Multi-objective optimization control method for high-speed aircraft
Through the NSGA-III multi-objective optimization algorithm and coefficient of variation method, the problem of poor convergence and diversity in multi-objective optimization control of high-speed aircraft is solved, and intelligent optimization control of high-speed aircraft is realized, ensuring the comprehensive optimization of multiple performance indicators is ensured. Effect.
Patent Information
- Application Number
- CN202411483459.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-23
- Publication Date
- 2024-11-22
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The prior art has problems with poor convergence and diversity of multi-objective optimization problems in the multi-objective optimization control of high-speed aircraft, especially when high-dimensional target optimization, parameter selection pressure is high, resulting in unsatisfactory optimization results.
Using NSGA-III multi-objective optimization algorithm and coefficient of variation method, multiple sets of optimization scenarios are designed, performance indicators and performance parameters are calculated based on these indicators and parameters. The index parameter set is set, and the comprehensive performance evaluation index is calculated using the coefficient of variation method to obtain the optimal performance indicators and parameters.
Intelligent optimization control of high-speed aircraft is realized, ensuring comprehensive optimization of tracking error, vibration, angle of attack amplitude and control energy performance, avoiding the difficulty of manual parameter adjustment, and improving control efficiency.
Smart Images

Figure CN119002292A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of intelligent optimization control technology, and in particular to a multi-objective optimization control method for a high-speed aircraft. Background Art
[0002] The optimal control method of aircraft can be divided into indirect method and direct method. The indirect method is widely used in trajectory planning and optimal guidance in the ascending phase. According to the Pontryagin maximum principle, the indirect method transforms the original problem into a two-point boundary value problem and uses the shooting method, finite difference method, etc. to optimize and solve it.
[0003] The indirect method is very sensitive to the initial value of the problem and has difficulty in dealing with complex problems, so the direct method has become the mainstream of trajectory optimization. The direct method discretizes the original problem into a nonlinear programming (NLP) problem and then solves the NLP problem.
[0004] Direct methods can be divided into two categories: direct target method and direct collocation method. The direct target method requires a large number of integral operations, and the solution efficiency is relatively low. Since the pseudo-spectral method has a low computational cost and high convergence, the pseudo-spectral method is not only used in offline trajectory optimization, but also in online trajectory generation.
[0005] Scholars have also conducted a lot of research on solving NLP problems. Among all the solving algorithms, Sequential Quadratic Programming (SQP) algorithm is the most widely used. Although the SQP algorithm has high solving accuracy, it is easy to enter the local optimal solution. Heuristic algorithms and intelligent algorithms are also used in solving NLP problems. Intelligent algorithms including genetic algorithms, particle swarm algorithms, simulated annealing algorithms, etc. have been widely used with the development of computing power.
[0006] There are generally two approaches to the multi-objective optimization problem of high-speed aircraft control: the first is to use the comprehensive performance index as the only optimization target, that is, to design the optimization target weight parameter and transform the multi-objective optimization problem into a single-objective optimization problem; the second method is to use the Pareto dominance relationship in the multi-objective space to obtain a set of mutually non-dominated optimal solutions. The first method requires the weight of the comprehensive performance index to be known in advance. If the weight is not selected reasonably, it is easy to produce a bad optimization result.
[0007] When faced with high-dimensional optimization problems, many methods have a reduced pressure on parameter selection due to the increase in dimensions, and the final optimization effect becomes unsatisfactory. The fast non-dominated sorting genetic algorithm (NSGA-II) is a multi-objective genetic optimization method based on the Pareto optimal solution. Compared with the basic genetic algorithm, the NSGA-II algorithm performs fast non-dominated sorting on individuals before the selection operation, thereby increasing the probability of retaining excellent individuals. However, when dealing with multi-objective optimization problems with 3 or more optimization objectives, the NSGA-II algorithm may have the disadvantages of poor algorithm convergence and diversity due to the use of the crowding idea. Summary of the invention
[0008] In order to overcome the deficiencies of the prior art, an object of the present invention is to provide a high-speed aircraft multi-objective optimization control method to solve the multi-objective optimization problem of high-speed aircraft control.
[0009] To achieve the above object, the present invention provides the following solutions: A multi-objective optimization control method for a high-speed aircraft, comprising: Obtain the dynamic model of the high-speed aircraft and the control instructions of the attitude tracking controller; According to the preset flight mission or control simulation requirements, multiple groups of optimization scenarios are designed and scenario parameters are determined; the scenario parameters include: flight speed, flight altitude and flight angle of attack; For each group of optimization scenarios, optimization simulation is performed in turn based on the NSGA-III multi-objective optimization algorithm to obtain performance indicators and performance parameters under each group of optimization scenarios, and a performance indicator parameter set is formed according to the performance indicators and the performance parameters; the performance indicators include: tracking error performance indicators, control jitter performance indicators, angle of attack amplitude performance indicators and control energy performance indicators; the performance parameters include: attitude tracking controller outer loop control gain and attitude tracking controller inner loop control gain; Based on the coefficient of variation method, the comprehensive performance evaluation index is calculated according to the performance index parameter set to obtain the optimal performance index and optimal performance parameter under each group of optimization scenarios; The high-speed aircraft is controlled using the optimal performance index and the optimal performance parameters.
