Dynamics analysis method of helicopter control system considering multiple uncertainties

Abstract

The present invention discloses a method for dynamic analysis of a helicopter control system taking into account multiple uncertain factors. First, a rigid-flexible coupling dynamic simulation model of a collective control system is established; then a parameterized model of corresponding uncertain parameters is established, and the dynamic response of a motion mechanism under different parameters is obtained through this calculation method; finally, an uncertainty analysis method is created with motion transfer accuracy as an optimization index. The overall parameters obtained through this optimization method can effectively reduce motion transfer errors, thereby improving control accuracy. The uncertainty analysis method of the control system of the present invention has universal applicability to uncertainty analysis and optimization problems of complex motion mechanisms, and can be more widely used in the field of aviation science.

Description

Technical Field

[0001] The invention belongs to the technical field of dynamics, and in particular relates to a dynamics analysis method for a helicopter control system. Background Art

[0002] Helicopters can take off and land vertically, can move forward, backward, left, and right in the air, and can hover at a certain height. They do not require high quality landing sites. These characteristics have made them increasingly widely used in various military and civilian fields. The control system is one of the important components of a helicopter. Its function is to change the flight attitude of the helicopter. Therefore, the motion mechanism of the control system must have high reliability and motion accuracy in order to achieve more precise control of the helicopter.

[0003] For a long time, the dynamics of motion mechanisms have been studied mainly by theoretical calculations combined with engineering practice and experimental verification. This traditional method requires a large amount of experimental calculations and relies solely on training helicopter pilots' control capabilities, and cannot provide reasonable and feasible guidance for the structural design and precision improvement of helicopter control systems.

[0004] With the development of computer virtual prototype technology, the use of multi-body dynamics simulation software to analyze the dynamics of complex motion mechanisms has become a common and reliable approach. However, there are a large number of uncertain factors in the operation of helicopter control systems. For complex, strongly nonlinear multi-body systems such as control systems, small parameter changes may lead to huge changes in dynamic response. At this time, the commonly used deterministic analysis method is not feasible. Therefore, it is urgent to develop a set of dynamic analysis methods that can take into account the various uncertain factors that may be encountered when the control system is working. Summary of the invention

[0005] In order to overcome the shortcomings of the prior art, the present invention provides a method for dynamic analysis of a helicopter control system that takes into account multiple uncertain factors. First, a rigid-flexible coupling dynamic simulation model of the collective control system is established; then a parameterized model of the corresponding uncertain parameters is established, and the dynamic response of the motion mechanism under different parameters is obtained through this calculation method; finally, an uncertainty analysis method is created with motion transfer accuracy as an optimization index. The overall parameters obtained through this optimization method can effectively reduce the motion transfer error, thereby improving the control accuracy. The uncertainty analysis method of the control system of the present invention is universal for uncertainty analysis and optimization problems of complex motion mechanisms, and can be more widely used in the field of aviation science.

[0006] The technical solution adopted by the present invention to solve the technical problem includes the following steps:

[0007] Step 1: Create an initial sample space of uncertain design variables of the helicopter control system; the uncertain design variables include the rotational dynamic friction coefficient μ of the lower collective pitch joystick down , the friction coefficient of the upper collective lever rotation μ up , the gap d1 at the first kinematic joint, the gap d2 at the second kinematic joint, the gap d3 at the third kinematic joint and the gap d4 at the fourth kinematic joint; the first kinematic joint, the second kinematic joint, the third kinematic joint and the fourth kinematic joint refer to the fixed joint, the revolving joint, the Hook joint and the spherical joint respectively;

[0008] Step 2: Conduct uncertain parameter analysis on the helicopter flight process;

[0009] The interval model is used to describe the distribution type of uncertain design variables, and the upper and lower a% of the initial value of the uncertain design variable is taken as the variation interval;

