Response surface and penalty function method-based steel bridge constrained damping design optimization method
By combining the response surface methodology and the penalty function method, the problem of multi-constraint optimization of the design parameters of the constrained damping layer was solved, achieving efficient vibration reduction and noise reduction for steel bridges.
Patent Information
- Application Number
- CN202410983204.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-22
- Publication Date
- 2024-11-12
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
In existing technologies, the design parameters of the constraint damping layer are limited by conditions such as the structure's self-weight should not be too large and the structure's fundamental frequency should not be too small, making it difficult to quickly and accurately obtain the optimal parameter combination to achieve the best vibration reduction and noise reduction effect.
The relationship between design parameters and target response is fitted using the response surface methodology, and the problem is transformed into an unconstrained optimization problem through iterative calculation using the penalty function method. The optimal combination of design parameters is then obtained using central composite design and numerical simulation methods.
It enables the rapid and accurate acquisition of the optimal design parameter combination of the constraint damping layer under multiple constraints, thereby improving the vibration reduction and noise reduction effect of steel bridges.
Smart Images

Figure CN118940372A_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the technical field of vibration reduction and noise reduction of rail transit bridges, and specifically relates to the design of a steel bridge constrained damping design optimization method based on a response surface and penalty function method. Background Art
[0002] Constrained damping layers are widely used in aerospace, shipbuilding, submarines, rail transit and other fields due to their excellent wide-spectrum vibration suppression characteristics. The vibration and noise reduction effect of the constrained damping layer laid on the steel bridge is related to design parameters such as material properties, laying position, laying size and steel bridge structure size. When designers adjust the combination of these parameters to achieve better vibration and noise reduction effects, they will be constrained by conditions such as the deadweight of the structure should not be too large and the fundamental frequency of the structure should not be too small. Therefore, how to obtain the optimal parameter combination quickly and accurately is a problem that needs to be solved urgently by technicians in this field. Summary of the invention
[0003] The purpose of the present invention is to solve the problem that the design parameters of the existing constrained damping layer are constrained by conditions such as the deadweight of the structure should not be too large and the fundamental frequency of the structure should not be too small, and a steel bridge constrained damping design optimization method based on response surface and penalty function method is proposed.
[0004] The technical solution of the present invention is: a steel bridge constrained damping design optimization method based on response surface and penalty function method, comprising the following steps:
[0005] S1. According to the actual situation of the steel bridge with the constrained damping layer and the requirements of vibration and noise reduction, the design parameters and target response of the constrained damping structure are selected.
[0006] S2. Determine the number of test points and the design parameter combination of the test points through central composite design.
[0007] S3. Obtain the target response value corresponding to each design parameter combination through experiments or numerical simulation methods.
[0008] S4. The fitting relationship between the design parameters and the target response values is obtained through the response surface method.
[0009] S5. An optimization objective function is established based on the fitting relationship. Constraints are introduced according to the actual situation of the steel bridge with the constrained damping layer and the requirements of vibration and noise reduction. The constraints are introduced into the objective function in the form of penalty terms, thereby transforming the constrained optimization problem into an unconstrained optimization problem.
[0010] S6. Select an initial iteration point that satisfies the constraint conditions, determine the initial value and change coefficient of the penalty factor, and the iteration termination condition.
[0011] S7. Use an unconstrained optimization method to iteratively calculate the extreme value of the objective function until the iteration termination condition is met.
[0012] Furthermore, the design parameters in step S1 include the material, thickness, arrangement position of the constrained damping structure and the size of the original structure, and the target response includes the sound power level and structural fundamental frequency of the constrained damping structure.
[0013] Furthermore, the number of test points n in step S2 is:
[0014] n=2 k +2k+1
[0015] where k represents the number of design parameters for the central composite design.
[0016] The star arm length γ of the central composite design satisfies the following equation:
[0017]
[0018] Furthermore, the numerical simulation method in step S3 includes a finite element method, a boundary element method and a statistical energy analysis method.
