Load data analysis method, device, and program

Abstract

To appropriately realize a reduction in memory capacity, a reduction in calculation time, and a real-time analysis when analyzing a load applied to an object based on the rain-flow method.SOLUTION: A load data analysis method for analyzing load data indicating a load repeatedly and irregularly applied to an object based on the rain-flow method includes: a first step of acquiring a predetermined amount of load data, and calculating the frequency of the load amplitude and load average related to the load applied to the object using load data that satisfies a predetermined condition in the predetermined amount of load data based on the rain-flow method; a second step of storing load data that did not satisfy the predetermined condition among the predetermined amount of load data; and a third step of generating the predetermined amount of load data for executing the first step by combining the load data stored by the second step with a newly acquired load data. The first to third steps are repeatedly executed.SELECTED DRAWING: Figure 15

Description

[Technical Field]

[0001] The present invention relates to a load data analysis method, device, and program for analyzing load data indicating a load applied to an object. [Background technology]

[0002] A technique has been known in the past that uses the rainflow method to analyze load data indicating loads (weight, stress, etc.) repeatedly and irregularly applied to objects such as various machines and buildings. This technique analyzes the load data using the rainflow method to calculate the frequency of the amplitude of the waveform of the load applied to the object and the average value of the load (hereinafter referred to as the "load amplitude" and "load average"). Typically, the damage value of an object due to the load applied to the object (in other words, the material damage level or fatigue damage level) is calculated based on the frequency of the load amplitude and load average calculated in this way. For example, Patent Document 1 discloses a technique for estimating the cause of deterioration of a specified device during operation based on data acquired in a deterioration section where the device is subject to deterioration. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] International Publication No. 2016 / 016956 Summary of the Invention [Problem to be solved by the invention]

[0004] However, in conventional techniques, all load data (i.e., all time-series load waveforms) over a long period of time is collected and stored, and then this all time-series load waveform is analyzed using the rainflow method. Therefore, a large memory capacity is required to record the huge amount of load data, and it takes a long time to process the huge amount of load data in a batch. Furthermore, such conventional techniques can only analyze load data collected over a long period of time, and cannot analyze load data acquired every moment in real time.

[0005] The present invention has been made to solve the problems of the conventional technology described above, and aims to provide a load data analysis method, device, and program that can reduce memory capacity, shorten calculation time, and appropriately realize real-time analysis when analyzing load data that indicates the load applied to an object based on the rainflow method. [Means for solving the problem]

[0006] In order to achieve the above object, the present invention provides a load data analysis method for analyzing load data indicating a load that is applied irregularly and repeatedly to an object based on the rainflow method, the method comprising: a first step of acquiring a predetermined amount of load data; and calculating a frequency related to the amplitude of the waveform of the load applied to the object (load amplitude) and / or an average value of the load (load average) using load data that satisfies a predetermined condition among the predetermined amount of load data based on the rainflow method; a second step of storing load data that does not satisfy the predetermined condition among the predetermined amount of load data; and a third step of generating a predetermined amount of load data for executing the first step by combining the load data stored in the second step with newly acquired load data, wherein the first to third steps are repeatedly executed; The second step deletes a portion of the load data from the predetermined amount of load data when the load data that does not satisfy the predetermined condition is equal to or greater than the limit amount so that the load data to be stored is less than the limit amount, and stores the load data remaining after this deletion; the second step deletes load data acquired later in time from the predetermined amount of load data when the load data that does not satisfy the predetermined condition is equal to or greater than the limit amount so that the load data to be stored is less than the limit amount; It is characterized by:

[0007] The present invention, configured as described above, processes only a predetermined amount of load data rather than collecting and processing load data over a long period of time. This allows for reduced memory capacity and shorter calculation times when analyzing load data based on the Rainflow method. Furthermore, according to the present invention, load data that meets a predetermined condition is used to calculate frequency, while load data that does not meet the predetermined condition is combined with new load data to create a new predetermined amount of load data, which is then used to calculate the next frequency. This process is repeated, allowing for real-time analysis of load data acquired from moment to moment while maintaining the same calculation accuracy as the conventional Rainflow method. As described above, according to the present invention, when analyzing load data based on the Rainflow method, it is possible to appropriately achieve reduced memory capacity (memory saving), shorter calculation times (high-speed calculation), and real-time analysis while maintaining the same calculation accuracy as the conventional Rainflow method. Furthermore, according to the present invention, by maintaining the number of load data stored in the second step below the limit amount, it is possible to reliably generate a new predetermined amount of load data from newly acquired load data in the third step. In particular, according to the present invention, load data acquired later in time that has a relatively small load amplitude is deleted from the predetermined amount of load data, thereby minimizing the effect (error) caused by deleting the load data. In another aspect, a load data analysis method for analyzing load data indicating a load that is applied irregularly and repeatedly to an object based on a rainflow method includes a first step of acquiring a predetermined amount of load data, and calculating a frequency related to the amplitude of a waveform of the load applied to the object and / or an average value of the load based on the rainflow method, using load data that satisfies a predetermined condition among the predetermined amount of load data; a second step of storing load data that does not satisfy the predetermined condition among the predetermined amount of load data; and a third step of generating a predetermined amount of load data for executing the first step by combining the load data stored in the second step with newly acquired load data, wherein the first to third steps are repeatedly executed, and the load data is calculated based on the first step. The method further includes a fourth step of calculating peak data corresponding to a point where the change in data switches from an increase to a decrease or from a decrease to an increase, and the first to third steps use the peak data calculated in the fourth step as the load data to be processed in each step, and the third step overwrites the first peak data with the second peak data if the relationship between the first peak data, which is the latest in time among the peak data corresponding to the load data stored in the second step, and the second peak data, which is the earliest in time among the peak data corresponding to the newly acquired load data, does not satisfy the relationship that the peak data alternately increases and decreases, thereby generating a predetermined amount of load data for executing the first step. According to the present invention configured in this manner, when the load data stored in the second step is combined with the newly acquired load data, a new predetermined amount of load data can be appropriately generated such that the load data (peak data) alternately increases and decreases, thereby enabling the rainflow method to be applied appropriately.

[0008] In the present invention, preferably, in the second step, when the load data that does not satisfy the predetermined conditions among the predetermined amount of load data is equal to or greater than the limit amount, some of the load data from the predetermined amount of load data is deleted so that the load data to be stored is less than the limit amount, and the load data remaining after this deletion is stored. According to the present invention configured in this manner, by maintaining the number of load data stored in the second step below the limit amount, it is possible to reliably generate a new predetermined amount of load data in the third step using newly acquired load data.

[0009] In the present invention, preferably, the second step is to delete load data acquired later in time from the predetermined amount of load data so that the load data to be stored is less than the limit amount when the load data that does not satisfy the predetermined condition is equal to or greater than the limit amount from the predetermined amount of load data. According to the present invention configured in this manner, among a predetermined amount of load data, load data that has a relatively small load amplitude and was acquired later in time is deleted, thereby minimizing the effect (error) caused by deleting the load data.

