阅读说明：本技术 一种滚动优化预测闭环控制器及系统 (Rolling optimization prediction closed-loop controller and system ) 是由 张园锁 李运华 袁璠 杨思幸 江春宇 于 2020-06-30 设计创作，主要内容包括：本发明涉及发动机推力控制领域,更具体的说,涉及一种滚动优化预测闭环控制器及系统。本发明提出的一种滚动优化预测闭环控制器,包括推力管理器；信号处理器；控制器系数调度器,接收信号处理器发送的信号,结合存储的离线获取的信息,向推力预测闭环控制器提供所需要的发动机动态模型信息；推力预测闭环控制器,根据发动机动态模型信息和信号处理器提供的第一发动机状态信息,进行推力预测闭环所需要的燃油控制量的计算；燃油控制量选择器,对燃油控制量进行选择,输出对应的燃油指令。本发明可以实现发动机动态运行推力预测的闭环控制,改进了发动机推力控制系统开发的效率,改善了发动机动态过程的控制性能指标。(The invention relates to the field of engine thrust control, in particular to a rolling optimization prediction closed-loop controller and system. The invention provides a rolling optimization prediction closed-loop controller, which comprises a thrust manager; a signal processor; the controller coefficient scheduler receives the signal sent by the signal processor, combines the stored offline acquired information and provides the required dynamic model information of the engine for the thrust prediction closed-loop controller; the thrust prediction closed-loop controller is used for calculating the fuel control quantity required by the thrust prediction closed loop according to the dynamic model information of the engine and the first engine state information provided by the signal processor; and the fuel control quantity selector is used for selecting the fuel control quantity and outputting a corresponding fuel instruction. The invention can realize the closed-loop control of the thrust prediction of the dynamic operation of the engine, improve the development efficiency of the thrust control system of the engine and improve the control performance index of the dynamic process of the engine.)

1. A rolling optimization prediction closed-loop controller is characterized by comprising a thrust manager, a signal processor, a controller parameter scheduler, a thrust prediction closed-loop controller and a fuel control quantity selector, wherein the thrust manager comprises a first module, a second module and a third module, wherein the thrust manager comprises a thrust prediction closed-loop controller and a fuel control quantity selector:

the thrust manager calculates an expected thrust command of the engine operation according to the throttle lever angle signal and the thrust mode signal and outputs the expected thrust command to the thrust prediction closed-loop controller;

the signal processor is used for completing the calculation of the real physical numerical value of the converted signal according to the acquired original sensor signal and outputting the real physical numerical value to the corresponding module;

the controller coefficient scheduler receives a turbofan engine high-pressure rotor converted rotating speed signal sent by the signal processor, and provides required engine dynamic model information for the thrust prediction closed-loop controller by combining stored offline acquired information;

the thrust prediction closed-loop controller is used for calculating the fuel control quantity required by the thrust prediction closed loop according to the engine dynamic model information provided by the controller coefficient scheduler and the first engine state information provided by the signal processor and sending the calculation result to the fuel control quantity selector;

and the fuel control quantity selector is used for selecting the fuel control quantity and outputting a corresponding fuel instruction.

2. The roll optimized predictive closed loop controller of claim 1, wherein the engine dynamics model is as follows:

X＝[π,Δ]^T，u＝[α_I,β]^T；π＝Ps3/P0；

W_f(t)＝α_I·Δ(t)+β(t)；

wherein f is the engine dynamic equation, X is the motion state of the dynamic equation, and u is the dynamic equation inputQuantity, π is the falling sub-ratio of the entire engine, α_IIs a linear coefficient, beta is a gain coefficient, Ps3 is the high-pressure compressor outlet static pressure, P0 is the engine working environment temperature, F_dFor engine operating thrust expectation, F_eActual thrust produced for engine operation, W_fIs the fuel control quantity.

3. The rolling optimization prediction closed-loop controller of claim 2, wherein the thrust prediction closed-loop controller comprises an initial value calculation module, a discrete optimization matrix calculation module, an initial control law calculation module, a sampling value update calculation module, a Jacobian matrix calculation module, a control law change rate calculation module, and a recursive least squares calculation module:

the initial value calculation module presets the initial value of the corresponding variable;

the discrete optimization matrix calculation module calculates the optimal matrix of the controller through a C matrix and an H matrix;

the initial control rule calculation module completes approximate solution calculation of the initial control rule in the least square sense based on the recursive least square calculation module according to the input parameters of the initial value calculation module and the H matrix and the C matrix input by the discrete optimal matrix calculation module, and outputs the result to the control rule calculation module;

the control rule calculation module calculates the fuel control quantity of each beat of unit time according to the initial control rule and the change rate of the control rule, outputs the result to the sampling value updating calculation module, and sends the final fuel control quantity result to the fuel control quantity selector;

the sampling value updating calculation module is used for updating and calculating the sampling value of the fuel control quantity in each beat of unit time according to the calculation value of the control rule calculation module and outputting a result to the Jacobian matrix calculation module;

the Jacobian matrix calculation module carries out updating calculation of the Jacobian matrix in each beat of unit time according to the input parameters of the initial value calculation module, and outputs the result to the control rule change rate calculation module;

the control rule change rate calculation module completes approximate solution calculation of the control rule change rate in the least square sense based on the recursive least square calculation module according to the input parameters of the initial value calculation module and outputs the control rule change rate to the control rule calculation module;

and the recursive least square calculation module completes the least square meaning equation solution calculation of the input equation by using a rotation transformation method.

