阅读说明：本技术 变旋翼转速直升机/涡轴发动机综合控制方法、装置 (Helicopter/turboshaft engine comprehensive control method and device with variable rotor rotation speed ) 是由 汪勇 宋劼 陈浩颖 郑前钢 张海波 于 2020-07-01 设计创作，主要内容包括：本发明公开了一种变旋翼转速直升机/涡轴发动机综合控制方法,包括以下步骤：超前预测直升机旋翼需求扭矩；根据直升机旋翼需求扭矩预测结果获得发动机期望输出扭矩,并与涡轴发动机真实输出扭矩作差,经比例环节的增益放大后生成前馈补偿项；根据动力涡轮相对转速与动力涡轮参考指令之间的误差和阈值之间的关系判断是否用所述前馈补偿项对涡轴发动机内环燃气涡轮转速的控制指令进行修正：如所述误差的绝对值大于等于阈值,则进行修正,否则,不进行修正。本发明还公开了一种变旋翼转速直升机/涡轴发动机综合控制装置。本发明可减小动力涡轮转速的超调与下垂量,抑制变旋翼转速过程中旋翼需求功率的干扰,实现直升机/涡轴发动机高品质综合控制。(The invention discloses a comprehensive control method for a variable rotor speed helicopter/turboshaft engine, which comprises the following steps: predicting the required torque of a helicopter rotor in advance; obtaining the expected output torque of the engine according to the prediction result of the required torque of the helicopter rotor, making a difference with the real output torque of the turboshaft engine, and generating a feedforward compensation term after gain amplification in a proportion link; judging whether the feedforward compensation term is used for correcting the control command of the rotating speed of the inner ring gas turbine of the turboshaft engine according to the relation between the error between the relative rotating speed of the power turbine and the reference command of the power turbine and a threshold value: and if the absolute value of the error is larger than or equal to the threshold value, correcting, otherwise, not correcting. The invention also discloses a comprehensive control device of the helicopter/turboshaft engine with the variable rotor rotating speed. The invention can reduce the overshoot and the sag of the rotating speed of the power turbine, inhibit the interference of the power required by the rotor wing in the process of changing the rotating speed of the rotor wing and realize the high-quality comprehensive control of the helicopter/turboshaft engine.)

1. A comprehensive control method for a variable rotor speed helicopter/turboshaft engine is characterized by comprising the following steps: predicting the required torque of a helicopter rotor in advance; obtaining the expected output torque of the engine according to the prediction result of the required torque of the helicopter rotor, making a difference with the real output torque of the turboshaft engine, and generating a feedforward compensation term after gain amplification in a proportion link; judging whether the feedforward compensation term is used for correcting the control command of the rotating speed of the inner ring gas turbine of the turboshaft engine according to the relation between the error between the relative rotating speed of the power turbine and the reference command of the power turbine and a threshold value: and if the absolute value of the error is larger than or equal to the threshold value, correcting, otherwise, not correcting.

2. A method for integrated control of a variable rotor speed helicopter/turboshaft engine according to claim 1 wherein said feed forward compensation term Δ is generated according to the equation:

wherein K is the gain amplification factor of the proportional element, T_qrFor the predicted result of the helicopter rotor torque demand, omega_MR、Ω_ptRespectively, the rotor rotation speed, the power turbine real physical rotation speed, T_qeAnd outputting the real torque for the turboshaft engine.

3. The integrated control method for a variable rotor speed helicopter/turboshaft engine according to claim 1, wherein the helicopter rotor demand torque is predicted in advance using a rotor demand torque prediction model based on a deep neural network; the rotor wing demand torque prediction model based on the deep neural network is obtained by taking the rotor wing total distance, the flight height, the forward flight speed and the rotor wing rotating speed at the historical moment as input, taking the rotor wing demand torque as output and adopting a deep neural network algorithm for off-line training.

4. The method for integrated control of a variable rotor speed helicopter/turboshaft engine according to claim 3, wherein the deep neural network based rotor demand torque prediction model hidden layer and output layer activation functions are a linear rectification function and a linear function, respectively.

5. The method for integrated variable rotor speed helicopter/turboshaft engine control according to claim 1 wherein said threshold value is 0.2.

