched historical monitoring data and thermal runaway warning event records, so as to obtain the mapping relationship between the thermal runaway warning event records and the historical monitoring data;

[0008] Step S200. Based on the mapping relationship between the thermal runaway warning event record and the historical monitoring data, feedback information of each thermal runaway warning event is obtained, and the thermal runaway warning events are classified into first warning events, second warning events and standard warning events; for the standard warning events, the standard monitoring features corresponding to the standard warning events are extracted in combination with the preset thresholds;

[0009] Step S300. For the first warning event, in combination with the preset threshold, extract the first monitoring feature corresponding to the first warning event; analyze the first monitoring feature and the standard monitoring feature, and calculate the first difference coefficient between the two; and obtain the dynamic thermal runaway warning threshold of the first warning event based on the first difference coefficient between the two;

[0010] Step S400. For the second warning event, in combination with the preset threshold, extract the second monitoring feature corresponding to the second warning event; analyze the second monitoring feature with the standard monitoring feature to calculate the second difference coefficient between the two; and based on the second difference coefficient between the two, obtain the dynamic thermal runaway warning threshold of the second warning event.

[0011] Furthermore, step S100 includes:

[0012] S101. Obtain the thermal runaway warning event record E and historical monitoring data D of the hydrogen-powered UAV in the past period of time, which are expressed as: E={e1,e2,...,en}, D={d1,d2,...,dm}, where e1 represents the first thermal runaway warning event of the hydrogen-powered UAV in the past period of time, e2 represents the second thermal runaway warning event of the hydrogen-powered UAV in the past period of time, and so on. en represents the nth thermal runaway warning event of the hydrogen-powered UAV in the past period of time, and n represents the total number of events in the thermal runaway warning event record E; d1 represents the number of thermal runaway warning events of the hydrogen-powered UAV in the past period of time. d1 represents the historical monitoring data corresponding to the first flight mission of the hydrogen-powered UAV in the past period of time, d2 represents the historical monitoring data corresponding to the second flight mission of the hydrogen-powered UAV in the past period of time, and so on, dm represents the historical monitoring data corresponding to the mth flight mission of the hydrogen-powered UAV in the past period of time, and m represents the total number of flight missions of the hydrogen-powered UAV in the past period of time; among them, the thermal runaway warning event record contains warning event information related to thermal runaway of the hydrogen-powered UAV, such as the timestamp, type and description of the event; the historical monitoring data records the monitoring data during the hydrogen-powered UAV flight mission, including various flight parameters and environmental data;

[0013] S102. Obtain the timestamp ti of each thermal runaway warning event ei in the thermal runaway warning event record E, and the flight mission time period Tj corresponding to the historical monitoring data dj, and match the thermal runaway warning event ei that satisfies ti∈Tj with the corresponding historical monitoring data dj; summarize the matching results between all thermal runaway warning events and historical monitoring data, establish a mapping relationship between the thermal runaway warning event records and the historical monitoring data, and form a corresponding mapping relationship table, and the structure of each mapping relationship in the mapping relationship table is: N(ei)→ei→dj, where N(ei) represents the index of the thermal runaway warning event ei in the mapping relationship table.

[0014] Further, step S200 includes:

[0015] S201. For each mapping relationship between the thermal runaway warning event record and the historical monitoring data in the mapping relationship table, the feedback information of the thermal runaway warning event is extracted from the thermal runaway warning event record, and the label of the corresponding thermal runaway warning event is obtained according to the feedback information of the thermal runaway warning event, wherein the labels are F, T and B respectively, and the label F indicates that the monitoring parameters in the historical monitoring data of the corresponding hydrogen-powered UAV exceed the preset threshold value, and the thermal runaway warning information is received. According to the analysis of relevant personnel, there is no thermal runaway risk; the label T indicates that the monitoring parameters in the historical monitoring data of the corresponding hydrogen-powered UAV do not exceed the preset threshold value, and the thermal runaway warning information is not received. According to the analysis of relevant personnel, there is a thermal runaway risk; the label B indicates that the monitoring parameters in the historical monitoring data of the corresponding hydrogen-powered UAV exceed the preset threshold value, and the thermal runaway warning information is received. According to the analysis of relevant personnel, there is a thermal runaway risk; according to the label of the thermal runaway warning event, the thermal runaway warning event corresponding to the label F is marked as the first warning event, the thermal runaway warning event corresponding to the label T is marked as the second warning event, and the thermal runaway warning event corresponding to the label B is marked as the standard warning event;

[0016] S202. For each standard warning event eb, obtain the index of the corresponding mapping relationship table, obtain the corresponding thermal runaway warning event record and historical monitoring data in the mapping relationship table, extract the thermal runaway warning trigger timestamp tb corresponding to the standard warning event eb based on the thermal runaway warning event record, and construct the analysis time window Tb with the thermal runaway warning trigger timestamp tb as the reference point, and Tb=[tb-Δtpr,t+Δtpo], wherein Δtpr represents the warning traceback time, and the statistical mode of the time when the parameters of all standard warning events before the triggering first break through the preset threshold is taken; Δtpo represents the warning delay time, and the statistical mode of the stable time after the risk of all standard warning events is confirmed is taken; according to the analysis time window Tb, obtain the historical monitoring data of each standard warning event eb within the analysis time window Tb, perform trend analysis on the historical monitoring data within the analysis time window Tb, thereby obtaining the corresponding trend curve Qb, and extract the standard monitoring features of the corresponding standard warning event from the trend curve Qb, and perform standardization on the standard monitoring features, thereby obtaining the standard monitoring feature vector Vb.

