CN-116384273-B - Dynamic vibration reduction control method

CN116384273BCN 116384273 BCN116384273 BCN 116384273BCN-116384273-B

Abstract

The invention discloses a dynamic vibration damping control method, which comprises the steps of obtaining parameter combination data of vibration damping equipment, inputting the parameter combination data into computational fluid dynamics software to calculate and output result combination data, training by using a deep neural network to obtain a vibration loss function by taking the result combination data as an independent variable and loss expectation as a dependent variable, obtaining parameter combination data of minimum loss expectation by using an optimizing algorithm, constructing a vibration damping optimal control strategy function, calculating a difference value between the loss expectation and the minimum loss expectation to obtain expected benefits, obtaining work required to overcome external force when an appendage is switched from a common working condition to an optimal control working condition to obtain a control cost function, obtaining a combination with the maximum comprehensive evaluation function by adopting an exhaustion or group intelligent method, obtaining the optimal comprehensive control strategy function under different flow rates, deploying the optimal comprehensive control strategy function into a controller of the vibration damping equipment, and dynamically adjusting the relative position of the appendage according to different actually measured incoming flow rates.

Inventors

HU CHUAN
XIANG XINTAO
JIANG XIAOLIANG

Assignees

上海鹰达信息科技有限公司

Dates

Publication Date: 20260505
Application Date: 20230320

Claims (3)

1. A dynamic vibration reduction control method, characterized by comprising the steps of: Step 1, acquiring parameter combination data of vibration damping equipment, inputting the parameter combination data into computational fluid dynamics software to calculate and output result combination data, wherein the corresponding relation between the parameter combination data and the result combination data is vortex-induced vibration relation map, the parameter combination data comprises that the flow velocity of the vibration damping equipment is recorded as U, the distance between a structural object and an accessory is recorded as r, and the included angle between the structural object and the accessory is recorded as The result combination data comprises a vibration amplitude marked as A and a vibration frequency marked as f; And 2, training by using the result combination data as an independent variable and a loss expected dependent variable and using a deep neural network to obtain a vibration loss function: Or (b) Where E is the loss expectation, As a function of the number of functions, h is a function; step 3, under the condition of the same flow velocity, obtaining the parameter combination data expected by minimum loss by utilizing an optimizing algorithm, namely And then constructing a vibration reduction optimal control strategy function: wherein E min is the minimum loss expectation, Is the minimum distance between the structure and the accessory, The included angle between the minimum structure and the appendage, p is a function; Step 4, calculating the difference value between the loss expected E and the minimum loss expected E min under the condition of the same flow rate to obtain expected benefit which is marked as EA, Wherein g is a function; And 5, under the condition of the same flow rate, acquiring work required to overcome external force when the control system of the vibration damping device switches the appendage from the common working condition to the optimal control working condition as W, and obtaining a control cost function according to the work: wherein P is a function, PR is a control cost, K is a proportionality coefficient, the value range is determined according to the actual application scene, Wherein F is the force of the center of the damping device body, F * is the minimum force of the center of the damping device body, M is the moment around the center of the damping device body, and M * is the minimum moment of the center of the damping device body; Step 6, under the condition of the same flow velocity, obtaining a combination with the maximum S by using the comprehensive evaluation function by adopting an exhaustion or group intelligent method 、 I.e. The combination is For a certain flow rate The corresponding comprehensive optimal control strategy is that the comprehensive evaluation function is: ; and 7, repeating the steps 3-6 to obtain the optimal comprehensive control strategy function under different flow rates: wherein q is a function; and 8, deploying the optimal comprehensive control strategy function into a controller of the vibration damping equipment, and dynamically adjusting the relative position of the appendage according to different actually measured incoming flow rates.
2. A dynamic vibration reducer control method as set forth in claim 1, wherein the input in step 1 is denoted as in, The output is noted as out, 。
3. The dynamic vibration reducer control method according to claim 2, wherein the optimizing algorithm in step3 comprises an enumeration method or a swarm intelligence method.

