Hydrogen-electric hybrid unmanned aerial vehicle energy management method, device, medium and product

Abstract

The application discloses a hydrogen-electric hybrid unmanned aerial vehicle energy management method, device, medium and product, relates to the hydrogen-electric hybrid unmanned aerial vehicle energy management field, and the method comprises the following steps: based on real-time state information of a hydrogen-electric hybrid unmanned aerial vehicle, predicting demand power in a future prediction time domain by using a working condition prediction model; based on the real-time state information, constructing, in each rolling time domain of model predictive control, an optimization problem with the demand power in the future prediction time domain as input and the minimum total hydrogen consumption as a target; based on the optimization problem, obtaining a control instruction by using the Pontryagin minimum principle; controlling according to the control instruction; and returning to the step of predicting the demand power in the future prediction time domain by using the working condition prediction model based on the real-time state information of the hydrogen-electric hybrid unmanned aerial vehicle to perform rolling optimization control. The application can improve the economy, real-time performance, robustness and adaptability of hydrogen-electric hybrid unmanned aerial vehicle energy management.

Description

Technical Field

[0001] This application relates to the field of energy management for hydrogen-electric hybrid drones, and in particular to a method, device, medium, and product for energy management of hydrogen-electric hybrid drones. Background Technology

[0002] In the field of energy management for hydrogen-electric hybrid unmanned aerial vehicles (UAVs), existing technical solutions mainly include the following representative methods: (1) rule-based control (such as fuzzy control) achieves real-time decision-making through a pre-set empirical rule base; (2) instantaneous optimization algorithms (represented by the Equivalent Consumption Minimization Strategy (ECMS)) use equivalent factors to instantaneously equate energy consumption to fuel consumption and perform static optimization; (3) global optimization algorithms (such as dynamic programming (DP)) search for the global optimal solution by traversing the state space under the premise of knowing the complete operating conditions; (4) model predictive control (MPC) uses predictive models to perform rolling optimization in a finite time domain. These methods constitute the main technical system of current energy management.

[0003] However, the main drawbacks of the above technologies are as follows: (1) Rule-based control relies on a preset empirical rule base and lacks an optimization theoretical basis. When faced with complex and undefined flight conditions, it has poor adaptability and limited optimization potential. (2) ECMS performs instantaneous static optimization through equivalent factors, only considering the optimization at the current moment, and cannot make the best decision based on future flight mission information. At the same time, its performance is highly sensitive to the value of the equivalent factor, which is difficult to adjust accurately in real time in practice, causing the system to easily deviate from the optimal operating point in practical applications. (3) Although global optimization algorithms (such as dynamic programming DP) can obtain the global optimal solution, they have the "curse of dimensionality" problem. Their computation and storage requirements increase exponentially with the increase of the dimension of the state variables. Therefore, it cannot be used for online real-time computation on airborne computing platforms and can only be used for offline simulation and performance benchmark testing. (4) Traditional MPC performs local optimization in the finite prediction time domain, and nonlinear MPC has the problem of heavy computational burden.

[0004] In conclusion, existing technologies cannot simultaneously meet the two core requirements of long-endurance UAV energy management: "global economy" and "online real-time performance".

[0005] Therefore, there is an urgent need to provide a hydrogen-electric hybrid UAV energy management method or system that can theoretically approach the global optimal economy of the UAV energy management system, while meeting the onboard real-time requirements in terms of computational efficiency. Summary of the Invention

[0006] The purpose of this application is to provide a method, device, medium, and product for energy management of hydrogen-electric hybrid drones, which can improve the economy, real-time performance, robustness, and adaptability of energy management for hydrogen-electric hybrid drones.

