Machine performance diagnostic equipment

Abstract

To provide a performance measurement diagnostic device which can more properly diagnose a state of a machine on the basis of a frequency analysis result obtained from state information of the machine.SOLUTION: A performance measurement diagnostic device for measuring and diagnosing a performance of a machine comprises: an information acquisition unit which acquires state information of the machine; an information pre-processing unit which performs pre-processing of the state information acquired by the information acquisition unit; a frequency feature discrimination unit which discriminates a state frequency feature candidate being a candidate of a frequency feature indicating a state of the machine and a frequency feature indicating noise by performing frequency analysis of data after pre-processing by the information pre-processing unit; a frequency feature estimation unit which estimates a frequency feature corresponding to the state of the machine from the state frequency feature candidates on the basis of the state frequency feature candidates and the frequency feature indicating the noise; a performance measurement diagnostic unit which specifies a feature on the basis of the number of frequency features estimated by the frequency feature estimation unit and the positional relation, and diagnoses the state of the machine; and a display unit which displays the diagnosed state of the machine on a screen.SELECTED DRAWING: Figure 7

Description

[Technical field]

[0001] The present invention relates to a performance diagnostic device. [Background technology]

[0002] As part of machine inspections, performance measurements are periodically conducted to quantitatively grasp the condition of each component of the machine. For example, if the machine is a construction machine, the worker will set the construction machine in a specified state, then perform a specified operation related to the performance measurement, measure the operation time, and compare it with a value indicating the standard performance (hereinafter referred to as the standard value) to diagnose the condition of the construction machine.

[0003] One method for diagnosing the performance of the traveling and drive system of construction machinery is to measure the traveling speed. The time it takes for the construction machinery to travel a specified distance at its maximum traveling speed is measured, and the traveling time is compared with a standard value to determine whether the traveling performance meets the standard. This method requires providing sufficient space for the construction machinery to travel, but there are problems in that it is difficult to secure an appropriate space for measurement at the work site, and it also takes time to prepare for the measurement.

[0004] In response to this, there is a method for diagnosing the performance of the traveling drive system of construction machinery by measuring the rotation speed of the tracks. After wrapping a cloth or the like around a part of the tracks of the construction machinery to mark a measurement, the worker lifts one side of the tracks while the construction machinery is stopped, tilting the construction machinery and allowing the lifted tracks to spin freely in place. Then, while checking the marks, the worker manually measures the time it takes for the tracks to rotate a specified number of times, for example using a stopwatch, and compares this with a predefined standard value to determine whether the standard performance is met.

[0005] This method does not require the construction machine to actually run, and does not require the time and effort of securing a space for measurement and preparation. However, this measurement is usually performed by a single worker. Therefore, it is necessary to operate the lever to spin the tracks, visually check that the mark on the tracks rotates the specified number of times, and measure the time with a stopwatch at the same time. This requires performing several different actions at the same time, so the measurement work is laborious, and there is also room for improvement in the measurement accuracy. Therefore, a method is desired to automatically measure the track rotation speed mechanically without human intervention.

[0006] One method for mechanically measuring the track rotation speed is to use the Control Area Network (CAN) data that flows through the network inside the construction machine. However, CAN data depends on the specifications and performance of the construction machine, so not all construction machines can acquire it, and installing a new sensor to acquire CAN data is costly. For this reason, it is considered desirable to use an external sensor that can be easily installed on the construction machine. Examples of external sensors include cameras, sound sensors, and acceleration sensors.

[0007] Among these, cameras and sound sensors are highly dependent on the environment, and require time-consuming advance preparations, such as the need to ensure the stability of the scaffolding at the installation site and the need to prevent external noise unrelated to the construction machinery being measured. Therefore, it is considered desirable to use sensors that are less dependent on the surrounding environment of the construction machinery so that measurements can be made with simple preparations. An acceleration sensor is an example of a desirable sensor in light of the above.

[0008] There are well-known methods for estimating the speed and state of an object based on vibration information of the object. Conventionally, there is a technology that utilizes an acceleration sensor as a method for measuring the vibration frequency of the wheels of a vehicle (Patent Document 1).

[0009] As a method for detecting vibrations when the track is idling in place, rather than while traveling, a method of detecting peak points as frequency features in frequency analysis is considered. Vibrations caused by collisions between track parts as the track rotates are detected, and the frequency of the vibrations is identified by frequency analysis. Patent Document 2 is an example of prior art in other fields that applies frequency analysis to vibration data. [Prior art documents] [Patent documents]

[0010] [Patent Document 1] Patent No. 6877403 [Patent Document 2] Patent No. 5565120 Summary of the Invention [Problem to be solved by the invention]

[0011] In Patent Document 1, a known method such as a Kalman filter is used in frequency analysis for calculating vibrations. This is because the purpose of this technique is to estimate the speed of an object, and the measurement target is limited to a moving vehicle.