[0010] Step 3: Establish a rigid-flexible coupling parameterized model of the helicopter control system, and flexibly import the tie rods in the mechanical line system with a force greater than 150N measured in the finite element analysis software ANSYS into the multi-body dynamics simulation software ADAMS to parameterize the uncertain design variables;

[0011] Step 4: Perform multi-body dynamics simulation in ADAMS, perform a collective control on the control system, including two driving commands: lowering and lifting, calculate the collective lever control torque and the angular displacement of the torsion shaft axial rotation, and use these two dynamic response quantities as target responses;

[0012] Step 5: Substitute the initial sample space of uncertain design variables into the multi-body dynamics simulation in step 4 to obtain the target response interval of the control system. The initial sample space and the target response interval together constitute the initial Kriging proxy model of the implicit problem of the control system during helicopter flight.

[0013] Step 6: Use the initial Kriging surrogate model to calculate the reliability sensitivity.

[0014] Furthermore, a=5.

[0015] Furthermore, the process of establishing the rigid-flexible coupling parameterized model of the helicopter control system in step 3 is as follows:

[0016] Step 3-1: Establish a three-dimensional geometric model of the control system;

[0017] Step 3-2: Import the 3D geometric model into the multi-body dynamics simulation software ADAMS, and add kinematic joints between the components of the control system, including fixed joints, revolute joints, Hooke joints and ball joints, to simulate the motion relationship between the components of the control system during driving;

[0018] Step 3-3: Write a helicopter vibration load spectrum program to obtain the vibration excitation data file;

[0019] Step 3-4: Analyze the load information of the control system, apply the hydraulic booster load at the end rocker arm, input the force element load data obtained from the test measurement, fit the spline curve, and apply it in the vertical downward direction of the three sets of rocker arms;

[0020] Step 3-5: Apply speed drive to the rotary pair of the lower collective pitch control lever to achieve the collective pitch lowering and lifting movements;

[0021] Step 3-6: Set the dynamics simulation parameters to enable the control system to achieve smooth collective pitch control;

[0022] Step 3-7: Export the deterministic simulation calculation result file.

[0023] Furthermore, the initial Kriging proxy model in step 5 is constructed by extracting b initial sample points from the initial sample space using the Latin hypercube sampling method, substituting them into the established rigid-flexible coupling parameterized model of the control system for calculation to obtain the initial Kriging proxy model.

[0024] Furthermore, b=10000.

[0025] Furthermore, the reliability sensitivity calculation in step 6 is specifically as follows:

[0026] Step 6-1: Set the reliability analysis convergence criteria;

[0027] Step 6-2: Use the adaptive learning function to select sample points that meet the requirements of the limit state function from the sample space to form the current sample set, which is used to update the Kriging proxy model until the set convergence condition is met;

[0028] Step 6-3: Substitute all samples in the sample space into the Kriging proxy model that meets the convergence conditions, and finally obtain the response quantity sample space;

[0029] Step 6-4: Perform failure mode analysis on the control system problems during driving;

[0030] Step 6-5: Substitute the response sample space into the limit state function, calculate the number of failure sample points, obtain the failure probability, and perform sensitivity analysis using the moment independent global sensitivity method.

[0031] Furthermore, the reliability analysis convergence criterion set in step 6-1 is that the error between the proxy model fitting output response and the simulation result does not exceed 5%.

[0032] Furthermore, the failure criteria for failure mode analysis in step 6-4 include taking a control torque greater than 100,000 N·mm as a failure criterion for difficulty in total distance control, and taking a torsion axis rotation angular displacement greater than 15d as a failure criterion for control motion accuracy.