6. A variable rotor speed helicopter/turboshaft engine integrated control device is characterized by comprising:

the rotor wing demand torque prediction model is used for predicting the helicopter rotor wing demand torque in advance;

the torque difference feedforward module is used for obtaining the expected output torque of the engine according to the predicted result of the required torque of the helicopter rotor wing, making a difference with the real output torque of the turboshaft engine, and generating a feedforward compensation term after gain amplification in a proportion link;

the self-adaptive feedforward structure is used for judging whether the feedforward compensation term is used for correcting the control command of the rotating speed of the inner ring gas turbine of the turboshaft engine according to the relation between the error between the relative rotating speed of the power turbine and the reference command of the power turbine and a threshold value: and if the absolute value of the error is larger than or equal to the threshold value, correcting, otherwise, not correcting.

7. The variable rotor speed helicopter/turboshaft engine complex control apparatus of claim 6 wherein said feed forward compensation term Δ is generated according to the following equation:

wherein K is the gain amplification factor of the proportional element, T_qrFor the predicted result of the helicopter rotor torque demand, omega_MR、Ω_ptRespectively, the rotor rotation speed, the power turbine real physical rotation speed, T_qeAnd outputting the real torque for the turboshaft engine.

8. The variable rotor speed helicopter/turboshaft engine integrated control apparatus according to claim 6, wherein the rotor demand torque prediction model is a deep neural network-based rotor demand torque prediction model obtained by offline training using a deep neural network algorithm, taking the rotor collective pitch, the altitude, the forward speed, and the rotor speed at the historical time as inputs, and taking the rotor demand torque as an output.

9. The variable rotor speed helicopter/turboshaft engine integrated control apparatus according to claim 8, wherein the deep neural network based rotor demand torque prediction model hidden layer and output layer activation functions are a linear rectification function and a linear function, respectively.

10. The variable rotor speed helicopter/turboshaft engine complex control apparatus of claim 6 wherein said threshold value is 0.2.

Technical Field

The invention relates to a comprehensive control method for a variable rotor speed helicopter/turboshaft engine, and belongs to the technical field of system control and simulation in aerospace propulsion theory and engineering.

Background

The variable rotor rotation speed technology enables the next generation of high-speed helicopters to break through the limit of flight speed, the maximum speed per hour can reach 463 km/h, and is 2 times of the speed of American active 'eagle' helicopters, and 1.5 times of the speed of 'Apache' helicopters. It is expected that the Variable rotor speed technology will become the key aircraft technology for the continuous development of military strong countries [ kalin DV. multistreamed continuous flight Variable Transmission Synthesis for Next-Generation Helicopters [ C ] ].

The existing research shows that: the variable-speed-variable-wing rotating speed technology is realized by two power schemes, namely, the variable rotating speed is realized by matching a multi-stage gearbox with a conventional fixed-rotating-speed turboshaft engine, and the variable rotating speed is realized by combining a conventional fixed-transmission-ratio speed change mechanism through a variable-power turboshaft engine. No matter what kind of speed change mode is adopted, the problem of coupling between the complex power output of the turboshaft engine and the aerodynamic characteristics of the helicopter [ Amri H, Feil R, Hajek M, et al. Possibilities and differences for rotorcraft using variable transmission drive [ J ]) needs to be solved, and research on the comprehensive control method of the variable-rotor-speed helicopter/turboshaft engine is developed. The traditional design method of the flight and engine control system does not consider the coupling among the airplane, the engine and the control system, the airplane or the engine is controlled independently, the stability of a subsystem is generally considered fully, relatively large safety margin is reserved, the performance of the whole system is sacrificed to a certain extent, and the requirement of the development of a variable rotor speed helicopter is difficult to adapt [ variable speed helicopter/transmission system/engine comprehensive modeling and control research [ D ] ].

At present, a turboshaft engine widely adopts a cascade PID control structure based on total pitch feedforward, and takes the total pitch of a rotor wing as a comprehensive cross-linking parameter of a helicopter and an engine to carry out comprehensive control on the helicopter/turboshaft engine [ Argentim L M, RezendeW C, Santos P E, et al. However, in the process of changing the rotating speed of the rotor, the rotating speed of the rotor changes in a large range, and the total distance is only used for representing the power required by the rotor, so that the method is not applicable any more, and needs to be combined with measurable parameters of a helicopter and an engine to optimize and select cross-linking parameters to form a new helicopter/engine comprehensive information cross-linking scheme so as to meet the requirement of high-quality control of the rotating speed of the variable rotor.

