Energy management strategy optimization method for hydrogen-lithium hybrid power system of aircraft

Abstract

The invention discloses an energy management strategy optimization method for an aircraft hydrogen-lithium hybrid power system, and relates to the technical field of energy management. The method comprises the steps of collecting and integrating prior information of flight path planning, load requirements and environmental conditions of an aircraft, and performing preprocessing to form a task profile comprehensive parameter set; a load characteristic model is constructed, a power curve is generated, and power sudden change and prediction characteristics are extracted; based on the feature identification power jump moment and the core task node, combining the characteristics of stable steady-state output of a hydrogen fuel cell and rapid response of a lithium battery, and making a core node energy collaborative distribution strategy; and through closed-loop comparison of the real-time state of the hydrogen-lithium system, the power distribution proportion and the execution time sequence are dynamically corrected. According to the method, the adaptability of the hybrid power system to task requirements and environment changes is improved, and the energy utilization efficiency is optimized.

Description

Technical Field

[0001] This invention relates to the field of energy management technology, and more specifically to an energy management strategy optimization method for an aircraft hydrogen-lithium hybrid power system. Background Technology

[0002] With the diversification of aircraft mission scenarios, the flight process encompasses multiple phases such as takeoff, cruise, and landing, and the energy requirements of each phase differ significantly: the takeoff phase requires instantaneous high power to support power output, the cruise phase requires stable power to maintain flight status, and the landing phase requires gradually reducing power to adapt to deceleration requirements. At the same time, the aircraft is also affected by flight path planning adjustments, the start and stop of payload equipment, and changes in environmental conditions during flight, resulting in dynamic fluctuations in overall power demand.

[0003] Hydrogen-lithium hybrid power systems benefit from the steady-state output characteristics of hydrogen fuel cells and the rapid response characteristics of lithium batteries. However, current energy management strategies for these systems rely on fixed, empirical allocation rules when dealing with the aforementioned dynamic changes. They fail to fully integrate prior information such as flight path planning, load requirements, and environmental conditions required for aircraft missions. Allocation schemes are formulated based on only a single dimension of data, resulting in a low degree of fit between the allocation criteria and the actual flight scenario, making it difficult to adapt to changes in power requirements at different mission stages in advance.

[0004] Therefore, the present invention provides an energy management strategy optimization method for an aircraft hydrogen-lithium hybrid power system. Summary of the Invention

[0005] The purpose of this invention is to provide an energy management strategy optimization method for an aircraft hydrogen-lithium hybrid power system, so as to solve at least one of the above-mentioned problems in the prior art.

[0006] An energy management strategy optimization method for an aircraft hydrogen-lithium hybrid power system includes the following steps:

[0007] Collect the prior information required for the spacecraft mission, and structurally integrate the prior information of the spacecraft to obtain a comprehensive parameter set for the mission profile;

[0008] A load characteristic model of the aircraft is constructed. The comprehensive parameter set is input into the load characteristic model, and the power curve of the aircraft in the future mission cycle is output. Based on the power curve, feature extraction is performed to obtain power mutation feature and power prediction feature.

[0009] Based on power prediction features, feature engineering is used to obtain the power jump time and extract the power change rate. Based on the power change rate and combined with the comprehensive parameter set, the capability pre-allocation processing of task nodes is carried out to identify the core nodes of the task and output the timing sequence of key nodes.

[0010] Based on the core nodes of the mission, and combining the dynamic response characteristics of hydrogen fuel cells and lithium batteries, an energy collaborative allocation strategy for the core nodes of the mission is constructed, and pre-allocated strategy parameters are output.

[0011] The pre-allocated strategy parameters are compared in a closed loop with the real-time status of the lithium hydrogen system to dynamically correct the power allocation ratio and execution timing in the pre-allocation strategy.

[0012] As a further technical solution of the present invention: the prior information includes: trajectory planning information, load requirement information and environmental condition information.

[0013] As a further technical solution of the present invention, the load characteristic model is constructed as follows:

[0014] Based on the load requirement information in the comprehensive parameter set of the mission profile, the core subsystems of propulsion and environmental control of the aircraft are decomposed, and power consumption sub-models adapted to the power consumption characteristics of each subsystem are established respectively.

[0015] The power consumption sub-model of each subsystem is bound to the aircraft mission phase divided in the integrated parameter set to achieve synchronization between the power consumption output of the sub-model and the timing of the actual mission actions.

[0016] Using the time axis of the integrated parameter set as a reference, a unified time granularity is set, and the output values ​​of the power consumption sub-models of each subsystem are superimposed at the same time to obtain the real-time total load power consumption of the entire system, and the power curve of the aircraft in the future mission cycle is plotted.

[0017] As a further technical solution of the present invention, the method for obtaining the real-time total load power consumption of the entire system is as follows:

[0018] The time axis of the comprehensive parameter set is used as the reference, and a uniform time granularity is set.

[0019] The time axis is extracted from the comprehensive parameter set of the task profile and the time granularity is made consistent with the output time step of the power consumption sub-model of each subsystem; the power consumption sub-model of each subsystem is called to obtain the instantaneous power consumption value of each subsystem respectively;

[0020] For each time node with a unified time granularity, the power consumption output values ​​of all subsystems at each time node are accumulated and calculated to obtain the real-time total load power consumption of the entire system at each time node.