4. The roll optimization predictive closed-loop controller of claim 3, further comprising a steady-state closed-loop controller and a thrust planning converter:

the thrust plan converter is connected with the thrust manager, and performs target calculation of steady-state control according to an expected engine running thrust command sent by the thrust plan converter, calculates to obtain a target control rotating speed of a low-pressure shaft rotating speed of the turbofan engine and a target control rotating speed of a high-pressure shaft rotating speed of the turbofan engine, and outputs the target control rotating speeds to the steady-state closed-loop controller;

the controller coefficient scheduler provides required control parameters and an existing rule calculation output function for the steady-state closed-loop controller;

and the steady-state closed-loop controller is connected with the thrust planning converter, calculates the steady-state fuel control quantity of the engine required by the rotating speed closed loop according to the steady-state controller control parameter information provided by the controller coefficient scheduler and the second engine state information provided by the signal processor, and sends the calculation result to the fuel control quantity selector.

5. The roll optimized predictive closed-loop controller of claim 4, wherein the method of calculating the steady state fuel control quantity of the engine uses a proportional-derivative-integral controller, a sliding mode controller and state feedback.

6. The roll optimization predictive closed-loop controller of claim 4, further comprising a transition state controller:

the thrust planning converter carries out target calculation of transition state control according to the expected thrust command of the engine operation, calculates the change rate of the target control rotating speed of the low-pressure shaft rotating speed of the turbofan engine and the change rate of the target control rotating speed of the high-pressure shaft rotating speed of the turbofan engine and outputs the change rates to the transition state controller

The controller coefficient scheduler provides required control parameters and an existing rule calculation output function for the transition state controller;

and the transition state controller is connected with the thrust planning converter, calculates the engine acceleration fuel control quantity and the engine deceleration fuel control quantity required by acceleration control according to the transition state controller control parameter information provided by the controller coefficient scheduler and the third engine state information provided by the signal processor, and sends the calculation result to the fuel control quantity selector.

7. The roll optimized predictive closed-loop controller of claim 6, wherein the engine acceleration fuel control amount and the engine deceleration fuel control amount are calculated using a proportional-derivative-integral controller, a sliding mode controller, and a feed-forward series integral.

8. The rolling optimization predictive closed-loop controller of claim 6, further comprising an engine state scheduler that determines a state of the engine based on information provided by the signal processor, and that completes setting of an operating strategy signal for the engine based on the current state, and outputs the operating strategy signal to the fuel control quantity selector.

9. The rolling optimization predictive closed-loop controller of claim 8, wherein the fuel control quantity selector selects the fuel control quantities sent by the thrust predictive closed-loop controller, the steady-state closed-loop controller and the transition-state controller according to an operating strategy signal from the engine state scheduler and outputs a corresponding fuel command.

10. The rolling optimization predictive closed-loop controller of claim 9, wherein the operating strategy selection logic of the fuel control quantity selector comprises:

when the working strategy is in a thrust prediction mode, the fuel instruction adopts the fuel control quantity sent by the thrust prediction closed-loop controller, and the steady state and the transition state are not distinguished;

and when the working strategy is in a normal rotating speed control mode, the fuel instruction calculation is completed by adopting a high-low selection rule based on the fuel control quantity sent by the transition state controller and the steady state closed loop controller.

11. The rolling optimization predictive closed-loop controller of claim 9, further comprising a servo actuator closed-loop controller for converting the fuel command output from the fuel control quantity selector into a desired position signal of the metering mechanism, and receiving an actual position feedback signal of the metering mechanism input from the signal processor, and calculating the servo control current quantity according to the deviation between the desired position signal and the actual position feedback signal.

12. An engine roll optimization predictive closed-loop control system comprising the roll optimization predictive closed-loop controller of claims 1-11 and on-board sensors, actuators:

the airborne sensor is connected with the rolling optimization prediction closed-loop controller, acquires an engine body signal and sends the engine body signal to the rolling optimization prediction closed-loop controller;

and the executing mechanism is connected with the rolling optimization prediction closed-loop controller and executes corresponding actions according to the control instructions output by the rolling optimization prediction closed-loop controller.

Technical Field

The invention relates to the field of engine thrust control, in particular to a rolling optimization prediction closed-loop controller and system.

Background

After the engineering application field of the aircraft engine is developed to the full-authority digital technology era, the requirements related to the safe operation protection of the engine are continuously improved, particularly the accuracy of thrust control in the dynamic operation process, and a large amount of research and development resource research is invested in various aircraft engine manufacturers and related scientific research institutions all over the world.

