JP3694749B1 - Wireless remote diagnosis device - Google Patents

Abstract

Even when a diagnosis object that is difficult for a person to approach is diagnosed over a long period of time, a highly reliable soundness diagnosis can be performed.
A wireless remote diagnostic apparatus is installed at a diagnostic base remote from a diagnostic target, and includes a central unit that displays at least a diagnostic result, an excitation unit that applies a predetermined local vibration to the diagnostic target, Measuring means 6 for measuring the response of vibration applied by the vibrating means 5, storage means 7 connected to the measuring means 6 by wire and storing measurement data output from the measuring means 6, and the storing means 7 The analysis means 9 for receiving the measurement data stored in the base station and diagnosing the soundness of the diagnosis object based on the measurement data, the base side transmission unit 4a provided on the diagnosis base side, and the diagnosis object side It consists of a diagnosis target side transmission unit 4b, and is configured by a transmission means 4 that transmits and receives at least one of measurement data and diagnosis results between them by radio.
[Selection] Figure 1

Description

  The present invention relates to a wireless remote diagnosis device that determines a diagnosis of a soundness of a diagnosis object by determining a change in the diagnosis object at a place away from a diagnosis object such as a concrete structure or a steel structure.

  Conventionally, diagnosis of soundness in steel objects such as steel bridges, ships, and railway facilities, and concrete structures such as tunnels, buildings, and power plants has been diagnosed by visual detection of a deformed state or hammering into the object to be diagnosed. And judging the presence or absence of deformation from the sound at that time.

  However, such a diagnostic method is likely to be overlooked, and because there are individual differences in the judgment, it is difficult to make a diagnosis with low reliability and reproducibility.

  Therefore, a piezoelectric element is affixed to the structure, and vibrations are applied in a certain frequency range, and changes in the entire structure are detected by observing changes in the vibration mode in the low frequency range. There has been proposed a method for detecting a local defect by detecting a change in the electrical impedance at (see Patent Document 1).

  However, originally, in order to identify a deformed part of a structure, a reflected wave caused by the deformation is detected by a piezoelectric element, and the position is determined from the position on the time axis of the point of interest in the detection signal and the propagation speed of the elastic wave. There is a problem that it takes a long time to identify the position.

In view of such a situation, the applicant of the present application has proposed a system in which a local vibration is applied to a structure, and the response of the object to be diagnosed is determined from the change in the vibration mode obtained by measuring the response. (See Patent Document 2 and Patent Document 3).
JP 2001-99760 A JP 2003-106931 A Japanese Patent Application No. 2003-97668

  However, each of the above proposals is a configuration in which the system components are connected to each other by wire. For example, if the subject to be diagnosed is an observer, such as a bridge, cannot be measured, a signal cable must be used. There is a problem that it is necessary to wire the system to the observer and configure the system, which increases the cost and frequently causes malfunctions and measurement errors.

  In general, the deformation of a structure progresses gradually over a long period of time, but the deformation often proceeds abruptly after a certain state. For example, considering metal fatigue, it is a well-known fact that deformation progresses gradually in the plastic region, but progresses exponentially beyond the yield point.

  The soundness diagnosis of a structure is used for recognition of the current situation to take preventive measures before such a change suddenly changes, and it is important to continue the diagnosis for a long period of time.

  However, the measured response vibration is repeatedly reflected at the edge of the structure, deformation points, etc., and the influence of these reflected waves tends to appear at high frequencies. In the configuration in which the signal cable is mounted or mounted, there is no guarantee that the same signal can be measured even when the diagnosis target is not deformed due to a deviation in the mounting position of the vibrator or the signal cable routing method.

Therefore, there is a problem that it is very difficult to maintain reliability in the conventional method in which the system must be configured for each diagnosis.
In addition, the method of configuring the system by wiring the signal cable every time diagnosis is performed has a problem that work load such as wiring labor is added to the diagnosis cost.

  Therefore, the present invention eliminates the need for signal cable wiring at the time of diagnosis, and enables wireless soundness diagnosis with high reliability even when diagnosing objects that are difficult for people to approach for a long time. An object is to provide a remote diagnosis apparatus.

In order to solve the above-mentioned problem, the invention according to claim 1 is installed in a diagnostic base that is distant from the diagnostic target that is the target of the health diagnosis, and at least a general means for displaying the diagnostic result, and the primary of the diagnostic target Vibration means having a plurality of laminated piezoelectric actuators for applying vibrations at a frequency higher than the natural frequency in the mode to a local position of the diagnosis target, and measurement means for measuring the response of vibrations applied by the vibration means A storage unit connected to the measurement unit by wire and storing measurement data output from the measurement unit; and receiving the measurement data stored in the storage unit, and the soundness of the diagnosis target based on the measurement data It consists of an analysis means for performing diagnosis, a base side transmission unit provided on the diagnosis base side, and a diagnosis target side transmission unit provided on the diagnosis target side. And a transmission means for transmitting and receiving 1 by radio, and the measuring means comprises a plurality of measurement unit for converting into an electric signal by detecting a response vibration, provided corresponding to the measurement portion of the plurality of An A / D converter that converts an analog signal into a digital signal, and the measured signal is amplified and output to the A / D converter. At that time, the A / D converter is amplified so that no overflow occurs. And an amplifying unit composed of a plurality of amplifiers that can be set individually, and the storage means has a data loss due to the slow transmission speed of the transmission means with respect to the frequency of the measurement data output from the measuring means In order to prevent this, it is a writing speed which forms a speed adjustment function between the sampling speed of the measurement data by the measuring means and the transmission speed to the control means .

