Analyzer, analysis method, program, trained machine learning model, and machine learning method

Abstract

To efficiently classify work on the basis of time-sequential data indicating the operation state of a work machine.SOLUTION: This analyzer comprises: an acquisition unit that acquires time-sequential data indicating at least the operation states of first and second movable parts which a work machine includes; a first operation state classification unit that classifies the operation state of the first movable part on the basis of the time-sequential data; a division unit that divides the time base of the time-sequential data into a plurality of sections on the basis of the classification result of the operation state of the first movable part; a second operation state classification unit that classifies the operation state of the second movable part for each section on the basis of the time-sequential data; and a work classification unit that classifies work performed by the work machine on the basis of the classification results of the operation states of the first and second movable parts.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present disclosure relates to an analysis device, an analysis method, a program, a trained machine learning model, and a machine learning method.

Background Art

[0002] Patent Document 1 discloses a technique for discriminating each operation mode included in a cycle operation (in this example, a loading operation) in which a wheel loader is operated by an operator to repeat each operation mode of forward travel with no load, excavation, backward travel with load, forward travel with load, loading, and backward travel with no load, based on numerical values representing the state of the wheel loader.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] By the way, the operations performed by a wheel loader include multiple types of operations such as a loading operation, a sorting operation of excavated materials, and a scraping operation, in addition to the loading operation. In addition, there is a need to efficiently analyze what operation a work machine is performing based on time-series data representing the operating state of the work machine.

[0005] The present disclosure has been made in view of the above circumstances, and an object thereof is to provide an analysis device, an analysis method, a program, a trained machine learning model, and a machine learning method capable of efficiently classifying operations based on time-series data representing the operating state of a work machine.

Means for Solving the Problems

[0006] The analysis device of the present disclosure includes an acquisition unit that acquires time-series data representing at least the operating state of the first movable part and the operating state of the second movable part of a working machine having the first movable part and the second movable part, a first operating state classification unit that classifies the operating state of the first movable part based on the time-series data, a division unit that divides the time axis of the time-series data into a plurality of sections based on the classification result of the operating state of the first movable part, a second operating state classification unit that classifies the operating state of the second movable part for each section based on the time-series data, and a work classification unit that classifies the work performed by the working machine based on the classification result of the operating state of the first movable part and the classification result of the operating state of the second movable part.

[0007] Further, the analysis method of the present disclosure includes a step of acquiring time-series data representing at least the operating state of the first movable part and the operating state of the second movable part of a working machine having the first movable part and the second movable part, a step of classifying the operating state of the first movable part based on the time-series data, a step of dividing the time axis of the time-series data into a plurality of sections based on the classification result of the operating state of the first movable part, a step of classifying the operating state of the second movable part for each section based on the time-series data, and a step of classifying the work performed by the working machine based on the classification result of the operating state of the first movable part and the classification result of the operating state of the second movable part.

[0008] Further, the program of the present disclosure causes a computer to execute a step of acquiring time-series data representing at least the operating state of the first movable part and the operating state of the second movable part of a working machine having the first movable part and the second movable part, a step of classifying the operating state of the first movable part based on the time-series data, a step of dividing the time axis of the time-series data into a plurality of sections based on the classification result of the operating state of the first movable part, a step of classifying the operating state of the second movable part for each section based on the time-series data, and a step of classifying the work performed by the working machine based on the classification result of the operating state of the first movable part and the classification result of the operating state of the second movable part.

[0009] In addition, the learned machine learning model of the present disclosure inputs predetermined input information based on time series data that at least represents the operating state of the first movable part and the operating state of the second movable part of a working machine having the first movable part and the second movable part, and outputs output information representing the result of classifying the operating state of the second movable part. The input information includes a plurality of predetermined feature amounts related to the operating state of the second movable part for each of the intervals obtained by dividing the time axis of the time series data into a plurality of intervals based on the classification result of the operating state of the first movable part, and the machine learning is performed using the input information in which the types of the operating states of the second movable part are labeled.

[0010] In addition, the machine learning method of the present disclosure inputs predetermined input information based on time series data that at least represents the operating state of the first movable part and the operating state of the second movable part of a working machine having the first movable part and the second movable part, and outputs output information representing the result of classifying the operating state of the second movable part. The input information includes a plurality of predetermined feature amounts related to the operating state of the second movable part for each of the intervals obtained by dividing the time axis of the time series data into a plurality of intervals based on the classification result of the operating state of the first movable part, and the machine learning model is machine-learned using the input information in which the types of the operating states of the second movable part are labeled.