[0010] Preferably, the formula of the kinetic model is: ; in, is the flight speed, is the pitch angle, is the yaw angle, is the roll angle, is the ballistic inclination angle, is the ballistic deflection angle, is the flight angle of attack, is the sideslip angle, is the velocity roll angle, For thrust, For resistance, For lift, is the lateral force, is the mass of the high-speed aircraft, is the acceleration due to gravity, is the rolling moment of inertia, is the pitch moment of inertia, is the yaw moment of inertia, , and They are the rolling moment, yaw moment and pitch moment acting on the high-speed aircraft, The high-speed aircraft along the body coordinate system x The angular velocity component of the shaft, The high-speed aircraft along the body coordinate system y The angular velocity component of the shaft, The high-speed aircraft along the body coordinate system z The angular velocity component of the shaft, t For time.
[0011] Preferably, the calculation formula of the control instruction is: ;in, is the expected value of the attitude angle, is the expected value of attitude angular rate, is the first-order derivative of the expected value of the attitude angle, is the first-order derivative of the expected value of the attitude angular rate, is the posture tracking error, is the uncertainty disturbance observation value of the outer loop of the high-speed aircraft attitude by the robust and uniformly convergent observer, is the nonlinear coefficient matrix of the outer loop of the high-speed aircraft attitude, , is the virtual control instruction of the attitude angle, is the attitude angular velocity tracking error, is the uncertainty disturbance observation value of the inner loop of the high-speed aircraft attitude by the robust and uniformly convergent observer, is the nonlinear parameter matrix of the inner loop of the high-speed aircraft attitude, , is the rolling moment of inertia, is the pitch moment of inertia, is the yaw moment of inertia, is the nonlinear coefficient matrix of the inner loop of the high-speed aircraft attitude, , and are the outer loop control gain of the attitude tracking controller and the inner loop control gain of the attitude tracking controller, respectively, u is the control input, , is the dynamic pressure, is the reference area, is the characteristic length, and are the dimensionless coefficients of the rolling moment with respect to the rolling rudder and the rolling angular velocity, respectively. , and are the dimensionless coefficients of the yaw moment with respect to the yaw rudder, yaw angular velocity, and sideslip angle, respectively. , and are the dimensionless coefficients of the pitch moment with respect to the angle of attack, elevator, and pitch angular velocity, respectively, is the elevator deflection angle, is the yaw rudder deflection angle, is the rolling rudder deflection angle.
[0012] Preferably, the mathematical expression of the tracking error performance indicator is: ;in, is the tracking error performance indicator, is the tracking error number, , is the tracking error, weight Satisfaction relationship ; The mathematical expression of the control jitter performance index is: ;in, is the control chattering performance index; is the roll rudder deflection command at the current moment, is the roll rudder deflection command at the last moment, is the yaw rudder deflection command at the current moment, The yaw rudder deflection command at the last moment, is the elevator deflection command at the current moment, is the elevator deflection instruction at the previous moment; the mathematical expression of the angle of attack amplitude performance index is: ;in, is the attack angle amplitude performance index, is the flight angle of attack, satisfying the conditions: , are the minimum and maximum values of the flight angle of attack amplitude, is the flight angle of attack constraint boundary value of the high-speed aircraft, , non-singular quantity ; The mathematical expression of the control energy performance index is: ;in, are the weights of each item, all greater than , and satisfies , is the control energy performance indicator.
[0013] Preferably, optimization simulation is performed in sequence based on the NSGA-III multi-objective optimization algorithm to obtain performance indicators and performance parameters under each group of optimization scenarios, and a performance indicator parameter set is formed according to the performance indicators and the performance parameters, including: Set the initial parameters of the NSGA-III algorithm and the initial state values of the attitude tracking controller used in the current state of the high-speed aircraft, and define the initial reference point; Set evolution generation , randomly generate the initial population , the initial population The corresponding attitude tracking controller of the high-speed aircraft is introduced to calculate the values of the tracking error performance index, the control buffeting performance index, the angle of attack amplitude performance index and the control energy performance index. A fast Pareto non-dominated sorting is performed according to the obtained values of each performance index, and random selection, simulated binary crossover and polynomial mutation are performed to obtain the offspring population. ; The initial population With offspring population Merge into new population ; For new populations , perform non-dominated sorting according to the Pareto dominance relationship, obtain the non-dominated layer, and use the non-dominated layer to construct a new population in turn , until the new population The size of the population equals or exceeds the initial population for the first time ; Based on the reference point idea, the values of various performance indicators calculated by the attitude tracking controller of the high-speed aircraft are used to generate the new population. Selection and Initial Population Individuals of equal size generate a new parent population ; judge Is it equal to the maximum evolutionary generation? If not, then the evolutionary generation And return to step "Set the initial population With offspring population Merge into new population "; If so, the performance indicator parameter set is determined according to the calculated performance indicator value.