[0010] In the present invention, preferably, the method further includes a fourth step of calculating peak data corresponding to a point where the change in the load data switches from an increase to a decrease or from a decrease to an increase, and the first to third steps use the peak data calculated in the fourth step as the load data to be processed in each step, and the third step overwrites the first peak data with the second peak data if the relationship between the first peak data that is the latest in time among the peak data corresponding to the load data stored in the second step and the second peak data that is the earliest in time among the peak data corresponding to the newly acquired load data does not satisfy the relationship that the peak data alternately increases and decreases, thereby generating a predetermined amount of load data for executing the first step. According to the present invention configured in this manner, when the load data stored in the second step is combined with the newly acquired load data, a new predetermined amount of load data can be appropriately generated such that the load data (peak data) alternately increases and decreases, thereby enabling the rainflow method to be applied appropriately.

[0011] In the present invention, preferably, the method further includes a fourth step of calculating peak data corresponding to a point where the change in the load data switches from an increase to a decrease or from a decrease to an increase, and the first to third steps use the peak data calculated in the fourth step as the load data to be processed in each step, and the first and second steps use, as a predetermined condition, a condition that the difference between the second peak data and the third peak data calculated in the fourth step, the subsequent second peak data, and the subsequent third peak data, is equal to or greater than the difference between the first peak data and the second peak data. According to the present invention configured as described above, load data with a relatively large amplitude can be reliably applied to the calculation of frequencies related to the load amplitude and the load average.

[0012] The present invention preferably further comprises a fifth step of calculating a damage value of the object due to the load applied to the object, based on at least the frequency calculated in the first step. According to the present invention configured as described above, it is possible to calculate highly accurate damage values ​​in real time while ensuring reduction in memory capacity and calculation time.

[0013] The present invention is preferably performed by a processing device mounted on the vehicle, which analyzes load data corresponding to the loads applied to the vehicle components. According to the present invention configured as described above, even with a relatively small memory capacity such as that used in a vehicle, it is possible to accurately analyze the load data applied to the vehicle parts based on the rainflow method. Furthermore, it is possible to analyze the load data applied to the vehicle parts in real time.

[0014] The present invention preferably further comprises a sixth step of displaying the frequency calculated in the first step or information related to the frequency. According to the present invention configured in this manner, it is possible to appropriately inform the user of the frequency of the load amplitude and load average, and information related to this frequency (for example, damage value and information corresponding to the damage value).

[0015] In another aspect, to achieve the above object, the present invention provides a load data analysis device that analyzes load data corresponding to a load that is applied irregularly and repeatedly to an object based on a rainflow method, the load data analysis device comprising: first means for acquiring a predetermined amount of load data, and calculating a frequency related to an amplitude of a waveform of the load applied to the object and / or an average value of the load based on the rainflow method, using load data that satisfies a predetermined condition among the predetermined amount of load data; second means for storing load data that does not satisfy the predetermined condition among the predetermined amount of load data; and third means for generating a predetermined amount of load data for executing the first means by combining the load data stored by the second means with newly acquired load data, and each of the first to third means repeatedly executes processing. death, the second means, when the load data that does not satisfy the predetermined condition among the predetermined amount of load data is equal to or greater than the limit amount, deletes a portion of the load data from the predetermined amount of load data so that the load data to be stored is less than the limit amount, and stores the load data remaining after this deletion; and the second means, when the load data that does not satisfy the predetermined condition among the predetermined amount of load data is equal to or greater than the limit amount, deletes load data acquired later in time from the predetermined amount of load data so that the load data to be stored is less than the limit amount; It is characterized by:

[0016] In still another aspect, to achieve the above object, the present invention provides a load data analysis program executed by a computer device for analyzing load data corresponding to a load that is applied irregularly and repeatedly to an object based on the rainflow method, the program causing the computer device to function as: first means for acquiring a predetermined amount of load data, and calculating a frequency related to the amplitude of a waveform of the load applied to the object and / or an average value of the load based on the rainflow method, using load data that satisfies a predetermined condition among the predetermined amount of load data; second means for storing load data that does not satisfy the predetermined condition among the predetermined amount of load data; and third means for generating a predetermined amount of load data for executing the first means by combining the load data stored by the second means with newly acquired load data; and causing the computer device to function so that each of the first to third means repeatedly executes processing; The second means causes the computer device to function in such a way that, when the load data that does not satisfy the predetermined condition among the predetermined amount of load data is equal to or greater than a limit amount, the second means deletes some of the load data from the predetermined amount of load data so that the load data to be stored is less than the limit amount, and stores the load data remaining after this deletion; and when the load data that does not satisfy the predetermined condition among the predetermined amount of load data is equal to or greater than the limit amount, the second means deletes load data that is acquired later in time from the predetermined amount of load data so that the load data to be stored is less than the limit amount. It is characterized by:

[0017] The load data analysis device and load data analysis program of the present invention configured in this manner can also appropriately reduce memory capacity, shorten calculation time, and perform real-time analysis when analyzing load data based on the rainflow method, while maintaining calculation accuracy equivalent to that of the conventional rainflow method. [Effects of the Invention]

[0018] According to the load data analysis method, device, and program of the present invention, when load data indicating the load applied to an object is analyzed based on the Rainflow method, it is possible to appropriately reduce memory capacity, shorten calculation time, and perform real-time analysis. [Brief explanation of the drawings]

[0019] [Figure 1] 1 is a schematic configuration diagram of a computer device that is an example of an entity that executes a load data analysis method according to an embodiment of the present invention. [Figure 2]FIG. 10 is an explanatory diagram of a basic calculation method of a damage value according to a load amplitude. [Figure 3] FIG. 10 is an explanatory diagram of a more specific calculation method (Minor's rule) of the damage value according to the load amplitude. [Figure 4] 10 shows an example of a histogram of load amplitude and load average obtained for an actual time-series load waveform. [Figure 5] FIG. 1 is an explanatory diagram of the load average, load amplitude, etc. used in the rainflow method. [Figure 6] FIG. 10 is an explanatory diagram of a method for processing a small loop in the Rainflow method. [Figure 7] FIG. 1 is an explanatory diagram of a method for processing a stopping wave in the rainflow method. [Figure 8] FIG. 1 is an explanatory diagram of peak data generation performed in the Rainflow method. [Figure 9] FIG. 1 is an explanatory diagram of small loop recovery performed in the rainflow method. [Figure 10] FIG. 1 is an explanatory diagram of stopping wave recovery performed in the rainflow method. [Figure 11] FIG. 1 is an explanatory diagram of why undetermined waves should not be collected as they are in the rainflow method. [Figure 12] FIG. 1 is an explanatory diagram of a calculation image of a load data analysis method according to an embodiment of the present invention. [Figure 13] 1 shows a specific example of load data in a calculation domain used in an embodiment of the present invention. [Figure 14] FIG. 10 is an explanatory diagram of small loop recovery and stop wave recovery for load data within a calculation domain in an embodiment of the present invention. [Figure 15] FIG. 10 is an explanatory diagram of inserting new load data into a free area in a calculation area in the embodiment of the present invention. [Figure 16] A specific example of the number of uncertain waves generated for each step in the embodiment of the present invention will be shown below. [Figure 17] FIG. 10 is an explanatory diagram illustrating partial deletion of load data of an uncertain wave in an embodiment of the present invention. [Figure 18]FIG. 10 is an explanatory diagram of a problem that occurs when an uncertain wave and new load data are connected. [Figure 19] FIG. 10 is an explanatory diagram of overwriting load data of an uncertain wave with new load data in the embodiment of the present invention. [Figure 20] 1 is a flowchart illustrating a method for analyzing load data according to an embodiment of the present invention. [Figure 21] 10 shows example results for memory usage and damage value error according to an embodiment of the present invention. [Figure 22] 10 shows example results in terms of memory usage and computation time according to an embodiment of the present invention. [Figure 23] 1 shows example results regarding damage values ​​according to an embodiment of the present invention and a conventional rainflow method. DETAILED DESCRIPTION OF THE INVENTION

[0020] Hereinafter, a load data analysis method, device, and program according to an embodiment of the present invention will be described with reference to the accompanying drawings.