[0017] Furthermore, step S300 includes:

[0018] S301. For each first warning event ef, refer to the analysis process of the historical monitoring data of the standard warning event eb, obtain the corresponding thermal runaway warning trigger timestamp tf and analysis time window Tf in turn, extract the corresponding historical monitoring data according to the analysis time window Tf, perform trend analysis on the historical monitoring data within the analysis time window Tf, thereby obtaining the corresponding trend curve Qf, and extract the first monitoring feature corresponding to the first warning event from the trend curve Qf, and perform standardization on the first monitoring feature, thereby obtaining the first monitoring feature vector Vf;

[0019] S302. Summarize the standard monitoring feature vectors Vb of all standard warning events, calculate the average feature vector Vavg of the standard warning events; calculate the difference coefficient between the first monitoring feature vector Vf of each first warning event ef and the average feature vector Vavg of the standard warning event, and the corresponding calculation formula is: C(Vf,Vavg)=[∑K k=1wk·(fk-avg_k) 2 ] (1 / 2) ; Wherein C(Vf,Vavg) represents the difference coefficient between the first monitoring feature vector Vf and the average feature vector Vavg of the standard warning event, fk represents the feature value of the kth dimension in the first monitoring feature vector Vf, avg_k represents the feature value of the kth dimension in the average feature vector Vavg, wk represents the weight of the kth feature dimension, and the weight can be determined based on historical data, expert evaluation or machine learning model; K represents the dimension of the first monitoring feature vector Vf and the average feature vector Vavg;

[0020] S303. Obtain the preset threshold, and calculate the dynamic thermal runaway warning threshold of the first warning event in combination with the difference coefficient C(Vf, Vavg) corresponding to each first warning event ef. The corresponding calculation formula is:

[0021] Rf=R0×[1+α×(C(Vf,Vavg)-C1_min) / (C1_max-C1_min)];

[0022] Among them, Rf represents the dynamic thermal runaway warning threshold of the first warning event, R0 represents the preset threshold, α represents the adjustment factor, and the value is greater than 0; C1_min represents the minimum value of the difference coefficient between the first monitoring feature vector Vf and the average feature vector Vavg of the standard warning event, and C1_max represents the maximum value of the difference coefficient between the first monitoring feature vector Vf and the average feature vector Vavg of the standard warning event.

[0023] Furthermore, step S400 includes:

[0024] S401. For each second warning event et, refer to the analysis process of the first warning event ef to analyze, so as to obtain the corresponding trend curve Qt, and extract the second monitoring feature corresponding to the second warning event from the trend curve Qt; and standardize the second monitoring feature to obtain the second monitoring feature vector Vt; refer to the calculation formula of the difference coefficient between the first monitoring feature vector Vf of the first warning event ef and the average feature vector Vavg of the standard warning event, and calculate the difference coefficient C(Vt,Vavg) between the second monitoring feature vector Vt of each second warning event et and the average feature vector Vavg of the standard warning event;

[0025] S402. Obtain a preset threshold value, and calculate the dynamic thermal runaway warning threshold value of the second warning event in combination with the difference coefficient C(Vt, Vavg) corresponding to each second warning event et. The corresponding calculation formula is:

[0026] Rt=R0×[1-β×(C(Vt,Vavg)-C2_min) / (C2_max-C2_min)];

[0027] Among them, Rt represents the dynamic thermal runaway warning threshold of the second warning event, β represents the adjustment factor, and its value is greater than 0; C2_min represents the minimum value of the difference coefficient between the second monitoring feature vector Vf and the average feature vector Vavg of the standard warning event, and C2_max represents the maximum value of the difference coefficient between the second monitoring feature vector Vf and the average feature vector Vavg of the standard warning event.

[0028] A hydrogen-powered UAV thermal runaway warning threshold dynamic adjustment system, comprising: a data acquisition and mapping module, a warning event classification module, a trend analysis and feature extraction module, and a threshold dynamic adjustment module;

[0029] The data acquisition and mapping module obtains the thermal runaway warning event records and historical monitoring data of the hydrogen-powered UAV in the past period of time, matches the corresponding historical monitoring data based on the timestamp of the thermal runaway warning event records; analyzes the matched historical monitoring data and thermal runaway warning event records, and obtains the mapping relationship between the thermal runaway warning event records and the historical monitoring data;

[0030] The warning event classification module obtains the feedback information of each thermal runaway warning event based on the mapping relationship between the thermal runaway warning event record and the historical monitoring data, and classifies the thermal runaway warning events into the first warning event, the second warning event and the standard warning event;

[0031] The trend analysis and feature extraction module extracts the standard monitoring features, the first monitoring features and the second monitoring features corresponding to the standard warning events, the first warning events and the second warning events respectively according to the preset thresholds; analyzes the first monitoring features and the second monitoring features with the standard monitoring features respectively, and calculates the first difference coefficient and the second difference coefficient;

[0032] The threshold dynamic adjustment module obtains dynamic thermal runaway warning thresholds corresponding to the first warning event and the second warning event respectively based on the first difference coefficient and the second difference coefficient.

