Automatic constant speed control method for fuel engine of rotary wing aircraft

Abstract

This invention relates to the field of control technology for fuel-powered rotorcraft, and in particular to an automatic speed control method for fuel-powered rotorcraft engines. Through bench testing and flight experiments, a multi-dimensional adaptive control parameter table is constructed using current power, target power, and altitude as three-dimensional indexes to quantify engine acceleration and deceleration characteristics. After the flight control system identifies the power demand in real time, it calls the parameter table to query the engine power response time and attitude loop response time, and calculates the dynamic lead based on the formula, achieving a synchronization effect where "different control start times, the same target arrival time." Simultaneously, a feedforward + incremental PID feedback composite control is employed, and the parameter table is iteratively corrected online based on actual operating data. This invention fundamentally cuts off the positive feedback coupling link between speed and attitude, significantly reducing speed overshoot and attitude fluctuations, and improving flight stability, control precision, and fuel economy.

Description

Technical Field

[0001] This invention relates to the field of control technology for fuel-powered rotorcraft, specifically to an automatic speed control method for a fuel-powered rotorcraft engine. Background Technology

[0002] In flight control, fuel-powered rotor unmanned aerial vehicles generally adopt variable pitch control, which achieves precise control of flight attitude and trajectory by adjusting the rotor pitch. On the engine control side, whether it is a turbine engine or a piston engine, constant speed control mode is generally used to ensure engine output stability and provide continuous and controllable power support for the rotor.

[0003] During actual flight, the power demand of an aircraft changes frequently and in a complex manner. This change stems from both the dynamic changes in the external flight environment, such as ambient temperature fluctuations, airflow disturbances, and changes in air density, as well as the active power adjustment requirements generated by the aircraft's active maneuvering flight (such as climbing, diving, and turning).

[0004] In existing technologies, the core process of traditional constant-speed control is as follows: flight state changes → flight control system provides attitude correction signal → pitch is adjusted (increased or decreased) → pitch change leads to torque change, which in turn causes engine speed to deviate from the constant-speed setpoint → flight control system sends throttle valve or throttle adjustment signal to the engine based on the speed deviation → engine adjusts throttle valve / throttle opening to return the speed to the normal setpoint. This control method is a typical pure deviation feedback control, which has significant technical defects: when power demand changes, it first causes the engine speed to deviate from the normal value, and then the throttle valve / throttle opening is adjusted in reverse to correct the deviation by detecting the magnitude of the speed deviation. However, the acceleration and deceleration processes of internal combustion engines (especially turbocharged engines) have significant time delays, and this lag can cause a series of problems.

[0005] In practical applications, when an aircraft performs violent maneuvers (such as rapid climbs and emergency turns) or when the flight environment changes drastically, traditional control methods cause the engine speed to oscillate around the set value. This speed oscillation is transmitted to the rotor system, causing aircraft attitude fluctuations. In severe cases, it can create a positive feedback coupling effect of "speed oscillation - attitude deviation - pitch adjustment - torque change - further speed oscillation", resulting in a significant deterioration of flight quality. If the coupled oscillation cannot be effectively converged, it will continue to exacerbate flight instability and may eventually lead to catastrophic consequences such as loss of control of the aircraft.

[0006] Furthermore, existing constant speed control strategies lack adaptive capabilities, cannot dynamically correct parameters based on individual engine characteristics, fuel quality changes, and long-term mechanical degradation, and also lack dynamic feedforward compensation mechanisms, thus failing to meet the aircraft's requirements for high stability and high precision flight under varying operating conditions such as high altitude and complex airflow.

[0007] In summary, existing constant speed control strategies for fuel-powered rotorcraft engines suffer from problems such as control lag, easy generation of speed oscillations, strong coupling with attitude control, lack of adaptive capability, and lack of dynamic feedforward. These issues fail to meet the high stability and high precision flight requirements of aircraft. Therefore, developing an automatic constant speed control method that can solve the above-mentioned technical problems has become an urgent technical issue to be addressed in this field. Summary of the Invention

[0008] To address the shortcomings of existing technologies, the present invention aims to provide an automatic speed control method for fuel-powered rotorcraft engines.

[0009] To achieve the above objectives, the present invention provides the following technical solution: an automatic speed control method for a fuel-powered rotorcraft engine, comprising the following steps:

[0010] The flight control system monitors changes in the aircraft's flight status in real time, identifies current power increase / decrease requirements, and determines the actual power output of the engine. With target power ;

[0011] The flight control system calls a preset multi-dimensional adaptive control parameter table, according to the... , and current flight altitude Query to obtain engine power response time Flight control attitude loop inherent response time and feedforward control reference opening ;

[0012] The flight control system is based on the formula Real-time calculation of power control advance time ;

[0013] The flight control system sends the target speed value to the engine control unit (ECU) via the CAN bus. Power target value and feedforward control reference opening The ECU performs engine power feedforward pre-adjustment control;

[0014] The delay mentioned Then, the flight control system sends attitude adjustment commands to the attitude control module to control the aircraft to adjust the propeller pitch to change the flight attitude;

[0015] During engine power adjustment, the ECU collects the actual engine speed in real time. The flight control system is based on the actual engine speed collected in real time by the ECU. Calculate the speed deviation Incremental PID feedback control generates feedback control to correct the opening degree. and the feedforward control reference opening degree The final control opening is obtained by combining the results. The closed-loop correction is performed by the ECU;

[0016] Real-time acquisition of actual operating condition data and calculation of correction coefficients The multi-dimensional adaptive control parameter table is updated online to synchronize engine power adjustment and aircraft attitude adjustment to achieve the target value, thus realizing automatic speed control.

[0017] In some embodiments, the multi-dimensional adaptive control parameter table is based on the current power. Target power Flight altitude This is a structured lookup table with a three-dimensional index, where each index cell contains four core data items: throttle body / expansion valve target opening. Estimated engine power arrival time Flight control attitude loop inherent response time Power control lead time .

[0018] In some embodiments, the multi-dimensional adaptive control parameter table is constructed using the following standardized test conditions:

[0019] For piston-type or turbofan-type fuel engines, tests are conducted on engine bench testing systems, high-altitude simulators, rotor load tables, and flight data acquisition systems.

[0020] The power range is based on the engine's rated power. Based on this, the equal-interval gradient method is used at 50%. ~100% Divide the range into steps of 5%. A test point includes a steady-state power point and a power step change point;

[0021] Altitude and environmental conditions are divided as follows: piston engine test altitudes include 0m, 500m, 1000m, 1500m, 2000m, 3000m, 4000m, and 5000m; turbine engine test altitudes include 0m, 1000m, 2000m, 3000m, 4000m, 5000m, 6000m, 7000m, and 8000m; ambient temperature includes −20℃, 0℃, 10℃, 20℃, 30℃, 40℃, and 50℃; humidity includes 30%RH, 50%RH, and 80%RH. Each set of operating conditions is an independent combination of altitude, temperature, and humidity.