[0019] Further, step S4 includes the following sub-steps:
[0020] S41. Let the influencing variable of a process be X = (x1, x2, ..., x k ), the response variable is y, and the second-order response surface model is constructed as follows:
[0021]
[0022] where x i represents the i-th influencing variable, x j represents the jth influencing variable, k represents the number of influencing variables, β0 represents the initial unknown coefficient, β i Influencing variable x i The unknown coefficient, β ij Influencing variable x i and x j The combined coefficient of unknown number, β ii Influencing variable x i The combined undetermined coefficients, ε represents the error between the fitted response and the true response, the influencing variable X corresponds to the design parameter, and the response variable y corresponds to the target response value;
[0023] S42. Convert the second-order response surface model into a matrix form to obtain:
[0024]
[0025] Where n represents the number of trials, y a represents the target response value of the ath trial, x iarepresents the i-th influencing variable of the a-th experiment, ε a represents the error between the fitted response and the true response of the a-th trial, a=1,2,...,n, i=1,2,...,k;
[0026] S43. According to the principle of least squares method, the coefficient vector β to be determined should make the variance L of the error take the minimum value:
[0027]
[0028] Expand L to get:
[0029] L=Y T Y-2β T X c T Y+β T X c T X c β
[0030] Where ε represents the error vector, Y represents the target response vector, and X c Represents the second-order combination matrix of influencing variables;
[0031] S44. Find the partial derivative of β and set it to 0:
[0032]
[0033] The fitting relationship between the design parameters and the target response value is obtained as follows:
[0034]
[0035] Furthermore, the objective function in step S5 is for:
[0036]
[0037] in represents the objective function, r (l) represents the penalty factor of the lth iteration, which is a decreasing positive sequence. f(X) represents the fitting relationship between the design parameters and the target response value. represents the penalty term, where g u (X)≥0 indicates the uth constraint, and m indicates the number of constraints.
[0038] Furthermore, the initial iteration point X in step S6 (0) Satisfy all constraints and stay away from the boundary. The initial value of the penalty factor is r (0) The value ranges from 1 to 50, the change coefficient C ranges from 0.1 to 0.5, and the iteration termination condition is expressed as:
[0039]
[0040] where ε x and ε y All are allowable errors, X * Represents the extreme points solved using the unconstrained optimization method.
[0041] Further, step S7 includes the following sub-steps:
[0042] S71. Set the number of iterations l=1.
[0043] S72, from X (l-1) Starting from this, we use the unconstrained optimization method to solve the extreme value of the objective function Where Ω represents the value range of the influencing variable X.
[0044] S73, determine whether the iteration termination condition is met, if so, terminate the iteration and obtain the optimal solution X of the objective function * (r (l) ), otherwise go to step S74.
[0045] S74, take r (l+1) =C·r (l) , X (0) =X * (r (l) ), and increase the number of iterations l by 1, and return to step S72.
[0046] The beneficial effects of the present invention are as follows: the present invention fits the relationship between the design parameters and the target response through the response surface method, and adopts the internal penalty function method to iteratively obtain the optimal design parameter combination under multiple constraint conditions, which can provide technical support for the multi-constraint vibration reduction and noise reduction parameter optimization design of the steel bridge structure with the constrained damping layer. BRIEF DESCRIPTION OF THE DRAWINGS
[0047] Figure 1 Shown is a flow chart of a steel bridge constrained damping design optimization method based on a response surface and penalty function method provided in an embodiment of the present invention.
[0048] Figure 2 Shown is a schematic diagram of test point distribution when the number of design parameters provided by an embodiment of the present invention is 2.
[0049] Figure 3 Shown is a schematic diagram of the dimensions of an I-beam and the laying of a constrained damping layer provided in an embodiment of the present invention.