[0033] Further, the data collection and mapping module includes a data collection unit and a mapping relationship establishment unit;

[0034] The data acquisition unit obtains the thermal runaway warning event records and historical monitoring data of the hydrogen-powered UAV in the past period of time; the mapping relationship establishment unit matches the thermal runaway warning event with the historical monitoring data based on the matching of the timestamp of the thermal runaway warning event with the flight mission time period, and establishes a mapping relationship between the thermal runaway warning event and the historical monitoring data.

[0035] Further, the warning event classification module includes a feedback information analysis unit and an event classification unit;

[0036] The feedback information analysis unit obtains the label of the corresponding thermal runaway warning event according to the feedback information of the thermal runaway warning event; the event classification unit classifies the thermal runaway warning event according to the label of the thermal runaway warning event into the first warning event, the second warning event and the standard warning event.

[0037] Further, the trend analysis and feature extraction includes a trend analysis unit and a feature extraction unit;

[0038] The trend analysis unit performs trend analysis on the historical monitoring data to obtain the trend curve of each warning event; the feature extraction unit extracts the corresponding monitoring features based on the trend curve of the warning event, standardizes the extracted monitoring features, and forms a corresponding monitoring feature vector.

[0039] Further, the threshold dynamic adjustment module includes a difference coefficient calculation unit and a threshold dynamic adjustment unit;

[0040] The difference coefficient calculation unit calculates the difference coefficient between the characteristic vector of each warning event and the average characteristic vector of the standard warning event, thereby obtaining a first difference coefficient and a second difference coefficient; the threshold dynamic adjustment unit calculates the dynamic thermal runaway warning thresholds of the first warning event and the second warning event based on the first difference coefficient and the second difference coefficient, respectively.

[0041] Compared with the prior art, the beneficial effects of the present invention are as follows: by dynamically adjusting the warning threshold based on the mapping relationship between historical thermal runaway warning events and monitoring data, the present invention can adapt to changes in flight status and environmental conditions, avoiding the false alarm and missed alarm problems caused by the traditional static threshold method; the strategy of dynamically adjusting the threshold ensures that the system can more accurately predict the risk of thermal runaway in different flight environments, thereby improving the accuracy of the warning. The present invention divides thermal runaway warning events into first warning events, second warning events and standard warning events, and adjusts the thresholds by warning events of different categories to effectively distinguish different risk levels, thereby avoiding the problem that a single warning threshold in traditional technology cannot accurately identify different risk levels; this not only improves the flexibility of the system, but also allows different response measures to be taken in different emergency situations. The present invention makes the response of the warning system more sensitive and accurate by comprehensively considering the flight status, environmental changes and the analysis of historical data, thereby effectively reducing the false alarm and missed alarm caused by the fixed threshold setting; this is of great significance for improving the safety and flight reliability of drones. Through dynamic threshold adjustment and precise trend analysis, the present invention provides a more intelligent thermal runaway warning method, which can timely identify potential thermal runaway risks under different flight phases and environmental conditions, effectively prevent safety accidents caused by thermal runaway, and thus greatly improve the flight safety of UAVs. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0043] Figure 1 It is a module schematic diagram of a hydrogen-powered UAV thermal runaway warning threshold dynamic adjustment system of the present invention. DETAILED DESCRIPTION

[0044] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0045] See also Figure 1 , the present invention provides a technical solution:

[0046] A hydrogen-powered UAV thermal runaway warning threshold dynamic adjustment system, comprising: a data acquisition and mapping module, a warning event classification module, a trend analysis and feature extraction module, and a threshold dynamic adjustment module;

[0047] The data acquisition and mapping module obtains the thermal runaway warning event records and historical monitoring data of the hydrogen-powered UAV in the past period of time, matches the corresponding historical monitoring data based on the timestamp of the thermal runaway warning event records; analyzes the matched historical monitoring data and thermal runaway warning event records, and obtains the mapping relationship between the thermal runaway warning event records and the historical monitoring data;

[0048] The warning event classification module obtains the feedback information of each thermal runaway warning event based on the mapping relationship between the thermal runaway warning event record and the historical monitoring data, and classifies the thermal runaway warning events into the first warning event, the second warning event and the standard warning event;

[0049] The trend analysis and feature extraction module extracts the standard monitoring features, the first monitoring features and the second monitoring features corresponding to the standard warning events, the first warning events and the second warning events respectively according to the preset thresholds; analyzes the first monitoring features and the second monitoring features with the standard monitoring features respectively, and calculates the first difference coefficient and the second difference coefficient;

[0050] The threshold dynamic adjustment module obtains dynamic thermal runaway warning thresholds corresponding to the first warning event and the second warning event respectively based on the first difference coefficient and the second difference coefficient.

[0051] The data acquisition and mapping module includes a data acquisition unit and a mapping relationship establishment unit;

[0052] The data acquisition unit obtains the thermal runaway warning event records and historical monitoring data of the hydrogen-powered UAV in the past period of time; the mapping relationship establishment unit matches the thermal runaway warning event with the historical monitoring data based on the matching of the timestamp of the thermal runaway warning event with the flight mission time period, and establishes a mapping relationship between the thermal runaway warning event and the historical monitoring data.