[0012] In addition, in Patent Document 2, the peak points are detected based on the value of the inverter's carrier frequency, and the positions at which the peak points occur must be known in advance. The periodic vibrations caused by the rotation of the crawler belt include a fundamental wave, which is a frequency corresponding to the rotation speed, and harmonics, which are integer multiples of the fundamental wave. The positions of the peak points corresponding to the fundamental wave and harmonics in the frequency band change dynamically according to the rotation speed of the crawler belt. Therefore, there is a problem that the positions at which the peak points occur cannot be limited in advance and the correct peak points cannot be identified. In measuring the rotation speed of the crawler belt, the positions and number of peak points in the frequency band change dynamically according to the rotation speed, so there is a problem that the correct peak points cannot be identified.

[0013] Another method for identifying frequency features associated with track vibrations is to identify the frequency feature with the largest amplitude from the frequency analysis results. However, since it depends on the case whether the fundamental wave, which is the frequency corresponding to the track rotation speed, or the harmonic, which is an integer multiple of the fundamental wave, has a larger amplitude, the frequency corresponding to the rotation speed cannot be identified by simply finding the maximum amplitude. If the amplitude of the harmonic is larger, there is a possibility that an inappropriate frequency feature will be identified by mistake.

[0014] The present invention has been made in consideration of the above-mentioned problems, and has an object to provide a performance measurement and diagnosis device that can more appropriately diagnose the condition of a machine based on the results of frequency analysis obtained from machine condition information. [Means for solving the problem]

[0015] An example of a machine performance measurement and diagnosis device according to the present invention includes: A performance measurement and diagnosis device for measuring and diagnosing the performance of a machine, An information acquisition unit that acquires status information of the machine; an information pre-processing unit that pre-processes the status information acquired by the information acquisition unit; a frequency feature discrimination unit that discriminates between state frequency feature candidates that are candidates of frequency features indicative of the state of the machine and frequency features indicative of noise by performing frequency analysis on the data that has been preprocessed by the information preprocessing unit; a frequency feature estimation unit that estimates a frequency feature corresponding to the state of the machine from among the state frequency feature candidates based on the state frequency feature candidates and the frequency feature indicating the noise; a performance measurement and diagnosis unit that identifies features based on the number and positional relationship of frequency features estimated by the frequency feature estimation unit and diagnoses a state of the machine; a display unit that displays the diagnosed state of the machine on a screen; Equipped with. Effect of the Invention

[0016] The performance measurement and diagnosis device according to the present invention can more appropriately diagnose the condition of a machine based on the results of frequency analysis obtained from machine condition information. [Brief description of the drawings]

[0017] [Figure 1] 1 is a diagram showing an example of a machine to which a first embodiment of the present invention can be applied; [Diagram 2] 1 is a schematic configuration diagram of a performance measurement and diagnosis device according to a first embodiment. [Diagram 3] 6 is a graph illustrating criteria used by a frequency characteristic estimating unit to estimate a frequency characteristic corresponding to a state of a construction machine. [Figure 4] 4 is a flowchart illustrating a procedure for the performance measurement diagnosis device to diagnose the condition of a crawler of a construction machine. [Diagram 5] FIG. 11 is a diagram for explaining a specific example of S103. [Figure 6] 10 is a flowchart illustrating a procedure for a frequency feature determination unit to determine a peak point. [Figure 7] FIG. 11 is a diagram for explaining specific examples of S204, S205, and S206. [Figure 8] 5 is a flowchart illustrating a procedure for a performance measurement and diagnosis unit to calculate a rotation speed and diagnose a state. [Figure 9] FIG. 11 is a diagram for explaining specific examples of S302, S305, and S306. [Figure 10] FIG. 11 is a schematic configuration diagram of a performance measurement and diagnosis device according to a second embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0018] Hereinafter, an embodiment of the present invention will be described with reference to the accompanying drawings. <Embodiment 1> Fig. 1 is a diagram showing an example of a machine to which the first embodiment of the present invention can be applied. This embodiment is applicable to a construction machine 50, and the condition of the construction machine 50 is diagnosed by measuring the performance of the construction machine 50. The construction machine 50 includes a component (part) that generates periodic vibrations. This component includes a track 51 (track portion).

[0019] In this embodiment, it is possible to perform measurement and diagnosis regarding periodic vibration in the construction machine 50. Note that in the modified example, the component that generates periodic vibration is not limited to the crawler belt 51. For example, it may be any other component that includes a rotating part.

[0020] The construction machine 50 is merely one example of a work machine to which this embodiment can be applied, and the configuration other than the tracks 51 can be designed as desired by a person skilled in the art, so details will be omitted. Also, while measurement and diagnosis based on periodic vibrations are generally useful in construction machines, machines to which this embodiment can be applied are not limited to construction machines, and may be other construction machines or any work machines other than construction machines. For example, this embodiment can be applied to machines that generate periodic vibrations, such as work machines such as industrial robots and other rotating machines.