[0021] <Computer equipment> First, a computer device that is an example of an entity that executes a load data analysis method according to an embodiment of the present invention will be described with reference to Fig. 1. As shown in Fig. 1, the computer device 10 mainly includes an input device 1 into which information is input by a user or the like, a processing device 3 that processes various types of information, and an output device 5 that outputs information.

[0022] The input device 1 is, for example, a mouse, keyboard, touch panel, microphone, etc., and the output device 5 is, for example, a display device, speaker, etc. The processing device 3 has one or more microprocessors 3a as a central processing unit (CPU) that executes programs, and memory 3b that is configured, for example, by RAM (Random Access Memory), ROM (Read Only Memory), hard disk, etc., and that stores programs and data.

[0023] A load sensor 7 that detects a load that is repeatedly and irregularly applied to a predetermined object is connected to the computer device 10, and load data corresponding to the load detected by this load sensor 7 is input to the computer device 10 (particularly the processing device 3). The load sensor 7 detects a load that corresponds to a load or stress applied to the predetermined object. For example, the load sensor 7 is configured by a strain gauge attached to the predetermined object.

[0024] The load data analysis method according to this embodiment is executed by the processing device 3 of the computer device 10. Specifically, a program corresponding to this load data analysis method (a load data analysis program of the present invention) is stored in the memory 3b of the processing device 3, and the microprocessor 3a of the processing device 3 reads out and executes this program from the memory 3b. This causes the computer device 10 (particularly the processing device 3) to function as the load data analysis device of the present invention.

[0025] In this embodiment, the computer device 10 (processing device 3) acquires load data indicating loads that are repeatedly and irregularly applied to objects such as various machines and buildings from the load sensor 7, and analyzes the load data based on the rainflow method. Specifically, in this embodiment, the computer device 10 analyzes such load data based on the rainflow method to calculate frequencies related to the load amplitude and load average of the waveform of the load applied to the object, and calculates damage values ​​(material damage level, fatigue damage level) of the object due to the load applied to the object based on the frequencies of the load amplitude and load average.

[0026] For example, the computer device 10 is configured by a processing device (such as an ECU (Electronic Control Unit)) mounted on a vehicle, and analyzes the load applied to a predetermined part of the vehicle. In one example, the computer device 10 analyzes the load applied to a gear case of a transmission to which engine torque is input, or a motor case of an EV vehicle. In this case, a strain gauge serving as the load sensor 7 described above is attached to the part of the vehicle to which the load analyzed by the computer device 10 is applied.

[0027] <Calculation method for damage value> Next, a general method for calculating damage values ​​will be described with reference to Figs. 2 to 4. First, Fig. 2 is an explanatory diagram of a basic method for calculating damage values ​​according to load amplitude. The left side of Fig. 2 shows the time change of the load (stress) applied to a specified object, that is, the time-series load waveform. Here, an example is shown in which a load with an amplitude of 100 MPa is applied to the object 60,000 times, that is, a load amplitude of 100 MPa occurs 60,000 times. The right side of Fig. 2 shows an SN diagram that shows the relationship between load amplitude and fatigue life (material properties). According to this SN diagram, for example, at a load amplitude of 100 MPa, the damage value is 2 × 10 5 This indicates that the material will break (fracture) in 60,000 times. Here, as mentioned above, if a load amplitude of 100 MPa occurs 60,000 times, the damage value D is "D = 60,000 / 2 × 10 5 =0.3". This indicates that the material will break when the damage value D reaches 1, so it is 30% of the way to breaking.

[0028] Next, Figure 3 is an explanatory diagram of a more specific method of calculating the damage value according to the load amplitude (Minor's rule). The left side of Figure 3 shows a time-series load waveform. Here, an example is shown in which a load amplitude of 80 MPa occurs 7,000 times and a load amplitude of 120 MPa occurs 5,000 times. The right side of Figure 3 shows an SN diagram that shows the relationship between load amplitude and fatigue life (material properties). According to this SN diagram, for example, at a load amplitude of 80 MPa, the damage value is 5 x 10 5 The material broke in 2×10 times at a load amplitude of 120 MPa. 5Here, as mentioned above, if a load amplitude of 80 MPa occurs 7000 times and a load amplitude of 120 MPa occurs 5000 times, the damage value D due to a load amplitude of 80 MPa occurring 7000 times is A "D A =70000 / (5×10 5 ) = 0.014, and the damage value D due to a load amplitude of 120 MPa that occurred 5,000 times B "D B =50000 / (2×10 5 )=0.025". Therefore, the total (final) damage value D at this time is "D A +D B =0.014+0.025" gives "0.039". This method of calculating the damage value is based on Minor's rule.

[0029] As explained in Figure 3, even when a load with a complex waveform is applied, it is possible to calculate the damage value by analyzing the load waveform and determining the load amplitude, load average, and number of cycles (i.e., frequency). For simplicity's sake, Figures 2 and 3 show an example in which the load average is nearly 0 (i.e., an example in which the positive peak value and the negative peak value in the load waveform are nearly the same), but in an actual load waveform, the load average will not be 0. Since this load average affects the damage value separately from the load amplitude, it is necessary to determine the load average when calculating the damage value. Typically, the load average is the median value between adjacent peak values ​​(the higher peak value and the lower peak value) in the load waveform.

[0030] Figure 4 shows an example of the frequency (number of cycles) of load amplitude and load average obtained from an actual time-series load waveform from this perspective. The left side of Figure 4 shows an example of a waveform (time-series load waveform) of a load applied to an object over a long period of time, and the right side of Figure 4 shows histograms representing the frequency of load amplitude and load average obtained by analyzing this time-series load waveform. Specifically, the histograms of load amplitude and load average are obtained by analyzing the time-series load waveform using the rainflow method (details will be described later). From the histograms of load amplitude and load average obtained in this way, the total damage value can be calculated from the number of cycles for various load amplitudes and load averages and the SN diagram based on the Minor's law described above.