[0053] The warning event classification module includes a feedback information analysis unit and an event classification unit;

[0054] The feedback information analysis unit obtains the label of the corresponding thermal runaway warning event according to the feedback information of the thermal runaway warning event; the event classification unit classifies the thermal runaway warning event according to the label of the thermal runaway warning event into the first warning event, the second warning event and the standard warning event.

[0055] Trend analysis and feature extraction include trend analysis unit and feature extraction unit;

[0056] The trend analysis unit performs trend analysis on the historical monitoring data to obtain the trend curve of each warning event; the feature extraction unit extracts the corresponding monitoring features based on the trend curve of the warning event, standardizes the extracted monitoring features, and forms a corresponding monitoring feature vector.

[0057] The threshold dynamic adjustment module includes a difference coefficient calculation unit and a threshold dynamic adjustment unit;

[0058] The difference coefficient calculation unit calculates the difference coefficient between the characteristic vector of each warning event and the average characteristic vector of the standard warning event, thereby obtaining a first difference coefficient and a second difference coefficient; the threshold dynamic adjustment unit calculates the dynamic thermal runaway warning thresholds of the first warning event and the second warning event based on the first difference coefficient and the second difference coefficient, respectively.

[0059] A method for dynamically adjusting a thermal runaway warning threshold of a hydrogen-powered UAV comprises the following steps:

[0060] Step S100. Obtain the thermal runaway warning event records and historical monitoring data of the hydrogen-powered UAV in the past period of time, match the corresponding historical monitoring data based on the timestamp of the thermal runaway warning event records; analyze the matched historical monitoring data and thermal runaway warning event records, so as to obtain the mapping relationship between the thermal runaway warning event records and the historical monitoring data;

[0061] Step S200. Based on the mapping relationship between the thermal runaway warning event record and the historical monitoring data, feedback information of each thermal runaway warning event is obtained, and the thermal runaway warning events are classified into first warning events, second warning events and standard warning events; for the standard warning events, the standard monitoring features corresponding to the standard warning events are extracted in combination with the preset thresholds;

[0062] Step S300. For the first warning event, in combination with the preset threshold, extract the first monitoring feature corresponding to the first warning event; analyze the first monitoring feature and the standard monitoring feature, and calculate the first difference coefficient between the two; and obtain the dynamic thermal runaway warning threshold of the first warning event based on the first difference coefficient between the two;

[0063] Step S400. For the second warning event, in combination with the preset threshold, extract the second monitoring feature corresponding to the second warning event; analyze the second monitoring feature with the standard monitoring feature to calculate the second difference coefficient between the two; and based on the second difference coefficient between the two, obtain the dynamic thermal runaway warning threshold of the second warning event.

[0064] Step S100 includes:

[0065] S101. Obtain the thermal runaway warning event record E and historical monitoring data D of the hydrogen-powered UAV in the past period of time, which are expressed as: E={e1,e2,...,en}, D={d1,d2,...,dm}, where e1 represents the first thermal runaway warning event of the hydrogen-powered UAV in the past period of time, e2 represents the second thermal runaway warning event of the hydrogen-powered UAV in the past period of time, and so on. en represents the nth thermal runaway warning event of the hydrogen-powered UAV in the past period of time, and n represents the total number of events in the thermal runaway warning event record E; d1 represents the number of thermal runaway warning events of the hydrogen-powered UAV in the past period of time. d1 represents the historical monitoring data corresponding to the first flight mission of the hydrogen-powered UAV in the past period of time, d2 represents the historical monitoring data corresponding to the second flight mission of the hydrogen-powered UAV in the past period of time, and so on, dm represents the historical monitoring data corresponding to the mth flight mission of the hydrogen-powered UAV in the past period of time, and m represents the total number of flight missions of the hydrogen-powered UAV in the past period of time; among them, the thermal runaway warning event record contains warning event information related to thermal runaway of the hydrogen-powered UAV, such as the timestamp, type and description of the event; the historical monitoring data records the monitoring data during the hydrogen-powered UAV flight mission, including various flight parameters and environmental data;

[0066] S102. Obtain the timestamp ti of each thermal runaway warning event ei in the thermal runaway warning event record E, and the flight mission time period Tj corresponding to the historical monitoring data dj, and match the thermal runaway warning event ei that satisfies ti∈Tj with the corresponding historical monitoring data dj; summarize the matching results between all thermal runaway warning events and historical monitoring data, establish a mapping relationship between the thermal runaway warning event records and the historical monitoring data, and form a corresponding mapping relationship table, and the structure of each mapping relationship in the mapping relationship table is: N(ei)→ei→dj, where N(ei) represents the index of the thermal runaway warning event ei in the mapping relationship table.