Advantages of the Invention

[0011] The analysis device, analysis method, program, learned machine learning model, and machine learning method of the present disclosure classify the operating state of the second movable part for each section obtained by dividing the time series data into a plurality of sections based on the classification result of the first movable part. Therefore, the operating state of the second movable part can be efficiently classified. Accordingly, based on the time series data, the operations of the working machine having the first movable part and the second movable part can be efficiently classified.

Brief Description of the Drawings

[0012]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Figure 13

Figure 14

Figure 15

Figure 16

Figure 17

Figure 18

Figure 19

Figure 20

Figure 21

Figure 22

Figure 23

Figure 24

Figure 25

Figure 26

Figure 27

Figure 28

Figure 29

Figure 30

Figure 31

Figure 32

[0013] Hereinafter, embodiments of the present disclosure will be described with reference to the drawings. In each figure, the same or corresponding components are denoted by the same reference numerals, and the description thereof will be omitted as appropriate.

[0014] (Analysis System) FIG. 1 is a schematic diagram showing a configuration example of an analysis system according to an embodiment of the present disclosure. The analysis system SY1 shown in FIG. 1 includes a wheel loader 1 as an example of a work machine to be analyzed, a data management server 20 communicatively connected to the wheel loader 1 via a communication network NW, and an analysis device 30. The wheel loader 1 includes a vehicle body control device 100 and an operation information recording device 101. The vehicle body control device 100 inputs sensor information measured by a plurality of sensors included in the wheel loader 1, and controls each part of the wheel loader 1 according to the operation of an operator. The operation information recording device 101 records time-series data representing the operation state of the wheel loader 1, such as a predetermined control signal and sensor information generated by the vehicle body control device 100, for a certain period (or for a certain amount of data), and transmits it to the data management server 20 via the communication network NW. The data management server 20 accumulates the time-series data received from the operation information recording device 101 as a time-series data file 21. In the present embodiment, the analysis device 30 is provided separately from the data management server 20, but the present invention is not limited to this. The analysis device 30 may be provided in the data management server 20.

[0015] (Work machine) Here, with reference to FIG. 2, the wheel loader 1 will be described. FIG. 2 is a side view showing a configuration example of a work machine (wheel loader 1) according to an embodiment of the present disclosure. As shown in FIG. 2, the wheel loader 1 has a vehicle body 2, a cab 3, a traveling device 4, and a work implement 10. The wheel loader 1 travels on a work site by the traveling device 4. The wheel loader 1 performs work using the work implement 10 at the work site. The wheel loader 1 can perform work such as excavation work, loading work, transportation work, and snow removal work using the work implement 10.

[0016] The cab 3 is supported by the vehicle body 2. Inside the cab 3, a driver's seat on which an operator sits, an operation device, a display input unit, etc. are arranged.

[0017] The traveling device 4 includes a power transmission device (not shown), rotatable wheels 5, etc. The wheels 5 support the vehicle body 2. The wheel loader 1 can travel on a road surface (or ground) RS by the traveling device 4. In FIG. 2, only the left front wheel 5F and rear wheel 5R are shown. The power transmission device transmits the driving force of a power source, which will be described later, to the wheels 5. In this embodiment, it is assumed that the power transmission device has a transmission with forward four speeds, neutral, and reverse four speeds (speed stages). Note that the traveling device 4 is an example of a configuration of the "traveling section" of the present disclosure.

[0018] The work implement 10 is supported by the vehicle body 2. The work implement 10 includes a bucket 12 as an example of a work tool, and a movable support portion 17 that changes the position and posture of the bucket 12. In the example shown in FIG. 2, the movable support portion 17 includes a boom 11, a boom cylinder 13, a bucket cylinder 14, a bell crank 15, and a link 16.

[0019] The boom 11 is rotatably supported with respect to the vehicle body 2 and moves in the vertical direction according to the extension and contraction of the boom cylinder 13. The boom cylinder 13 is an actuator that generates power for moving the boom 11. One end is connected to the vehicle body 2, and the other end is connected to the boom 11. When the operator operates a boom operating device (not shown), the boom cylinder 13 extends and contracts. Thereby, the boom 11 moves in the vertical direction. The boom cylinder 13 is, for example, a hydraulic cylinder. In this embodiment, as shown by an arrow A1 for the angle of the boom 11, with the horizontal direction being 0 [deg], when it faces upward, it is represented by a positive value, and when it faces downward, it is represented by a negative value.