[0014] Preferably, based on the coefficient of variation method, the comprehensive performance evaluation index is calculated according to the performance index parameter set to obtain the optimal performance index and optimal performance parameter in each group of optimization scenarios, including: The indicators in the performance indicator parameter set are subjected to positive processing to obtain each positively processed indicator; the formula for the positive processing is: ;in, represents the processed performance indicator sample, represents the performance indicator sample before processing, Indicates the performance indicator number, , represents the sequence number of the Pareto optimal solution set, is a normal number, Indicates the number of samples among all The maximum absolute value of the negative indicator; Each index after positive processing is standardized to obtain the standardized index; the formula for standardization is: , represents the standardized performance index sample after the standardization process, Indicates the number of Pareto optimal solution sets; The weights of each standardized indicator are calculated according to the coefficient of variation; the weight calculation formula is: ;in Indicates The weight of each performance indicator, Indicates The coefficient of variation of item weights, ; The comprehensive performance evaluation index is determined according to each standardized index and the corresponding weight; the calculation formula of the comprehensive performance evaluation index is: , represents the weight of the tracking error performance indicator, represents the weight of the control chattering performance index, represents the weight of the attack angle amplitude performance index, represents the weight of the control energy performance index; The optimal performance index and optimal performance parameters under each group of optimization scenarios are determined according to the comprehensive performance evaluation index.
[0015] The present invention discloses the following technical effects: The present invention provides a multi-objective optimization control method for a high-speed aircraft, comprising: obtaining a dynamic model of the high-speed aircraft and a control instruction of an attitude tracking controller; designing multiple groups of optimization scenarios according to preset flight missions or control simulation requirements, and determining scenario parameters; the scenario parameters include: flight speed, flight altitude and flight angle of attack; for each group of the optimization scenarios, optimizing simulation is performed in sequence based on the NSGA-III multi-objective optimization algorithm to obtain performance indicators and performance parameters under each group of optimization scenarios, and a performance indicator parameter set is formed according to the performance indicators and the performance parameters; the performance indicators include: tracking error performance indicators, control buffeting performance indicators, angle of attack amplitude performance indicators and control energy performance indicators; the performance parameters include: attitude tracking controller outer loop control gain and attitude tracking controller inner loop control gain; based on the coefficient of variation method, a comprehensive performance evaluation index is calculated according to the performance indicator parameter set to obtain the optimal performance indicators and optimal performance parameters under each group of optimization scenarios; and the high-speed aircraft is controlled using the optimal performance indicators and the optimal performance parameters. The present invention can realize intelligent optimization control of high-speed aircraft. BRIEF DESCRIPTION OF THE DRAWINGS
[0016] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.
[0017] Figure 1 A flow chart of a method provided by an embodiment of the present invention; Figure 2 An algorithm flow chart provided for an embodiment of the present invention; Figure 3 Flowchart of the NSGA-III optimization algorithm provided by an embodiment of the present invention. DETAILED DESCRIPTION
[0018] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0019] The purpose of the present invention is to provide a multi-objective optimization control method for a high-speed aircraft, which can realize intelligent optimization control of the high-speed aircraft.
[0020] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.
[0021] Figure 1 A flow chart of a method provided by an embodiment of the present invention, such as Figure 1 As shown, the present invention provides a multi-objective optimization control method for a high-speed aircraft, comprising: Step 100: Obtaining a dynamic model of the high-speed aircraft and control instructions of an attitude tracking controller; Step 200: design multiple sets of optimization scenarios according to preset flight missions or control simulation requirements, and determine scenario parameters; the scenario parameters include: flight speed, flight altitude and flight attack angle; Step 300: For each group of optimization scenarios, optimization simulation is performed in turn based on the NSGA-III multi-objective optimization algorithm to obtain performance indicators and performance parameters under each group of optimization scenarios, and a performance indicator parameter set is formed according to the performance indicators and the performance parameters; the performance indicators include: tracking error performance indicators, control jitter performance indicators, angle of attack amplitude performance indicators and control energy performance indicators; the performance parameters include: attitude tracking controller outer loop control gain and attitude tracking controller inner loop control gain; exemplary, performance parameters are parameters in the controller. In general control, control parameters need to be manually adjusted by designers. In this embodiment, the optimization algorithm is used to adjust the parameters.
[0022] Step 400: Based on the coefficient of variation method, a comprehensive performance evaluation index is calculated according to the performance index parameter set to obtain the optimal performance index and optimal performance parameter in each group of optimization scenarios; Step 500: Controlling the high-speed aircraft using the optimal performance index and the optimal performance parameters.
