Hybrid power system energy management method during VTOL unmanned aerial vehicle dual-mode switching

Abstract

The invention discloses a hybrid power system energy management method during VTOL unmanned aerial vehicle dual-mode switching, and the method comprises the steps: S1, dual-mode switching prediction parameters are collected, the dual-mode switching prediction parameters are obtained from a flight control system in real time, and the dual-mode switching prediction parameters comprise a mode switching instruction, a pitch angle change rate and a height change rate; s2, pre-judging a flight switching mode: pre-judging the flight switching mode of the VTOL unmanned aerial vehicle based on the dual-mode switching prediction parameters, the flight switching mode including vertical horizontal switching and horizontal vertical switching; and S3, energy strategy adjustment: triggering a corresponding energy adjustment strategy 5-10 seconds in advance based on the flight switching mode, and dynamically adjusting the power output of the hydrogen fuel cell and the sodium cell to realize progressive power handover. Energy distribution is adjusted in advance through flight trend pre-judgment, the control belongs to feed-forward-feedback composite control, and efficient cooperation of the hydrogen fuel cell and the sodium cell of the hybrid power system is achieved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of hybrid unmanned aerial vehicle energy management, in particular to a hybrid system energy management method for VTOL unmanned aerial vehicle dual-mode switching. BACKGROUND

[0002] The vertical take-off and landing fixed-wing unmanned aerial vehicle (VTOL) combines the vertical take-off and landing capability of multi-rotor and the long endurance advantage of fixed-wing, and has become an important platform in the fields of industrial inspection and logistics transportation. To improve the endurance mileage, the hybrid power system combining hydrogen fuel cells and high-power sodium batteries has become the mainstream solution. However, when the VTOL unmanned aerial vehicle switches between vertical mode and horizontal cruising mode, the power demand of the whole machine will change dramatically by more than 5 times, which puts high requirements on the dynamic response and coordinated control capability of the energy management system.

[0003] In the prior art, there are some schemes about energy management of hybrid unmanned aerial vehicles. For example, the patent with publication number CN120134964A proposes an energy management system based on hydrogen fuel cells and lithium batteries. The lithium battery provides instantaneous power supply during the take-off stage of the unmanned aerial vehicle. When there is no high power demand, the hydrogen fuel cell is used as the power supply device. In the hybrid power supply mode, the unmanned aerial vehicle adopts a power supply mode mainly using the hydrogen fuel cell and supplemented by the lithium battery. This scheme provides instantaneous power compensation through the lithium battery when the power demand suddenly increases, and alleviates the insufficient dynamic response of the fuel cell. However, this scheme still relies on pre-set power distribution rules and cannot be optimized in real time according to the flight task and energy state.

[0004] For example, the patent with publication number CN119176269A implements certain energy management. If the second electric quantity of the hydrogen fuel cell is less than the required power for the unmanned aerial vehicle to fly, the power supply quantity of the hydrogen fuel cell is increased or the lithium battery is started for hybrid power supply. If the second electric quantity of the hydrogen fuel cell is still less than the required power for the unmanned aerial vehicle to fly after increasing the power supply quantity of the hydrogen fuel cell, the lithium battery is started for hybrid power supply. If the second electric quantity of the hydrogen fuel cell is greater than the required power for the unmanned aerial vehicle to fly, the hydrogen fuel cell is used for single power supply. The essence of this scheme is still passive response, which relies on the feedback signal of the current state and has inherent control delay. It cannot solve the problem of pre-judgment and preparation before mode switching, so there will still be a significant power gap or voltage oscillation when dealing with severe transient working conditions.

[0005] Further analysis shows that during the flight mode switching process of the VTOL unmanned aerial vehicle, the energy management system not only needs to achieve fast response, but also needs to consider the working efficiency of the hydrogen fuel cell, the life loss of the sodium battery, and the flight safety boundary, which requires the control system to have multi-objective collaborative optimization capability.

[0006] In summary, the currently disclosed technical solutions are mainly based on rule-based feedback control or simple power source switching. In the face of such power mutations, there are usually three major problems of response lag, current impact and safety risk. Specifically, as a slow response system, the power output of the hydrogen fuel cell cannot quickly match the instantaneous changes of the load, resulting in system voltage fluctuations; the sodium battery is forced to discharge or charge at a large current during mode switching, seriously affecting its cycle life; if the energy distribution is improper, it may even cause power shortage during switching, leading to a crash. SUMMARY

[0007] To solve the problems in the prior art, the present application provides a hybrid system energy management method for VTOL unmanned aerial vehicle dual-mode switching, which is based on the energy management method of "flight mode pre-identification" and "gradual power handover", realizing fast response, smooth transition, safety and stability, and efficiency optimization.

[0008] The technical solution adopted by the present application to solve its technical problems is as follows:

[0009] A hybrid system energy management method for VTOL unmanned aerial vehicle dual-mode switching, comprising:

[0010] S1: Dual-mode switching prediction parameter acquisition, real-time acquisition of dual-mode switching prediction parameters from the flight control system, the dual-mode switching prediction parameters including mode switching instructions, pitch angle change rate and height change rate;

[0011] S2: Flight switching mode pre-judgment, pre-judgment of the flight switching mode of the VTOL unmanned aerial vehicle based on the dual-mode switching prediction parameters, the flight switching mode including vertical to horizontal switching and horizontal to vertical switching;

[0012] S3: Energy strategy adjustment, triggering the corresponding energy adjustment strategy based on the flight switching mode 5-10 seconds in advance, dynamically adjusting the power output of the hydrogen fuel cell and the sodium battery, and realizing gradual power handover;

[0013] The energy adjustment strategy includes: in vertical to horizontal switching, reducing the output power of the hydrogen fuel cell and gradually increasing the energy storage of the sodium battery; in horizontal to vertical switching, controlling the output power of the hydrogen fuel cell to rise to a high-power standby state, while dynamically adjusting the SOC of the sodium battery to the optimal output interval.