[0007] To achieve the above objectives, this application provides the following solution: Firstly, this application provides an energy management method for a hydrogen-electric hybrid unmanned aerial vehicle (UAV), the energy management method comprising: Based on the real-time status information of the hydrogen-electric hybrid drone, the power demand in the future prediction time domain is predicted using a working condition prediction model; the status information includes: battery state of charge, flight altitude, flight airspeed, environmental information, and flight plan instructions; Based on real-time state information, an optimization problem is constructed in each rolling time domain of model predictive control, with the power demand in the future predicted time domain as input and the goal of minimizing total hydrogen consumption. Based on the aforementioned optimization problem, the control commands are obtained using the Pontryagin minimum principle. Control is performed according to the control commands; and real-time status information based on the hydrogen-electric hybrid drone is returned, and rolling optimization control is performed by using the operating condition prediction model to predict the power demand in the future prediction time domain.

[0008] Secondly, this application provides a computer device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the energy management method for the hydrogen-electric hybrid drone.

[0009] Thirdly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the energy management method for the hydrogen-electric hybrid unmanned aerial vehicle described above.

[0010] Fourthly, this application provides a computer program product, including a computer program that, when executed by a processor, implements the aforementioned hydrogen-electric hybrid drone energy management method.

[0011] According to the specific embodiments provided in this application, this application has the following technical effects: This application provides an energy management method, device, medium, and product for a hydrogen-electric hybrid unmanned aerial vehicle (UAV). It offers a collaborative energy management method (PMP-MPC) based on Pontryagin's Minimum Principle (PMP) as the optimization kernel and Model Predictive Control (MPC) as the rolling framework. By deeply integrating the global optimality concept of PMP with the feedback optimization mechanism of MPC, the hydrogen-electric hybrid UAV theoretically approaches the global optimal economic efficiency while simultaneously meeting the onboard real-time requirements in terms of computational efficiency. Attached Figure Description

[0012] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0013] Figure 1 This is a schematic diagram of a hydrogen-electric hybrid drone energy management method according to one embodiment of this application; Figure 2 This is a schematic diagram of the overall architecture of an energy management method for a hydrogen-electric hybrid unmanned aerial vehicle in one embodiment of this application; Figure 3 This is a schematic diagram illustrating the deep integration of PMP and MPC. Figure 4 This is a diagram of the overall algorithm architecture of PMP-MPC; Figure 5 Here is the overall algorithm flowchart for PMP-MPC; Figure 6 This is a timing diagram for real-time control execution. Detailed Implementation

[0014] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0015] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0016] like Figure 2As shown, the overall architecture of this application includes an information perception layer, a prediction layer, a core control layer (PMP-MPC controller), and an execution layer; Specifically, the information perception layer includes a battery state of charge (SOC) sensor, a flight speed sensor, an altitude sensor, and an environmental sensor; these sensors are used to collect the real-time status of the UAV (such as SOC, flight speed, etc.). Among them, the battery SOC sensor is generally integrated into the battery management system (BMS) or directly installed on the power battery pack, while the flight speed, altitude, and environmental sensors are generally distributed on the UAV's fuselage and wings. Specifically, the environmental sensors mainly collect data such as wind speed, wind direction, ambient temperature, and atmospheric pressure. Wind speed and wind direction serve as multi-dimensional input features for the Long Short-Term Memory (LSTM) network to improve speed prediction accuracy, while temperature and air pressure data are used to calculate air density and thus accurately solve for the UAV's power demand. At the same time, temperature information is also transmitted to the battery management system (BMS) to optimize the battery charging and discharging strategy, ensuring the system operates safely and efficiently in complex environments. The prediction layer contains a working condition prediction model to predict the flight power demand over a future period. The core control layer receives information from the information perception layer and the prediction layer, executes the PMP-MPC optimization algorithm online, and outputs the optimal fuel cell power command (control command). The execution layer includes the fuel cell controller, lithium battery management system, etc., and executes the control commands of the core control layer.