[0033] The beneficial effects of the present invention are as follows:

[0034] 1. The present invention performs deterministic dynamic simulation on the helicopter control system during flight based on the multi-body dynamics of the motion mechanism, adopts the reliability analysis method, analyzes the uncertainty of random parameters, and provides a design basis for effectively improving the control accuracy of the control system and reducing mechanical failures;

[0035] 2. The present invention proposes a complete set of analysis processes for uncertain parameters of complex motion mechanisms such as helicopter control systems, which includes rigid-flexible coupling multi-body dynamics modeling criteria and a reliability analysis process based on active learning Kriging proxy model. The use of efficient reliability algorithms can solve the contingency caused by small sample impact analysis;

[0036] 3. The uncertainty joint simulation platform established by the present invention solves the problem of insufficient experimental design caused by the inability of current multi-body dynamics simulation software to achieve effective sampling;

[0037] 4. The uncertainty analysis scheme provided by the present invention can provide a technical solution for the uncertainty problem of complex motion mechanisms, including an agent model, an active learning optimization model, and repeated simplified operations. It is suitable for uncertainty analysis of the motion function of a helicopter control system. It has a great reference value for reducing mechanical failures such as control system jams and improving control accuracy, and can be widely used in helicopter inspection and maintenance research. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 It is a schematic diagram of the structure of a helicopter control system of the present invention.

[0039] Figure 2 This is a flow chart of helicopter vibration load spectrum conversion according to the present invention.

[0040] Figure 3 This is a diagram showing the uncertainty calculation results of the control system of the present invention.

[0041] Figure 4 It is a flow chart of the main scheme of the present invention.

[0042] Figure 5 This is a flow chart of the basic content of the present invention. DETAILED DESCRIPTION

[0043] The present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0044] The purpose of the present invention is to propose a dynamic analysis method for the operation of a helicopter control system under the combined effects of multiple uncertain factors. The method first establishes a kinematic and dynamic simulation model of the collective control system through the analysis of the motion law of the collective channel and the connection analysis between the collective line system and the onboard components; secondly, the influencing factors of the motion function of the collective system are taken into account in the simulation model, and the dynamic response quantity in the result is extracted to analyze the cause of the system motion failure; finally, the reliability analysis of the motion function of the collective system is carried out, and the optimization design is carried out according to the analysis results, and improvement suggestions are put forward. This method can intuitively obtain the influence of various uncertain factors on the motion function of the control system, and provide a reference for reducing mechanical failures and improving motion accuracy.

[0045] The present invention performs deterministic dynamics simulation on a helicopter control system during flight based on multi-body dynamics of a kinematic mechanism, and adopts a reliability analysis method to analyze the uncertainty of random parameters, thereby providing a design basis for effectively improving the control accuracy of the control system and reducing mechanical failures.

[0046] The technical ideas of the present invention include:

[0047] (1) Based on the helicopter flight vibration load spectrum, the vibration acceleration time domain spectrum is obtained through inverse fast Fourier transform, and the vibration environment of the control system is established.

[0048] (2) Conduct uncertainty analysis on the working process of the helicopter control system to obtain the uncertainty type and range of the model input variables, integrate the reliability sensitivity algorithm of the complex motion mechanism, establish an ADAMS-MATLAB joint simulation platform, and carry out the tedious sampling process.

[0049] (3) A reliability optimization design method for the control system based on the Kriging model is established. This optimization method can obtain the optimal design goal, effectively improve the control accuracy of the control system, reduce the occurrence of mechanical failures, and provide a technical reference for the structural optimization design of the control system.

[0050] A helicopter control system dynamics analysis method considering multiple uncertainties includes the following steps:

[0051] Step 1: Create an initial sample space of uncertain design variables of the helicopter control system; the uncertain design variables include the rotational dynamic friction coefficient μ of the lower collective pitch joystick down , the friction coefficient of the upper collective lever rotation μ up , a gap d1 at the first kinematic joint, a gap d2 at the second kinematic joint, a gap d3 at the third kinematic joint and a gap d4 at the fourth kinematic joint;

[0052] Step 2: Conduct uncertain parameter analysis on the helicopter flight process;

[0053] The interval model is used to describe the distribution type of uncertain design variables, and the upper and lower 5% of the initial value of the uncertain design variable is taken as the variation interval;