[0031] <Rainflow method> Next, the basic concept of a general rainflow method will be explained with reference to FIGS. 5 to 7. First, FIG. 5 is an explanatory diagram of the load average, load amplitude, and the like used in the rainflow method. FIG. 5 shows approximately 0.5 cycles of a load waveform applied to an object. The symbols PV1 and PV2 are values ​​(peak values) corresponding to the peaks of the load waveform (points where the change in the load waveform switches from an increase to a decrease or from a decrease to an increase). Also, 0.5 cycles indicates the interval between adjacent peak values ​​PV1 and PV2. In the rainflow method, calculations are performed based on the peak values ​​PV1 and PV2 of the load waveform. In this case, the load average is defined as "(PV1 + PV2) / 2," and the load amplitude (particularly, the load half amplitude) is defined as "r1 / 2" using the difference r1 between the peak values ​​PV1 and PV2 (hereinafter referred to as the "PV difference"). r1 = |PV1 - PV2| is used.

[0032] Next, FIG. 6 is an explanatory diagram of a method for processing small loops present in a load waveform in the Rainflow method. This FIG. 6 shows a load waveform with multiple peaks. This load waveform contains a small loop (0.5 cycle) in which the load changes slightly (reverses) within a loop (0.5 cycle) in which the load changes significantly. The Rainflow method does not ignore such small loops but uses them in the calculation, i.e., "recovers" them. This ensures high calculation accuracy. Here, in this specification, the term "recover" is used to mean, in the Rainflow method, updating (adding) the histograms of the load amplitude and load average, as well as updating (adding) the damage value based on the load waveform data (i.e., load data, specifically peak data), and removing the data after these processes.

[0033] Furthermore, the condition for establishing a small loop (hereinafter referred to as the "small loop condition") is defined as "|r1|>|r2|≦|r3|" based on the three adjacent PV differences r1, r2, and r3 in the load waveform. In the example shown in Figure 6, the PV difference r2 satisfies the small loop condition, so it is recovered. Furthermore, r2 and r3 are removed during this recovery; specifically, the process "|r1|←|r3|-|r2|+|r1|" is performed.

[0034] Next, Figure 7 is an explanatory diagram of the stop wave in the rainflow method. This figure also shows a load waveform with multiple peaks, specifically, a load waveform with peak values ​​PV1, PV2, PV3, and PV4. In this load waveform, the amplitude from peak value PV3 to peak value PV4 is larger than the amplitude from peak value PV1 to peak value PV4. Therefore, the latter amplitude must be fully recovered. To achieve this, the rainflow method prioritizes recovery of the larger amplitude, and stops the amplitude from peak value PV1 at peak value PV2. Specifically, in the rainflow method, the load wave from peak value PV1 to peak value PV2 is defined as the stop wave, and the recovery process described above is performed at this stop wave. In other words, the stop wave is specified to ensure the recovery process. In this case, the condition for the stop wave to be met (hereinafter referred to as the "stop wave condition") is defined as "|r2|≧|r1|" based on the PV differences r1 and r2 between two adjacent load waveforms.

[0035] Next, the specific calculation procedure of the general rainflow method will be explained with reference to Figures 8 to 10. Figure 8 shows an explanatory diagram of the generation of peak data, which is the first step in the rainflow method, Figure 9 shows an explanatory diagram of the small loop recovery, which is the next step in the rainflow method, and Figure 10 shows an explanatory diagram of the stop wave recovery, which is the next step in the rainflow method.

[0036] First, as shown in FIG. 8, peak values ​​are searched for in the raw data of continuously acquired load data (time-series load waveform), and data consisting of peak values ​​(peak data) is generated. This peak data is generated by searching for points (black circles in FIG. 8) where the load waveform changes from increasing to decreasing or from decreasing to increasing. Next, as shown in FIG. 9, small loops are identified for the obtained peak data (FIG. 8) using the small loop condition "|r1|>|r2|≦|r3|" for the three adjacent PV differences r1, r2, and r3. These small loops are then collected (including removed) sequentially. Specifically, collection of small loops is repeated until there are no small loops in the load waveform. Next, as shown in FIG. 10, for the data from which small loops no longer exist (FIG. 9), stop waves are identified using the stop wave condition "|r2|≧|r1|" for the two adjacent PV differences r1 and r2. These stop waves are then collected (including removed) sequentially.

[0037] Here, the wave surrounded by the dashed line area in Figure 10 does not satisfy the stop wave condition, i.e., the difference between the two adjacent PVs r1 and r2 is "|r2|<|r1|", so this wave will remain without being recovered. Hereinafter, such a wave that remains without being recovered will be referred to as an "undetermined wave." In contrast, a wave that is recovered because it is a stop wave will be referred to as a "determined wave." Since an undetermined wave is a wave that occurs after a small loop recovery, its amplitude is relatively large. Therefore, it can be said that a wave that has a large impact on intensity will remain without being recovered. While it is possible to immediately recover such a remaining undetermined wave with its waveform, this is not desirable. The reason for this will be explained with reference to Figure 11.

[0038] Figure 11 explains why undetermined waves should not be collected directly in the rainflow method. Figure 11 shows a load waveform obtained over a relatively long period of time, specifically, a waveform after small-loop collection of peak data obtained from load data. This example illustrates a case in which an undetermined wave remains uncollected for a long period of time because a large load was not applied for a long period of time. Specifically, a load that satisfies the stop-wave condition (|r2| ≥ |r1|) between two adjacent PV differences r1 and r2 did not appear for a long period of time. This undetermined wave is collected after a long period of time when a stop wave, as indicated by arrows A11 and A12, appears, and is correctly evaluated. However, if the undetermined wave is collected using a stop wave (a wave preceding the undetermined wave), as indicated by arrows A13 and A14, it will be underestimated. Therefore, it is recommended that an undetermined wave not be collected immediately after its occurrence, but rather that an appropriate stop wave (a wave that satisfies the stop-wave condition) appears and then collected using this stop wave.

[0039] In the conventional rainflow method, all load data over a long period of time is collected and temporarily stored before the load data is analyzed. Therefore, a large memory capacity is required to record the huge amount of load data, and the calculation time is long because the huge amount of load data must be processed in bulk. Furthermore, such conventional rainflow methods can only analyze load data collected over a long period of time, and cannot analyze load data acquired every moment in real time. Therefore, in this embodiment, when analyzing load data based on the rainflow method, it is intended to appropriately achieve reduced memory capacity (memory saving), shorter calculation time (high-speed calculation), and real-time analysis while maintaining the same calculation accuracy as the conventional rainflow method.

[0040] <Load Data Analysis Method According to This Embodiment> Next, a load data analysis method for analyzing load data based on the Rainflow method, which is executed by the processing device 3 (see FIG. 1) of the computer device 10 in this embodiment, will be described.

[0041] First, with reference to FIG. 12, a calculation image of the load data analysis method according to this embodiment will be described. FIG. 12 shows an example of a waveform (time-series load waveform) of a load applied to an object over a long period of time. As described above, in the conventional rainflow method, the entire load data over a long period of time (entire time-series load waveform) is collected and temporarily stored, and then the load data is analyzed. In contrast, in this embodiment, as shown in FIG. 12, the processing device 3 of the computer device 10 analyzes the load data by the rainflow method using a predetermined amount of load data, rather than the entire load data, that is, using load data within a partial region of the entire time-series load waveform (hereinafter referred to as the "calculation region" as appropriate).

