A multi-source aviation electric propulsion power system frequency domain decoupling and dynamic optimization energy management method for external environment disturbance

Abstract

The application discloses a kind of multi-source aviation electric propulsion power systems frequency domain decoupling and dynamic optimization energy management methods for external environment disturbance, belong to aviation hybrid electric propulsion system control technical field, first obtain flight state, environmental disturbance and propulsion system operating parameter, combine with propulsion model and solve direct current bus total power demand in real time;Total power demand frequency domain decoupling, separate low-frequency reference power and high-frequency transient power and construct disturbance intensity characteristic quantity;According to the dynamic response capability of slow response main power source, a soft constraint with the rate of change of slack variable is constructed, and a power boundary is constructed according to the state of charge of fast response auxiliary power source;An optimization objective function containing a high-frequency power tracking cost term is established, the weight coefficient is adjusted online according to the disturbance intensity characteristic quantity, and the optimal power distribution result is solved under the constraint;The optimal result is converted into control command for execution.The application improves the energy supply safety and control robustness of multi-source aviation electric propulsion power system in complex flight environment.

Description

Technical Field

[0001] This invention belongs to the field of control technology for aviation hybrid electric propulsion systems, and particularly relates to a frequency domain decoupling and dynamic optimization energy management method for multi-source aviation electric propulsion power systems oriented towards external environmental disturbances. Background Technology

[0002] In multi-source aero-electric propulsion systems, a combined power supply configuration consisting of a slow-response main power source and a fast-response auxiliary power source has become an important technological approach. In such systems, the slow-response main power source typically provides the primary continuous power output, while the fast-response auxiliary power source handles transient power compensation and dynamic bus support. During flight, aircraft are affected by factors such as gusts, turbulence, wind shear, and changes in maneuvering conditions, causing propulsion loads to fluctuate rapidly with the external environment and flight status, manifesting as continuous changes in DC bus power demand. In a preferred application, the slow-response main power source can be a fuel cell system, an eddy current generator system, or other continuous power source, while the fast-response auxiliary power source can be a lithium battery system, a supercapacitor system, or other high-dynamic-response energy storage unit.

[0003] Model predictive control (MRC) methods can comprehensively handle power balance, state of charge evolution, and power source operating constraints within the prediction time domain. Most methods directly use total power demand as the optimization input, simultaneously solving for the output power of the slow-response active power source, the output power of the fast-response auxiliary power source, and related state variables within a unified objective function. While this approach can improve the coordinated control capability of multiple power sources, it does not decouple the low-frequency reference power and high-frequency disturbance power in the total power demand at the input side. When external disturbances intensify, the optimizer may still drive the slow-response active power source to participate in high-frequency power regulation to reduce the overall power tracking error. Slow-response active power sources have inherent dynamic lag; their operating characteristics make them more suitable for handling gradually changing power outputs. Frequent tracking of high-frequency power commands can cause reactant supply mismatch, increased voltage fluctuations, and consequently, accelerated component aging, affecting system lifespan.

[0004] Data-driven optimization methods, such as deep reinforcement learning, possess strong expressive power in nonlinear system control and decision-making under complex operating conditions. They typically establish the mapping relationship between state variables and control variables through sample training, with low dependence on precise mechanistic models. However, aerospace propulsion systems have high requirements for the determinism, verifiability, and safety under extreme conditions of control boundaries. Data-driven methods struggle to directly provide strict physical boundary guarantees when training sample coverage is insufficient or actual disturbances deviate from the training distribution. Under strong disturbance conditions, the controller may generate control commands exceeding the power variation capability of the slow-response active power source or the charging and discharging capability of the fast-response auxiliary power source, increasing the risk of power source over-limit operation.

[0005] Decomposing power demand using signal processing methods such as low-pass filtering and wavelet analysis, and combining this with optimized allocation or state feedback control for energy management, can identify different frequency components in the power demand to a certain extent, reducing the direct impact of high-frequency disturbances on the main power source. However, the signal decomposition stage and subsequent control stages are usually relatively independent. The decomposition results are mostly used to assist in correcting the power allocation strategy, rather than serving as the core adjustment variable in a unified optimization framework. When the disturbance intensity changes, the controller lacks a mechanism to adjust the control weights online based on the high-frequency disturbance level, making it difficult to adaptively respond to external environmental shocks. Under continuous strong disturbances, fast-response auxiliary power sources may also lose their adjustment margin due to output saturation or near-limit state of charge. Existing methods typically lack controlled overrun and safety fallback mechanisms for extreme operating conditions, making it difficult to balance the lifespan protection of slow-response main power sources with the continuity of bus power supply.