[0017] In one exemplary embodiment, such as Figure 1 As shown, a method for energy management of a hydrogen-electric hybrid unmanned aerial vehicle is provided, which includes the following steps S101 to S104. Wherein: S101, based on the real-time status information of the hydrogen-electric hybrid drone, uses a condition prediction model to predict the power demand in the future prediction time domain; the status information includes: battery state of charge, flight altitude, flight airspeed, environmental information, and flight plan instructions; the condition prediction model includes, but is not limited to: long short-term memory network, time series model (ARIMA), dynamic filtering algorithm (Kalman filter), probabilistic model (Markov chain) or machine learning regression model (SVR, GPR).

[0018] Specifically, based on flight missions, historical flight data, and real-time environmental information, a working condition prediction model is used to obtain the future prediction time domain of hydrogen-electric hybrid drones. The flight speed sequence within the range is converted into a power demand sequence P. req ; As a specific example, when the operating condition prediction model is a Long Short-Term Memory (LSTM) network, the LTM network can effectively capture long-term dependencies in flight speed time-series data, provide accurate short-term predictions, and provide high-quality forward-looking information for subsequent optimization.

[0019] S101 also includes: S11, obtain real-time status information of the hydrogen-electric hybrid drone; S12, Based on real-time state information, a mathematical model of a hybrid power system is constructed using a battery equivalent circuit model and UAV dynamics; the mathematical model of the hybrid power system uses the battery state of charge as the state variable, the fuel cell power as the control variable, and the demand power as the disturbance variable.

[0020] The state equation of the mathematical model of the hybrid power system is: dx / dt = f(x, u, w); where x = SOC is the state variable (battery state of charge), u = P fc As the control variable (fuel cell power), w=P req For the disturbance variable (demand power); S102 and S103 are PMP-MPC optimization algorithms, and as... Figure 2 , Figure 3 as well as Figure 4 As shown; S102, based on real-time state information, in each rolling time domain of model predictive control, an optimization problem is constructed with the power demand in the future predicted time domain as input and the total hydrogen consumption as the objective, that is, a finite time domain optimization problem is constructed. Using formula Determine the cost function ; in, Sampling time, For future time-domain prediction, for Hydrogen consumption rate at time t, for The state of charge of the battery at any given time. for Reference trajectory at any moment For the terminal cost function, , As a reference trajectory for future predictions in the time domain, , and All are weighting coefficients, with min representing the minimum value. To predict the state of charge of the battery in the future time domain.

[0021] The purpose and design principles of the terminal cost function in this application are as follows: 1. Prevent the "terminal effect": avoid excessive consumption of battery energy by the optimizer at the end of the prediction time domain; 2. Achieve energy conservation: force the PMP-MPC controller composed of PMP-MPC to maintain the battery SOC within a reasonable range at the end of the future prediction time domain; 3. Selection of weight coefficients: Achieving a balance between instantaneous optimization and terminal constraints; 4. Reference trajectory design: Dynamically generated based on global energy management strategy, that is, the target SOC value at the end of the predicted time domain is calculated in real time based on the total flight range, the distance already flown, and the expected flight distance in the predicted time domain. S103, Based on the optimization problem, the control command is obtained by using the Pontryagin minimum principle; S103 specifically includes: H = instantaneous hydrogen consumption + λ d(SOC) / dt represents the architecture construction. The Hamiltonian function H; such as Figure 3 As shown, the Hamiltonian function system is used to achieve the synergy of global and local optimization, realizing the deep integration of PMP and MPC. like Figure 3 As shown, the MPC framework uses predictive models to estimate future system behavior and solves for the optimal control sequence in each control cycle. MPC's rolling optimization mechanism ensures the system's real-time performance and adaptability. The PMP kernel transforms the energy management problem into an optimal control problem by constructing a Hamiltonian function. By incorporating global economic information into local optimization decisions using costate equations, it overcomes the "short-sightedness" problem of traditional MPC.

