Aero-engine dynamic control law combination optimization method

Abstract

The application discloses an aero-engine dynamic control law combination optimization method, and belongs to the field of aero-engine control law optimization design; the method steps comprise the following steps: determining the control parameters and the controlled parameters of the starting point and the ending point of the engine dynamic process; discretizing the dynamic process of the engine into a plurality of sub-control points according to time steps or other independent variables; constructing an optimization objective function and a constraint of the discrete points; determining the optimization variables of the point-by-point optimization process as the control parameters; solving the discrete optimization problem by carrying out the point-by-point optimization; fitting to obtain the dynamic control law curve of the engine; and judging whether the optimization termination condition is reached. The application solves the problems in the prior art, such as the complex engine dynamic control law design process, the large difficulty in solving the optimization problem when facing multiple adjustable parameters, the numerical fluctuation of the obtained control law and the difficulty in directly applying the control law to engineering.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of aero-engine control law optimization design, and particularly relates to a dynamic control law combination optimization method for aero-engine. BACKGROUND

[0002] The maneuverability of an aircraft depends largely on the dynamic response capability of the engine. For the same engine, different dynamic control laws will make a big difference in its dynamic response capability. Therefore, the optimization design of the engine dynamic control law is of great significance to the dynamic performance of the engine. There are mainly two kinds of current mainstream control law optimization methods: point-by-point optimization technology and global optimization technology. The point-by-point optimization technology simplifies the construction and solution difficulty of the optimization problem by discretizing the dynamic process. However, since the point-by-point optimization technology only focuses on the fastest response of each independent discrete process, it cannot obtain the global optimal solution of the entire dynamic process, and the optimization result obtained by the point-by-point optimization technology may also have numerical fluctuations, which cannot be directly used as a control law.

[0003] For a complex configuration like a variable cycle engine, the number of adjustable parameters reaches more than 10. Assuming that each adjustable parameter is discretized into 10 time-arranged control point sequences, the optimization variables will reach more than 100. For such a high-dimensional nonlinear optimization problem, the conventional gradient-based optimization algorithm will no longer be applicable, and a swarm intelligence optimization algorithm is generally used for solution. Then, multiplied by the number of populations, the dynamic model of the engine needs to be called thousands of times for each generation. Therefore, for a configuration like a variable cycle engine with a large number of adjustable parameters, the optimization time and calculation cost of the global optimization technology are extremely high.

[0004] Therefore, it is necessary to develop a new aero-engine dynamic control law optimization design method to obtain a usable optimal engine dynamic control law under the premise of ensuring certain global convergence and calculation speed. SUMMARY

[0005] The technical problem to be solved is:

[0006] In order to avoid the shortcomings of the prior art, the present application provides a dynamic control law combination optimization method for aero-engine, which combines the global optimization idea on the basis of the traditional point-by-point optimization technology. The method of the present application reduces the construction and solution difficulty of the optimization problem by using the point-by-point optimization technology, and improves the comprehensive performance of the dynamic control law optimization design method by combining the global convergence of the global optimization technology. The method of the present application at least solves the problems in the prior art that the engine dynamic control law design process is complex, the optimization problem is difficult to solve when facing multiple adjustable parameters, and the control law obtained has numerical fluctuations and cannot be directly applied to engineering.

[0007] The technical scheme of the present application is: an aero-engine dynamic control law combination optimization method, and the specific steps are as follows:

[0008] The control parameters and the controlled parameters of the starting point and the ending point of the engine dynamic process are determined; the control parameters include the combustion chamber fuel supply amount and the adjustable geometry parameters, and the controlled parameters are the performance parameters of the engine;

[0009] The dynamic process of the engine is discretized into a plurality of sub-control points according to the time step or the remaining independent variables;

[0010] The optimization objective function and the constraint of the discrete points are constructed;

[0011] The optimization variables of the point-by-point optimization process are determined as the control parameters;

[0012] The point-by-point optimization is solved to solve the discrete optimization problem; the discrete optimization problem is constructed and solved point by point using a series of quadratic programming algorithms to obtain the dynamic control law data of the engine;

[0013] The dynamic control law curve of the engine is fitted; the dynamic control law curve of all the control parameters is fitted using the dynamic control law data of the engine;

[0014] Whether the optimization termination condition is reached is judged.