Description

Dynamic vibration reduction control method Technical Field The invention relates to the field of dynamic vibration reduction control, in particular to a dynamic vibration reduction control method. Background The cable, pipeline and cable of large bridge of ocean platform are long and thin body structures and can cause fatigue damage because of vortex-induced vibration, can all cause huge maintenance cost each year, at present, the method for inhibiting vortex-induced vibration adopted for protecting related equipment is only suitable for a plurality of limited working conditions and is difficult to adapt to changeable actual working environment, so the protection effect on the related equipment is limited, therefore, an intelligent dynamic vibration reduction control method capable of adapting to various working conditions is needed, the vortex-induced vibration of long and thin body structures can be effectively inhibited under complex and changeable environments, the related equipment is practically protected, and the maintenance cost of the equipment is greatly reduced. Disclosure of Invention The invention aims to solve the technical problems that at present, a vortex-induced vibration suppression method adopted for protecting related equipment is only suitable for a plurality of limited working conditions and is difficult to adapt to changeable actual working environments. The invention provides the following technical scheme for solving the technical problems: A dynamic vibration reduction control method comprises the following steps: Step 1, acquiring parameter combination data of vibration damping equipment, inputting the parameter combination data into computational fluid dynamics (Computational Fluid Dynamics, CFD) software to calculate and output result combination data, wherein the corresponding relation between the parameter combination data and the result combination data is vortex-induced vibration relation map, the parameter combination data comprises that the flow speed of the vibration damping equipment is recorded as U, the distance between a structure and an appendage is recorded as r, the included angle between the structure and the appendage is recorded as theta, and the result combination data comprises that the vibration amplitude is recorded as A and the vibration frequency is recorded as f; And 2, training by using the result combination data as an independent variable and a loss expected dependent variable and using a deep neural network to obtain a vibration loss function: Or e=h (r, θ), where E is the loss expectation, As a function of the number of functions, h is a function; step 3, under the condition of the same flow rate, utilizing an optimizing algorithm to obtain parameter combination data of a minimum loss expectation, namely E min＝h(r*,θ*), and then constructing a vibration reduction optimal control strategy function (r *,θ*) =p (U), wherein E min is the minimum loss expectation, r * is the distance between a minimum structure and an appendage, and the included angle between a theta * minimum structure and the appendage is p is a function; Step 4, calculating the difference value between the loss expected E and the minimum loss expected E min under the condition of the same flow rate to obtain expected benefits, wherein the expected benefits are expressed as EA, EA=E-E min＝h(r,θ)-h(r*,θ*) =g (r, θ), and g is a function; Step 5, under the condition of the same flow rate, acquiring a control system of the vibration damping device, wherein work required by overcoming external force when an accessory is switched from a common working condition to an optimal control working condition is recorded as W, and a control cost function is obtained according to the work, wherein PR=KW (r, θ), P is a function, PR is a control cost, K is a proportionality coefficient, a value range is (the value range is required to be determined according to a scene of actual application), W (r, θ) =0.5 x (F-F *)*(r-r*)+0.5*(M-M*)*(θ-θ*), wherein F is force of the center of the main body of the vibration damping device, F * is minimum force of the center of the main body of the vibration damping device, M is moment around the center of the main body of the vibration damping device, and M * is minimum moment of the center of the main body of the vibration damping device; Step 6, under the condition of the same flow rate, adopting an exhaustion or group intelligent method to obtain a combination r best、θbest with a maximum comprehensive evaluation function of S, namely S max＝g(rbset,θbesr)-KW(rbset,θbesr), wherein the combination (r best,θbest) is a comprehensive optimal control strategy corresponding to a certain flow rate U *, and the comprehensive evaluation function is S=EA-PR=g (r, theta) -KW (r, theta); and 7, repeating the steps 3-6 to obtain the optimal comprehensive control strategy function under different flow rates: (r best,θbest) =q (U), where q is a function; and 8, deploying the optimal