[0042] To put it simply, the processing device 3 sequentially executes the Rainflow method while gradually shifting the calculation domain, and sequentially collects wave information of the load data. In reality, the processing device 3 does not use a partial calculation domain of a long-term time-series load waveform as shown in Figure 12, but analyzes the load data using a calculation domain obtained in real time, that is, a calculation domain consisting of multiple load data acquired every moment. More specifically, the processing device 3 sequentially analyzes the load data in such a calculation domain using the Rainflow method, thereby updating (adding) the histograms of the load amplitude and load average, and updating (adding) the damage value.

[0043] Next, Fig. 13 shows a specific example of load data in the calculation domain used in this embodiment. Fig. 13 shows load data corresponding to the peak value of the load (i.e., peak data, hereinafter referred to as "load peak data"), and N ALL The horizontal axis of Fig. 13 shows the calculation domain consisting of N (35 in one example) load peak data. ALLIn this embodiment, the processing device 3 uses the load peak data in the calculation domain as shown in FIG. ALL Only the load peak data having peak values ​​is stored in the memory 3b, and this load peak data is analyzed by the Rainflow method. Specifically, the processing device 3 sequentially repeats the storage of the load peak data in the current calculation domain and the analysis of this load peak data by the Rainflow method. ALL This load peak data corresponds to the "predetermined amount of load data" in the present invention.

[0044] Next, Fig. 14 shows a specific example of small loop recovery and stop wave recovery for load data (load peak data) within a calculation domain in this embodiment. The left side of Fig. 14 shows load peak data within the same calculation domain as Fig. 13. When small loops in the load peak data within such a calculation domain are recovered (see Figs. 6 and 9 for small loop recovery), as shown in the upper right of Fig. 14, an amplified wave in which the load amplitude gradually increases and a decaying wave in which the load amplitude gradually decreases appear in the waveform of the load peak data within the calculation domain. Usually, the amplified wave appears before the decaying wave.

[0045] When stop wave recovery is performed on the load peak data after such small loop recovery (see Figures 7 and 10 for stop wave recovery), as shown in the lower right of Figure 14, all of the amplified waves are recovered, while the decaying waves remain unrecovered. In other words, the load peak data constituting the amplified waves satisfy the stop wave condition (corresponding to the "predetermined condition" in this invention) of "|r2| ≧ |r1|" for the two adjacent PV differences r1 and r2 in the load waveform, and therefore become stop waves (determined waves) and are recovered. On the other hand, the load peak data constituting the decaying waves do not satisfy the stop wave condition and therefore remain unrecovered as undetermined waves. In this case, the processing device 3 updates (adds) the histograms of the load amplitude and load average based on the load peak data constituting the amplified waves, and updates (adds) the damage value, and then removes this load peak data without storing it. In contrast, the processing device 3 basically stores the load peak data constituting the decaying waves in memory 3b without performing the calculations performed on such amplified waves.

[0046] Note that an undetermined wave is usually a decaying wave whose load amplitude gradually decreases. Furthermore, if a wave larger than the undetermined wave appears in the future, the undetermined wave will be collected as a stopping wave. Therefore, the undetermined wave is stored in the memory 3b until such a large wave appears.

[0047] Next, FIG. 15 is an explanatory diagram of the processing that is performed after storing the undetermined waves in this embodiment. The left side of FIG. 15 shows the same undetermined waves as in FIG. 14. In this embodiment, first, the processing device 3 pushes the undetermined waves to the beginning of the calculation area, as shown in the upper right of FIG. 15. Specifically, the processing device 3 compresses the undetermined waves and puts them left-justified in the calculation area. In this case, the processing device 3 calculates the number of waves (N ALL15) so that the undetermined wave occupies the number of peak values ​​that the undetermined wave has (six in the example shown in FIG. 15). When the undetermined wave is pushed to the front of the calculation domain in this way, an empty area where no load peak data exists is generated at the rear of the calculation domain. In this embodiment, the processing device 3 inserts newly read load peak data into such an empty area of ​​the calculation domain, as shown in the lower right of FIG. 15. In this way, the processing device 3 ALL A new calculation domain is generated that is filled with load peak data. Then, the processing device 3 performs the analysis by the Rainflow method again on the load peak data in the new calculation domain thus generated, that is, performs the processing shown in Figures 14 and 15 again, particularly, repeats the processing in real time.

[0048] Here, the number of undetermined waves as described above does not continue to increase, but rather increases and decreases. Therefore, storing undetermined waves does not unilaterally burden the memory 3b. FIG. 16 shows a specific example of the number of undetermined waves generated for each step in this embodiment. FIG. 16 shows the number of undetermined waves for each step. From this figure, it can be seen that the number of undetermined waves increases and decreases for each step. Therefore, it can be said that undetermined waves do not burden the memory 3b.

[0049] Next, Fig. 17 is an explanatory diagram of the upper limit of the number of load data (number of load peak data) of an undetermined wave in this embodiment. An example of an undetermined wave in the calculation domain is shown on the left side of Fig. 17. Here, as a result of repeating the processes shown in Figs. 14 and 15, the number of load peak data constituting the undetermined wave reaches N ALL 17 illustrates a case where the number of load peak data constituting the uncertain wave reaches N. In this case, the calculation area is filled with load peak data of the uncertain wave, and new load peak data cannot be read. In other words, new load peak data cannot be inserted into the free space in the calculation area to generate a new calculation area. Therefore, in this embodiment, as shown on the right side of FIG. 17, the processing device 3 performs the calculation when the number of load peak data constituting the uncertain wave reaches N. unknownWhen the number of load peak data constituting the uncertain wave becomes N (corresponding to the "limit amount" in the present invention, 31 in one example), unknown In other words, the processing device 3 deletes some of the load peak data that are acquired later in time from the load peak data that constitute the undetermined wave so that the number of load peak data that constitute the undetermined wave is less than N. unknown In this case, the processor 3 collects the load peak data to be deleted at its current size and then deletes it.

[0050] According to this embodiment, the number of load peak data constituting the uncertain wave is set to N unknown By keeping the number of load peaks to less than 1, new load peak data can be reliably read. Therefore, this new load peak data can be inserted into an empty space in the calculation area to generate a new calculation area. Furthermore, in this embodiment, load peak data acquired later in time among the load peak data constituting the undetermined wave is deleted, so the influence (error) caused by deleting this load peak data can be minimized. This is because the load peak data constituting the undetermined wave are arranged in descending order of load amplitude (because the undetermined wave is a damped wave), and therefore the load peak data at the rear of the undetermined wave have smaller load amplitudes.

[0051] In this embodiment, the processing device 3 calculates the damage value (hereinafter, the damage value calculated from the deterministic wave is referred to as the “damage value D ”) from the load amplitude and the frequency of the load average when collecting the small loop and the stop wave. known In this case, the processing device 3 calculates the calculated damage value D known The damage value so far is D known Add to the damage value D knownOn the other hand, the processing device 3 recovers (temporarily recovers) the load peak data constituting the uncertain wave with the magnitude of the current load peak data, and updates the damage value (hereinafter, the damage value calculated from the uncertain wave) as the "damage value D unknown In this case, the processing device 3 calculates the damage value D unknown The processing device 3 recalculates the damage value D known and damage value D unknown By adding this, the current damage value (hereinafter referred to as "Damage Value D") total ") is calculated.