[0014] Further, the S2 uses a multi-threshold triggering mechanism to pre-judge the flight switching mode of the VTOL unmanned aerial vehicle;

[0015] When the mode switching instruction is detected as "preparing to hover", or the pitch angle rate is greater than 8° / s, or the height rate suddenly changes from zero to negative, it is determined that the VTOL UAV is about to enter the vertical take-off and landing mode, and the horizontal-to-vertical switching is determined.

[0016] When the mode switching instruction is detected as "preparing to hover", or the pitch angle rate is greater than 8° / s, or the height rate suddenly changes from zero to negative, it is determined that the VTOL UAV is about to enter the vertical take-off and landing mode, and the horizontal-to-vertical switching is determined.

[0017] Further, the S2 calculates a mode switching confidence through an information fusion algorithm, and the mode switching confidence is used to quantify the probability of the VTOL UAV switching from the current flight mode to the target flight mode;

[0018] The mode switching confidence includes a first confidence of horizontal-to-vertical switching and a second confidence of vertical-to-horizontal switching; wherein the first confidence is calculated by a weighted sum model, and the second confidence is calculated by a product model;

[0019] When the first confidence exceeds a first trigger threshold, the horizontal-to-vertical switching is determined, and when the second confidence exceeds a second trigger threshold, the vertical-to-horizontal switching is determined; the first trigger threshold is less than the second trigger threshold.

[0020] Further, the calculation formula of the first confidence is:

[0021]

[0022] Wherein, C1 is the first confidence; c cmd is the mode switching instruction; R pitch is the pitch angle rate; R high is the height rate; f 11 (c cmd ) is a first instruction sub-confidence calculated based on the mode switching instruction; f 12 (R pitch ) is a first pitch angle sub-confidence calculated based on the pitch angle rate; f 13 (R high ) is a first height sub-confidence calculated based on the height rate; w1, w2, w3 are weight coefficients corresponding to the first instruction sub-confidence, the first pitch angle sub-confidence and the first height sub-confidence respectively, and w1+w2+w3=1, and w1>w2>w3.

[0023] Further, when the mode switching instruction is detected as "preparing to hover", f 11 (c cmd) = 1.0, when the mode switching instruction is not detected as "preparing for hovering", then f 11 (c cmd ) = 0.0.

[0024] When the pitch angle rate of change is less than 5° / s, then f 12 (R pitch ) = 0.0, when the pitch angle rate of change is between 5~8° / s, then the first pitch angle sub-confidence is linearly mapped between 0.0 and 1.0; when the pitch angle rate of change is greater than 8° / s, then f 12 (R pitch ) = 1.0.

[0025] When the height rate of change is greater than -0.2m / s, then f 13 (R high ) = 0.0, when the height rate of change is between -0.2~-1.0m / s, then the first height sub-confidence is linearly mapped between 0.0 and 1.0; when the height rate of change is less than -1.0m / s, then f 13 (R high ) = 1.0.

[0026] Further, the calculation formula of the second confidence is:

[0027]

[0028] Wherein, C2 is the second confidence; c cmd is the mode switching instruction; R pitch is the pitch angle rate of change; R high is the height rate of change; f 21 (c cmd ) is the second instruction sub-confidence calculated based on the mode switching instruction; f 22 (|R pitch |) is the second pitch angle sub-confidence calculated based on the pitch angle rate of change absolute value, which decreases with the increase of the pitch angle rate of change absolute value; f 23 (|R high |) is the second height sub-confidence calculated based on the height rate of change absolute value, which decreases with the increase of the height rate of change absolute value.

[0029] Further, when the mode switching instruction is detected as "preparing for level flight", then f 21 (c cmd ) = 1.0, when the mode switching instruction is not detected as "preparing for level flight", then f 21 (c cmd ) = 0.8.

[0030] When the pitch angle rate of change absolute value is less than 3° / s, then f22 (|R pitch |) = 1.0, when the pitch angle rate of change is between 3-5° / s, the second pitch angle sub-confidence is linearly mapped between 1.0 and 0; when the pitch angle rate of change is greater than 5° / s, f 22 (|R pitch |) = 0.0.

[0031] when the height rate of change is less than 0.3m / s, f 23 (|R high |) = 1.0, when the height rate of change is between 0.3-0.8m / s, the second height sub-confidence is linearly mapped between 1.0 and 0; when the height rate of change is greater than 0.8m / s, f 23 (|R high |) = 0.0.

[0032] Further, in the S3: when switching from vertical to horizontal, trigger the corresponding energy adjustment strategy 5 seconds in advance; when switching from horizontal to vertical, trigger the corresponding energy adjustment strategy 10 seconds in advance.

[0033] Further, in the S3:

[0034] when switching from vertical to horizontal, set the first fixed time window to 3 seconds, complete the vertical to horizontal power handover in the first fixed time window, the total power demand of the VTOL unmanned aerial vehicle during the vertical to horizontal power handover is borne by the hydrogen fuel cell and the sodium battery, the fuel cell power decreases at a predetermined slope, and the power gap generated is immediately dynamically made up by the sodium battery increasing the discharge power;

[0035] when switching from horizontal to vertical, set the second fixed time window to 5 seconds, complete the horizontal to vertical power reserve in the second fixed time window, the hydrogen fuel cell output power is increased to a high power standby state at a safety slope, and the sodium battery SOC is adjusted to the interval of 60%-80%.