[0050] Figure 4 Shown is a schematic diagram of test point distribution when the number of design parameters provided by an embodiment of the present invention is 3.
[0051] Figure 5Shown is a schematic diagram of a target response iteration process provided by an embodiment of the present invention. DETAILED DESCRIPTION
[0052] Now, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings. It should be understood that the embodiments shown and described in the accompanying drawings are only exemplary and are intended to explain the principles and spirit of the present invention, rather than to limit the scope of the present invention.
[0053] The embodiment of the present invention provides a steel bridge constrained damping design optimization method based on response surface and penalty function method, such as Figure 1 As shown, the following steps S1 to S7 are included:
[0054] S1. According to the actual situation of the steel bridge with the constrained damping layer and the requirements of vibration and noise reduction, the design parameters and target response of the constrained damping structure are selected.
[0055] In the embodiment of the present invention, the design parameters include the material, thickness, arrangement position and size of the original structure of the constrained damping structure, and the target response includes the sound power level and structural fundamental frequency of the constrained damping structure.
[0056] S2. Determine the number of test points and the design parameter combination of the test points through central composite design.
[0057] In the embodiment of the present invention, according to the central composite design, for k design parameters, the number of test points n is:
[0058] n=2 k +2k+1
[0059] In the embodiment of the present invention, the test point on the axis is called a star point, and the distance between the star point and the center point is called the star arm length. For the central composite design, the star arm length γ must satisfy the following equation to ensure the orthogonality between the test points:
[0060]
[0061] In the embodiment of the present invention, Figure 2 As shown in the figure, if the number of design parameters is 2, the number of test points is 9, namely: (1,1), (γ,0), (1,-1), (0,γ), (0,0), (0,-γ), (-1,1), (-γ,0), (-1,-1). Design parameters usually need to be converted into standard variables through linear transformation. For example, when the value range of a design parameter T1 is 12 to 16, the center point value is 14, and the standard variable X1 corresponding to the design parameter T1 is obtained by the following formula:
[0062]
[0063] S3. Obtain the target response value corresponding to each design parameter combination through experiments or numerical simulation methods.
[0064] In the embodiment of the present invention, the numerical simulation method includes a finite element method, a boundary element method and a statistical energy analysis method.
[0065] S4. The fitting relationship between the design parameters and the target response values is obtained through the response surface method.
[0066] Step S4 includes the following sub-steps S41 to S44:
[0067] S41. Let the influencing variable of a process be X = (x1, x2, ..., x k ), the response variable is y, and the second-order response surface model is constructed as follows:
[0068]
[0069] where x i represents the i-th influencing variable, x j represents the jth influencing variable, k represents the number of influencing variables, β0 represents the initial unknown coefficient, β i Influencing variable x i The unknown coefficient, β ij Influencing variable x i and x j The combined coefficient of unknown number, β ii Influencing variable x i The combined undetermined coefficients, ε represents the error between the fitted response and the true response, the influencing variable X corresponds to the design parameter, and the response variable y corresponds to the target response value;
[0070] S42. Convert the second-order response surface model into a matrix form to obtain:
[0071]
[0072] Where n represents the number of trials, y a represents the target response value of the ath trial, x ia represents the i-th influencing variable of the a-th experiment, ε a represents the error between the fitted response and the true response of the a-th trial, a=1,2,...,n, i=1,2,...,k;
[0073] S43. According to the principle of least squares method, the coefficient vector β to be determined should make the variance L of the error take the minimum value:
[0074]
[0075] Expand L to get:
[0076] L=Y T Y-2β T X c T Y+β T X c T X c β
[0077] Where ε represents the error vector, Y represents the target response vector, and X c Represents the second-order combination matrix of influencing variables;
[0078] S44. Find the partial derivative of β and set it to 0:
[0079]
[0080] The fitting relationship between the design parameters and the target response value is obtained as follows:
[0081]
[0082] S5. An optimization objective function is established based on the fitting relationship. Constraints are introduced according to the actual situation of the steel bridge with the constrained damping layer and the requirements of vibration and noise reduction. The constraints are introduced into the objective function in the form of penalty terms, thereby transforming the constrained optimization problem into an unconstrained optimization problem.