[0021] 2 is a schematic diagram showing a machine performance measurement and diagnosis device 1 according to embodiment 1. The performance measurement and diagnosis device 1 is a performance measurement and diagnosis device that measures and diagnoses the performance of a machine, and in this embodiment in particular, is a device for calculating the rotational speed and diagnosing the state of a crawler 51 of a construction machine 50. In this embodiment, the performance measurement and diagnosis device 1 is composed of a mobile terminal 10 that can be placed in any location where a certain degree of stability can be ensured during measurement, such as a cab or a body frame equipped in the construction machine 50.

[0022] The performance measurement diagnosis device 1 includes an acceleration sensor 11 , an information acquisition section 12 , an information preprocessing section 13 , a frequency characteristic discrimination section 14 , a frequency characteristic estimation section 15 , a performance measurement diagnosis section 16 , and a display section 17 .

[0023] Vibrations applied to the performance measurement diagnosis device 1 are measured, for example, as three-axis acceleration, using an acceleration sensor 11 mounted on the mobile terminal 10. By installing the performance measurement diagnosis device 1 at any location within the construction machine 50, the acceleration sensor 11 can indirectly measure vibrations generated by the tracks 51.

[0024] The information acquisition unit 12 acquires status information of the construction machine 50. For example, the information acquisition unit 12 receives information on vibrations caused by collisions between components constituting the track when the track rotates from the acceleration sensor 11, and stores the information as time-series data for each axis. Note that, although the status information of the construction machine 50 is information representing acceleration in this embodiment, in a modified example, it is also possible to use information on a physical quantity other than acceleration.

[0025] The information preprocessing unit 13 preprocesses the state information acquired by the information acquiring unit 12. For example, the information preprocessing unit 13 performs a process of extracting time series data within a time interval corresponding to a rotation interval of the crawler belt 51 from the acquired time series data.

[0026] The frequency feature discrimination unit 14 performs frequency analysis on the data preprocessed by the information preprocessing unit 13 to obtain a frequency spectrum. Then, it captures peak points present on the frequency spectrum as frequency features, and thereby discriminates between frequency feature candidates (state frequency feature candidates) indicating the state of the construction machine 50 and frequency features indicating noise. For example, for all peak points present in a section in the frequency spectrum where features related to the operation of the construction machine 50 occur, it discriminates whether each peak point is a state frequency feature candidate related to the state of the construction machine 50 or a frequency feature indicating other noise.

[0027] Based on the state frequency feature candidates and the frequency feature indicating noise, the frequency feature estimation unit 15 estimates a frequency feature corresponding to the state of the construction machine 50 from among the state frequency feature candidates. For example, a threshold is calculated from the amplitude at the peak point determined to be noise by the frequency feature discrimination unit 14, and based on this threshold, a frequency feature corresponding to the state of the construction machine 50 is estimated from among the state frequency feature candidates.

[0028] The performance measurement diagnosis unit 16 identifies features based on the number and positional relationship of frequency features estimated by the frequency feature estimation unit 15, and diagnoses the condition of the construction machine 50. For example, it identifies a frequency feature corresponding to the rotational speed of the crawler 51, and calculates the rotational speed while the crawler 51 is rotating. Then, based on the calculated rotational speed, it is determined whether the condition of the construction machine 50 is normal or not. In this way, it becomes possible to judge the condition of the construction machine 50 based on concrete grounds.

[0029] The display unit 17 has a screen and displays the diagnosed state of the machine on the screen. For example, the display unit 17 displays the rotation speed of the crawler 51 calculated by the performance measurement diagnosis unit 16 and the judgment result of the state of the construction machine 50 on the mobile terminal. At this time, the frequency spectrum diagram acquired by the frequency feature discrimination unit 14, the discrimination result of the peak point (state frequency feature candidate or noise), the threshold value acquired by the frequency feature estimation unit 15, etc. may be displayed together.

[0030] 3 is an example of a graph illustrating the criteria by which the frequency characteristic estimation unit 15 estimates the frequency characteristic corresponding to the state of the construction machine 50. This graph is a frequency spectrum in which the horizontal axis (X axis) indicates frequency and the vertical axis (Y axis) indicates the energy of each frequency component. By frequency analyzing the vibration data (for example, by converting a time domain representation into a frequency domain representation), it is possible to detect the frequency components contained in the vibration.

[0031] When vibrations are measured while the crawler 51 is rotating, the working devices of the construction machine 50 other than the crawler 51 are stopped, and so it is assumed that the vibration characteristics related to the crawler 51 will be prominent. Therefore, when vibration information is expressed as a frequency spectrum in which each frequency component is expressed as energy, it can be assumed that the frequency characteristics related to the crawler 51 will be prominent. Based on this assumption, in this embodiment, a threshold value is calculated to distinguish between locations where the power of the frequency characteristics is high and locations where this power is not high, and the frequency characteristics related to the crawler 51 are estimated based on the difference from this threshold value.

