A hierarchical optimization method for dynamic control laws of aircraft engines

Abstract

The present invention provides a hierarchical optimization method for the dynamic control law of an aero-engine, which belongs to the field of aero-engines. The method comprises the following steps: determining the control parameter values ​​and the controlled parameter values ​​at the starting and ending points of the engine dynamic process; optimizing the dynamic control law of the engine without considering the controller; solving the optimization problem using a differential evolution algorithm, and obtaining an optimization result which is a theoretical optimal curve of the dynamic control law of the engine without considering the response of the controller; utilizing the output theoretical optimal solution of the controller obtained above, combining a control system simulation, selecting a control parameter, and optimizing the controller input instruction of the parameter; obtaining an optimization result which is a dynamic control law curve of the engine of the combined controller; and determining whether the control law curves of all control parameters have been obtained to complete the optimization. The present invention reduces the difficulty of solving the optimization problem, improves the optimization efficiency, and solves the technical difficulty of oscillation in the optimization result when there is only one optimization problem.

Description

Technical Field

[0001] The present invention belongs to the field of aero-engines, and in particular relates to a hierarchical optimization method for dynamic control laws of aero-engines. Background Art

[0002] For aircraft engines, optimizing the design of dynamic control laws is crucial for their ultimate performance. Currently, optimizing dynamic control laws for engines generally involves two steps: first, optimizing the control law based on overall performance. Once the optimal control law is obtained, it is passed to the control system. If the control result output by the controller meets the requirements, the control law is considered acceptable. Otherwise, the control law needs to be readjusted and then tested with the control system until the control system output meets the response requirements of the engine's dynamic process. This approach separates overall performance simulation from control system simulation, failing to consider the control system's response characteristics during the control law optimization process. Consequently, it is impossible to fully realize the engine's performance potential during dynamic processes.

[0003] Because the response characteristics of the control system are taken into account, the optimization results obtained when optimizing dynamic control laws often exhibit oscillations, making them incapable of direct application to engine dynamic process control. Therefore, it is necessary to develop new methods for optimizing dynamic engine control laws that can not only take the response characteristics of the control system into account when optimizing dynamic control laws, but also reduce the difficulty of solving the optimization problem and address the technical difficulties of oscillations in the optimization results. Summary of the Invention

[0004] Technical issues to be solved:

[0005] To overcome the shortcomings of the prior art, the present invention provides a hierarchical optimization method for aircraft engine dynamic control laws. This method decomposes the problem of optimizing the aircraft engine dynamic control laws of a joint controller into two sub-optimization problems: minimizing the error between the actual controller output and the theoretical optimal solution as the first optimization objective, and ensuring smoothness of the controller input commands as the second optimization objective. This multi-objective optimization problem is then solved using a differential evolution algorithm. The resulting controller input commands are the aircraft engine dynamic control law curves for the joint controller. This method allows the response characteristics of the control system to be considered during the optimization design of the engine dynamic control law, addressing the issue of oscillation in the optimization results.

[0006] The technical solution of the present invention is: a hierarchical optimization method for dynamic control laws of an aircraft engine, the specific steps of which are as follows:

[0007] Determine the control parameters and controlled parameters of the engine dynamic process;

[0008] Carry out steady-state performance optimization of aircraft engines to obtain the control parameter values ​​and controlled parameter values ​​at the starting and ending points of the engine dynamic process;

[0009] Construct optimization problems and optimize the dynamic control laws of the engine without considering the controller;

[0010] Using a differential evolution algorithm to solve the engine dynamic control law optimization problem without considering the controller, the optimization result is the theoretical optimal curve of the engine dynamic control law without considering the controller response. If the dynamic control law is optimized by a combined controller, the control parameter output after the controller response must be based on this theoretical optimal curve.

