Vibration method, design method, vibration device, program

Abstract

We propose a simpler vibration excitation method that differs from conventional methods. [Solution] A vibration excitation method using automatic control, wherein the automatic control is a control to match the output signal of the vibrator that drives the object to a reference signal, and the excitation is performed by a vibration excitation method that does not use the characteristics of the vibrator when the object is installed.

Description

[Technical Field]

[0001] This invention relates to a vibration excitation method using automatic control, etc. [Background technology]

[0002] For example, when conducting vibration tests by controlling a vibration table used in seismic tests or road simulators, if the test specimen placed on the vibration table has nonlinear characteristics, it is necessary to control the vibration table while considering the effects of the nonlinear characteristics of the test specimen. Therefore, various control methods have been studied. For example, Patent Document 1 discloses a method for controlling a vibration table on which a test specimen with nonlinear characteristics is placed, based on a model called NSBC (Nonlinear Signal Based Control). [Prior art documents] [Patent Documents]

[0003] [Patent Document 1] Patent No. 7287660 [Overview of the Initiative] [Problems that the invention aims to solve]

[0004] Beyond the vibration tests described above, generally speaking, when an object such as a structure (the object to be vibrated) is placed in a vibrator, the vibrator may not function as intended. Therefore, there is a demand for a way to ensure that the vibrator functions as intended even when the object is placed in it.

[0005] One of the objectives of this invention is to propose a new vibration excitation method that differs from conventional methods. [Means for solving the problem]

[0006] According to a first aspect of the present invention, there is a vibration excitation method using automatic control, wherein the automatic control is a control for matching the output signal of a vibrator that drives an object with a reference signal, and the vibration excitation method does not use the characteristics of the vibrator when the object is installed. According to a second aspect of the present invention, a vibration exciter using automatic control is configured such that the automatic control is a control that matches the output signal of a vibrator that drives an object to a reference signal, and the vibration exciter does not use the characteristics of the vibrator when the object is installed. According to a third aspect of the present invention, a program for realizing vibration using automatic control is provided, wherein the automatic control is a control to match the output signal of a vibrator that drives an object to a reference signal, and the program may be configured to cause a computer to execute a control to realize vibration that does not use the characteristics of the vibrator when the object is installed. [Effects of the Invention]

[0007] According to the present invention, it becomes possible to achieve simple and sufficiently accurate vibration excitation using automatic control. [Brief explanation of the drawing]

[0008] [Figure 1] This figure shows an example configuration of a vibration testing system according to the embodiment. [Figure 2] A diagram illustrating a vibrator according to an embodiment. [Figure 3] A block diagram showing an example of a vibration table control device according to the embodiment. [Figure 4] (a) shows the test specimen not placed on the table, and (b) shows the test specimen placed on the table. [Figure 5] A diagram showing an example of the functional configuration of the computer section of a vibration table control system. [Figure 6] A diagram illustrating PID control using composite filtering technology. [Figure 7](a) is a figure showing the time history of a conventional shaking table experiment, (b) is a figure showing the FFT (Fourier amplitude spectrum) of a conventional shaking table experiment, and (c) is a figure showing the hysteresis characteristics of the test specimen in a conventional shaking table experiment. [Figure 8] (a) is a figure showing the time history of the vibration table experiment using the method of the embodiment, (b) is a figure showing the FFT (Fourier amplitude spectrum) of the vibration table experiment using the method of the embodiment, and (c) is a figure showing the hysteresis characteristics of the test specimen in the vibration table experiment using the method of the embodiment. [Modes for carrying out the invention]

[0009] Hereinafter, an example of an embodiment to which the present invention is applied will be described with reference to the attached drawings. However, the embodiments to which the present invention can be applied are not limited to those described below. In this embodiment, as an embodiment in which a method of vibrating an object via a vibration table is applied, an embodiment of performing a vibration test on a test specimen having nonlinear characteristics will be described.

[0010] Figure 1 shows an example of the configuration of the vibration testing system 1 in this embodiment.

[0011] The vibration testing system 1 comprises, for example, a vibration table 3 on which a test specimen 2 can be placed, and a vibration table control device 10 that controls the vibration state of the vibration table 3 (e.g., displacement, velocity, acceleration, etc.). The vibration testing system performs vibration testing on the test specimen 2 placed on the vibration table 3 by controlling the vibration state of the vibration table 3 with the vibration table control device 10.

[0012] Test specimen 2 has nonlinear characteristics and may be, for example, a structure such as a building or vehicle, or a structure that simulates them. Other test specimens with nonlinear characteristics may also be used. For example, the plastic deformation of the structure occurs in the test specimen 2 due to the excitation of the vibration table 3, and the nonlinear behavior of the test specimen 2 may affect the control performance of the vibration table 3 by the vibration table control device 10.

[0013] The vibration table 3 comprises, for example, a base 30, a table 32 supported on the base 30 so as to be movable in the horizontal direction via a bearing 31, an actuator 33 that drives the connected table 32 in the horizontal direction, and a detection sensor 34 that detects the vibration state of the table 32.

[0014] In this embodiment, the actuator 33 is described as being a single-axis actuator that drives the table 32 in the horizontal direction, but it may also be composed of a multi-axis actuator (for example, three axes: X-axis, Y-axis, and Z-axis) that drives the table 32 in both the horizontal and vertical directions.

