Hypergravity centrifuge vibration control system with autonomous learning capability

Abstract

The present application relates to a kind of supergravity centrifuge vibration control systems with autonomous learning ability.The present application is applicable to building structure vibration control field.The technical problem to be solved by the present application is to provide a kind of supergravity centrifuge vibration control systems with autonomous learning ability.The technical scheme adopted by the present application is: a kind of supergravity centrifuge vibration control systems with autonomous learning ability, characterized in that, including: damping module, installed on building structure in the vibration influence range of supergravity centrifuge, with at least one variable stiffness damper;Monitoring module, installed on building structure in the vibration influence range of supergravity centrifuge, for monitoring the amplitude data of the building structure;Central processing module is connected with the damping module, monitoring module and supergravity centrifuge by data transmission line, for obtaining the operating parameter of supergravity centrifuge.

Description

TECHNICAL FIELD

[0001] The present application relates to a supergravity centrifuge vibration control system with autonomous learning ability. It is suitable for building structure vibration control field. BACKGROUND

[0002] Supergravity centrifuge is a high-speed rotating equipment, which can generate centrifugal acceleration exceeding gravity acceleration through high-speed rotation. Supergravity centrifuge is an important test platform for studying multiphase medium material science and engineering technology under supergravity conditions, and is an indispensable test means for studying large space-time evolution of rock-soil body and deep earth material and accelerating material phase separation. Supergravity centrifuge is called revolutionary engineering tool, which was first built in Columbia University, USA in 1931, and began to develop vigorously in 1960s. Its development shows the characteristics of larger and larger model scale of test object, more and more extreme on-board device, larger and larger capacity and centrifugal acceleration. The maximum design centrifugal acceleration of supergravity centrifuge simulation and experimental device of Zhejiang University has reached 1500g, which exceeds other existing supergravity centrifuges in the world. The vibration of the equipment will inevitably cause the vibration of the building structure, which is a practical engineering problem that must be solved.

[0003] Supergravity centrifuge is often installed underground, and the upper part is a supporting office and laboratory. Whether it is office or test, the specification has corresponding restrictions on the vibration from outside. The core evaluation parameter of centrifuge is the centrifugal acceleration of rotating equipment, and the acceleration is in quadratic relationship with the rotation speed, and the speed is proportional to the frequency. The greater the centrifugal acceleration of rotating equipment, the greater the vibration frequency range of the equipment rotation, which brings greater difficulty to the vibration control of the structure.

[0004] The existing control technology mainly includes three kinds, namely passive control, semi-active control and active control. Passive control technology uses TMD (tuned mass damper) or MTMD (tuned mass damper) and other pre-installed damping devices, which can only reduce the vibration of a certain frequency or several frequencies; semi-active control technology can change the stiffness or damping characteristics of the damper according to the vibration source frequency, so as to realize vibration control in a certain frequency range; active control technology uses actuators to actively apply actuating force to offset the vibration brought by vibration source according to the actual structure vibration observed by the monitoring equipment.

[0005] Theoretically, semi-active control and active control can achieve better vibration control, but there are also problems in application. For example, although the vibration frequency of the centrifuge is known, when the vibration is transmitted to the upper structure through the foundation, the soil and possibly the vibration damping device of the centrifuge itself, the vibration frequency has changed. In addition, due to differences in construction and design, the material may also change its own frequency as the service life increases; the existing semi-active and active control based on structural analysis in the design stage cannot adapt to changes in the actual use stage. Active control is less used in practical engineering because the actuator requires a large amount of energy input and the entire system has a time delay effect. SUMMARY

[0006] The technical problem to be solved by the present application is to provide a supergravity centrifuge vibration control system with self-learning ability to solve the above problems.

[0007] The technical solution adopted by the present application is: a supergravity centrifuge vibration control system with self-learning ability, characterized in that it comprises:

[0008] A damping module is installed on the building structure within the vibration influence range of the supergravity centrifuge and has at least one variable stiffness damper.

[0009] A monitoring module is installed on the building structure within the vibration influence range of the supergravity centrifuge and is used to monitor and obtain the amplitude data of the building structure.