[0083] In the embodiment of the present invention, the fitting relationship between the design parameter and the target response value is taken as f(X), and the constraint condition g is introduced. u (X)≥0, then the objective function is:
[0084]
[0085] in represents the objective function, r (l) Represents the penalty factor for the lth iteration, which is a decreasing positive sequence, i.e. r (0) >r (1) >...>r (l) >r (l+1) >...>0, and represents the penalty term, where g u (X)≥0 indicates the uth constraint condition, and m indicates the number of constraints. When the design point X moves from the inside of the feasible region away from the constraint boundary to the boundary (g u (X)=0), the penalty term It increases rapidly and tends to infinity. Therefore, the objective function will not touch the constraint boundary during the optimization iteration process. After the constraint conditions are introduced into the objective function in the form of penalty terms, the constrained optimization problem is transformed into an unconstrained optimization problem.
[0086] S6. Select an initial iteration point that satisfies the constraint conditions, determine the initial value and change coefficient of the penalty factor, and the iteration termination condition.
[0087] In this embodiment of the present invention, the initial iteration point X (0) Satisfy all constraints and stay away from the boundary. The initial value of the penalty factor is r (0) The value ranges from 1 to 50, the change coefficient C ranges from 0.1 to 0.5, and the iteration termination condition is expressed as:
[0088]
[0089] where ε x and ε y All are allowable errors, X * Represents the extreme points solved using the unconstrained optimization method.
[0090] S7. Use an unconstrained optimization method to iteratively calculate the extreme value of the objective function until the iteration termination condition is met.
[0091] Step S7 includes the following sub-steps S71 to S74:
[0092] S71. Set the number of iterations l=1.
[0093] S72, from X (l-1) Starting from this, we use the unconstrained optimization method to solve the extreme value of the objective function Where Ω represents the value range of the influencing variable X.
[0094] S73, determine whether the iteration termination condition is met, if so, terminate the iteration and obtain the optimal solution X of the objective function * (r (l) ), otherwise go to step S74.
[0095] S74, take r (l+1) =C·r (l) , X (0) =X * (r (l) ), and increase the number of iterations l by 1, and return to step S72.
[0096] The technical effects of the present invention are described in detail below in conjunction with specific embodiments.
[0097] In this embodiment, Figure 3 Take the Q355B I-beam shown in the figure as an example. The I-beam is 2.5m long, 0.5m high, with a flange thickness of 12mm and a web thickness of 8mm. It is fixed to the support platform by 8 bolts. The I-beam is made of Q355B steel, with an elastic modulus of 206GPa, a Poisson's ratio of 0.3, and a density of 7850kg / m 3The material of the constraint layer is aluminum, with an elastic modulus of 70 GPa, a Poisson's ratio of 0.3, and a density of 2700 kg / m 3 . Viscoelastic material density 1500kg / m 3 , Poisson's ratio 0.495.
[0098] (1) According to the actual situation of the steel bridge with constrained damping layer and the requirements for vibration and noise reduction, the design parameters are selected as follows: flange thickness T1, damping layer thickness T2, and constrained layer thickness T3. The target response is the structural fundamental frequency f0, the sound power level W, and the total mass m of the I-beam with constrained damping.
[0099] (2) Determine the number of test points and the design parameter combination of the test points through central composite design, n = 2 3 +3×2+1=15, The test points are distributed as follows Figure 4 As shown. The design parameters are converted into standard variables through linear transformation:
[0100]
[0101] (3) According to step S3, 15 groups of response data are obtained through numerical simulation as shown in Table 1.