HPW3000 engine control [ Smith B.J, Zagreski R.D. Next Generation control System for Helicopter Engines [ C ] ] predicts the transient change of the rotor torque through the neural network, is used as the engine feed-forward compensation link, and has a high-speed slow-vehicle operation control mode with guide vane advance, thereby further improving the slow-vehicle-high-power state acceleration response capability of the system. The power turbine rotating speed controller of the turboshaft engine is designed based on an internal model principle by a turboshaft engine internal model control method [ J ] based on predicted torque feedforward. Aiming at the interference of the torque change of the main rotor on the rotating speed of the power turbine, a torque prediction method based on an extreme learning machine is provided, and the effective compensation of the interference on the load change of the engine is realized. The Sunpiko provides a helicopter torque dynamic advanced prediction model design method based on an iterative reduction least square support vector regression algorithm and a model prediction mechanism, and a torque-based compensation controller of a turboshaft engine is designed according to torque advanced prediction information, so that adverse effects of large torque disturbance on the turboshaft engine in the maneuvering process of the helicopter are favorably inhibited. The above studies were essentially developed around fixed rotor speed, not involving variable rotor situations; in addition, when the rotating speed of the rotor wing changes in a large range, the required torque of the main rotor wing cannot accurately reflect the required power of the helicopter to a certain extent, so that the method cannot be applied to the comprehensive control of the helicopter/turboshaft engine with the variable rotating speed of the rotor wing.

Disclosure of Invention

The invention aims to overcome the defects of the prior art and provide a comprehensive control method for a variable-rotor-speed helicopter/turboshaft engine, which can reduce the overshoot and the droop of the power turbine speed, effectively inhibit the interference of the power required by a rotor in the process of changing the rotor speed and realize the high-quality comprehensive control of the helicopter/turboshaft engine on the premise of ensuring the stable operation of the helicopter/engine.

The invention specifically adopts the following technical scheme to solve the technical problems:

a comprehensive control method for a variable rotor speed helicopter/turboshaft engine comprises the following steps: predicting the required torque of a helicopter rotor in advance; obtaining the expected output torque of the engine according to the prediction result of the required torque of the helicopter rotor, making a difference with the real output torque of the turboshaft engine, and generating a feedforward compensation term after gain amplification in a proportion link; judging whether the feedforward compensation term is used for correcting the control command of the rotating speed of the inner ring gas turbine of the turboshaft engine according to the relation between the error between the relative rotating speed of the power turbine and the reference command of the power turbine and a threshold value: and if the absolute value of the error is larger than or equal to the threshold value, correcting, otherwise, not correcting.

Preferably, the feedforward compensation term Δ is generated according to the following equation:

wherein K is the gain amplification factor of the proportional element, T_qrFor the predicted result of the helicopter rotor torque demand, omega_MR、Ω_ptRespectively, the rotor rotation speed, the power turbine real physical rotation speed, T_qeAnd outputting the real torque for the turboshaft engine.

Preferably, the helicopter rotor demand torque is predicted in advance by using a rotor demand torque prediction model based on a deep neural network; the rotor wing demand torque prediction model based on the deep neural network is obtained by taking the rotor wing total distance, the flight height, the forward flight speed and the rotor wing rotating speed at the historical moment as input, taking the rotor wing demand torque as output and adopting a deep neural network algorithm for off-line training.

Further preferably, the deep neural network-based rotor demand torque prediction model hidden layer and the output layer activation function are a linear rectification function and a linear function respectively.

Preferably, the threshold is 0.2.

The following technical solutions can also be obtained according to the same inventive concept:

a variable rotor speed helicopter/turboshaft engine integrated control device comprises:

the rotor wing demand torque prediction model is used for predicting the helicopter rotor wing demand torque in advance;

the torque difference feedforward module is used for obtaining the expected output torque of the engine according to the predicted result of the required torque of the helicopter rotor wing, making a difference with the real output torque of the turboshaft engine, and generating a feedforward compensation term after gain amplification in a proportion link;

the self-adaptive feedforward structure is used for judging whether the feedforward compensation term is used for correcting the control command of the rotating speed of the inner ring gas turbine of the turboshaft engine according to the relation between the error between the relative rotating speed of the power turbine and the reference command of the power turbine and a threshold value: and if the absolute value of the error is larger than or equal to the threshold value, correcting, otherwise, not correcting.

