Verification method of vibration characteristic parameters at key positions of complex equipment

Abstract

The present invention discloses a method for verifying vibration characteristic parameters of key positions of complex equipment, comprising the following steps: 1) constructing a model of parts of complex equipment; 2) establishing a dynamic model of complex equipment in simulation software; 3) obtaining the connection mode and constraint relationship between the parts during the movement of the physical complex equipment; 4) pre-simulating the complex equipment model in dynamic simulation software; 5) determining the vibration characteristic parameters of the key positions of the complex equipment to be verified, and post-processing the vibration characteristic parameters under different levels of road spectra and vehicle speeds; 6) using a neural network model to train the stiffness damping coefficient and vibration characteristic parameters of the selected key positions; 7) comparing and verifying the vibration characteristic parameters obtained by the dynamic model of the complex equipment in the simulation process with the vibration characteristic parameters obtained by the neural network training model. The present invention can simplify the analysis cost of vibration characteristics, thereby reducing the manufacturing cost of complex equipment.

Description

Technical Field

[0001] The present invention belongs to the technical field of equipment simulation, and in particular relates to a method for analyzing and verifying vibration characteristic parameters at key positions of complex equipment. Background Art

[0002] With the rapid development of scientific information technology, the manufacturing methods of complex equipment are becoming more and more sophisticated, and there are more and more ways to combine parts. There is an urgent need for an efficient design method to analyze the vibration characteristics of key positions. The traditional design method is mainly based on practical experiments. Through continuous debugging, the corresponding optimization design scheme is proposed. This method consumes a lot of manpower and material resources. With the emergence of virtual prototype technology, the transformation of the research and development mode of complex equipment has been promoted. Establishing virtual prototype models, analyzing and studying the key positions of complex equipment, and completing the optimization design of parameters in manufacturing have become an indispensable technical means in the equipment development process. Therefore, it is very important to study the influence of certain important parameters on the vibration characteristics of key positions and realize the analysis and verification of vibration characteristic parameters. Summary of the invention

[0003] The technical problem to be solved by the present invention is that the vibration characteristics of key positions such as suspension devices and balance elbow torsion angles during the actual movement of complex equipment are difficult to measure. In the process of improving the vibration reduction performance of complex equipment systems, how to verify the vibration characteristics of key positions of complex equipment is a key issue.

[0004] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0005] A method for verifying vibration characteristic parameters of key positions of complex equipment includes the following steps:

[0006] 1) Based on the principle of multi-body system dynamics, build models of complex equipment parts and perform symbolic conventions for dynamic analysis;

[0007] 2) Establish a complex equipment dynamics model in the simulation software, determine the constraint relationship and force of each component of the complex equipment under the dynamic response of different levels of road spectra, find the corresponding constraint relationship and force in the Professional column of the dynamics simulation software, and add them to each component of the simulation model;

[0008] 3) During the process of physical complex equipment, the connection mode and constraint relationship between the various parts are obtained. The connection mode is used to assemble the virtual prototype model, and the constraint relationship is used to make the virtual prototype model simulate correctly;

[0009] When the complex equipment is a tracked vehicle, the component constraints include:

[0010] The rotating pair between the vehicle body and the driving wheel, road wheel and track roller;

[0011] The contact relationship between the ground and the track shoe;

[0012] Initial angle of balance axis;

[0013] Translation pair on the tensioning device;

[0014] 4) In the dynamics simulation software, the simulation time and step length are given, and the complex equipment model is pre-simulated. In the post-processing result module, the validity of the complex equipment model is verified by viewing the output chart. After the validity verification is passed, the simulation parameters are set to simulate the complex equipment model; the simulation parameters include simulation time, step length and frame number, etc.;

[0015] 5) Determine the vibration characteristic parameters of the key positions of the complex equipment that needs to be verified, and post-process the vibration characteristic parameters under different levels of road spectra and vehicle speeds;

[0016] 6) Using a neural network model to train the stiffness damping coefficient and vibration characteristic parameters of the selected key positions, and obtaining a fitting relationship between the stiffness damping coefficient and vibration characteristic parameters of the key positions;

[0017] 7) The stiffness and damping coefficient of the suspension device in the neural network training data is transmitted back to the dynamics simulation software, and the vibration characteristic parameters are recalculated in the post-processing module. The vibration characteristic parameters obtained by the complex equipment dynamics model in the simulation process are compared and verified with the vibration characteristic parameters obtained by the neural network training model.