[0102] Table 1
[0103]
[0104] (4) According to step S4, the fitting function relationship between the structural fundamental frequency, the sound power level, the total mass of the constrained damping I-beam and the design parameters is calculated by the response surface method:
[0105]
[0106]
[0107] m(x)=m s (x1)+m d (x2)+m c (x3) = ρ s A s (x1)L s +ρ d A d T2(x2)+ρ c A c T3(x3)
[0108] =215.8+30.12x1+1.742x2+1.568x3
[0109] Where m(x) is the total mass of the I-beam with the constrained damping layer under different design parameter combinations, ms (x1), m d (x2) and m c (x3) are the masses of the structural layer (steel beam), damping layer and constraint layer, respectively, ρ s , d and ρ c are the densities of the structural layer (steel beam), damping layer and constraint layer, A s (x1) and L s are the length and cross-sectional area of the structural layer (steel beam), A d and A c are the areas of the damping layer and the constraining layer, respectively; T2(x2) and T3(x3) are the thicknesses of the damping layer and the constraining layer, respectively.
[0110] (5) The optimization goal of this engineering embodiment is to minimize the sound power level. The optimization objective function f(x)=W(x) is established, and then the constraint condition is introduced:
[0111]
[0112] Where f0 = 9.87 Hz, m0 = 180.7 kg. The constraints are introduced into the optimization objective function in the form of penalty terms, written as form.
[0113] (6) Select the initial iteration point (0,0,0) that satisfies the constraints, set the initial value of the penalty factor to 1, the change coefficient to 0.5, and the iteration termination condition to:
[0114]
[0115] (7) According to the iterative method of step S7, the optimal parameter combination under the constraint conditions can be obtained. The optimization of each objective during the iterative process is as follows: Figure 5 shown.
[0116] In summary, the present invention obtains the fitting function of the target response and the design parameters through central composite design and response surface method, then introduces the constraint conditions, and uses the internal penalty function method to calculate the optimal design parameter combination of the constrained damping structure, which can provide technical support for the multi-constraint vibration and noise reduction parameter optimization design of the steel bridge structure with the constrained damping layer.
[0117] Those skilled in the art will appreciate that the embodiments described herein are intended to help readers understand the principles of the present invention, and should be understood that the protection scope of the present invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific variations and combinations that do not deviate from the essence of the present invention based on the technical revelations disclosed by the present invention, and these variations and combinations are still within the protection scope of the present invention.
Claims
1. A steel bridge constrained damping design optimization method based on response surface and penalty function method, characterized in that: The following steps are involved: S1. According to the actual situation of the steel bridge with the constrained damping layer and the requirements for vibration and noise reduction, the design parameters and target response of the constrained damping structure are selected; S2. Determine the number of test points and the design parameter combination of the test points through central composite design; S3. Obtain the target response value corresponding to each design parameter combination through experiments or numerical simulation methods; S4, obtaining the fitting relationship between the design parameters and the target response values by using the response surface method; S5. Establish an optimization objective function based on the fitting relationship, introduce constraints according to the actual situation of the steel bridge with the constrained damping layer and the requirements of vibration and noise reduction, and introduce the constraints into the objective function in the form of penalty terms, thereby transforming the constrained optimization problem into an unconstrained optimization problem; S6. Select an initial iteration point that satisfies the constraint conditions, determine the initial value and change coefficient of the penalty factor, and the iteration termination condition; S7. Use an unconstrained optimization method to iteratively calculate the extreme value of the objective function until the iteration termination condition is met.
2. The steel bridge constrained damping design optimization method based on response surface and penalty function method according to claim 1 is characterized in that: The design parameters in step S1 include the material, thickness, arrangement position and size of the original structure of the constrained damping structure, and the target response includes the sound power level and structural fundamental frequency of the constrained damping structure.