[0032] 4 is a flow chart for explaining the procedure for diagnosing the condition of the crawler 51 of the construction machine 50 by the performance measurement diagnosis device 1. Each step in the figure will be explained below.

[0033] (Figure 4: Step S101) The worker places the performance measurement diagnostic device 1 in an appropriate place on the construction machine 50. The worker may check whether any of the detection axes of the acceleration sensor 11 is parallel and perpendicular to the horizontal plane. Doing so enables more accurate measurement.

[0034] (Figure 4: Step S102) The worker starts the operation of the construction machine 50. With the construction machine 50 stopped in place, the worker tilts the construction machine 50 so that one of the tracks 51 is suspended in the air. The suspended track 51 is then rotated (idled). The acceleration sensor 11 acquires vibration information as three-axis acceleration from the time the mobile terminal is placed on the construction machine 50 until just after the rotation of the track ends, and the information acquisition unit 12 acquires the information as time-series data.

[0035] (FIG. 4: Step S103) The information preprocessing unit 13 extracts a section in which the crawler belt 51 was rotating from the time-series data acquired in S102. A specific example of this step will be described later.

[0036] (FIG. 4: Step S104) The frequency characteristic discrimination unit 14 performs frequency analysis on the time series data acquired in S103 to acquire a frequency spectrum. At this time, before the frequency analysis, an autocorrelation function may be applied to emphasize the periodicity of the vibration, or a window function may be applied to suppress the influence of the side lobe. Moreover, the information preprocessing unit 13 may perform these processes before the frequency analysis.

[0037] (FIG. 4: Step S105) The frequency characteristic discrimination unit 14 detects all peak points from the frequency spectrum acquired in S104. A peak point is a point whose amplitude is greater than the adjacent points before and after it. A specific example of the process of this step will be described later.

[0038] (FIG. 4: Step S106) The frequency feature discrimination unit 14 treats the frequency spectrum acquired in S105 as a peak point, and discriminates the frequency feature into a state frequency feature candidate indicating the state of the construction machine 50 and a frequency feature indicating noise. Specific processing and examples of this step will be described later.

[0039] (FIG. 4: Step S107) The frequency feature estimation unit 15 calculates a threshold value from the peak points determined to be noise in S106, and estimates features corresponding to the state of the construction machine 50 from the state frequency feature candidates. For example, as shown in FIG. 3, the frequency feature corresponding to the state of the construction machine 50 is estimated based on the difference between the amplitude value of the state frequency feature candidate and the threshold value calculated based on the amplitude of the frequency feature indicating noise. As a specific example, when the difference is 0 or more (i.e., when the amplitude of the state frequency feature candidate is equal to or more than the threshold value calculated based on the amplitude of the frequency feature indicating noise), the state frequency feature candidate is estimated as the frequency feature corresponding to the state of the construction machine 50. According to such processing, the threshold value can be dynamically changed according to the magnitude of the noise. State frequency feature candidates having an amplitude smaller than the threshold value (or equal to or less than a reference value) are eliminated as noise.

[0040] As an example of a method for calculating the threshold, the threshold is set based on the assumption that a feature that is an outlier compared to noise is a feature corresponding to a noteworthy condition of the construction machine 50. Specifically, the average of the amplitudes of peak points determined to be noise + 3σ may be set as the reference value. Note that σ represents the standard deviation of the amplitudes of peak points corresponding to noise. This is because, if the distribution of the amplitudes of peak points determined to be noise is based on a normal distribution, 99% of the noise will be included in the range of average + 3σ or less, and outliers that are not included in this range can be considered to be frequency features corresponding to the condition of the construction machine 50. Note that this calculation method is one example, and the average value or median may simply be set as the threshold.

[0041] It should be noted that a person skilled in the art can appropriately design a process to be performed when no peak point is determined to be noise in S106. For example, the threshold value may be set to 0.

[0042] (FIG. 4: Step S108) The performance measurement diagnosis unit 16 identifies frequency features corresponding to the rotational speed of the crawler belt 51 based on the number and positional relationship of frequency features corresponding to the state of the construction machine 50 estimated in S107, calculates the rotational speed of the crawler belt 51, and diagnoses the state of the construction machine 50. Specific processing and examples of this step will be described later.

[0043] 5(A) and (B) are diagrams for explaining a specific example in S103. The time-series acceleration sensor data acquired in S102 includes both sections where the track is stopped and sections where the track is rotating, as shown in FIG. 5(A). Therefore, it is necessary to extract the sections where the track is rotating from the data.