[0054] Step 3: Establish a rigid-flexible coupling parameterized model of the helicopter control system, flexibly import the tie rods with large force in the mechanical line system from the finite element analysis software ANSYS into the multi-body dynamics simulation software ADAMS, and parameterize the uncertain design variables;

[0055] Step 3-1: Establish a three-dimensional geometric model of the control system;

[0056] Step 3-2: Import the 3D geometric model into the multi-body dynamics simulation software ADAMS, and add kinematic joints between the components of the control system, including fixed joints, revolute joints, Hooke joints and ball joints, to simulate the motion relationship between the components of the control system during driving;

[0057] Step 3-3: Write a helicopter vibration load spectrum program to obtain the vibration excitation data file;

[0058] Step 3-4: Analyze the load information of the control system, apply the hydraulic booster load at the end rocker arm, input the force element load data obtained from the test measurement, fit the spline curve, and apply it in the vertical downward direction of the three sets of rocker arms;

[0059] Step 3-5: Apply speed drive to the rotary pair of the lower collective pitch control lever to achieve the collective pitch lowering and lifting movements;

[0060] Step 3-6: Set the dynamics simulation parameters to enable the control system to achieve smooth collective pitch control;

[0061] Step 3-7: Export the deterministic simulation calculation result file;

[0062] Step 4: Perform multi-body dynamics simulation in ADAMS, perform a collective control on the control system, including two driving commands: lowering and lifting, calculate the collective lever control torque and the angular displacement of the torsion shaft axial rotation, and use these two dynamic response quantities as target responses;

[0063] Step 5: Substitute the initial sample space of uncertain design variables into the multibody dynamics simulation of step 4 to obtain the target response range of the control system. The initial sample space and the target response range together constitute the initial Kriging proxy model of the implicit problem of the control system during helicopter flight. Specifically, 10,000 initial sample points are extracted from the initial sample space using the Latin hypercube sampling method, and substituted into the established rigid-flexible coupling parameterized model of the control system for calculation to obtain the initial Kriging proxy model.

[0064] Step 6: Use the initial Kriging surrogate model to calculate the reliability sensitivity.

[0065] Step 6-1: Set the reliability analysis convergence criteria; the error between the surrogate model fitting output response and the simulation result should not exceed 5%;

[0066] Step 6-2: Use the adaptive learning function to select sample points that meet the requirements of the limit state function from the sample space to form the current sample set, which is used to update the Kriging proxy model until the set convergence condition is met;

[0067] Step 6-3: Substitute all samples in the sample space into the Kriging proxy model that meets the convergence conditions, and finally obtain the response quantity sample space;

[0068] Step 6-4: Perform failure mode analysis on the control system problems during driving; take the control torque greater than 100,000 N·mm as the failure criterion for the difficulty of total distance control, and take the torsion axis rotation angle displacement greater than 15d as the failure criterion for the control motion accuracy;

[0069] Step 6-5: Substitute the response sample space into the limit state function, calculate the number of failure sample points, obtain the failure probability, and perform sensitivity analysis using the moment independent global sensitivity method. Specific embodiment:

[0071] The invention provides an uncertainty analysis method for a helicopter control system in operation, which is composed of a main scheme and two basic contents.

[0072] 1. Main plan

[0073] like Figure 4 As shown in the figure, a motion mechanism uncertainty analysis platform based on the agent model is built. The specific steps are as follows:

[0074] (1) Creating an initial sample space of uncertain design variables of the helicopter control system. The six uncertain design variables used in the present invention include the rotational kinetic friction coefficient μ of the lower collective pitch joystick down , the friction coefficient of the upper collective lever rotation μ up , a gap d1 at the first kinematic joint, a gap d2 at the second kinematic joint, a gap d3 at the third kinematic joint and a gap d4 at the fourth kinematic joint;

[0075] (2) Uncertainty analysis of the control process of the helicopter control system is carried out. Since it is not convenient to determine the parameter type of the uncertain parameters of complex motion mechanisms such as the control system through a large number of experiments, the interval model is used to describe the distribution type of uncertain variables. This method only requires knowing the upper and lower limits of the variable, and takes 5% above and below the initial value of the variable as the variation interval;

[0076] (3) Establish a rigid-flexible coupling parametric model of the helicopter control system based on multi-body dynamics. The modeling process is shown in Basic Content 1. The pull rods in the control line system are flexibly imported from ANSYS into ADAMS, and the design variables are parameterized so that the model modification operation can be performed in the cmd model information file.