[0021] As a further technical solution of the present invention, the feature extraction is performed as follows:

[0022] Extract the power abrupt change points from the power curve, obtain the power change amplitude, duration and triggering scenario corresponding to the power abrupt change points, and integrate the data to obtain power abrupt change characteristics;

[0023] By identifying the changing patterns in the comprehensive parameter set of the mission profile, integrating the parameter-power correlation patterns, historical power curve trends, and periodic nodes, power prediction characteristics are established.

[0024] As a further technical solution of the present invention, the method for obtaining the power jump moment is as follows:

[0025] Extract time-series power data from the power prediction features, and identify abrupt change features based on the time-series power data to obtain the power jump time.

[0026] As a further technical solution of the present invention: the method for identifying the core node of the task is as follows:

[0027] Based on the subsystem capability parameters in the comprehensive parameter set of the mission profile, the energy supply boundary of the hydrogen-lithium system is defined.

[0028] Based on the energy supply boundary, power jump-supply matching degree analysis is performed to screen candidate core nodes;

[0029] Candidate core nodes are screened based on their irreplaceability and energy reserve constraints to obtain the core nodes of the mission.

[0030] As a further technical solution of the present invention, the method for performing the energy reserve constraint correlation screening is as follows:

[0031] Obtain the retained candidate core nodes, and extract the initial hydrogen quantity, initial battery power, and predicted energy consumption data for the entire mission cycle from the comprehensive parameter set of the mission profile. Calculate the energy consumption percentage corresponding to each retained candidate node.

[0032] Candidate core nodes whose energy consumption ratio exceeds the average consumption ratio throughout the entire cycle are selected and determined as core nodes of the task.

[0033] As a further technical solution of the present invention: the energy collaborative allocation strategy is constructed as follows:

[0034] Based on the power demand characteristics of each core node of the mission, allocation rules are formulated in combination with the response characteristics of the two types of energy, and a collaborative energy allocation strategy for the core nodes of the mission is constructed.

[0035] As a further technical solution of the present invention, the closed-loop comparison processing is performed as follows:

[0036] The pre-allocated parameters are compared with the real-time status data in real time to calculate the percentage of deviation between the actual power and the pre-allocated power, and the power deviation rate is obtained.

[0037] The real-time SOC of the lithium battery and the planned remaining amount of the lithium battery pre-allocated are obtained respectively. The deviation rate between the lithium battery SOC and the planned remaining amount is calculated to obtain the remaining deviation degree of the lithium battery SOC.

[0038] Obtain the remaining hydrogen quantity and the planned remaining hydrogen quantity after pre-allocation, respectively, calculate the deviation rate between the hydrogen quantity and the planned remaining hydrogen quantity after pre-allocation, and obtain the deviation degree of the remaining hydrogen quantity.

[0039] The weighted sum of the deviation of the remaining capacity of the lithium battery SOC and the deviation of the remaining capacity of hydrogen is used to obtain the comprehensive deviation.

[0040] Establish a deviation criterion for the output state of the power system. If the power deviation rate and the overall deviation satisfy the deviation criterion, it is determined that there is a deviation in the output state of the power system.

[0041] If a deviation is found, a dynamic correction mechanism is immediately triggered, and the new parameters are then compared with the updated real-time status.

[0042] The beneficial effects of this invention are:

[0043] 1. By collecting three types of prior information—track planning, load requirements, and environmental conditions—and combining them with preprocessing to remove outliers and fill in missing data, and integrating them into a comprehensive parameter set along a unified timeline, a structured data foundation that fits the actual flight scenario can be provided for the construction of subsequent energy management strategies. This reduces the interference of scattered data on subsequent analysis and makes subsequent model building and strategy formulation more data-supported.

[0044] 2. By breaking down core subsystems such as propulsion and environmental control, and constructing adapted power consumption sub-models, combined with power curves generated from comprehensive parameter sets, it is beneficial to obtain power demand changes within future mission cycles. Simultaneously, the extracted power mutation and prediction features help to perceive power fluctuations and trends in advance, providing characteristic basis for subsequent identification of key power nodes. By setting judgment thresholds to identify power jump moments and rates of change, and combining task node information and subsystem capability parameters to formulate pre-allocation rules, energy supply can be initially matched to the power demand differences of different task nodes. Furthermore, by screening core nodes based on multi-dimensional indicators and outputting key timing sequences, energy allocation can be focused on stages that play a crucial role in mission completion, improving the targeting of energy allocation.

[0045] 3. Combining the stable steady-state output of hydrogen fuel cells and the fast response speed of lithium batteries, differentiated energy allocation rules are formulated for different core nodes. This helps to leverage the advantages of both energy sources, ensuring that the different demands of core nodes, such as instantaneous high power and steady-state power, are met with appropriate energy supply. This reduces the limitations of a single energy source in handling complex power demands and provides a strategic framework for efficient energy allocation. By collecting real-time state data such as output power and remaining capacity of the hydrogen-lithium system, comparing them with pre-allocation parameters, and calculating the deviation, dynamic correction is triggered when a deviation occurs. This allows the power allocation ratio and execution timing to adjust according to the actual system state, reducing energy supply mismatch problems caused by system state fluctuations and ensuring that energy allocation is aligned with the real-time operation of the hydrogen-lithium system. Attached Figure Description

[0046] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0047] Figure 1 This is a flowchart of an energy management strategy optimization method for an aircraft hydrogen-lithium hybrid power system provided in an embodiment of the present invention;

[0048] Figure 2 This is a schematic diagram of the structure of an energy management strategy optimization system for an aircraft hydrogen-lithium hybrid power system provided in an embodiment of the present invention. Detailed Implementation

[0049] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0050] Example 1

[0051] like Figure 1 As shown in the figure, the energy management strategy optimization method for a hydrogen-lithium hybrid power system for aircraft provided by this invention specifically includes the following steps:

[0052] S1. Collect the prior information required for the spacecraft mission, and structurally integrate the prior information of the spacecraft to obtain the comprehensive parameter set of the mission profile.