  In the invention according to claim 2, the analyzing means is connected to the controlling means by wire, and the storage means analyzes the measurement data stored in response to the data request command from the controlling means via the transmitting means and the controlling means. The soundness diagnosis is performed on the basis of the measurement data transmitted to the means and received by the analyzing means.

  Further, in the invention according to claim 3, the analysis means is connected to the storage means by wire, the analysis means receives the measurement data stored in the storage means, performs the soundness diagnosis, and requests the data from the overall means. According to the command, the diagnosis result is sent to the overall unit via the transmission unit.

  In the invention according to claim 4, at least the diagnosis start time and the excitation condition are stored in advance in the vibration unit, and when the diagnosis start time is reached, the diagnosis target is vibrated and the response is measured by the measurement unit. The measurement data is stored in the storage means, and when the diagnosis object side transmission unit receives a data request from the control means, the measurement data stored in the storage means is transmitted to the control means.

  The invention according to claim 5 is characterized in that a transmission medium for wireless communication performed between the base-side transmission unit and the diagnosis target-side transmission unit is radio waves or infrared rays.

  The invention according to claim 6 is characterized in that the vibration applied to the object to be diagnosed by the vibration means changes continuously with a constant amplitude.

  The invention according to claim 7 is characterized in that the amplitude of the vibration applied to the object to be diagnosed by the vibration means continuously changes at a constant period.

  In the invention according to claim 8, the analysis means analyzes the measurement data from the storage means and obtains the natural frequency of the diagnosis object as the experimental natural frequency, and one change of the diagnosis object Define the state and define multiple individuals that use genetic information as information that characterizes the transformation, such as the location of the transformation point, and calculate the natural frequency of each individual based on the genetic information to obtain the theoretical natural frequency. Search for individuals with the theoretical natural frequency that matches the experimental natural frequency with a predetermined degree of coincidence with the theoretical analysis unit, identify the individual with the highest degree of coincidence, and estimate the deformed state from the genetic information of the individual And an analysis database that stores the individual searched by the evaluation unit and enables the theoretical analysis unit to calculate the theoretical natural frequency using the stored individual in the next diagnosis. Features.

  According to the present invention, since at least one of the measurement data or the diagnosis result is wirelessly transmitted by the transmission unit and the measurement data is stored in the storage unit, the labor for performing the wiring of the signal cable or the like for each diagnosis becomes unnecessary. Even if it is difficult for the observer to get close to the diagnosis target, long-term diagnosis can be easily performed, and the storage means functions as a buffer between the sampling rate and transmission rate of the measurement data, so the reliability of the acquired measurement data Will improve.

  Embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram of a wireless remote diagnosis apparatus 2 according to the present invention, which starts diagnosis and instructs diagnosis contents, and receives and displays a diagnosis result and the like, a base side transmission unit 4a and a diagnosis object side A transmission unit 4 comprising a transmission unit 4b for transmitting and receiving various types of information; an excitation unit 5 for applying a predetermined vibration to a diagnosis target; a measurement unit 6 for measuring a response to the vibration applied by the excitation unit 5; The main components are a storage unit 7 that temporarily stores measurement data measured by the measurement unit 6, an analysis unit 9 that diagnoses the soundness of a diagnosis target based on the measurement data, and the like.

  The vibration means 5, the measurement means 6, the storage means 7, and the diagnosis target side transmission unit 4b are arranged on the diagnosis target side, and the control means 3, the analysis means 9 and the base side transmission are provided in the diagnosis base away from the diagnosis target. The part 4a is arranged.

The supervising means 3 is composed of a personal computer or the like, and supervises the whole diagnosis process by instructing the excitation condition for diagnosing the soundness of the diagnosis target most appropriately and displaying the diagnosis result. At the start, control data such as a diagnosis start command and vibration conditions are transmitted to the vibration means 5 via the transmission means 4.
In addition, a data request command is transmitted to the storage unit 7 via the transmission unit 4, and the received measurement data is transmitted to the analysis unit 9, and the diagnosis result is received from the analysis unit 9 and displayed on the screen.

The transmission means 4 transmits such control data and measurement data using a transmission medium such as radio waves and infrared rays, and the transmission method is not particularly defined in the present invention.
That is, PAN (Personal Area Network) such as Bluetooth, which is one of communication standards used when wireless communication between mobile devices such as notebook personal computers, mobile phones, PDAs and other various electronic devices as an alternative to communication cables. IrDA (Infrared Data Association), which is one of wireless data communication techniques using infrared rays and infrared rays, and a wireless LAN can be used.

  The transmission method to be adopted is determined by the distance between the diagnostic object and the diagnostic base and the surrounding conditions. For example, the diagnostic object is an iron bridge, and a place that is several tens of meters away from the iron bridge is the diagnostic base. In such a case, a radio wave system that can easily secure a communication distance of about 100 m and has low directivity is preferable.

  On the other hand, since infrared rays consume less power and do not cause radio interference, it is suitable to be applied to a system in which a large number of electronic devices are provided in close proximity, such as piping diagnosis in a nuclear power plant. .

  Wireless LANs and PANs have the advantage that they can be used even when there are obstacles in the communication path, depending on the degree of failure, because omnidirectional radio waves are used as transmission media. At this time, a wireless LAN having a radio wave interference avoidance function by spread spectrum is preferable so that external radio interference can be suppressed.

  Furthermore, the same wireless LAN preferably conforms to the international standard IEEE802.11b using a frequency band of 2.4 GHz. This is because the standard is designed in consideration of connectivity with a personal computer. For example, even if a personal computer is used as the overall means 3, no special interface is required, that is, the versatility is high. is there.