At present, few reports are reported on OEM (Original Equipment Manufacturer) plans of various large engines, an onboard controller is adopted to directly perform closed-loop control on thrust of the engine, the relation between the calibrated thrust and the rotating speed on a test bed (ground, flight and the like) is generally adopted in engineering, and the aim of controlling the thrust is achieved by adopting a conversion rotating speed table and a relation curve of the thrust to control the conversion rotating speed.

The method has the characteristics that various running scenes of the engine need to be subjected to trial run calibration, the development cost is high, and the dynamic self-adaptive capacity is poor.

The invention CN201710931188.5 of China adopts the idea of model predictive control, and provides a hybrid power ship energy management method based on model predictive control; a Markov model is selected to predict the required power of the hybrid power ship, and a dynamic programming method is used to optimize power distribution in a prediction time domain by taking minimization of fuel consumption as a target in combination with the principle of model prediction control. The method has the problem that the steps of a theoretical solving method are complicated.

The Chinese invention CN201711134813.X discloses an algorithm for estimating thrust of an aircraft engine by using a self-adaptive RBF neural network, and the self-adaptive RBF neural network is provided based on a particle swarm algorithm. In the improved particle swarm optimization, the local optimal solution with the same number as the number of the hidden layer nodes is set according to the number of the hidden layer nodes of different networks. The method also has the problem of complicated steps of a theoretical solving method.

Disclosure of Invention

The invention aims to provide a rolling optimization prediction closed-loop controller and a system, which solve the problems of complex solving and poor performance of the dynamic thrust prediction control of a transmitter in the prior art.

In order to achieve the above object, the present invention provides a rolling optimization prediction closed-loop controller, comprising a thrust manager, a signal processor, a controller parameter scheduler, a thrust prediction closed-loop controller and a fuel control quantity selector:

the thrust manager calculates an expected thrust command of the engine operation according to the throttle lever angle signal and the thrust mode signal and outputs the expected thrust command to the thrust prediction closed-loop controller;

the signal processor is used for completing the calculation of the real physical numerical value of the signal after calibration and conversion according to the acquired original sensor signal and outputting the signal to the corresponding module;

the controller coefficient scheduler receives a turbofan engine high-pressure rotor converted rotating speed signal sent by the signal processor, and provides required engine dynamic model information for the thrust prediction closed-loop controller by combining stored offline acquired information;

the thrust prediction closed-loop controller is used for calculating the fuel control quantity required by the thrust prediction closed loop according to the engine dynamic model information provided by the controller coefficient scheduler and the first engine state information provided by the signal processor and sending the calculation result to the fuel control quantity selector;

and the fuel control quantity selector is used for selecting the fuel control quantity and outputting a corresponding fuel instruction.

In one embodiment, the engine dynamics model is as follows:

X＝[π,Δ]^T，u＝[α_I,β]^T；π＝Ps3/PO；

W_f(t)＝α_I·Δ(t)+β(t)；

wherein f is the engine dynamic equation, X is the motion state of the dynamic equation, u is the input quantity of the dynamic equation, pi is the falling-to-inferior ratio of the whole engine, alpha_IIs a linear coefficient, beta is a gain coefficient, Ps3 is the high-pressure compressor outlet static pressure, P0 is the engine working environment temperature, F_dFor engine operating thrust expectation, F_eActual thrust produced for engine operation, W_fIs the fuel control quantity.

In one embodiment, the thrust prediction closed-loop controller includes an initial value calculation module, a discrete optimization matrix calculation module, an initial control law calculation module, a sampling value update calculation module, a jacobian matrix calculation module, a control law change rate calculation module, and a recursive least square calculation module:

the initial value calculation module presets the initial value of the corresponding variable;

the discrete optimization matrix calculation module calculates the optimal matrix of the controller through a C matrix and an H matrix;

the initial control rule calculation module completes approximate solution calculation of the initial control rule in the least square sense based on the recursive least square calculation module according to the input parameters of the initial value calculation module and the H matrix and the C matrix input by the discrete optimal matrix calculation module, and outputs the result to the control rule calculation module;

the control rule calculation module calculates the fuel control quantity of each beat of unit time according to the initial control rule and the change rate of the control rule, outputs the result to the sampling value updating calculation module, and sends the final fuel control quantity result to the fuel control quantity selector;

the sampling value updating calculation module is used for updating and calculating the sampling value of the fuel control quantity in each beat of unit time according to the calculation value of the control rule calculation module and outputting a result to the Jacobian matrix calculation module;

the Jacobian matrix calculation module carries out updating calculation of the Jacobian matrix in each beat of unit time according to the input parameters of the initial value calculation module, and outputs the result to the control rule change rate calculation module;

the control rule change rate calculation module completes approximate solution calculation of the control rule change rate in the least square sense based on the recursive least square calculation module according to the input parameters of the initial value calculation module and outputs the control rule change rate to the control rule calculation module;

and the recursive least square calculation module completes the least square meaning equation solution calculation of the input equation by using a rotation transformation method.