[0036] Further, the specific steps of the sodium battery increasing the discharge power are:

[0037] based on the difference between the total power demand and the actual output power of the hydrogen fuel cell, and corrected through closed-loop feedback of the direct current bus voltage, calculate the discharge power of the sodium battery;

[0038] The calculation formula of the actual output power of the hydrogen fuel cell is:

[0039] wherein, P fc-actual is the actual output power of the hydrogen fuel cell, P fc-initialThe initial output power of the hydrogen fuel cell at the moment of triggering the energy adjustment strategy, i.e., t=0; k is a predetermined slope, the predetermined slope is not greater than 10% / s, and t is time, the time elapsed since triggering the energy adjustment strategy;

[0040] The calculation formula of the discharge power of the sodium battery is:

[0041] Wherein, P batt-ref is the discharge power of the sodium battery, P tatal is the total power demand, PID() is a proportional-integral-derivative correction function, Vc is a DC bus voltage reference value, and Va is a DC bus voltage actual value.

[0042] Further, S4: dynamic correction is further included.

[0043] Real-time monitoring of dynamic correction parameters, the dynamic correction parameters include: DC bus voltage fluctuation rate, sodium battery SOC, attitude angle deviation, and the predetermined slope and the first fixed time window are adjusted according to the dynamic correction parameters;

[0044] When the DC bus voltage change rate is greater than 5% / s, the predetermined slope is reduced, and the first fixed time window is extended;

[0045] When the sodium battery SOC is less than 50%, the predetermined slope is reduced, the fuel cell power reduction time is shortened, and the sodium battery is prevented from overdischarging;

[0046] When the attitude angle deviation is greater than 5°, return to step S1.

[0047] The beneficial effects of the present application are:

[0048] The traditional rule-based energy management system is a kind of after-the-fact remedial feedback control, which only adjusts after detecting power imbalance or voltage fluctuation, and has inherent control delay. The present application is based on mode switching recognition of multi-source information (switching instruction, attitude angle change rate, height change rate) fusion, which can recognize the intention within 5-10 seconds before the actual mode switching occurs, and trigger the energy scheduling strategy in advance. Through the pre-judgment mechanism, the response speed is improved by more than 40%, so that the unmanned aerial vehicle can establish a new power balance faster, the synchronization of flight action and power output is greatly enhanced, and the advantage is obvious especially in high maneuverability tasks.

[0049] The power change during mode switching is the main factor of the power source life loss. The present application sets a safe slope limit of ≤10% / s for the power change of the hydrogen fuel cell through the "progressive power handover" strategy, so that the internal electrochemical reaction and auxiliary system (such as air compressor) can follow smoothly, avoiding hydrogen deficiency and output voltage drop. At the same time, through the feedforward-feedback algorithm, the compensation power of the sodium battery is accurately controlled, avoiding the sodium battery from bearing severe current impact, significantly prolonging the life of the key components, reducing the sodium battery current impact by 50%, and making the fuel cell run more efficiently.

[0050] The present application constitutes multiple insurance through the coordinated control of dual power sources and the dynamic correction mechanism based on voltage closed loop. First, the prediction and preloading ensure that the power is in place in advance; second, the accurate compensation of the sodium battery fills any temporary gap; and finally, the voltage feedback loop can inhibit disturbances in real time to prevent system oscillation.

[0051] All the advantages of the present application are realized through the optimization of algorithms and strategies, and the core increment is only a software module embedded in the existing flight control or energy management unit (EMU). It makes full use of the existing sensors (IMU, altimeter) and controllers (FCU, BMS) signals on the unmanned aerial vehicle, without adding any additional expensive hardware and complex models, realizing lightweight layout. BRIEF DESCRIPTION OF DRAWINGS

[0052] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are only a part of the present application, and other drawings can be obtained by those skilled in the art without creative labor on the basis of the provided drawings.

[0053] Figure 1 Flowchart of a hybrid system energy management method for a VTOL unmanned aerial vehicle during dual mode switching;

[0054] Figure 2 Flowchart of an energy management method containing a dynamic correction step provided for another embodiment. DETAILED DESCRIPTION

[0055] The technical solutions in the embodiments of the present application will be described clearly and completely in the embodiments of the present application in combination with the drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0056] The currently disclosed technical solutions are mainly based on rule-based feedback control or simple power source switching. In the face of such power mutations, there are usually three problems of response lag, current impact and safety risk. In order to solve the problems existing in the prior art, the present application provides a hybrid system energy management method for VTOL unmanned aerial vehicle dual-mode switching, which is based on the energy management method of "flight mode pre-identification" and "gradual power handover", and realizes fast response, smooth transition, safety and stability and efficiency optimization.

[0057] A hybrid system energy management method for VTOL unmanned aerial vehicle dual-mode switching, as shown in Figure 1 The method comprises the following steps:

[0058] S1: Dual-mode switching prediction parameter acquisition, real-time acquisition of dual-mode switching prediction parameters from the flight control system, the dual-mode switching prediction parameters including mode switching instruction, pitch angle change rate and height change rate.

[0059] The mode switching instruction is directly obtained from the flight control system (such as the pre-instruction of "ready to hover" or "ready to fly horizontally"), the pitch angle change rate and the height change rate are data obtained from the attitude sensor. If the unmanned aerial vehicle changes from horizontal cruising to vertical mode, the pitch angle will increase rapidly (such as >10° / s), and when landing vertically, the height continues to decrease and the rate increases (such as the height change rate increases from -1m / s to -3m / s), and when changing from horizontal to vertical, the height change rate tends to zero (entering the hovering state).

[0060] S2: Flight switching mode prediction, based on the dual-mode switching prediction parameters, the flight switching mode of the VTOL unmanned aerial vehicle is predicted, and the flight switching mode includes: vertical to horizontal switching and horizontal to vertical switching.

[0061] S3: Energy strategy adjustment, based on the flight switching mode, the corresponding energy adjustment strategy is triggered 5-10 seconds in advance, the power output of the hydrogen fuel cell and the sodium battery is dynamically adjusted, and gradual power handover is realized.