[0044] When the crawler belt 51 is stationary, the displacement of the acceleration sensor data is close to 0 (i.e., the amplitude is small), and when the crawler belt 51 is rotating, the displacement of the acceleration sensor is large. Therefore, as shown by the thick rectangle in FIG. 5(B), a section where the displacement of the acceleration sensor data is large and continues is extracted as a section where the crawler belt 51 is rotating. Then, time-series acceleration sensor data for the section where the crawler belt 51 is rotating (rotation section) is obtained. At this time, all the data for the rotation section may be used, or a part of the data for the rotation section may be further extracted and used. Furthermore, the section where the amplitude gradually changes at the start and end of the rotation of the crawler belt 51 may be excluded from the rotation section.

[0045] Fig. 6 is a flow chart for explaining specific examples of the processing in S104 to S106. The frequency feature discrimination unit 14 regards peak points present in the frequency spectrum after frequency analysis as frequency features, and discriminates them into state frequency feature candidates indicating the state of the construction machine 50 and frequency features indicating noise. Each step in the figure will be explained below.

[0046] (Figure 6: Step S201) The time series data after the preprocessing in step S103 is subjected to frequency analysis to obtain a frequency spectrum. This step corresponds to step S104 in FIG.

[0047] Steps S202 and S203 correspond to step S105 in FIG. (Figure 6: Step S202) For the frequency spectrum acquired in S201, a predetermined range of the frequency band in which candidates for frequency characteristics indicating the state of the construction machine 50 may occur (hereinafter referred to as the search range) is set as the range for searching for peak points. The search range may be determined based on the results of a prior verification, or may be determined based on the range of the rotational speed to be measured. As an example, in this embodiment, the track rotational speed to be measured is assumed to be within the range of 1 Hz to 12 Hz based on its possible values, so the frequency spectrum is acquired with this range as the search range.

[0048] (Figure 6: Step S203) In S202, all peak points existing within the search range are extracted. Here, a peak point refers to a point whose amplitude value is greater than the adjacent points before and after it.

[0049] Steps S204 to S206 correspond to step S106 in FIG. (FIG. 6: Step S204) Among the multiple peak points (for example, all peak points) extracted in S203, the peak point having the maximum amplitude is determined as a state frequency feature candidate, which is a candidate for a frequency feature indicating the state of the construction machine 50 (state frequency feature candidate determination process). At this time, peak points that have already been determined as state frequency feature candidates or noise are excluded from the search for the peak point having the maximum amplitude.

[0050] (Figure 6: Step S205) Peak points existing within a predetermined range (for example, within a range of 1 Hz each way) before and after the peak point determined as a state frequency feature candidate in S204 are determined as frequency features indicating noise (noise determination process). In this embodiment, feature determination is performed under the condition that the peak points corresponding to the state frequency feature candidates are separated by 1 Hz or more. Therefore, peak points existing within 1 Hz each way before and after the feature candidate are determined as noise. The condition of 1 Hz is determined so that the fundamental frequency and the harmonics have an integer multiple relationship. The lower limit of the search range in this embodiment is 1 Hz, and when the frequency corresponding to the rotation speed of the crawler belt 51 is 1 Hz, which is the smallest in the search range, it is assumed that the harmonics are generated at 1 Hz intervals. In other words, since it is necessary to search for feature candidates at a minimum interval of 1 Hz, the interval before and after the peak point is set to 1 Hz. By using such a range, appropriate determination can be performed for crawlers generally used in construction machines.

[0051] (Figure 6: Step S206) It is determined whether all peak points existing in the search range have been discriminated. If there is an undetermined peak point, the process returns to S204 and continues (S206: No). In the second S204 (additional discrimination process), among the multiple peak points, the peak point having the maximum amplitude is further discriminated as a state frequency feature candidate, excluding the peak points already discriminated as state frequency feature candidates and the peak points discriminated as frequency features indicating noise. Then, in the second S205 (additional discrimination process), the peak points existing within 1 Hz before and after the newly discriminated state frequency feature candidate are discriminated as noise. By repeating such a process of discriminating between state frequency feature candidates and noise, it is possible to appropriately discriminate between multiple state frequency feature candidates. In this way, the frequency feature discrimination unit 14 performs additional state frequency feature candidate discrimination and noise discrimination processes as additional discrimination processes for peak points not discriminated within a predetermined range of the frequency band.

[0052] If all peak points have been identified, the processing in the frequency feature discrimination unit 14 is completed (S206: Yes). Thus, in this embodiment, the additional discrimination processing of S204 and S205 is repeatedly executed until all of the multiple peak points in the search range are discriminated as either state frequency feature candidates or frequency features indicating noise. Then, a set of peak points that are finally discriminated as state frequency feature candidates and a set of peak points that are discriminated as noise are output. By such repetitive processing, it is possible to more appropriately discriminate state frequency feature candidates, and the processing of the above-mentioned step S107 is executed using these discriminated peak points.

[0053] As a modified example, steps S204 and S205 may be repeated a prescribed number of times. In this case, the number of repetitions is sufficient if a certain number of peak points are determined to be either state frequency feature candidates or noise, and therefore, the peak points need only be determined at least once.