[0011] Using the theoretical optimal solution of the controller output obtained above, combined with the control system simulation, a control parameter is selected and the controller input instruction optimization of the parameter is carried out;

[0012] A differential evolution algorithm is used to solve the controller input instruction optimization problem, and the obtained optimization result is an engine dynamic control law curve of the joint controller;

[0013] Determine whether the control law curves of all control parameters have been obtained. If so, the optimization is completed, and the control law curves of the control parameters obtained are the dynamic control law curves of the aircraft engine of the joint controller; if not, continue to select control parameters and use the differential evolution algorithm to solve the controller input instruction optimization problem. Repeat this cycle until the control law curves of all control parameters are obtained and the optimization is completed.

[0014] A further technical solution of the present invention is: the control parameters refer to parameters that can control the working state of the engine, including the engine's adjustable geometric parameters and fuel flow; the controlled parameters are the engine's performance parameters, including thrust and high-pressure / low-pressure physical speed.

[0015] A further technical solution of the present invention is that the steady-state performance optimization of the aero-engine is carried out using a differential evolution algorithm.

[0016] A further technical solution of the present invention is that the objective function expression of the engine dynamic control law without considering the controller is as follows:

[0017]

[0018] Where J[k] represents the objective function of the kth time step; P[k+1] represents the performance parameter value after the kth time step; P obj Indicates the target value of the performance parameter; P start Indicates the starting value of the performance parameter; X[j] k+1 represents the value of the jth adjustable parameter after the kth time step; X[j]obj represents the target value of the jth adjustable parameter, which is equal to the value of the steady-state adjustable parameter corresponding to the target value of the performance parameter; X[j] up and X[j] low denote the upper and lower limits of the jth adjustable parameter respectively; D denotes the number of adjustable parameters; ω denotes the weight factor; the first term on the right side of the equal sign of this expression indicates that P[k+1] needs to be as close to its target value as possible within each time step, and the second term on the right side of the equal sign indicates that X[j] k+1 As close to its target value as possible; when X[j] k+1 When it approaches its target value, it will also cause P[k+1] to approach its target value.

[0019] A further technical solution of the present invention is that the optimization parameters of the controller input instruction optimization process are the input instructions of each parameter controller, and their starting value and ending value are the steady-state input instructions of the controller.

[0020] A further technical solution of the present invention is: the objective function expression of the controller input instruction optimization problem is as follows:

[0021]

[0022] Where J[k] represents the objective function of the kth time step; P[k+1] represents the actual output value of the controller after the kth time step; P best [k+1] represents the theoretical optimal output value of the controller after the kth time step; P obj The target value of the actual output value of the controller, that is, the terminal value of the control parameter; P start represents the starting value of the control parameter; R[k+1] represents the rate of change of the controller input command after the kth time step, R obj Indicates the target value of the controller input command change rate; ω E The weight factor representing the relative error between the actual output value of the controller and the theoretical optimal output value, ω R Represents the weight factor of the controller input command change rate; the first term on the right side of the equal sign in this expression indicates that the actual output value of the controller in each time step is as close as possible to its theoretical optimal value; the second term on the right side of the equal sign indicates that the controller input command change rate is as close as possible to its target value, in order to ensure the smoothness of the input command; the third term on the right side of the equal sign indicates that the actual output value of the controller is as close as possible to its target value, which is to ensure the rapidity of the dynamic control law.

[0023] A hierarchical optimization system for dynamic control laws of an aero-engine, comprising an engine dynamic process simulation model, an aero-engine steady-state performance optimization model, an engine dynamic control law optimization model, and a controller input instruction optimization model;

[0024] Determine engine dynamic process control parameter values ​​and controlled parameter values ​​based on an engine dynamic process simulation model;

[0025] Through the aircraft engine steady-state performance optimization model, the control parameter values ​​and controlled parameter values ​​of the engine dynamic process starting point and ending point are obtained;

[0026] The engine dynamic control law optimization model is used to solve the engine dynamic control law optimization problem without a controller, and the theoretical optimal curve of the engine dynamic control law without considering the controller response is obtained;

[0027] The controller input instruction optimization problem is solved through the controller input instruction optimization model, and the optimization result obtained is the engine dynamic control law curve of the joint controller.