[0020] The bucket 12 has a cutting edge 12T and is a working tool for excavating objects to be excavated such as earth and sand and for loading. The bucket 12 is pivotally connected to the boom 11 and is also pivotally connected to one end of the link 16. The other end of the link 16 is pivotally connected to one end of the bell crank 15. The central part of the bell crank 15 is pivotally connected to the boom 11, and the other end is pivotally connected to one end of the bucket cylinder 14. The other end of the bucket cylinder 14 is pivotally connected to the vehicle body 2. The bucket 12 is actuated by the power generated by the bucket cylinder 14. The bucket cylinder 14 is an actuator that generates power for moving the bucket 12. When the operator operates a bucket operating device (not shown), the bucket cylinder 14 expands and contracts. As a result, the bucket 12 swings. The bucket cylinder 14 is, for example, a hydraulic cylinder. The cutting edge 12T has a shape such as a mountain edge or a flat edge and is detachably attached to the end of the bucket 12. In this embodiment, the angle of the bucket 12 (the angle in the direction in which the cutting edge 12T faces (the cutting edge direction)) is represented by a value that is positive when facing upward and negative when facing downward, with the horizontal direction being 0 [deg] as shown by the arrow A2.

[0021] In this embodiment, the posture of the bucket 12c with the cutting edge 12T facing downward is referred to as the dump posture. The dump posture is, for example, a posture (dumping posture) in which the excavated material in the bucket 12 can be loaded into a transport vehicle, a hopper, or the like. Further, the posture of the bucket 12a with the cutting edge 12T facing upward is referred to as the tilt posture (holding posture). The tilt posture is, for example, a posture (transport posture) in which the excavated material can be held in the bucket 12. Also, the posture of the bucket 12b with the cutting edge 12T facing in the horizontal direction (including substantially the horizontal direction) with respect to the road surface RS is referred to as the excavation posture (or the traveling posture during excavation). The excavation posture is, for example, a posture when starting to excavate an excavation target such as earth and sand or when traveling toward the excavation target (or a posture suitable for starting excavation or traveling). Further, the posture of the bucket 12 with the cutting edge 12T in contact with the road surface RS with the boom 11 lowered is referred to as the grounding posture. The wheel loader 1 starts excavating an excavation target located in front by, for example, setting the bucket 12 in the excavation posture (or a posture in which the cutting edge 12T is lower than the road surface RS from the excavation posture) and traveling in the forward direction. In the wheel loader 1, the excavation posture can also be called the horizontal posture because the cutting edge direction is substantially horizontal with respect to the road surface RS.

[0022] Note that the wheel loader 1 includes components such as a power source (not shown), a PTO (Power Take Off), a hydraulic pump, a control valve, an operating device, and a display input unit. The power source generates a driving force for operating the wheel loader 1. Examples of the power source include an internal combustion engine and an electric motor. The PTO transmits at least a part of the driving force of the power source to the hydraulic pump. The PTO distributes the driving force of the power source to the traveling device 4 and the hydraulic pump. The hydraulic pump is driven by the power source and discharges hydraulic oil. At least a part of the hydraulic oil discharged from the hydraulic pump is supplied to the boom cylinder 13 and the bucket cylinder 14 via the control valve. The control valve controls the flow rate and direction of the hydraulic oil supplied from the hydraulic pump to the boom cylinder 13 and the bucket cylinder 14 respectively. The working machine 10 operates by the hydraulic oil from the hydraulic pump.

[0023] The operating device is arranged inside the cab 3. The operating device is operated by the operator. The operator operates the operating device to adjust the traveling direction and traveling speed of the wheel loader 1, switch between forward and reverse, and operate the work implement 10. The operating device includes, for example, a steering, a shift lever, an accelerator pedal, a brake pedal, and an operating device for operating the boom 11 and bucket 12 of the work implement 10.

[0024] Also, the wheel loader 1 is provided with sensors such as a GNSS (Global Navigation Satellite System) receiver, an engine speed sensor, a vehicle speed sensor, a fuel consumption sensor, a work implement load sensor, a boom angle sensor, and a bucket angle sensor. The GNSS receiver has two receivers (antennas) and acquires information on position information (latitude, longitude, and altitude) respectively. Based on the two pieces of position information, the orientation of the wheel loader 1 can be calculated. The engine speed sensor detects the rotational speed of the engine (power source). The vehicle speed sensor detects the traveling speed of the wheel loader 1. The fuel consumption sensor detects the fuel consumption of the wheel loader 1. The work implement load sensor detects, for example, the weight (load) of the excavated material held by the bucket 12. The boom angle sensor detects the angle of the boom 11. The bucket angle sensor detects the angle of the bucket 12.