[0062] The energy adjustment strategy includes: when switching from vertical to horizontal, reducing the output power of the hydrogen fuel cell and gradually increasing the energy storage of the sodium battery; when switching from horizontal to vertical, controlling the output power of the hydrogen fuel cell to be raised to a high-power standby state, and dynamically adjusting the SOC of the sodium battery to the optimal output interval.

[0063] When switching from horizontal to vertical (e.g. from cruising to hovering), the power demand of the whole system will increase dramatically (3-5 times of cruising power). This huge power gap needs to be filled by the hydrogen fuel cell, as the sodium battery has limited power and can only serve as an instant supplement. When switching from vertical to horizontal (e.g. from hovering to cruising), the power demand of the whole system will decrease significantly. At this time, the power output of the hydrogen fuel cell needs to be reduced to cruising level, while the discharge of the sodium battery needs to be reduced and the energy storage of the sodium battery needs to be increased.

[0064] In this embodiment, based on the mode switching recognition of multi-source information (switching instruction, attitude angle rate of change, height rate of change) fusion, the intention can be recognized within 5-10 seconds before the actual mode switching occurs, and the energy scheduling strategy is triggered in advance. Through the prediction mechanism, the response speed is improved by more than 40%, so that the unmanned aerial vehicle can establish a new power balance faster, the synchronization of flight action and power output is greatly enhanced, and the control is more convenient, especially in high mobility tasks.

[0065] In order to realize high-precision recognition of flight mode, S2 adopts a multi-threshold triggering mechanism to predict the flight switching mode of the VTOL unmanned aerial vehicle.

[0066] When it is detected that the mode switching instruction is "ready to hover", or the pitch angle rate of change is greater than 8° / s, or the height rate of change suddenly changes from zero to negative, it is determined that the VTOL unmanned aerial vehicle will enter the vertical take-off and landing mode, and it is judged as horizontal to vertical switching.

[0067] The "horizontal to vertical" switching (e.g. preparing to hover in high-speed cruising) is a very high-risk operation. The power demand increases sharply, and if the system fails to provide sufficient power in time, it will cause the unmanned aerial vehicle to stall and crash. Therefore, the sensitivity of mode prediction must be improved as much as possible to ensure that it is never missed. In this embodiment, a multi-redundant triggering mode is adopted, that is, any one of the switching instruction ("ready to hover"), or the attitude change > 8° / s, or the height change becomes negative, is enough to indicate that the unmanned aerial vehicle is likely to enter the vertical mode.

[0068] The use of "or" logic establishes three independent safety lines. Even if the switching instruction is not received due to communication problems, as long as the IMU sensor detects that the nose begins to rise sharply, the system can still give an early warning and start energy preparation. This greatly enhances the robustness and reliability of the system.

[0069] When it is detected that the mode switching instruction is "ready to fly horizontally", and the pitch angle rate of change is less than 5° / s, and the height rate of change is less than 0.5m / s, it is determined that the VTOL unmanned aerial vehicle will enter the horizontal cruising mode, and it is judged as vertical to horizontal switching.

[0070] The risk of "vertical→horizontal" switching (e.g. from hover to forward flight) is relatively low. The power demand is decreasing, and even if the switch is slightly slow or delayed, the worst case is usually a less smooth transition and a less pleasant experience, but not a direct crash. Therefore, the primary goal is to ensure absolute accuracy in the timing of the switch, avoiding unnecessary and frequent mode oscillation.

[0071] In this embodiment, three conditions must be met simultaneously for vertical-to-horizontal switching: receiving the flight control command and the attitude tending to be horizontal and the height being stable. This ensures that the UAV is indeed in a stable and controllable state, ready to enter cruising. It can effectively prevent false positives.

[0072] In another embodiment, S2 calculates the mode switching confidence through an information fusion algorithm, and the mode switching confidence is used to quantify the probability of the VTOL UAV switching from the current flight mode to the target flight mode;

[0073] The mode switching confidence includes a first confidence for horizontal-to-vertical switching and a second confidence for vertical-to-horizontal switching; wherein the first confidence is calculated using a weighted sum model, and the second confidence is calculated using a product model;

[0074] When the first confidence exceeds a first trigger threshold, it is determined to be horizontal-to-vertical switching, and when the second confidence exceeds a second trigger threshold, it is determined to be vertical-to-horizontal switching; the first trigger threshold is less than the second trigger threshold.

[0075] In this embodiment, the system can be embedded as a software module in the energy management unit (EMU) or advanced flight control of the UAV, and real-time multi-source data is received to output a continuous "mode switching confidence" value, which is used to trigger energy adjustment in advance.

[0076] The system calculates two confidences in parallel and makes independent judgments. The first confidence is used for horizontal-to-vertical switching, and the calculation formula of the first confidence is:

[0077]

[0078] Wherein, C1 is the first confidence; c cmd is the mode switching command; R pitch is the pitch angle rate of change; R high is the height rate of change; f 11 (c cmd ) is the first instruction sub-confidence calculated based on the mode switching command; f 12 (R pitch ) is the first pitch angle sub-confidence calculated based on the pitch angle rate of change; f 13 (R high) is the first height sub-confidence calculated based on the rate of height change; w1, w2, w3 are weight coefficients corresponding to the first instruction sub-confidence, the first pitch angle sub-confidence, and the first height sub-confidence respectively, and w1+w2+w3=1, and w1>w2>w3.

[0079] The weight distribution in this embodiment is as follows: the flight control instruction is the most direct representation of the intention of flight mode switching, and has the highest weight, w1=0.5; the attitude change is strong association evidence, w2=0.3; and the height change is auxiliary evidence, w3=0.2.

[0080] When the mode switching instruction is detected as “prepare to hover”, f 11 (c cmd )=1.0, and when the mode switching instruction is not detected as “prepare to hover”, f 11 (c cmd )=0.0, and the instruction exists, that is, 100% confidence.