[0077] (4) Perform multi-body dynamics simulation in ADAMS, perform a collective maneuver on the control system, calculate the collective drive torque and the torsion axis rotation angle displacement, and use them as the target response;

[0078] (5) Substitute the initial sample space of uncertain design variables into (4) and calculate the dynamic response interval by modifying the corresponding parameter values. The target response interval of the initial sample space together constitutes the initial Kriging model of the implicit problem of the control system during helicopter driving;

[0079] (6) Use this proxy model to perform reliability sensitivity calculation. The reliability calculation process can be found in Basic Content 2.

[0080] 2. Basic content 1

[0081] like Figure 5 As shown in the figure, the specific steps of the rigid-flexible coupling multi-body dynamics modeling process of the helicopter flight control system are as follows:

[0082] (1) Establish the geometric model of the control system based on kinematics, multi-rigid body dynamics and multi-flexible body dynamics;

[0083] (2) Simplify the imported geometric model in the multi-body dynamics simulation software ADAMS and delete some insignificant parts. Add kinematic joints between the components of the control system, including fixed joints, revolute joints, Hooke joints, and ball joints, to simulate the motion relationship between the components of the control system during driving;

[0084] (3) Using the helicopter vibration load spectrum, write a power spectrum density inverse fast Fourier transform program to obtain the vibration acceleration load in the time domain;

[0085] (4) Analyze the load information of the control system, apply the hydraulic assist load at the end rocker arm, input the force element load data obtained by the test measurement, fit the spline curve with the function CUBSPL (time, 0, Gforce1Y, 0), and apply it in the vertical downward direction of the three groups of rocker arms;

[0086] (5) Apply the speed drive function STEP(time,0,0,3,15d)+STEP(time,3,0,6,-15d) to the rotary pair of the lower collective pitch joystick to achieve the collective pitch lowering and lifting movements;

[0087] (6) Dynamic simulation parameters are set, with a simulation time of 6 seconds, an analysis step of 1000 steps, and all model units using Newton-Kilogram-Millimeter-Second, so that the control system can achieve the working requirements of smooth collective pitch control.

[0088] (7) Export the deterministic simulation calculation result file.

[0089] 3. Basic content 2

[0090] The specific steps of the reliability analysis process based on the active learning Kriging agent model are as follows:

[0091] (1) From the initial sample space consisting of 5% of the initial values ​​of the six uncertain design variables as the upper and lower limits of the interval, 10,000 initial sample points were extracted using the Latin hypercube sampling method, and substituted into the rigid-flexible coupling model established previously for calculation to obtain the initial proxy model in the main scheme;

[0092] (2) Setting the reliability analysis convergence criteria, where the convergence condition is that the error between the output response fitted by the surrogate model and the simulation result does not exceed 5%;

[0093] (3) Using the adaptive learning function, sample points that meet the requirements of the limit state function are selected from the sample space to form the current sample set, so as to update the Kriging model until the preset convergence conditions are met;

[0094] (4) Substitute all samples in the sample space into the Kriging model that meets the convergence conditions, and finally obtain the response quantity sample space;

[0095] (5) Conduct failure mode analysis on the control problems of the control system during operation, and use a control torque greater than 100,000 N·mm as the failure criterion for collective control difficulty, and a torsion axis rotation angular displacement greater than 15d as the failure criterion for control motion accuracy;

[0096] (6) Substitute the response sample space into the limit state function, calculate the number of failure sample points, obtain the failure probability, and use the moment independent global sensitivity method to perform sensitivity analysis.