[0023] Preferably, the formula of the kinetic model is: ; in, is the flight speed, is the pitch angle, is the yaw angle, is the roll angle, is the ballistic inclination angle, is the ballistic deflection angle, is the flight angle of attack, is the sideslip angle, is the velocity roll angle, For thrust, For resistance, For lift, is the lateral force, is the mass of the high-speed aircraft, is the acceleration due to gravity, is the rolling moment of inertia, is the pitch moment of inertia, is the yaw moment of inertia, , and They are the rolling moment, yaw moment and pitch moment acting on the high-speed aircraft, The high-speed aircraft along the body coordinate system x The angular velocity component of the shaft, The high-speed aircraft along the body coordinate system y The angular velocity component of the shaft, The high-speed aircraft along the body coordinate system z The angular velocity component of the shaft, t For time.
[0024] Preferably, the calculation formula of the control instruction is: ;in, is the expected value of the attitude angle, is the expected value of attitude angular rate, is the first-order derivative of the expected value of the attitude angle, is the first-order derivative of the expected value of the attitude angular rate, is the posture tracking error, is the uncertainty disturbance observation value of the outer loop of the high-speed aircraft attitude by the robust and uniformly convergent observer, is the nonlinear coefficient matrix of the outer loop of the high-speed aircraft attitude, , is the virtual control instruction of the attitude angle, is the attitude angular velocity tracking error, is the uncertainty disturbance observation value of the inner loop of the high-speed aircraft attitude by the robust and uniformly convergent observer, is the nonlinear parameter matrix of the inner loop of the high-speed aircraft attitude, , is the rolling moment of inertia, is the pitch moment of inertia, is the yaw moment of inertia, is the nonlinear coefficient matrix of the inner loop of the high-speed aircraft attitude, , and are the outer loop control gain of the attitude tracking controller and the inner loop control gain of the attitude tracking controller, respectively, u is the control input, , is the dynamic pressure, is the reference area, is the characteristic length, and are the dimensionless coefficients of the rolling moment with respect to the rolling rudder and the rolling angular velocity, respectively. , and are the dimensionless coefficients of the yaw moment with respect to the yaw rudder, yaw angular velocity, and sideslip angle, respectively. , and are the dimensionless coefficients of the pitch moment with respect to the angle of attack, elevator, and pitch angular velocity, respectively, is the elevator deflection angle, is the yaw rudder deflection angle, is the rolling rudder deflection angle.
[0025] Preferably, the mathematical expression of the tracking error performance indicator is: ;in, is the tracking error performance indicator, is the tracking error number, , is the tracking error, weight Satisfaction relationship ; The mathematical expression of the control jitter performance index is: ;in, is the control chattering performance index; is the roll rudder deflection command at the current moment, is the roll rudder deflection command at the last moment, is the yaw rudder deflection command at the current moment, The yaw rudder deflection command at the last moment, is the elevator deflection command at the current moment, is the elevator deflection instruction at the previous moment; the mathematical expression of the angle of attack amplitude performance index is: ;in, is the attack angle amplitude performance index, is the flight angle of attack, satisfying the conditions: , are the minimum and maximum values of the flight angle of attack amplitude, is the flight angle of attack constraint boundary value of the high-speed aircraft, , non-singular quantity ; The mathematical expression of the control energy performance index is: ;in, are the weights of each item, all greater than , and satisfies , is the control energy performance indicator.
[0026] Specifically, Figure 2As shown, this embodiment first selects optimization parameters and optimization indicators, and designs a high-speed aircraft optimization control algorithm model.
[0027] According to the trajectory control requirements of high-speed aircraft, the following performance indicators are selected as the optimization objective function: (1) During the flight of a high-speed aircraft, the first thing to do is to ensure stable tracking of the desired command, that is, the cumulative value of the aircraft's tracking error of the desired command is minimized, as shown in the following formula: , In the formula, is the tracking error number, , is the tracking error, weight Satisfaction relationship , the total tracking error .
[0028] (2) The controller has the smallest accumulated chattering, that is, In the formula, is the roll rudder deflection command at the current moment, is the roll rudder deflection command at the last moment, is the yaw rudder deflection command at the current moment, The yaw rudder deflection command at the last moment, is the elevator deflection command at the current moment, The elevator deflection command and the vibration command deviation value at the previous moment .
[0029] (3) When a high-speed aircraft is in flight, when the angle of attack is too large, it is easy to cause aeroelastic oscillation problems and instability. Therefore, the actual flight angle of attack of a high-speed aircraft cannot be too large. At the same time, in the cruise phase, the scramjet engine works to provide thrust for the high-speed aircraft to maintain the aircraft's uniform flight speed. However, due to the intake constraints of the scramjet engine, the actual flight angle of attack of a high-speed aircraft cannot be too large, otherwise it is easy to cause the engine to stall.