In one embodiment, the system further comprises a steady state closed loop controller and a thrust force planning converter:

the thrust plan converter is connected with the thrust manager, and performs target calculation of steady-state control according to an expected engine running thrust command sent by the thrust plan converter, calculates to obtain a target control rotating speed of a low-pressure shaft rotating speed of the turbofan engine and a target control rotating speed of a high-pressure shaft rotating speed of the turbofan engine, and outputs the target control rotating speeds to the steady-state closed-loop controller;

the controller coefficient scheduler provides required control parameters and an existing rule calculation output function for the steady-state closed-loop controller;

and the steady-state closed-loop controller is connected with the thrust planning converter, calculates the steady-state fuel control quantity of the engine required by the rotating speed closed loop according to the steady-state controller control parameter information provided by the controller coefficient scheduler and the second engine state information provided by the signal processor, and sends the calculation result to the fuel control quantity selector.

In one embodiment, the method for calculating the steady-state fuel control quantity of the engine adopts a proportional-derivative-integral controller, a sliding mode controller and state feedback.

In one embodiment, the system further comprises a transition state controller:

the thrust planning converter carries out target calculation of transition state control according to the expected thrust command of the engine operation, calculates the change rate of the target control rotating speed of the low-pressure shaft rotating speed of the turbofan engine and the change rate of the target control rotating speed of the high-pressure shaft rotating speed of the turbofan engine and outputs the change rates to the transition state controller

The controller coefficient scheduler provides required control parameters and an existing rule calculation output function for the transition state controller;

and the transition state controller is connected with the thrust planning converter, calculates the engine acceleration fuel control quantity and the engine deceleration fuel control quantity required by acceleration control according to the transition state controller control parameter information provided by the controller coefficient scheduler and the third engine state information provided by the signal processor, and sends the calculation result to the fuel control quantity selector.

In one embodiment, the calculation method of the engine acceleration fuel control amount and the engine deceleration fuel control amount adopts a proportional-derivative-integral controller, a sliding mode controller and a feedforward series integration.

In one embodiment, the fuel control system further comprises an engine state scheduler which judges the state of the engine according to the information provided by the signal processor, completes the setting of the working strategy signal of the engine according to the current state and outputs the working strategy signal to the fuel control quantity selector.

In one embodiment, the system further comprises a fuel control quantity selector, and the fuel control quantity selector is used for selecting the fuel control quantities sent by the thrust prediction closed-loop controller, the steady-state closed-loop controller and the transition-state controller according to the working strategy signals from the engine state scheduler and outputting corresponding fuel instructions.

In one embodiment, the operating strategy selection logic of the fuel control quantity selector includes:

when the working strategy is in a thrust prediction mode, the fuel instruction adopts the fuel control quantity sent by the thrust prediction closed-loop controller, and the steady state and the transition state are not distinguished;

and when the working strategy is in a normal rotating speed control mode, the fuel instruction calculation is completed by adopting a high-low selection rule based on the fuel control quantity sent by the transition state controller and the steady state closed loop controller.

In one embodiment, the fuel metering device further comprises a servo actuator closed-loop controller, the servo actuator closed-loop controller is used for converting the fuel command output by the fuel control quantity selector into a position expected signal of the metering mechanism, receiving an actual position feedback signal of the metering mechanism input by the signal processor and completing calculation of the servo control current quantity according to the deviation of the actual position feedback signal and the metering mechanism actual position feedback signal.

In order to achieve the above object, the present invention provides an engine rolling optimization prediction closed-loop control system, which includes the rolling optimization prediction closed-loop controller, an onboard sensor, and an execution mechanism as described above:

the airborne sensor is connected with the rolling optimization prediction closed-loop controller, acquires an engine body signal and sends the engine body signal to the rolling optimization prediction closed-loop controller;

and the executing mechanism is connected with the rolling optimization prediction closed-loop controller and executes corresponding actions according to the control instructions output by the rolling optimization prediction closed-loop controller.

The invention provides a rolling optimization prediction closed-loop controller and an engine rolling optimization prediction closed-loop control system, which adopt a model-based thrust closed-loop prediction control algorithm to predict fuel control quantity and implement closed-loop thrust control regulation, and have the following advantages:

1) the prediction control algorithm can realize closed-loop control of engine dynamic operation thrust prediction, the form of the slave controller is more simplified, and one thrust prediction closed-loop controller can realize the functions of the traditional steady-state controller and the traditional transition-state controller;

2) the development efficiency of the engine thrust control system is improved, a part of test run work can be replaced by a model-based mode, and the development time period of the control system is shortened;

3) the control performance index of the dynamic process of the engine is improved, the finite step iteration solving algorithm based on the initial guess value is adopted in the mode of optimizing the performance index of the dynamic process, and the real-time performance of the fuel control quantity calculation is ensured on the basis of meeting the thrust closed-loop control performance index.

Drawings

The above and other features, properties and advantages of the present invention will become more apparent from the following description of the embodiments with reference to the accompanying drawings in which like reference numerals denote like features throughout the several views, wherein:

FIG. 1 discloses a schematic diagram of a model-based engine roll optimization predictive closed-loop control system according to an embodiment of the invention;

FIG. 2 discloses an internal flow diagram of a rolling optimization predictive closed-loop controller according to an embodiment of the invention;

FIG. 3 discloses an internal schematic diagram of a thrust predictive closed-loop controller according to an embodiment of the invention.