[0097] The following will be described in more detail with specific reference to an example of a helicopter control system and a specific implementation method, so that designers can use the present invention more conveniently and extend it to the analysis of other complex motion mechanisms.

[0098] The specific implementation of this example is as follows:

[0099] (1) Establish an initial proxy model for the implicit problem of helicopter control system during flight. Call the DACE toolbox in MATLAB;

[0100] (2) Use the DACE toolbox to create the initial sample space of the design variables when the control system is working. It includes 6 design variables such as Figure 3 As shown. Using the Latin hypercube sampling method, a sample space containing 10,000 sets of design variables is extracted;

[0101] (3) Referring to the control system model of a single-rotor helicopter, CAD modeling is performed in CATIA modeling software to obtain a three-dimensional model of the control system;

[0102] (4) In order to reduce the connection relationship of the dynamic model and improve the calculation speed, the control system model is reasonably simplified in Solidworks. The top automatic tilt device and the periodic control line system are deleted, and only the collective control line system and the compound rocker arm torsion shaft are retained. The simplified model structure is as follows: Figure 1 As shown;

[0103] (5) Import the simplified control system model into ADAMS software, set the material properties of each component of the control system, and select aluminum alloy as the component material;

[0104] (6) There are many connections between tie rods and rocker arms in the control system. Since the control system itself is relatively complex, the connection between the two needs to use Hook joints and ball joints to release a certain degree of freedom to avoid errors during simulation. Use three sets of equivalent forces to replace the effect of the hydraulic booster on the control system, establish a flange bearing on the torsion shaft, and install the webs on both sides on the torsion shaft;

[0105] (7) Create a square vibration table directly below the support. The vibration table is connected to the ground through a fixed pair. There are four pin holes on the support. A bushing (damper) is added in the middle of the four holes to constrain the vibration table and the support. The connection relationship between the two is defined by the stiffness and damping in the equivalent translation and rotation directions. The vibration table can simulate the corresponding vibration environment of the helicopter by applying Motion. It only needs to set the vibration acceleration load spectrum in the XYZ direction in the form of a spline curve in Motion.

[0106] (8) According to the helicopter vibration environment power spectrum density in my country's GJB150.16A-2009 document, and the control system is installed close to the transmission system, based on the fast Fourier inverse transform method of the vibration signal, the frequency domain to time domain program is written using MATLAB to generate the helicopter random vibration acceleration time domain signal, and the values ​​in the range of 0 to 6s are extracted and imported into the software to generate the corresponding spline curve according to the unit system in ADAMS multiplied by 1000;

[0107] (9) Import the x_t files of the lower and upper collective tie rods into ANSYS for flexibility. The mesh adopts ten-node tetrahedral elements. According to the constraints applied on the tie rods, MPC and rigid regions are established on the flexible tie rods, and the MNF flexibility file is output;

[0108] (10) Apply a rotary drive STEP(time,0,0,3,15d)+STEP(time,3,0,6,-15d) to the lower collective pitch rod to realize the collective pitch control movement of the helicopter;

[0109] (11) Apply hydraulic assist load at the end rocker arm, input the force element load data obtained by test measurement, fit the spline curve with the function CUBSPL (time, 0, Gforce1Y, 0), and apply it in the vertical downward direction of the three groups of rocker arms;

[0110] (12) Perform a deterministic rigid-flexible coupling motion mechanism dynamics simulation in ADAMS, set the response parameters, output the required total control torque and torsion axis response (angular displacement), and export the batch processing bat file;

[0111] (13) For the control system model studied in this design, the distribution type of the selected parameters cannot be determined in general, so the normal distribution is used to describe the uncertain variables. The uncertain parameter variable values ​​used in this invention are shown in Table 1;

[0112] Table 1 Uncertain parameter variable values

[0113]