[0054] A specific process procedure in S204, S205, and S206 will be described with reference to Fig. 7. Fig. 7 shows a frequency spectrum in which the vertical axis represents the power of the frequency feature and the horizontal axis represents the frequency.

[0055] First, in S204, the peak point with the maximum amplitude is searched for in the search range determined in S202, and peak point p1 is identified as a state frequency feature candidate. The result is shown in FIG. 7(A). Next, in S205, peak point p2 existing within a range of 1 Hz before and after p1 is identified as noise. The result is shown in FIG. 7(B). Then, in S206, it is determined whether all peak points existing in the search range have been identified. Since identification of all peak points has not been completed at the time of FIG. 7(B), the process returns to S204 and the process of identifying state frequency feature candidates and noise is repeated.

[0056] In the second S204, the peak point p3 with the maximum amplitude is searched for among the peak points excluding p1 and p2 for which discrimination has already been completed, and it is discriminated as a state frequency feature candidate. In the subsequent second S205, peak points existing 1 Hz before and after peak point p3 are searched for, but since no peak points are found, noise discrimination processing is not performed. The result after the second S204 and S205 processing is performed is shown in Figure 7(C). In this way, the peak points existing in the search range are discriminated as either state frequency feature candidates or noise.

[0057] Figure 7(D) shows that all peak points in the frequency spectrum were determined to be either state frequency feature candidates or noise, with peak points surrounded by circular frames determined to be state frequency feature candidates and peak points surrounded by square frames determined to be noise.

[0058] Fig. 8 is a flowchart explaining the procedure for the performance measurement diagnosis unit 16 to calculate the rotation speed and diagnose the state. The performance measurement diagnosis unit 16 specifies a frequency corresponding to the rotation speed of the crawler 51 based on the number and positional relationship of frequency features corresponding to the state of the construction machine estimated by the frequency feature estimation unit 15. Each step in Fig. 8 will be explained below.

[0059] (FIG. 8: Steps S301 to S303) The number of frequency features estimated by the frequency feature estimation unit 15 is confirmed. If the number of frequency features is 0 (S301:0), it is determined that the frequency feature corresponding to the state of the construction machine 50 could not be acquired, and the process proceeds to S310 (as a modified example, the process may be terminated). In this case, the display unit 17 may display an instruction to perform measurement again.

[0060] When the number of frequency features estimated by the frequency feature estimation unit 15 is one (S301:1), the state of the construction machine 50 is diagnosed based on the frequency feature. Specifically, the process proceeds to step S302, where the frequency of the frequency feature is output as a frequency corresponding to the rotation speed of the crawler belt 51.

[0061] When a plurality of frequency features are estimated by the frequency feature estimation unit 15 (S301: 2 or more), the state of the construction machine 50 is diagnosed based on the positional relationship of each of the frequency features. Specifically, the process proceeds to step S303, in which the frequency feature having the maximum amplitude is acquired from among the frequency features.

[0062] In this way, by carrying out different processes depending on the number of frequency features estimated by the frequency feature estimating unit 15, more flexible diagnostic processes are possible.

[0063] (FIG. 8: Steps S304 to S306) In step S304, it is determined whether the frequency of the frequency feature acquired in S303 is an integer multiple of the frequency of the other frequency feature. It is not necessary to determine whether the frequency feature is an integer multiple strictly, and a predetermined error range may be taken into consideration. If there are frequency features whose frequencies are an integer multiple of each other (S304: Yes), it can be assumed that the two are in a fundamental wave and harmonic relationship. Therefore, the process proceeds to step S305, in which the smallest frequency among those frequency features is output as the frequency corresponding to the rotational speed of the crawler belt 51. If there are no other frequency features that correspond to the integer multiple (S304: No), the process proceeds to step S306, in which the frequency acquired in S303 (the frequency of the frequency feature with the maximum amplitude) is output as the frequency corresponding to the rotational speed of the crawler belt 51.

[0064] (FIG. 8: Steps S307 to S309) In step S307, the time required for the crawler 51 to rotate once is calculated from the frequency identified in S302, S305, or S306, and this is set as the rotational speed of the crawler 51. The rotational speed is calculated, for example, by the following formula (1). The calculated rotational speed is checked to see if it satisfies the standard performance, and it is determined whether the condition of the construction machine 50 is normal or abnormal. Rotational speed (sec / revolution) = rotation time per revolution (sec) = track links that make up the track (revolutions / revolution) ÷ frequency (revolutions / sec) …(1) If the rotation speed satisfies the standard performance, the process proceeds to step S308 where the condition of the construction machine 50 is diagnosed as normal. If the rotation speed does not satisfy the standard performance, the process proceeds to step S309 where the condition of the construction machine 50 is diagnosed as abnormal.