[0022] The data acquisition channels include engine speed, torque, power, throttle / expansion valve opening, altitude, ambient temperature, intake air pressure, coolant temperature, and engine oil temperature. The acquisition frequency is 100Hz. After first-order low-pass filtering and moving average filtering, each set of operating conditions is tested at least 5 times, and samples are discarded. For outliers, the arithmetic mean is taken and stored in the parameter table.

[0023] In some embodiments, the parameters in the multi-dimensional adaptive control parameter table have the following mathematical mapping relationship:

[0024] Power-aperture mapping: ;

[0025] Power response time model: ,in ;

[0026] Dynamic lead time calculation formula: .

[0027] In some embodiments, the feedforward pre-adjustment control and feedback closed-loop correction constitute a composite control architecture, wherein the feedforward path rapidly outputs a hysteresis-free reference opening based on a multi-dimensional adaptive control parameter table. It provides over 90% adjustment to compensate for engine delay; the feedback path is based on an incremental PID algorithm according to the speed deviation. Real-time fine-tuning of the opening eliminates model errors and external disturbances, ensuring stable rotational speed.

[0028] In some embodiments, the online iterative update includes the following steps:

[0029] Real-time acquisition of engine actual response time Compared with the estimated response time in the parameter table Deviation;

[0030] Calculate the correction factor ;

[0031] Update the data of the corresponding working point in the parameter table according to the gradient, without overwriting the original calibration value;

[0032] For non-test point operating conditions, a three-dimensional linear interpolation method is used to output continuous control quantities, so that the parameter table can be adaptively optimized as engine degradation, fuel differences and individual characteristics change.

[0033] In some embodiments, safety and fault-tolerant control logic is also included:

[0034] Real-time monitoring of engine unresponsiveness, excessive power deviation, sudden attitude changes, and communication failures;

[0035] When an anomaly is detected, dynamically execute at least one of the following degradation measures: recalculation Increase the weight of feedback control, interrupt the current posture action, and enter the safe constant speed mode;

[0036] When the measured speed deviation e exceeds ±20 r / min to ±50 r / min, the PID feedback path immediately fine-tunes the valve / throttle valve opening to quickly stabilize the speed; when the measured response time deviation from the theoretical response time exceeds ±0.1 s to 0.2 s, it automatically updates. And calibrate .

[0037] In some embodiments, the flight control system and the ECU communicate and interact via a CAN bus, with a control cycle of 10ms and a synchronization cycle of 1ms. The instruction message ID sent by the flight control system to the ECU is 0x110, which includes the target power value, target speed value, feedforward opening degree, and mode word. The status message ID fed back by the ECU to the flight control system is 0x210, which includes the actual speed, actual power, actual opening degree, and status word.

[0038] In some embodiments, when the current operating point queried by the flight control system is not a test point in the multi-dimensional adaptive control parameter table, the target opening degree is calculated using a three-dimensional linear interpolation method. Power response time and power control lead time The interpolation dimension is , , This is to ensure the continuity and smoothness of the control output.

[0039] In some embodiments, the method is applicable to rotorcraft unmanned aerial vehicles equipped with piston-type or turbine-type fuel engines; when adapting to new engine models, a basic model plus rapid calibration method is adopted, requiring only 3 sets of typical operating condition tests to generate a dedicated multi-dimensional adaptive control parameter table, and automatically adapting to the individual characteristics of different engines through online learning.

[0040] Compared with the prior art, the beneficial effects of the present invention are:

[0041] 1. Address the root causes of control lag and speed oscillation issues.

[0042] This invention constructs a multi-dimensional adaptive control parameter table to achieve advance prediction and precise pre-adjustment of engine power, eliminating the need to wait for engine speed deviations before correction. This effectively avoids control lag caused by acceleration and deceleration delays in fuel engines (especially turbocharged engines). Through a feedforward + feedback composite control architecture, speed overshoot is reduced from ±15%~±25% in traditional control methods to within ±2%, a reduction of over 85%, completely solving the engine speed oscillation problem and significantly improving constant speed control accuracy.

[0043] 2. Completely sever the power-attitude positive feedback coupling link.

[0044] This invention achieves precise temporal alignment between engine power adjustment and aircraft attitude adjustment through dynamic lead time T_lead, ensuring that the two have "different control start times but the same target arrival time," synchronously reaching the target required value. This avoids attitude fluctuations caused by speed oscillations transmitted to the rotor system. The attitude fluctuation amplitude is reduced from ±5°~±8° in traditional control methods to within ±0.8°, a reduction of more than 80%. This fundamentally severs the positive feedback coupling link of "speed deviation → attitude adjustment → torque change → speed deviation," eliminating the flight quality deterioration problem caused by coupled oscillations and significantly improving the aircraft's flight stability and flight quality.

[0045] 3. Significantly improves response speed and adaptability to complex environments.

[0046] This invention employs a composite control strategy that combines feedforward pre-adjustment and closed-loop correction, reducing the system convergence time from 4s in the traditional method to less than 1.5s, a reduction of more than 60%. In harsh flight environments such as complex airflow disturbances and drastic altitude changes, the aircraft's attitude maintenance capability is improved by more than 70%, demonstrating excellent robustness and environmental adaptability, and is especially suitable for severe maneuvering flight scenarios.

[0047] 4. Extends engine life and improves fuel economy

[0048] Because this invention effectively suppresses frequent oscillations in engine speed and mechanical shocks, engine mechanical shocks are reduced by more than 60%, mechanical wear and fatigue damage caused by speed fluctuations are reduced, and engine service life is extended; at the same time, precise power control reduces fuel consumption by 5% to 12%, balancing engine durability and fuel economy.

[0049] 5. Possesses adaptive learning and broad adaptability.

[0050] This invention supports online learning and iteration of parameter tables, automatically correcting control parameters based on actual engine response errors. It adaptively adapts to individual engine characteristics, fuel quality variations, and long-term mechanical degradation, eliminating the need for frequent factory recalibration. For new engine models, only three sets of typical operating condition tests are required to quickly generate a dedicated parameter table, demonstrating excellent universal adaptability. This method can be widely applied to various rotorcraft unmanned aerial vehicles equipped with piston and turbine internal combustion engines, without requiring significant modifications to the original engine structure. It can be achieved solely through software algorithm optimization, significantly reducing system modification costs and enhancing the practicality and promotional value of the technical solution.