[0053] It should be noted that the prior information of the mission profile includes: trajectory planning information, payload requirement information, and environmental condition information;

[0054] The method for collecting prior information about the spacecraft's mission profile and then structurally integrating this prior information is as follows:

[0055] In some embodiments, the trajectory planning information is obtained by setting the key waypoints (take-off point, cruise point, landing point, etc.) of the aircraft according to the mission objectives using ground mission planning software (such as FlightPlanProV5.0), calculating the planned altitude, speed and time for each segment, and forming preliminary trajectory planning information.

[0056] It should be noted that the aircraft includes core subsystems such as avionics, propulsion, and environmental control.

[0057] Load demand information is obtained by measuring the base power consumption and peak power consumption of each subsystem in real time through power sensors in the power supply circuits of the core subsystems of avionics, propulsion, and environmental control.

[0058] Then, the startup time and duration of the subsystem are recorded synchronously (e.g., satellite communication starts once every 30 minutes and lasts for 8 minutes); finally, the collected power consumption data is bound to the corresponding mission stage in combination with the mission stage of the aircraft (takeoff, cruise, landing) to obtain the load requirement information for each mission stage.

[0059] Environmental condition information is collected in real time by airborne equipment and combined with ground data: for example, the weather radar on the wings detects wind speed, wind direction and turbulence level, and the temperature and pressure sensors on the fuselage measure the ambient temperature and air pressure at the current altitude; the environmental data is aligned with the flight timestamp to ensure that it matches the flight status;

[0060] The systematic processing method implemented in the structural integration phase is as follows:

[0061] Extreme outliers in trajectory planning information, load demand information, and environmental condition information are removed using the 3σ principle, and short-term missing data is completed by linear interpolation; long-term missing data is completed by matching historical data in the same scenario, thus realizing the preprocessing of prior information of the mission profile.

[0062] Based on the prior information of the preprocessed mission profile, with the flight start time as 0, time nodes are divided according to a preset granularity to form a unified time axis.

[0063] For example, 1 second / millisecond is used as the preset granularity;

[0064] Extract the trajectory planning information, load demand information, and environmental condition information for each time point; match by timestamp:

[0065] By mapping the time, flight path planning information, load demand information, and environmental condition information at the same time point, data in four dimensions are obtained and a correlation table is constructed.

[0066] Cross-validation and physical logic verification are performed on the data of the four dimensions of the association table to obtain a structured parameter set containing key parameters;

[0067] The structured parameter set is filtered according to mission requirements, core parameters are selected, data is associated according to mission time / stage, quantitative integration is combined with the spacecraft mission objectives, and after verification, it is summarized according to the profile framework and transformed into a comprehensive parameter set of the mission profile.

[0068] For example, the core parameters required for the mission (such as the 6000m altitude during the cruise phase) are selected from the structured parameter set; then, the corresponding parameters are bound to the phases according to the mission sequence of takeoff, cruise, observation, and landing; finally, the parameter logic is verified (such as the takeoff speed not exceeding the upper limit of the aircraft type), and the parameters are summarized according to the mission phase-track-load-environment framework to form the profile integrated parameter set for the mission.

[0069] S2. Construct a load characteristic model of the aircraft, input the comprehensive parameter set into the load characteristic model, and output the power curve of the aircraft in the future mission cycle. Based on the power curve, perform feature extraction to obtain power mutation features and power prediction features.

[0070] The process of constructing the aircraft load characteristic model is as follows:

[0071] S201. Based on the load requirement information in the comprehensive parameter set of the mission profile, the core subsystems of propulsion and environmental control of the aircraft are decomposed, and power consumption sub-models adapted to the power consumption characteristics of each subsystem are established respectively.

[0072] Preferably, the method for establishing a power consumption sub-model adapted to the power consumption characteristics of each subsystem is as follows:

[0073] Propulsion sub-model: associates trajectory planning information (altitude, speed, atmospheric density) with the integrated parameter set;

[0074] Set the aerodynamic drag-power coupling equation: P 推 =0.5×ρ×v³×S×Cx to obtain the steady-state power consumption P of the propulsion system. 推 ;

[0075] Where ρ is the atmospheric density in the integrated parameter set, v is the airspeed, S is the wing area of ​​the aircraft, Cx is the drag coefficient, and 0.5 is the inherent coefficient of the dynamic pressure formula in the aerodynamic drag calculation;

[0076] For example, during the cruise phase, at an altitude of 8000m (ρ=0.54kg / m³) and a speed of 420km / h, the steady-state propulsion power consumption is calculated to be 180kW using the aerodynamic drag-power coupling formula. During the takeoff phase, at an acceleration speed of 270km / h, the peak propulsion power consumption is calculated to be 290kW.

[0077] Environmental control sub-model: Based on the environmental conditions (temperature, humidity) in the comprehensive parameter set, an adaptive adjustment unit is designed, and power consumption adjustment rules corresponding to temperature and humidity are set. Under specific environmental conditions, the output environmental control power consumption can be automatically calculated.