[0006] In summary, existing technologies lack an energy management method for multi-source aero-electric propulsion systems accommodating external environmental disturbances. This method should effectively decouple the high-frequency transient components and low-frequency reference components in the total power demand at the control input level, dynamically adjust the fast-response auxiliary power source state and physical boundaries during the optimization process, and retain controlled safety fallback capabilities under extreme conditions to balance the protection requirements of the slow-response active power source and the continuous power supply requirements for flight. Summary of the Invention

[0007] To address the aforementioned technical problems, this invention provides a frequency domain decoupling and dynamic optimization energy management method for a multi-source aero-electric propulsion system oriented towards external environmental disturbances. The multi-source aero-electric propulsion system includes a slow-response active power source system, a fast-response auxiliary power source system, a DC bus, a propulsion motor load, a unidirectional power conversion device, a bidirectional power conversion device, and an energy management controller, comprising: The system acquires the aircraft's flight status parameters, external environmental disturbance parameters, and propulsion system operating parameters. Based on the flight status parameters, external environmental disturbance parameters, and propulsion system operating parameters, and combined with the propulsion model, it calculates the total power demand of the DC bus in real time. The total power demand is decoupled in the frequency domain to separate the low-frequency reference power and the high-frequency transient power, and a disturbance intensity characterization quantity is constructed based on the high-frequency transient power. The output power change rate constraint is constructed based on the dynamic response capability of the slow-response active power source system. Relaxation variables are introduced into the output power change rate constraint to form a controlled soft constraint. The output power boundary is constructed based on the state of charge and operating state of the fast-response auxiliary power source system. An optimization objective function containing a high-frequency power tracking cost term is established in the prediction time domain. The weight coefficients of the high-frequency power tracking cost term are adjusted online according to the disturbance intensity characterization quantity. Under the joint constraints of the controlled soft constraint and the output power boundary, the optimization objective function is solved to obtain the optimal power allocation result at the current time. The optimal power allocation result is converted into control commands for the unidirectional power conversion device and the bidirectional power conversion device, and then sent to the slow-response active power source system and the fast-response auxiliary power source system for execution, respectively, to achieve coordinated power supply.

[0008] Optionally, the aircraft's flight status parameters, external environmental disturbance parameters, and propulsion system operating parameters are acquired. Based on the flight status parameters, external environmental disturbance parameters, and propulsion system operating parameters, combined with the propulsion model, the total power demand of the DC bus is calculated in real time, specifically including: The flight status parameters include one or more of the following: flight speed, flight altitude, attitude angle, rate of climb, acceleration, and flight mode command; the external environmental disturbance parameters include one or more of the following: gust speed, turbulence intensity, wind shear characteristic quantity, and airflow change rate; and the propulsion system operating parameters include one or more of the following: propulsion motor speed, output torque, propeller operating status, and inverter operating status. Based on the flight state parameters and the external environmental disturbance parameters, the propulsion power requirement is calculated by combining the flight dynamics model, aerodynamic load model and propulsion characteristic model. The power requirement of the airborne auxiliary electrical load is obtained. The sum of the propulsion power requirement and the power requirement of the airborne auxiliary electrical load is divided by the propulsion link integrated efficiency to obtain the total power requirement of the DC bus.

[0009] Optionally, the total power demand is decoupled in the frequency domain to separate the low-frequency reference power and the high-frequency transient power, specifically including: The total power demand is filtered using a low-pass filter to extract the low-frequency components from the total power demand, thereby obtaining the low-frequency reference power. The high-frequency transient power is obtained by subtracting the low-frequency reference power from the total power demand.

[0010] Optionally, a disturbance intensity characterization quantity is constructed based on the high-frequency transient power, specifically including: Within a preset sliding time window, the instantaneous value of the high-frequency transient power is integrally squared or discretized and summed squared. The integral or summation result is used as the high-frequency power energy accumulation value, and the high-frequency power energy accumulation value is used as the disturbance intensity characterization quantity.

[0011] Optionally, an output power change rate constraint is constructed based on the dynamic response capability of the slow-response active power source system. Relaxation variables are introduced into the output power change rate constraint to form a controlled soft constraint, specifically including: The operating status of the slow-response active power source system is obtained, and the boundary of the rate of change of output power increase and the boundary of the rate of change of output power decrease are determined based on the operating status. The output power change rate constraint is limited to: the output power change rate of the slow-response active power source system is greater than or equal to the negative value of the output power decrease rate of change boundary minus the relaxation variable, and less than or equal to the output power increase rate of change boundary plus the relaxation variable, wherein the relaxation variable is a non-negative value, so as to allow the output power change rate of the slow-response active power source system to exceed the output power increase rate of change boundary or the output power decrease rate of change boundary in a controlled manner under extreme operating conditions.

[0012] Optionally, the output power boundary is constructed based on the state of charge and operating state of the fast-response auxiliary power source system, specifically including: Obtain the current state of charge and operating status of the fast-response auxiliary power source system, and determine the maximum and minimum output power boundaries allowed at the current moment based on the current state of charge and operating status. The output power of the fast-response auxiliary power source system is limited between the minimum output power boundary and the maximum output power boundary.