[0018] The beneficial effects achieved by the present invention are as follows: the present invention uses three-dimensional modeling software to import the built components of complex equipment into dynamics simulation software, simulates the vibration model of key positions of the complex equipment, obtains the vibration characteristic parameters of the key positions, analyzes and predicts the obtained parameters in combination with small sample deep learning, transmits the data obtained from the training of the neural network model back to the simulation model for comparison and verification, and analyzes the optimal situation of the vibration characteristics of the key positions of the complex equipment while ensuring that the quality and moment of inertia of each component are in line with the actual situation. Compared with the traditional complex equipment design and manufacturing process, the consumption of manpower and material resources can be greatly reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 It is a flow chart of a method for analyzing and verifying vibration characteristic parameters at key positions of complex equipment provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0020] The present invention will be further described below in conjunction with the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and cannot be used to limit the protection scope of the present invention.

[0021] like Figure 1As shown, a method for verifying vibration characteristic parameters of key positions of complex equipment includes the following steps:

[0022] 1) Based on the principle of multi-body system dynamics, build models of complex equipment parts and perform symbolic conventions for dynamic analysis;

[0023] 2) Establish a complex equipment dynamics model in the simulation software, determine the constraint relationship and force of each component of the complex equipment under the dynamic response of different levels of road spectra, find the corresponding constraint relationship and force in the Professional column of the dynamics simulation software, and add them to each component of the simulation model;

[0024] 3) During the process of physical complex equipment, the connection mode and constraint relationship between the various parts are obtained. The connection mode is used to assemble the virtual prototype model, and the constraint relationship is used to make the virtual prototype model simulate correctly;

[0025] When the complex equipment is a tracked vehicle, the component constraints include:

[0026] The rotating pair between the vehicle body and the driving wheel, road wheel and track roller;

[0027] The contact relationship between the ground and the track shoe;

[0028] The initial angle of the balance axis;

[0029] Translation pair on the tensioning device;

[0030] 4) In the dynamics simulation software, the simulation time and step length are given, and the complex equipment model is pre-simulated. In the post-processing result module, the validity of the complex equipment model is verified by viewing the output chart. After the validity verification is passed, the simulation parameters are set to simulate the complex equipment model; the simulation parameters include simulation time, step length and frame number, etc.;

[0031] 5) Determine the vibration characteristic parameters of the key positions of the complex equipment that needs to be verified, and post-process the vibration characteristic parameters under different levels of road spectra and vehicle speeds;

[0032] 6) Using a neural network model to train the stiffness damping coefficient and vibration characteristic parameters of the selected key positions, and obtaining a fitting relationship between the stiffness damping coefficient and vibration characteristic parameters of the key positions;

[0033] 7) The stiffness and damping coefficient of the suspension device in the neural network training data is transmitted back to the dynamics simulation software, and the vibration characteristic parameters are recalculated in the post-processing module. The vibration characteristic parameters obtained by the complex equipment dynamics model in the simulation process are compared and verified with the vibration characteristic parameters obtained by the neural network training model.

[0034] Furthermore, in step 1), in the process of constructing the complex equipment component model, a topological diagram of each component is constructed to represent the connection method between each component, and each component is assembled into a model identical to the physical complex equipment in the dynamic simulation software.

[0035] In the process of assembling various components into a model identical to the physical complex equipment in the dynamic simulation software, the various components of the complex equipment constructed by the 3D modeling software are imported into the dynamic simulation software, and the model is established according to the geometric position relationship of each component, including the revolute pairs added to the driving wheel, road wheel, and traction wheel, the translation pairs required for the suspension device, and the contact relationship between the balance elbow and the road wheel.

[0036] According to the driving mode of complex equipment such as physical tracked vehicles, as well as the connection method and contact and collision method between each component, symbol conventions are made for the components. Components with mass are body elements, represented by circles, and components without mass are hinge elements, represented by triangles. The connection method between components is represented by arrows.