[0081] f 12 ), f 13 ) are mapped using a Sigmoid or piecewise linear function:

[0082] When the pitch angle rate of change is less than 5° / s, f 12 (R pitch )=0.0, when the pitch angle rate of change is between 5 and 8° / s, the first pitch angle sub-confidence is linearly mapped between 0.0 and 1.0; and when the pitch angle rate of change is greater than 8° / s, f 12 (R pitch )=1.0.

[0083] When the height rate of change is greater than -0.2 m / s, f 13 (R high )=0.0, when the height rate of change is between -0.2 and -1.0 m / s, the first height sub-confidence is linearly mapped between 0.0 and 1.0; and when the height rate of change is less than -1.0 m / s, f 13 (R high )=1.0.

[0084] The first trigger threshold is set to 70%, and when the first confidence is greater than 70%, it is determined that the horizontal-to-vertical switching is required, and the fuel cell needs to be preloaded. The first trigger threshold is low, which can ensure the sensitivity of recognition.

[0085] The second confidence is used for the judgment of vertical-to-horizontal switching, and the calculation formula of the second confidence is:

[0086]

[0087] wherein C2 is the second confidence; ccmd is a mode switching instruction; R pitch is a pitch rate; R high is an altitude rate; f 21 (c cmd ) is a second instruction sub-confidence calculated based on the mode switching instruction; f 22 (|R pitch |) is a second pitch angle sub-confidence calculated based on the pitch rate absolute value, which decreases as the pitch rate absolute value increases; f 23 (|R high |) is a second altitude sub-confidence calculated based on the altitude rate absolute value, which decreases as the altitude rate absolute value increases.

[0088] When the mode switching instruction is detected as "prepare to fly horizontally", f 21 (c cmd ) = 1.0, and when the mode switching instruction is not detected as "prepare to fly horizontally", f 21 (c cmd ) = 0.8. Even without instructions, a basic confidence possibility of 0.8 is retained (allowing pure manual control).

[0089] F 22 (), f 23 () uses a Sigmoid or piecewise linear function mapping:

[0090] When the pitch rate absolute value is less than 3° / s, f 22 (|R pitch |) = 1.0, when the pitch rate absolute value is between 3~5° / s, the second pitch angle sub-confidence is linearly mapped between 1.0 and 0; when the pitch rate is greater than 5° / s, f 22 (|R pitch |) = 0.0;

[0091] When the altitude rate absolute value is less than 0.3m / s, f 23 (|R high |) = 1.0, when the altitude rate absolute value is between 0.3~0.8m / s, the second altitude sub-confidence is linearly mapped between 1.0 and 0; when the altitude rate absolute value is greater than 0.8m / s, f 23 (|R high |) = 0.0.

[0092] The second trigger threshold is set to 90% at the same time, when the second confidence > 90%, it is judged as vertical to horizontal switching, and the fuel cell power needs to be smoothly reduced. The first trigger threshold is lower, and the threshold is higher, to ensure stability and prevent false triggering.

[0093] To prevent the confidence from fluctuating around the threshold, the instructions frequently switch. In this embodiment, once triggered, the triggered state is only released when the confidence falls below the threshold -20% (i.e. the first confidence <50% or the second confidence <70%).

[0094] On the other hand, the energy management system can not only be triggered, but also dynamically adjust the slope of power change according to the growth rate of confidence.

[0095] For example: The first confidence soars from 10% to 80% in 1 second, indicating that the switch is very urgent. At this time, the system can slightly speed up the power ramp-up slope of the hydrogen fuel cell to better match the urgent demand.

[0096] After completing the flight switch mode prediction, enter S3, which dynamically adjusts the power output of the hydrogen fuel cell and the sodium battery according to the flight switch mode prediction result, and realizes gradual power handover.

[0097] In S3: 5 seconds in advance when switching from vertical to horizontal, trigger the corresponding energy adjustment strategy; 10 seconds in advance when switching from horizontal to vertical, trigger the corresponding energy adjustment strategy.

[0098] In this embodiment, when switching from "horizontal to vertical", the overall power demand will increase sharply (possibly 3-5 times the cruising power). This huge power gap is mainly borne by the hydrogen fuel cell, but the output power of the hydrogen fuel cell (especially the proton exchange membrane fuel cell) cannot be instantly increased, and the response of its internal electrochemical reaction, air / hydrogen supply system (such as air compressor) also needs a certain time. If the power request is too fast, it will cause "hydrogen starvation" and "oxygen starvation", causing the output voltage to drop sharply and unable to provide the required power. Therefore, in this embodiment, the corresponding energy adjustment strategy is triggered 10 seconds in advance. This 10-second prediction window allows the control system to gradually increase the power of the fuel cell from the high-efficiency cruising state to the "high-power standby" state (such as 90% of the maximum power) at a safe slope. When the UAV really starts to enter the vertical mode and the power demand reaches the peak, the fuel cell has basically reached the required power, and the sodium battery only needs to provide the difference, thereby avoiding the risk of power interruption caused by the fuel cell "not keeping up".

[0099] When switching from "vertical to horizontal" (for example, from hovering to cruising), the overall power demand will decrease significantly, allowing the system to reduce power output much faster than increasing power. In this process, the role of the sodium battery is to switch from "high-power discharge" to "small-power charging or standby". Reducing the discharge current or even charging can be completed instantaneously for the sodium battery.

[0100] Therefore, the corresponding energy adjustment strategy is triggered 5 seconds in advance in the embodiment, and the 5-second time window is enough for the fuel cell to reduce the power at a gentle slope (to avoid voltage fluctuation). At the same time, the time is not too long, so that the system can enter the cruise state as soon as possible to avoid unnecessary energy waste.

[0101] The scheme is not mechanically applied to a single rule, but based on deep understanding of physical characteristics of different power sources and accurate management for different flight stages, so as to realize the comprehensive goals of safety, efficiency and long life. This is a significant intelligent and advanced place compared with the traditional single feedback control strategy.