  It should be noted that the transmission means 4 so that the command from the generalization means 3 received by the diagnosis object side transmission unit 4b can be directly input to the vibration means 5, and the measurement data in the storage means 7 can be transmitted to the generalization means 3 as it is. It is preferable that a digital signal can be transmitted / received as described above, but a D / A converter is provided in the base side transmission unit 4a and an A / D converter is provided in the diagnosis target side transmission unit 4b so that the base side transmission unit 4a and the diagnosis target side transmission unit 4b are provided. It is also possible to communicate with an analog signal that has been subjected to FM modulation or the like, and the present invention does not exclude such a case.

  In addition, when communication between the base side transmission unit 4a and the diagnosis target side transmission unit 4b is performed with a digital signal, it is possible to perform the communication by packet communication. In this case, however, TCP / IP (Transmission Control Protocol / A method in which information indicating the packet order is added and transmitted as in Internet Protocol) is preferable. This is to prevent the occurrence of data loss due to a communication error and improve the reliability of diagnosis.

  The excitation means 5 is based on the excitation condition from the overall means 3, a signal generator 12 that generates a waveform signal of vibration to be applied to the diagnosis target, and a signal generated by the signal generator 12 to amplify the drive signal And a vibration unit 14 for applying a vibration corresponding to the drive signal to the diagnosis target.

  The signal generator 12 includes a function generator, a multifunction synthesizer, and the like, and generates and outputs a waveform signal of an analog signal corresponding to the excitation condition from the overall unit 3.

  In general, multi-function generators and multi-function synthesizers can generate general waveform signals such as sine waves, rectangular waves, and sawtooth waves themselves, but any waveform signal can be created according to the diagnosis target and purpose. It is preferable to have a function.

Here, the multi-function is used so that the vibration unit 14 is configured by a plurality of vibrators so that multi-channel vibration can be performed.
As a result, the excitation timing is shifted for each vibrator, the phase is changed, the amplitude, frequency, and excitation time are changed, and sweep vibration (vibration in which the amplitude and frequency are changed with time) And can be vibrated according to the diagnosis object and purpose.

  The drive unit 13 amplifies the waveform signal from the signal generation unit 12 with a predetermined gain, and outputs this to the corresponding vibrator as a drive signal. The gain may be specified from the signal generator 12.

  The vibration unit 14 is configured by a vibration exciter composed of a laminated piezoelectric actuator or the like, and this vibration exciter is attached to a diagnosis target with a fixed load by a fixture (see FIG. 2C).

  As the fixture, an example in which a leaf spring is used as the reaction force plate 17 is shown, but the present invention is not limited to this, as long as vibration can be reliably applied to a diagnosis target, For example, configurations such as double-sided tape, screwing, and magnetized magnetic bonding can be exemplified.

  Further, the vibrator is not necessarily limited to the case where the position is fixed, and a configuration in which the vibrator is moved to a desired position by a three-dimensional arm and pressed against a diagnosis target may be employed. Of course, it is preferable that the excitation position can be designated with high accuracy (for example, 10 microns). The same applies to measuring instruments described later.

  In addition, in order to observe the deformation | transformation of a diagnostic object, the vibrator which can give a high frequency (for example, 500 Hz or more) vibration is preferable. When considering nuclear power plants, chemical plants, iron bridges, ships, etc. as diagnostic targets, their natural frequencies (natural frequencies in the first-order mode) are very low (heavy weight). This is because the natural frequency of the deformation point has a high frequency so that a change at the frequency can be measured.

  The measurement means 6 includes a measurement unit 21 that measures a response to vibration applied to a diagnosis target, an amplification unit 22 that amplifies a measurement signal that is an analog signal output from the measurement unit 21, and a measurement signal from the amplification unit 22. An A / D converter 23 for converting into a digital signal is provided.

  The measuring unit 21 measures response vibrations to be diagnosed using a plurality of laser-type displacement measuring devices, piezoelectric elements, and the like. The arrangement method is an arrangement method of an exciter, an expected deformation point position, and the like. It is set appropriately according to

  It should be noted that the arrangement of a plurality of vibrators and measuring instruments can be arranged at regular intervals or arranged according to certain rules, but can be set as appropriate according to the diagnosis target as described above. It is.

  In other words, soundness diagnosis is particularly required in systems that cause tremendous human and property damage when accidents occur, such as in nuclear power plants, chemical plants, iron bridges, ships, etc., and long-term diagnosis is required. The

  In structural materials, deformation is likely to occur at rivet joints and welded parts due to corrosion and metal fatigue due to repeated loads, etc., and piping is severely work-hardened and is subject to large impacts due to water hammer. There is a feature that deformation is likely to occur in the portion, the pipe connection portion, and the like.

  Moreover, even if it says a welding location, for example, the said welding location does not deteriorate uniformly and a crack etc. do not generate | occur | produce, but it generate | occur | produces normally from the welding edge part side.

  In this way, there is a feature in the part that is subject to deformation depending on the diagnosis target, and diagnosing around such a part eliminates useless diagnosis and concentrates on the high risk part, It becomes possible to quickly detect the initial deformation in the diagnosis target with high accuracy.

  Therefore, it is important to arrange a vibrator and a measuring instrument and set the number so that a site where such a deformation is expected can be accurately diagnosed.

  In addition to the above-described meanings, a plurality of vibrators and measuring instruments may be provided in order to accurately measure vibration modes due to symmetric vibration, inverse symmetry, torsional vibration, and the like.

  The amplification unit 22 is composed of an amplifier such as a signal conditioner provided corresponding to each measuring instrument, and mainly plays a role of adjusting the level of the signal level input to the A / D conversion unit 23, and the gain of each amplifier is measured. It is set depending on the configuration of the instrument and the configuration of the diagnosis target.