[0013] Optionally, the state of charge of the fast-response auxiliary power source system is updated in the following manner: The output power and control cycle duration of the fast-response auxiliary power source system are obtained, and the equivalent energy change is determined based on the output power and control cycle duration, combined with the charging and discharging efficiency. The current state of charge is obtained by subtracting the ratio of the equivalent energy change to the available energy of the fast-response auxiliary power source system from the previous state of charge.

[0014] Optionally, an optimization objective function including a high-frequency power tracking cost term is established in the prediction time domain. The weight coefficients of the high-frequency power tracking cost term are adjusted online according to the disturbance intensity characterization quantity. Under the joint constraints of the controlled soft constraint and the output power boundary, the optimization objective function is solved to obtain the optimal power allocation result at the current time, specifically including: In the prediction time domain, the sum of the output power of the slow-response active power source system and the output power of the fast-response auxiliary power source system is equal to the sum of the total power demand and the equivalent terms of power conversion and power supply link loss, which serves as the power balance constraint condition. An optimization objective function is established, which includes a fuel consumption cost term for the slow-response active power source system, a state-of-charge maintenance cost term for the fast-response auxiliary power source system, a tracking error cost term between the output power of the fast-response auxiliary power source system and the high-frequency transient power, and a penalty term for the slack variables. An adaptive weighting coefficient is configured in the tracking error cost term, and the adaptive weighting coefficient increases monotonically as the disturbance intensity characterization quantity increases; Under the combined constraints of the power balance constraint, the controlled soft constraint, and the output power boundary, a power allocation sequence that minimizes the objective function is solved, and the optimal power allocation result at the current moment is extracted from the power allocation sequence.

[0015] Optionally, the adaptive weighting coefficients monotonically increase with the increase of the disturbance intensity characterization quantity, specifically: The adaptive weighting coefficient is equal to the sum of the base weight, the disturbance sensitivity coefficient, and the disturbance intensity characterization quantity, wherein the disturbance sensitivity coefficient is greater than zero.

[0016] Optionally, the optimal power allocation result is converted into control commands for the unidirectional power conversion device and the bidirectional power conversion device, and sent to the slow-response active power source system and the fast-response auxiliary power source system for execution, respectively, specifically including: Obtain the equivalent terminal voltage of the slow-response active power source system, divide the optimal output power of the slow-response active power source system in the optimal power allocation result by the equivalent terminal voltage of the slow-response active power source system, and obtain the target control current of the slow-response active power source system. Obtain the equivalent terminal voltage of the fast-response auxiliary power source system, divide the optimal output power of the fast-response auxiliary power source system in the optimal power allocation result by the equivalent terminal voltage of the fast-response auxiliary power source system, and obtain the target control current of the fast-response auxiliary power source system. The target control current of the slow-response active power source system and the target control current of the fast-response auxiliary power source system are respectively sent to the corresponding power conversion devices for execution.

[0017] Compared with the prior art, the present invention has the following advantages and technical effects: The present invention provides a frequency domain decoupling and dynamic optimization energy management method for multi-source aero-electric propulsion systems accommodating external environmental disturbances. By performing frequency domain decoupling on the total power demand before optimization, the low-frequency reference power under the flight mission is separated from the high-frequency transient power caused by external environmental disturbances. This allows the slow-response active propulsion system to primarily handle the gently changing low-frequency reference power, while the fast-response auxiliary propulsion system prioritizes handling the high-frequency transient power. This weakens the coupling relationship between the slow-response active propulsion system and high-frequency power fluctuations at the control input level, effectively reducing the degree to which the slow-response active propulsion system participates in high-frequency power regulation. It also alleviates reactant supply mismatch and voltage fluctuation problems caused by frequent tracking of high-frequency commands, helps to delay component aging, and improves system lifespan.

[0018] This invention further constructs a disturbance intensity characterization quantity based on high-frequency transient power and uses this disturbance intensity characterization quantity as the basis for adaptive weight adjustment of the high-frequency power tracking cost term in the optimization objective function. This enables the energy management strategy to change the control focus online according to the actual strength of the external environmental disturbance: when the disturbance is weak, the controller takes into account both fuel consumption economy and state of charge maintenance; when the disturbance is strong, the controller automatically strengthens the tracking capability of the fast-response auxiliary power source system for high-frequency transient power, thereby improving the adaptability and response flexibility of the energy management method to different disturbance conditions.