[0114] (14) Substitute the parameters of the initial sample space into the simulation calculation in ADAMS to obtain the initial component response space. According to the bat batch command file exported previously, it is used as a computer language command tool to realize automatic dynamic calculation. After the simulation is completed, the above two response information are printed in a text file in sequence;

[0115] (15) Substituting the initial sample space of the design variables and the obtained target response back into the initial surrogate model, updating the surrogate model so that the surrogate model can better fit the implicit relationship;

[0116] (16) Determine whether the accuracy of the proxy model meets the requirements of autonomous learning accuracy. If not, proceed to step (14) and repeat in sequence;

[0117] (17) obtaining an accurate and effective proxy model, and using the proxy model to obtain a response quantity sample space;

[0118] (18) Substitute the response quantity sample space into the functional function assumed above, calculate the number of failure sample points, and obtain the failure probability. Finally, use the moment independent global sensitivity analysis method to calculate the reliability sensitivity, and obtain the failure probability of the control system when it is working and the sensitivity of each design variable.

[0119] (19) The final calculated sensitivity result is as follows Figure 3 shown.

[0120] The helicopter control system has many faults such as difficulty in lowering the collective pitch during the flight control process. The present invention can simulate the dynamic response of the helicopter control system under various uncertain factors through virtual prototype technology, saving a lot of test costs and facilitating the optimization design of the control system structure to improve the reliability of the motion mechanism. The helicopter vibration environment under actual working conditions is taken into account, and the dynamic friction coefficient and wear clearance size at the key kinematic joint are used as uncertain design variables. An ADAMS-MATLAB joint simulation platform is established, and the computer batch command language is called to obtain a large amount of data to fit the active learning Kriging proxy model. Finally, the failure probability of the control system based on two failure modes and the sensitivity of each design variable are calculated with the help of this model. Through this method, a design basis is provided for effectively improving the control accuracy of the control system and reducing mechanical failures, and it has a high engineering application value.

[0121] The innovation of the present invention is:

[0122] 1. Based on the multi-body dynamics method and virtual prototype technology, a dynamic simulation calculation process for the collective pitch control of the control system in the helicopter working environment is proposed. The mechanical wire system and the automatic tilt device part of the control system are innovatively disconnected for research. Three sets of equivalent forces are used to replace the effect of the hydraulic booster on the mechanical wire system below it. The vibration acceleration load of the helicopter during flight is transmitted by setting bushing dampers and Motion. Rigid-flexible coupling dynamic deterministic simulation is performed in ADAMS to obtain the dynamic response of key components. This method does not require experiments, is easy to operate and has low cost.

[0123] 2. Build an ADAMS-MATLAB joint simulation platform, call the computer batch command language, and then perform parametric modeling based on the above-mentioned deterministic simulation model. Substitute the values ​​in the design variable range for multiple calculations, and import the obtained response into MATLAB to fit the active learning Kriging proxy model. Finally, use the Kriging model to calculate the failure probability of the control system under the two failure modes and the sensitivity of each design variable.

Claims

1. A helicopter control system dynamics analysis method considering multiple uncertainties, characterized in that: The steps include: Step 1: Create an initial sample space of uncertain design variables of the helicopter control system; the uncertain design variables include the rotational dynamic friction coefficient μ of the lower collective pitch joystick down , the friction coefficient of the upper collective lever rotation μ up , the gap d1 at the first kinematic joint, the gap d2 at the second kinematic joint, the gap d3 at the third kinematic joint and the gap d4 at the fourth kinematic joint; the first kinematic joint, the second kinematic joint, the third kinematic joint and the fourth kinematic joint refer to the fixed joint, the revolving joint, the Hook joint and the spherical joint respectively; Step 2: Conduct uncertain parameter analysis on the helicopter flight process; The interval model is used to describe the distribution type of uncertain design variables, and the upper and lower a% of the initial value of the uncertain design variable is taken as the variation interval; Step 3: Establish a rigid-flexible coupling parameterized model of the helicopter control system, and flexibly import the tie rods in the mechanical line system with a force greater than 150N measured in the finite element analysis software ANSYS into the multi-body dynamics simulation software ADAMS to parameterize the uncertain design variables; Step 4: Perform multi-body dynamics simulation in ADAMS, perform a collective control on the control system, including two driving commands: lowering and lifting, calculate the collective lever control torque and the angular displacement of the torsion shaft axial rotation, and use these two dynamic response quantities as target responses; Step 5: Substitute the initial sample space of uncertain design variables into the multi-body dynamics simulation in step 4 to obtain the target response interval of the control system. The initial sample space and the target response interval together constitute the initial Kriging proxy model of the implicit problem of the control system during helicopter flight. Step 6: Use the initial Kriging surrogate model to calculate the reliability sensitivity.