[0025] (Relationship with the first movable part and the second movable part of the present disclosure) As described in the section of means for solving the problems, in the present disclosure, the working machine to be analyzed by the analysis device has a first movable part and a second movable part. The first movable part is, for example, a traveling part (traveling body), which corresponds to the traveling device 4 in the present embodiment. The second movable part is, for example, a working machine, which corresponds to the working machine 10, the boom 11, and the bucket 12 in the present embodiment. In addition, when the working machine to be analyzed is, for example, a hydraulic excavator, for example, the first movable part corresponds to a slewing part (slewing body), and the second movable part corresponds to a working machine (bucket, arm, and boom). Alternatively, the first movable part corresponds to the slewing part and the traveling part, and the second movable part corresponds to the working machine. Alternatively, the first movable part corresponds to the traveling part, and the second movable part corresponds to the slewing part and the working machine. In addition, examples of the working machine of the present disclosure include dump trucks, bulldozers, and the like.

[0026] (Time-series data) FIG. 3 is a schematic diagram showing a configuration example of a time-series data file according to an embodiment of the present disclosure. The time-series data file 21 shown in FIG. 3 includes working machine information 22, operator information 23, and time-series data 24. The working machine information 22 includes, for example, information such as the individual identification information, model, and manufacturing date of the wheel loader 1. The operator information 23 includes identification information of the operator of the wheel loader 1 and the like. The time-series data 24 is time-series data representing the operating states of each part of the wheel loader 1. In the example shown in FIG. 3, the time-series data 24 includes data representing date and time, position information, engine speed, gear position, vehicle speed, fuel consumption, load, boom angle, bucket angle, etc. at a predetermined sampling period (for example, every 1 second). Each time-series data file 21 can be created, for example, for each time from when the engine key is turned on to when it is turned off, or in a form of being divided into a plurality of files at predetermined time intervals. In addition, for example, the gear position and the vehicle speed are examples of data representing the operating state of the first movable part, and the boom angle and the bucket angle are examples of data representing the operating state of the second movable part. In this case, the time-series data 24 is data representing at least the operating state of the first movable part and the operating state of the second movable part of the working machine (wheel loader 1) having the first movable part (traveling device 4) and the second movable part (working machine 10).

[0027] (Configuration of the analysis device) The analysis device 30 can be configured using a computer such as a personal computer or a tablet terminal. The analysis device 30 includes the following components as a functional configuration composed of a combination of hardware such as a computer and its peripheral devices, and software executed by the computer. That is, the analysis device 30 includes, as a functional configuration, an acquisition unit 31, a first operation state classification unit 32, a classification unit 33, a second operation state classification unit 34, a learned machine learning model 35, a work classification unit 36, a detailed analysis unit 37, a storage unit 38, and an output unit 39.

[0028] The acquisition unit 31 acquires a time-series data file 21 including the time-series data 24 to be analyzed and stores it in the storage unit 38. The time-series data file 21 may be automatically acquired from the data management server 20 and stored in the storage unit 38, or may be manually acquired and stored in the storage unit 38.

[0029] The first operation state classification unit 32 classifies the operation state of the traveling device 4 (the first movable part) based on the time-series data 24. FIG. 4 is a schematic diagram showing an example of a behavior discrimination flag of the traveling device 4 (the undercarriage of the working machine, and in this embodiment, the traveling device 4 is also simply referred to as the traveling device) according to the embodiment of the present disclosure. As shown in FIG. 4, the first operation state classification unit 32 classifies the behavior of the traveling device into four types (stop, forward, reverse, or neutral coasting), and sets a behavior discrimination flag (1, 2, 3, or 4) of the traveling device for each type. Here, neutral coasting is a state in which the gear position is set to neutral, and for example, the vehicle speed is limited while lightly stepping on the brake and the vehicle is traveling. In the example shown in FIG. 4, when the value of the vehicle speed is "0" [km / h], the operation state of the traveling device 4 is classified as "stop" regardless of the value of the gear position. When the value of the vehicle speed is not "0", it is "forward" when the gear position is forward 1 to 4 speeds (F1 to F2), "reverse" when the gear position is reverse 1 to 4 speeds (R1 to R2), and "neutral coasting" when the gear position is neutral (N).