[0019] This invention also introduces a relaxation variable with a high penalty term into the output power change rate constraint of the slow-response active power source system, forming a controlled soft constraint mechanism. Under normal operating conditions, the system operates strictly according to the preset change rate boundary to protect the slow-response active power source. However, under extreme disturbance conditions, the relaxation variable is allowed to be activated to a limited extent, so that the power change rate of the slow-response active power source system is relaxed in a controlled manner. This avoids the optimization problem becoming unsolvable or the feasible region shrinking excessively due to rigid constraints. Thus, while ensuring the life protection requirements of the slow-response active power source system, it also retains the continuous power supply capability under extreme operating conditions, improving the power supply safety and control robustness of the multi-source aerospace electric propulsion system in complex flight environments. Attached Figure Description

[0020] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a schematic diagram of the multi-source aero-electric propulsion system according to an embodiment of the present invention; Figure 2 This is a flowchart of the frequency domain decoupling and dynamic optimization energy management method for multi-source aero-electric propulsion systems oriented towards external environmental disturbances, according to an embodiment of the present invention. Figure 3This is a schematic diagram illustrating the frequency domain decoupling and dynamic optimization control principle of total power demand in an embodiment of the present invention. Detailed Implementation

[0021] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0022] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0023] Example 1 This embodiment provides a frequency domain decoupling and dynamic optimization energy management method for multi-source aero-electric propulsion systems accommodating external environmental disturbances, such as... Figure 1 As shown, the multi-source aero-electric propulsion system employs a fuel cell system as a slow-response active power source system and a lithium battery system as a fast-response auxiliary power source system. The multi-source aero-electric propulsion system includes a fuel cell system, a lithium battery system, a DC bus, a propulsion motor load, a unidirectional power converter, a bidirectional power converter, and an energy management controller. The fuel cell system is connected to the DC bus via the unidirectional power converter to provide continuous output power. The lithium battery system is connected to the DC bus via the bidirectional power converter to provide transient power compensation. The propulsion motor load draws power from the DC bus via an inverter. The energy management controller is communicatively connected to the fuel cell system, the lithium battery system, the unidirectional power converter, the bidirectional power converter, and the airborne flight control system, respectively, to receive flight status information and external environmental information, and to execute the energy management method described in this invention.

[0024] Specifically, the method proposed in this embodiment includes the following steps: The system acquires the aircraft's flight status parameters, external environmental disturbance parameters, and propulsion system operating parameters. Based on the flight status parameters, external environmental disturbance parameters, and propulsion system operating parameters, and combined with the propulsion model, it calculates the total power demand of the DC bus in real time. The total power demand is decoupled in the frequency domain to separate the low-frequency reference power and the high-frequency transient power, and a disturbance intensity characterization quantity is constructed based on the high-frequency transient power. The output power change rate constraint is constructed based on the dynamic response capability of the slow-response active power source system. Relaxation variables are introduced into the output power change rate constraint to form a controlled soft constraint. The output power boundary is constructed based on the state of charge and operating state of the fast-response auxiliary power source system. An optimization objective function containing a high-frequency power tracking cost term is established in the prediction time domain. The weight coefficients of the high-frequency power tracking cost term are adjusted online according to the disturbance intensity characterization quantity. Under the joint constraints of the controlled soft constraint and the output power boundary, the optimization objective function is solved to obtain the optimal power allocation result at the current time. The optimal power allocation result is converted into control commands for the unidirectional power conversion device and the bidirectional power conversion device, and then sent to the slow-response active power source system and the fast-response auxiliary power source system for execution, respectively, to achieve coordinated power supply.

[0025] In this embodiment, the aircraft is in typical flight conditions, and the propulsion system has a baseline power demand that varies with the flight status. External environmental disturbances such as gusts, turbulence, or wind shear cause the propulsion load to fluctuate rapidly based on the baseline power. The energy management controller operates according to a preset control cycle. The method of the present invention is executed in a rolling manner. Under the same disturbance scenario, the method sequentially goes through a weak disturbance stage, a stronger disturbance stage, and an extreme disturbance stage. The energy management logic in each stage is adaptively adjusted according to the disturbance intensity and the state of the power source.

[0026] As a possible implementation method, such as Figures 2-3 As shown, the method proposed in this embodiment achieves coordinated power supply between the slow-response active power source system and the fast-response auxiliary power source system under complex flight disturbance conditions by analyzing, decomposing, constraining, optimizing, and executing the total power demand. The process involves several steps: Step S1 combines flight status, external environmental disturbances, and the propulsion model to calculate the total power demand of the power system in real time, providing an input basis for subsequent energy management; Step S2 decouples the total power demand in the frequency domain, separating the low-frequency reference power and the high-frequency transient power, and constructs a characterization quantity reflecting the disturbance intensity, providing a disturbance perception basis for subsequent adaptive optimization; Step S3 constructs physical constraints for the power source based on the dynamic characteristics and state boundaries of the slow-response active power source system and the fast-response auxiliary power source system, defining a feasible control space for optimization; Step S4 establishes a dynamic optimization model in the prediction time domain and adjusts the control weights online according to the disturbance intensity to find the optimal power allocation result that meets the current operating conditions; Step S5 converts the optimal power allocation result into control commands executable by unidirectional and bidirectional power conversion devices, completing the coordinated control of the slow-response active power source system and the fast-response auxiliary power source system.