The meanings of the reference symbols in the figures are as follows:

10 rolling optimization prediction closed-loop controller;

11 a thrust manager;

12 a thrust planning converter;

13 a signal processor;

14 a controller parameter scheduler;

15 thrust predictive closed loop controller;

151 initial value calculation module;

152 a discrete optimization matrix calculation module;

153 initial control law calculation module;

154 control law calculation module;

155 sampling value updating calculation module;

156 Jacobian matrix calculation module;

157 a control law change rate calculation module;

158 recursive least squares computation module;

16 a steady state closed loop controller;

17 a transition state controller;

18 an engine state scheduler;

19 a fuel control quantity selector;

110 servo actuator closed loop controller;

20 on-board sensors;

30 an actuator;

40 engine rolling optimization prediction closed-loop control system;

50 an engine body;

60 airplane hard line signals;

70 aircraft bus signals.

Detailed Description

In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

The invention provides an engine rolling optimization prediction closed-loop control system based on a model, which solves the problem of performance overshoot of closed-loop control in the engine control process, particularly in transition state control, and further improves the predictability of engine operation closed-loop thrust control.

Fig. 1 discloses a schematic diagram of a model-based engine rolling optimization prediction closed-loop control system according to an embodiment of the present invention, and as shown in fig. 1, the present invention provides a model-based engine rolling optimization prediction closed-loop control system, which mainly includes a rolling optimization prediction closed-loop controller 10, an on-board sensor 20 of an engine, and an actuator 30.

And the onboard sensor 20 acquires signals of the engine body.

The rolling optimization prediction closed-loop controller 10 collects and receives airplane hard wire signals 60 and airplane bus signals 70.

The rolling optimization prediction closed-loop controller 10 is suitable for online and real-time control.

The rolling optimization prediction closed-loop controller 10 adopts an engine dynamic model which can be obtained through off-line theoretical calculation or experiment.

Aiming at the rolling optimization prediction closed-loop controller 10, the invention provides a thrust closed-loop prediction control algorithm based on a model, and the operation of an engine is ensured to move along an expected thrust curve on the premise of ensuring a certain optimization performance index to be optimal.

Fig. 2 discloses an internal flowchart of a rolling optimization prediction closed-loop controller according to an embodiment of the present invention, and as shown in fig. 2, the rolling optimization prediction closed-loop controller 10 of the present invention mainly includes:

a thrust manager 11, a thrust plan converter 12, a signal processor 13, a controller parameter scheduler 14, a thrust predictive closed-loop controller 15, a steady-state closed-loop controller 16, a transition state controller 17, an engine state scheduler 18, a fuel control quantity selector 19, and a servo actuator closed-loop controller 110.

The thrust manager 11 is connected to a thrust planning converter 12, a signal processor 13, and a thrust prediction closed-loop controller 15.

The thrust manager 11 calculates an expected engine running thrust command Fd according to the throttle lever angle signal PLA and the thrust mode signal ThrustMode sent by the signal processor 13, and outputs the expected engine running thrust command Fd to the thrust plan converter 12 and the thrust predictive closed-loop controller 15. The unit of the throttle lever angle signal PLA is.

The thrust plan converter 12 is connected with the steady-state closed-loop controller 16, performs target calculation of steady-state control according to an expected thrust command Fd for engine operation, calculates a target control rotating speed N1Cmd of the rotating speed of the low-pressure shaft of the turbofan engine and a target control rotating speed N2Cmd of the rotating speed of the high-pressure shaft of the turbofan engine, and outputs the target control rotating speeds to the steady-state closed-loop controller 16;

the thrust planning converter 12 is connected to the transient state controller 17, performs target calculation of the transient state control according to the engine operation thrust expectation command Fd, calculates a change rate N1dotCmd of a target control rotation speed of the low-pressure shaft rotation speed of the turbofan engine and a change rate N2dotCmd of a target control rotation speed of the high-pressure shaft rotation speed of the turbofan engine, and outputs the calculated change rates to the transient state controller 17.

The signal processor 13 is connected to the thrust manager 11, the controller parameter scheduler 14, the thrust prediction closed-loop controller 15, the steady-state closed-loop controller 16, the transition-state controller 17, the engine state scheduler 18, and the servo actuator closed-loop controller 110.

And the signal processor 13 is used for completing the calculation of the real physical numerical value of the calibrated and converted signal according to the acquired original sensor signal, and outputting the real physical numerical value to the other modules for calculation.

The controller coefficient scheduler 14 is connected to the signal processor 13, the thrust predictive closed-loop controller 15, the transition state controller 17, and the steady-state closed-loop controller 16.

The controller coefficient scheduler 14 receives a turbofan engine high-pressure rotor converted rotating speed signal N2r25 sent by the signal processor 13, combines the stored offline acquired information, provides required control parameters and an existing rule calculation output function for the transition state controller 17 and the steady state closed-loop controller 16, and provides required engine dynamic model information for the thrust prediction closed-loop controller 15.

The control parameters provided by the controller coefficient scheduler 14 to the transition state controller 17 include predicted controller parameters Kptr and Kitr.