[0052] Next, a method of linking an undetermined wave with new load peak data in this embodiment will be described with reference to Fig. 18 and Fig. 19. First, problems that may occur when linking an undetermined wave with new load peak data will be described with reference to Fig. 18. On the left side of Fig. 18, when the number of load peak data is N, unknown In this case, as described above, the processing device 3 determines whether the number of load peak data constituting the undetermined wave reaches N. unknown The processing device 3 deletes the load peak data acquired later in time from the load peak data constituting the undetermined wave so that the number of undetermined wave data is less than 1. Then, the processing device 3 reads new load peak data and inserts this new load peak data into the free space in the calculation area. At this time, the processing device 3 connects the latest load peak data D1 (hereinafter referred to as "first load peak data" as appropriate) in time from the undetermined wave after the load peak data deletion with the earliest load peak data D2 (hereinafter referred to as "second load peak data" as appropriate) in time from the newly read load peak data.

[0053] In Case 1 shown in the upper right of Figure 18, the load waveform after connecting the first load peak data D1 and the second load peak data D2 is a waveform in which the load peak data alternately increases and decreases. Specifically, the value decreases (negative slope) from the load peak data D0 before the first load peak data D1 to the first load peak data D1, and the value increases (positive slope) from the first load peak data D1 to the second load peak data D2. This means that the load peak data switches from decreasing to increasing at the first load peak data D1. The rainflow method can be correctly applied to the load waveform in Case 1 in which the load peak data alternates between increasing and decreasing.

[0054] On the other hand, in Case 2 shown in the lower right of Figure 18, the load waveform after connecting the first load peak data D1 and the second load peak data D2 does not show alternating increases and decreases in the load peak data. Specifically, the value decreases (negative slope) from load peak data D0 to first load peak data D1, and also decreases (negative slope) from the first load peak data D1 to the second load peak data D2. Therefore, it can be said that there is a continuous decrease in the load peak data before and after the first load peak data D1. In this way, the load waveform in Case 2, where increases and decreases in the load peak data do not alternate, cannot be used correctly with the Rainflow method.

[0055] In this embodiment, if the load waveform obtained by connecting the first load peak data D1 and the second load peak data D2 does not result in a waveform in which the load peak data alternately increases and decreases, the processing device 3 overwrites the first load peak data D1 with the second load peak data D2, as shown in FIG. 19. In other words, the processing device 3 deletes the first load peak data D1 and connects the load peak data D0 preceding the first load peak data D1 with the second load peak data D2. The processing device 3 then uses the new load peak data thus overwritten as the load peak data for the new calculation domain for the next processing. The load peak data for this new calculation domain results in a load waveform in which the load peak data alternately increases and decreases, allowing the Rainflow method to be correctly applied.

[0056] Next, the overall flow of the load data analysis method according to this embodiment will be described with reference to Fig. 20. Fig. 20 is a flowchart showing the load data analysis method according to this embodiment. This flow is repeatedly executed at a predetermined cycle by the processing device 3 of the computer device 10.

[0057] First, in step S101, the processing device 3 determines whether initialization has been performed. In one example, if the computer device 10 is mounted on a vehicle and analyzes the load applied to a predetermined part of the vehicle, the initialization determined in step S101 corresponds to when the vehicle is shipped from a factory, when data initialization is performed by a dealer, or when the memory 3b that collects data is initialized by turning off the ignition.

[0058] If it is determined that the initialization has been performed (step S101: Yes), the processing device 3 proceeds to step S102. In step S102, the processing device 3 calculates the damage value D known is set to "0", and the variable i used for increment processing in this flow is set to "0", that is, the damage value D knownand the variable i are initialized. Then, the processing device 3 proceeds to step S103. On the other hand, if it is determined that the initialization has not been performed (step S101: No), the processing device 3 proceeds to step S103 without performing step S102.

[0059] In step S103, the processing device 3 acquires load data corresponding to the load (weight, stress, etc.) applied to a predetermined object, detected by the load sensor 7. Then, in step S104, the processing device 3 calculates load peak data by searching for peak values ​​in the continuously acquired load data. Specifically, the processing device 3 calculates the load peak data by searching for points (peak values) where the change in the load waveform switches from an increase to a decrease or from a decrease to an increase.

[0060] Next, in step S105, the processing device 3 determines whether the variable i is not 0 (i≠0). As a result, if it is determined that the variable i is not 0 (step S105: Yes), the processing device 3 proceeds to step S106, where it adds the load peak data calculated in step S104 to the calculation domain. On the other hand, if it is not determined that the variable i is not 0 (step S105: Yes), that is, if the variable i is 0, the processing device 3 proceeds to step S107, where it links the load peak data calculated in step S104 to the undetermined wave already in the calculation domain. Specifically, the processing device 3 links the load peak data calculated in step S104 to the chronologically latest (end) load peak data in the undetermined wave.

[0061] Next, in step S108, the processing device 3 increments the variable i (i=i+1). Then, in step S109, the processing device 3 determines whether the number of load peak data in the calculation domain is N ALL In other words, it is determined whether the calculation area is filled with load peak data. As a result, the processing device 3 determines whether the number of load peak data in the calculation area is N ALLIf it is determined that the number of load peak data in the calculation domain has reached N (step S109: Yes), the process proceeds to step S110. ALL If it is not determined that the number of pieces has been reached (step S109: No), the process returns to step S103, and the processes from step S103 onward are performed again.

[0062] Next, in step S110, the processing device 3 identifies small loops for the load peak data within the calculation domain using the small loop condition "|r1|>|r2|≦|r3|" for three adjacent PV differences r1, r2, and r3, and recovers the identified small loops. Specifically, the processing device 3 updates (adds) the histograms of the load amplitude and the load average based on the load peak data constituting the small loops, and calculates the damage value D known Update (add)

[0063] Next, in step S111, the processing device 3 specifies a stop wave for the load peak data in the calculation domain after the small loop recovery using the stop wave condition "|r2|≧|r1|" for two adjacent PV differences r1 and r2, and recovers the specified stop wave. Specifically, the processing device 3 updates (adds) the histograms of the load amplitude and the load average based on the load peak data constituting the stop wave, and calculates the damage value D known Update (add)

[0064] Next, in step S112, the processing device 3 calculates the damage value D based on the load peak data of the undetermined wave that has not been collected as a stop wave within the calculation domain. unknown Then, in step S113, the processing device 3 calculates the damage value D known and damage value D unknown By adding and, the current damage value D total The processing device 3 calculates the current damage value D total may be displayed on the display device of the output device 5. At this time, the processing device 3 may also display a histogram of the load amplitude and the load average. totalIf the fatigue life is less than a predetermined value (e.g., 0.2 (20%)), a display may be displayed to notify the user that the fatigue life is approaching. In one example, the processing device 3 may display a message that a vehicle part may fail or that the vehicle part needs to be replaced.