[0027] Specifically, the following steps are included: In step S1, the energy management controller first acquires the aircraft's current flight status parameters, external environmental disturbance parameters, and propulsion system operating parameters, and then combines these with the propulsion model to calculate the total power demand on the DC bus side in real time. The flight status parameters include flight speed. Flight altitude Attitude angle Climb rate acceleration and one or more of the flight mode commands, wherein the external environmental disturbance parameters include gust speed. turbulence intensity Wind shear characteristics And one or more of the airflow change rate, the propulsion system operating parameters include propulsion motor speed. Output torque One or more of the following: propeller operating state and inverter operating state. The total power demand is determined jointly by flight state, environmental disturbances, and propulsion model, and converted into the power to be allocated on the DC bus side, which can be expressed as: ; in, This represents the total electrical power demand on the DC bus side. The propulsion power requirement is calculated from flight state parameters, environmental disturbance parameters, and the propulsion model. To meet the power requirements of airborne auxiliary electrical loads, To improve overall link efficiency. The aforementioned It can be calculated jointly by the flight dynamics model, aerodynamic load model, and propulsion characteristic model, or it can be obtained by equivalent conversion of the propulsion demand output by the flight control system. Through this step, the external flight mission requirements and disturbances are uniformly mapped into the power input required for subsequent control.

[0028] In step S2, the total power requirement is first determined. Frequency domain decoupling is performed to obtain the low-frequency reference power. and high-frequency transient power And based on the high-frequency transient power, a disturbance intensity characterization quantity is constructed. The total power requirement satisfies: ; In this embodiment, the low-frequency reference power The high-frequency transient power is obtained by low-pass filtering. It is obtained from the difference between the total power demand and the low-frequency reference power, that is: ; After completing the separation of high and low frequency power components, in a length of Energy accumulation of high-frequency transient power is performed within a sliding time window to construct a disturbance intensity characterization quantity. : ; in, The length of the sliding time window. This step is used to characterize the intensity of high-frequency power surges caused by external environmental disturbances during the current period. It explicitly extracts the high-frequency disturbance from the total power demand and introduces it as a separate control object into the subsequent optimization process. This weakens the coupling between the active power source system and high-frequency power fluctuations at the control input level, creating conditions for lifetime protection of the slow-response active power source system.

[0029] In step S3, firstly, a constraint on the rate of change of fuel cell output power is constructed based on the dynamic response capability of the fuel cell system. Then, a boundary value for the lithium battery output power is constructed based on the state of charge and operating state of the lithium battery system. Relaxation variables are introduced into the fuel cell rate of change constraint to form a controlled soft constraint. Let the fuel cell output power be... Its rate of change constraint is expressed as: ; in, and These are the boundary values ​​for the rate of increase and the rate of decrease of the fuel cell output power, respectively, and these boundaries are related to the fuel cell operating state. Related to the operating state This includes one or more of the following: fuel cell output power, stack temperature, gas supply status, and health status. It is a non-negative slack variable and satisfies: ; Assume the output power of the lithium battery is In this embodiment, it is stipulated that when the lithium battery outputs power to the DC bus... When lithium batteries absorb power The output power boundary of a lithium battery is expressed as: ; in, and These represent the minimum and maximum output power boundaries of the lithium battery under its current state of charge and operating conditions, respectively. These boundaries can be determined by the lithium battery's state of charge. The battery temperature, allowable charge / discharge current boundaries, and health status are all considered together to determine its performance.

[0030] The state of charge of a lithium battery satisfies the state update relationship: ; in, The equivalent energy change function that takes into account the effect of charge and discharge efficiency. This provides usable energy for the lithium battery. In this step, the slow-response active power source system employs a soft constraint on the rate of change with relaxation variables, while the fast-response auxiliary power source system employs a hard constraint on power related to the state of charge. This ensures that the power source boundary reflects normal operating limits while also preserving controlled adjustment space for feasible control under extreme disturbance conditions.

[0031] In this embodiment, the equivalent energy change function Determined by segmenting according to the charging and discharging direction of the lithium battery, when At that time, the lithium battery is in a discharging state, and the equivalent energy change function is taken as: ; when At that time, the lithium battery is in a charging state, and the equivalent energy change function is taken as: ; in, For lithium battery discharge efficiency. For lithium battery charging efficiency, both are parameters greater than 0 and not greater than 1. Substituting the above piecewise form into the state-of-charge update relation, we get: ; Without distinguishing between charging efficiency and discharging efficiency, a uniform equivalent efficiency parameter can be used for approximate calculation.