[0067] Step S200 includes:

[0068] S201. For each mapping relationship between the thermal runaway warning event record and the historical monitoring data in the mapping relationship table, the feedback information of the thermal runaway warning event is extracted from the thermal runaway warning event record, and the label of the corresponding thermal runaway warning event is obtained according to the feedback information of the thermal runaway warning event, wherein the labels are F, T and B respectively, and the label F indicates that the monitoring parameters in the historical monitoring data of the corresponding hydrogen-powered UAV exceed the preset threshold value, and the thermal runaway warning information is received. According to the analysis of relevant personnel, there is no thermal runaway risk; the label T indicates that the monitoring parameters in the historical monitoring data of the corresponding hydrogen-powered UAV do not exceed the preset threshold value, and the thermal runaway warning information is not received. According to the analysis of relevant personnel, there is a thermal runaway risk; the label B indicates that the monitoring parameters in the historical monitoring data of the corresponding hydrogen-powered UAV exceed the preset threshold value, and the thermal runaway warning information is received. According to the analysis of relevant personnel, there is a thermal runaway risk; according to the label of the thermal runaway warning event, the thermal runaway warning event corresponding to the label F is marked as the first warning event, the thermal runaway warning event corresponding to the label T is marked as the second warning event, and the thermal runaway warning event corresponding to the label B is marked as the standard warning event;

[0069] S202. For each standard warning event eb, obtain the index of the corresponding mapping relationship table, obtain the corresponding thermal runaway warning event record and historical monitoring data in the mapping relationship table, extract the thermal runaway warning trigger timestamp tb corresponding to the standard warning event eb based on the thermal runaway warning event record, and construct the analysis time window Tb with the thermal runaway warning trigger timestamp tb as the reference point, and Tb=[tb-Δtpr,t+Δtpo], wherein Δtpr represents the warning traceback time, and the statistical mode of the time when the parameters of all standard warning events before the triggering first break through the preset threshold is taken; Δtpo represents the warning delay time, and the statistical mode of the stable time after the risk of all standard warning events is confirmed is taken; according to the analysis time window Tb, obtain the historical monitoring data of each standard warning event eb within the analysis time window Tb, perform trend analysis on the historical monitoring data within the analysis time window Tb, thereby obtaining the corresponding trend curve Qb, and extract the standard monitoring features of the corresponding standard warning event from the trend curve Qb, and perform standardization on the standard monitoring features, thereby obtaining the standard monitoring feature vector Vb.

[0070] In this embodiment, the analysis process of the standard monitoring feature vector Vb is as follows:

[0071] Assume that within the time window Tw, all historical monitoring data related to the standard warning event are collected; assuming that the time window is 30 minutes (Δtpr=10 minutes, Δtpo=20 minutes), the temperature, pressure, current and other monitoring data within this time period will be extracted. Perform trend analysis on the collected monitoring data. Trend analysis can use methods such as linear regression and time series analysis to obtain the trend curve Qb; extract standard monitoring features from the trend curve, such as the temperature rise rate and the current change rate. Standardize the extracted standard monitoring features so that they have the same influence at different scales; for example, use methods such as Z-score standardization or minimum-maximum normalization to map the feature values ​​to a unified standard range.

[0072] Assuming that within the time window Tw, the monitored temperature data shows a steady upward trend and the current data shows fluctuations, after trend analysis and standardization, the following standard monitoring feature vector is obtained:

[0073] Vb=[v1,v2,v3,v4]=[0.85,0.92,0.75,0.88]; where: v1 represents the trend characteristics of temperature, v2 represents the current fluctuation rate, v3 represents the pressure change rate, and v4 represents the stability of certain key parameters.

[0074] Step S300 includes:

[0075] S301. For each first warning event ef, refer to the analysis process of the historical monitoring data of the standard warning event eb, obtain the corresponding thermal runaway warning trigger timestamp tf and analysis time window Tf in turn, extract the corresponding historical monitoring data according to the analysis time window Tf, perform trend analysis on the historical monitoring data within the analysis time window Tf, thereby obtaining the corresponding trend curve Qf, and extract the first monitoring feature corresponding to the first warning event from the trend curve Qf, and perform standardization on the first monitoring feature, thereby obtaining the first monitoring feature vector Vf;

[0076] S302. Summarize the standard monitoring feature vectors Vb of all standard warning events, calculate the average feature vector Vavg of the standard warning events; calculate the difference coefficient between the first monitoring feature vector Vf of each first warning event ef and the average feature vector Vavg of the standard warning event, and the corresponding calculation formula is: C(Vf,Vavg)=[∑K k=1wk·(fk-avg_k) 2 ] (1 / 2); Wherein C(Vf,Vavg) represents the difference coefficient between the first monitoring feature vector Vf and the average feature vector Vavg of the standard warning event, fk represents the feature value of the kth dimension in the first monitoring feature vector Vf, avg_k represents the feature value of the kth dimension in the average feature vector Vavg, wk represents the weight of the kth feature dimension, and the weight can be determined based on historical data, expert evaluation or machine learning model; K represents the dimension of the first monitoring feature vector Vf and the average feature vector Vavg;

[0077] S303. Obtain the preset threshold, and calculate the dynamic thermal runaway warning threshold of the first warning event in combination with the difference coefficient C(Vf, Vavg) corresponding to each first warning event ef. The corresponding calculation formula is:

[0078] Rf=R0×[1+α×(C(Vf,Vavg)-C1_min) / (C1_max-C1_min)];

[0079] Among them, Rf represents the dynamic thermal runaway warning threshold of the first warning event, R0 represents the preset threshold, α represents the adjustment factor, and the value is greater than 0; C1_min represents the minimum value of the difference coefficient between the first monitoring feature vector Vf and the average feature vector Vavg of the standard warning event, and C1_max represents the maximum value of the difference coefficient between the first monitoring feature vector Vf and the average feature vector Vavg of the standard warning event.