[0078] For example, based on the environmental conditions in the comprehensive parameter set, the basic power consumption is set to 8kW when the temperature is ≥25℃, and increases by 1.5kW for every 10℃ below the temperature, and increases by an additional 1kW when the humidity exceeds 60%. When the comprehensive parameter set records the ambient temperature of -32℃ and humidity of 55% during the landing phase, the output environmental control power consumption is automatically calculated to be 13.8kW.

[0079] S202. Bind the power consumption sub-model of each subsystem to the aircraft mission phase divided in the integrated parameter set to realize the synchronization of the sub-model power consumption output with the actual mission action timing.

[0080] Preferably, based on the task phase division and timing of the trajectory planning information in the integrated parameter set, the sub-model activation rules are designed as follows:

[0081] The sub-model activation rules include:

[0082] Takeoff phase (0-10 minutes): Simultaneously activate the peak power consumption mode of the propulsion sub-model (output 270-290kW) and the start-up transition mode of the environmental control sub-model (2kW higher than the basic power consumption for the first 2 minutes, and returns to the basic value for the next 8 minutes) to match the high power demand of the takeoff phase and the start-up characteristics of the subsystem;

[0083] Climbing phase (10-30 minutes): The propulsion sub-model switches to a gradual power consumption mode (i.e., gradually reducing from 290kW to 180kW), and the environmental control sub-model gradually increases power consumption based on the temperature drop caused by the increase in altitude and with reference to the ambient temperature gradient of the integrated parameter set.

[0084] Cruise phase (30-150 minutes): The propulsion sub-model activates steady-state power consumption mode (i.e., output 180kW), and the environmental control sub-model enters constant temperature stable mode (fixed temperature -34℃, output 12kW).

[0085] During the descent phase (150–180 minutes): the propulsion sub-model switches to a decreasing power consumption mode (from 180kW to 120kW), and the environmental control sub-model gradually reduces power consumption as altitude decreases and temperature increases;

[0086] S203. Using the time axis of the integrated parameter set as a reference, set a unified time granularity, superimpose the output values ​​of the power consumption sub-models of each subsystem at the same time to obtain the real-time total load power consumption of the entire system, and draw the power curve of the aircraft in the future mission cycle.

[0087] The method for obtaining the real-time total load power consumption of the entire system is as follows:

[0088] The time axis is extracted from the comprehensive parameter set of the task profile and the time granularity is consistent with the output time step of the power consumption sub-model of each subsystem.

[0089] Call the power consumption sub-model of each subsystem to obtain the instantaneous power consumption value of each subsystem;

[0090] For each time node under a unified time granularity, the power consumption output values ​​of all subsystems at each time node are accumulated and calculated to obtain the real-time total load power consumption of the entire system at each time node;

[0091] The power curve is obtained in the following way:

[0092] Traversing all time points, a real-time total load power consumption time series of the entire system is formed. Using the time axis of the integrated parameter set as the horizontal axis and the real-time total load power consumption of the entire system as the vertical axis, the total load power consumption value of each time point is mapped onto the coordinate plane. By linearly connecting all data points on the coordinate plane, the power curve of the aircraft in the entire mission cycle is generated.

[0093] Information extracted from the power curve is used to obtain power mutation characteristics and power prediction characteristics;

[0094] The power mutation feature is obtained in the following way:

[0095] Based on the generated power curves for the future mission cycle of the aircraft, obtain the time-series power data for the entire cycle;

[0096] Set a power change judgment threshold (the threshold is determined based on the power fluctuation tolerance range of the core subsystem of the aircraft), traverse the time-series power data, calculate the power difference between the current moment and the previous moment, and mark the current moment as a power change point if the absolute value of the difference reaches the judgment threshold.

[0097] For each marked power mutation point, the corresponding power change amplitude (the power difference between the current moment and the previous moment), duration (the time interval from the start of the mutation to the power recovery and stabilization) and triggering scenario (matching the flight mission phase corresponding to the mutation point with the comprehensive parameter set of the mission profile, such as takeoff switching and load equipment startup) are extracted, and the above data are integrated to form power mutation features.

[0098] The power prediction features are obtained in the following way:

[0099] Key parameters affecting power requirements (such as ambient temperature, speed, and altitude) are extracted from the comprehensive parameter set of the mission profile. Combined with historical mission data, the changing patterns of key parameters (such as the inflection point of temperature change with altitude and the adjustment point of speed) are identified.

[0100] By calling the power models of core subsystems such as propulsion and environmental control of the aircraft, the correlation points between key parameters in the model and the power output of the subsystems (such as the correspondence between changes in airspeed and the power output of the propulsion subsystem) are extracted.

[0101] Simultaneously, trend nodes (such as the inflection point of power increase during takeoff) and periodic nodes (such as the power stabilization node at the same stage in repeated missions) of power changes are extracted from the power curves of historical missions; key parameter change patterns, parameter-power correlation patterns, historical power curve trends and periodic nodes are integrated to form power prediction features.

[0102] Example 2

[0103] like Figure 1 As shown in the figure, the energy management strategy optimization method for a hydrogen-lithium hybrid power system for aircraft provided by this invention specifically includes the following steps:

[0104] S3. Based on the power prediction features, the power jump moment is obtained and the power change rate is extracted through feature engineering. Based on the power change rate and the comprehensive parameter set, the capability pre-allocation processing of the task nodes is carried out, the core nodes of the task are identified and the timing of the key nodes is output.

[0105] The methods for obtaining the power jump moment and the power change rate are as follows:

[0106] Extract time-series power data from power prediction features, identify abrupt change features based on time-series power data, and obtain the power jump time.