[0010] A central processing module is connected to the damping module, the monitoring module and the supergravity centrifuge through a data transmission line and is used to obtain the operating parameters of the supergravity centrifuge, retrieve the stiffness parameters of each variable stiffness damper in the damping module pre-stored and obtained through optimized training based on the operating parameters, and control and change the stiffness of each variable stiffness damper based on the stiffness parameters of each variable stiffness damper.

[0011] The stiffness parameters of each variable stiffness damper in the damping module obtained through optimized training include:

[0012] The operating parameters of the supergravity centrifuge when it is running are obtained, and the amplitude data of the supergravity centrifuge when it is running under the operating parameters are obtained by the monitoring module, and the stiffness parameters of each variable stiffness damper in the damping module are optimized with the minimum maximum amplitude in the amplitude data as the optimization target.

[0013] The stiffness parameters of each variable stiffness damper in the damping module are optimized with the minimum maximum amplitude in the amplitude data as the optimization target, which includes using genetic algorithm, simulated annealing method or enumeration method for optimization.

[0014] The optimization using genetic algorithm includes:

[0015] S1, generate several individuals to constitute an initial population, and each chromosome in the individual corresponds to the stiffness parameter of each variable stiffness damper in the damping module;

[0016] S2, control the stiffness of each variable stiffness damper in the damping module based on the stiffness parameter of each variable stiffness damper in the individual;

[0017] S3, after the stiffness of each variable stiffness damper in the damping module is controlled in step S2, the amplitude data of the building structure corresponding to each individual under the same operating parameters of the super gravity centrifuge is monitored and obtained through the monitoring module;

[0018] S4, the individual whose amplitude data meets the preset condition is retained, and genetic operation is performed on the retained individual to form a new generation population;

[0019] S5, steps S2-S4 are repeated based on the individual in the new generation population until the genetic iteration requirement is met, and the final population is obtained;

[0020] S6, the individual with the smallest maximum amplitude in the corresponding building structure amplitude data in the final population is taken as the optimal individual under the corresponding operating parameters of the super gravity centrifuge.

[0021] The central processing module is further configured to:

[0022] After the stiffness of each variable stiffness damper is controlled, the amplitude data of the building structure is monitored and obtained through the monitoring module:

[0023] Based on the comparison between the monitored amplitude data and the original amplitude data under the same operating parameters and the same stiffness parameters, and when the difference between the two exceeds the preset value, the optimization training of the stiffness parameters of each variable stiffness damper in the damping module under the operating parameters is re-performed.

[0024] Each variable stiffness damper in the damping module is installed at each position where the structural stiffness can be changed on the building structure.

[0025] The monitoring module comprises at least one three-way vibration displacement meter.

[0026] The three-way vibration displacement meter is arranged at the intersection of the beam and the column of the building structure.

[0027] The beneficial effects of the present application are: the damping module, the monitoring module and the central processing module cooperate to perform optimization training autonomously, and the optimal control scheme of each variable stiffness damper in the damping module corresponding to each operating parameter of the super gravity centrifuge is obtained through training, so that the optimal control scheme of each variable stiffness damper in the damping module can be recalled based on the operating parameters of the super gravity centrifuge in the subsequent operation process, and automatic vibration control is realized.

[0028] The application judges whether the natural vibration frequency of the building structure changes based on amplitude data under the same operating parameters of the super gravity centrifuge and the same control scheme of the variable stiffness damper, and re-optimizes training after the natural vibration frequency changes to obtain the optimal control scheme of each variable stiffness damper in the damping module, thereby improving the problem that the effect of the previous optimal control scheme is weakened due to factors such as stiffness degradation. BRIEF DESCRIPTION OF DRAWINGS

[0029] Figure 1 The system block diagram of the vibration control system in the embodiment.

[0030] Figure 2 、 3 The overall arrangement schematic diagram of the super gravity centrifuge, the building structure and the vibration control system in the embodiment.

[0031] 1, monitoring module; 2, central processing module; 3, damping module; 5, soil body; 6, super gravity centrifuge; 7, centrifuge foundation; 11, three-direction vibration displacement meter; 12, data transmission line; 21, data analysis module; 22, storage module; 31, variable stiffness damper; 41, column; 42, roof; 43, diagonal brace; 44, beam. DETAILED DESCRIPTION

[0032] The embodiment is a super gravity centrifuge vibration control system with autonomous learning ability, which comprises a damping module, a monitoring module and a central processing module, wherein the damping module and the monitoring module are arranged on the building structure in the vibration influence range of the super gravity centrifuge, and the building structure comprises a column, a roof, a diagonal brace and a beam.