[0030] FIG. 5 and FIG. 6 are diagrams showing examples of time-series data related to the behavior determination of a traveling device according to an embodiment of the present disclosure. FIGS. 5 and 6 take time on the horizontal axis, vehicle speed and gear position on the vertical axis (left side), and a behavior determination flag of the traveling device on the vertical axis (right side), and show changes in the values of the vehicle speed, gear position, and behavior determination flag of the traveling device. Note that the vehicle speed is shown as an absolute value regardless of forward or backward movement, and the gear position is shown as +1 to +4 for forward first to fourth gears, 0 for neutral, and -1 to -4 for reverse first to fourth gears. In the example shown in FIG. 5, the vehicle speed becomes 0 at 10 seconds and 15 seconds, and the value of the behavior determination flag of the traveling device is 1 ("stopped"). As it is, the behavior will be frequently determined as "stopped" when forward and backward are switched. Therefore, in the example shown in FIG. 6, correction processing is performed so as not to stop when the stops are not continuous. Although the vehicle speed becomes 0 at 10 seconds and 15 seconds, when the time of becoming 0 does not continue (for example, when it becomes 0 only for 1 sampling), the value of the behavior determination flag of the traveling device is not set to 1 ("stopped"), but is changed to the value of the behavior before the vehicle speed becomes 0.

[0031] The dividing unit 33 divides the time axis of the time-series data 24 into a plurality of sections based on the classification result of the operating state of the traveling device 4 (first movable part) by the first operating state classification unit 32. FIG. 7 is a diagram showing an example in which the time-series data according to the embodiment of the present disclosure is divided into a plurality of sections. In the example shown in FIG. 7, the dividing unit 33 divides the time axis of the time-series data so that a period classified as "forward", "backward", or "stop" by the first operating state classification unit 32 becomes one section. For the period classified as "neutral coasting", it is combined with the previous classification period and included in the previous type of period. For the classification result of the period (0 to 42 seconds) shown in FIG. 6, the dividing unit 33 divides the period (0 to 42 seconds) into five sections, sections (k) to (k + 4) as shown in FIG. 7 (k is an arbitrary natural number). In this embodiment, each section is divided every 1 second, and the last time of each section is the first time of the next section. For this reason, for example, at time 43 seconds in the period 0 to 43 seconds, it is divided into the next section (k + 5) not shown in the figure. Also, as will be described later, the period (0 to 42 seconds) will be classified as a loading operation. In the example shown in FIG. 7, sections (k) to (k + 4) are the periods for one loading operation, section (k) corresponds to forward (1), section (k + 1) corresponds to backward (1), section (k + 2) corresponds to forward (2), section (k + 3) corresponds to stop, and section (k + 4) corresponds to backward (2).

[0032] The second operating state classification unit 34 classifies the operating state of the working machine 10 (second movable part) for each section based on the time-series data 24. The second operating state classification unit 34 is, for example, a machine learning model that inputs input information including a plurality of predetermined feature amounts related to the operating state of the working machine 10 (second movable part) for each section and outputs output information representing the result of classifying the operating state of the working machine 10 (second movable part). Using the learned machine learning model 35 that has been machine-learned with input information in which the types of the operating states of the working machine 10 (second movable part) are labeled as teacher data, the operating state of the working machine 10 (second movable part) is classified. FIG. 8 is a diagram showing an example of time-series data related to the behavior determination of the working machine 10 (in this embodiment, the working machine 10 is also simply referred to as the working machine) according to the embodiment of the present disclosure.

[0033] The learned machine learning model 35 is composed of, for example, a combination of a program that performs operations from input to output and weighting coefficients (parameters) used in the operations. Also, the coefficients used in the calculation by machine learning are optimized so that the solutions obtained for a large number of input data are output. In this embodiment, it has been confirmed that good accuracy can be obtained by using the k-nearest neighbor method as the algorithm of the machine learning model. However, there is no limitation specific to this embodiment regarding the algorithm of the machine learning model.

[0034] FIG. 8 shows the time on the horizontal axis, the vehicle speed, boom angle, and bucket angle on the vertical axis (left side), and the behavior discrimination flag of the traveling device on the vertical axis (right side), and shows the changes in the values of the vehicle speed, boom angle, bucket angle, and the behavior discrimination flag of the traveling device. The values on the time axis, the vehicle speed, and the value of the behavior discrimination flag of the traveling device are the same as those in FIG. 7. FIGS. 9 to 13 are diagrams showing examples of the operating states of the working machine 10 in the sections (k) to (k + 4) shown in FIGS. 7 and 8. The horizontal axis represents the boom angle, the vertical axis represents the bucket angle, and each point based on the boom angle and bucket angle at the same time within the target section is plotted.