[0107] Understandably, the method for identifying abrupt change features is as follows: to determine the moment of a power jump, a power jump judgment threshold must first be set, which is determined based on the maximum tolerable power fluctuation value of the subsystem in the comprehensive parameter set. The time series power data is traversed, and the difference between the predicted power at the current moment and the predicted power 1 second ago is calculated. If the absolute value of the difference is not lower than the jump threshold, the current moment is marked as the moment of a power jump.

[0108] The method for obtaining the maximum tolerable power fluctuation value of a subsystem is as follows: based on the parameters of the subsystem components, the minimum tolerable value of each subsystem is taken as the basis, and then combined with the mission safety factor of the aircraft (such as 1.2) is used to calculate the power jump judgment threshold = minimum tolerable power fluctuation value of the subsystem × 1.2.

[0109] For each power jump moment, calculate the power change rate and record the mission stage corresponding to the jump moment (matched from the mission stage-time axis mapping relationship in the integrated parameter set, such as the jump moment 00:05 corresponding to the takeoff stage).

[0110] The pre-allocation method for task node capabilities is as follows:

[0111] In some embodiments, task node information and subsystem capability parameters are first extracted from the comprehensive parameter set; the power change rate is divided into three levels: high, medium, and low (high: change rate > 10kW / s, medium: 3-10kW / s, low: < 3kW / s). Based on the predicted power demand of the task nodes, a capacity pre-allocation rule is formulated: nodes with high change rate (such as the takeoff phase 00:02-00:05) are given priority in being allocated full-load generator output capacity (the maximum generator output power of 200kW is extracted from the comprehensive parameter set), and 10% battery redundancy capacity is reserved.

[0112] At nodes with medium rate of change (such as the satellite communication start time of the cruise phase at 00:30), 70% of the generator output capacity is allocated, along with 5% battery redundancy; at nodes with low rate of change (such as the stable cruise phase), only 50% of the generator output capacity is allocated; at the same time, referring to the subsystem correlation relationship in the integrated parameter set, the pre-allocated capacity is split according to the power consumption ratio of the subsystem to ensure that the capacity supply of each subsystem matches the power demand fluctuations.

[0113] It should be noted that the mission node information includes takeoff, climb, cruise, landing, and special mission nodes such as circling observation. Each node is marked with its planned duration and associated subsystems. The subsystem capability parameters are mainly the battery rated capacity, the generator's maximum output power, and the subsystem's basic power consumption.

[0114] The core node identification and key information output methods for the task are as follows:

[0115] S301. Define the energy supply boundary of the hydrogen-lithium system based on the subsystem capability parameters in the comprehensive parameter set of the mission profile.

[0116] Preferably, based on the subsystem capability parameters (battery rated capacity, generator maximum output power) in the comprehensive parameter set of the task profile, combined with the steady-state output upper limit of the hydrogen fuel cell, the instantaneous response peak of the lithium battery, and the synergistic supply limit of the two, an energy supply boundary model of the hydrogen-lithium hybrid system is constructed to determine the maximum supply power range of the system at different time points, which serves as the energy supply boundary of the lithium system.

[0117] S302. Based on the energy supply boundary, perform power jump-supply matching degree analysis to screen candidate core nodes;

[0118] Preferably, the power change rate at the power jump moment is correlated with the energy supply boundary model to calculate the matching difference between the power demand and the system supply power at each jump node; if the matching difference is in the critical range of the supply boundary (i.e., the power demand is close to the system supply limit), then the jump node and the associated task periods before and after it are marked as candidate core nodes.

[0119] S303. Select the core nodes of the task by considering their irreplaceability and energy reserve constraints.

[0120] Preferably, by combining the relationship between the mission objectives and subsystems in the comprehensive parameter set of the mission profile, the power demand characteristics of each candidate core node are analyzed. If the power fluctuation characteristics of the node (such as instantaneous high power or continuous stable power) cannot be met independently by a single hydrogen energy source or a single lithium energy source, and the mission must rely on the synergistic supply of hydrogen and lithium to ensure the normal progress of the mission, then the irreplaceable screening is carried out, and the candidate core nodes retained after the screening are extracted.

[0121] For example, the non-substitutability screening method is as follows: instantaneous power demand > peak output power of lithium battery (e.g., when the peak output power of lithium battery is 200kW, the instantaneous demand of node is 220kW); continuous power demand > rated output power of hydrogen fuel cell (e.g., when the rated output power of hydrogen fuel cell is 180kW, the continuous demand of node is 200kW); power fluctuation frequency > hydrogen fuel cell response frequency (e.g., when the response frequency of hydrogen fuel cell is 0.5Hz, the fluctuation frequency of node is 1Hz); if any of these conditions are met, it is determined that the demand cannot be met independently by a single energy source, and the candidate core node is retained.

[0122] Extract the initial hydrogen quantity, initial battery capacity, and predicted energy consumption data for the entire mission cycle from the comprehensive parameter set of the mission profile, and calculate the energy consumption ratio corresponding to each retained candidate node.

[0123] Candidate core nodes whose energy consumption ratio exceeds the average consumption ratio throughout the entire cycle are selected and ultimately determined to be the core nodes of the mission.

[0124] The method for obtaining the timing sequence of key nodes is as follows:

[0125] Based on the timeline of the task profile, the planned start time and duration of the core nodes are extracted. Combined with the coordinated response time requirements in the energy supply boundary model, the energy connection logic between the core nodes and the preceding and following task nodes is sorted out and integrated in chronological order to form the key node time sequence containing supply response constraints.