[0015] The vibration table control device 10 (an example of an excitation device) is a device that controls the vibration table 3 by generating an operation signal u for the actuator 33 so that the output signal y, which indicates the vibration state of the table 32 on which the test specimen 2 with nonlinear characteristics is placed, matches the input signal r (an example of a reference signal) which indicates the target vibration waveform of the vibration table 3. In other words, the vibration table control device 10 drives the actuator 33 based on the operation signal u so that the output signal y matches the input signal r, and controls the vibration table 3 by sequentially updating the operation signal u based on the output signal y detected by the detection sensor 34 when the actuator 33 is driven.

[0016] As described above, the output signal y is an output signal indicating the vibration state of the table 32 on which the test specimen 2, which has nonlinear characteristics, is placed. The output signals of the table 32, such as displacement, velocity, and acceleration, become the output signals of the control system 100.

[0017] In the following, the displacement of table 32 is defined as "y d (=y0) is denoted as "y d It can be expressed as the first time derivative and the second time derivative of .

[0018] In this embodiment, the test specimen 2 (an example of an object) is vibrated via the vibration table 3 (more precisely, the table 32 as a movable part driven by the actuator 33; sometimes referred to as the "vibration table"). As will be explained in detail below, considering that the characteristics of the vibration table 3 change when the test specimen 2 is placed on it, in this embodiment, the test specimen 2 is vibrated without using the characteristics of the vibration table 3 in the state in which the test specimen 2 is placed. This is one example of an excitation method that does not use the characteristics of the exciter in the state in which the object is placed.

[0019] As mentioned above, when an object such as a structure is installed (including being placed on) the vibrator, the vibrator may not function as intended. Therefore, there is a demand for the vibrator to function as intended even when an object is installed on it. Specifically, the characteristics of the vibrator change when an object is installed on it, and the characteristics of the vibrator also change when the installed object is damaged and its characteristics change. For this reason, there is a need to control the vibrator with high precision without using the characteristics of the vibrator when an object is installed on it.

[0020] The characteristics of the vibrator can be defined as, for example, the input-output relationship of the vibrator, and not using the characteristics of the vibrator with the object installed can be defined as, for example, not using the input-output relationship of the vibrator with the object installed. The input-output relationship of the vibrator can be expressed as, for example, a transfer function, and not using the input-output relationship of the vibrator with the object installed can be defined as, for example, not using the transfer function of the vibrator with the object installed. However, in the case of a system (control system 100) having nonlinearity as in this embodiment, it cannot be strictly expressed as a transfer function, so it may include not using the input-output relationship corresponding to the transfer function of the vibrator with the object installed.

[0021] In this embodiment, when the test specimen 2 is placed on the vibration table 3, the characteristics of the vibration table 3 (input / output relationship: the relationship between the operation signal u for the actuator 33 and the output signal y indicating the vibration state of the table 32) change. More specifically, when the test specimen 2 is placed on the vibration table 3, the characteristics (input / output relationship) of the actuator 33 change, and the characteristics (input / output relationship) of the table 32 also change. Therefore, it is required to control the vibration table 3 with high precision without using the characteristics of the vibration table 3 in the state in which the test specimen 2 is placed. In this embodiment, for example, as shown in Figure 2(a), the vibrator (a vibrator that drives the object) is a vibration table 3, and the test object 2 is vibrated without using the characteristics (input / output relationship) of the vibration table 3 with the test object 2 placed on it. In a broad sense, the vibrator may include a vibration table or the like, to which accessories such as movable parts (e.g., a table 32) are attached to an actuator 33 or other device generally called a vibrator or drive device. It can also be considered that the "actuator 33 + table 32" (driver + movable part) functions as a vibrator, or that the table 32 (movable part) functions as a vibrator.

[0022] Furthermore, as mentioned above, the vibration table control device 10 controls the vibration table 3 by generating an operation signal u for the actuator 33 so that the output signal y, which indicates the vibration state of the table 32 on which the test specimen 2 with nonlinear characteristics is placed, matches the input signal r (an example of a reference signal) which indicates the target vibration waveform of the vibration table 3. In other words, the vibration table control device 10 does not monitor the output signal of the actuator 33 itself. However, since the input signal r can also be considered a signal of the target operation of the actuator 33, the vibration table control device 10 generating the operation signal u for the actuator 33 can also be considered a control to match the output signal of the actuator 33 to the reference signal. Therefore, as shown in Figure 2(b), for example, devices generally called vibrators or drive devices, such as the actuator 33, may also be included in the vibrator of the present invention.

[0023] In any case, the excitation method of this embodiment is characterized by not using the characteristics (input / output relationship) of the exciter in the state in which the test specimen 2 is installed.

[0024] The vibration table control device 10 is composed of, for example, a general-purpose computer, and its specific configuration includes, for example, an input unit composed of a keyboard, mouse, touch panel, etc., a storage unit composed of an HDD, memory, etc., a control unit composed of a processor such as a CPU (Central Processing Unit), a connection unit (which may include a communication unit) that connects to the actuators 33 and detection sensors 34 of the vibration table 3 and to an external network, etc., and a display unit composed of a display, etc. It is not necessary to have all of these, and it is also acceptable if some of them are not provided.