The control parameters provided by the controller coefficient scheduler 14 to the steady-state closed loop controller 16 include conventional controller parameters Kpstd and Kistd.

The controller coefficient scheduler 14 supplies the thrust predictive closed-loop controller 15 with required engine dynamic model information, which is an expression coefficient of an engine dynamic model function f (X, u).

The engine dynamic model is specifically as follows:

wherein X is [ pi, Delta ]]^T，u＝[α_I,β]^T；π＝Ps3/PO；

W_f(t)＝α_I·Δ(t)+β(t)；

Wherein f is the engine dynamic equation, X is the motion state of the dynamic equation, u is the input quantity of the dynamic equation, pi is the falling-to-inferior ratio of the whole engine, alpha_IIs a linear coefficient, beta is a gain coefficient, and Ps3 is the high pressure compressor outlet static pressure in unitskPa, P0 engine operating ambient temperature, F_dFor engine operating thrust expectation, F_eActual thrust produced for engine operation in kN, W_fThe unit is kg/h for fuel control quantity.

The thrust force prediction closed-loop controller 15 is connected to the thrust force manager 11, the signal processor 13, the controller coefficient scheduler 104, and the fuel control amount selector 19.

The thrust force prediction closed-loop controller 15 calculates the fuel control amount Wf _ MPC required by the thrust force prediction closed-loop according to the engine dynamic model information provided by the controller coefficient scheduler 104 and the first engine state information provided by the signal processor 13, and sends the obtained Wf _ MPC signal to the fuel control amount selector 19. The unit of the fuel control amount Wf _ MPC is kg/h.

The first engine state information provided by the signal processor 13 includes a total inlet temperature T25 of the high-pressure compressor, a total outlet temperature T3 of the high-pressure compressor, a static outlet pressure Ps3 of the high-pressure compressor, an engine working environment temperature P0, a low-pressure shaft rotation speed N1 of the turbofan engine, and a high-pressure shaft rotation speed N2 of the turbofan engine.

The unit of the total inlet temperature T25 of the high-pressure compressor, the total outlet temperature T3 of the high-pressure compressor and the working environment temperature P0 of the engine is K. The high pressure compressor discharge static pressure Ps3 is in kPa. The units of turbofan engine low pressure shaft speed N1 and turbofan engine high pressure shaft speed N2 are rpm.

The steady-state closed-loop controller 16 is connected to the thrust schedule converter 12, the signal processor 13, the controller coefficient scheduler 104, and the fuel control amount selector 19.

The Steady-state closed-loop controller 16 calculates the Steady-state fuel control amount Wf _ Steady of the engine required for the closed-loop rotation speed according to the Steady-state controller control parameter information provided by the controller coefficient scheduler 104 and the second engine state information provided by the signal processor 13, and outputs the obtained Wf _ Steady signal to the fuel control amount selector 19. The unit of the fuel control amount Wf _ Steady is kg/h.

The second engine state information provided by the signal processor 13 includes the turbofan engine low pressure shaft speed N1 and the turbofan engine high pressure shaft speed N2. Wherein the units of N1 and N2 are rpm.

The calculation method of the Steady-state fuel control amount Wf _ Steady of the engine can adopt a traditional proportional-derivative-integral controller, a sliding mode controller, state feedback and the like.

The transition state controller 17 is connected to the thrust planning converter 12, the signal processor 13, the controller coefficient scheduler 104, and the fuel control amount selector 19.

The transient controller 17 calculates an engine acceleration fuel control amount Wf _ Acc and an engine deceleration fuel control amount Wf _ Dec required for acceleration control based on the transient controller control parameter information provided from the controller coefficient scheduler 104 and the third engine state information provided from the signal processor 13, and outputs the obtained Wf _ Acc and Wf _ Dec signals to the fuel control amount selector 19. The unit of the fuel control amounts Wf _ Acc and Wf _ Dec is kg/h.

The third engine state information provided by the signal processor 13 includes a converted rotation speed N2r25 of the high-pressure rotor of the turbofan engine and a change rate N2dot of the rotation speed of the high-pressure shaft of the turbofan engine, where N2r25 is in rpm, and N2dot is in rpm/s.

The calculation method of the engine acceleration fuel control amount Wf _ Acc and the engine deceleration fuel control amount Wf _ Dec can adopt a traditional proportional-derivative-integral controller, a sliding mode controller, a feedforward series integration and the like.

The engine state scheduler 18 is connected to the signal processor 13 and the fuel controller selector 109.

The engine state scheduler 18 judges the state of the engine based on the information provided by the signal processor 13, sets the operation strategy Stratgy signal of the engine based on the current state, and outputs the Stratgy signal to the fuel control quantity selector 19.

The engine is in a normal state, a standby state, a safe state and the like.

The setting of the working strategy Stratgy can be designed based on the state mechanism theory.

The fuel control amount selector 19 is connected to the thrust predictive closed-loop controller 15, the steady-state closed-loop controller 16, the transition-state controller 17, and the engine state scheduler 18.