[0037] Furthermore, in step 4), during the pre-simulation process of the complex equipment model, a driving force is added to the complex equipment model with constraints. When the complex equipment is a tracked vehicle, a motion attribute is added to the rotation pair on the driving wheel of the tracked vehicle, and a step function is added as the driving force to replace the engine module of the real complex equipment. The end time, step size and number of frames are selected for pre-simulation. In the post-processing module, click Plot to view the output results of the center of mass of each component of the virtual prototype model. The output results include the components of the center of mass velocity, acceleration, displacement, torque, etc. of the vehicle body and the balance elbow in the x, y, and z coordinate directions.

[0038] Furthermore, in step 5), when the complex equipment is a tracked vehicle, the target vibration characteristic parameters are the root mean square value of the vibration acceleration of the vehicle body and the balance elbow center of mass and the balance elbow torsion angle, and the analysis target is the influence of the active wheel drive parameters, suspension device stiffness and damping on the target vibration characteristic parameters.

[0039] Furthermore, in step 5), the vibration characteristic parameters are obtained from the post-processing module of the complex equipment model simulation results, and the components of the vertical acceleration and displacement of the center of mass of the vehicle body and the balance elbow in the y-coordinate direction are output to the mathematical tool for post-processing, and the following formula is used for calculation:

[0040] SQRT(SUMSQ(A:B) / N)

[0041] Where A is the starting point of the data, B is the end point of the data, and N is the number of data. The above formula can be used to calculate the root mean square value of the vertical acceleration of the center of mass, where SQRT represents the square root of the returned value, and SUMSQ represents the sum of the squares of the returned values.

[0042] Further, in step 6), the analysis process specifically includes the following steps:

[0043] In the post-processing module of the simulation software, the output data and input data are exported in text form, and the exported data is filled with data through the generative adversarial neural network; the data filled in the adversarial generative network is sent to the fully connected neural network;

[0044] The generator and the discriminator in the adversarial generative network are both composed of a multi-layer residual neural network, the activation function in the residual neural network adopts the ReLu function, and the activation function of the fully connected neural network adopts the ReLu function or the Sigmoid function;

[0045] Define the input layer, hidden layer and output layer of the fully connected neural network. The input layer is the speed, suspension stiffness and damping coefficient, and road surface label of the complex equipment dynamics model.

[0046] The hidden layer is the mapping of the input layer under the action of the activation function;

[0047] The output layer is the selected vibration characteristic parameters, such as the vertical acceleration of the center of mass of the vehicle body, the root mean square value of the vertical acceleration of the center of mass of the balancing elbow, and the torsion angle of the balancing elbow as the output layer;

[0048] The data of the input layer is input into the Sigmoid activation function in the hidden layer. The formula is as follows:

[0049]

[0050] The linear combination relationship is used to input g(x i )=w*x i +b is transformed into a nonlinear relationship, where x i is the stiffness and damping coefficient of the selected suspension device, which can be set according to the actual situation through the properties of the simulation software components. w is the weight, b is the bias, e is the exponential function, g(x i ) is the input layer data, f[g(x i )] means passing the input layer data into the Sigmoid function expression; the output results are the vertical acceleration of the vehicle center of mass, the root mean square value of the vertical acceleration of the balance elbow center of mass, and the torsion angle of the balance elbow. The optimal weight and bias are fitted through a fully connected neural network.

[0051] In step 7), the stiffness and damping coefficient of the suspension device obtained by neural network training is transmitted back to the dynamics simulation software for verification. The end time, step size, number of frames and vibration characteristic parameters of the selected key positions are defined in the dynamics simulation software, and then simulation is performed. Through the post-processing module of the dynamics simulation software, the vertical acceleration of the center of mass of the vehicle body and the balance elbow and the component of the displacement in the y-coordinate direction are output to the mathematical tool for post-processing to obtain the root mean square value of the vertical acceleration of the center of mass. The root mean square value of the vertical acceleration of the center of mass obtained by the simulation process is compared and verified with the root mean square value of the vertical acceleration of the center of mass obtained by the neural network training model.