[0025] The target vibration waveform for the vibration table 3 may be, for example, a waveform that reproduces vibrations caused by past earthquakes when the test specimen 2 is a building. Alternatively, for example, when the test specimen 2 is a vehicle, the waveform may reproduce vibrations when the vehicle travels on rough roads. The input signal r representing the vibration waveform may be input, for example, via the input unit, or as vibration waveform data stored in the memory unit or vibration waveform data acquired from an external network (vibration waveform data received by the communication unit).

[0026] Figure 3 is a block diagram showing an example of the vibration table control device 10 according to this embodiment. The vibration table control device 10 is a controller that controls the output signal y(s) to match the input signal r(s), for example, a feedforward controller 11(K) that performs feedforward control. r (s)), a first feedback controller 12(K) performs first feedback control. σ (s)), and a second feedback controller 13(K) that performs second feedback control. e (s)) is provided below, K r "K" σ "K"e The "" is appropriately referred to as ""gain"" (control gain). The shaking table control device 10 includes, for example, an adder 14, a first subtractor 15, a second subtractor 16, and a stability evaluation unit 17.

[0027] As shown in the following formula (1), the adder 14 adds the output of the feedforward controller 11(K r (s)), the output of the first feedback controller 12(K σ (s)), and the output of the second feedback controller 13(K e (s)) to output an operation signal u(s).

[0028]

Number

[0029] As shown in the following formula (2), when the operation signal u(s) is input to the control system 100 through the dead time element 102(e -τs ) due to the dead time τ, the first subtractor 15 subtracts the output signal y^(s) of the linear model 101 for the control system 100 when the operation signal u(s) is input to the linear model 101 for the control system 100 through the estimated dead time element 103(e -τ^s ) based on the estimated value τ^ of the dead time τ from the output signal y(s) of the control system 100, and outputs a first deviation signal σ(s).

[0030]

Number

[0031] As shown in the following formula (3), the second subtractor 16 subtracts the output signal y(s) of the control system 100 when the operation signal u(s) is input to the control system 100 through the dead time element 102(e -τs ) from the input signal r(s), and outputs a second deviation signal e(s).

[0032]

Number

[0033] Feedforward controller 11(K r (s)) performs feedforward control by using the inverse transfer function for the transfer function G^(s) of the linear model 101 described later, with respect to the input signal r(s).

[0034] First feedback controller 12(K σ (s)) performs the first feedback control on the first deviation signal σ(s).

[0035] In the following, the vibration table control device 10 controls the first feedback controller 12(K σ The provision of (s)) or the application of a first feedback control to the first deviation signal σ(s) is referred to as Nonlinear Signal Based Control (NSBC).

[0036] Second feedback controller 13(K e (s)) performs a second feedback control on the second deviation signal e(s).

[0037] Here, when applying NSBC to the control of the vibration table 3, the operation signal u(s) to the control system 100 is obtained as shown in equation (1) above by considering the test specimen 2 and the vibration table 3 as a single control target. The second deviation signal e(s) is obtained by equation (3) above, and the feedforward controller 11(K) that achieves zero error in equation (3) r (s)), first feedback controller 12(K σ (s)), second feedback controller 13(K e For (s), a controller that satisfies equation (4) below is suitable. This corresponds to designing the controller so that the second deviation signal e(s) in equation (3) above becomes zero when the dead time τ and its estimated value τ^ are equal to zero (τ=τ^=0).

[0038]

number

[0039] In the above equation (4), the feedforward controller 11(K r (s)) is a controller that uses the inverse transfer function for the transfer function G^(s) of the linear model 101.

[0040] For convenience, while mathematical formulas use an upperline (overline) above the letters, the specification will use a superscript "^".

[0041] The first feedback controller 12(Kσ(s)) controls the control system 100, which has nonlinear characteristics, by performing feedback control with the first deviation signal σ(s), which is a nonlinear signal, as input (feedback input).

[0042] Second feedback controller 13(K e (s)) is the first feedback controller 12(K) which functions as an NSBC. σ (s)) plays an auxiliary role. Therefore, the second feedback controller 13(K e By properly designing (s), the second deviation signal e is forcibly reduced, thereby further improving control accuracy. Furthermore, the first feedback controller 12(K σ If sufficient control accuracy can be achieved by (s), the second feedback controller 13(K e (s)) can be omitted.

[0043] The stability evaluation unit 17 applies Nyquist's stability theory based on the transfer function of the linear model 101 and the dynamic characteristics of the control system 100 that are not modeled by the linear model 101, thereby determining the stability of the first feedback controller 12(K σ Evaluate the stability of (s). First feedback controller 12(K σThe stability of (s)) may be evaluated in accordance with the method disclosed in Japanese Patent No. 7287660, which is incorporated herein by reference.

[0044] (Methods for obtaining transfer functions of linear models and designing controllers)

[0045] (Conventional method) First, as an example of a conventional method, we will explain the method disclosed in Japanese Patent Publication No. 7287660. Japanese Patent Publication No. 7287660 is incorporated herein by reference.

[0046] Figure 4 shows the test specimen 2 and vibration table 3 according to this embodiment, with (a) showing the state without the test specimen 2 and (b) showing the state with the test specimen 2 placed on it.

[0047] First, as a model of the vibration table 3, as shown in Figure 4(a), when the test specimen 2 is not placed on it, the dynamic characteristics of the vibration table 3 under displacement control are expressed as a second-order transfer function shown in equation (5) below.

[0048]

number

[0049] Equation (5) above is equivalent to equation (6) below in the time domain.