[0051] Details of one or more embodiments of this application are set forth in the following drawings and description to make other features, objects and advantages of this application more readily apparent. The embodiments of this application will provide a detailed description and understanding of the application. Attached Figure Description

[0053] Figure 1 This is the overall system architecture and control flowchart of the automatic speed control method for fuel-powered rotorcraft engines of the present invention;

[0054] Figure 2 A comparison chart showing the speed and attitude control effects of traditional constant speed control and the automatic constant speed control of this invention;

[0055] Figure 3 This is a schematic diagram of the three-dimensional index structure and data flow of the multi-dimensional adaptive control parameter table of the present invention;

[0056] Figure 4 This is the calculation model for the dynamic lead T_lead of this invention and the timing alignment diagram of engine power and aircraft attitude control.

[0057] Figure 5 This is a schematic diagram of the CAN bus communication architecture and message interaction between the flight control system and the ECU of this invention. Detailed Implementation

[0059] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. 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 are within the scope of protection of the present invention.

[0060] The automatic speed control method for fuel-powered rotorcraft engines described in this invention is a composite control system for fuel-powered rotorcraft unmanned aerial vehicles, employing a combination of "feedforward prediction, closed-loop correction, and online learning." This system is based on CAN bus communication between the Flight Control System (FCS) and the Engine Control Unit (ECU). By constructing a multi-dimensional adaptive control parameter table specifically for the power-attitude coupling scenario of rotorcraft, it achieves precise temporal alignment between engine power adjustment and aircraft attitude adjustment, thereby fundamentally severing the positive feedback coupling link of "speed deviation → attitude adjustment → torque change → speed deviation."

[0061] In the entire control system, the flight control system undertakes core functions such as flight status perception, power demand identification, parameter table lookup and interpolation calculation, advance dynamic calculation, feedforward command issuance, attitude timing control, feedback closed-loop correction, and online learning iteration; while the ECU is responsible for receiving the target speed value sent by the flight control system through the CAN bus. Power target value and feedforward control reference opening Based on its built-in speed closed-loop control algorithm, it executes precise actuation of the throttle or flow control valve, while simultaneously feeding back the actual speed to the flight control system in real time. Actual power, actual opening degree, and status word.

[0062] The multi-dimensional adaptive control parameter table is the data foundation for achieving precise advance control in this invention. This parameter table is not a general engine operating condition data table, but rather a multi-dimensional coupling-specific model built specifically for fuel-powered rotorcraft to solve the particular technical challenge of power-attitude coupling oscillations. It uses the current actual engine power... Target power The current flight altitude H is a three-dimensional index, and each index cell fully records four core data items: throttle / boost valve target opening θ, estimated engine power arrival time, and so on. (Also recorded as) ), Flight control attitude loop inherent response time And the power control lead time T_lead (also denoted as The parameter table is constructed by integrating the three coupled characteristics of engine dynamics, rotor aerodynamics, and aircraft kinematics, and is based on a large number of targeted bench tests and real flight tests. It is only applicable to rotorcraft scenarios involving variable pitch, sudden torque changes, engine delay compensation, and synchronous constant speed control.

[0063] The calculation of dynamic lead is the core of this invention, distinguishing it from traditional deviation feedback control. After identifying the power requirement, the flight control system calculates the lead time according to the formula... Real-time calculation of power control advance time This lead time is not a fixed constant, but rather a dynamic function output that changes in real time with the operating conditions. ,in air density, Engine temperature status. This is the engine's real-time response rate coefficient. Within each control cycle (10ms), the flight control system re-queries the parameter table and calculates... This ensures that the lead time is always precisely matched with the current operating conditions.

[0064] The control execution phase adopts a three-stage timing architecture of "feedforward pre-adjustment priority, attitude delay synchronization, and feedback real-time correction". In the first stage, the flight control system sends a power feedforward command, including the target speed value, to the ECU via the CAN bus. Power target value and feedforward control reference opening The ECU immediately executes engine power feedforward pre-adjustment control, driving the throttle valve / throttle body according to the target opening given in the parameter table, causing the engine to begin the acceleration and deceleration process; in the second stage, the flight control system has an internal delay. Then, an attitude adjustment command is sent to the attitude control module to control the aircraft to adjust the rotor pitch to change its flight attitude. At this time, engine power adjustment has already been performed. During the first stage, both will reach the target state simultaneously; in the second stage, throughout the entire engine power adjustment process, the ECU will collect the actual engine speed in real time. The flight control system is based on the actual engine speed collected in real time by the ECU. Calculate the speed deviation Incremental PID feedback control generates feedback control to correct the opening degree. and the feedforward control reference opening degree The final control opening is obtained by combining the results. The ECU performs closed-loop correction to eliminate model errors and external disturbances.

[0065] Furthermore, this invention also features an online learning iterative update mechanism and a safety fault-tolerant control logic. Online learning is achieved by real-time acquisition of the engine's actual response time. Compared with the estimated response time in the parameter table Calculate the correction factor for the deviation. The system updates the data for the corresponding operating point in the parameter table according to the gradient, without overwriting the original calibration values, so that the parameter table can adaptively optimize as the engine degrades, fuel differences, and individual characteristics change. The safety fault-tolerant logic monitors abnormal situations such as engine unresponsiveness, excessive power deviation, sudden attitude changes, and communication failures in real time, and dynamically performs recalculation. At least one degraded measure, such as increasing the weight of feedback control, interrupting the current attitude maneuver, or entering a safe constant speed mode, should be taken to ensure that the flight does not lose control.

[0066] The technical solution of the present invention will be further described in detail below through several specific embodiments.

[0067] Example 1: Rapid Climbing Operation of a Piston-Type Fuel-Powered Rotor Vehicle

[0068] This embodiment is applied to a small rotorcraft unmanned aerial vehicle equipped with a piston-type fuel engine, simulating the complete control process of the aircraft performing a rapid climb mission in an airspace at an altitude of 1500m, and fully demonstrating the entire process of querying and applying multi-dimensional adaptive control parameter tables, dynamic lead calculation, feedforward pre-adjustment and attitude synchronization control.

[0069] Prior to the flight mission, the multi-dimensional adaptive control parameter table for the aircraft had been constructed through standardized bench testing. For this piston engine, test altitudes covered 0m, 500m, 1000m, 1500m, 2000m, 3000m, 4000m, and 5000m, with the power range based on the rated power P_n, at 50... ~100% Within the range, at 5% The test points were divided into steps, with ambient temperatures covering -20℃, 0℃, 10℃, 20℃, 30℃, 40℃, and 50℃, and humidity covering 30%RH, 50%RH, and 80%RH. The test was conducted at an altitude of 1500m and with power ranging from 80%. The step increase reached 95%. Under the test conditions, after no less than 5 repeated tests, first-order low-pass filtering (cutoff frequency 20Hz) and 5-point moving average filtering, 3 were eliminated. Outliers are taken and their arithmetic mean is calculated. The data recorded in the parameter table is: throttle target opening. =82%, estimated engine power delivery time =1.0s, inherent response time of the flight control attitude loop =0.8s, power control lead time =0.2s.