[0030] In order to achieve safe flight of high-speed aircraft, the flight angle of attack Limiting is performed to ensure that the angle of attack of the aircraft is always maintained within a safe range during flight. In the formula, are the minimum and maximum values of the angle of attack limit respectively.
[0031] Design the performance index of the high-speed aircraft flight angle of attack as shown in the following formula In the formula, the non-singular quantity , is the normal flight attack angle constraint of the aircraft, Total Amplitude of Angle of Attack .
[0032] As shown in the formula of the high-speed aircraft flight angle of attack performance index, when the angle of attack From limit value The farther away, The larger the value of At the same time, due to the positive number The existence of can avoid the occurrence of singularities when the angle of attack reaches the limiting value.
[0033] When the flight angle of attack exceeds the angle of attack limit range, if the flight angle of attack is too large, the high-speed aircraft's air-breathing engine will not work properly, causing the aircraft to lose control. At this time, the performance index There is no valid value. Therefore, this embodiment will use the program to run the error judgment method. When the aircraft loses control due to excessive flight angle of attack, the performance index will be assigned a larger value, such as .
[0034] (4) The controller minimizes the control energy, that is, In the formula, are the weights of each item, all greater than , and satisfies , control the total amount of energy .
[0035] Then, this embodiment selects the high-speed aircraft flight condition to perform multi-objective optimization.
[0036] According to the control instructions, consider the influence of the control parameters of the high-speed aircraft attitude control instructions on the performance indicators and select the controller parameters. and as optimization parameters.
[0037] The flight speed, flight altitude, pitch angle, roll angle, and yaw angle of the high-speed aircraft are selected as the multi-objective optimization flight conditions of the NSGA-III high-speed aircraft, and the flight conditions are set by random acquisition. flight conditions, and . Select the optimization scenario according to the following constraints: This embodiment then obtains the optimal performance index parameters under each working condition based on the NSGA-III multi-objective optimization algorithm.
[0038] Considering the four selected performance indicators Conflicts and contradictions, so the optimization control of high-speed aircraft is a typical multi-objective optimization problem (Multi-objective Optimization Problem, MOP), that is, there is no solution that makes each performance indicator reach the optimal level, and the improvement of any sub-goal may cause the performance of other sub-goals to decrease. Therefore, the optimal solution that can be selected under any flight condition is not unique, but a set of equilibrium solutions, that is, the optimal non-inferior solution set or the Pareto optimal solution set. In the end, the most ideal solution needs to be obtained through manual screening. This solution can only be coordinated and compromised between each sub-goal to ensure that the four indicators can obtain the ideal value range as much as possible.
[0039] Preferably, optimization simulation is performed in sequence based on the NSGA-III multi-objective optimization algorithm to obtain performance indicators and performance parameters under each group of optimization scenarios, and a performance indicator parameter set is formed according to the performance indicators and the performance parameters, including: Set the initial parameters of the NSGA-III algorithm and the initial state values of the attitude tracking controller used in the current state of the high-speed aircraft, and define the initial reference point; Set evolution generation , randomly generate the initial population , the initial population The corresponding attitude tracking controller of the high-speed aircraft is introduced to calculate the values of the tracking error performance index, the control buffeting performance index, the angle of attack amplitude performance index and the control energy performance index. A fast Pareto non-dominated sorting is performed according to the obtained values of each performance index, and random selection, simulated binary crossover and polynomial mutation are performed to obtain the offspring population. ; The initial population With offspring population Merge into new population ; For new populations , perform non-dominated sorting according to the Pareto dominance relationship, obtain the non-dominated layer, and use the non-dominated layer to construct a new population in turn , until the new population The size of the population equals or exceeds the initial population for the first time ; Based on the reference point idea, the values of various performance indicators calculated by the attitude tracking controller of the high-speed aircraft are used to generate the new population. Selection and Initial Population Individuals of equal size generate a new parent population ; judge Is it equal to the maximum evolutionary generation? If not, then the evolutionary generation And return to step "Set the initial population With offspring population Merge into new population "; If so, the performance indicator parameter set is determined according to the calculated performance indicator value.
[0040] like Figure 3 As shown, this embodiment adopts the iterative process of the NSGA-III optimization algorithm. The multi-objective optimization steps for designing high-speed aircraft control based on the NSGA-III algorithm can be summarized as follows: (1) Initialization: Set the initial parameters of the NSGA-III algorithm and the initial state values of the tracking controller used in the current state of the high-speed aircraft, and define the initial reference point.
[0041] (2) Setting the evolutionary generation , randomly generate the initial population According to the flight phase and control scheme, the initial population is introduced into the corresponding high-speed aircraft controller, and the four performance indicators of the aircraft system are calculated. According to the obtained performance indicator values, a fast Pareto non-dominated sorting is performed, and then random selection, SBX crossover and polynomial mutation are performed to obtain the offspring population. .
[0042] (3) The parent population With offspring population Merge into new population .