[0028] An electronic device comprises at least one processor and a memory communicatively connected to the at least one processor; wherein the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the hierarchical optimization method of the dynamic control law of the aircraft engine.

[0029] A computer-readable digital storage medium stores computer instructions, which are used to enable a processor to implement the hierarchical optimization method for dynamic control laws of an aircraft engine when the instructions are executed.

[0030] Beneficial effects

[0031] The beneficial effects of the present invention are as follows: the hierarchical optimization method for the dynamic control laws of aircraft engines of the present invention decomposes the optimization problem of the dynamic control laws of aircraft engines of the joint controller into two sub-optimization problems. After completing the first sub-optimization problem and obtaining the optimal solution of the theoretical output of the controller, indicators such as the rate of change and smoothness of the control law curve are introduced into the objective function of the second sub-optimization problem. By decomposing the joint optimization problem of the two systems and reconstructing the objective function of the reverse optimization of the control system, the overall difficulty of solving the optimization problem is reduced, and the technical difficulty of oscillation in the optimization results can be effectively resolved. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 This is a flow chart of a hierarchical optimization method for dynamic control laws of an aircraft engine, which is optional in an embodiment of the present invention;

[0033] Figure 2The curve of the control law of the combustion chamber fuel supply quantity during the acceleration process of a single-shaft turbojet engine of a joint controller is obtained by directly solving the optimization problem without using the method proposed by the present invention;

[0034] Figure 3 The curve of the fuel supply control law of the combustion chamber during the acceleration process of a single-shaft turbojet engine without considering the controller response is the theoretical optimal solution of the fuel controller output.

[0035] Figure 4 The invention adopts the hierarchical optimization method proposed in the present invention to solve the optimization problem and obtains the control law curve of the combustion chamber command fuel supply quantity of the joint controller during the acceleration process of the single-axis turbojet engine. DETAILED DESCRIPTION

[0036] The embodiments described below with reference to the accompanying drawings are exemplary and are intended to explain the present invention, but should not be construed as limiting the present invention.

[0037] Based on the technical problems such as the difficulty in solving existing optimization problems and the oscillation of optimization results, the present invention provides a hierarchical optimization method for dynamic control laws of aircraft engines. The specific steps are as follows:

[0038] Step 1: Determine the control parameters and controlled parameters of the engine dynamic process;

[0039] Step 2: Optimize the steady-state performance of the aircraft engine to obtain the control parameter values ​​and controlled parameter values ​​at the starting and ending points of the engine dynamic process;

[0040] Step 3: Construct an optimization problem and optimize the dynamic control law of the engine without considering the controller.

[0041] Step 4: Use the differential evolution algorithm to solve the optimization problem in step 3. The optimization result obtained is the theoretical optimal curve of the engine dynamic control law without considering the controller response. If the dynamic control law is optimized by the joint controller, the output result of the control parameters after the controller response must be based on this theoretical optimal curve. Therefore, the optimization result obtained in this step is the theoretical optimal solution of the controller output.

[0042] Step 5: Using the theoretical optimal solution of the controller output in step 4, combined with the control system simulation, select a control parameter and carry out the optimal design of the controller input instruction for the parameter.

[0043] Step 6: Use the differential evolution algorithm to solve the optimization problem in step 5. The optimization result is the engine dynamic control law curve of the joint controller;

[0044] Step 7: Determine whether the control law curves of all control parameters are obtained. If so, go to step 8; otherwise, go to step 5.

[0045] Step 8: The control law curve of the optimized control parameters is the dynamic control law curve of the aircraft engine of the joint controller.

[0046] The present invention provides a hierarchical optimization system for an aero-engine dynamic control law, comprising an engine dynamic process simulation model, an aero-engine steady-state performance optimization model, an engine dynamic control law optimization model, and a controller input instruction optimization model. The system determines engine dynamic process control parameter values ​​and controlled parameter values ​​based on the engine dynamic process simulation model. The system obtains control parameter values ​​and controlled parameter values ​​at the starting and ending points of the engine dynamic process through the aero-engine steady-state performance optimization model. The system solves the engine dynamic control law optimization problem without a controller through the engine dynamic control law optimization model to obtain a theoretical optimal curve of the engine dynamic control law without considering the controller response. The system solves the controller input instruction optimization problem through the controller input instruction optimization model to obtain an optimization result that is an engine dynamic control law curve of a combined controller.