[0065] In addition, when all the load peak data in the calculation domain are collected as stop waves, that is, when no undetermined waves occur, the processing device 3 does not need to perform steps S112 and S113. In this case, the damage value D total is the damage value D known This becomes:

[0066] Next, in step S114, the processing device 3 determines whether the number of load peak data of the undetermined wave is N unknown As a result, it is determined whether the number of load peak data of the undetermined wave is N unknown If it is determined that the number of load peak data constituting the undetermined wave is N or more (step S114: Yes), the processing device 3 proceeds to step S115. unknown The processor 3 deletes the load peak data acquired later in time from the load peak data constituting the uncertain wave so that the number of the uncertain wave is less than 1. More specifically, the processor 3 calculates the damage value D based on the load peak data to be deleted, and sets the damage value D as the above-mentioned damage value D. unknown Then, the processing device 3 proceeds to step S116. On the other hand, if the number of load peak data of the undetermined wave is N unknown If it is not determined that the number is equal to or greater than the number (step S114: No), the processing device 3 proceeds to step S116 without performing step S115.

[0067] Next, in step S116, the processing device 3 compresses the load peak data of the undetermined wave (the load peak data of the undetermined wave that was not deleted when step S115 was performed) to the beginning of the calculation area. Specifically, the processing device 3 compresses the undetermined wave and justifies it to the left in the calculation area. Then, the processing device 3 proceeds to step S117, where it sets the variable i to "0", that is, it initializes the variable i. After this, the processing device 3 exits the flow shown in FIG. 20 and performs the processes from step S101 onwards again.

[0068] <Action and effect> Next, the actions and effects of the load data analysis method according to this embodiment will be specifically described with reference to Figures 21 to 23. Figures 21 to 23 show an example of the results when the load data analysis method according to this embodiment is applied to the load waveform shown in Figure 12.

[0069] Specifically, Fig. 21 shows the memory usage (graph G11) according to this embodiment, and the error (graph G12) of the damage value according to this embodiment compared to the conventional rainflow method (a method of analyzing the load data after collecting all of the load data). The horizontal axis of Fig. 21 shows the length of the array corresponding to the calculation domain (i.e., the number N of load peak data constituting the calculation domain). ALL ) Naturally, the longer the length of this array, the more memory is used, but the smaller the error in the damage value. Graph G11 shows that the load data analysis method according to this embodiment uses less than 1 KB of memory, meaning that very little memory is required for calculation. Graph G12 also shows that this embodiment has a much smaller error in the damage value than the conventional rainflow method, ensuring calculation accuracy equivalent to that of the conventional rainflow method.

[0070] FIG. 22 shows the memory usage (graph G11) and the calculation time per step (graph G13) according to this embodiment. As in FIG. 21, the horizontal axis of FIG. 22 shows the length of the array. Graph G13 shows that according to the load data analysis method according to this embodiment, the calculation time per step is less than 10 ms, which is extremely short. Note that the results shown in FIG. 22 were obtained in debug mode (without machine language conversion), which has a relatively slow calculation speed.

[0071] FIG. 23 shows damage values ​​obtained by this embodiment (graph G21) and the conventional rainflow method (graph G22). The horizontal axis of FIG. 23 shows steps corresponding to the time from the start of calculation. In the conventional rainflow method, the load data is analyzed after all the load data is collected, so only the final state of the damage value can be obtained (graph G22). In contrast, the load data analysis method according to this embodiment (N ALL In this method (assuming that the value of the damage coefficient is set to 35), as shown in graph G21, analysis is performed in real time on load data acquired every moment, so that the damage value history can be obtained. This makes it possible to grasp the timing at which damage will increase. Moreover, according to this embodiment, the damage value at the later step (approximately 800 steps) matches the damage value at the final state in the conventional rainflow method. In other words, according to this embodiment, it is possible to ensure calculation accuracy equivalent to that of the conventional rainflow method.

[0072] In summary, the load data analysis method according to this embodiment is as follows: (1) A first step of acquiring load data in a calculation domain and calculating frequencies related to load amplitude and load average using load data that satisfy the stopping wave condition among the load data in the calculation domain based on the rainflow method; (2) a second step of storing the load data that does not satisfy the stopping wave condition among the load data in the calculation domain; (3) a third step of generating load data of the calculation domain for executing the first step by combining the load data stored in the second step with newly acquired load data; The first to third steps are repeatedly executed.

[0073] According to this embodiment, instead of collecting and processing load data over a long period of time, only load data in a specified calculation domain is processed. This allows for reduced memory capacity and shorter calculation times when analyzing load data based on the Rainflow method (Figures 21 and 22). Furthermore, according to this embodiment, load data that satisfies the stop wave condition is used to calculate frequency, while load data that does not satisfy the stop wave condition is combined with new load data to create load data in a new calculation domain, and this load data in the new calculation domain is then processed. This allows for real-time analysis of load data acquired from moment to moment while maintaining the same calculation accuracy as the conventional Rainflow method (Figure 23). As described above, according to this embodiment, when analyzing load data based on the Rainflow method, reduced memory capacity (memory saving), shorter calculation times (high-speed calculation), and real-time analysis can be appropriately achieved while maintaining the same calculation accuracy as the conventional Rainflow method.

[0074] Furthermore, according to this embodiment, in the second step, N load data that does not satisfy the stop wave condition among the load data in the calculation domain is selected. unknown If the number of load data is more than the limit, the number of load data to be stored is N. unknown The load data acquired later in time among the load data in the calculation domain is deleted so that the number of load data is less than N, and the load data remaining after this deletion is stored. unknown By keeping the number of load data to less than 1, new load data can be reliably read. Therefore, this new load data can be combined to reliably generate a new calculation domain. Furthermore, in this embodiment, since load data acquired later in time among the load data in the calculation domain is deleted, that is, load data with a relatively small load amplitude is deleted, the effect (error) caused by the deletion of load data can be minimized.

[0075] Furthermore, according to this embodiment, if the relationship between the first load peak data that is the latest in time among the load peak data corresponding to the load data stored in the second step and the second load peak data that is the earliest in time among the newly acquired load peak data does not satisfy the relationship that the load peak data alternately increases and decreases, the third step overwrites the first load peak data with the second load peak data to generate load data for the calculation domain for executing the first step. This makes it possible to appropriately generate new load data for the calculation domain in which the load data alternates in increase and decrease when the load data stored in the second step is combined with the newly acquired load data, thereby enabling the Rainflow method to be applied appropriately.

[0076] Furthermore, in this embodiment, the first and second steps use a stop wave condition that, for two adjacent PV differences r1 and r2 in the load waveform, the latter PV difference r2 is equal to or greater than the former PV difference r1. This ensures that load peak data with a relatively large load amplitude can be reliably used to calculate the frequencies related to the load amplitude and load average.

[0077] Furthermore, this embodiment further includes a fifth step of calculating a damage value of the object due to the load applied to the object based on at least the frequency calculated in the first step, thereby making it possible to calculate a highly accurate damage value in real time while ensuring reductions in memory capacity and calculation time.

[0078] Furthermore, this embodiment further includes a sixth step of displaying the frequency calculated in the first step or a value related to the frequency, thereby making it possible to appropriately notify the user of the frequency, damage value, etc. related to the load amplitude and load average.

[0079] <Modification> In the above-described embodiment, the present invention is applied to a vehicle as an example, but the present invention can also be applied to various objects to which a load is applied, such as electrical appliances, ships, aircraft, and large buildings, in addition to vehicles.