[0070] When the aircraft is flying at an altitude of 1500m and at 80% of its rated power, the flight control system monitors the aircraft's flight status in real time through the inertial measurement unit (IMU), GPS / BeiDou integrated navigation module, and air data computer. When the flight control system detects that the aircraft needs to climb rapidly based on mission planning or control commands, it immediately determines the actual power of the current engines. =80% Target power t=95% The current flight altitude is H=1500m.

[0071] The flight control system calls up a preset multi-dimensional adaptive control parameter table, according to... =80% , =95% as well as =1500m three-dimensional index, directly query to obtain: engine power response time =1.0s, inherent response time of flight control attitude loop =0.8s, feedforward control reference opening =82%. Subsequently, the flight control system calculated the dynamic lead using the formula. Perform real-time calculations to obtain =1.0s-0.8s=0.2s.

[0072] At the start of the control cycle At any given moment, the flight control system sends a command message (message ID: 0x110) to the ECU via the CAN bus. This message contains the target speed value. (For example, set to a rated operating speed of 6500 r / min), target power value =95% and feedforward control reference opening =82%. After receiving the command, the ECU immediately leaves the idle or original steady-state control mode and enters the feedforward pre-adjustment control mode. It drives the throttle valve to quickly adjust from the current opening (assuming it is 65%) to the target opening of 82%. The engine intake air volume increases, and the mixture concentration in the combustion chamber and the ignition timing are optimized in coordination by the ECU's internal pulse spectrum. The engine output torque begins to rise, and the power adjustment process is initiated.

[0073] While the ECU performs power feedforward pre-adjustment, the flight control system internally uses a precise 10ms control cycle for timing. When the delay reaches... At 0.2s (t_0 + 0.2s), the flight control system sends a climb attitude adjustment command to the attitude control module. Upon receiving the command, the attitude control module increases the rotor pitch via servos or electronic speed controllers, causing the rotor thrust vector to deflect vertically, generating additional lift to overcome gravity and achieve climb. Since power control was initiated 0.2s in advance, and the engine power reaches 95%,... It takes 1.0s for attitude control to go from issuing the command to adjusting the pitch to the correct position, and 0.8s for both. At +1.0s, the target state is reached synchronously: the engine power stabilizes at 95% of the rated power, and the aircraft attitude is synchronously adjusted to the climb state.

[0074] Throughout the power adjustment process, the ECU sends a status message (message ID: 0x210) to the flight control system at a frequency of 100Hz, including the actual engine speed. Actual power, actual opening degree, and status words. The ECU collects the engine's actual speed in real time. The flight control system is based on the actual engine speed collected in real time by the ECU. Calculate the speed deviation Assuming that during the power ramp-up process, a slight downward surge in actual engine speed occurs due to a transient change in intake air temperature, resulting in a speed deviation... Upon reaching -30 r / min, the incremental PID controller in the flight control system responds immediately. Let the current sampling time be k, then the incremental PID output is... Generate feedback control to correct the opening degree This correction is appended to the ECU via the CAN bus, and the ECU will then feed forward the control reference opening. With feedback control to correct opening degree The composite process yields the final control opening degree. This causes the throttle valve to be finely adjusted from its base opening of 82% (for example, to 83.2%), allowing the engine speed to quickly return to normal. Eliminate overshoot and undershoot.

[0075] Finally, after 1.0 seconds, the engine power stabilized at 95% of its rated power, and the actual speed... With target speed With deviations controlled within ±2%, the aircraft smoothly transitions to a climb state, engine speed remains stable, and flight quality is significantly improved. Compared with traditional pure deviation feedback control, in this embodiment, engine speed overshoot is reduced from ±15%~±25% to within ±2%, and attitude fluctuation is reduced from ±5°~±8° to within ±0.8°, completely avoiding the coupling effect of "engine speed oscillation - attitude deviation - pitch adjustment - torque change - further engine speed oscillation".

[0076] Example 2: Airflow disturbance conditions of a turbine-type fuel-powered rotorcraft

[0077] This embodiment is applied to a medium-sized rotorcraft unmanned aerial vehicle equipped with a turbine-powered fuel engine. It simulates the power reduction and attitude stabilization control process of the aircraft when it encounters sudden airflow disturbances in the airspace at an altitude of 4000m. It focuses on demonstrating the advance prediction control, feedforward + feedback composite control and anti-disturbance capability under the delay characteristics of the turbine engine.

[0078] The multi-dimensional adaptive control parameter table for the turbine engine mounted on this medium-sized rotorcraft UAV was constructed jointly using a high-altitude simulation cabin and a flight data acquisition system. For the turbine engine, test altitudes covered 0m, 1000m, 2000m, 3000m, 4000m, 5000m, 6000m, 7000m, and 8000m, with the power range also divided in 5% increments based on the rated power P_n. Under test conditions at an altitude of 4000m, where the power decreased in a step from 90% P_n to 75% P_n, the parameter table recorded the following data: throttle valve target opening θ = 62%, and estimated engine power arrival time. =1.8s, inherent response time of flight control attitude loop =1.5s, power control lead time 0.3s.

[0079] When the aircraft is flying stably at 90% of its rated power at an altitude of 4000m, the flight control system detects a sudden increase in lift due to a sudden airflow disturbance via the pitot tube, IMU, and external airflow sensing sensors. To maintain flight altitude and attitude stability, the flight control system recognizes the need to reduce power to match the new aerodynamic balance state and determines the current actual engine power. =90% Target power =75% Current flight altitude =4000m.

[0080] The flight control system calls the multi-dimensional adaptive control parameter table to obtain... =1.8s =1.5s =62%. According to the formula... Calculated =1.8s-1.5s=0.3s.

[0081] exist At any given time, the flight control system sends a power reduction command message (ID: 0x110) to the ECU via the CAN bus, containing... , =75% and =62%. After receiving the data, the ECU immediately controls the throttle valve to decrease from the current opening (assuming 78%) to the target opening of 62%. This reduces the fuel supply to the turbocharged engine, causing the turbine speed to drop and initiating the power adjustment process. Due to the large rotor inertia and significant thermal inertia of the turbocharged engine, there is a noticeable delay in its deceleration process. If a traditional control method is used, the speed will have already begun to deviate from the set value at this point, causing attitude fluctuations.

[0082] The flight control system precisely delays the time after sending the power reduction command. =0.3s. In At +0.3s, the flight control system sends an attitude stabilization signal to the attitude control module, controlling the pitch to fine-tune and counteract the lift changes caused by airflow disturbances. At this time, the turbine engine has already begun deceleration 0.3s in advance, while attitude adjustment requires 1.5s to complete and power reduction requires 1.8s. Both will... At +1.8s, the target state is reached synchronously: the engine power stabilizes at 75% of the rated power, and the aircraft attitude returns to stability.