[0052] The present invention sets the values ​​of stiffness and damping coefficients. On the one hand, the parameters are trained by a neural network to obtain the best fitting curve, and the vertical acceleration of the center of mass of the vehicle body, the root mean square value of the vertical acceleration of the center of mass of the balancing elbow, and the torsion angle of the balancing elbow are obtained. Then, the values ​​of stiffness and damping coefficients are substituted into the simulation software, and the stiffness and damping coefficient of the suspension device are set through the properties to perform simulation. In the simulation post-processing module, the vertical acceleration of the center of mass of the vehicle body, the root mean square value of the vertical acceleration of the center of mass of the balancing elbow, and the torsion angle of the balancing elbow obtained by output are observed and compared with the values ​​obtained by the training model for verification.

[0053] The above embodiments are only used to illustrate the technical solutions of the invention rather than to limit them. Researchers in the relevant field can still modify or make equivalent substitutions to the specific implementation modes of the present invention with reference to the above embodiments. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention are within the scope of protection of the claims of the present invention to be approved.

Claims

1. A method for verifying vibration characteristic parameters at key locations of complex equipment. It is characterized in that The following steps are involved: 1) Based on the principle of multi-body system dynamics, build models of complex equipment parts and perform symbolic conventions for dynamic analysis; 2) Establish a complex equipment dynamics model in the simulation software, determine the constraint relationship and force of each component of the complex equipment under the dynamic response of different levels of road spectra, find the corresponding constraint relationship and force in the Professional column of the dynamics simulation software, and add them to each component of the simulation model; 3) During the process of physical complex equipment, the connection mode and constraint relationship between the various parts are obtained. The connection mode is used to assemble the virtual prototype model, and the constraint relationship is used to make the virtual prototype model simulate correctly; 4) In the dynamics simulation software, the simulation time and step length are given, and the complex equipment model is pre-simulated. In the post-processing result module, the validity of the complex equipment model is verified by viewing the output chart. After the validity verification is passed, the simulation parameters are set to simulate the complex equipment model; the simulation parameters include simulation time, step length and frame number; 5) Determine the vibration characteristic parameters of the key positions of the complex equipment that needs to be verified, and post-process the vibration characteristic parameters under different levels of road spectra and vehicle speeds; 6) Using a neural network model to train the stiffness damping coefficient and vibration characteristic parameters of the selected key positions, and obtaining a fitting relationship between the stiffness damping coefficient and vibration characteristic parameters of the key positions; 7) The stiffness and damping coefficient of the suspension device in the neural network training data is transmitted back to the dynamics simulation software, and the vibration characteristic parameters are recalculated in the post-processing module. The vibration characteristic parameters obtained by the complex equipment dynamics model in the simulation process are compared and verified with the vibration characteristic parameters obtained by the neural network training model.

2. The method for verifying vibration characteristic parameters of key positions of complex equipment according to claim 1, Features: In step 1), in the process of constructing the complex equipment component model, a topological diagram of each component is constructed to represent the connection method between each component, and each component is assembled into a model identical to the physical complex equipment in the dynamic simulation software.

3. The method for verifying vibration characteristic parameters of key positions of complex equipment according to claim 2, Features: In the process of assembling various components into a model identical to the physical complex equipment in the dynamic simulation software, the various components of the complex equipment constructed by the 3D modeling software are imported into the dynamic simulation software, and the model is established according to the geometric position relationship of each component, including the revolute pairs added to the driving wheel, road wheel, and traction wheel, the translation pairs required for the suspension device, and the contact relationship between the balance elbow and the road wheel.

4. The method for verifying vibration characteristic parameters of key positions of complex equipment according to claim 1, Features: In step 3), when the complex equipment is a tracked vehicle, the component constraint relationships include: The rotating pair between the vehicle body and the driving wheel, road wheel and track roller; The contact relationship between the ground and the track shoe; The initial angle of the balance axis; Translation pair on the tensioning device.