[0080] In the above embodiment, an example is shown in which the frequencies (histograms) of both the load amplitude and the load average are calculated, but the frequency of only one of the load amplitude and the load average (for example, only the load amplitude) may be calculated. In the above embodiment, an example is shown in which a damage value is calculated from such frequencies, but the present invention is not limited to calculating a damage value. [Explanation of symbols]

[0081] 1. Input Device 3 Processing equipment 3a microprocessor 3b Memory 5 Output Devices 7 Load Sensor 10 Computer equipment

Claims

1. A load data analysis method for analyzing load data indicating loads applied irregularly and repeatedly to an object based on a rainflow method, comprising: a first step of acquiring a predetermined amount of load data, and calculating a frequency related to an amplitude of a waveform of a load applied to the object and / or an average value of the load based on the Rainflow method, using load data that satisfies a predetermined condition among the predetermined amount of load data; a second step of storing load data that does not satisfy the predetermined condition among the predetermined amount of load data; a third step of generating the predetermined amount of load data for executing the first step by combining the load data stored in the second step with newly acquired load data; and The first to third steps are repeatedly executed, the second step, when the load data that does not satisfy the predetermined condition among the predetermined amount of load data is equal to or greater than a limit amount, deletes a part of the load data from the predetermined amount of load data so that the load data to be stored is less than the limit amount, and stores the load data remaining after the deletion; the second step, when the load data that does not satisfy the predetermined condition among the predetermined amount of load data is equal to or greater than the limit amount, deletes the load data that is acquired later in time among the predetermined amount of load data so that the load data to be stored is less than the limit amount; A load data analysis method comprising:

2. A load data analysis method for analyzing load data indicating loads applied irregularly and repeatedly to an object based on a rainflow method, comprising: a first step of acquiring a predetermined amount of load data, and calculating a frequency related to an amplitude of a waveform of a load applied to the object and / or an average value of the load based on the Rainflow method, using load data that satisfies a predetermined condition among the predetermined amount of load data; a second step of storing load data that does not satisfy the predetermined condition among the predetermined amount of load data; a third step of generating the predetermined amount of load data for executing the first step by combining the load data stored in the second step with newly acquired load data; and The first to third steps are repeatedly executed, a fourth step of calculating peak data corresponding to a point where the change in the load data switches from an increase to a decrease or from a decrease to an increase; the first to third steps use the peak data calculated in the fourth step as load data to be processed in each step; In the third step, if a relationship between first peak data that is the latest in time among the peak data corresponding to the load data stored in the second step and second peak data that is the earliest in time among the peak data corresponding to the newly acquired load data does not satisfy a relationship in which peak data increases and decreases alternately, the first peak data is overwritten with the second peak data to generate the predetermined amount of load data for executing the first step. A load data analysis method comprising:

3. 3. The load data analysis method according to claim 2, wherein the second step, when the load data that does not satisfy the specified condition among the specified amount of load data is equal to or greater than a limit amount, deletes some of the load data from the specified amount of load data so that the load data to be stored is less than the limit amount, and stores the load data remaining after this deletion.

4. 4. The load data analysis method according to claim 3, wherein the second step deletes load data acquired later in time from the specified amount of load data when the load data that does not satisfy the specified condition among the specified amount of load data is equal to or greater than the limit amount so that the load data to be stored is less than the limit amount.

5. a fourth step of calculating peak data corresponding to a point where the change in the load data switches from an increase to a decrease or from a decrease to an increase; the first to third steps use the peak data calculated in the fourth step as load data to be processed in each step; In the third step, if a relationship between first peak data that is the latest in time among the peak data corresponding to the load data stored in the second step and second peak data that is the earliest in time among the peak data corresponding to the newly acquired load data does not satisfy a relationship in which peak data increases and decreases alternately, the first peak data is overwritten with the second peak data to generate the predetermined amount of load data for executing the first step. The load data analysis method according to claim 1 .

6. a fourth step of calculating peak data corresponding to a point where the change in the load data switches from an increase to a decrease or from a decrease to an increase; the first to third steps use the peak data calculated in the fourth step as load data to be processed in each step; the first and second steps use, as the predetermined condition, a condition that, for adjacent first peak data, subsequent second peak data, and subsequent third peak data calculated in the fourth step, a difference between the second peak data and the third peak data is equal to or greater than a difference between the first peak data and the second peak data; The load data analysis method according to any one of claims 1 to 5.

7. 7. The load data analysis method according to claim 1, further comprising a fifth step of calculating a damage value of the object due to the load applied to the object based on at least the frequency calculated in the first step.

8. 8. The load data analysis method according to claim 1, which is executed by a processing device mounted on a vehicle, and analyzes load data corresponding to loads applied to components of the vehicle.

9. The load data analysis method according to claim 1 , further comprising a sixth step of displaying the frequency calculated in the first step or information related to the frequency.

10. A load data analysis device that analyzes load data corresponding to loads that are repeatedly and irregularly applied to an object based on a rainflow method, a first means for acquiring a predetermined amount of load data, and calculating a frequency related to the amplitude of a waveform of a load applied to the object and / or an average value of the load based on the Rainflow method, using load data that satisfies a predetermined condition among the predetermined amount of load data; a second means for storing load data that does not satisfy the predetermined condition among the predetermined amount of load data; a third means for generating the predetermined amount of load data for executing the first means by combining the load data stored by the second means with newly acquired load data; and each of the first to third means repeatedly executes a process; the second means, when the load data that does not satisfy the predetermined condition among the predetermined amount of load data is equal to or greater than a limit amount, deletes a portion of the load data from the predetermined amount of load data so that the load data to be stored is less than the limit amount, and stores the load data remaining after the deletion; When the load data that does not satisfy the predetermined condition among the predetermined amount of load data is equal to or greater than the limit amount, the second means deletes the load data that is acquired later in time among the predetermined amount of load data so that the load data to be stored is less than the limit amount. A load data analysis device characterized by:

11. A load data analysis program executed by a computer device for analyzing load data corresponding to loads applied irregularly and repeatedly to an object based on a rainflow method, comprising: The computer device a first means for acquiring a predetermined amount of load data, and calculating a frequency related to the amplitude of a waveform of a load applied to the object and / or an average value of the load based on the Rainflow method, using load data that satisfies a predetermined condition among the predetermined amount of load data; a second means for storing load data that does not satisfy the predetermined condition among the predetermined amount of load data; a third means for generating the predetermined amount of load data for executing the first means by combining the load data stored by the second means with newly acquired load data; It functions as causing the computer device to function so that each of the first to third means repeatedly executes a process; the second means, when the load data that does not satisfy the predetermined condition among the predetermined amount of load data is equal to or greater than a limit amount, deletes a portion of the load data from the predetermined amount of load data so that the load data to be stored is less than the limit amount, and stores the load data remaining after the deletion; When the load data that does not satisfy the predetermined condition among the predetermined amount of load data is equal to or greater than the limit amount, the second means deletes the load data that is acquired later in time among the predetermined amount of load data so that the load data to be stored is less than the limit amount. and causing the computer device to function so as to:

Images