Preferably, the feedforward compensation term Δ is generated according to the following equation:

wherein K is the gain amplification factor of the proportional element, T_qrFor the predicted result of the helicopter rotor torque demand, omega_MR、Ω_ptRespectively, the rotor rotation speed, the power turbine real physical rotation speed, T_qeAnd outputting the real torque for the turboshaft engine.

Preferably, the rotor demand torque prediction model is a rotor demand torque prediction model based on a deep neural network, and is obtained by taking the rotor total pitch, the flying height, the forward flying speed and the rotor speed at the historical moment as inputs, taking the rotor demand torque as an output and performing offline training by adopting a deep neural network algorithm.

Further preferably, the deep neural network-based rotor demand torque prediction model hidden layer and the output layer activation function are a linear rectification function and a linear function respectively.

Preferably, the threshold is 0.2.

Compared with the prior art, the technical scheme of the invention has the following beneficial effects:

the method combines the rotor dynamics characteristics of the output shaft end of the variable-rotor-wing-rotating-speed turboshaft engine, performs feedforward compensation correction on the control instruction of the rotating speed of the inner ring gas turbine of the turboshaft engine based on the error of the predicted output torque and the real output torque of the engine, reduces the overshoot and the droop of the rotating speed of the power turbine on the premise of ensuring the stable operation of the helicopter/engine, effectively inhibits the interference of the required power of the rotor wing in the variable-rotor-wing-rotating-speed process, and realizes the high-quality comprehensive control of the helicopter/turboshaft engine.

Drawings

FIG. 1 is a block diagram showing a specific structure of the integrated control device for a variable rotor speed helicopter/turboshaft engine according to the present invention;

FIG. 2 is a plot of rotor demand torque prediction model training errors;

FIG. 3 is a graph of the change in forward flight speed of a helicopter;

FIG. 4 is a graph comparing the variation of the power turbine speed of the turboshaft engine;

FIG. 5 is a graph comparing rotor speed profiles;

FIG. 6 is a graph comparing the fuel flow rate variation of the turboshaft engine;

FIG. 7 is a graph comparing the variation of the relative rotational speed of the gas turbine of the turboshaft engine;

FIG. 8 is a graph comparing the output torque curves of a turboshaft engine;

FIG. 9 is a comparison of rotor demand torque curves;

figure 10 is a comparison of rotor collective pitch curves.

Detailed Description

Aiming at the defects in the prior art, the solution idea of the invention is to combine the rotor dynamics characteristics of the output shaft end of the variable-rotor-speed turboshaft engine and perform feedforward compensation correction on the control instruction of the rotating speed of the inner ring gas turbine of the turboshaft engine based on the error between the predicted output torque and the real output torque of the engine.

Specifically, the comprehensive control method for the helicopter/turboshaft engine with variable rotor rotation speed provided by the invention comprises the following steps: predicting the required torque of a helicopter rotor in advance; obtaining the expected output torque of the engine according to the prediction result of the required torque of the helicopter rotor, making a difference with the real output torque of the turboshaft engine, and generating a feedforward compensation term after gain amplification in a proportion link; judging whether the feedforward compensation term is used for correcting the control command of the rotating speed of the inner ring gas turbine of the turboshaft engine according to the relation between the error between the relative rotating speed of the power turbine and the reference command of the power turbine and a threshold value: and if the absolute value of the error is larger than or equal to the threshold value, correcting, otherwise, not correcting.

To facilitate understanding of the public, the technical solution of the present invention is further described in detail by a specific embodiment in combination with the attached drawings:

as shown in fig. 1, the integrated control device for a variable rotor speed helicopter/turboshaft engine in the present embodiment mainly comprises 3 parts, namely a rotor demand torque prediction model, a torque difference feedforward module and an adaptive feedforward structure. The rotor wing demand torque prediction model is used for predicting the helicopter rotor wing demand torque in advance, and can be established in an off-line or on-line mode by adopting various existing helicopter rotor wing demand torque advance prediction schemes; the torque difference feedforward module is used for obtaining the expected output torque of the engine according to the predicted result of the required torque of the helicopter rotor wing, making a difference with the real output torque of the turboshaft engine, and generating a feedforward compensation term after gain amplification in a proportion link; the self-adaptive feedforward structure is used for judging whether the feedforward compensation term is used for correcting the control command of the rotating speed of the inner ring gas turbine of the turboshaft engine according to the relation between the error between the relative rotating speed of the power turbine and the reference command of the power turbine and the threshold value: and if the absolute value of the error is larger than or equal to the threshold value, correcting, otherwise, not correcting. As shown in fig. 1, in a single step, the rotor demand torque prediction model predicts the rotor demand torque on line according to the input quantity, and generates a torque difference feedforward compensation term by combining the real physical rotation speed of the power turbine, the rotation speed of the rotor, the output torque of the engine and the proportion link, and automatically selects whether to use the self-adaptive feedforward structure or notAnd connecting a torque difference feedforward and a command for correcting the relative rotating speed of the inner ring gas turbine. The dynamic response of the rotating speed of the turboshaft engine is improved, and the stable control precision is considered, so that the helicopter/turboshaft engine with variable rotating speed of the rotor wing is comprehensively controlled. The structure shown in figure 1 has two different modes to realize the goal of changing the rotating speed of the rotor wing, one is that a transmission command is set as a fixed transmission ratio, and a command N for changing the relative rotating speed of a power turbine_prChanging the rotating speed of the rotor wing, setting a variable transmission instruction to adjust the reduction ratio of the transmission mechanism, and then, setting an instruction N of the relative rotating speed of the power turbine_prKeep 100% unchanged. In the figure, H, v_c、Ω_MR、T_qrRespectively the flying height, the forward flying speed, the rotating speed of the rotor wing and the torque required by the rotor wing; w_fb,、N_g、N_p、Ω_pt、T_qeRepresenting fuel flow, gas turbine relative speed, power turbine true physical speed, and engine true output torque, respectively.

The following is a further detailed description of the main components:

1) rotor wing demand torque prediction model

The main rotor serves as the most important part of the helicopter, the transient torque at the blade tip not only contains the manipulated variable information and the dynamic information of the helicopter body and represents the power requirement of the helicopter, but also serves as the external excitation of a turboshaft engine, and the transient change of the torque can influence the dynamic response of the engine. However, since helicopters basically use flexible or semi-rigid and semi-flexible blades, the measured value of the torque at the shaft of the blade has obvious lag relative to the change of the manipulated variable, and the method cannot be directly applied, so that the rotor demand torque needs to be predicted in advance. The dynamic characteristics of the original helicopter rotor system need to be maintained as much as possible when predicting the rotor demand torque. The invention preferably adopts a deep neural network for off-line training to improve the accuracy of the rotor wing demand torque prediction model.

Let rotor torque prediction model be 2-order system, the expression is as follows, k represents sampling time.

Wherein the output is T of the rotor at the current moment_qrThe input quantity is H and v at the current and historical moments_c、θ₀And omega of historical time_MR. Thus, the entire rotor prediction model contains 11 inputs and 1 output. In addition, when the deep neural network is adopted for offline training, 2 hidden layers are arranged, and the number of nodes of each hidden layer is respectively set to be 30 and 18. In order to prevent gradient disappearance or explosion during deep neural training, the hidden layer and the output layer activation functions are respectively set as a linear rectification function (ReLU) and a linear function (purelin).

Under different operating conditions, fully exciting the helicopter/turboshaft engine comprehensive model with the variable rotor rotating speed, and carrying out normalization processing on the acquired data to serve as training data of the deep neural network. Fig. 2 shows relative errors of the rotor demand torque prediction model established by the invention, wherein the relative errors are less than 0.1%, the precision is high, and the rotor demand torque prediction model can be used for predicting the rotor demand torque in advance.

2) Self-adaptive feedforward compensation method based on output torque difference and proportion link of turboshaft engine

The inertia of the rotor wing is large, the rotating speed of the power turbine cannot respond to the change of the required power of the helicopter in time, and therefore the rotating speed has certain delay. Especially in the process of changing the rotating speed of the rotor, the rotating speed of the rotor is changed in a large range, and the response lag of the rotating speed of the power turbine to the required power is more serious. Therefore, it is necessary to optimally select the relevant parameters of the helicopter and the turboshaft engine to form a new cross-linking method and a new comprehensive control method.