[0050]

number

[0051] Equation (6) above is equivalent to the equation of motion for a linear single-mass system. When this system alone is used for excitation, the input signal r, which represents the target vibration displacement waveform, becomes the control signal u.

[0052] As an extension of the above, when the shaking table 3 supports the test specimen 2, which is a nonlinear multi-mass point system with N degrees of freedom, as shown in Figure 4(b), its equation of motion is expressed by equation (7) below.

[0053]

number

[0054] If test specimen 2 exhibits linearity, its governing equation is given by equation (8) below.

[0055]

number

[0056] Furthermore, when the vibration table 3 supports a linear test specimen 2, the operation signal u and the i-th level output signal y are used. i The relationship between them is expressed by equation (9) below.

[0057]

number

[0058] However, "i" is an index representing the degrees of freedom of test specimen 2, and if test specimen 2 is a hierarchical structure, it may represent the hierarchy number "i". For example, if test specimen 2 is an N-hierarchical structure, it may be written as "i=1,2,···,N".

[0059] From equations (8) and (9) above, the i-th order transfer function G i (s) is expressed by the following equation (10).

[0060]

number

[0061] Equations (8) to (9) above are valid when the test specimen 2 has linearity, but they can be applied to the design of the linear model 101 which is essential in NSBC. That is, the linear model 101 of the control system 100 is expressed by the following equation (11).

[0062]

number

[0063] Then, in the linear model 101, the output signal y of the i-th order i And the i-1th output signal y i-1 The transfer function between them is given by equation (12) below.

[0064]

number

[0065] Based on equation (12), "G^ N (s), G^ N-1 By calculating (s), ..., G^0(s) in this order, the dynamic characteristics of the vibration table 3 with the nonlinear test specimen 2 placed on it, that is, the transfer function "G^(s)=G^0(s)" of the linear model 101 for the control system 100 is obtained. Therefore, based on the equation (4) described above, the feedforward controller 11(K r (s)) and the first feedback controller 12(K σ (s)) can be designed.

[0066] For example, for the displacement and acceleration of Table 32, the transfer function of the linear model 101 is expressed as shown in equation (13) below.

[0067]

number

[0068] In vibration table experiments, acceleration control is primarily important. This is because the object is excited by the inertial force generated on the table 32. When NSBC is applied to its acceleration control, the control signal u(s) is expressed by the following equation (14).

number

[0069] Here, the second derivative of time, "y d " is the output signal of the acceleration in table 32 (table acceleration signal), and the second derivative with respect to time is "y^ d This is the acceleration signal corresponding to the Table 32 portion of the linear model 101. In this case, the input signal "r" is the acceleration signal of Table 32 that we aim to reproduce.

[0070] Based on the above equation (13), G^(s) = G^ a As (s), the controller is designed as shown in equation (15) below.

number

[0071] (Method of this embodiment) As mentioned above, the characteristics of the vibration table 3 (including the input-output relationship of the vibration table 3 and the dynamic characteristics of the vibration table 3 (more precisely, the dynamic characteristics related to the output (displacement, velocity, acceleration) of the table 32)) change depending on the test specimen 2 placed on the vibration table 3. In this case, the larger the mass of the placed test specimen 2, the greater the effect. Furthermore, if the test specimen 2 is damaged by vibration, this also changes the characteristics of the vibration table 3, making it difficult to control the vibration table 3.

[0072] Therefore, one could consider, for example, conducting a system identification test in advance to obtain the characteristics of the test specimen 2 to be placed on the table (characteristics here refer to physical properties including mechanical properties such as robustness, hardness, and toughness), and then understanding the characteristics of the vibration table 3 when the test specimen 2 is placed on the vibration table 3. However, in this case, conducting a system identification test is essential, and if the test specimen 2 is damaged by vibration, the characteristics of the vibration table 3 will change, making it difficult to control the vibration table 3.

[0073] Therefore, the inventors of this application devised a method for designing a controller using the transfer function of a simple linear model 101. Specifically, they considered designing a controller using the transfer function of the linear model 101, represented by equation (16) below, as G^(s).

[0074]

number

[0075] The upper equation represents the transfer function corresponding to the displacement of table 32, and the lower equation represents the transfer function corresponding to the acceleration of table 32. The transfer function corresponding to the velocity of table 32 can be formulated similarly. These transfer functions are collectively denoted as "G^" or "G^(s)".

[0076] Furthermore, {ω^0, ζ^0} in equation (16) is given by equation (17) below.

[0077]

number

[0078] "m0" is the mass of table 32, and "ω0" is the natural frequency of table 32 (when nothing is placed on it). "m sp " is the mass of test specimen 2. "ρ" is the tuning coefficient.

[0079] Furthermore, if test specimen 2 is an N-level structure, then as mentioned above, "i=1,2,···,N" is acceptable, and "m sp This is represented by the sum of the masses at each level.

[0080] The transfer function G^(s) expressed by equation (16) is a function based on the characteristics of the vibration table 3 in an empty state, and takes into account the effect that the placed test specimen 2 has on the vibration table 3. Specifically, the mass "m sp This function takes into account that when test specimen 2 is placed on the table, the natural frequency "ω0" of table 32 changes to "ω^0" (ω0→ω^0).

[0081] In this case, the mass "m0" and natural frequency "ω0" of table 32, and the mass "m" of test specimen 2. sp Based on this, the natural frequency "ω^0" of the table 32 with the test specimen 2 placed on it can be determined, and the corresponding damping ratio "ζ^0" can be fixed to a constant value (for example, "1") along with the tuning coefficient "ρ" to obtain the transfer function G^(s).