[0126] S4. Based on the core node of the mission and combined with the dynamic response characteristics of hydrogen fuel cells and lithium batteries, construct an energy collaborative allocation strategy for the core node of the mission and output the pre-allocated strategy parameters.

[0127] The energy collaborative allocation strategy for constructing the core nodes of the task is as follows:

[0128] Extract the core nodes of the mission from the comprehensive parameters of the mission profile, and combine them with the power curve data corresponding to each node to clarify the power demand characteristics of each core node (such as the instantaneous high power of 290kW required at the initial takeoff node).

[0129] The inherent characteristics of the two types of energy are summarized as follows: Hydrogen fuel cells have stable steady-state output but slow response speed, making them suitable for continuous and stable loads; lithium batteries have fast response speed, making them suitable for handling instantaneous power surges.

[0130] Based on the power demand characteristics of each core node of the task, and combined with the response characteristics of the two types of energy, allocation rules are formulated to construct an energy collaborative allocation strategy for the core nodes of the task.

[0131] For example, at the initial takeoff node (instantaneous high power 290kW): the lithium battery handles the 110kW instantaneous surge, while the hydrogen fuel cell handles the 180kW base stable portion; at the cruise start node (steady-state 180kW): the hydrogen fuel cell independently handles 180kW, while the lithium battery is in standby peak-shaving mode; at the landing switching node (power decreases from 180kW to 120kW): the hydrogen fuel cell's dominant power decreases (gradually from 180kW to 120kW), and the lithium battery assists in compensating for minor fluctuations during the decrease (such as instantaneous adjustments within ±5kW).

[0132] The method for outputting the pre-allocated strategy parameters is as follows:

[0133] The power ratio of hydrogen fuel cells at each core node (e.g., 62% at the initial takeoff node and 100% at the cruise node), the lithium battery response trigger threshold (e.g., lithium battery intervention when the power change rate exceeds 100kW / s), and the power switching time window (e.g., the hydrogen fuel cell-dominated reduction is completed within 30 seconds when the power drops from 180kW to 120kW at the landing node) are output as pre-allocated strategy parameters.

[0134] S5. Perform closed-loop comparison processing on the pre-allocated strategy parameters in conjunction with the real-time status of the lithium hydrogen system to dynamically correct the power allocation ratio and execution timing in the pre-allocation strategy.

[0135] The method for collecting real-time status data of the lithium hydrogen system is as follows:

[0136] Install corresponding monitoring devices on hydrogen fuel cells to monitor their current output power, remaining hydrogen quantity, and operating temperature in real time.

[0137] Install corresponding monitoring devices on lithium batteries to detect their current output power, remaining power, and charge / discharge rate in real time.

[0138] The monitoring device transmits the detected data to the data acquisition module in the system in real time. The acquisition module then aggregates and integrates the data to ensure that the data is updated at a set frequency (such as once per second) to form real-time status data that can be directly used for comparison.

[0139] The closed-loop comparison process is as follows:

[0140] The pre-allocated parameters are compared with the real-time status data in real time to calculate the percentage of deviation between the actual power and the pre-allocated power, and the power deviation rate is obtained.

[0141] The real-time SOC of the lithium battery and the planned remaining amount of the lithium battery pre-allocated are obtained respectively. The deviation rate between the lithium battery SOC and the planned remaining amount is calculated to obtain the remaining deviation degree of the lithium battery SOC.

[0142] Obtain the remaining hydrogen quantity and the planned remaining hydrogen quantity after pre-allocation, respectively, calculate the deviation rate between the hydrogen quantity and the planned remaining hydrogen quantity after pre-allocation, and obtain the deviation degree of the remaining hydrogen quantity.

[0143] The weighted sum of the deviation of the remaining capacity of the lithium battery SOC and the deviation of the remaining capacity of hydrogen is used to obtain the comprehensive deviation.

[0144] It should be noted that the weight of the overall deviation is derived from the preset output ratio of the two energy sources in the hydrogen-lithium hybrid system. If the preset hydrogen output ratio is α and the lithium output ratio is β (α+β=1), then the weight of the hydrogen balance deviation is α and the weight of the lithium battery SOC balance deviation is β. For example, if the preset hydrogen output ratio is 70% and lithium output ratio is 30% during the cruise phase, then the overall deviation = 0.7 × hydrogen balance deviation + 0.3 × lithium battery SOC balance deviation.

[0145] Establish deviation criteria for the output state of the power system. If the power deviation rate and the overall deviation meet the deviation criteria, it is determined that there is a deviation in the output state of the power system. If they do not meet the criteria, the changes in the power deviation rate and the overall deviation are continuously monitored.

[0146] For example, the deviation criterion can be established as follows: Takeoff phase: power deviation rate ±5%, overall deviation 0.06.

[0147] Cruise phase: Power deviation rate ±8%, overall deviation 0.10;

[0148] Landing phase: Power deviation rate ±6%, overall deviation 0.07;

[0149] Judgment logic: If the power deviation in any stage exceeds the threshold and lasts for ≥0.5 seconds, or if the overall deviation meets the standard and the two deviations exceed the 50% threshold, then a deviation is judged;

[0150] If a deviation is found, a dynamic correction mechanism is immediately triggered. After correction, the new parameters are compared with the updated real-time state again, forming a cycle of pre-allocation - real-time comparison - correction - re-comparison to ensure that the parameters always match the actual state of the system.