[0102] In S3, different energy adjustment strategies are adopted in different switching modes. When switching from vertical to horizontal, the UAV changes from hovering (high power) to forward flight (low power), and the total power demand of the whole machine decreases greatly. In the process of total power reduction, the continuity and stability of power output are ensured to avoid voltage fluctuation. When switching from horizontal to vertical, the UAV changes from cruise (low power) to hovering (high power), and the total power demand of the whole machine increases sharply.

[0103] In the embodiment, when switching from vertical to horizontal, the first fixed time window is set to 3 seconds, and the vertical-to-horizontal power handover is completed within the first fixed time window. The total power demand of the VTOL UAV is borne by the hydrogen fuel cell and the sodium battery during the vertical-to-horizontal power handover, the fuel cell power decreases at a predetermined slope, and the power gap generated is immediately dynamically compensated by the sodium battery increasing the discharge power.

[0104] For the fuel cell (slow response system): since its output cannot be suddenly reduced, it is instructed to linearly reduce the power at a safe and predetermined slope, which means that the current change rate is limited within a safe range, effectively preventing the direct current bus voltage from suddenly dropping or rising due to current impact, and protecting the working stability of all electrical equipment (such as flight control, rudder, and task load) connected to the bus.

[0105] For the sodium battery (fast response system): while the fuel cell power is reduced, in order to maintain the total power to meet the current flight demand (to prevent power surplus from causing overspeed), it is instructed to increase the discharge power synchronously to "take over" or "offset" part of the load. This is a "regulation" command.

[0106] When switching from horizontal to vertical, the second fixed time window is set to 5 seconds, and the horizontal-to-vertical power reserve is completed within the second fixed time window. The hydrogen fuel cell output power is increased to the high power standby state at a safe slope, and the sodium battery SOC is adjusted to the interval of 60% to 80%.

[0107] For fuel cell (slow response system): instruct it to increase power output in advance with a safe slope, enter high-power standby state.

[0108] For sodium battery (fast response system): instruct it to prepare to provide instantaneous peak power to make up for the power gap in the early stage of fuel cell rise, and possibly adjust its SOC to the optimal output interval.

[0109] The first fixed window time and the second window time are the time windows for completing energy exchange in two switching modes respectively. This time parameter can be adjusted according to the UAV model and power system parameters.

[0110] When switching vertically to horizontally, the discharge power of the sodium battery needs to be increased synchronously. In order to ensure the accuracy of process control, a collaborative control scheme based on power closed loop and state feedback is designed. The specific steps for the sodium battery to increase discharge power are as follows:

[0111] Based on the difference between the total power demand and the actual output power of the hydrogen fuel cell, and through the closed-loop feedback of the DC bus voltage, the discharge power of the sodium battery is calculated;

[0112] The calculation formula of the actual output power of the hydrogen fuel cell is:

[0113] Where, P fc-actual is the actual output power of the hydrogen fuel cell, P fc-initial is the initial output power of the hydrogen fuel cell at the moment of triggering the energy adjustment strategy, i.e. t=0; k is the predetermined slope, which is not greater than 10%P fc_initial / s, t is the time, which is the time elapsed since triggering the energy adjustment strategy.

[0114] Different UAV models have different initial output powers (P fc-initial ) of fuel cells. Using percentage makes this control strategy automatically adapt to power systems of different power levels, and 10%P fc-initial / s is an empirical value based on the characteristics of fuel cells (typical dynamic response capability of PEMFC) and system experimental verification, which can ensure the safe and stable reduction of fuel cell output in most working conditions.

[0115] Suppose a UAV's fuel cell has a maximum power P max =100kW. The initial power P fc-initial =100kW when hovering vertically.

[0116] The predetermined slope k=10%*100kW / s=10kW / s

[0117] Apply the formula:

[0118] Hydrogen fuel cell power reduction process: when t=0s (start switching): P fc-actual (0)=100kW, when t=1s: P fc-actual (1)=100-10*1=90kW, when t=2s: P fc-actual (2)=100-10*2=80kW, when t=3s (switching complete): P fc-actual (3)=100-10*3=70kW (reaching the target cruise power of 70kW).

[0119] At the same time, the energy management system (EMS) will synchronously increase the discharge power of the sodium battery to accurately fill in this "10kW per second" power gap, ensuring that the total power supplied to the motor remains stable.

[0120] The calculation formula of the discharge power of the sodium battery is: where P batt-ref is the discharge power of the sodium battery, P tatal is the total power demand, PID() is the proportional-integral-derivative correction function, Vc is the DC bus voltage reference value, and Va is the DC bus voltage actual value.

[0121] The power change during mode switching is the main factor that causes the loss of the power source life. In this embodiment, the "gradual power handover" strategy is used to set a safety slope limit of ≤10%P fc-initia / s for the power change of the hydrogen fuel cell, so that the internal electrochemical reaction and auxiliary system (such as the air compressor) can smoothly follow, avoiding hydrogen deficiency and sudden drop of output voltage. At the same time, through the feedforward-feedback algorithm, the compensation power of the sodium battery is accurately controlled, avoiding the sodium battery from bearing severe current impact, significantly prolonging the life of key components, reducing the sodium battery current impact by 50%, and making the fuel cell run more efficiently.

[0122] In order to ensure the robustness of the vertical to horizontal switching, the predetermined slope k is not fixed and will be adjusted in real time according to the system state to realize truly intelligent management.

[0123] In this embodiment, the predetermined slope adjustment is based on the sodium battery SOC:

[0124] When SOC<50% (low power): automatically reduce the slope k from 10% / s to 5% / s. The purpose of this is to slow down the power reduction speed of the fuel cell, thereby reducing the discharge power requirement of the low power sodium battery, preventing over-discharge of the sodium battery, and protecting its life.