[0083] During the power reduction process, the ECU collects the actual engine speed in real time. The flight control system is based on the actual engine speed collected in real time by the ECU. Due to their time-delay characteristics, turbine engines may experience a brief surge in engine speed during the initial stages of deceleration. The flight control system calculates this speed deviation. When detected When the speed exceeds +40 r / min, the incremental PID controller generates negative feedback to correct the opening. The signal is sent to the ECU via the CAN bus. The ECU then performs composite control of the opening degree. Further fine-tuning the flow valve opening (e.g., adjusting it to 60.5%) accelerates the speed return process.

[0084] Meanwhile, the flight control system monitors the actual engine response time in real time. If the actual time for the measured power to drop from 90% to 75% is 1.9 seconds, which deviates from the estimated 1.8 seconds in the parameter table by 0.1 seconds, the online learning module calculates a correction factor. =1.9 / 1.8≈1.056, and update the parameter table for that operating point according to the gradient. The data (not overwriting the original calibration values) is available for subsequent flight operations.

[0085] Ultimately, after 1.8 seconds, the engine power stabilized at 75% of its rated power, the aircraft attitude remained stable, and the engine speed did not fluctuate, effectively avoiding flight instability caused by airflow disturbances. This embodiment fully demonstrates that, given the significant acceleration and deceleration delays of turbine engines, predicting deceleration time and throttle valve opening in advance using multi-dimensional parameter tables can effectively solve the control lag problem of turbine engines and ensure the flight stability of medium-sized rotorcraft in complex airflow environments.

[0086] Example 3: Standardized Construction and Calibration of Multi-Dimensional Adaptive Control Parameter Table

[0087] This embodiment details the complete construction process of the multi-dimensional adaptive control parameter table from test design to data entry, corresponding to the technical features of claim 3, and serves as the data foundation for all subsequent embodiments.

[0088] Test Objects and Platform Preparation: A certain type of piston-type fuel engine and a certain type of turbofan-type fuel engine were selected as test objects. The test platform included an engine bench testing system (equipped with full-range torque / speed / power sensors), a high-altitude simulation chamber (capable of simulating air pressure and temperature environments from 0m to 8000m), a rotor load test bench (simulating rotor aerodynamic load characteristics), and a flight data acquisition system (airborne 100Hz multi-channel data logger). The acquisition channels covered speed, torque, power, throttle / expansion valve opening, altitude, atmospheric temperature, intake pressure, coolant temperature, and engine oil temperature, with the acquisition frequency uniformly set to 100Hz.

[0089] Power range division: based on engine rated power Based on this, the equal-interval gradient method is used at 50%. ~100% Divide the area into test points, with a step size of 5%. One test point. The test point includes steady-state power points (50%, 55%, 60%...100%) and power step change points (power increase step: 50%→55%, 55%→60%...95%→100%; power decrease step: 100%→95%, 95%→90%...55%→50%). For each step change point, the response time from command issuance to power reaching the target value within a ±2% error band must be recorded. .

[0090] Altitude and environmental condition classification: For piston engines, test altitudes are set at 0m, 500m, 1000m, 1500m, 2000m, 3000m, 4000m, and 5000m; for turbine engines, the test altitudes are extended to 0m, 1000m, 2000m, 3000m, 4000m, 5000m, 6000m, 7000m, and 8000m. Ambient temperature is set at −20℃, 0℃, 10℃, 20℃, 30℃, 40℃, and 50℃; humidity is set at 30%RH, 50%RH, and 80%RH. Each operating condition is an independent combination of altitude, temperature, and humidity to ensure the completeness of the three-dimensional index.

[0091] Engine initial condition control: Before testing a piston engine, the coolant temperature should be stable at 80℃~90℃, the engine oil temperature at 90℃~110℃, and the stable idling time after startup should be no less than 30 seconds. The intake pressure and power supply voltage should be stable within the nominal range. Before testing a turbocharged engine, the exhaust temperature should be stable within the operating range, the lubricating oil temperature at 90℃~120℃, and the warm-up time after startup should be no less than 60 seconds.

[0092] Data Acquisition and Processing: Under each independent operating condition, the engine... Step change to The high-speed data acquisition system records the entire process at a frequency of 100Hz. The raw data is first processed by a first-order low-pass filter with a cutoff frequency set at 20Hz to eliminate high-frequency electromagnetic interference; then, a 5-point moving average filter is applied to further smooth random noise. Each operating condition is tested at least 5 times, and the data obtained from these repeated tests are... , , Parameters, using 3 Outliers are removed according to the criteria (i.e., data points that deviate from the average by more than 3 times the standard deviation). The arithmetic mean of the remaining data is taken as the final calibration value for that operating point and stored in the parameter table.

[0093] Parameter table structure storage: The parameter table is stored in the flight control system's non-volatile memory in the form of a structured database. The three-dimensional index axes are respectively... (50%) ~100% (Step size 5%) (50%) ~100% (Step size 5%) (The altitude sequence above is used based on engine type). Each cell contains four core data points: throttle body / expansion valve target opening. Estimated engine power arrival time Flight control attitude loop inherent response time Power control lead time .in Direct calculation from the calibration stage It can be stored later and recalculated in real time by the flight control system during flight.

[0094] Verification of mathematical mapping relationships: During calibration, the mathematical mapping relationships between the parameters in the parameter table are verified simultaneously. The power-aperture mapping relationship satisfies... ,in , , These are the fitting coefficients. For constant terms, The ambient temperature is considered. The power response time model satisfies... ,in 'a' and 'b' are engine characteristic coefficients. For piston engines, 'a' is typically small (approximately 0.05 s / %), while for turbocharged engines, 'a' is larger (approximately 0.12 s / %). The dynamic lead calculation formula strictly follows... Through fitting with a large amount of measured data, it was confirmed that the above mathematical mapping relationship has good linearity within the calibration interval and can serve as the theoretical basis for three-dimensional linear interpolation.

[0095] Example 4: Application of Feedforward + Feedback Composite Control in Complex Maneuvering Flight

[0096] This embodiment uses the scenario of a rotorcraft equipped with a piston engine performing an emergency turn maneuver as an example to illustrate in detail the composite control architecture consisting of feedforward pre-adjustment control and feedback closed-loop correction, corresponding to the technical features of claim 5.

[0097] When the aircraft is cruising at 70% of its rated power at an altitude of 2000m, the flight control system receives an emergency turn command and recognizes the need to rapidly increase power to provide additional yaw torque and thrust reserves. =70% , =90% , =2000m. (This was obtained by querying the parameter table.) =79%, =1.1s, =0.85s, calculate =0.25s.

[0098] The feedforward path rapidly outputs a hysteresis-free reference opening based on a multi-dimensional adaptive control parameter table. =79%. This feedforward opening is directly mapped from massive bench test data and can provide more than 90% adjustment range. Its physical essence is pre-compensation for engine delay characteristics. At that moment, the flight control system sends a feedforward command to the ECU via the CAN bus. The ECU immediately drives the throttle to 79% opening, causing a surge in engine intake air, accelerating the combustion process, and rapidly increasing torque. The role of the feedforward path is to "take the lead," applying control parameters to the engine based on the predictive model before any deviation in engine speed occurs, fundamentally eliminating the lag root cause of "waiting for deviation to occur before taking action" in traditional control.