  That is, it is necessary to set the gain of each amplifier to be constant (so that at least the same level signal is output when the same level signal is input), and the signal from the amplifying unit 22 is described later. In this case, the digital signal is converted by the A / D converter 23. At this time, it is necessary to set so as not to cause an overflow.

However, when it is desired to adjust the vibration energy to be applied according to the object to be diagnosed or to continuously change the vibration energy to be applied, an overflow may occur if the amplifier has a fixed gain.
From such a viewpoint, it is preferable that the gain can be adjusted, and the gain adjustment may be manual adjustment or may be made possible by the overall unit 3 or the like.

  Since the signal from the measurement unit 21 is an analog signal, the A / D conversion unit 23 converts the signal into a digital signal so that the storage unit 7 can store it. As a matter of course, the sampling speed in the A / D converter 23 is required to be faster than the dominant frequency of the response vibration.

  The storage unit 7 includes a measurement data storage unit 24 including a rewritable memory such as a DRAM (Dynamic Random Access Memory), and stores measurement data from the A / D conversion unit 23. When the data request is received from the diagnosis target side transmission unit 4b, the stored measurement data is output, and when the data transmission completion notification is received from the diagnosis target side transmission unit 4b, the data is erased.

  In addition, it is possible to transmit the measurement data directly to the overall unit 3 via the transmission unit 4 without using the storage unit 7 as described above, but the transmission speed is slow with respect to the frequency of the measurement data. There is a risk of missing measurement data.

  In general, since it is practically difficult to make the entire diagnosis target uniform, it is possible to specify an expected deformation point in advance and perform measurement centering on that point. In a large-scale diagnosis target such as a facility, there may be a huge number of such assumed deformation places, and if remeasurement is performed every time measurement data is lost, the measurement time is prolonged and the workability is remarkably reduced.

  In addition, since the observer cannot recognize the occurrence of data loss in most cases, it must be diagnosed as having no data loss. In order to cope with such a situation, an apparatus for monitoring data loss is introduced. A configuration in which the monitoring result by the device and the operation of the vibration means 5 are linked to perform re-measurement is required, and the diagnostic device becomes very complicated and expensive.

  However, by providing the storage means 7, the storage means 7 functions to adjust the sampling speed of the measurement data and the transmission speed to the control means, so that data loss can be prevented and the apparatus can be simplified and reduced in cost. There is an advantage that becomes possible.

  Note that a semiconductor element capable of reading and writing at high speed such as DRAM is preferable as the storage means 7, but even an information recording medium such as a hard disk can be applied as long as requirements such as writing speed are satisfied.

  The analysis means 9 includes an experimental data analysis unit 25 for obtaining a natural frequency (hereinafter referred to as an experimental natural frequency) in a primary to n-th vibration mode of a diagnosis target from measurement data, and theoretically the primary to n of the diagnosis target. Theoretical analysis unit 26 for calculating the natural frequency in the next vibration mode (hereinafter referred to as the theoretical natural frequency), judging whether or not the theoretical natural frequency matches the experimental natural frequency under a predetermined condition. An evaluation unit 27 that performs diagnosis, an analysis database (analysis DB) 28 that stores information such as diagnosis information, theoretical natural frequency, and experimental natural frequency are provided.

  In the following, as shown in FIG. 2, an H-shaped steel material 39 is supported by a support column 40 through an angle 41, and the H-shaped steel material 39 and the angle 41 are fixed by a plurality of bolts 42. A system fixed by welding or the like will be described as an example (see FIGS. 2A and 2B).

  Then, assuming that the deformation is caused by the loosening of the bolt 42, the vibrator 45 is pressed against the H-shaped steel material 39 by the reaction force of the reaction force plate 17, and the measuring instrument 46 is affixed to the H-shaped steel material 39 with double-sided tape or the like. Suppose that

  In such a system, when the response of the applied vibration is measured and its power spectrum is obtained, a power spectrum as shown in FIG. 13 is obtained. 13A shows a power spectrum when all the bolts 42 are in a tight state, FIG. 13B shows a power spectrum when any of the bolts 42 is in a relaxed state, the horizontal axis is the frequency, and the vertical axis is the spectrum intensity. Is shown.

When these power spectra are compared, a large peak appears at about 310 Hz in FIG. 13B (point indicated by P1 in the figure), and the peak appears to be separated at about 150 Hz (point indicated by P2 in the figure). . Thus, the presence or absence of deformation can be determined by whether or not there is a difference in the power spectrum.
Therefore, when the soundness diagnosis aims at the presence / absence of deformation, it is possible to compare the power spectra.

  However, even if the occurrence of deformation can be determined, it cannot be determined until where and what kind of deformation has occurred. Therefore, in order to make the contents of the diagnosis and the contents of the deformation point the diagnosis contents, an analysis algorithm such as GA described later is used in the present invention.

  Genetic algorithms, neuronetwork regression analysis, multivariate analysis, pattern recognition analysis, etc. can be applied as such analysis algorithms. Among them, genetic algorithms that are considered to be very effective for optimization problems ( A case where GA (Genetic Algorithm) is used will be described as an example.

  An outline of GA will be briefly described. GA is a technique originally devised by simulating the laws of heredity in the living world, and is a technique for finding a better solution while genetically changing multiple solutions. This solution is expressed (coded) in the form of a gene.

FIG. 3 is a flowchart showing such a GA algorithm. In the GA, first, an initial group (individual group) that is a group of solutions (individuals) is created (step S1).
This individual corresponds to one deformed state to be diagnosed, and the position of the deformed point, the spring stiffness value, etc. are used as genetic information characterizing this individual.