[0125] When SOC > 90% (high power): automatically increase the slope k to 15% / s. This allows the fuel cell power to drop faster, encouraging the sodium battery to participate more in discharging, thus freeing up capacity for subsequent regenerative braking (energy recovery) or charging of the fuel cell.

[0126] To establish a complete closed-loop control framework and achieve adaptive adjustment based on multi-parameter feedback, in another embodiment, as shown in Figure 2 A hybrid system energy management method for VTOL UAV dual-mode switching is provided:

[0127] S1: Dual-mode switching prediction parameter acquisition, real-time acquisition of dual-mode switching prediction parameters from the flight control system, including mode switching instructions, pitch angle rate and height rate.

[0128] S2: Flight switching mode prediction, based on dual-mode switching prediction parameters, the flight switching mode of the VTOL UAV is predicted, including vertical to horizontal switching and horizontal to vertical switching.

[0129] S3: Energy strategy adjustment, based on the flight switching mode, the corresponding energy adjustment strategy is triggered 5-10 seconds in advance, and the power output of the hydrogen fuel cell and the sodium battery is dynamically adjusted to realize gradual power handover.

[0130] When switching from vertical to horizontal, set the first fixed time window to 3 seconds, and complete the vertical to horizontal power handover within the first fixed time window. When the VTOL UAV switches from vertical to horizontal, the total power demand is borne by the hydrogen fuel cell and the sodium battery, and the fuel cell power decreases at a predetermined slope. The resulting power gap will be immediately compensated by the sodium battery increasing the discharge power;

[0131] When switching from horizontal to vertical, set the second fixed time window to 5 seconds, and complete the horizontal to vertical power storage within the second fixed time window. The hydrogen fuel cell output power is increased to a high power standby state at a safety slope, and the sodium battery SOC is adjusted to the interval of 60%-80%.

[0132] S4: Dynamic correction: real-time monitoring of dynamic correction parameters, including DC bus voltage fluctuation rate, sodium battery SOC, attitude angle deviation, adjusting the predetermined slope and the first fixed time window according to the dynamic correction parameters to ensure that the switching process can be safely and smoothly completed under any working condition.

[0133] When the DC bus voltage change rate is greater than 5% / s, the predetermined slope is reduced and the first fixed time window is extended;

[0134] When the sodium battery SOC is lower than 50%, the predetermined slope is reduced, and the fuel cell power reduction time is shortened to avoid over-discharge of the sodium battery.

[0135] When the attitude angle deviation is greater than 5°, return to step S1.

[0136] In this embodiment, through the coordinated control of dual power sources and the dynamic correction mechanism based on voltage closed loop, multiple insurance is constituted. First, pre-judgment and pre-loading ensure that the power is in place in advance; second, the precise compensation of the sodium battery fills any temporary gap; finally, the voltage feedback loop can inhibit disturbances in real time and prevent system oscillation.

[0137] All the advantages of this embodiment are realized through the optimization of algorithms and strategies, and the only core increment is a software module embedded in the existing flight control or energy management unit (EMU). It makes full use of the existing sensors (IMU, altimeter) and controller (FCU, BMS) signals on the UAV, without adding any additional expensive hardware and complex models, realizing lightweight layout.

[0138] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by programs instructing relevant hardware, and the programs can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the above-mentioned embodiment methods. The storage medium can be a magnetic disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM), etc.

[0139] The above only describes the preferred examples of the application and is not intended to limit the application. Although the application has been described in detail with reference to the foregoing examples, those skilled in the art can modify the technical solutions described in the foregoing examples or make equivalent replacements to some technical features. Any modification, equivalent replacement, etc. within the spirit and principles of the application shall be included in the protection scope of the application.

Claims

1. A method for energy management of a hybrid system during dual-mode switching of a VTOL unmanned aerial vehicle (UAV), characterized in that, include: S1: Dual-mode switching prediction parameter acquisition: acquire dual-mode switching prediction parameters in real time from the flight control system. The dual-mode switching prediction parameters include mode switching command, pitch angle change rate and altitude change rate. S2: Flight switching mode prediction. Based on the dual-mode switching prediction parameters, the flight switching mode of the VTOL UAV is predicted. The flight switching mode includes: vertical to horizontal switching and horizontal to vertical switching. S3: Energy strategy adjustment. Based on the flight switching mode, the corresponding energy adjustment strategy is triggered 5 to 10 seconds in advance to dynamically adjust the power output of the hydrogen fuel cell and sodium battery to achieve gradual power handover. The energy adjustment strategy includes: reducing the output power of the hydrogen fuel cell and gradually increasing the energy storage of the sodium battery when switching from vertical to horizontal; and controlling the output power of the hydrogen fuel cell to a high-power standby state when switching from horizontal to vertical, while dynamically adjusting the SOC of the sodium battery to the optimal output range.

2. The energy management method for a hybrid system during dual-mode switching of a VTOL UAV according to claim 1, characterized in that, The S2 uses a multi-threshold triggering mechanism to predict the flight switching mode of the VTOL UAV; When the mode switching command is detected as "prepare to hover", or the pitch angle change rate is greater than 8° / s, or the altitude change rate changes abruptly from zero to a negative value, the VTOL UAV is about to enter the vertical take-off and landing mode, and it is determined to be the horizontal to vertical switch. When the mode switching command is detected as "prepare for level flight", and the pitch angle change rate is less than 5° / s and the altitude change rate is less than 0.5m / s, the VTOL UAV is about to enter the horizontal cruise mode, which is determined to be the vertical-to-horizontal switch.