[0065] (FIG. 8: Step S310) This step does not necessarily have to be executed, but after diagnosing whether the construction machine 50 is normal or not based on the rotational speed of the crawler 51, it may be determined whether other abnormalities (abnormalities other than speed) have occurred. For example, when there are multiple features that are not in an integer multiple relationship with each other, in S306, only the frequency feature with the largest amplitude is treated as corresponding to the rotational speed, and the other frequency features are not treated as corresponding to the rotational speed. When it is known that the frequency feature that was not treated as corresponding to the rotational speed is related to some other abnormality, the performance measurement diagnosis unit 16 may diagnose the cause and possibility of the abnormality based on the frequency feature in S310, and the display unit 17 may present the cause and possibility of the abnormality.

[0066] Fig. 9 is a diagram for explaining specific examples of S302, S305, and S306. The dashed lines in each graph indicate threshold values. Fig. 9(A) is a diagram for explaining a specific example of S302. This diagram shows a case where only one frequency feature corresponding to the state of the construction machine 50 has been estimated. In this case, the estimated frequency F1 is selected as the one corresponding to the rotational speed of the crawler 51 (i.e., the frequency feature corresponding to the state of the construction machine 50).

[0067] 9(B) is a diagram for explaining a specific example in S305. At a frequency corresponding to an integer multiple of the frequency corresponding to the rotational speed of the crawler belt 51, a frequency characteristic may appear as a harmonic. Therefore, depending on the rotational speed, multiple peak points may appear in the search range, including a frequency F2 corresponding to the rotational speed of the crawler belt 51 and a frequency F3 corresponding to the harmonic. In this way, when multiple frequency characteristics having an integer multiple relationship exist, the smaller frequency F2 is selected as the one corresponding to the rotational speed (i.e., the frequency characteristic corresponding to the state of the construction machine 50).

[0068] Fig. 9(C) is a diagram for explaining a specific example in S306. There are cases where noise that could not be completely eliminated by the frequency feature estimation unit 15 or frequency features due to vibrations of other components of the construction machine 50 appear (e.g., frequency F5 in Fig. 9(C)). Since vibrations that occur during rotation of the crawler belt 51 are mainly specific to the crawler belt, it is expected that the frequency features will be larger than other features. Therefore, the frequency feature with the largest amplitude is obtained, and the frequency F4 is selected as the one corresponding to the rotation speed (i.e., the frequency feature corresponding to the state of the construction machine 50).

[0069] <Embodiment 1: Summary> The performance measuring and diagnosing device 1 according to the first embodiment estimates the frequency feature corresponding to the state of the construction machine 50 by distinguishing between candidates of frequency features (state frequency feature candidates) indicating the state of the construction machine 50 and features corresponding to noise from the frequency analysis results acquired from vibration information caused by the crawler 51, and identifies the frequency corresponding to the rotational speed of the crawler 51. By dynamically distinguishing between state frequency feature candidates and noise, the rotational speed of the crawler 51 can be calculated more accurately without depending on the position of the frequency corresponding to the rotational speed or the number of its harmonics, and this makes it possible to more appropriately diagnose the state of the construction machine 50. Furthermore, in the first embodiment, measurements are performed in a state in which components other than the crawler 51 of the construction machine 50 are stationary, and the running performance can be evaluated without actually running the construction machine 50.

[0070] <Embodiment 2> In the second embodiment of the present invention, as shown in Fig. 10, the performance measurement diagnosis device 1 comprises a mobile terminal 10 and a cloud 20. An acceleration sensor 11, an information acquisition unit 12, and a display unit 17 are provided on the mobile terminal 10, and an information preprocessing unit 13, a frequency characteristic discrimination unit 14, a frequency characteristic estimation unit 15, and a performance measurement diagnosis unit 16 are provided on the cloud 20. The mobile terminal 10 and the cloud 20 may be provided with a configuration for transmitting and receiving information to each other via wireless communication. In the above points, the second embodiment differs from the above-mentioned first embodiment, but the other configurations are the same as those of the first embodiment.

[0071] Since the mobile terminal 10 and the cloud 20 are configured to be able to communicate with each other, for example, acquired sensor data can be uploaded to the cloud 20 and frequency analysis and performance measurement diagnosis can be performed on the cloud 20. Then, the performance measurement diagnosis results calculated on the cloud 20 can be transmitted to the mobile terminal 10 and can be confirmed on the display unit 17 of the mobile terminal 10.

[0072] The performance measurement diagnosis device 1 configured as above can obtain the same effects as those of the above-mentioned embodiment 1. Furthermore, since the information preprocessing unit 13, the frequency feature discrimination unit 14, the frequency feature estimation unit 15, and the performance measurement diagnosis unit 16 are provided on the cloud 20, it becomes possible to perform performance measurement diagnosis that is not dependent on the performance of the mobile terminal 10, as compared with the case where all the processing is performed on the mobile terminal 10. Furthermore, it becomes possible to perform performance measurement diagnosis and check the results at any time.