[0015] The further technical scheme of the present application is: the control parameters refer to the parameters capable of controlling the working state of the engine, wherein the adjustable geometry parameters include the fan inlet guide vane angle, the fan inlet stator blade angle, the compressor inlet guide vane angle, the turbine inlet guide vane area, and the nozzle throat area; the controlled parameters include the thrust and the high / low pressure rotor physical speed.

[0016] The further technical scheme of the present application is: the dynamic process of the engine is discretized according to the time step, and the direct optimization objective is the shortest dynamic response time, and the optimization objective after discretization becomes the fastest response in each time step.

[0017] The further technical scheme of the present application is: the optimization objective function expression of the discrete points is as follows:

[0018]

[0019] In the formula, J[k] represents the objective function of the kth time step; P[k+1] represents the performance parameter value after the end of the kth time step; P obj represents the target value of the performance parameter; P start represents the starting value of the performance parameter; X[j] k+1 represents the value of the jth adjustable parameter after the end of 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 respectively represent the upper limit value and the lower limit value of the jth adjustable parameter; D represents the number of adjustable parameters; ω represents the weight factor; the first term on the right side of the expression represents that P[k+1] needs to be as close as possible to its target value within each time step, and the second term on the right side of the expression represents that X[j] k+1 is as close as possible to its target value within each time step; when X[j] k+1 is close to its target value, P[k+1] is also close to its target value.

[0020] A further technical solution of the present application is that the optimization constraint condition of the discrete points is that the rotational speed is not higher than a given value, the combustion chamber outlet temperature is not higher than a given value, the compression component surge margin is not lower than a given value, and the control parameter change rate is not higher than a given value.

[0021] A further technical solution of the present application is that the process of solving the discrete optimization problem by point-by-point optimization is:

[0022] A series of quadratic programming algorithms are used to sequentially solve the optimization objective function of each discrete point until the control parameters and the controlled parameters both reach a steady state and last for a sufficient time or the simulation time reaches an upper limit of a dynamic process, that is, the optimization result of each discrete point is obtained;

[0023] The optimization result of each discrete point is arranged in time sequence, that is, a discrete dynamic control law data sequence is obtained;

[0024] If the point-by-point optimization count variable i = 1, it is the first time to carry out point-by-point optimization, at this time, the optimization variables are the combustion chamber fuel supply amount and all adjustable geometric parameters, and the point-by-point solving of the optimization problem obtains a discrete dynamic control law data sequence of the combustion chamber fuel supply amount and all adjustable geometric parameters; otherwise, after global optimization, the adjustable geometric parameters have become control variables known in advance, and are no longer optimization variables for point-by-point optimization, at this time, the optimization variables are only the combustion chamber fuel supply amount, and therefore, only a discrete dynamic control law data sequence of the combustion chamber fuel supply amount can be obtained after the point-by-point solving of the optimization problem.

[0025] A further technical solution of the present application is that the process of fitting the dynamic control law curve of the engine is: if the point-by-point optimization count variable i = 1, the discrete dynamic control law data sequence is used to fit the dynamic control law curve of the combustion chamber fuel supply amount and the dynamic control law curve of the adjustable geometric parameters; otherwise, the adjustable geometric parameters have been obtained as control variables, and the dynamic control law curve of the adjustable geometric parameters has been obtained before the point-by-point optimization, and therefore, only the dynamic control law curve of the combustion chamber fuel supply amount needs to be fitted.

[0026] A further technical solution of the present application is that the process of judging whether the optimization termination condition is reached is:

[0027] If the optimization termination condition is reached, the optimization is completed; if not, the next step is performed.

[0028] If it is judged that the optimization termination condition is not reached, the differential evolution algorithm is used to carry out global optimization, the positions of the characteristic points of the adjustable geometry parameter control law are optimized, and the new characteristic points are fitted using linear fitting to obtain a new adjustable geometry parameter control law curve.

[0029] The new adjustable geometry parameter control law curve is transmitted to the engine model, at this time the adjustable geometry parameter is no longer used as an optimization parameter in the next point-by-point optimization, but as a known control parameter, and the point-by-point optimization technique is used to obtain the control law data of the combustion chamber fuel supply, and the cycle is repeated until the termination condition is reached.