[0047] The present invention provides an electronic device, comprising at least one processor and a memory communicatively connected to the at least one processor; wherein the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the hierarchical optimization method for dynamic control laws of an aircraft engine.

[0048] The present invention provides a computer-readable digital storage medium, wherein the computer-readable digital storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the hierarchical optimization method for dynamic control laws of an aircraft engine when the processor executes the computer instructions.

[0049] The above technical solution is further described below with reference to the accompanying drawings and examples:

[0050] Reference Figure 1 As shown, this embodiment provides a hierarchical optimization method for dynamic control laws of an aircraft engine, and the specific steps are as follows:

[0051] Step 1: Determine the control parameters and controlled parameters of the engine dynamic process. The control parameters refer to the parameters that can control the working state of the engine. Taking a single-spool turbojet engine as an example, assuming that it has no adjustable geometric structure, the control parameter of the single-spool turbojet engine at this time is only the fuel supply to the combustion chamber, and the controlled parameter is thrust.

[0052] Step 2: Carry out steady-state performance optimization of the single-shaft turbojet engine to obtain the optimal fuel flow and thrust values ​​when the engine is in slow and intermediate states.

[0053] Step three: Carry out optimization design of the control law of the acceleration process of the single-axis turbojet engine with a joint fuel controller. The optimization parameters in the acceleration process are the input instructions of the fuel controller, hereinafter referred to as the combustion chamber command fuel supply, and the output instructions of the fuel controller are referred to as the combustion chamber fuel supply. The starting value and the ending value of the combustion chamber command fuel supply are the values ​​of the combustion chamber fuel supply in the slow state and the intermediate state respectively. In order to ensure stable and safe operation of the engine during acceleration, the following constraints are set: the relative physical speed and the relative converted speed of the compressor do not exceed 100%, the compressor surge margin is not less than 15%, the total temperature at the combustion chamber outlet does not exceed 1700K, and the change rate of the combustion chamber command fuel supply does not exceed 10kg / s 2 , The fuel supply rate of the combustion chamber does not exceed 2kg / s 2 The objective function for optimizing the control law of the acceleration process of a single-axis turbojet engine is constructed as follows:

[0054]

[0055] Where J[k] represents the objective function of the kth time step; P[k+1] represents the thrust value after the kth time step; P obj Indicates the target value of thrust; P start represents the starting value of thrust; X[k+1] represents the value of the fuel supply of the combustion chamber after the kth time step; X obj Indicates the target value of the fuel supply to the combustion chamber; X up and X low They represent the upper and lower limits of the j-th adjustable parameter respectively; ω represents the weight factor. When the thrust change does not reach 95% of its total change, ω is taken as 0, and ω is taken as 0.8 at other times.

[0056] Step 4: Directly use the differential evolution algorithm to solve the optimization problem in step 3. The optimization result is the control law curve of the combustion chamber fuel supply quantity of the single-axis turbojet engine acceleration process of the joint controller, as shown in Figure 2 As can be seen from the figure, the combustion chamber command fuel supply curve has very serious oscillations and cannot be directly input to the fuel controller, so the hierarchical optimization method proposed in the present invention must be used.

[0057] Step 5: Optimize the control law design for the acceleration process of a single-shaft turbojet engine without considering the controller response. In this case, the fuel supply to the combustion chamber is the command fuel supply to the combustion chamber. Therefore, it is only necessary to replace the optimization parameter in step 3 with the fuel supply to the combustion chamber and remove the restriction on the rate of change of the command fuel supply to the combustion chamber in the constraint. In this case, X in the formula in step 3 represents the value of the fuel supply to the combustion chamber.