[0099] However, feedforward control, based on a calibration model, cannot fully cover all uncertainties in actual flight, such as fuel octane fluctuations, partial intake blockages, and decreased volumetric efficiency due to engine aging. Therefore, a feedback path is indispensable. During engine power adjustment, the flight control system collects data in real time at a frequency of 100Hz. ,calculate Suppose the incremental PID parameters are tuned to... =0.015% / r / min, =0.003% / r / min, =0.001% / r / min, then when detected When the actual speed is 25 r / min lower than the target speed, the PID controller outputs... =+0.375%, ECU executes =79% + 0.375% = 79.375%. Although this feedback correction is small, it can effectively eliminate model errors and external disturbances, ensuring the steady-state accuracy of the rotational speed.

[0100] Delay =0.25s later, the flight control system sends a steering command to the attitude control module, controlling the differential pitch change of each rotor blade to generate yaw torque. Since the feedforward path has already boosted the engine power to near the target state in advance, the additional torque required for attitude adjustment is provided by the engine in time, avoiding the stall phenomenon of "attitude requires force, engine lacks power". Finally, the engine power and the aircraft attitude reach the target state synchronously, the steering process is smooth and without jerking, and the speed overshoot is controlled within ±2%.

[0101] Example 5: Online Learning, Iterative Updates, and Adaptive Optimization

[0102] This embodiment takes the adaptive optimization of parameter tables in long-term flight missions as a scenario, and elaborates on the online learning iterative update mechanism, corresponding to the technical features of claim 6.

[0103] After multiple missions, the aircraft's engine experienced a decrease in mechanical efficiency due to carbon buildup and piston ring wear, resulting in discrepancies in the original parameter tables. Gradually deviating from reality. The flight control system's built-in online learning module continuously monitors the actual response performance for each control cycle.

[0104] In a time from =60% Increase to P_target=80% During flight missions, parameter table predictions =0.9s, but the flight control system determines the actual time to reach the target power by collecting the engine's actual power curve in real time. =1.05s. The online learning module calculates the correction factor. =1.05 / 0.9≈1.167.

[0105] Following the principle of gradient-based nearest-neighbor updates, the online learning module does not directly overwrite the original calibration values ​​(to protect the safety of the base model). Instead, it establishes a dynamic correction layer in the parameter table for this operating point (60%→80%, current altitude), recording a correction coefficient of 1.167. In subsequent flights, when this operating point is queried again, the flight control system first reads the original calibration values. =0.9s, then multiplied by the correction factor of 1.167, the adaptive predicted response time is 1.05s, and the calculation is based on this. This two-tiered architecture of "original calibration + dynamic correction" not only preserves the benchmark reliability of bench testing but also endows the system with adaptive capabilities.

[0106] For non-test point conditions, such as =65% , =85% , =1750m (between the parameter table grid points), the flight control system uses three-dimensional linear interpolation to output continuous control quantities. The interpolation dimension is... , , Linear interpolation is performed in three dimensions: firstly in... Interpolation is performed on two adjacent power points (60% and 70%) along the dimension to obtain the parameters of the intermediate surface; subsequently... Interpolate adjacent target power points (80% and 90%) along the dimension; finally, Interpolation was performed between 1500m and 2000m along the dimensional axis to obtain the result for this non-test point. , and Three-dimensional linear interpolation ensures the continuity and smoothness of the control output, avoiding the control quantity jumps caused by parameter table discretization.

[0107] After approximately 50 online iterations during flight missions, the parameter table dynamic correction layer has covered commonly used operating conditions. The engine power control accuracy has improved by about 15% compared to the initial state, and fuel economy has improved by about 8%, fully demonstrating the adaptive optimization capability of the online learning mechanism for engine degradation, fuel differences, and individual characteristics.

[0108] Example 6: Safety Fault Tolerance Control and Anomaly Degradation Handling

[0109] This embodiment uses the scenario of communication interference and engine transient stall during flight as an example to elaborate on the safety and fault-tolerant control logic, corresponding to the technical features of claim 7.

[0110] When the aircraft was performing a cruise mission at an altitude of 3000m, the flight control system sent power maintenance commands to the ECU via the CAN bus. Suddenly, the airborne CAN bus was subjected to electromagnetic interference, and no status message (ID: 0x210) was received from the ECU for three consecutive control cycles (30ms). The flight control system's safety monitoring module immediately identified the "communication failure" anomaly and triggered a dynamic degradation process.

[0111] Level 1 Degradation: Flight Control System Recalculation Due to communication delays causing a missynchronization between command issuance and status feedback, the flight control system will... A 0.1s safety margin is added to compensate for timing uncertainties caused by communication jitter.

[0112] Second-level degradation: Increase the weight of feedback control. The flight control system will use incremental PID... The coefficient is temporarily increased by 30%. The feedback path is improved by 20% to enhance its correction capability and address the risk that feedforward commands may not be accurately delivered to the ECU.

[0113] Level 3 Degradation: Abort the current attitude maneuver. The flight control system sends a "maintain current pitch" command to the attitude control module, suspending any active maneuvers to prevent the power-attitude mismatch from worsening due to attitude adjustments during communication anomalies.

[0114] If the communication failure persists for more than 100ms (10 control cycles), the flight control system enters the fourth level of degradation: safe speed control mode. In this mode, the flight control system sends a fixed safe speed command (e.g., 85% of rated speed) and a fixed safe power command (e.g., 75% of rated power) to the ECU. The ECU enters an independent speed control state, no longer responding to feedforward opening commands, and maintains basic operation solely through its own speed closed loop. Simultaneously, the flight control system guides the aircraft into a gentle descent path, prioritizing flight safety.

[0115] In another scenario, suppose the engine's actual speed is reduced due to a transient interruption in fuel supply. Sudden drop, speed deviation Exceeding +50 rpm (beyond the normal threshold of ±20 rpm to ±50 rpm), the PID feedback path immediately and significantly fine-tunes the valve opening. The positive correction amount reached +5%. The opening is significantly increased from the baseline to restore speed as quickly as possible. Simultaneously, the timing dynamic correction module records this abnormal response; when the measured response time deviates from the theoretical response time by more than ±0.1s to 0.2s, it automatically updates the parameter table for that operating point. And calibrate This makes subsequent control more conservative and stable.

[0116] Example 7: Real-time calculation of three-dimensional linear interpolation under non-test point conditions

[0117] This embodiment details how the flight control system calculates control parameters in real time using a three-dimensional linear interpolation method when the aircraft is in a non-test point condition not directly calibrated in the parameter table, corresponding to the technical features of claim 9.