[0043] (4) For populations , perform non-dominated sorting based on Pareto dominance, obtain the non-dominated layer, and use the non-dominated layer to construct new populations in sequence , until the population The size of the population equals or exceeds the original population for the first time .
[0044] (5) Based on the reference point concept, the four performance index values calculated by the high-speed aircraft tracking controller are obtained from the population Select suitable individuals to generate a new parent population .
[0045] (6) Judgment Is it equal to the maximum evolutionary algebra? If not, then the evolutionary algebra And return (3); otherwise, the algorithm ends.
[0046] In order to improve the operating efficiency of the algorithm of this embodiment and avoid unreasonable situations such as the hypersonic aircraft system being out of control and the indicator value being null or infinite due to unreasonable population selection, an operation error judgment function is deliberately added to the algorithm, that is, when the state quantity of the aircraft is null or infinite, or there is no constraint angle of attack tracking error When it is a plural number, it is determined that the system operation is erroneous, the control program stops iterating and assigns a larger value to the performance indicator value.
[0047] Preferably, based on the coefficient of variation method, the comprehensive performance evaluation index is calculated according to the performance index parameter set to obtain the optimal performance index and optimal performance parameter in each group of optimization scenarios, including: The indicators in the performance indicator parameter set are subjected to positive processing to obtain each positively processed indicator; the formula for the positive processing is: ;in, represents the processed performance indicator sample, represents the performance indicator sample before processing, Indicates the performance indicator number, , represents the sequence number of the Pareto optimal solution set, is a normal number, Indicates the number of samples among all The maximum absolute value of the negative indicator; Each index after positive processing is standardized to obtain the standardized index; the formula for standardization is: , represents the standardized performance index sample after the standardization process, Indicates the number of Pareto optimal solution sets; The weights of each standardized indicator are calculated according to the coefficient of variation; the weight calculation formula is: ;in Indicates The weight of each performance indicator, Indicates The coefficient of variation of item weights, ; The comprehensive performance evaluation index is determined according to each standardized index and the corresponding weight; the calculation formula of the comprehensive performance evaluation index is: , represents the weight of the tracking error performance indicator, represents the weight of the control chattering performance index, represents the weight of the attack angle amplitude performance index, represents the weight of the control energy performance index; The optimal performance index and optimal performance parameters under each group of optimization scenarios are determined according to the comprehensive performance evaluation index.
[0048] Specifically, this embodiment selects the optimal performance indicator based on the coefficient of variation method.
[0049] Although the NSGA-III optimization algorithm can obtain performance index parameters, there are too many parameters and screening is still required. The subjective judgment method requires a lot of effort and is less efficient. This embodiment will use the coefficient of variation method to obtain the comprehensive performance index weight online. , and then obtain the optimal performance index.
[0050] The coefficient of variation method is an objective weighting method. Compared with the subjective weighting method, the coefficient of variation method calculates the degree of change of each indicator of the system based on statistical methods, directly uses the information contained in each indicator, and obtains the weight of the indicator through calculation. Importantly, the coefficient of variation method can reflect the difference and fluctuation of the data by depicting the deviation trend of each indicator. In terms of numerical value, it is equal to the standard deviation divided by the mean.
[0051] Assume that the data set obtained through ground simulation has a total of groups, each group includes 4 indicators. In this case, the performance indicator data set can be expressed as Since the four groups of performance indicators are set in this embodiment, , , The three items are negative indicators, that is, the smaller the value, the better the performance, so they must be positively converted before calculating the weight: In the formula, , Indicates the number of samples among all The maximum absolute value of the negative indicator.
[0052] At the same time, considering that the dimensions of the four groups of performance indicators set in this embodiment are quite different, especially the controller energy One of the indicators is much larger than the other three indicators, so the weight cannot be set directly. It is necessary to pre-process and unify the dimensions of each item, that is, At this point, the standardized sample data becomes Therefore, the weight of each indicator is In the formula, is the coefficient of variation of each indicator, The coefficient of variation method assigns weights to each indicator according to the degree of difference between the current value and the target value of the performance indicator in different scenarios. That is, if the numerical variation of an indicator is large and can clearly distinguish the evaluated objects, it means that the evaluation information of the indicator is rich, and thus the indicator should be given a larger weight; conversely, if the change of an indicator in a scenario is small, then the distinguishing ability of this indicator is weak, and thus the indicator should be given a smaller weight.
[0053] Based on the coefficient of variation method, the weight of the aircraft performance index under the current flight conditions can be calculated. , and then a comprehensive performance evaluation index can be determined: , By comparing the comprehensive performance evaluation indicators , a set of optimal performance parameters can be selected.
[0054] Finally, return to "Select high-speed aircraft flight conditions and perform multi-objective optimization" until the optimal performance parameters under all flight conditions are obtained.