[0058] Step 6: Use the differential evolution algorithm to solve the optimization problem in step 5. The optimization result is the control law curve of the fuel supply to the combustion chamber during the acceleration process of the single-shaft turbojet engine without considering the controller response, that is, the theoretical optimal solution of the controller output, such as Figure 3 shown.

[0059] Step 7: Use the theoretical optimal solution for the fuel controller output obtained in Step 6 to optimize the fuel controller input command. The optimization parameter is the fuel controller input command, i.e., the combustion chamber command fuel supply, with the same starting and ending values ​​as in Step 3. Since this optimization does not require the use of the engine performance calculation model, the only remaining constraint parameter is: the combustion chamber command fuel supply change rate does not exceed 10kg / s 2 The objective function of the optimization problem is as follows:

[0060]

[0061] Where J[k] represents the objective function of the kth time step; P[k+1] represents the value of the combustion chamber command fuel supply after the kth time step; P best [k+1] represents the theoretical optimal output value of the fuel controller after the kth time step; P obj Indicates the target value of the actual output value of the fuel controller, that is, the end value of the fuel supply command of the combustion chamber; P start represents the starting value of the fuel supply quantity of the combustion chamber; R[k+1] represents the value of the rate of change of the fuel supply quantity of the combustion chamber after the kth time step. obj Indicates the target value of the rate of change of the combustion chamber command fuel supply; ω E The weight factor representing the relative error between the actual output value of the fuel controller and the theoretical optimal output value, ω R Indicates the weight factor of the change rate of the combustion chamber command fuel supply. When the thrust change does not reach 95% of its total change, ω E Take it as 0.7, ω R Take it as 0.2, and ω for the rest of the time E and ω R All are taken as 0.

[0062] Step 8: Use the differential evolution algorithm to solve the optimization problem in step 7. The optimization result is the control law curve of the combustion chamber command fuel supply, that is, the control law curve of the combustion chamber fuel supply during the acceleration process of the single-axis turbojet engine combined with the fuel controller, as shown in Figure 4 As shown in the figure, the combustion chamber fuel supply control law curve is very smooth, with only a slight "overshoot phenomenon" at the end of the optimization. Therefore, the hierarchical optimization method proposed in the present invention can effectively solve the problem of oscillation of the optimization results in step 4.

[0063] Although the embodiments of the present invention have been shown and described above, it will be understood that the above embodiments are illustrative and are not to be construed as limitations on the present invention. A person skilled in the art may change, modify, replace and modify the above embodiments within the scope of the present invention without departing from the principles and purpose of the present invention.

Claims

1. A hierarchical optimization method for dynamic control laws of aircraft engines, characterized in that The specific steps are as follows: Determine the control parameters and controlled parameters of the engine dynamic process; Carry out steady-state performance optimization of aircraft engines to obtain the control parameter values ​​and controlled parameter values ​​at the starting and ending points of the engine dynamic process; Construct optimization problems and optimize the dynamic control laws of the engine without considering the controller; Using a differential evolution algorithm to solve the engine dynamic control law optimization problem without considering the controller, the optimization result is the theoretical optimal curve of the engine dynamic control law without considering the controller response. If the dynamic control law is optimized by a combined controller, the control parameter output after the controller response must be based on this theoretical optimal curve. Using the theoretical optimal solution of the controller output obtained above, combined with the control system simulation, a control parameter is selected and the controller input instruction optimization of the parameter is carried out; A differential evolution algorithm is used to solve the controller input instruction optimization problem, and the obtained optimization result is an engine dynamic control law curve of the joint controller; Determine whether the control law curves of all control parameters have been obtained. If so, the optimization is completed, and the control law curves of the control parameters obtained are the dynamic control law curves of the aircraft engine of the joint controller; if not, continue to select control parameters and use the differential evolution algorithm to solve the controller input instruction optimization problem. Repeat this cycle until the control law curves of all control parameters are obtained and the optimization is completed.