[0082] The tuning coefficient "ρ" is a parameter whose value can be adjusted, and for example, it may be set to "1" (ρ=1). However, it is not limited to this, and a value other than "1", for example, a value within a predetermined range near "1", may also be set. Furthermore, as will be discussed later in the vibration table experiment, the damping ratio "ζ^0" may be set to, for example, "1" (ζ^0=1).

[0083] The transfer function of equation (16) above is G^(s) = G^ a By using (s), for example, according to equation (15), the feedforward controller 11(K r (s)) and the first feedback controller 12(K σ (s)) can be designed.

[0084] According to equations (16) and (17) above, in order to design the controller (to obtain the transfer function G^(s) of the linear model 10²), • Information on vibration table 3 (table 32) when nothing is placed on it: Mass "m0", natural frequency "ω0" • Information on the test specimen 2 placed on the vibration table 3 (table 32): Mass "m sp " Although necessary, the controller can be designed without needing to understand the characteristics of the vibration table 3 when the test specimen 2 is placed on it. Furthermore, no information other than mass is required for the test specimen 2. Therefore, system identification testing is also unnecessary. However, this does not preclude conducting system identification tests.

[0085] Furthermore, although the linear model 101 represented by the transfer function G^(s) has a modeling gap with the actual control system 100, it was confirmed that by using a controller designed with this transfer function G^(s) in the NSBC, control performance equivalent to or better than conventional methods for the NSBC can be obtained, as will be described later in the shaking table experiments.

[0086] As mentioned above, the second feedback controller 13(K e (s)) is the first feedback controller 12(K σ (s)) is a controller that plays an auxiliary role to the first feedback controller 12(K σ If sufficient control accuracy can be achieved by (s), the second feedback controller 13(K e (s)) is not provided (or gain "K e You can also set "" to zero. Also, the second feedback controller 13(K e Even when (s)) is provided, the influence of the second feedback control is substantially zero, so the second feedback controller 13(K e (s)) may be designed. Specifically, the gain "K e The value of " may be set to a predetermined value, for example, a predetermined value that is very close to zero.

[0087] Furthermore, using the transfer function G^(s) in equation (16), the feedforward controller 11(Kr (s)) and the first feedback controller 12(K σ You may design either (s)) or both controllers.

[0088] Furthermore, unlike the above, the second feedback controller 13(K) is constructed using the transfer function G^(s) of equation (16). e ) could be designed.

[0089] (Functional configuration of the vibration table control device 10) Figure 5 is a block diagram showing an example of the functional configuration of the vibration table control device 10 in this embodiment. Here, an example of a block diagram focusing on the functional parts of the computer is shown. In this example, the vibration table control device 10 is assumed to have the function of a controller design device.

[0090] As mentioned above, the vibration table control device 10 may be composed of a general-purpose computer or the like, and may include, for example, a control unit 110, an input unit 120, a display unit 130, a connection unit 140, and a storage unit 190.

[0091] The memory unit 190 stores, as programs, for example, a vibration table control program 191 which is read by the control unit 110 and executed as vibration table control processing, and a controller design program 192 which is read by the control unit 110 and executed as controller design processing.

[0092] Furthermore, the memory unit 190 stores various types of data 193, such as various types of data used in vibration table control processing and various types of data used in controller design processing. These data include, for example, all data related to the control of the vibration table 3, including the vibration waveform data mentioned above, and parameter values ​​related to the test specimen 2 mentioned above (for example, mass "m" sp This may include all data related to the design of the controller, such as parameter values ​​for the vibration table 3 (e.g., mass "m0", natural frequency "ω0"), the transfer function of the calculated linear model 101, and the set controller gain.

[0093] In the vibration table control process, the control unit 110 calculates the transfer function G^(s) of the linear model 101 according to the method described above, and uses the calculated transfer function G^(s) to design a controller according to the method described above. Then, the control unit 110 uses the designed controller to control the vibration table 3.

[0094] Alternatively, the user may calculate the transfer function G^(s) of the linear model 101, and the control unit 110 may design the controller based on the transfer function G^(s) input via the input unit or the transfer function G^(s) received from an external device via the connection unit 140.

[0095] Furthermore, the display unit 130 may be configured to display the various types of information mentioned above. Additionally, the various types of information mentioned above may be transmitted and received between the device and an external device via the connection unit 140.

[0096] Alternatively, the vibration table control device 10 and the controller design device may be configured as separate devices, or the vibration test system 1 may include both the vibration table control device 10 and the controller design device.

[0097] (Vibration table experiment using vibration table control device 10) The vibration table control using NSBC in this embodiment can be applied to one or more layers of structures having various nonlinear characteristics as the test specimen 2.

[0098] The following is an example of the results of a shaking table experiment conducted with the aim of reproducing the JMA Kobe wave observed by the Japan Meteorological Agency (JMA) during the Great Hanshin Earthquake, similar to the results of Japanese Patent Publication No. 7287660.

[0099] The evaluation of the control performance of the vibration table control device 10 may be carried out in accordance with the method disclosed in Japanese Patent No. 7287660, for example, the evaluation index value S disclosed as formula (21) in Japanese Patent No. 7287660. t S fEvaluation may be performed using [a specific method]. Japanese Patent No. 7287660 is incorporated herein by reference.