[0073] <Modifications of each embodiment> In each of the above-described embodiments and the modified examples thereof, the performance measurement diagnosis device 1 diagnoses the state of the construction machine 50 (for example, whether it is normal or not) based on the rotational state of the crawler 51, but it is also possible to diagnose the state of the components of the construction machine 50. For example, when used as a device for diagnosing the state of the crawler 51, it is also possible to interpret it as calculating the rotational speed or the like as a physical quantity that represents the state of the crawler 51.

[0074] The present invention is not limited to the above-described embodiments and their modifications, but includes various other modifications. For example, the above-described embodiments have been described in detail to clearly explain the present invention, and are not necessarily limited to those having all of the configurations described. In addition, it is possible to replace a part of the configuration of one embodiment with the configuration of another embodiment, and it is also possible to add the configuration of another embodiment to the configuration of one embodiment. In addition, it is possible to add, delete, or replace a part of the configuration of each embodiment with another configuration. [Explanation of symbols]

[0075] 1...Performance measurement and diagnosis device 10. Mobile terminals 11...Accelerometer 12…Information acquisition department 13...Information preprocessing section 14...Frequency characteristic discrimination section 15...Frequency characteristic estimation unit 16…Performance Measurement and Diagnosis Department 17...Display section 20. Cloud 50…Construction machinery 51…Track F1…Frequency (frequency characteristics corresponding to the machine condition) F2: Frequency (frequency characteristics corresponding to the machine condition) F3…Frequency F4…Frequency (frequency characteristics corresponding to the machine condition) F5…Frequency p1: Peak point (state frequency feature candidate) p2: Peak point (noise) p3: Peak point (state frequency feature candidate)

Claims

1. A performance measurement and diagnosis device for measuring and diagnosing the performance of a machine, An information acquisition unit that acquires status information of the machine; an information pre-processing unit that pre-processes the status information acquired by the information acquisition unit; a frequency feature discrimination unit that discriminates between state frequency feature candidates that are candidates of frequency features indicative of the state of the machine and frequency features indicative of noise by performing frequency analysis on the data that has been preprocessed by the information preprocessing unit; a frequency feature estimation unit that estimates a frequency feature corresponding to the state of the machine from among the state frequency feature candidates based on the state frequency feature candidates and the frequency feature indicating the noise; a performance measurement and diagnosis unit that identifies features based on the number and positional relationship of frequency features estimated by the frequency feature estimation unit and diagnoses a state of the machine; a display unit that displays the diagnosed state of the machine on a screen, The performance measurement and diagnosis unit includes: When the number of frequency features estimated by the frequency feature estimation unit is one, diagnosing the state of the machine based on the frequency of the frequency feature; when the frequency features estimated by the frequency feature estimation unit are multiple and the frequency of the frequency feature having the maximum amplitude among the multiple frequency features is an integer multiple of the frequency of the other frequency features, diagnosing the state of the machine based on the minimum frequency among the multiple frequency features; a frequency characteristic estimation unit that estimates a frequency characteristic of a machine based on a frequency characteristic having a maximum amplitude when the frequency characteristic estimated by the frequency characteristic estimation unit is multiple and the frequency characteristic having a maximum amplitude among the multiple frequency characteristics is not an integer multiple of the frequencies of the other frequency characteristics;

2. 2. The machine performance measurement and diagnosis device according to claim 1, wherein said information acquisition section acquires, as the machine status information, status information of a component generating periodic vibration.

3. 3. The machine performance measurement and diagnosis device according to claim 2, The component includes a track of a construction machine; The performance measurement diagnosis unit calculates a rotational speed while the crawler portion is rotating. Machine performance measurement and diagnostic equipment.

4. 2. The machine performance measurement and diagnosis device according to claim 1, The frequency characteristic determination unit is a state frequency feature candidate discrimination process for discriminating a peak point having a maximum amplitude from among a plurality of peak points included within a predetermined range in a frequency band as the state frequency feature candidate; a noise discrimination process for discriminating, as a frequency feature indicating noise, a peak point existing within a predetermined range around the peak point discriminated as a state frequency feature candidate; Targeting undistinguished peak points within a predetermined range of the frequency band, and an additional discrimination process for performing the additional state frequency feature candidate discrimination process and the noise discrimination process.

5. 5. A machine performance measurement and diagnosis device according to claim 4, wherein the frequency feature discrimination unit repeatedly executes the additional discrimination process until all of the plurality of peak points are discriminated to be either the state frequency feature candidates or the frequency features indicating noise.

6. 2. The machine performance measurement and diagnosis device according to claim 1, wherein the frequency feature estimation unit estimates a frequency feature corresponding to the state of the machine based on a difference between an amplitude value of the state frequency feature candidate and a threshold value calculated based on an amplitude of the frequency feature indicating the noise. Machine performance measurement and diagnostic equipment.

7. 5. A machine performance measurement and diagnosis apparatus according to claim 4, wherein said predetermined range is a range of 1 Hz each before and after said peak point.

Images