The self-adaptive feedforward compensation method based on the output torque difference and proportion link of the turboshaft engine combines the rotor rotation speed and the real physical rotation speed (omega) of the power turbine according to the rotor dynamics characteristic of the output shaft end of the turboshaft engine on the basis of the rotor demand torque advance prediction_pt) Obtaining the expected output torque of the engine and the real output torque (T) of the turboshaft engine_qe) Performing difference, amplifying by gain K of proportion link to generate feedforward compensation term (delta), and online correcting control instruction of inner ring gas turbine rotation speed of turboshaft engine to inhibitThe interference of the required power of the helicopter in the process of changing the rotating speed of the rotor. In addition, in order to further reduce the influence of the possible torque prediction error on the steady-state control of the rotating speed of the turboshaft engine, the invention introduces an adaptive feedforward compensation structure according to the relative rotating speed (N) of the power turbine_p) And its reference instruction (N)_pr) The error of the method is automatically selected whether to access a feedforward compensation item, so that the purpose of self-adaptive feedforward compensation is achieved. The specific adaptive feedforward compensation can be expressed by the following formula:

wherein, T_qrThe predicted value of the rotor wing required torque is a threshold value; the sat function is defined as follows:

here, 0.2 is set. When N is present_pAnd N_prWhen the absolute value of the error is more than or equal to 0.2, a torque difference feedforward compensation item is accessed for online correcting the instruction of the relative rotating speed of the gas turbine, the fuel flow of the engine is compensated in advance, the overshoot and the droop of the rotating speed of the power turbine in the process of changing the rotating speed of the rotary wing are reduced, and the high-quality comprehensive control of the helicopter/turboshaft engine with the changed rotating speed of the rotary wing is realized; on the contrary, a feedforward compensation term is not accessed, and at the moment, the turboshaft engine only adopts a cascade control structure based on the relative rotating speed of the gas turbine.

In order to verify the effect of the technical scheme, on the basis of realizing variable rotor rotation speed by variable transmission ratio, a comprehensive control effect simulation test based on torque difference feedforward compensation is carried out, and at the moment, a reference instruction N of the relative rotation speed of the power turbine is used_prSet to 100%. The flying height H of the helicopter is 1000m, and the front flying speed v of the helicopter is_cAs shown in fig. 3. V when t is 10s_cLinear acceleration from 118m/s to 123 m/s; between 15 and 45s, v_cKeeping 123m/s unchanged; followed by 5s, v_cLinearly decelerating to 118m/s and keeping until the simulation is finished, wherein the specific simulation results are shown in fig. 4-10.

As can be seen from fig. 3 and 5, when t is 10s, the helicopter flies forward and accelerates; thereafter, the rotor speed is rapidly reduced by 38%. When t is 33s, the rotating speed of the rotor wing is rapidly increased to 18.5rad/s, and the helicopter flies forward and accelerates after 10 s. In the rotor speed reduction/forward flight acceleration phase, the power turbine relative speed shown in fig. 4 is 100% below the set point. This is because the pull increases during acceleration, and the rotor speed decreases at this time, and the rotor collective pitch needs to be increased in order to keep the forward flight speed constant (as shown in fig. 10). As can be seen from fig. 4, the feedforward control method based on the torque difference effectively reduces the overshoot of the relative rotational speed of the power turbine by about 13.8% as compared to the collective feedforward control. In addition, the power turbine has shorter stabilization time relative to the rotating speed, faster response speed and better dynamic performance. This is because the power demand of the rotor changes sharply when the rotational speed of the rotor is changed, and when the rotational speed of the rotor decreases, the power demand of the rotor decreases, as can be seen from equation (2), and at this time, the desired output torque of the engine decreases accordingly, and the error from the actual output torque of the engine is negative. The negative error is amplified by a proportion link and then used as a feedforward term, when the absolute value of the relative rotating speed of the power turbine and the reference instruction error is more than or equal to 0.2, the feedforward compensation term corrects the instruction of the relative rotating speed of the gas turbine on line, the fuel flow is compensated in advance, the output torque of the engine is reduced, the overshoot of the relative rotating speed of the power turbine is obviously reduced, and vice versa. Fig. 6 shows the fuel flow rate profile, for a rotor with reduced speed and reduced power demand, resulting in a rapid reduction in fuel flow and a reduction in output power from the turboshaft engine. As can be seen from fig. 7 and 8, in the stage of the upshift of the rotor speed, the relative rotation speed of the gas turbine does not exceed 102% by using the two feedforward control methods, the peak value of the output torque of the engine does not exceed 5kN · m, and the service life of the engine is not affected. As shown in fig. 9, when the rotor speed is continuously reduced, the total rotor pitch is increased to increase the torque required by the rotor in order to ensure that the rotor can provide sufficient tension.