[0080] Step S400 includes:

[0081] S401. For each second warning event et, refer to the analysis process of the first warning event ef to analyze, so as to obtain the corresponding trend curve Qt, and extract the second monitoring feature corresponding to the second warning event from the trend curve Qt; and standardize the second monitoring feature to obtain the second monitoring feature vector Vt; refer to the calculation formula of the difference coefficient between the first monitoring feature vector Vf of the first warning event ef and the average feature vector Vavg of the standard warning event, and calculate the difference coefficient C(Vt,Vavg) between the second monitoring feature vector Vt of each second warning event et and the average feature vector Vavg of the standard warning event;

[0082] S402. Obtain a preset threshold value, and calculate the dynamic thermal runaway warning threshold value of the second warning event in combination with the difference coefficient C(Vt, Vavg) corresponding to each second warning event et. The corresponding calculation formula is:

[0083] Rt=R0×[1-β×(C(Vt,Vavg)-C2_min) / (C2_max-C2_min)];

[0084] Among them, Rt represents the dynamic thermal runaway warning threshold of the second warning event, β represents the adjustment factor, and its value is greater than 0; C2_min represents the minimum value of the difference coefficient between the second monitoring feature vector Vf and the average feature vector Vavg of the standard warning event, and C2_max represents the maximum value of the difference coefficient between the second monitoring feature vector Vf and the average feature vector Vavg of the standard warning event.

[0085] In this embodiment, real-time monitoring data of the hydrogen-powered UAV triggering a preset threshold is obtained, and trend analysis is performed on the real-time monitoring data to obtain a trend curve of the real-time monitoring data, and the corresponding real-time trend features are extracted to obtain a real-time trend feature vector Vs; the real-time trend feature vector Vs is calculated for similarity with the monitoring feature vectors of the first warning event, the standard warning event, and the second warning event, and the event with the greatest similarity is selected as the category of the current warning event. Assuming that the event with the greatest similarity is the first warning event, the corresponding dynamic thermal runaway warning threshold Rf is obtained, and the real-time monitoring data is further compared. If the real-time monitoring data is less than the dynamic thermal runaway warning threshold Rf, notification information that the current event is the first monitoring event is output; if the real-time monitoring data is greater than or equal to the dynamic thermal runaway warning threshold Rf, the corresponding notification information is output to relevant personnel, who will perform further processing.

[0086] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0087] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein by equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for dynamically adjusting the thermal runaway warning threshold of a hydrogen-powered UAV, characterized in that: The method comprises the following steps: Step S100. Obtain the thermal runaway warning event records and historical monitoring data of the hydrogen-powered UAV in the past period of time, match the corresponding historical monitoring data based on the timestamp of the thermal runaway warning event records; analyze the matched historical monitoring data and thermal runaway warning event records, so as to obtain the mapping relationship between the thermal runaway warning event records and the historical monitoring data; Step S200. Based on the mapping relationship between the thermal runaway warning event record and the historical monitoring data, feedback information of each thermal runaway warning event is obtained, and the thermal runaway warning events are classified into first warning events, second warning events and standard warning events; for the standard warning events, the standard monitoring features corresponding to the standard warning events are extracted in combination with the preset thresholds; Step S300. For the first warning event, in combination with the preset threshold, extract the first monitoring feature corresponding to the first warning event; analyze the first monitoring feature and the standard monitoring feature, and calculate the first difference coefficient between the two; and obtain the dynamic thermal runaway warning threshold of the first warning event based on the first difference coefficient between the two; Step S400. For the second warning event, in combination with the preset threshold, extract the second monitoring feature corresponding to the second warning event; analyze the second monitoring feature with the standard monitoring feature to calculate the second difference coefficient between the two; and based on the second difference coefficient between the two, obtain the dynamic thermal runaway warning threshold of the second warning event.

2. A method for dynamically adjusting the thermal runaway warning threshold of a hydrogen-powered UAV according to claim 1, characterized in that: The step S100 includes: S101. Obtain the thermal runaway warning event record E and historical monitoring data D of the hydrogen-powered UAV in the past period of time, which are expressed as: E={e1,e2,...,en}, D={d1,d2,...,dm}, where e1 represents the first thermal runaway warning event of the hydrogen-powered UAV in the past period of time, e2 represents the second thermal runaway warning event of the hydrogen-powered UAV in the past period of time, and so on. en represents the nth thermal runaway warning event of the hydrogen-powered UAV in the past period of time, and n represents the total number of events in the thermal runaway warning event record E; d1 represents the number of thermal runaway warning events of the hydrogen-powered UAV in the past period of time. d1 represents the historical monitoring data corresponding to the first flight mission of the hydrogen-powered UAV in the past period of time, d2 represents the historical monitoring data corresponding to the second flight mission of the hydrogen-powered UAV in the past period of time, and so on, dm represents the historical monitoring data corresponding to the mth flight mission of the hydrogen-powered UAV in the past period of time, and m represents the total number of flight missions of the hydrogen-powered UAV in the past period of time; among them, the thermal runaway warning event record contains warning event information related to thermal runaway of the hydrogen-powered UAV, such as the timestamp, type and description of the event; the historical monitoring data records the monitoring data during the hydrogen-powered UAV flight mission, including various flight parameters and environmental data; S102. Obtain the timestamp ti of each thermal runaway warning event ei in the thermal runaway warning event record E, and the flight mission time period Tj corresponding to the historical monitoring data dj, and match the thermal runaway warning event ei that satisfies ti∈Tj with the corresponding historical monitoring data dj; summarize the matching results between all thermal runaway warning events and historical monitoring data, establish a mapping relationship between the thermal runaway warning event records and the historical monitoring data, and form a corresponding mapping relationship table, and the structure of each mapping relationship in the mapping relationship table is: N(ei)→ei→dj, where N(ei) represents the index of the thermal runaway warning event ei in the mapping relationship table.