2. A helicopter control system dynamics analysis method considering multiple uncertainties according to claim 1, characterized in that: Said a=5.

3. A helicopter control system dynamics analysis method considering multiple uncertainties according to claim 1, characterized in that: The process of establishing the rigid-flexible coupling parameterized model of the helicopter control system in step 3 is as follows: Step 3-1: Establish a three-dimensional geometric model of the control system; Step 3-2: Import the 3D geometric model into the multi-body dynamics simulation software ADAMS, and add kinematic joints between the components of the control system, including fixed joints, revolute joints, Hooke joints and ball joints, to simulate the motion relationship between the components of the control system during driving; Step 3-3: Write a helicopter vibration load spectrum program to obtain the vibration excitation data file; Step 3-4: Analyze the load information of the control system, apply the hydraulic booster load at the end rocker arm, input the force element load data obtained from the test measurement, fit the spline curve, and apply it in the vertical downward direction of the three sets of rocker arms; Step 3-5: Apply speed drive to the rotary pair of the lower collective pitch control lever to achieve the collective pitch lowering and lifting movements; Step 3-6: Set the dynamics simulation parameters to enable the control system to achieve smooth collective pitch control; Step 3-7: Export the deterministic simulation calculation result file.

4. A helicopter control system dynamics analysis method considering multiple uncertainties according to claim 1, characterized in that: The initial Kriging proxy model in step 5 is constructed by extracting b initial sample points from the initial sample space using the Latin hypercube sampling method, substituting them into the established rigid-flexible coupling parameterized model of the control system for calculation to obtain the initial Kriging proxy model.

5. A helicopter control system dynamics analysis method considering multiple uncertainties according to claim 4, characterized in that: Said b=10000.

6. A helicopter control system dynamics analysis method considering multiple uncertainties according to claim 1, characterized in that: The reliability sensitivity calculation in step 6 is specifically as follows: Step 6-1: Set the reliability analysis convergence criteria; Step 6-2: Use the adaptive learning function to select sample points that meet the requirements of the limit state function from the sample space to form the current sample set, which is used to update the Kriging proxy model until the set convergence condition is met; Step 6-3: Substitute all samples in the sample space into the Kriging proxy model that meets the convergence conditions, and finally obtain the response quantity sample space; Step 6-4: Perform failure mode analysis on the control system problems during driving; Step 6-5: Substitute the response sample space into the limit state function, calculate the number of failure sample points, obtain the failure probability, and perform sensitivity analysis using the moment independent global sensitivity method.

7. A helicopter control system dynamics analysis method considering multiple uncertainties according to claim 6, characterized in that: The reliability analysis convergence criterion set in step 6-1 is that the error between the proxy model fitting output response and the simulation result does not exceed 5%.

8. A helicopter control system dynamics analysis method considering multiple uncertainties according to claim 6, characterized in that: The failure criteria for failure mode analysis in step 6-4 include taking a control torque greater than 100,000 N·mm as a failure criterion for difficulty in total distance control, and taking a torsion axis rotation angular displacement greater than 15d as a failure criterion for control motion accuracy.

Images