[0100] <Experimental conditions> In this vibration table experiment, the specifications for Table 32 were set as follows: "Size = 1.2m × 1.2m", "Mass (m0) = 200kg", "Stroke = ±300mm", and "Maximum speed = 1.0ms".

[0101] Furthermore, in this shaking table experiment, a displacement transducer and two acceleration sensors were assumed to be equipped for control. This basic operation involves an internal error "e" as shown in Figure 6. in =r in -y cf "C" applies to " in The experiment relies on a PID controller described as (s) = 5 + 20·1 / s + 0.2·s / (0.01s+1). Since composite filtering is useful for reducing noise in the feedback signal, this technique was adopted in this vibration table experiment. The feedback signal "y" shown in Figure 6 cf This is a signal based on a composite filtering technique that acts on the displacement and acceleration of the vibration table 3, and is described, for example, by equation (18) below.

[0102]

number

[0103] Here, "ω cf " is the output of the displacement of the vibration table 3 "y d ", and the output of the displacement of the vibration table 3 "y d This is the switching frequency used to determine the contribution of the second time derivative (=output acceleration) of . Also, "ε0" is a fixed value, and for example, a value of approximately zero is set.

[0104] Based on the fact that the displacement transducer (P) attached to the vibration table 3 is sufficiently reliable up to a frequency of "3.0 Hz", the switching frequency "ω cf Let "" be "3.0·2πrad / s" (ω cf(=3.0·2π), and we set "ε0 = 0.01".

[0105] It should be noted that the above-mentioned PID controller was used by the inventor of this invention solely as a controller for the foundation of vibration table 3 in conducting this vibration table experiment, and is not essential. Furthermore, the above-mentioned PID controller is not directly related to NSBC.

[0106] Furthermore, in this shaking table experiment, test specimen 2 was a three-story (three-tiered) steel structure. The mass of this structure was as follows: "Mass of the first floor (m1) = 165 kg", "Mass of the second floor (m2) = 165 kg", and "Mass of the third floor (m3) = 160 kg". Additionally, because experimental fixtures were placed on table 32, the mass (m0) of table 32 became "250 kg".

[0107] (System identification test) Conventionally, system identification tests were performed on the vibration table 3 when nothing was placed on it, and on the vibration table 3 when the test specimen 2 was placed on it. However, as mentioned above, system identification tests are not essential in the method of this embodiment. This system identification test was performed using random band-limited excitation including frequency components up to 60.0 Hz.

[0108] (Vibration table 3 with nothing placed on it) First, the dynamic characteristics of the vibration table 3 (more specifically, the vibration table 3 + PID controller) in an empty state are given by the second-order transfer function G shown in equation (19) below. a Modeled as (s).

[0109]

number

[0110] This system identification test revealed that the natural frequency "ω0" and corresponding damping ratio "ζ0" of table 32 in an empty state are "ω0 = 5.8·2π (rad / s)" and "ζ0 = 1.0," respectively. It was also found that there is a time delay of "τ = 4.0 ms."

[0111] (Vibration table 3 with test specimen 2 placed on it) System identification tests yielded the frequency response function of a three-layer structure designated as Test Specimen 2. As a result, it was found that the natural frequencies and corresponding damping ratios of Test Specimen 2 are {f1, f2, f3} = {1.48 Hz, 3.71 Hz, 5.55 Hz} (expressed in Hz) and {ζ1, ζ2, ζ3} = {0.006, 0.004, 0.003}. However, the indices "1," "2," and "3" correspond to the 1st, 2nd, and 3rd floors, respectively.

[0112] Furthermore, in order to obtain more detailed values ​​regarding the damping and stiffness of specimen 2, SPLiTS (Simple Piecewise Linearisation in Time Series), an identification method in the time domain approach, was applied to the structural response obtained by the system identification test. The damping coefficients and stiffness of each layer of specimen 2 were then identified as {c1, c2, c3} = {571.0, 464.4, 468.5} Ns / m and {k1, k2, k3} = {81.16, 66.01, 48.38} kN / m, respectively.

[0113] After identifying these parameters, the damping coefficient and stiffness of the shaking table 3 were determined to be c0 = 3.14 kNs / m and k0 = 69.09 kN / m, respectively. Then, using the values ​​of these parameters, "m0 = 250 (kg)", and the masses of each of the three layers of the test specimen 2, "{m1, m2, m3} = {165, 165, 160} (kg)", the dynamic characteristics of the vibration table 3 with the test specimen 2 on it were modeled as an eighth-order transfer function shown in equation (20) below.

[0114]

number

[0115] Similar to vibration table 3 with nothing placed on it, vibration table 3 supporting a three-story structure was found to have a time delay of "τ = 4.0 ms", so "τ^ = 4.0 ms" was used.

[0116] (Experimental results using conventional methods) The transfer function G in equation (20) above a The results of a vibration table experiment using (s) as G^(s), where NSBC is not applied (K r ) and when NSBC is applied (K r , K σ ) and the evaluation index value S for each of them t S f The calculations yielded the results shown in Tables 1 and 2 below.

[0117] [Table 1]

[0118] [Table 2]

[0119] However, in this shaking table experiment, filter F e (s) to 0(F e (s)=0.0), filter F r (s) to 1(F r (s) = 1.0) Also, NSBC(K r , K σ ) filter F σ (s) is a second-order Butterworth bandpass filter with a bandpass range of "0.2~20Hz", and NSBC(K r ) filter F σ (s) was set to "0" (F σ (s=0.0).