3. The method for dynamically adjusting the thermal runaway warning threshold of a hydrogen-powered UAV according to claim 2 is characterized in that: The step S200 includes: S201. For each mapping relationship between the thermal runaway warning event record and the historical monitoring data in the mapping relationship table, the feedback information of the thermal runaway warning event is extracted from the thermal runaway warning event record, and the label of the corresponding thermal runaway warning event is obtained according to the feedback information of the thermal runaway warning event, wherein the labels are F, T and B respectively, and the label F indicates that the monitoring parameters in the historical monitoring data of the corresponding hydrogen-powered UAV exceed the preset threshold value, and the thermal runaway warning information is received. According to the analysis of relevant personnel, there is no thermal runaway risk; the label T indicates that the monitoring parameters in the historical monitoring data of the corresponding hydrogen-powered UAV do not exceed the preset threshold value, and the thermal runaway warning information is not received. According to the analysis of relevant personnel, there is a thermal runaway risk; the label B indicates that the monitoring parameters in the historical monitoring data of the corresponding hydrogen-powered UAV exceed the preset threshold value, and the thermal runaway warning information is received. According to the analysis of relevant personnel, there is a thermal runaway risk; according to the label of the thermal runaway warning event, the thermal runaway warning event corresponding to the label F is marked as the first warning event, the thermal runaway warning event corresponding to the label T is marked as the second warning event, and the thermal runaway warning event corresponding to the label B is marked as the standard warning event; S202. For each standard warning event eb, obtain the index of the corresponding mapping relationship table, obtain the corresponding thermal runaway warning event record and historical monitoring data in the mapping relationship table, extract the thermal runaway warning trigger timestamp tb corresponding to the standard warning event eb based on the thermal runaway warning event record, and construct the analysis time window Tb with the thermal runaway warning trigger timestamp tb as the reference point, and Tb=[tb-Δtpr,t+Δtpo], wherein Δtpr represents the warning traceback time, and the statistical mode of the time when the parameters of all standard warning events before the triggering first break through the preset threshold is taken; Δtpo represents the warning delay time, and the statistical mode of the stable time after the risk of all standard warning events is confirmed is taken; according to the analysis time window Tb, obtain the historical monitoring data of each standard warning event eb within the analysis time window Tb, perform trend analysis on the historical monitoring data within the analysis time window Tb, thereby obtaining the corresponding trend curve Qb, and extract the standard monitoring features of the corresponding standard warning event from the trend curve Qb, and perform standardization on the standard monitoring features, thereby obtaining the standard monitoring feature vector Vb.

4. A method for dynamically adjusting the thermal runaway warning threshold of a hydrogen-powered UAV according to claim 3, characterized in that: The step S300 includes: S301. For each first warning event ef, refer to the analysis process of the historical monitoring data of the standard warning event eb, obtain the corresponding thermal runaway warning trigger timestamp tf and analysis time window Tf in turn, extract the corresponding historical monitoring data according to the analysis time window Tf, perform trend analysis on the historical monitoring data within the analysis time window Tf, thereby obtaining the corresponding trend curve Qf, and extract the first monitoring feature corresponding to the first warning event from the trend curve Qf, and perform standardization on the first monitoring feature, thereby obtaining the first monitoring feature vector Vf; S302. Summarize the standard monitoring feature vectors Vb of all standard warning events, and calculate the average feature vector Vavg of the standard warning events; calculate the difference coefficient between the first monitoring feature vector Vf of each first warning event ef and the average feature vector Vavg of the standard warning event, and the corresponding calculation formula is: C(Vf,Vavg)=[∑K k=1wk·(fk-avg_k)2](1 / 2); wherein C(Vf,Vavg) represents the difference coefficient between the first monitoring feature vector Vf and the average feature vector Vavg of the standard warning event, fk represents the feature value of the kth dimension in the first monitoring feature vector Vf, avg_k represents the feature value of the kth dimension in the average feature vector Vavg, and wk represents the weight of the kth feature dimension, which can be determined based on historical data, expert evaluation or machine learning model; K represents the dimension of the first monitoring feature vector Vf and the average feature vector Vavg; S303. Obtain the preset threshold, and calculate the dynamic thermal runaway warning threshold of the first warning event in combination with the difference coefficient C(Vf, Vavg) corresponding to each first warning event ef. The corresponding calculation formula is: Rf=R0×[1+α×(C(Vf,Vavg)-C1_min) / (C1_max-C1_min)]; Among them, Rf represents the dynamic thermal runaway warning threshold of the first warning event, R0 represents the preset threshold, α represents the adjustment factor, and the value is greater than 0; C1_min represents the minimum value of the difference coefficient between the first monitoring feature vector Vf and the average feature vector Vavg of the standard warning event, and C1_max represents the maximum value of the difference coefficient between the first monitoring feature vector Vf and the average feature vector Vavg of the standard warning event.