[0033] The damping module in the embodiment has at least one variable stiffness damper, which is respectively installed at each position on the building structure in the vibration influence range of the super gravity centrifuge, which can obviously change the stiffness of the building structure.

[0034] The monitoring module in the embodiment comprises at least one three-direction vibration displacement meter, which is installed at a position with large amplitude of the building structure obtained through structural analysis, and is used to obtain the amplitude data of the building structure. In the embodiment, the three-direction vibration displacement meter is installed at the intersection of the beam and the column of the building structure.

[0035] The central processing module in the embodiment is a PC or desktop computer that can perform data analysis and processing, and comprises a data analysis module and a storage module. The central processing module is connected with each variable stiffness damper in the damping module, each three-direction vibration displacement meter in the monitoring module and the super gravity centrifuge through a data transmission line.

[0036] The stiffness control scheme of each variable stiffness damper in the damping module corresponding to different operating parameters (different operating parameters correspond to different working conditions) of the supergravity centrifuge is stored in the storage module of the central processing module, and the stiffness control scheme of each variable stiffness damper corresponding to different operating parameters is obtained by optimized training of the vibration control system.

[0037] The optimization training method of the vibration control system in the embodiment includes: obtaining the operating parameters of the supergravity centrifuge during operation, and the amplitude data of the supergravity centrifuge obtained by the monitoring module during operation under the operating parameters, and taking the minimum maximum amplitude in the amplitude data as the optimization target, and using genetic algorithm, simulated annealing method, enumeration method and other algorithms to optimize the stiffness parameters of each variable stiffness damper in the damping module.

[0038] In this example, the stiffness parameters of each variable stiffness damper in the damping module are optimized by using genetic algorithm, including:

[0039] S1, randomly generate a plurality of individuals to form an initial population, and each chromosome in the individual corresponds to the stiffness parameter of each variable stiffness damper in the damping module; the stiffness parameter of each variable stiffness damper in the individual belongs to the stiffness range of the variable stiffness damper;

[0040] S2, based on the stiffness parameters of each variable stiffness damper in the individual, the stiffness of each variable stiffness damper in the damping module is changed;

[0041] S3, after the stiffness of each variable stiffness damper in the damping module is changed in step S2, the amplitude data of the building structure corresponding to each individual under the same operating parameters of the supergravity centrifuge is obtained by the monitoring module;

[0042] S4, the individuals whose amplitude data meet the preset condition are retained, and genetic operation is performed on the retained individuals to form a new generation population; the preset condition can be that 30% of the individuals with the minimum maximum amplitude in the retained population are retained;

[0043] S5, based on the individuals in the new generation population, steps S2-S4 are repeated until the genetic iteration requirement is met, and the final population is obtained;

[0044] S6, the individual with the minimum maximum amplitude in the corresponding building structure amplitude data in the final population is taken as the optimal individual under the corresponding operating parameters of the supergravity centrifuge, that is, the optimal control scheme of each variable stiffness damper in the damping module, and is saved to the storage module of the central processing module.

[0045] In this embodiment, the stiffness parameters of each variable stiffness damper in the damping module are optimized by using enumeration method, including:

[0046] The stiffness that each variable stiffness damper can reach is divided into 21 discrete stiffness values in equal difference according to the stiffness range of each variable stiffness damper;

[0047] Randomly assign values to each variable stiffness damper in the vibration reduction module based on 21 discrete stiffness values, form multiple control schemes, each control scheme contains the stiffness parameters of each variable stiffness damper in the vibration reduction module;

[0048] Each working condition of the super gravity centrifuge lasts for 1050s, within 1000s after the centrifuge stabilizes, every 10 / f time (f is the minimum centrifuge frequency under this working condition) is a period, before the start of each period, control according to one of the multiple control schemes, at the end of the period, record the maximum vibration amplitude;

[0049] Stop running when the maximum optimization target is reached or the specified running time is reached, training is complete. The storage module stores the optimal control scheme under each working condition.