The fuel control quantity selector 19 selects the fuel control quantity Wf _ MPC sent by the thrust predictive closed-loop controller 15, the fuel control quantity Wf _ Acc and the fuel control quantity Wf _ Dec sent by the transient controller 17, and the fuel control quantity WfSteady sent by the steady-state closed-loop controller 16 according to the working strategy Stratgy from the engine state scheduler 18, and outputs a corresponding fuel command WfCmd, and the specific selection logic is as follows:

a) when the working strategy Stratgy is in a thrust prediction mode, the fuel control quantity Wf _ MPC is adopted by the fuel command Wfcmd, and the steady state and the transition state are not distinguished;

b) and when the working strategy Stratgy is in a normal rotating speed control mode, calculating the fuel command WfCmd by adopting a high-low selection rule based on the fuel control quantities Wf _ Acc, Wf _ Dec and wfSteady, wherein the high-low selection rule is that the lowest value of the three is adopted for acceleration, and the highest value of the three is selected for deceleration.

And the servo actuator closed-loop controller 110 is connected with the signal processor 13 and the fuel controller selector 109.

The servo actuator closed-loop controller 110 converts the fuel command WfCmd output from the fuel control amount selector 19 into a metering mechanism position expected signal lactmd, receives a metering mechanism actual position feedback signal Lactfbk input from the signal processor 13, and calculates the servo control current amount iac according to the deviation between the two.

The thrust prediction closed-loop controller provided by the invention is based on a prediction control method, and adopts an improved optimization algorithm to obtain the fuel control quantity of the thrust closed-loop prediction control of the engine through a dynamic process index optimization idea.

Fig. 3 discloses an internal schematic diagram of a thrust predictive closed-loop controller according to an embodiment of the present invention, and as shown in fig. 3, the thrust predictive closed-loop controller 15 includes an initial value calculation module 151, a discrete optimization matrix calculation module 152, an initial control law calculation module 153, a control law calculation module 154, a sampling value update calculation module 155, a jacobian matrix calculation module 156, a control law change rate calculation module 157, and a recursive least squares calculation module 158.

The initial value calculation module 151 is connected to the discrete optimization matrix calculation module 152 and the initial control law calculation module 153, and performs initial value calculation to calculate initial values of the variables in the thrust prediction closed-loop controller 15.

The variables in thrust predictive closed loop controller 15 include:

a) t is 0, wherein t is time and is in the unit of s;

b) Δ T ═ Ts, where Ts is the sampling time in units of s;

c) x (k) ═ X _ sample (k Δ T), where X _ sample (k Δ T) is a state value sampled at time k Δ T, and k is 0,1, 2.;

d) x0 ═ X (0), where X0 is the initial variable and X (0) is the value at the first position of the state vector X;

e)wherein the content of the first and second substances,is U_tRate of change over time, U_tIs an augmented control quantity matrix;

f) a and b, wherein A is [1, 0; 0,1, b is [ 1; 1 ];

g) 0 is set in the firstflag1, wherein the firstflag1 is a first intermediate variable;

h) and 0 is set in the firstflag2, wherein the firstflag2 is a second intermediate variable.

The discrete optimization matrix calculation module 152 is connected to the initial control law calculation module 153, and performs discrete optimization matrix calculation to complete the calculation of the controller optimal matrix F according to the following formula.

Calculating the C matrix, and converting the inequality into an equation:

wherein the content of the first and second substances,in this embodiment, the symbol ": by definition action, i.e. toDefined as the subsequent formula.

The superscript T denotes transpose.

Wherein X is the motion state of the dynamic equation,u is a dynamic equation input quantity, u 'for a defined augmented control quantity'₁Is composed ofU 'as a third element of'₂Is composed ofU's fourth element of'₃Is composed ofA fifth element of (a)_IIs a linear coefficient, alpha_IminIs the minimum value of the linear coefficient, W_fFor controlling fuel quantity, W_fmaxMaximum fuel control quantity, W_fminIs the minimum value of the fuel control quantity.

Calculating an H matrix:

wherein X is the motion state of the dynamic equation,for defined augmented control quantities, mu and lambda are the parameters of the calculation process, f is the engine dynamic equation, r₁、r₂And r₃To calculate process parameters, F_dFor engine operating thrust expectation, F_eActual thrust, alpha, generated for engine operation_IIs a linear coefficient, beta is a gain coefficient, t is a time, and N is an amplification control quantity matrix U_tOf (c) is calculated.

Calculating a controller optimal matrix F through the C matrix and the H matrix:

wherein the content of the first and second substances,for defined control quantity after augmentation, mu and lambda are parameters of calculation process, X is motion state of dynamic equation, and N is U_tOf (c) is calculated.

The initial control rule calculation module 153 is connected to the control rule calculation module 154 and the recursive least square calculation module 158, performs initial control rule calculation according to the input parameters of the initial value calculation module and the H matrix and the C matrix input by the discrete optimal matrix calculation module, completes approximate solution calculation of the initial control rule in the least square sense based on the recursive least square calculation module, and outputs the result to the control rule calculation module 154.