3. The hybrid system energy management method for dual-mode switching of a VTOL UAV according to claim 1, characterized in that, The S2 calculates the mode switching confidence score through an information fusion algorithm. The mode switching confidence score is used to quantify the probability that the VTOL UAV will switch from the current flight mode to the target flight mode. The mode switching confidence level includes a first confidence level for horizontal to vertical switching and a second confidence level for vertical to horizontal switching; wherein, the first confidence level is calculated using a weighted sum model and the second confidence level is calculated using a product model; When the first confidence level exceeds the first trigger threshold, it is determined to be a horizontal to vertical switch; when the second confidence level exceeds the second trigger threshold, it is determined to be a vertical to horizontal switch; the first trigger threshold is less than the second trigger threshold.

4. The energy management method for a hybrid system during dual-mode switching of a VTOL UAV according to claim 3, characterized in that, The formula for calculating the first confidence level is: Where C1 is the first confidence level; c cmd This is a mode switching command; R pitch R is the rate of change of pitch angle; high f represents the rate of change of height. 11 (c) cmd ) is the first instruction sub-confidence calculated based on the mode switching instruction; f 12 (R) pitch The first pitch angle sub-confidence score is calculated based on the pitch angle change rate; f 13 (R) high The first altitude sub-confidence is calculated based on the altitude change rate; w1, w2, and w3 are the first command sub-confidence, the first pitch angle sub-confidence, and the first altitude sub-confidence, respectively, and are weighting coefficients, with w1+w2+w3=1 and satisfying w1>w2>w3.

5. The hybrid system energy management method for dual-mode switching of a VTOL UAV according to claim 4, characterized in that, When the mode switching command is detected to be "prepare to hover", then f 11 (c) cmd If f = 1.0, and no "prepare to hover" mode switching command is detected, then f 11 (c) cmd =0.0; When the pitch angle change rate is less than 5° / s, then f 12 (R) pitch When the pitch angle change rate is between 5 and 8° / s, the first pitch angle sub-confidence level is linearly mapped between 0.0 and 1.0; when the pitch angle change rate is greater than 8° / s, f = 0.

0. 12 (R) pitch =1.0; When the height change rate is greater than -0.2 m / s, then f 13 (R) high When the rate of change of altitude is between -0.2 and -1.0 m / s, the first altitude sub-confidence score is linearly mapped between 0.0 and 1.0; when the rate of change of altitude is less than -1.0 m / s, f = 0.

0. 13 (R) high =1.

0.

6. The energy management method for a hybrid system during dual-mode switching of a VTOL UAV according to claim 3, characterized in that, The formula for calculating the second confidence level is: Where C2 is the second confidence level; c cmd This is a mode switching command; R pitch R is the rate of change of pitch angle; high f represents the rate of change of height. 21 (c) cmd () is the second instruction sub-confidence calculated based on the mode switching instruction; f 22 (|R) pitch |) is the second pitch angle sub-confidence level calculated based on the absolute value of the pitch angle change rate, and its value decreases as the absolute value of the pitch angle change rate increases; f 23 (|R) high |) is the second height sub-confidence level calculated based on the absolute value of the altitude change rate, and its value decreases as the absolute value of the altitude change rate increases.

7. The energy management method for a hybrid system during dual-mode switching of a VTOL UAV according to claim 6, characterized in that, When the mode switching command is detected to be "prepare for level flight", then f 21 (c) cmd If f = 1.0, and no "prepare for level flight" command is detected, then f 21 (c) cmd =0.8; When the absolute value of the pitch angle change rate is less than 3° / s, then f 22 (|R) pitch |) = 1.0, when the absolute value of the pitch angle change rate is between 3 and 5° / s, the second pitch angle sub-confidence level is linearly mapped between 1.0 and 0; when the pitch angle change rate is greater than 5° / s, then f 22 (|R) pitch |)=0.0; When the absolute value of the height change rate is less than 0.3 m / s, then f 23 (|R) high |) = 1.

0. When the absolute value of the altitude change rate is between 0.3 and 0.8 m / s, the second altitude sub-confidence level is linearly mapped between 1.0 and 0; when the absolute value of the altitude change rate is greater than 0.8 m / s, then f 23 (|R) high |)=0.

0.

8. The energy management method for a hybrid system during dual-mode switching of a VTOL UAV according to claim 1, characterized in that, In S3: when switching from vertical to horizontal, the corresponding energy adjustment strategy is triggered 5 seconds in advance; when switching from horizontal to vertical, the corresponding energy adjustment strategy is triggered 10 seconds in advance.

9. The energy management method for a hybrid system during dual-mode switching of a VTOL UAV according to claim 1, characterized in that, In S3: When switching from vertical to horizontal, a first fixed time window is set to 3 seconds. The vertical-to-horizontal power handover is completed within the first fixed time window. During the vertical-to-horizontal power handover, the total power demand of the VTOL UAV is shared by the hydrogen fuel cell and the sodium battery. The power of the fuel cell decreases at a predetermined slope, and the resulting power gap will be dynamically made up by the sodium battery increasing its discharge power. When switching from horizontal to vertical, a second fixed time window is set to 5 seconds. Within the second fixed time window, the horizontal and vertical power reserves are completed, the output power of the hydrogen fuel cell is increased to a high power standby state at a safe slope, and the SOC of the sodium battery is adjusted to the range of 60% to 80%.

10. The energy management method for a hybrid system during dual-mode switching of a VTOL UAV according to claim 9, characterized in that, The method further includes S4: dynamic correction, real-time monitoring of dynamic correction parameters, the dynamic correction parameters including: DC bus voltage fluctuation rate, sodium battery SOC, attitude angle deviation, and adjusting the predetermined slope and the first fixed time window according to the dynamic correction parameters; When the DC bus voltage change rate is greater than 5% / s, the predetermined slope is reduced and the first fixed time window is extended. When the sodium battery's state of charge (SOC) is below 50%, the predetermined slope is reduced to shorten the fuel cell's power reduction time and prevent the sodium battery from being over-discharged. If the attitude angle deviation is greater than 5°, return to step S1.

Images