[0055] Furthermore, the NSGA_III multi-objective optimization algorithm obtains several sets of solutions, all of which are Pareto optimal. However, since the weights of the performance indicators are unknown, it is impossible to know which set has the best comprehensive performance. Then, the coefficient of variation method is used. The coefficient of variation method can calculate the weight of each performance indicator, and the weight is substituted into the Pareto solution obtained by the NSGA_III multi-objective optimization algorithm to obtain the comprehensive performance indicator of each set of solutions. Finally, by comparing the size of the comprehensive performance indicators, the optimal solution can be obtained.
[0056] The beneficial effects of the present invention are as follows: (1) Compared with the general optimization control method, the multi-objective optimization control method designed by the present invention comprehensively considers the influence of tracking error, buffeting, angle of attack amplitude, and control energy on the control of high-speed aircraft, ensuring the optimal comprehensive control performance of high-speed aircraft; (2) The present invention can also be used in gain scheduling control, and the automatic construction of the gain scheduling table can be realized through programming without human intervention.
[0057] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.
[0058] The principles and implementation methods of the present invention are described in this article using specific examples. The description of the above embodiments is only used to help understand the method and core idea of the present invention. At the same time, for those skilled in the art, according to the idea of the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.
Claims
1. A multi-objective optimization control method for a high-speed aircraft, characterized in that: include: Obtain the dynamic model of the high-speed aircraft and the control instructions of the attitude tracking controller; Design multiple sets of optimization scenarios and determine scenario parameters according to preset flight missions or control simulation requirements; The scenario parameters include: flight speed, flight altitude and flight attack angle; For each group of optimization scenarios, optimization simulation is performed in turn based on the NSGA-III multi-objective optimization algorithm to obtain performance indicators and performance parameters under each group of optimization scenarios, and a performance indicator parameter set is formed according to the performance indicators and the performance parameters; the performance indicators include: tracking error performance indicators, control jitter performance indicators, angle of attack amplitude performance indicators and control energy performance indicators; the performance parameters include: attitude tracking controller outer loop control gain and attitude tracking controller inner loop control gain; Based on the coefficient of variation method, the comprehensive performance evaluation index is calculated according to the performance index parameter set to obtain the optimal performance index and optimal performance parameter under each group of optimization scenarios; The high-speed aircraft is controlled using the optimal performance index and the optimal performance parameter.
2. The multi-objective optimization control method for a high-speed aircraft according to claim 1, characterized in that: The formula of the kinetic model is: ; in, is the flight speed, is the pitch angle, is the yaw angle, is the roll angle, is the ballistic inclination angle, is the ballistic deflection angle, is the flight angle of attack, is the sideslip angle, is the velocity roll angle, For thrust, For resistance, For lift, is the lateral force, is the mass of the high-speed aircraft, is the acceleration due to gravity, is the rolling moment of inertia, is the pitch moment of inertia, is the yaw moment of inertia, , and They are the rolling moment, yaw moment and pitch moment acting on the high-speed aircraft, The high-speed aircraft along the body coordinate system x The angular velocity component of the shaft, The high-speed aircraft along the body coordinate system y The angular velocity component of the shaft, The high-speed aircraft along the body coordinate system z The angular velocity component of the shaft, t For time.
3. The multi-objective optimization control method for high-speed aircraft according to claim 2, characterized in that: The calculation formula of the control instruction is: ;in, is the expected value of the attitude angle, is the expected value of attitude angular rate, is the first-order derivative of the expected value of the attitude angle, is the first-order derivative of the expected value of the attitude angular rate, is the posture tracking error, is the uncertainty disturbance observation value of the outer loop of the high-speed aircraft attitude by the robust and uniformly convergent observer, is the nonlinear coefficient matrix of the outer loop of the high-speed aircraft attitude, , is the virtual control instruction of the attitude angle, is the attitude angular velocity tracking error, is the uncertainty disturbance observation value of the inner loop of the high-speed aircraft attitude by the robust and uniformly convergent observer, is the nonlinear parameter matrix of the inner loop of the high-speed aircraft attitude, , is the rolling moment of inertia, is the pitch moment of inertia, is the yaw moment of inertia, is the nonlinear coefficient matrix of the inner loop of the high-speed aircraft attitude, , and are the outer loop control gain of the attitude tracking controller and the inner loop control gain of the attitude tracking controller, respectively, u is the control input, , is the dynamic pressure, is the reference area, is the characteristic length, and are the dimensionless coefficients of the rolling moment with respect to the rolling rudder and the rolling angular velocity, respectively. , and are the dimensionless coefficients of the yaw moment with respect to the yaw rudder, yaw angular velocity, and sideslip angle, respectively. , and are the dimensionless coefficients of the pitch moment with respect to the angle of attack, elevator, and pitch angular velocity, respectively, is the elevator deflection angle, is the yaw rudder deflection angle, is the rolling rudder deflection angle.