2. The hierarchical optimization method for aircraft engine dynamic control laws according to claim 1, characterized in that: The control parameters refer to parameters that can control the working state of the engine, including the engine's adjustable geometric parameters and fuel flow; the controlled parameters are the engine's performance parameters, including thrust and high-pressure / low-pressure physical speed.

3. The hierarchical optimization method for aircraft engine dynamic control laws according to claim 2, characterized in that: The steady-state performance optimization of the aero-engine is carried out using a differential evolution algorithm.

4. The hierarchical optimization method for aircraft engine dynamic control laws according to claim 3, characterized in that: The objective function expression of the engine dynamic control law without considering the controller is as follows: Where J[k] represents the objective function of the kth time step; P[k+1] represents the performance parameter value after the kth time step; P obj Indicates the target value of the performance parameter; P start Indicates the starting value of the performance parameter; X[j] k+1 represents the value of the jth adjustable parameter after the kth time step; X[j] obj represents the target value of the jth adjustable parameter, which is equal to the value of the steady-state adjustable parameter corresponding to the target value of the performance parameter; X[j] up and X[j] low denote the upper and lower limits of the jth adjustable parameter respectively; D denotes the number of adjustable parameters; ω denotes the weight factor; the first term on the right side of the equal sign of this expression indicates that P[k+1] needs to be as close to its target value as possible within each time step, and the second term on the right side of the equal sign indicates that X[j] k+1 Get as close as possible to its target value; When X[j] k+1 When it approaches its target value, it will also cause P[k+1] to approach its target value.

5. The hierarchical optimization method for aircraft engine dynamic control laws according to claim 4, characterized in that: The optimization parameters of the controller input instruction optimization process are the input instructions of each parameter controller, and the starting value and the ending value are the steady-state input instructions of the controller.

6. The hierarchical optimization method for aircraft engine dynamic control laws according to claim 5, characterized in that: The objective function expression of the controller input instruction optimization problem is as follows: Where J[k] represents the objective function of the kth time step; P[k+1] represents the actual output value of the controller after the kth time step; P best [k+1] represents the theoretical optimal output value of the controller after the kth time step; P obj The target value of the actual output value of the controller, that is, the terminal value of the control parameter; P start represents the starting value of the control parameter; R[k+1] represents the rate of change of the controller input command after the kth time step, R obj Indicates the target value of the controller input command change rate; ω E The weight factor representing the relative error between the actual output value of the controller and the theoretical optimal output value, ω R The weight factor representing the rate of change of the controller input command; The first term on the right side of the equal sign in this expression indicates that the actual output value of the controller in each time step is as close as possible to its theoretical optimal value; the second term on the right side of the equal sign indicates that the rate of change of the input command of the controller is as close as possible to its target value, in order to ensure the smoothness of the input command; the third term on the right side of the equal sign indicates that the actual output value of the controller is as close as possible to its target value, which is to ensure the rapidity of the dynamic control law.

7. A hierarchical optimization system for dynamic control laws of an aircraft engine, characterized by: Used to implement the hierarchical optimization method for dynamic control laws of an aircraft engine according to any one of claims 1 to 6; Including engine dynamic process simulation model, aviation engine steady-state performance optimization model, engine dynamic control law optimization model, controller input instruction optimization model; Determine engine dynamic process control parameter values ​​and controlled parameter values ​​based on an engine dynamic process simulation model; Through the aircraft engine steady-state performance optimization model, the control parameter values ​​and controlled parameter values ​​of the engine dynamic process starting point and ending point are obtained; The engine dynamic control law optimization model is used to solve the engine dynamic control law optimization problem without a controller, and the theoretical optimal curve of the engine dynamic control law without considering the controller response is obtained; The controller input instruction optimization problem is solved through the controller input instruction optimization model, and the optimization result obtained is the engine dynamic control law curve of the joint controller.

8. An electronic device, characterized in that: It includes at least one processor and a memory communicatively connected to the at least one processor; wherein the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the hierarchical optimization method for dynamic control laws of an aircraft engine as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the hierarchical optimization method for dynamic control laws of an aircraft engine as described in any one of claims 1 to 6 when executed.

Images