The calculation is done according to the following formula:

the initial control law is mu₀According to X0 and u inputted from the initial value calculation module 151₀And the H matrix and C matrix input by the discrete optimal matrix calculation module 152, based on the recursive least square calculation module 158, perform the approximate solution mu in the least square sense of the following equation₀And (4) calculating.

H^T(x₀,u₀,0,μ₀)＝0

C^T(x₀,u₀)＝0

Wherein x is₀Taking input X0 as a preset initial value of the state quantity;

u₀is a preset initial value of the control quantity.

And the control rule calculation module 154 is connected with the sampling value update calculation module 155, the control rule change rate calculation module 157 and the fuel control quantity selector 19, performs control rule calculation according to the initial control rule and the control rule change rate, completes the fuel control quantity calculation of the thrust prediction closed-loop controller in each beat of unit time, outputs the result to the sampling value update calculation module 155, and sends the finally completed fuel control quantity result to the fuel control quantity selector 19.

The following pseudo code calculation formula is a control amount calculation method for each beat, and the Firstflag1 is 1 as an initial time, the calculation of the first branch is adopted, and the other times are the calculation of the second branch.

Inputting parameters: initial time value U of amplification control quantity U₀[0]，U_tRate of change over time

Outputting parameters: the fuel control amount Wf (0) at the initial time, and the fuel control amount Wf (τ) at the time τ.

And the sampling value updating calculation module 155 is connected with the Jacobian matrix calculation module 156, performs updating calculation on the sampling value, completes updating calculation on the sampling value of each beat of the thrust prediction closed-loop controller according to the calculation value of the control rule calculation module, and outputs the result to the Jacobian matrix calculation module 156.

According to U_tRate of change over timeThen, by integration of a team time step Δ t, according to the current U_tCalculating the value U of the next sampling time_t+△t。

The following formula is a calculation method of the sampling value of each beat:

X_t＝X(t+△t)；

X_t+△Tx (t +. DELTA.t); (t is the sample value calculated at the current time, t +. DELTA.t is the value calculated at the next sample time)

The jacobian matrix calculation module 156 is connected to the control law change rate calculation module 157, and according to the motion state parameter X of the dynamic equation input from the initial value calculation module 151, the following formula is used to complete the update calculation of the jacobian matrix for each beat of the thrust prediction closed-loop controller, and the result is output to the control law change rate calculation module 157.

The following pseudo code calculation formula is an update calculation method of a Jacobian matrix of each beat, and Firstflag2 is taken as an initial moment, the calculation of a first branch is adopted, and the calculation of a second branch is taken at other moments.

Wherein the content of the first and second substances,is time tζ, b, a are jacobian matrix parameters.

The control law change rate calculation module 157 is connected to the control law calculation module 154 and the recursive least square calculation module 158, performs update calculation of the control law change rate, and outputs the control law change rate to the control law calculation module 154.

Controlling the rate of change of law toTo perform an approximate solution in the least squares sense of the following equation based on the recursive least squares calculation module 158 based on A and b input from the initial value calculation module 151And (4) calculating.

Wherein the content of the first and second substances,is U_tAnd b and A are Jacobian matrix parameters of the change rate of time.

The recursive least square calculation module 158 completes the least square equation solution calculation of the input equation by using a round robin transformation method.

The invention provides a rolling optimization prediction closed-loop controller and an engine rolling optimization prediction closed-loop control system, which adopt a model-based thrust closed-loop prediction control algorithm to predict fuel control quantity and implement closed-loop thrust control regulation, and have the following advantages:

1) the prediction control algorithm can realize closed-loop control of engine dynamic operation thrust prediction, the form of the slave controller is more simplified, and one thrust prediction closed-loop controller can realize the functions of the traditional steady-state controller and the traditional transition-state controller;

2) the development efficiency of the engine thrust control system is improved, a part of test run work can be replaced by a model-based mode, and the development time period of the control system is shortened;

3) the control performance index of the dynamic process of the engine is improved, the finite step iteration solving algorithm based on the initial guess value is adopted in the mode of optimizing the performance index of the dynamic process, and the real-time performance of the fuel control quantity calculation is ensured on the basis of meeting the thrust closed-loop control performance index.

While, for purposes of simplicity of explanation, the methodologies are shown and described as a series of acts, it is to be understood and appreciated that the methodologies are not limited by the order of acts, as some acts may, in accordance with one or more embodiments, occur in different orders and/or concurrently with other acts from that shown and described herein or not shown and described herein, as would be understood by one skilled in the art.

As used in this application and the appended claims, the terms "a," "an," "the," and/or "the" are not intended to be inclusive in the singular, but rather are intended to be inclusive in the plural unless the context clearly dictates otherwise. In general, the terms "comprises" and "comprising" merely indicate that steps and elements are included which are explicitly identified, that the steps and elements do not form an exclusive list, and that a method or apparatus may include other steps or elements.

The embodiments described above are provided to enable persons skilled in the art to make or use the invention and that modifications or variations can be made to the embodiments described above by persons skilled in the art without departing from the inventive concept of the present invention, so that the scope of protection of the present invention is not limited by the embodiments described above but should be accorded the widest scope consistent with the innovative features set forth in the claims.