  For example, when there are two bolts A and B, the bolt A is “01”, the bolt B is “10”, and when each of the bolts A and B is in a tight state, “1”, when in a relaxed state Is associated with “0”. In this case, the genetic information is “011” (bolt A is tight), “010” (bolt A is loose), “101” (bolt B is tight), “100” (bolt B is loose) Can be coded as Therefore, four individuals can be defined as the deformed state (a healthy state is also an aspect of the deformed state).

Next, evaluation is performed. In the evaluation, the fitness is obtained for all individuals (solutions), and individuals to be left in the next generation are determined based on the fitness (steps S2 and S3). This goodness of fit is like the height of the evaluation of the solution, and the evaluation function is set so that the better the better the solution, the higher the goodness of fit.
In the present invention, the theoretical natural frequency of each individual is calculated, and how well this theoretical natural frequency matches (matches) with the experimental natural frequency (is a good solution) by an evaluation function. evaluate. The evaluation function will be described later.

When the abnormal state is such as bolt relaxation, the optimal solution can often be found among the assumed individuals, but in the case of a crack, the optimal solution can be found from the assumed individuals. There are things you can't do.
Therefore, in GA, crossover and mutation operations called GA operators are performed (step S4).

  This crossover and mutation operation is based on the laws of heredity. In crossover, new individuals (children) that inherit genes from multiple parents (generally two individuals) are generated with a certain probability, and mutations are made. Then, it sets so that an individual may occur with a lower probability than crossover.

  For example, when there is a plurality of bolts and a group of individuals considering relaxation of the bolts every other bolt is assumed, it is assumed that the bolts adjacent to the assumed bolts are relaxed. If the bolt assumed at this time is relaxed, an optimal solution can be obtained by exploration.

  However, when the bolt adjacent to the assumed bolt is relaxed, the optimum solution cannot be found by the search, and therefore, a new individual is created by the crossover operation with the high-fitness 2 bolt (assumed bolt) as a parent. If this individual assumes that the adjacent bolt is deformed, the optimum solution is obtained by the search.

However, if the deformation is not a bolt relaxation but a crack, the optimal solution cannot be obtained even if the crossover operation is repeated. Will be generated and explored.
Such an exploration is considered as one generation, and an optimal solution is obtained by performing a predetermined number of times.

  The procedure for making a diagnosis using such GA will be described with reference to FIG. First, a vibration is applied to the diagnosis target (step SA1). Thereby, a response vibration having a vibration mode of the first order to the nth order (n: integer) is measured from the diagnosis target.

  Therefore, the experiment data analysis unit 25 performs power spectrum analysis of the measurement data, and if the outstanding amplitude changes due to the deformation or a new outstanding amplitude is generated, this superior point is searched for and used as the natural frequency of the experiment. (Step SA2, Step SA3).

  For example, when a sweep vibration in which the frequency is continuously changed from the start frequency to the stop frequency as shown in FIG. 5 is given to the diagnosis target, measurement data as shown in FIG. 6 is observed. Therefore, the experimental data analysis unit 25 performs a power spectrum analysis on the measurement data to obtain a power spectrum as shown in FIG. 7, and a place where a vertical line is drawn in FIG.

  Note that frequency analysis is performed by Fourier transforming the measured data, and the point where the amplitude at each frequency is excellent may be the natural frequency of the experiment. Furthermore, power spectrum analysis is performed on the data obtained by Fourier transform. By doing so, the amplitude change may be clarified, and thereby the experimental natural frequency may be identified with high accuracy.

  The reason for applying the sweep vibration as shown in FIG. 5 is to resonate the deformation point so that information such as the position, size, and type of the deformation point is included in the measurement data.

  That is, since the position of the deformation point is unknown at the time of diagnosis, the deformation point is resonated by vibrating at a wide frequency so that the measurement data includes the information.

  Of course, when the deformation point can be identified or when the deformation point can be predicted with high accuracy, the vibration may be a constant frequency instead of a sweep vibration, or a vibration whose frequency changes within a narrow range. May be.

  For example, in a system in which a plurality of members are integrated by bolts, when bolt relaxation is predicted to occur empirically or statistically (the relaxation degree in this case is binary) ) Since the position of the bolt is known in advance, vibration of a specific frequency or a frequency in the vicinity thereof may be applied so as to detect the deformation of the bolt.

  In addition, the vibration given in FIG. 5 is considered to be a surface acoustic wave propagating on or near the surface to be diagnosed. For example, internal defects such as corrosion of reinforcing bars in concrete structures and deformation of concrete due to acid rain. Thus, when measuring deformation at a deep position, it is necessary to have enough energy to propagate the vibration to that depth and return the response.

  For this reason, for example, it is necessary to apply vibration of energy corresponding to the depth of the detected deformation point, such as applying sweep vibration whose amplitude increases with time. The present invention includes such a case.

  In the case of a diagnosis target as shown in FIG. 5, (1) bolt relaxation, (2) crack, (3) rust / corrosion, etc. can be assumed as deformation modes. Explained.

  The theoretical analysis unit 26 calculates a theoretical natural vibration to be diagnosed using a finite element method or the like. In the finite element method, the diagnosis target as shown in FIG. 2 is divided into a plurality of elements as shown in FIG. 8, for example, and expressed by boundary conditions of each element, Young's modulus, cross-sectional area, bolt tightening length, and the like. Calculate by changing the stiffness value of the spring element.

  For this reason, element division corresponding to the presumed deformation point is important. For example, when bolt deformation or corrosion is assumed as the deformation, the element is divided so that the bolt or corrosion position is included in one of the elements. In addition, when a crack is assumed in a welded part or the like, the element is divided so that the welded part is located at the boundary of the element.