3. The steel bridge constrained damping design optimization method based on response surface and penalty function method according to claim 1 is characterized in that: The number of test points n in step S2 is: <h2 style=";text-align:left;direction:ltr">n=2<h2 style=";text-align:left;direction:ltr"> k <h2 style=";text-align:left;direction:ltr"> +2k+1 Where k represents the number of design parameters of the central composite design; The star arm length γ of the central composite design satisfies the following equation:
4. The steel bridge constrained damping design optimization method based on response surface and penalty function method according to claim 1 is characterized in that: The numerical simulation method in step S3 includes finite element method, boundary element method and statistical energy analysis method.
5. The steel bridge constrained damping design optimization method based on response surface and penalty function method according to claim 1 is characterized in that: The step S4 comprises the following sub-steps: S41. Let the influencing variable of a process be X = (x1, x2, ..., x k ), the response variable is y, and the second-order response surface model is constructed as follows: where x i represents the i-th influencing variable, x j represents the jth influencing variable, k represents the number of influencing variables, β0 represents the initial unknown coefficient, β i Influencing variable x i The unknown coefficient, β ij Influencing variable x i and x j The combined coefficient of unknown number, β ii Influencing variable x i The combined undetermined coefficients, ε represents the error between the fitted response and the true response, the influencing variable X corresponds to the design parameter, and the response variable y corresponds to the target response value; S42. Convert the second-order response surface model into a matrix form to obtain: Where n represents the number of trials, y a represents the target response value of the ath trial, x ia represents the i-th influencing variable of the a-th experiment, ε a represents the error between the fitted response and the true response of the a-th trial, a=1,2,...,n, i=1,2,...,k; S43. According to the principle of least squares method, the coefficient vector β to be determined should make the variance L of the error take the minimum value: Expand L to get: L=Y T Y-2β T X c T Y+β T X c T X c β Where ε represents the error vector, Y represents the target response vector, and X c Represents the second-order combination matrix of influencing variables; S44. Find the partial derivative of β and set it to 0: The fitting relationship between the design parameters and the target response value is obtained as follows: β=(X c T X c ) -1 (X c T Y)。 6. The steel bridge constrained damping design optimization method based on response surface and penalty function method according to claim 5 is characterized in that: The objective function in step S5 is: in represents the objective function, r (l) represents the penalty factor of the lth iteration, which is a decreasing positive sequence. f(X) represents the fitting relationship between the design parameters and the target response value. represents the penalty term, where g u (X)≥0 indicates the uth constraint, and m indicates the number of constraints.
7. The steel bridge constrained damping design optimization method based on response surface and penalty function method according to claim 6 is characterized in that: The initial iteration point X in step S6 (0) Satisfy all constraints and stay away from the boundary. The initial value of the penalty factor is r (0) The value ranges from 1 to 50, the change coefficient C ranges from 0.1 to 0.5, and the iteration termination condition is expressed as: where ε x and ε y All are allowable errors, X * Represents the extreme points solved using the unconstrained optimization method.
8. The steel bridge constrained damping design optimization method based on response surface and penalty function method according to claim 7 is characterized in that: The step S7 comprises the following sub-steps: S71, set the number of iterations l = 1; S72, from X (l-1) Starting from this, we use the unconstrained optimization method to solve the extreme value of the objective function Where Ω represents the value range of the influencing variable X; S73, determine whether the iteration termination condition is met, if so, terminate the iteration and obtain the optimal solution X of the objective function * (r (l) ), otherwise go to step S74; S74, take r (l+1) =C·r (l) , X (0) =X * (r (l) ), and increase the number of iterations l by 1, and return to step S72.
Citation Information
Patent Citations
Vibration radiated sound power layered optimization design method for constraint damping structure
CN108595789A
Sound-vibration response calculation method of constrained damping structure considering frequency-dependent characteristic
CN115130347A
Bridge reliability prediction method based on dynamic characteristics and intelligent algorithm response surface method
CN115270239A
Method for constructing composite response surfaces by combining neural networks with polynominal interpolation or estimation techniques
US7191161B1