[0030] The optimization is completed, and the dynamic control law curve of the combustion chamber fuel supply and the adjustable geometry parameter is obtained.

[0031] An aero-engine dynamic control law combined optimization system, comprising an engine dynamic process simulation model, a discrete model, a discrete point optimization model, a point-by-point optimization model, and a fitting model.

[0032] The control parameter values and the controlled parameter values of the engine dynamic process starting point and the termination point are determined based on the engine dynamic process simulation model.

[0033] The engine dynamic process is discretized into a plurality of control points according to the time step through the discrete model.

[0034] The optimization objective function and the constraint of the discrete points are constructed through the discrete point optimization model.

[0035] The point-by-point optimization model is used to carry out point-by-point optimization to solve the discrete optimization problem.

[0036] The dynamic control law curve of the engine is obtained through the fitting model.

[0037] Advantages

[0038] The beneficial effects of the present application are that the aero-engine dynamic control law combination optimization method of the present application combines the global optimization thought on the basis of the traditional point-by-point optimization technology, uses the differential evolution algorithm to carry out global optimization on the characteristic point position of the adjustable geometry parameter control law curve, and directly obtains the adjustable geometry parameter control law curve through data fitting, which can overcome the problems that the point-by-point optimization design method cannot guarantee to obtain the global optimal control law and the optimization model in the global optimization design method is difficult to solve, and solve the technical difficulties that the optimization results exist numerical fluctuations when the traditional point-by-point optimization design method is used and are difficult to be directly applied to engineering. BRIEF DESCRIPTION OF DRAWINGS

[0039] Figure 1 is an aero-engine dynamic control law combination optimization method flowchart of an embodiment of the present application;

[0040] Figure 2 is a detailed flowchart of the combination of point-by-point technology and global technology in the aero-engine dynamic control law combination optimization method of an embodiment of the present application;

[0041] Figure 3 is a dynamic control law curve of the combustor fuel supply quantity in the reduction process of a mixed-flow turbofan engine obtained by carrying out point-by-point optimization only once by using the method of the present application;

[0042] Figure 4 is a dynamic control law curve of the nozzle throat area in the reduction process of a mixed-flow turbofan engine obtained by carrying out point-by-point optimization only once by using the method of the present application;

[0043] Figure 5 is a curve of the change of thrust with time in the reduction process of a mixed-flow turbofan engine obtained by carrying out point-by-point optimization only once by using the method of the present application;

[0044] Figure 6 is a dynamic control law curve of the combustor fuel supply quantity in the reduction process of a mixed-flow turbofan engine obtained by optimization by using the method of the present application;

[0045] Figure 7 is a dynamic control law curve of the nozzle throat area in the reduction process of a mixed-flow turbofan engine obtained by optimization by using the method of the present application;

[0046] Figure 8 is a curve of the change of thrust with time in the reduction process of a mixed-flow turbofan engine obtained by optimization by using the method of the present application. DETAILED DESCRIPTION

[0047] The embodiments described below with reference to the drawings are exemplary and are intended to explain the present application, and cannot be understood as a limitation of the present application.

[0048] Based on the variable parameter configuration such as the variable cycle engine, there is a problem of high optimization time and high calculation cost of the global optimization technology, and the application provides an aero-engine dynamic control law combination optimization method, and the specific steps are as follows:

[0049] Step 1: determining the control parameters and the controlled parameters of the starting point and the ending point of the engine dynamic process; the control parameters include the combustion chamber oil supply amount and the adjustable geometry parameters, and the controlled parameters are the performance parameters of the engine;

[0050] Step 2: discretizing the dynamic process of the engine into a plurality of sub-control points according to the time step or the remaining independent variables;

[0051] Step 3: constructing the optimization objective function and the constraint of the discrete point;

[0052] Step 4: determining the optimization variable of the point-by-point optimization process as the control parameter;

[0053] Step 5: carrying out point-by-point optimization to solve the discrete optimization problem; constructing the discrete optimization problem and using a series of quadratic programming algorithms to solve point by point to obtain the dynamic control law data of the engine;

[0054] Step 6: fitting to obtain the dynamic control law curve of the engine; using the dynamic control law data of the engine to fit to obtain the dynamic control law curve of all the control parameters;

[0055] Step 7: judging whether the optimization termination condition is reached.