[0120] Figure 7 shows, as an example, the experimental results corresponding to the 60% excitation experiment of (Kr, Kσ) in Table 2. (a) shows the time history of acceleration, (b) shows the FFT (Fourier amplitude spectrum) of acceleration, and (c) shows the history characteristics of the control system. In (a) and (b), the solid line is the input signal of acceleration (r = r a ), and the dashed line is the output signal of acceleration (y = y a ).

[0121] When NSBC is not applied {K r}, as shown in Table 1, the control accuracy is low. This is considered to be due to the modeling gap and non-linear characteristics of the control system 100 causing insufficient control performance because the control system does not have a feedback control system. On the other hand, when NSBC is applied {Kr, Kσ}, relatively excellent control is achieved as shown in Table 2 and Figure 7. At this time, the rigidity of the second-order part of the test body decreases, and severe non-linear characteristics occur.

[0122] (Experimental results applying the method of this embodiment) In the method of this embodiment, in the above formula (17), "ω^0 = 3.25·2π [rad / s]" and "ζ^0 = 1.0" are used, and based on the above formula (16), the transfer function G^(s) of the linear model 101 represented by the following formula (21) is obtained. [Number]

[0123] Based on the transfer function G^(s) of this linear model, the controller {K r , K σ} of NSBC is designed. The filter F σ (s) is a second-order Butterworth band-pass filter with a band-pass band of "0.2~20Hz" as in the conventional method.

[0124] When the method of this embodiment is applied, when the evaluation index values S t , S f are calculated, the results shown in the following Table 3 are obtained.

[0125] [Table 3]

[0126] In comparison with Table 2, it can be seen that, despite using a controller designed with a simple method, the control performance of this embodiment is comparable to that of the conventional method at excitation levels of 30% or less, and is equivalent to or better than that of the conventional method at excitation levels of 40% or more. For example, in a 60% vibration experiment, the evaluation index value in the time domain is "S" in the conventional method. t While the figure is "97.69 (%)", the method of this embodiment is "S t This represents 98.35%, which surpasses conventional methods.

[0127] Figure 8 shows, as an example, the experimental results corresponding to the 80% excitation experiment for (Kr, Kσ) in Table 3, where (a) shows the time history of acceleration, (b) shows the FFT (Fourier amplitude spectrum) of acceleration, and (c) shows the hysteresis characteristics of the control system. In (a) and (b), the solid line represents the input signal of acceleration (r=r a ) and the dashed line represents the output signal of acceleration (y=y a )

[0128] This demonstrates that, despite using a controller designed with a simple method, the method of this embodiment maintains stability and achieves excellent control performance. According to Table 3, the evaluation index value in the frequency domain is "S f =99.38(%)" and the evaluation index value in the time domain is "S f The results were 98.34% and all were high values ​​(although nonlinear characteristics were observed, the results were good according to Figures 8(a) and (b)).

[0129] In this embodiment, the vibrator can be controlled as required without needing to understand the characteristics of the vibrator in the state in which the object is installed. Furthermore, it is not necessary to understand any information about the object other than its mass. Moreover, in this embodiment, even if the object is damaged by the vibration, the vibrator can be controlled as required without needing to understand the extent of the damage. Another feature of this invention is that, when the object is installed on the vibrator, it is sufficient to monitor the output signal of the vibrator, and there is no need to monitor the output signal of the object itself.

[0130] (Regarding the scope of the present invention) The above describes an example of an embodiment to which the present invention is applied. The present invention uses control to match the output signal of the exciter that drives the object to a reference signal (control to match the output signal of the exciter that drives the object to the reference signal) as an automatic control, and is applicable to all types of excitation that do not use the characteristics of the exciter when the object is installed. As mentioned above, a major feature of the present invention is that it does not utilize the characteristics of the vibrator in the state in which the object is installed. Regardless of the specific method, any method of vibration that does not utilize the characteristics of the vibrator in the state in which the object is installed is included within the scope of the present invention.

[0131] Furthermore, generally speaking, the difficulty of a vibration test is determined by factors such as the mass of the object placed in the exciter and the strength of its nonlinear characteristics. The vibration test described in the above embodiment is arguably one of the most difficult vibration tests, but as demonstrated by the experimental results, it was confirmed that sufficient performance could be obtained. Therefore, it is believed that sufficient performance can be obtained even in very general vibration tests by using the method of the present invention, and even considering this point, the scope of the present invention is not limited to the scope of the embodiments described above.

[0132] (Effects of the embodiment) In the vibration excitation method using automatic control in this embodiment, the automatic control is a control that matches the output signal (e.g., output signal y) of a vibrator that drives an object (e.g., a structure such as a building) to a reference signal (e.g., input signal r), and does not use the characteristics of the vibrator in the state in which the object is installed (e.g., the input / output relationship of the vibrator). This eliminates the need for time-consuming tasks such as system identification tests in vibration excitation using automatic control, as it does not utilize the characteristics of the exciter with the object installed. This makes vibration excitation easier to implement. Furthermore, it is possible to achieve vibration excitation with sufficient accuracy even without utilizing the characteristics of the exciter with the object installed.