5. The method for verifying vibration characteristic parameters of key positions of complex equipment according to claim 1, Features: In step 4), during the pre-simulation process of the complex equipment model, a driving force is added to the complex equipment model with constraints. When the complex equipment is a tracked vehicle, a motion attribute is added to the rotation pair on the driving wheel of the tracked vehicle, and a step function is added as the driving force to replace the engine module of the real complex equipment. The end time, step size and number of frames are selected for pre-simulation. In the post-processing module, click Plot to view the output results of the center of mass of each component of the virtual prototype model. The output results include the center of mass velocity, acceleration, displacement, and torque of the vehicle body and the balance elbow in the x, y, and z coordinate directions.

6. The method for verifying vibration characteristic parameters of key positions of complex equipment according to claim 1, Features: In step 5), when the complex equipment is a tracked vehicle, the target vibration characteristic parameters are the root mean square value of the vibration acceleration of the vehicle body and the balance elbow center of mass and the balance elbow torsion angle, and the analysis target is the influence of the active wheel drive parameters, suspension device stiffness and damping on the target vibration characteristic parameters.

7. The method for verifying vibration characteristic parameters of key positions of complex equipment according to claim 6, It is characterized by: In step 5), the vibration characteristic parameters are obtained from the post-processing module of the complex equipment model simulation results, and the vertical acceleration of the center of mass of the vehicle body and the balance elbow and the component of the displacement in the y-coordinate direction are output to the mathematical tool for post-processing, and the following formula is used for calculation: SQRT(SUMSQ(A:B) / N) Where A is the starting point of the data, B is the end point of the data, and N is the number of data. The above formula can be used to calculate the root mean square value of the vertical acceleration of the center of mass, where SQRT represents the square root of the returned value, and SUMSQ represents the sum of the squares of the returned values.

8. The method for verifying vibration characteristic parameters of key positions of complex equipment according to claim 1, Features: In step 6), the analysis process specifically includes the following steps: In the post-processing module of the simulation software, the output data and input data are exported in text form, and the exported data is filled with data through the generative adversarial neural network; the data filled in the adversarial generative network is sent to the fully connected neural network; The generator and the discriminator in the adversarial generative network are both composed of a multi-layer residual neural network, the activation function in the residual neural network adopts the ReLu function, and the activation function of the fully connected neural network adopts the ReLu function or the Sigmoid function; The input layer, hidden layer and output layer of the fully connected neural network are defined. The input layer is the speed, suspension stiffness and damping coefficient, and road surface label of the complex equipment dynamics model; the hidden layer is the mapping of the input layer under the action of the activation function; the output layer is the selected vibration characteristic parameters, such as the vertical acceleration of the center of mass of the vehicle body, the root mean square value of the vertical acceleration of the center of mass of the balancing elbow, and the torsion angle of the balancing elbow as the output layer.

9. According to the complex equipment key position vibration characteristic parameter verification method of claim 8, the data of the input layer is input into the Sigmoid activation function in the hidden layer, and the formula is as follows: The linear combination relationship is used to input g(x i )=w*x i +b is transformed into a nonlinear relationship, where x i is the stiffness and damping coefficient of the selected suspension device, which can be set according to the actual situation through the properties of the simulation software components. w is the weight, b is the bias, e is the exponential function, and g(x i ) is the input layer data, f[g(x i )] means passing the input layer data into the Sigmoid function expression; the output results are the vertical acceleration of the vehicle center of mass, the root mean square value of the vertical acceleration of the balance elbow center of mass, and the torsion angle of the balance elbow. The optimal weight and bias are fitted through a fully connected neural network.

10. The method for verifying vibration characteristic parameters of key positions of complex equipment according to claim 1, Features: In step 7), the stiffness and damping coefficient of the suspension device obtained by neural network training is transmitted back to the dynamics simulation software for verification. The end time, step size, number of frames and vibration characteristic parameters of the selected key positions are defined in the dynamics simulation software, and then simulation is performed. Through the post-processing module of the dynamics simulation software, the vertical acceleration of the center of mass of the vehicle body and the balance elbow and the component of the displacement in the y-coordinate direction are output to the mathematical tool for post-processing to obtain the root mean square value of the vertical acceleration of the center of mass. The root mean square value of the vertical acceleration of the center of mass obtained by the simulation process is compared and verified with the root mean square value of the vertical acceleration of the center of mass obtained by the neural network training model.

Images