[0056] Specifically, the control parameter refers to the parameter capable of controlling the working state of the engine, wherein the adjustable geometry parameters include the fan inlet guide vane angle, the fan inlet stator blade angle, the compressor inlet guide vane angle, the turbine inlet guide vane area and the nozzle throat area; and the controlled parameter includes the thrust and the high / low pressure rotor physical speed.

[0057] Specifically, the dynamic process of the engine is discretized according to the time step, and the direct optimization objective is the shortest dynamic response time (such as the acceleration time and the starting time), and the optimization objective after discretization becomes the fastest response in each time step.

[0058] Specifically, the optimization objective function expression of the discrete point is as follows:

[0059]

[0060] In the formula, J[k] represents the objective function of the kth time step; P[k+1] represents the performance parameter value after the end of the kth time step; P obj represents the target value of the performance parameter; P start represents the starting value of the performance parameter; X[j] k+1X[j] represents the value of the jth adjustable parameter at the end of the kth time step; X[j] obj X[j] 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 X[j] 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] low X[j] and X[j] represent the upper and lower limit values of the jth adjustable parameter, respectively; D represents the number of adjustable parameters; ω represents a weight factor; the first term on the right side of the expression represents that P[k+1] needs to be as close as possible to its target value within each time step, and the second term on the right side of the expression represents that X[j] k+1 is as close as possible to its target value within each time step; when X[j] k+1 is close to its target value, it also causes P[k+1] to be close to its target value.

[0061] Specifically, the optimization constraint conditions of the discrete points are that the rotational speed is not higher than a given value, the combustion chamber outlet temperature is not higher than a given value, the compression component surge margin is not lower than a given value, and the control parameter change rate is not higher than a given value.

[0062] Specifically, the process of solving the discrete optimization problem by point-by-point optimization is as follows:

[0063] A series of quadratic programming algorithms are used to sequentially solve the optimization objective function of each discrete point until the control parameters and the controlled parameters both reach a steady state and last for a sufficient time or the simulation time reaches the upper limit T Dyn of the dynamic process (i.e., dt×m>T Dyn ), that is, the optimization results of each discrete point are obtained;

[0064] The optimization results of each discrete point are arranged in chronological order, which is the discrete dynamic control law data sequence;

[0065] If the point-by-point optimization count variable i=1, it is the first time to carry out point-by-point optimization, at this time, the optimization variables are the combustion chamber fuel supply amount and all adjustable geometric parameters, and the discrete dynamic control law data sequence of the combustion chamber fuel supply amount and all adjustable geometric parameters is obtained by solving the optimization problem point by point; otherwise, after global optimization, the adjustable geometric parameters have become control variables known in advance and are no longer optimization variables for point-by-point optimization, at this time, the optimization variables are only the combustion chamber fuel supply amount, therefore, only the discrete dynamic control law data sequence of the combustion chamber fuel supply amount can be obtained after solving the optimization problem point by point.

[0066] Specifically, the process of fitting the dynamic control law curve of the engine is as follows: if the point-by-point optimization count variable i = 1, the dynamic control law curve of the combustion chamber fuel supply amount and the dynamic control law curve of the adjustable geometry parameter are fitted by using the discrete dynamic control law data sequence; otherwise, the adjustable geometry parameter has been obtained as the control variable, and the dynamic control law curve of the adjustable geometry parameter has been obtained before the point-by-point optimization, so that only the dynamic control law curve of the combustion chamber fuel supply amount needs to be fitted.

[0067] Specifically, the process of judging whether the optimization termination condition is reached is as follows:

[0068] If the optimization termination condition is reached, the optimization is completed; otherwise, the next step is performed.

[0069] If it is judged that the optimization termination condition is not reached, the differential evolution algorithm is used to carry out global optimization, the position of the adjustable geometry parameter control law characteristic point is optimized, and the linear fitting method is used to fit the new characteristic point to obtain a new adjustable geometry parameter control law curve.

[0070] The new adjustable geometry parameter control law curve is transmitted to the engine model, at this time, the adjustable geometry parameter is no longer used as the optimization parameter in the next point-by-point optimization, but as a known control parameter, and the point-by-point optimization technique is used to obtain the control law data of the combustion chamber fuel supply amount, and the cycle is repeated until the termination condition is reached.