When assuming bolt relaxation as deformation, set the spring element stiffness value in the gene information as described above, and when assuming cracks, the boundary condition that there is no continuity with adjacent elements is set. Set to genetic information. In the case of corrosion, the elastic coefficient is set in the gene information.
In this way, the spring element stiffness value or the like is set in the gene information according to the assumed property of deformation, and this is used as a calculation parameter.

  FIG. 9 is a diagram illustrating a deformed state in which the bolt is relaxed, in which CASE 1 is in a healthy state and CASE 2 to CASE 7 are in a deformed state. The bolt position is indicated by a circle, and the bolt hatched in the circle indicates that the bolt is relaxed.

  The calculation parameters are set for the diagnosis target divided into elements. At this time, it is possible to use the spring constant, elastic coefficient, damping constant, etc. of the members constituting the diagnosis target. However, since it is difficult to properly model the diagnosis target when applying the theory, it is impossible to calculate a reliable natural frequency.

  Therefore, in the present invention, at the first diagnosis (K = 1), the theoretical natural frequency is calculated using the spring constant of the member as a calculation parameter, and this theoretical natural frequency matches the experimental natural frequency under a predetermined condition. The calculation parameter is changed as follows. Then, the calculation parameter when the values match is identified as a spring element stiffness value or the like to be diagnosed.

  As a result, a theoretical natural frequency (high accuracy) close to the natural frequency of the actual diagnosis target can be easily obtained, and the time required for calculation can be shortened.

The first diagnosis (K = 1) means the first diagnosis when the apparatus is installed or when the analysis database data is initialized after installation.
Since the change of the diagnosis target is performed by detecting the state change, in the first diagnosis (K = 1), there is no data regarding the original state, so the diagnosis cannot be performed, and no individual group is set in GA. Because it is in a state.

  Therefore, at the time of the first diagnosis, an individual group in GA is set (step SA4, step SA6), and the genetic information of the individual is coded (step SA7). FIG. 10 shows the genetic information of each individual coded in this way.

  Thereafter, the theoretical natural frequency is calculated by the finite element method (step SA8). At this time, the spring element stiffness value and deformation point position constituting the gene information are calculated as calculation parameters, and a curve of amplitude (vertical axis) with respect to frequency (horizontal axis) as shown in FIG. 11 is obtained.

  It is assumed that the maximum value of this curve corresponds to the natural frequency in the first to nth vibration modes, and this natural frequency is affected by the occurrence of the deformation and appears as a frequency change or the like.

  The theoretical natural frequency fa (i) and the experimental natural frequency fe (i) obtained in this way are substituted for the preset evaluation function δ to evaluate the fitness (step SA9). Here, i indicates the natural frequency of the i-th vibration mode.

As the evaluation function δ, for example, a function that adds the square of the difference between the theoretical natural frequency fa (i) and the experimental natural frequency fe (i) in each vibration mode as shown in Equation 1 can be used. Of course, various evaluation functions δ can be defined. For example, the absolute value of the difference may be added in each vibration mode.

  On the other hand, if it is determined in step SA4 that it is not the first diagnosis (K = 1), the data D (K-1) as the previous diagnosis result is read from the analysis DB 28, and this data D (K-1) The fitness is evaluated using the theoretical natural frequency included in the above and the experimental natural frequency obtained in step SA3 (step SA9).

  It is determined whether or not the fitness value obtained in this way is smaller than a certain fitness value reference value (step SA10). If the fitness value is larger than the fitness value reference value (when the fitness value is low), it is set in advance. Genetic manipulation is performed based on the crossover probability Pc and the mutation probability Pm (step SA11). Such a process is set as one generation, and the optimum solution is obtained by repeating up to a preset number of generations.

  When the optimum solution is obtained, it is determined whether or not the diagnosis is first (K = 1). When K = 1, the individual group including the obtained optimum solution is stored in the analysis DB 28 as data D (K). (Step SA15).

  On the other hand, if the diagnosis is not the first (K ≠ 1), the deformed position and the contents of the deformation (whether the deformation is due to bolt loosening or cracking, etc.) are determined from the genetic information in the individual with the optimal solution. Thus, the soundness diagnosis is performed (step SA13, step SA14), and the result is stored (step SA15).

  Next, a verification test of the wireless remote diagnosis apparatus of the present invention will be described. As a test object, an H-shaped steel having a steel material structure as shown in FIG. 2 and having a length of 350 mm × 350 mm × 12 mm × 19 mm and a length L = 2200 mm is a diagnosis object, and the diagnosis object and the support column 40 are not in contact with each other.

  A vibration means 5, a measurement means 6, a storage means 7, and an analysis means 9 as shown in FIG. As shown in FIG. 2, the actuator exciter (1 unit) and the accelerometer measuring unit (1 unit) are attached by the reaction force plate 17 so that the reaction force is about 100 N. The measuring instrument was attached with double-sided tape or the like.

  Next, it was confirmed that the measurement data is the same for both wireless and wired data. FIG. 12 shows the result. FIG. 12A shows a wired case, and FIGS. 12B to 12E show a case where a wireless LAN is used. FIGS. 12B to 12E show cases where different wireless LAN devices (transmission means) are used. In each figure, the time axis of the waveform is shifted due to a shift in sampling timing.

  As can be seen from these figures, the waveforms in the figures show the same waveform, and it is shown that the reliability equivalent to that in the wired case can be obtained even in the case of wireless.

  In the analysis means, the bolt joint part to be diagnosed was modeled as a spring element as a preparation, and natural vibration analysis was performed by the finite element method. At this time, in order to analytically evaluate the stiffness of the spring element representing the bolt joint, the stiffness of the spring is identified using GA (corresponding to the result of the first diagnosis (K = 1) in the analysis procedure described above). The possibility of identification of deformation points was examined.