[0022] Using the formula d(SOC) / dt = H / λ determines the state equations (inherent dynamics); Using the formula dλ / dt = - H / (SOC) determines the costate equation (making λ a messenger of "future cost impact," introducing a global perspective into local optimization, and capturing the impact of future states on total cost); that is, the costate variable λ is updated according to the costate equation during online rolling optimization. At the beginning of each control cycle, the optimal costate variable value obtained in the previous time step is used... (k) is used as the initial value λ(k) of the costate variable at the current time to achieve a smooth transition in the optimization process.

[0023] Using formula =arg min H determines the minimum condition, and the minimum condition is directly used to find the value in [P]. fcmin ,P fcmax Find the P that minimizes H within the range.fc This method efficiently handles constraints, directly finding the optimal solution within the control constraints, significantly improving solution efficiency and stability. Using formula Determine the cross-sectional conditions; Where x is the state variable, u is the control variable, λ is the co-state variable, and t is the time. Let be the hydrogen consumption rate at time t. Let be the battery state of charge at time t. for The reference trajectory at time t, d(SOC) / dt is the rate of change of the battery's state of charge, P fc For fuel cell power, For the optimal fuel cell power, argmin is the function that minimizes it. E / x is the partial derivative of the terminal cost function with respect to the terminal state. H / λ is the partial derivative of H with respect to λ, and dλ / dt is the derivative of λ with respect to t. H / (SOC) is the partial derivative of H with respect to SOC. For optimal fuel cell power, Let λ be the costate variable that makes U optimal.

[0024] This application, within the rolling time domain of each MPC (Multi-Process Planning) cycle, no longer employs a traditional programming solver. Instead, it constructs the energy management problem as an optimal control problem and uses Process Management Programming (PMP) to solve it. By introducing PMP's "costate variables" and "costate equations," the impact of future states on total cost (global information) is incorporated into the current decision-making process, overcoming the "short-sightedness" of traditional MPC. Through PMP's "minimum conditions," the optimal solution is directly sought within the control constraints. This is a fundamental solution method that is more efficient and stable than traditional numerical programming.

[0025] like Figure 3 As shown, this application achieves dual-timescale optimization through PMP-MPC fusion, specifically in the following ways: (1) Global optimization layer: Based on the global perspective of PMP, it optimizes the total hydrogen consumption in the future prediction time domain and provides a long-term energy allocation strategy.

[0026] (2) Local optimization layer: Combined with the rolling mechanism of MPC, the control commands are finely adjusted within the current control cycle to ensure real-time response to changes in flight conditions.

[0027] Numerical methods are used to solve for the minimum and transverse conditions to obtain the optimal fuel cell power sequence, i.e., the control command. These numerical methods include the gradient method and the conjugate gradient method.

[0028] As another example, the solution to the necessary conditions of PMP can be flexibly selected according to the model accuracy and real-time requirements: (1) Numerical solution (preferred): For high-precision nonlinear models, numerical optimization algorithms such as the gradient method and the conjugate gradient method are used for iterative solution. This method is robust and reliable because it integrates the state and costate equations and calculates the gradient.

[0029] (2) Analytical / Approximate Analytical Solution (Alternative): After reasonably simplifying the model (such as polynomial fitting, piecewise linearization), it can be solved by... H / P fc By setting the value to 0 and combining it with constraints, the optimal solution can be obtained directly, resulting in extremely high computational efficiency.

[0030] S104, control is performed according to the control command; and the real-time status information based on the hydrogen-electric hybrid drone is returned, and the rolling optimization control is performed by using the working condition prediction model to predict the power demand in the future prediction time domain.

[0031] The optimal fuel cell power sequence is { }, the first element in the optimal fuel cell power sequence As an actual control command, it is issued to the fuel cell system. At the next sampling time, the state of the hydrogen-electric hybrid drone is updated, and S101-S103 is repeated to achieve rolling optimization.

[0032] Combination Figure 5 The overall algorithm flowchart of PMP-MPC and Figure 6 The real-time control execution timing diagram shows its online working process as follows: Flight begins, system initialization.