[0071] The optimization is completed, and the dynamic control law curves of the combustion chamber fuel supply amount and the adjustable geometry parameter are obtained.

[0072] The aviation engine dynamic control law combined optimization system provided by the application comprises an engine dynamic process simulation model, a discrete model, a discrete point optimization model, a point-by-point optimization model and a fitting model; the control parameter values and the controlled parameter values of the engine dynamic process starting point and the engine dynamic process ending point are determined based on the engine dynamic process simulation model; the engine dynamic process is discretized into a plurality of control points according to the time step through the discrete model; the optimization objective function and the constraint of the discrete point are constructed through the discrete point optimization model; the point-by-point optimization is carried out to solve the discrete optimization problem through the point-by-point optimization model; and the dynamic control law curve of the engine is fitted through the fitting model.

[0073] The above technical solutions are further described in combination with the drawings and examples as follows:

[0074] Referring to Figure 1 The aviation engine dynamic control law combined optimization method provided by the embodiment is as follows:

[0075] Step 1: determining the control parameter values and the controlled parameter values of the engine dynamic process starting point and the engine dynamic process ending point.

[0076] The control parameter value and the controlled parameter value of the starting point and the ending point of the engine dynamic process need to be determined. The control parameter refers to the parameter that can control the working state of the engine, and the controlled parameter is usually the performance parameter of the engine. In this embodiment, a mixed-flow turbofan engine is taken as an example, and it is assumed that the adjustable geometry parameter is only the nozzle throat area. The control parameter is selected as the combustion chamber fuel supply quantity and the nozzle throat area, and the controlled parameter is the thrust;

[0077] Step two: deceleration process of the discrete mixed-flow turbofan engine.

[0078] The dynamic process of the engine is discretized into m sub-control points according to the time step 0.1 s. It should be noted that the time and parameter value of the starting point are known, and the parameter value of the ending point is known but the time is unknown, so the starting point is not included in the sub-control points, but the ending point is included;

[0079] Step three: constructing the optimization objective function and constraints of the discrete points.

[0080] The direct optimization objective of the dynamic process is to minimize the dynamic response time (such as acceleration time, start-up time). The optimization objective after discretization becomes the fastest response in each time step. In this embodiment, in order to ensure the stable and safe operation of the mixed-flow turbofan engine during the deceleration process, there are some constraints in the optimization problem: the relative physical speed and the relative converted speed of the fan and the compressor are not higher than 102%, the fan surge margin is not less than 10%, the compressor surge margin is not less than 15%, the total temperature at the outlet of the combustion chamber is not higher than 2050K, and the change rate of the combustion chamber fuel supply quantity is not higher than 0.8 kg / s 2 and the like.

[0081] Step four: determining the optimization variables of the point-by-point optimization process.

[0082] The dynamic process of the engine is mainly adjusted by the combustion chamber fuel supply quantity, and the adjustable geometry parameter is mainly used to match the combustion chamber fuel supply quantity, so that the engine can work in the best state. The optimization variables of the first point-by-point optimization of the mixed-flow turbofan engine deceleration process are the combustion chamber fuel supply quantity and the nozzle throat area, and the point-by-point optimization count variable i is assigned a value of 1, which is used to record the number of times of carrying out point-by-point optimization;

[0083] Step five: carrying out point-by-point optimization to solve the discrete optimization problem.

[0084] Referring to Figure 2As shown, the series quadratic programming algorithm is used to solve the optimization objective function of each discrete point in turn until the control parameters and controlled parameters reach steady state and last for a sufficient time or the simulation time reaches the upper limit of the dynamic process (6s). The optimization results of each discrete point are arranged in time sequence to obtain the discrete dynamic control law data sequence. It should be noted that if the point-by-point optimization count variable i = 1, it is the first time to carry out point-by-point optimization, at this time the optimization variables of the mixed-flow turbofan engine deceleration process are the combustor fuel supply and the nozzle throat area, and the discrete dynamic control law data sequence of the combustor fuel supply and the nozzle throat area is obtained by solving the optimization problem point by point; otherwise, after global optimization, the nozzle throat area has become a control variable known in advance and is no longer an optimization variable for point-by-point optimization, at this time the optimization variable is only the combustor fuel supply, so only the discrete dynamic control law data sequence of the combustor fuel supply can be obtained after solving the optimization problem point by point;

[0085] Step six: fitting to obtain the dynamic control law curve of the mixed-flow turbofan engine.