5. A method for dynamically adjusting the thermal runaway warning threshold of a hydrogen-powered UAV according to claim 4, characterized in that: The step S400 includes: S401. For each second warning event et, refer to the analysis process of the first warning event ef to analyze, so as to obtain the corresponding trend curve Qt, and extract the second monitoring feature corresponding to the second warning event from the trend curve Qt; and standardize the second monitoring feature to obtain the second monitoring feature vector Vt; refer to the calculation formula of the difference coefficient between the first monitoring feature vector Vf of the first warning event ef and the average feature vector Vavg of the standard warning event, and calculate the difference coefficient C(Vt,Vavg) between the second monitoring feature vector Vt of each second warning event et and the average feature vector Vavg of the standard warning event; S402. Obtain a preset threshold value, and calculate the dynamic thermal runaway warning threshold value of the second warning event in combination with the difference coefficient C(Vt, Vavg) corresponding to each second warning event et. The corresponding calculation formula is: Rt=R0×[1-β×(C(Vt,Vavg)-C2_min) / (C2_max-C2_min)]; Among them, Rt represents the dynamic thermal runaway warning threshold of the second warning event, β represents the adjustment factor, and its value is greater than 0; C2_min represents the minimum value of the difference coefficient between the second monitoring feature vector Vf and the average feature vector Vavg of the standard warning event, and C2_max represents the maximum value of the difference coefficient between the second monitoring feature vector Vf and the average feature vector Vavg of the standard warning event.

6. A hydrogen-powered UAV thermal runaway warning threshold dynamic adjustment system, applied to a hydrogen-powered UAV thermal runaway warning threshold dynamic adjustment method according to any one of claims 1 to 5, characterized in that: The system includes: a data collection and mapping module, a warning event classification module, a trend analysis and feature extraction module, and a threshold dynamic adjustment module; The data acquisition and mapping module obtains the thermal runaway warning event records and historical monitoring data of the hydrogen-powered UAV in the past period of time, matches the corresponding historical monitoring data based on the timestamp of the thermal runaway warning event records; analyzes the matched historical monitoring data and thermal runaway warning event records, thereby obtaining a mapping relationship between the thermal runaway warning event records and the historical monitoring data; The warning event classification module obtains feedback information of each thermal runaway warning event based on the mapping relationship between the thermal runaway warning event record and the historical monitoring data, and classifies the thermal runaway warning events into first warning events, second warning events and standard warning events; The trend analysis and feature extraction module extracts the standard monitoring features, the first monitoring features and the second monitoring features corresponding to the standard warning events, the first warning events and the second warning events respectively according to the preset thresholds; analyzes the first monitoring features and the second monitoring features with the standard monitoring features respectively, and calculates the first difference coefficient and the second difference coefficient; The threshold dynamic adjustment module obtains dynamic thermal runaway warning thresholds corresponding to the first warning event and the second warning event respectively based on the first difference coefficient and the second difference coefficient.

7. A hydrogen-powered UAV thermal runaway warning threshold dynamic adjustment system according to claim 6, characterized in that: The data acquisition and mapping module includes a data acquisition unit and a mapping relationship establishment unit; The data acquisition unit obtains the thermal runaway warning event records and historical monitoring data of the hydrogen-powered UAV in the past period of time; the mapping relationship establishment unit matches the thermal runaway warning event with the historical monitoring data based on the matching of the timestamp of the thermal runaway warning event with the flight mission time period, and establishes a mapping relationship between the thermal runaway warning event and the historical monitoring data.

8. The hydrogen-powered UAV thermal runaway warning threshold dynamic adjustment system according to claim 6 is characterized by: The warning event classification module includes a feedback information analysis unit and an event classification unit; The feedback information analysis unit obtains a label of a corresponding thermal runaway warning event according to the feedback information of the thermal runaway warning event; The event classification unit classifies the thermal runaway warning events according to the labels of the thermal runaway warning events into first warning events, second warning events and standard warning events.

9. A hydrogen-powered UAV thermal runaway warning threshold dynamic adjustment system according to claim 6, characterized in that: The trend analysis and feature extraction comprises a trend analysis unit and a feature extraction unit; The trend analysis unit performs trend analysis on the historical monitoring data to obtain a trend curve for each warning event; The feature extraction unit extracts corresponding monitoring features based on the trend curve of the warning event, and standardizes the extracted monitoring features to form a corresponding monitoring feature vector.

10. The hydrogen-powered UAV thermal runaway warning threshold dynamic adjustment system according to claim 6 is characterized by: The threshold dynamic adjustment module includes a difference coefficient calculation unit and a threshold dynamic adjustment unit; The difference coefficient calculation unit calculates the difference coefficient between the characteristic vector of each warning event and the average characteristic vector of the standard warning event, thereby obtaining a first difference coefficient and a second difference coefficient; the threshold dynamic adjustment unit calculates the dynamic thermal runaway warning thresholds of the first warning event and the second warning event respectively based on the first difference coefficient and the second difference coefficient.

Images