  In modeling, the contact state between the H-shaped steel and the angle by the bolt is expressed by spring elements in three directions (X, Y, Z), and the constraint conditions (θx, θY, θz) in the rotational direction of the spring element nodes are expressed. Fixed.

  A predetermined number of these spring elements are arranged at the nodes on the contact surface with the angle, and the stiffness value of the spring element at the relaxed portion is obtained. In GA, since the theoretical natural frequency is repeatedly calculated, the time required for this is increased. Therefore, in this embodiment, the GA analysis model is simplified as much as possible to shorten the time required for the analysis.

  Then, as shown in FIG. 5, the sine wave vibration having a constant amplitude, a start frequency of 0.1 Hz, and a stop frequency of 400 Hz is linearly changed from the integrated means to a diagnosis target in a sweep time of 15 seconds. An instruction having the diagnostic contents to be applied is transmitted to the vibration means via the transmission means. The vibration means generates a vibration waveform signal based on this command and vibrates the vibrator.

  Response vibration due to this excitation was measured by a measuring instrument, amplified by adjusting the gain of the amplifying unit so as not to cause overflow in the A / D conversion unit 23, and stored in the measurement data storage unit 24.

  When the measurement is completed, a data request command is issued from the diagnosis target side transmission unit 4b, and in response to this, the measurement data stored in the measurement data storage unit 24 is sent to the diagnosis target side transmission unit 4b, and the diagnosis target side transmission is performed. The data is sent from the unit 4b to the experimental data analysis unit 25 in the analysis unit 9 via the base side transmission unit 4a and the overall unit 3.

  FIG. 13 is a power spectrum obtained by analyzing the power spectrum based on the measurement data. FIG. 13A corresponds to the healthy state (CASE1), and FIG. 13B corresponds to CASE3.

  FIG. 14 shows the results of the experimental natural frequencies of the vibration modes of the respective orders in CASE 1 to CASE 7, and FIG. 15 shows the change in the experimental natural frequencies of CASE 2 to CASE 7 in terms of a decrease rate based on the normal state (CASE 1). It is a figure. If attention is paid to the rate of decrease in these figures, it can be seen that the degree of decrease in the change in the natural frequency differs depending on the loosening bolt.

  Subsequently, the theoretical analysis unit 26 and the evaluation unit 27 identify the theoretical natural frequency using GA with the experimental natural frequency at the time of sound (that is, based on measurement data of CASE 1) as a target value.

  FIG. 16 is a diagram showing the theoretical natural frequency and the experimental natural frequency by simple GA when the crossover probability Pc = 60%, the mutation probability Pm = 5%, and the population number 50. However, the results of the third and fourth orders are very close to the experimental values.

  Next, assuming that CASE 3 shown in FIG. At this time, only the spring element stiffness value in the Y direction was decreased for the corresponding element, and the spring element stiffness value of 60% in the healthy state was obtained. And the spring element rigidity value of the Y direction was identified, and it was investigated whether the result was 60% of healthy.

  FIG. 17 shows the result as a whole, although there are some spring elements that are evaluated to be smaller than 60% in a healthy state, and variation is recognized, but it is found that the whole is effective. That is, CASE 3 can be diagnosed as a deformed state. Note that simple GA was applied as GA at this time, and the crossover rate was 60%, the mutation rate was 5%, and the population was 50.

  As described above, even when wireless measurement data is transmitted, it can be proved that the same result as in the case of priority is obtained, so that even if it is difficult for the observer to get close to the diagnosis target, long-term diagnosis is possible. Can be easily performed, and it is possible to eliminate the labor of wiring the signal cable for each diagnosis.

  Next, a second embodiment of the present invention will be described. In the first embodiment, the vibration means 5, the measurement means 6, the storage means 7 and the diagnosis object side transmission unit 4b are arranged on the diagnosis object side, and the overall means 3, analysis means 9 and Although the base side transmission unit 4a is arranged, the present invention is not limited to this, and as shown in FIG. 18, the analysis unit 9 is provided on the diagnosis target side and wired to the storage unit 7. However, the configuration may be such that the diagnosis result and measurement data are transmitted to the overall unit 3.

  In addition, as shown in FIG. 19, the vibration control unit including the vibration unit 5, the measurement unit 6, the storage unit 7, and the diagnosis target side transmission unit 4 b as a unit, and the vibration unit 5 includes a timer unit 30 such as a timer. In addition, a configuration is also possible in which a diagnosis start time is stored in the vibration control unit 31, and vibration and measurement are automatically performed when the diagnosis start time is reached, and the result is transmitted to the overall unit 3.

  In this way, by automatically oscillating and diagnosing, for example, in a large-scale facility such as a nuclear power plant, there is no need to issue a diagnosis start command one by one, and an abnormality is detected at any place in the facility. System monitoring such as issuing an alarm becomes possible.

  Such system monitoring is currently indispensable for the purpose of monitoring the movement of components (for example, abnormal flow rate of cooling water), but has not yet been monitored for structures, and its means In view of the current situation where there is no such thing, it is very useful in ensuring safety.