[0086] If the point-by-point optimization count variable i = 1, the dynamic control law curve of the combustor fuel supply (as shown in Figure 3 ) and the dynamic control law curve of the nozzle throat area (as shown in Figure 4 ) are obtained by fitting the discrete dynamic control law data sequence in step five, if the optimization is ended at this time, it is equivalent to carrying out the control law optimization of the mixed-flow turbofan engine deceleration process by using the traditional point-by-point optimization method, at this time the deceleration time calculated according to the 95% thrust change is 3.3s (the thrust change rule over time is as shown in Figure 5 ), and the deceleration process is as shown in Figure 4 It can be seen that the control law curve of the nozzle throat area presents high-dimensional characteristics and is difficult to use as a control law; otherwise if i≠1, the nozzle throat area has been used as a control variable and the dynamic control law curve of the nozzle throat area has been obtained before point-by-point optimization, so only the dynamic control law curve of the combustor fuel supply needs to be fitted. Considering the realizability in engineering, the linear fitting method is used to fit the dynamic control law curve of the mixed-flow turbofan engine deceleration process in this step.

[0087] Step seven: judge whether the optimization termination condition is reached, if yes, go to step ten, otherwise go to step eight.

[0088] Step eight: since the optimization results of the nozzle throat area obtained by the point-by-point optimization technique are difficult to be directly applied in engineering, the differential evolution algorithm is used to carry out global optimization in this step to optimize the characteristic point position of the nozzle throat area control law curve, and the obtained characteristic points are fitted by using linear fitting to obtain a new nozzle throat area control law curve.

[0089] Step nine: the nozzle throat area control law curve obtained in step eight is transmitted to the engine model, so that the nozzle throat area is no longer an optimization parameter in the next point-by-point optimization, but a known control parameter, and the point-by-point optimization count variable i=i+1, go to step five.

[0090] Step ten: the optimization is completed, and the dynamic control law curves of the combustion chamber fuel supply and the nozzle throat area are obtained (as shown in Figures 6-7 It can be seen from the figure that the dynamic control law curve of the combustion chamber fuel supply obtained by applying the method of the present application basically changes linearly, and the dynamic control law curve of the nozzle throat area completely changes linearly, which is conducive to the implementation of the control system. In addition, Figure 8 The figure shows the variation curve of the thrust of the mixed-flow turbofan engine during the deceleration process obtained by applying the method of the present application, and the deceleration time calculated according to the 95% thrust change is 3.2s, which is 0.1s shorter than the deceleration time obtained by the traditional point-by-point optimization technique.

[0091] Although the embodiments of the present application have been shown and described above, it should be understood that the above embodiments are exemplary and should not be construed as limiting the present application, and those skilled in the art can make changes, modifications, replacements and variations to the above embodiments without departing from the principles and purposes of the present application within the scope of the present application.