[0133] In this case, the information about the object is its mass (for example, mass "m"). sp You may choose not to use any information other than the above. Because the vibration excitation process utilizes the characteristics of the vibrator in the state in which the object is installed, and does not use any information other than mass as information about the object, vibration excitation can be implemented more easily.

[0134] Furthermore, the object may be, for example, a structure with one or more floors. Also, the exciter may be, for example, a shaking table.

[0135] Furthermore, a vibrator may be configured to drive the object in the above-described vibration method. This makes it possible to provide a vibrator that drives an object.

[0136] Alternatively, a computer may be used to design a controller for automatic control in the above-described vibration excitation method. This makes it possible to design controllers for automatic control.

[0137] In this case, the controller to be designed is a feedforward controller that performs feedforward control (for example, feedforward controller 11(K r (s))) may be included. This makes it possible to design feedforward controllers that perform feedforward control.

[0138] Furthermore, the controller to be designed may include a feedback controller that performs feedback control on a deviation signal (e.g., a first deviation signal σ(s)) between the output signal (e.g., output signal y) and the output signal (e.g., output signal y^) of a linear model (e.g., linear model 101) for the exciter (e.g., vibration table 3). This makes it possible to design a feedback controller that performs feedback control on the deviation signal between the output signal of the vibrator and the output signal of the linear model applied to the vibrator.

[0139] Furthermore, in this case, these controllers may be designed without using the characteristics of the exciter in the state in which the object is installed. Since the characteristics of the exciter in the state in which the object is installed are not used, the controller can be designed simply.

[0140] Furthermore, these controllers may be designed to use no information other than mass as information about the object. In addition to the characteristics of the vibrator with the object installed, the controller can be designed more simply because it does not use any information other than mass as information about the object.

[0141] Furthermore, the controller to be designed is a feedback controller (for example, a second feedback controller 13(K)) that performs feedback control on the deviation signal between the reference signal and the output signal (for example, a second deviation signal e(s)). e (s))) may be included. This makes it possible to design a feedback controller that performs feedback control on the deviation signal between the reference signal and the output signal. e (s))) is not necessarily required, and its impact may be minimized in the design.

[0142] <Other Embodiments> The embodiments to which the present invention can be applied are not limited to those described above. Several other embodiments are described below.

[0143] (1) In the above embodiment, the vibration table control device 10 controls the first feedback controller 12(K σ NSBC has been described as having (s)) or performing first feedback control on the first deviation signal σ(s). However, the present invention is not applicable only to NSBC, and the vibration table control device 10 may also be a first feedback controller 12(K σ The same applies when (s) is not provided or when the first feedback control is not performed on the first deviation signal σ(s).

[0144] In this case, for example, the second feedback controller 13(K e (s)) Gain "K e For example, a value determined in advance through experiments or calculated by a computer may be set as a value that can ensure a certain level of control performance. On the other hand, feedforward controller 11(K r (s)) may be designed, for example, by the method of the above embodiment.

[0145] (2) Unlike the above embodiments, the method of the above embodiments can also be applied in the case of vibrating an object rather than as a vibration test.

[0146] (3) There may be cases where it is undesirable to vibrate the object, such as in the event of an actual earthquake. In such cases, it is conceivable to not provide an input signal r, which can be considered equivalent to setting the input signal r to zero (or zero voltage in the case of voltage control), and such cases may also be included in the present invention. [Explanation of Symbols]

[0147] 1. Vibration Testing System 2 Test specimens 3. Vibration Table 10. Vibration table control device 11 Feedforward Controller 12. First Feedback Controller 13. Second Feedback Controller 14 Adder 15. The First Subtractor 16. The second subtractor 17 Stability Evaluation Department 30 bases 31 Bearings 32 tables 33 Actuators 34 detection sensors 100 control systems 101 Linear Models 102 Wasted Time Elements 103 Estimated waste time elements

Claims

1. A vibration excitation method using automatic control, The aforementioned automatic control is a control mechanism to match the output signal of the vibrator that drives the object to a reference signal. A vibration method that does not utilize the characteristics of the vibrator when the object is installed.

2. The vibration method according to claim 1, wherein no information other than mass is used as information about the object.

3. The aforementioned object is a structure with one or more floors. The vibration method according to claim 1.

4. The aforementioned vibrator is a vibration table. The vibration method according to claim 1.

5. A design method for causing a computer to design a controller for automatic control in the vibration excitation method described in claim 1.

6. The controller includes a feedforward controller that performs feedforward control. The design method according to claim 5.

7. The controller includes a feedback controller that performs feedback control on the deviation signal between the output signal and the output signal of the linear model for the vibrator. The design method according to claim 5.

8. The controller is designed without using the characteristics of the vibrator in the state in which the object is installed. The design method according to claim 6 or 7.

9. The controller is designed without using any information other than mass as information about the object. The design method according to claim 8.

10. The controller includes a feedback controller that performs feedback control on the deviation signal between the reference signal and the output signal. The design method according to claim 5.

11. A vibration excitation device using automatic control, The aforementioned automatic control is a control mechanism to match the output signal of the vibrator that drives the object to a reference signal. A vibration device that does not utilize the characteristics of the vibrator when the object is installed.

12. A program for achieving vibration excitation using automatic control, The aforementioned automatic control is a control mechanism to match the output signal of the vibrator that drives the object to a reference signal. A program that causes a computer to execute control to achieve vibration without using the characteristics of the vibrator in the state in which the object is installed.

Images