It is a block diagram which shows schematic structure of the radio | wireless remote diagnosis apparatus concerning this invention. It is a figure which illustrates a diagnostic object. It is a flowchart explaining the algorithm of GA. It is a flowchart which shows a diagnostic analysis procedure. It is a figure which illustrates a sweep vibration as a vibration waveform provided to a diagnostic object. It is a figure which illustrates measurement data. It is the figure which calculated | required the power spectrum from measurement data. It is the figure which divided | segmented the diagnostic object shown in FIG. 2 into elements for application of the finite element method. FIG. 3 is a diagram illustrating a deformed state of the diagnosis target illustrated in FIG. 2. FIG. 10 is a diagram illustrating gene information when the deformed state in FIG. 9 is an individual. It is the waveform of the vibration mode calculated in the theoretical analysis part. It is a figure which shows the measurement data at the time of using the case where a transmission means is wired, and setting it as wireless LAN. It is the figure which showed how a power spectrum changes with the difference in the presence or absence of deformation. It is the figure which showed the experiment natural frequency of the individual | organism | solid shown in FIG. 9, and its fluctuation rate in the vibration mode. FIG. 10 is a graph showing the result of FIG. 9. It is the figure which compared the theoretical natural frequency and experimental natural frequency in the time of healthy (CASE1) in the vibration mode of each order. It is a figure which shows a diagnostic result when CASE3 is assumed as a deformed state. It is a block diagram which shows schematic structure of the radio | wireless remote diagnosis apparatus instead of FIG. 1 applied to description of other embodiment. It is a block diagram which shows schematic structure of the radio | wireless remote diagnosis apparatus instead of FIG. 1 applied to description of other embodiment.

Explanation of symbols

DESCRIPTION OF SYMBOLS 2 Wireless remote diagnosis apparatus 3 Supervision means 4 Transmission means 4a Base side transmission part 4b Diagnosis object side transmission part 5 Excitation means 6 Measurement means 7 Storage means 9 Analysis means 12 Signal generation part 13 Drive part 14 Excitation part 17 Reaction force plate 21 Measurement Unit 22 Amplification Unit 23 A / D Conversion Unit 24 Measurement Data Storage Unit 25 Experimental Data Analysis Unit 26 Theoretical Analysis Unit 27 Evaluation Unit 30 Timekeeping Unit 31 Excitation Control Unit

Claims (8)

A centralized means for displaying at least the diagnostic result, installed at a diagnostic base that is distant from the diagnostic target that is the target of the health diagnosis ,
A vibration means comprising a plurality of laminated piezoelectric actuators for applying a vibration at a frequency higher than the natural frequency in the primary mode of the diagnosis object to a local position of the diagnosis object ;
Measuring means for measuring a response of vibration applied by the vibration means;
Storage means connected to the measurement means by wire and storing measurement data output from the measurement means;
Analysis means for receiving the measurement data stored in the storage means, and performing a diagnosis of the soundness of the diagnosis object based on the measurement data;
A base-side transmission unit provided on the diagnostic base side and a diagnostic-subject side transmission unit provided on the diagnostic target side, and transmission means for wirelessly transmitting and receiving at least one of the measurement data and the diagnostic result between them Prepared and
A plurality of measuring units that detect response vibrations and convert them into electrical signals and output; and an A / D conversion unit that is provided corresponding to the plurality of measuring units and converts analog signals into digital signals. And amplifying the measured signal and outputting the amplified signal to the A / D converter, and at this time, a plurality of amplification levels can be set individually so that overflow does not occur in the A / D converter And an amplifying unit comprising an amplifier,
In order to prevent data loss due to a slow transmission speed of the transmission means relative to the frequency of the measurement data output from the measurement means, the storage means sends the measurement data sampling speed by the measurement means to the overall means. A wireless remote diagnostic apparatus characterized in that the writing speed forms a function of adjusting the speed with the transmission speed .   The analysis means is wired to the control means, and the storage means transmits measurement data stored in response to a data request command from the control means to the analysis means via the transmission means and the control means. The wireless remote diagnosis apparatus according to claim 1, wherein a soundness diagnosis is performed based on the measurement data received by the analyzing means.   The analysis unit is wired to the storage unit, the analysis unit receives measurement data stored in the storage unit and performs a soundness diagnosis, and a diagnosis result according to a data request command from the control unit 2. The wireless remote diagnosis apparatus according to claim 1, wherein a diagnosis result is sent to the control unit via the transmission unit.   At least the diagnosis start time and the excitation condition are stored in advance in the vibration means, and when the diagnosis start time is reached, the diagnosis target is vibrated, the response is measured by the measurement means, and stored in the storage means. 4. The diagnostic data transmission unit according to any one of claims 1 to 3, wherein when the diagnosis target side transmission unit receives a data request from the integrated unit, the measurement data stored in the storage unit is transmitted to the integrated unit. Wireless remote diagnostic device.   The wireless remote diagnosis apparatus according to any one of claims 1 to 4, wherein a transmission medium for wireless communication performed between the base side transmission unit and the diagnosis target side transmission unit is radio waves or infrared rays.   6. The wireless remote diagnosis device according to claim 1, wherein the vibration applied to the diagnosis target by the vibration means changes continuously with a constant amplitude. The vibration to be applied to the diagnostic object by vibration means, wireless remote diagnostic apparatus according to claim 1 to 5 any one of claims, characterized in that changes in amplitude continuously at a constant period. The analysis means analyzes the measurement data from the storage means and obtains the natural frequency of the diagnosis target as the experimental natural frequency;
Defines one abnormal state to be diagnosed, defines multiple individuals whose genetic information is information that characterizes the deformation, such as the position of the deformation point, and calculates the natural frequency of each individual based on the genetic information And the theoretical analysis part to make the theoretical natural frequency,
Evaluation of searching for the individual having the theoretical natural frequency that matches the experimental natural frequency with a predetermined degree of matching, identifying the individual having the highest degree of matching, and estimating the deformed state from the genetic information of the individual And
An analysis database is provided that stores an individual searched by the evaluation unit and enables the theoretical analysis unit to calculate a theoretical natural frequency using the stored individual in a next diagnosis. Item 8. The wireless remote diagnosis device according to any one of Items 1 to 7.

Images