[0032] In step S4, an optimization objective function is first established within the prediction time domain, incorporating hydrogen consumption costs, state-of-charge maintenance costs, high-frequency power tracking costs, and out-of-bounds penalty costs. Then, the weighting coefficients are adjusted online based on the disturbance intensity characterization quantity to solve for the optimal power allocation sequence between the fuel cell and the lithium battery. Let the prediction time domain length be... In the discrete time domain, the first The power balance relationship at each moment satisfies: ; in, This is the equivalent term for power conversion and power supply link losses. The optimization objective function is expressed as: ; in, Let the hydrogen consumption cost function of the fuel cell be... The cost function for maintaining the state of charge of a lithium battery. This is to compensate for the tracking error of lithium batteries in response to high-frequency transient power. , and These are weighting coefficients. The disturbance intensity characterization quantity... Through adaptive weight function This applies to the high-frequency power tracking cost term. In this embodiment, the adaptive weighting function uses a weighting function with respect to... The monotonically increasing function can be expressed as: ; in, Based on the weights, Since is the disturbance sensitivity coefficient, we have: ; In this embodiment, the hydrogen consumption cost function The output power of the fuel cell can be approximated by a convex function, or it can be derived from the fuel cell efficiency mapping relationship; the state-of-charge maintenance cost function The state of charge can be used relative to the target state of charge. The deviation function, for example: ; in, The target state of charge is preset. The optimization problem can be constructed as a convex optimization problem, a quadratic programming problem, or an equivalent real-time solvable form, and is executed in a rolling time domain manner.

[0033] In step S5, the optimal fuel cell output power for the current moment is first extracted. and optimal lithium battery output power The optimal power command is then converted into a corresponding power conversion device control quantity and sent to the fuel cell side and lithium battery side for execution. The fuel cell side control command can be expressed as: ; The lithium battery-side control command can be expressed as: ; in, and These are the target control currents for the fuel cell side and the lithium battery side, respectively. and These are the equivalent terminal voltages on the fuel cell side and the lithium battery side, respectively. The control commands can also be issued in the form of power reference values, current reference values, duty cycle reference values, or other equivalent forms compatible with the power conversion device. After executing the control command in the current control cycle, the energy management controller enters the next control cycle, again collecting flight status parameters, external environmental disturbance parameters, and propulsion system operating parameters, and repeating steps S1 to S5. Through this step, the optimization results are converted into control commands that the power system can directly execute, realizing coordinated power supply from the slow-response active power source system and the fast-response auxiliary power source system.

[0034] In this embodiment, the method execution process under the same typical disturbance scenario can be summarized as the following three-stage state evolution: In the stage of weak disturbance, the high-frequency transient power component in the total power demand is small, and the disturbance intensity characterization quantity is small. Maintaining a low level, the fuel cell system primarily provides low-frequency reference power, while the lithium battery system performs high-frequency compensation and relaxation. It remains at zero; during the disturbance enhancement phase, the high-frequency transient power component increases. Increase, adaptive weight function As the high-frequency transient power increases, the lithium battery system preferentially handles more high-frequency transient power, while the fuel cell output remains relatively stable. During extreme disturbances, if the high-frequency transient power continues to increase and the lithium battery approaches its power boundary, the optimizer allows for slack variables under high penalty conditions. It is activated in a limited manner, which allows for controlled relaxation of the rate of change of fuel cell output power in order to meet the continuous power supply requirements under extreme operating conditions.

[0035] In this embodiment, each parameter can be configured according to the specific system object, including the control cycle. Predicted time domain length Sliding time window length Basic weights Disturbance sensitivity coefficient Cost function weights , and All of these can be set by combining the system's dynamic response characteristics and control objectives. The value is greater than the weight corresponding to the conventional cost term, the The value is greater than zero, indicating the charging efficiency of the lithium battery. and discharge efficiency The values ​​can be determined based on the selected lithium battery type and the efficiency of the power conversion device. These parameters can be obtained through engineering calibration or adjusted according to the actual system requirements.

[0036] In this embodiment, the present invention weakens the coupling relationship between the slow-response active power source system and high-frequency disturbances in the control structure by decoupling the total power demand in the frequency domain before optimization and treating the low-frequency reference power and high-frequency transient power as different control objectives. By constructing a disturbance intensity characterization quantity based on high-frequency transient power and using the disturbance intensity characterization quantity to adjust the high-frequency tracking cost weight online, the controller can adaptively change the power allocation focus according to the impact intensity of the external environment. By introducing a high-penalty relaxation variable into the rate of change constraint of the slow-response active power source system, the controller prioritizes meeting the life protection requirements of the slow-response active power source system under normal operating conditions and retains the controlled over-limit capability under extreme operating conditions, thereby taking into account the lifespan of the power source, the utilization of the high-frequency support capability of the fast-response auxiliary power source system, and the continuous power supply requirements.