Claims

1. A method for combined optimization of dynamic control laws for aero-engines, characterized in that... The specific steps are as follows: The control parameters and controlled parameters for determining the start and end points of the engine's dynamic process are defined. The control parameters include the combustion chamber fuel supply and adjustable geometric parameters, while the controlled parameters are the engine's performance parameters. The control parameters refer to parameters that can control the engine's operating state. The adjustable geometric parameters include the fan inlet guide vane angle, fan inlet stator blade angle, compressor inlet guide vane angle, turbine inlet guide vane area, and nozzle throat area. The controlled parameters include thrust and the physical speed of the high-pressure / low-pressure rotor. The dynamic process of the engine is discretized into several sub-control points according to the time step or other independent variables; the dynamic process of the engine is discretized according to the time step, and its direct optimization objective is to minimize the dynamic response time. The optimization objective after discretization becomes to maximize the response within each time step. Construct the objective function and constraints for the discrete points; the expression for the objective function for the discrete points is as follows: In the formula, Indicates the first k The objective function for each time step; Indicates the first k Performance parameter values ​​after each time step; Indicates the target value of the performance parameter; Indicates the initial value of the performance parameter; Indicates the first j The adjustable parameter is at the first k The value after each time step ends; Indicates the first j The target value of an adjustable parameter is equal to the steady-state adjustable parameter value corresponding to the target value of the performance parameter. and They represent the first j The upper and lower limits of each adjustable parameter; D Indicates the number of adjustable parameters; Indicates the weighting factor; The optimization constraints for the discrete points are: the rotational speed is not higher than a given value, the combustion chamber outlet temperature is not higher than a given value, the surge margin of the compression components is not lower than a given value, and the rate of change of the control parameters is not higher than a given value. The optimization variables in the point-by-point optimization process are determined as control parameters; The discrete optimization problem is solved point by point; the discrete optimization problem is constructed and solved point by point using a series of quadratic programming algorithms to obtain the dynamic control law data of the engine. The dynamic control law curve of the engine is obtained by fitting; the dynamic control law curves of all control parameters are obtained by fitting the dynamic control law data of the engine. Determine whether the optimization termination condition has been met; The process of performing point-by-point optimization to solve the discrete optimization problem is as follows: A series of quadratic programming algorithms are used to solve the optimization objective function of each discrete point in turn until the control parameters and the controlled parameters reach a steady state and remain so for a sufficient time or the simulation time reaches the upper limit of the dynamic process, thus obtaining the optimization result of each discrete point. Arranging the optimization results of each discrete point in chronological order yields the discrete dynamic control law data sequence. If the counting variable i=1 in the point-by-point optimization, it is the first point-by-point optimization. At this time, the optimization variables are the combustion chamber fuel supply and all adjustable geometric parameters. Solving the optimization problem point by point yields the discrete dynamic control law data sequence of the combustion chamber fuel supply and all adjustable geometric parameters. Otherwise, after global optimization, the adjustable geometric parameters have become known control variables and are no longer used as optimization variables for point-by-point optimization. At this time, the only optimization variable left is the combustion chamber fuel supply. Therefore, after solving the optimization problem point by point, only the discrete dynamic control law data sequence of the combustion chamber fuel supply can be obtained.

2. The method for combined optimization of dynamic control laws for aero-engines according to claim 1, characterized in that: The process of fitting the dynamic control law curve of the engine is as follows: if the counting variable i=1 is optimized point by point, the dynamic control law curve of the combustion chamber fuel supply and the dynamic control law curve of the adjustable geometric parameters are obtained by fitting the discrete dynamic control law data sequence; otherwise, the adjustable geometric parameters have already been used as control variables, and the dynamic control law curve of the adjustable geometric parameters has been obtained before point by point optimization. Therefore, it is only necessary to fit the dynamic control law curve of the combustion chamber fuel supply.

3. The method for combined optimization of dynamic control laws for aero-engines according to claim 2, characterized in that: The process for determining whether the optimization termination condition has been met is as follows: Determine if the optimization termination condition has been met. If yes, complete the optimization; otherwise, proceed to the next step. If the optimization termination condition is not met, the differential evolution algorithm is used to carry out global optimization, optimize the position of the feature points of the adjustable geometric parameter control law, and use linear fitting to fit the new feature points to obtain a new adjustable geometric parameter control law curve. The new adjustable geometric parameter control law curve is passed to the engine model. At this time, the adjustable geometric parameter is no longer used as the optimization parameter in the next point-by-point optimization, but as the known control parameter. The control law data of the combustion chamber fuel supply is then obtained by using point-by-point optimization technology. This process is repeated until the termination condition is reached. Optimization is complete, and dynamic control curves of combustion chamber fuel supply and adjustable geometric parameters are obtained.

4. A combined optimization system for dynamic control laws of an aero-engine, characterized in that: This method is used to implement the combined optimization method for dynamic control laws of aero-engines as described in any one of claims 1-3; it includes an engine dynamic process simulation model, a discrete model, a discrete point optimization model, a point-by-point optimization model, and a fitting model; The control parameter values ​​and controlled parameter values ​​at the start and end points of the engine dynamic process are determined based on the engine dynamic process simulation model. The engine dynamic process is discretized into several control points according to the time step using the aforementioned discrete model. The objective function and constraints for discrete points are constructed using the discrete point optimization model. The discrete optimization problem is solved by point-by-point optimization using the aforementioned point-by-point optimization model. The dynamic control law curve of the engine is obtained by fitting the fitting model.

Images