[0151] The verification method for the corrected data is as follows:

[0152] Calculate the corrected deviation rate = (corrected real-time power - target power) / target power × 100%, and require ≤ ±3%;

[0153] For example, if the corrected hydrogen fuel power is adjusted from 170kW to 175kW, with a target of 180kW, the deviation rate is -2.8%.

[0154] The effect of time series correction is evaluated using mean squared error (MSE). The MSE is calculated and is required to be ≤5 (the smaller the value, the better the synchronicity).

[0155] For example, taking the landing phase as an example, the initial pre-allocated parameters are that the hydrogen fuel cell will decrease from 180kW to 120kW within 30 seconds (deceleration rate of 2kW / second), and the lithium battery will be in standby mode (compensation ±5kW, SOC≥70%). At the 5th second, the real-time power of the hydrogen fuel is collected at 175kW (the pre-allocated power should decrease to 170kW in the 5th second, which is 5kW behind schedule, a deviation of 50% exceeding the 20% standard), and the lithium battery SOC is 78% (above the threshold). The hydrogen fuel deceleration rate is then corrected to 3kW / second. At the 10th second, the data is collected and compared again. The hydrogen fuel power has decreased to 160kW (meeting the requirement of a 15kW decrease in 5 seconds under the new rate), and the lithium battery SOC is 76%. If there is no deviation, the process continues. Subsequently, the process of collecting real-time data every second, comparing it with the current parameters, correcting any deviations, and maintaining the current parameters is repeated. This forms a closed-loop comparison cycle to ensure that the parameters always match the actual state of the system.

[0156] The dynamic adjustment of the power allocation ratio and execution timing in the pre-allocation strategy is as follows:

[0157] Obtain the allocation ratio and execution sequence of hydrogen and lithium from historical flight logs, as well as the current strategy optimization dataset constructed by comparing the remaining deviation of lithium battery SOC with the remaining deviation of hydrogen.

[0158] A policy optimization model is constructed based on the random forest algorithm. The policy optimization dataset is input into the policy optimization model, and the output is the power allocation ratio and execution time sequence.

[0159] Those skilled in the art will understand that the strategy optimization model constructed based on the random forest algorithm, inputting the strategy optimization dataset into the model, and outputting the power allocation ratio and execution sequence are as follows: using the hydrogen-lithium power allocation ratio and execution sequence from historical flight logs as labels, and the lithium battery SOC deviation, hydrogen balance deviation, mission phase (takeoff / cruise / landing), and real-time power deviation rate as feature variables, after removing outliers, the model is divided into a training set and a validation set in a 7:3 ratio; the number of random forest decision trees is set to 100, and the minimum number of samples for node splitting is 5; the model is trained using the training set, and the model accuracy is verified using the validation set with mean squared error (MSE≤3); the real-time collected strategy optimization dataset is input into the trained model, and the model outputs the hydrogen-lithium power allocation ratio (e.g., 62% hydrogen / 38% lithium during takeoff) and execution sequence (e.g., lithium battery response delay ≤0.2 seconds) adapted to the current mission phase based on the correlation between feature variables and historical best cases.

[0160] Adjustments are made to the aircraft based on the output power allocation ratio and execution timing.

[0161] The revised power allocation ratios and execution sequences are compiled and used as the basis for real-time execution of the energy system, while the reasons for the revisions are recorded.

[0162] Example 3

[0163] like Figure 2 As shown, an energy management strategy optimization system for an aircraft hydrogen-lithium hybrid power system further includes the following modules:

[0164] Parameter acquisition module: used to collect the prior information required for the spacecraft mission, and to structurally integrate the prior information of the spacecraft to obtain a comprehensive parameter set of the mission profile;

[0165] Feature extraction module: used to build the load characteristic model of the aircraft. The comprehensive parameter set is input into the load characteristic model and the power curve of the aircraft in the future mission cycle is output. Based on the power curve, feature extraction is performed to obtain power change feature and power prediction feature.

[0166] Node Analysis Module: Based on power prediction features, it uses feature engineering to obtain the power jump moment and extract the power change rate. Based on the power change rate and combined with the comprehensive parameter set, it performs pre-allocation of task node capabilities, identifies core task nodes, and outputs the timing of key nodes.

[0167] Strategy Allocation Module: Based on the core nodes of the task and combined with the dynamic response characteristics of hydrogen fuel cells and lithium batteries, this module constructs a collaborative energy allocation strategy for the core nodes of the task and outputs pre-allocated strategy parameters.

[0168] Dynamic correction module: Used to perform closed-loop comparison processing of pre-allocated strategy parameters with the real-time status of the lithium hydrogen system, and dynamically correct the power allocation ratio and execution timing in the pre-allocation strategy.

[0169] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the claims of this invention should still fall within the patent coverage of this invention.

Claims

1. A method for optimizing energy management strategies in an aircraft hydrogen-lithium hybrid power system, characterized in that, Includes the following steps: Collect the prior information required for the spacecraft mission, and structurally integrate the prior information of the spacecraft to obtain a comprehensive parameter set for the mission profile; A load characteristic model of the aircraft is constructed. The comprehensive parameter set is input into the load characteristic model, and the power curve of the aircraft in the future mission cycle is output. Based on the power curve, feature extraction is performed to obtain power mutation feature and power prediction feature. Based on power prediction features, feature engineering is used to obtain the power jump time and extract the power change rate. Based on the power change rate and combined with the comprehensive parameter set, the capability pre-allocation processing of task nodes is carried out to identify the core nodes of the task and output the timing sequence of key nodes. Based on the core nodes of the mission, and combining the dynamic response characteristics of hydrogen fuel cells and lithium batteries, an energy collaborative allocation strategy for the core nodes of the mission is constructed, and pre-allocated strategy parameters are output. The pre-allocated strategy parameters are compared in a closed loop with the real-time status of the lithium hydrogen system to dynamically correct the power allocation ratio and execution timing in the pre-allocation strategy.