[0037] In this embodiment, the present invention decouples the total power demand, which is jointly determined by flight status, external environmental disturbances, and propulsion model, in the frequency domain. This separates the low-frequency reference power under the flight mission and the high-frequency transient power caused by external environmental disturbances. The slow-response active power source system primarily handles the gently changing low-frequency reference power, while the fast-response auxiliary power source system prioritizes the high-frequency transient power. This weakens the coupling between the slow-response active power source system and high-frequency power fluctuations at the control input level, effectively reducing the degree to which the slow-response active power source system participates in high-frequency power regulation. It alleviates reactant supply mismatch and voltage fluctuation problems caused by frequent tracking of high-frequency commands, improves the targeting and rationality of power source allocation under complex disturbance conditions, helps delay component aging, and extends system lifespan.

[0038] This invention also constructs a disturbance intensity characterization quantity based on high-frequency transient power and introduces this disturbance intensity characterization quantity into the adaptive weight adjustment process of the high-frequency power tracking cost term in the optimization objective function. This enables the energy management strategy to change the control focus online according to the actual strength of the external environmental disturbance: when the external disturbance is weak, the controller takes into account both fuel consumption economy and state of charge maintenance; when the external disturbance is strong, the controller automatically strengthens the tracking capability of the fast-response auxiliary power source system for high-frequency transient power, while limiting the slow-response active power source system to undertake the output task mainly based on low-frequency reference power. This improves the adaptability and response flexibility of the energy management method to different disturbance conditions.

[0039] Furthermore, this invention introduces a relaxation variable with a high penalty term into the constraint on the output power change rate of a slow-response active power source system, forming a controlled soft constraint mechanism. Under normal operating conditions, the system operates strictly according to the preset change rate boundary to protect the slow-response active power source. However, under extreme disturbance conditions, the relaxation variable is allowed to be activated to a limited extent, causing the power change rate of the slow-response active power source system to be relaxed in a controlled manner. This reduces the risk of the optimization problem becoming unsolvable or the feasible region shrinking excessively under rigid constraints. This method can meet the continuous power supply requirements under extreme conditions while taking into account the protection needs of the slow-response active power source system, thus improving the power supply safety and control robustness of multi-source aerospace electric propulsion power systems in complex flight environments.

[0040] The above are merely preferred embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A frequency domain decoupling and dynamic optimization energy management method for a multi-source aero-electric propulsion system oriented towards external environmental disturbances, wherein the multi-source aero-electric propulsion system includes a slow-response active power source system, a fast-response auxiliary power source system, a DC bus, a propulsion motor load, a unidirectional power conversion device, a bidirectional power conversion device, and an energy management controller, characterized in that, include: The system acquires the aircraft's flight status parameters, external environmental disturbance parameters, and propulsion system operating parameters. Based on the flight status parameters, external environmental disturbance parameters, and propulsion system operating parameters, and combined with the propulsion model, it calculates the total power demand of the DC bus in real time. The total power demand is decoupled in the frequency domain to separate the low-frequency reference power and the high-frequency transient power, and a disturbance intensity characterization quantity is constructed based on the high-frequency transient power. The output power change rate constraint is constructed based on the dynamic response capability of the slow-response active power source system. Relaxation variables are introduced into the output power change rate constraint to form a controlled soft constraint. The output power boundary is constructed based on the state of charge and operating state of the fast-response auxiliary power source system. An optimization objective function containing a high-frequency power tracking cost term is established in the prediction time domain. The weight coefficients of the high-frequency power tracking cost term are adjusted online according to the disturbance intensity characterization quantity. Under the joint constraints of the controlled soft constraint and the output power boundary, the optimization objective function is solved to obtain the optimal power allocation result at the current time. The optimal power allocation result is converted into control commands for the unidirectional power conversion device and the bidirectional power conversion device, and then sent to the slow-response active power source system and the fast-response auxiliary power source system for execution, respectively, to achieve coordinated power supply.

2. The method according to claim 1, characterized in that, The system acquires the aircraft's flight status parameters, external environmental disturbance parameters, and propulsion system operating parameters. Based on these parameters and the propulsion system operating parameters, and in conjunction with the propulsion model, it calculates the total power demand of the DC bus in real time. Specifically, this includes: The flight status parameters include one or more of the following: flight speed, flight altitude, attitude angle, rate of climb, acceleration, and flight mode command; the external environmental disturbance parameters include one or more of the following: gust speed, turbulence intensity, wind shear characteristic quantity, and airflow change rate; and the propulsion system operating parameters include one or more of the following: propulsion motor speed, output torque, propeller operating status, and inverter operating status. Based on the flight state parameters and the external environmental disturbance parameters, the propulsion power requirement is calculated by combining the flight dynamics model, aerodynamic load model and propulsion characteristic model. The power requirement of the airborne auxiliary electrical load is obtained. The sum of the propulsion power requirement and the power requirement of the airborne auxiliary electrical load is divided by the propulsion link integrated efficiency to obtain the total power requirement of the DC bus.