[0118] Assuming the aircraft is equipped with a piston engine, the current operating condition is as follows: =73% (not a multiple of 5%) =88% (not a multiple of 5%) =1250m (non-calibrated elevation point). The directly calibrated adjacent grid points in the parameter table are: ∈{70%,75%}, ∈{85%,90%}, ∈{1000m,1500m}, forming a total of 8 corner points.

[0119] The flight control system is first fixed and ,exist Linear interpolation is performed along the dimension. Let's assume 70% → 88% @ 1000m points. for 75%→88%@1000m point for The interpolation result for the 73% point is: Similarly, calculate the interpolation result for a height of 1500m. .

[0120] Subsequently, fixed =73%, in Interpolation in dimensionality: .

[0121] for Dimensionally, since 88% falls between 85% and 90%, the flight control system also needs to consider the above. - Based on the two-dimensional interpolation results, further along Linear interpolation is performed on the axis. This yields the complete parameter set for the non-test point. =80.6%, =1.12s, =0.82s, 0.30s.

[0122] The computational complexity of the three-dimensional linear interpolation method is O(1), which can be completed within 0.1ms on the main frequency processor of the flight control system, far less than the control cycle of 10ms, thus meeting the real-time requirements. The interpolation results ensure the continuity and smoothness of the control output, avoiding control quantity jumps and actuator jitter caused by parameter table discretization.

[0123] Example 8: Rapid Calibration and Universal Adaptation of New Engine Models

[0124] This embodiment illustrates the rapid adaptation process of the method of the present invention to a new engine model, corresponding to the technical features of claim 10.

[0125] A new type of turbine engine needs to be adapted to an existing medium-sized rotorcraft platform. Traditional methods require a complete bench calibration (approximately 200 operating conditions, taking 2 weeks), while this invention uses a "basic model + rapid calibration" approach, requiring only 3 sets of typical operating condition tests to generate a dedicated multi-dimensional adaptive control parameter table.

[0126] The three typical operating conditions selected are: ① steady-state point at 100% rated power at sea level (0m); ② maximum power increase step point at 50%→100% at sea level; ③ power increase step point at 75%→100% at the highest commonly used altitude (6000m). Through these three sets of tests, the basic response coefficients a and b of the engine (used for...) were obtained. and basic opening coefficient , , (for) ).

[0127] The flight control system fills the basic model parameters into the parameter table framework. For untested operating points, the theoretical values ​​are first calculated from the basic model, and then a large initial safety margin is assigned (e.g., ...). (Increase by 20%). In subsequent actual flights, the online learning module automatically collects the actual response error at each operating point and gradually corrects the parameter table. Usually, after 10 to 15 flight missions, the accuracy of the parameter table can reach a level comparable to that of a complete test bench calibration.

[0128] This rapid calibration method shortens the engine adaptation cycle for new engine models from 2 weeks to 2 days, and automatically adapts to the individual characteristics of different engines through online learning, demonstrating good universal adaptability.

[0129] Example 9: Detailed Timing and Message Design for CAN Bus Communication

[0130] This embodiment details the design of the physical layer, data link layer, and application layer of CAN bus communication between the flight control system and the ECU, corresponding to the technical features of claim 8.

[0131] Physical layer and data link layer: The flight control system and ECU communicate and interact via a CAN bus conforming to the ISO11898 standard, with a bus baud rate set to 1Mbps and a terminating resistor of 120Ω. Twisted-pair shielded transmission is used to suppress aviation electromagnetic interference.

[0132] Control cycle and synchronization mechanism: The control cycle is 10ms (100Hz), that is, the flight control system sends a control command to the ECU every 10ms; the synchronization cycle is 1ms, and the flight control system and the ECU ensure that the phase difference between their control clocks is less than 0.5ms through hardware synchronization signals (such as PWM synchronization pulse or CAN timestamp alignment), so as to ensure the accuracy of timing control.

[0133] Command message (flight controller → ECU, ID: 0x110): The message length is 8 bytes, and the data structure is as follows: Bytes 0~1 represent the power target value. (Resolution 0.1%, offset 0, range 0%~100%); bytes 2~3 are the target rotational speed values. (Resolution 1 r / min, range 0~65535 r / min); Byte 4 is the feedforward control reference opening. (Resolution 0.5%, range 0%~100%); Byte 5 is the mode word (bit 0: feedforward enable; bit 1: feedback enable; bit 2: security mode; bit 3: learning mode; bits 4~7: reserved); bytes 6~7 are the checksum (CRC16 low 8 bits, high 8 bits).

[0134] Status message (ECU→Flight Controller, ID: 0x210): The message length is 8 bytes, and the data structure is as follows: Bytes 0~1 represent the actual rotational speed. (Resolution 1 r / min); Bytes 2-3 are actual power (resolution 0.1%); Byte 4 is actual opening degree (resolution 0.5%); Byte 5 is the status word (bit 0: engine ready; bit 1: power meets target; bit 2: speed stable; bit 3: fault indicator; bit 4: overheat warning; bits 5-7: reserved); Bytes 6-7 are checksums.

[0135] Communication timing example: In At 0ms, the flight control system sends command message 0x110, which the ECU receives and parses within 1ms; the ECU then... Throttle valve actuation begins at 2ms; At 5ms, the ECU sends a status message 0x210 to report the current actual status; the flight control system then... The status message is received at 6ms and used for feedback calculation in the next control cycle. The entire closed-loop delay is approximately 6ms, which is much less than the 10ms control cycle, meeting the requirements for real-time control.

[0136] Abnormal communication handling: If the flight control system does not receive a 0x210 message for three consecutive cycles, or receives a CRC check error message, the security degradation process described in Example 6 will be triggered to ensure that the communication failure does not affect flight safety.

[0137] Through the detailed description of the above embodiments, the automatic speed control method for a fuel-powered rotorcraft engine of the present invention has been fully presented. This method achieves advance prediction and precise pre-adjustment of engine power adjustment by constructing a multi-dimensional adaptive control parameter table; and achieves this through dynamic advance... It achieves precise time-domain alignment of power control and attitude control; through a feedforward + incremental PID feedback composite control architecture, it balances response speed and steady-state accuracy; through an online learning iterative update mechanism, it endows the system with the ability to adapt to individual engine characteristics, fuel quality changes, and long-term mechanical degradation; through a comprehensive safety fault-tolerant control logic, it ensures flight safety under complex operating conditions and abnormal states.

[0138] Compared with traditional pure deviation feedback control, this invention reduces speed overshoot from ±15%~±25% to within ±2%, attitude fluctuation from ±5°~±8° to within ±0.8°, system convergence time from 4s to within 1.5s, attitude maintenance capability in complex environments is improved by more than 70%, engine mechanical shock is reduced by more than 60%, and fuel consumption is reduced by 5%~12%. This method can be widely applied to various rotorcraft unmanned aerial vehicles equipped with piston and turbine fuel engines, without requiring significant modifications to the original engine structure. It can be achieved solely through software algorithm optimization and parameter table construction, significantly reducing system modification costs and possessing outstanding substantive features and significant progress.