[0050] In this embodiment, the central processing module monitors the building structure through the monitoring module for a long time, and determines whether the structural material causes the natural frequency of the structure to change due to degradation through the judgment statement in the central processing module 2.

[0051] In this example, when the vibration amplitude under the same working condition and the control scheme of the same variable stiffness damper exceeds 10% of the original vibration amplitude, it is considered that the natural frequency of the structure has changed. At this time, the working condition of the same centrifuge frequency is considered as a new working condition, the system is optimized again, a new optimal control scheme is found, and it is stored, so as to improve the problem that the effect of the previous optimization control scheme is weakened due to the stiffness degradation and other factors.

Claims

1. A high gravity centrifuge vibration control system with autonomous learning capability, characterized in that, The application relates to a vibration damping system for a high-gravity centrifuge, comprising: a damping module installed on a building structure within a vibration influence range of the high-gravity centrifuge, and comprising at least one variable stiffness damper; a monitoring module installed on the building structure within the vibration influence range of the high-gravity centrifuge, and used for monitoring and obtaining amplitude data of the building structure; a central processing module connected to the damping module, the monitoring module and the high-gravity centrifuge through a data transmission line, and used for obtaining operation parameters of the high-gravity centrifuge, calling pre-stored stiffness parameters of the variable stiffness dampers in the damping module based on the operation parameters, and controlling and changing the stiffness of the variable stiffness dampers based on the stiffness parameters of the variable stiffness dampers. The pre-stored stiffness parameters of the variable stiffness dampers in the damping module are obtained by: obtaining operation parameters of the high-gravity centrifuge, and amplitude data of the high-gravity centrifuge obtained by the monitoring module when the high-gravity centrifuge operates under the operation parameters, and optimizing the stiffness parameters of the variable stiffness dampers in the damping module with the minimum maximum amplitude in the amplitude data as the optimization target.

2. The supercentrifuge vibration control system with autonomous learning capability according to claim 1, wherein, The optimization of the stiffness parameters of the variable stiffness dampers in the damping module with the minimum maximum amplitude in the amplitude data as the optimization target comprises:

3. The supercentrifuge vibration control system with autonomous learning capability according to claim 2, wherein, using a genetic algorithm, a simulated annealing method or an enumeration method to optimize. The optimization using the genetic algorithm comprises: S1, generating a plurality of individuals to form an initial population, and each chromosome in the individual corresponds to the stiffness parameter of each variable stiffness damper in the damping module; S2, controlling and changing the stiffness of the variable stiffness dampers in the damping module based on the stiffness parameters of the variable stiffness dampers in the individual; S3, after the stiffness of the variable stiffness dampers in the damping module is controlled and changed in step S2, monitoring and obtaining amplitude data of the building structure corresponding to each individual under the same operation parameters of the high-gravity centrifuge through the monitoring module; S4, retaining the individual whose amplitude data meet a preset condition, and performing genetic operation on the retained individual to form a new generation population; S5, repeating steps S2-S4 based on the individual in the new generation population until a genetic iteration requirement is met, and obtaining a final population; 4. The supercentrifuge vibration control system with autonomous learning capability according to claim 1, wherein S6, taking the individual corresponding to the building structure whose maximum amplitude is the smallest in the amplitude data in the final population as the optimal individual under the corresponding operation parameters of the high-gravity centrifuge. The central processing module is further used for: monitoring and obtaining amplitude data of the building structure through the monitoring module after the stiffness of the variable stiffness dampers is controlled and changed; 5. The supercentrifuge vibration control system with autonomous learning capability according to claim 1, characterized in that: comparing the monitored amplitude data with original amplitude data under the same operation parameters and the same stiffness parameters, and re-performing optimization training of the stiffness parameters of the variable stiffness dampers in the damping module under the operation parameters when the difference between the two exceeds a preset value.

6. The supercentrifuge vibration control system with autonomous learning capability of claim 1, wherein: Each variable stiffness damper in the damping module is installed at a position where the structure stiffness can be changed on the building structure.

7. The supercentrifuge vibration control system with autonomous learning capability according to claim 6, characterized in that: The monitoring module comprises at least one three-way vibration displacement meter. The three-way vibration displacement meter is arranged at the intersection of a beam and a column of the building structure.

Images