3. The method according to claim 1, characterized in that, Frequency domain decoupling is performed on the total power demand to separate the low-frequency reference power and the high-frequency transient power, specifically including: The total power demand is filtered using a low-pass filter to extract the low-frequency components from the total power demand, thereby obtaining the low-frequency reference power. The high-frequency transient power is obtained by subtracting the low-frequency reference power from the total power demand.

4. The method according to claim 1, characterized in that, The disturbance intensity characterization quantity is constructed based on the high-frequency transient power, specifically including: Within a preset sliding time window, the instantaneous value of the high-frequency transient power is integrally squared or discretized and summed squared. The integral or summation result is used as the high-frequency power energy accumulation value, and the high-frequency power energy accumulation value is used as the disturbance intensity characterization quantity.

5. The method according to claim 1, characterized in that, Based on the dynamic response capability of the slow-response active power source system, an output power change rate constraint is constructed. Relaxation variables are introduced into the output power change rate constraint to form a controlled soft constraint, specifically including: The operating status of the slow-response active power source system is obtained, and the boundary of the rate of change of output power increase and the boundary of the rate of change of output power decrease are determined based on the operating status. The output power change rate constraint is limited to: the output power change rate of the slow-response active power source system is greater than or equal to the negative value of the output power decrease rate of change boundary minus the relaxation variable, and less than or equal to the output power increase rate of change boundary plus the relaxation variable, wherein the relaxation variable is a non-negative value, so as to allow the output power change rate of the slow-response active power source system to exceed the output power increase rate of change boundary or the output power decrease rate of change boundary in a controlled manner under extreme operating conditions.

6. The method according to claim 1, characterized in that, The output power boundary is constructed based on the state of charge and operating state of the fast-response auxiliary power source system, specifically including: Obtain the current state of charge and operating status of the fast-response auxiliary power source system, and determine the maximum and minimum output power boundaries allowed at the current moment based on the current state of charge and operating status. The output power of the fast-response auxiliary power source system is limited between the minimum output power boundary and the maximum output power boundary.

7. The method according to claim 6, characterized in that, The state of charge of the fast-response auxiliary power source system is updated in the following manner: The output power and control cycle duration of the fast-response auxiliary power source system are obtained, and the equivalent energy change is determined based on the output power and control cycle duration, combined with the charging and discharging efficiency. The current state of charge is obtained by subtracting the ratio of the equivalent energy change to the available energy of the fast-response auxiliary power source system from the previous state of charge.

8. The method according to claim 1, characterized in that, An optimization objective function incorporating a high-frequency power tracking cost term is established within the prediction time domain. The weighting coefficients of the high-frequency power tracking cost term are adjusted online based on the disturbance intensity characterization. Under the combined constraints of the controlled soft constraint and the output power boundary, the optimization objective function is solved to obtain the optimal power allocation result at the current time. Specifically, this includes: In the prediction time domain, the sum of the output power of the slow-response active power source system and the output power of the fast-response auxiliary power source system is equal to the sum of the total power demand and the equivalent terms of power conversion and power supply link loss, which serves as the power balance constraint condition. An optimization objective function is established, which includes a fuel consumption cost term for the slow-response active power source system, a state-of-charge maintenance cost term for the fast-response auxiliary power source system, a tracking error cost term between the output power of the fast-response auxiliary power source system and the high-frequency transient power, and a penalty term for the slack variables. An adaptive weighting coefficient is configured in the tracking error cost term, and the adaptive weighting coefficient increases monotonically as the disturbance intensity characterization quantity increases; Under the combined constraints of the power balance constraint, the controlled soft constraint, and the output power boundary, a power allocation sequence that minimizes the objective function is solved, and the optimal power allocation result at the current moment is extracted from the power allocation sequence.

9. The method according to claim 8, characterized in that, The adaptive weighting coefficients monotonically increase with the increase of the disturbance intensity characterization quantity, specifically: The adaptive weighting coefficient is equal to the sum of the base weight, the disturbance sensitivity coefficient, and the disturbance intensity characterization quantity, wherein the disturbance sensitivity coefficient is greater than zero.

10. The method according to claim 1, characterized in that, The optimal power allocation result is converted into control commands for the unidirectional power conversion device and the bidirectional power conversion device, and then sent to the slow-response active power source system and the fast-response auxiliary power source system for execution, respectively. Specifically, this includes: Obtain the equivalent terminal voltage of the slow-response active power source system, divide the optimal output power of the slow-response active power source system in the optimal power allocation result by the equivalent terminal voltage of the slow-response active power source system, and obtain the target control current of the slow-response active power source system. Obtain the equivalent terminal voltage of the fast-response auxiliary power source system, divide the optimal output power of the fast-response auxiliary power source system in the optimal power allocation result by the equivalent terminal voltage of the fast-response auxiliary power source system, and obtain the target control current of the fast-response auxiliary power source system. The target control current of the slow-response active power source system and the target control current of the fast-response auxiliary power source system are respectively sent to the corresponding power conversion devices for execution.

Images