[0139] The above embodiments merely illustrate several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

[0140] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An automatic speed control method for a fuel-powered rotorcraft engine, characterized in that, Includes the following steps: The flight control system monitors changes in the aircraft's flight status in real time, identifies current power increase / decrease requirements, and determines the actual power output of the engine. With target power ; The flight control system calls a preset multi-dimensional adaptive control parameter table, according to the... , and current flight altitude Query to obtain engine power response time Flight control attitude loop inherent response time and feedforward control reference opening ; The flight control system is based on the formula Real-time calculation of power control advance time ; The flight control system sends the target speed value to the engine control unit (ECU) via the CAN bus. Power target value and feedforward control reference opening The ECU performs engine power feedforward pre-adjustment control; The delay mentioned Then, the flight control system sends attitude adjustment commands to the attitude control module to control the aircraft to adjust the propeller pitch to change the flight attitude; During engine power adjustment, the ECU collects the actual engine speed in real time. The flight control system is based on the actual engine speed collected in real time by the ECU. Calculate the speed deviation Incremental PID feedback control generates feedback control to correct the opening degree. and the feedforward control reference opening degree The final control opening is obtained by combining the results. The closed-loop correction is performed by the ECU; Real-time acquisition of actual operating condition data and calculation of correction coefficients The multi-dimensional adaptive control parameter table is updated online to synchronize engine power adjustment and aircraft attitude adjustment to achieve the target value, thus realizing automatic speed control.

2. The automatic speed control method for a fuel-powered rotorcraft engine according to claim 1, characterized in that, The multi-dimensional adaptive control parameter table is based on the current power... Target power Flight altitude This is a structured lookup table with a three-dimensional index, where each index cell contains four core data items: throttle body / expansion valve target opening. Estimated engine power arrival time Flight control attitude loop inherent response time Power control lead time .

3. The automatic speed control method for a fuel-powered rotorcraft engine according to claim 1, characterized in that, The multi-dimensional adaptive control parameter table was constructed using the following standardized test conditions: For piston-type or turbofan-type fuel engines, tests are conducted on engine bench testing systems, high-altitude simulators, rotor load tables, and flight data acquisition systems. The power range is based on the engine's rated power. Based on this, the equal-interval gradient method is used at 50%. ~100% Divide the range into steps of 5%. A test point includes a steady-state power point and a power step change point; Altitude and environmental conditions are divided as follows: piston engine test altitudes include 0m, 500m, 1000m, 1500m, 2000m, 3000m, 4000m, and 5000m; turbine engine test altitudes include 0m, 1000m, 2000m, 3000m, 4000m, 5000m, 6000m, 7000m, and 8000m; ambient temperature includes −20℃, 0℃, 10℃, 20℃, 30℃, 40℃, and 50℃; humidity includes 30%RH, 50%RH, and 80%RH. Each set of operating conditions is an independent combination of altitude, temperature, and humidity. The data acquisition channels include engine speed, torque, power, throttle / expansion valve opening, altitude, ambient temperature, intake air pressure, coolant temperature, and engine oil temperature. The acquisition frequency is 100Hz. After first-order low-pass filtering and moving average filtering, each set of operating conditions is tested at least 5 times, and samples are discarded. For outliers, the arithmetic mean is taken and stored in the parameter table.

4. The automatic speed control method for a fuel-powered rotorcraft engine according to claim 1, characterized in that, The parameters in the multi-dimensional adaptive control parameter table have the following mathematical mapping relationship: Power-aperture mapping: ; Power response time model: ,in ; Dynamic lead time calculation formula: .

5. The automatic speed control method for a fuel-powered rotorcraft engine according to claim 1, characterized in that, The feedforward pre-adjustment control and feedback closed-loop correction constitute a composite control architecture, wherein the feedforward path rapidly outputs a lag-free reference opening based on a multi-dimensional adaptive control parameter table. It provides over 90% adjustment to compensate for engine delay; The feedback path is based on an incremental PID algorithm according to the speed deviation. Real-time fine-tuning of the opening eliminates model errors and external disturbances, ensuring stable rotational speed.

6. The automatic speed control method for a fuel-powered rotorcraft engine according to claim 1, characterized in that, The online iterative update includes the following steps: Real-time acquisition of engine actual response time Compared with the estimated response time in the parameter table Deviation; Calculate the correction factor ; Update the data of the corresponding working point in the parameter table according to the gradient, without overwriting the original calibration value; For non-test point operating conditions, a three-dimensional linear interpolation method is used to output continuous control quantities, so that the parameter table can be adaptively optimized as engine degradation, fuel differences and individual characteristics change.

7. The automatic speed control method for a fuel-powered rotorcraft engine according to claim 1, characterized in that, It also includes safety and fault-tolerant control logic: Real-time monitoring of engine unresponsiveness, excessive power deviation, sudden attitude changes, and communication failures; When an anomaly is detected, dynamically execute at least one of the following degradation measures: recalculation Increase the weight of feedback control, interrupt the current posture action, and enter the safe constant speed mode; When the measured speed deviation e exceeds ±20 r / min to ±50 r / min, the PID feedback path immediately fine-tunes the valve / throttle valve opening to quickly stabilize the speed; when the measured response time deviation from the theoretical response time exceeds ±0.1 s to 0.2 s, it automatically updates. And calibrate .

8. The automatic speed control method for a fuel-powered rotorcraft engine according to claim 1, characterized in that, The flight control system and the ECU communicate and interact via a CAN bus, with a control cycle of 10ms and a synchronization cycle of 1ms. The command message ID sent by the flight control system to the ECU is 0x110, which includes the target power value, target speed value, feedforward opening degree, and mode word. The status message ID sent by the ECU to the flight control system is 0x210, which includes the actual speed, actual power, actual opening degree, and status word.

9. The automatic speed control method for a fuel-powered rotorcraft engine according to claim 2, characterized in that, When the current operating point queried by the flight control system is not a test point in the multi-dimensional adaptive control parameter table, the target opening is calculated using the three-dimensional linear interpolation method. Power response time and power control lead time The interpolation dimension is , , This is to ensure the continuity and smoothness of the control output.

10. The automatic speed control method for a fuel-powered rotorcraft engine according to claim 1, characterized in that, The method is applicable to rotorcraft unmanned aerial vehicles equipped with piston-type or turbofan-type fuel engines. When adapting to new engine models, a basic model plus rapid calibration method is adopted, requiring only 3 sets of typical operating condition tests to generate a dedicated multi-dimensional adaptive control parameter table, and automatically adapting to the individual characteristics of different engines through online learning.

Images