2. The energy management strategy optimization method for an aircraft hydrogen-lithium hybrid power system according to claim 1, characterized in that, The prior information includes: trajectory planning information, load requirement information, and environmental condition information.

3. The energy management strategy optimization method for an aircraft hydrogen-lithium hybrid power system according to claim 1, characterized in that, The load characteristic model is constructed as follows: Based on the load requirement information in the comprehensive parameter set of the mission profile, the core subsystems of propulsion and environmental control of the aircraft are decomposed, and power consumption sub-models adapted to the power consumption characteristics of each subsystem are established respectively. The power consumption sub-model of each subsystem is bound to the aircraft mission phase divided in the integrated parameter set to achieve synchronization between the power consumption output of the sub-model and the timing of the actual mission actions. Using the time axis of the integrated parameter set as a reference, a unified time granularity is set, and the output values ​​of the power consumption sub-models of each subsystem are superimposed at the same time to obtain the real-time total load power consumption of the entire system, and the power curve of the aircraft in the future mission cycle is plotted.

4. The energy management strategy optimization method for an aircraft hydrogen-lithium hybrid power system according to claim 3, characterized in that, The method for obtaining the real-time total load power consumption of the entire system is as follows: The time axis of the comprehensive parameter set is used as the reference, and a uniform time granularity is set. The time axis is extracted from the comprehensive parameter set of the task profile and the time granularity is consistent with the output time step of the power consumption sub-model of each subsystem. Call the power consumption sub-model of each subsystem to obtain the instantaneous power consumption value of each subsystem; For each time node with a unified time granularity, the power consumption output values ​​of all subsystems at each time node are accumulated and calculated to obtain the real-time total load power consumption of the entire system at each time node.

5. The energy management strategy optimization method for an aircraft hydrogen-lithium hybrid power system according to claim 1, characterized in that, The feature extraction is performed as follows: Extract the power abrupt change points from the power curve, obtain the power change amplitude, duration and triggering scenario corresponding to the power abrupt change points, and integrate the data to obtain power abrupt change characteristics; By identifying the changing patterns in the comprehensive parameter set of the mission profile, integrating the parameter-power correlation patterns, historical power curve trends, and periodic nodes, power prediction characteristics are established.

6. The energy management strategy optimization method for an aircraft hydrogen-lithium hybrid power system according to claim 1, characterized in that, The method for obtaining the power jump moment is as follows: Extract time-series power data from the power prediction features, and identify abrupt change features based on the time-series power data to obtain the power jump time.

7. The energy management strategy optimization method for an aircraft hydrogen-lithium hybrid power system according to claim 1, characterized in that, The method for identifying the core nodes of the task is as follows: Based on the subsystem capability parameters in the comprehensive parameter set of the mission profile, the energy supply boundary of the hydrogen-lithium system is defined. Based on the energy supply boundary, power jump-supply matching degree analysis is performed to screen candidate core nodes; Candidate core nodes are screened based on their irreplaceability and energy reserve constraints to obtain the core nodes of the mission.

8. The energy management strategy optimization method for an aircraft hydrogen-lithium hybrid power system according to claim 7, characterized in that, The method for performing the energy reserve constraint correlation screening is as follows: Obtain the retained candidate core nodes, and extract the initial hydrogen quantity, initial battery power, and predicted energy consumption data for the entire mission cycle from the comprehensive parameter set of the mission profile. Calculate the energy consumption percentage corresponding to each retained candidate node. Candidate core nodes whose energy consumption ratio exceeds the average consumption ratio throughout the entire cycle are selected and determined as core nodes of the task.

9. The energy management strategy optimization method for an aircraft hydrogen-lithium hybrid power system according to claim 1, characterized in that, The energy collaborative allocation strategy is constructed as follows: Based on the power demand characteristics of each core node of the mission, allocation rules are formulated in combination with the response characteristics of the two types of energy, and a collaborative energy allocation strategy for the core nodes of the mission is constructed.

10. The energy management strategy optimization method for an aircraft hydrogen-lithium hybrid power system according to claim 1, characterized in that, The closed-loop comparison process is performed as follows: The pre-allocated parameters are compared with the real-time status data in real time to calculate the percentage of deviation between the actual power and the pre-allocated power, and the power deviation rate is obtained. The real-time SOC of the lithium battery and the planned remaining amount of the lithium battery pre-allocated are obtained respectively. The deviation rate between the lithium battery SOC and the planned remaining amount is calculated to obtain the remaining deviation degree of the lithium battery SOC. Obtain the remaining hydrogen quantity and the planned remaining hydrogen quantity after pre-allocation, respectively, calculate the deviation rate between the hydrogen quantity and the planned remaining hydrogen quantity after pre-allocation, and obtain the deviation degree of the remaining hydrogen quantity. The weighted sum of the deviation of the remaining capacity of the lithium battery SOC and the deviation of the remaining capacity of hydrogen is used to obtain the comprehensive deviation. Establish a deviation criterion for the output state of the power system. If the power deviation rate and the overall deviation satisfy the deviation criterion, it is determined that there is a deviation in the output state of the power system. If a deviation is found, a dynamic correction mechanism is immediately triggered, and the new parameters are then compared with the updated real-time status.

Images