[0033] Enter each control cycle (e.g., once per second): Step 1. Status Acquisition: Read the current battery SOC, flight speed, etc., and acquire the status information of the hydrogen-electric hybrid drone through the sensor system.

[0034] Step 2. Predicting Operating Conditions: Input historical operating conditions into the LSTM prediction model to obtain the future power demand sequence P. req The specific inputs and outputs of the LSTM prediction model are as follows: (1) Input variables: a time series window, which may include historical demand power series, historical flight status (such as altitude, flight speed) and predetermined flight plan instructions.

[0035] (2) Output format: The demand power sequence {P} over a future predicted time domain N. req(k) , Preq(k+1) , ...,P req(k+N)}

[0036] Step 3. Start the PMP-MPC controller (see...) Figure 5 ): (1) Task information analysis: Analyze task requirements based on current status information and prediction information.

[0037] (2) Global optimization: By constructing and solving the optimization problem, the globally optimal control strategy is obtained.

[0038] (3) Local optimization: Combine real-time status to further optimize control commands.

[0039] (4) Optimization parameter update: Adjust optimization parameters in real time based on system feedback.

[0040] Step 4. Instruction Issuance: Issue the first element of the control instruction. It was distributed to the fuel cell system.

[0041] Step 5. Rolling update: At time k+1, return to step 1, and repeat the cycle.

[0042] Step 6. Feedback and Learning: Based on the current control effect and performance data, adjust and optimize parameters in real time to further improve control performance (see...). Figure 6 ).

[0043] This application has the following technical advantages: (1) Global economic efficiency is significantly better than traditional real-time strategies: Due to the introduction of costate equations derived from PMP, the decision-making mechanism of this application theoretically fully considers the impact of future flight mission information on the current moment. Its optimization performance can approach the dynamic programming (DP) result as the performance benchmark, and is significantly better than the equivalent consumption minimization strategy (ECMS) that only considers the current moment and the fuzzy logic control that lacks optimization theory. (2) High computational efficiency, meeting the requirements of online real-time applications: This application finds the optimal solution by solving the differential equation system derived from PMP, which fundamentally avoids the "curse of dimensionality" problem of dynamic programming (DP). At the same time, compared with the complex online planning iteration in traditional nonlinear MPC, this principle-based solution method has a lower computational load and is easier to implement on an airborne computing platform. (3) Stronger robustness and adaptability: This application integrates the rolling time domain and feedback correction mechanism of MPC, which can effectively overcome the impact of model mismatch, external disturbances and prediction errors. It is more adaptable to the complex and ever-changing environment in actual flight than the open-loop PMP strategy.

[0044] In one exemplary embodiment, a computer device is provided, which may be a server or a terminal. The computer device includes a processor, memory, input / output interfaces (I / O), and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is connected to the system bus via the I / O interfaces. The processor of the computer device provides computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The I / O interfaces of the computer device are used for exchanging information between the processor and external devices. The communication interface of the computer device is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements a hydrogen-electric hybrid unmanned aerial vehicle (UAV) energy management method.

[0045] In one exemplary embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0046] In one exemplary embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0047] In one exemplary embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0048] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0049] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).

[0050] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0051] In this application, all actions to acquire signals, information, or data are carried out in compliance with the relevant data protection laws and policies of the country where the location is situated, and with the authorization granted by the owner of the relevant device.