4. The multi-objective optimization control method for high-speed aircraft according to claim 1, characterized in that: The mathematical expression of the tracking error performance index is: ;in, is the tracking error performance indicator, is the tracking error number, , is the tracking error, weight Satisfaction relationship ; The mathematical expression of the control vibration performance index is: ;in, is the control chattering performance index; is the roll rudder deflection command at the current moment, is the roll rudder deflection command at the last moment, is the yaw rudder deflection instruction at the current moment, The yaw rudder deflection command at the last moment, is the elevator deflection command at the current moment, is the elevator deflection instruction at the previous moment; the mathematical expression of the angle of attack amplitude performance index is: ;in, is the attack angle amplitude performance index, is the flight angle of attack, satisfying the conditions: , are the minimum and maximum values of the flight angle of attack amplitude, is the flight angle of attack constraint boundary value of the high-speed aircraft, , non-singular quantity ; The mathematical expression of the control energy performance index is: ;in, are the weights of each item, all greater than , and satisfies , is the control energy performance indicator.
5. The multi-objective optimization control method for high-speed aircraft according to claim 3, characterized in that: Based on the NSGA-III multi-objective optimization algorithm, optimization simulation is performed in sequence to obtain performance indicators and performance parameters under each group of optimization scenarios, and a performance indicator parameter set is formed according to the performance indicators and the performance parameters, including: Set the initial parameters of the NSGA-III algorithm and the initial state values of the attitude tracking controller used in the current state of the high-speed aircraft, and define the initial reference point; Set evolution generation , randomly generate the initial population , the initial population The corresponding attitude tracking controller of the high-speed aircraft is introduced to calculate the values of the tracking error performance index, the control buffeting performance index, the angle of attack amplitude performance index and the control energy performance index. A fast Pareto non-dominated sorting is performed according to the obtained values of each performance index, and random selection, simulated binary crossover and polynomial mutation are performed to obtain the offspring population. ; The initial population With offspring population Merge into new population ; For new populations , perform non-dominated sorting according to the Pareto dominance relationship, obtain the non-dominated layer, and use the non-dominated layer to construct a new population in turn , until the new population The size of the population equals or exceeds the initial population for the first time ; Based on the reference point idea, the values of various performance indicators calculated by the attitude tracking controller of the high-speed aircraft are used to generate the new population. Selection and Initial Population Individuals of equal size generate a new parent population ; judge Is it equal to the maximum evolutionary generation? If not, then the evolutionary generation And return to step "set the initial population With offspring population Merge into new population "; If so, the performance indicator parameter set is determined according to the calculated performance indicator value.
6. The multi-objective optimization control method for high-speed aircraft according to claim 3, characterized in that: Based on the coefficient of variation method, the comprehensive performance evaluation index is calculated according to the performance index parameter set to obtain the optimal performance index and optimal performance parameter under each group of optimization scenarios, including: The indicators in the performance indicator parameter set are forward processed to obtain each forward processed indicator; the forward processing formula is: ;in, represents the processed performance indicator sample, represents the performance indicator sample before processing, Indicates the performance indicator number, , represents the sequence number of the Pareto optimal solution set, is a positive constant, Indicates the number of samples among all The maximum absolute value of the negative indicator; Each index after positive processing is standardized to obtain the standardized index; the formula for the standardized processing is: , represents the standardized performance index sample after the standardization process, Indicates the number of Pareto optimal solution sets; The weights of each standardized indicator are calculated according to the coefficient of variation; the weight calculation formula is: ;in Indicates The weight of each performance indicator, Indicates The coefficient of variation of item weights, ; The comprehensive performance evaluation index is determined according to each standardized index and the corresponding weight; the calculation formula of the comprehensive performance evaluation index is: , represents the weight of the tracking error performance indicator, represents the weight of the control chattering performance index, represents the weight of the attack angle amplitude performance index, represents the weight of the control energy performance index; The optimal performance index and optimal performance parameters under each group of optimization scenarios are determined according to the comprehensive performance evaluation index.
Citation Information
Patent Citations
Multi-target control parameter tuning method of unmanned aerial vehicle cluster control system
CN112783210A
Unmanned aerial vehicle energy-saving flight control method and system based on intelligent deformation decision
CN114637312A
Hypersonic aircraft attitude control method and system, electronic equipment and medium
CN115793696A
Aircraft guidance and control integrated design method based on preset performance
CN117369509A
High-speed aircraft intelligent control method based on deep forest algorithm
CN118034345A
Cited By
Guidance control method based on particle optimization algorithm and differential game in tripartite game scene
CN119847179A
Aircraft multi-parameter joint design method and system based on deep learning and storage medium
CN119862658A
Multi-parameter Joint Design Method, System and Storage Medium for Aircraft Based on Deep Learning
CN119862658B
Aircraft landing control parameter multi-objective optimization method based on NSGA-II algorithm
CN120161758A
Plasma virtual control surface-oriented cooperative group optimization control distribution method for reentry flight of high-speed aircraft
CN120215548A