[0052] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0053] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for energy management of a hydrogen-electric hybrid unmanned aerial vehicle, characterized in that, The energy management method for the hydrogen-electric hybrid drone includes: Based on the real-time status information of the hydrogen-electric hybrid drone, the power demand in the future prediction time domain is predicted using a working condition prediction model; the status information includes: battery state of charge, flight altitude, flight airspeed, environmental information, and flight plan instructions; Based on real-time state information, an optimization problem is constructed in each rolling time domain of model predictive control, with the power demand in the future predicted time domain as input and the goal of minimizing total hydrogen consumption. Based on the aforementioned optimization problem, the control commands are obtained using the Pontryagin minimum principle. Control is performed according to the control instructions; and real-time status information based on the hydrogen-electric hybrid drone is returned, and rolling optimization control is performed by using the working condition prediction model to predict the power demand in the future prediction time domain. Based on real-time state information, within each rolling time domain of model predictive control, an optimization problem is constructed with the power demand in the future predicted time domain as input and the goal of minimizing total hydrogen consumption. Specifically, this includes: Using formula Determine the cost function ; in, Sampling time, For future time-domain prediction, for Hydrogen consumption rate at time t, for The state of charge of the battery at any given time. for Reference trajectory at any moment For the terminal cost function, , As a reference trajectory for future predictions in the time domain, , and All are weighting coefficients, with min representing the minimum value. To predict the state of charge of the battery in the future time domain.

2. The energy management method for a hydrogen-electric hybrid unmanned aerial vehicle according to claim 1, characterized in that, Based on the real-time status information of the hydrogen-electric hybrid drone, the power demand in the future forecast time domain is predicted using a working condition prediction model. Previous methods also included: Obtain real-time status information of hydrogen-electric hybrid drones; Based on real-time state information, a mathematical model of a hybrid power system is constructed using a battery equivalent circuit model and UAV dynamics. The mathematical model of the hybrid power system uses the battery state of charge as the state variable, the fuel cell power as the control variable, and the demand power as the disturbance variable.

3. The energy management method for a hydrogen-electric hybrid unmanned aerial vehicle according to claim 1, characterized in that, The operating condition prediction model includes: long short-term memory network, time series model, dynamics-based filtering algorithm, probabilistic model or machine learning regression model.

4. The energy management method for a hydrogen-electric hybrid unmanned aerial vehicle according to claim 1, characterized in that, Based on the optimization problem, the control commands are obtained using the Pontryagin minimum principle, specifically including: Using formula Determine the Hamiltonian function H; Using the formula d(SOC) / dt = H / λ determines the state equation; Using the formula dλ / dt = - H / (SOC) determines the costate equation; Using formula =argmin H determines the minimum condition; Using formula Determine the cross-sectional conditions; Where x is the state variable, u is the control variable, λ is the co-state variable, and t is the time. Let be the hydrogen consumption rate at time t. Let be the battery state of charge at time t. for The reference trajectory at time t, d(SOC) / dt is the rate of change of the battery's state of charge, P fc For fuel cell power, For the optimal fuel cell power, argmin is the function that minimizes it. E / x is the partial derivative of the terminal cost function with respect to the terminal state. H / λ is the partial derivative of H with respect to λ, and dλ / dt is the derivative of λ with respect to t. H / (SOC) is the partial derivative of H with respect to SOC. For optimal fuel cell power, Let λ be the costate variable that makes U optimal.

5. The energy management method for a hydrogen-electric hybrid unmanned aerial vehicle according to claim 4, characterized in that, The process of obtaining control commands based on the optimization problem and using the Pontryagin minimum principle further includes: Numerical methods are used to solve for the minimum condition and the transverse condition.

6. The energy management method for a hydrogen-electric hybrid unmanned aerial vehicle according to claim 5, characterized in that, The numerical methods include the gradient method and the conjugate gradient method.

7. A computer device, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor executes the computer program to implement the energy management method for a hydrogen-electric hybrid unmanned aerial vehicle according to any one of claims 1-6.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the energy management method for the hydrogen-electric hybrid unmanned aerial vehicle as described in any one of claims 1-6.

9. A computer program product, comprising a computer program, characterized in that, When executed by a processor, the computer program implements the energy management method for the hydrogen-electric hybrid unmanned aerial vehicle as described in any one of claims 1-6.

Images