[0035] FIG. 14 is a diagram showing an example of a feature amount according to an embodiment of the present disclosure. Note that the feature amount is a numerical value that quantitatively represents the feature of the analysis target. In the present embodiment, four types of feature amounts are defined. For example, for each plot for each interval shown in FIGS. 9 to 13, the first one is "area" (FV1), and for the purpose of grasping the position of the work machine 10, it can be represented by the maximum and minimum values of the boom angle and the maximum and minimum values of the bucket angle. The second one is "inlet / outlet coordinates" (FV2), and for the purpose of grasping the movement, it can be represented by the initial value and the final value of the boom angle in the interval and the initial value and the final value of the bucket angle in the interval. The third one is "center of gravity coordinates" (FV3), and for the purpose of grasping the movement inside the area, it can be represented by the average value of the boom angle and the average value of the bucket angle. The fourth one is "number of peak mountains" (FV4), and for the purpose of grasping the repetition inside the area, it can be represented by the number of changes when there is a peak-like change (local maximum value, change in the value of the slope of the curve exceeding a predetermined threshold value, etc.) in the curve connecting the plots of the boom angle and the bucket angle in time series. FIG. 15 is a diagram showing an example of the feature amount of the interval (k) according to an embodiment of the present disclosure. FIG. 16 is a diagram showing an example of the feature amount according to an embodiment of the present disclosure. FIG. 15 is a diagram in which figures for explaining the feature amounts FV1 to FV3 are added to FIG. 9. Further, FIG. 16 shows an example including a peak (an example during warm-up operation).

[0036] FIG. 17 is a schematic diagram showing an example of a behavior determination flag of a working machine according to an embodiment of the present disclosure. As a result of classifying the operating state of the working machine 10 (second movable part), the second operating state classification unit 34 sets the value of the behavior determination flag of the working machine shown in FIG. 17 for the corresponding section. Further, in the present embodiment, the learned machine learning model 35 inputs the four types of feature amounts shown in FIG. 14 and outputs the value of the behavior determination flag of the working machine shown in FIG. 17. In the example shown in FIG. 17, the behaviors of the working machine are classified into 13 types. The association between the behavior determination flag of the working machine shown in FIG. 17 and the respective contents of "classification", "bucket" "posture" and "movement", and "boom" "posture" and "movement" is, for example, a rough standard when labeling input information in supervised learning. Note that "leveler" is a state in which a function for automatically adjusting to a horizontal state is used. FIGS. 18 to 21 are diagrams showing examples of feature amounts in sections (k + 1) to (k + 4) according to an embodiment of the present disclosure. FIGS. 18 to 21 are diagrams in which figures for explaining the feature amounts FV1 to FV3 are added to FIGS. 10 to 13. When the feature amount shown in FIG. 15 is input, the learned machine learning model 35 outputs, for example, the behavior determination flag "20" of the working machine. The behavior determination flag "20" of the working machine corresponds to the behavior of the working machine in the "scooping" operation. The "scooping" operation is, for example, an operation in which the boom 11 is lowered while moving forward to scrape the excavated material into the bucket 12 and then the bucket 12 is held. Further, when the feature amounts shown in FIGS. 18, 19, 20, and 21 are input, the learned machine learning model 35 outputs, for example, the behavior determination flags "31", "51", "12", and "41" of the working machine.

[0037] FIG. 22 is a diagram showing an example of teacher data according to an embodiment of the present disclosure. The teacher data (data set) 351 shown in FIG. 22 includes a plurality of sets of data in which the initial value, maximum value, minimum value, average value, number of peak mountains, and final value of the boom angle in the target section, the initial value, maximum value, minimum value, average value, number of peak mountains, and final value of the bucket angle, and the value of the behavior determination flag of the working machine (this is the value of the label) are associated with each other.

[0038] The learned machine learning model 35 of this embodiment can be understood as follows. That is, the learned machine learning model 35 inputs predetermined input information based on time-series data that at least represents the operating state of the first movable part and the operating state of the second movable part of a machine tool having a first movable part and a second movable part, and outputs output information representing the result of classifying the operating state of the second movable part. The input information includes a plurality of predetermined feature quantities regarding the operating state of the second movable part for each section obtained by dividing the time axis of the time-series data into a plurality of sections based on the classification result of the operating state of the first movable part, and is a learned machine learning model that has been machine-learned using input information in which the types of the operating states of the second movable part are labeled. Note that the second movable part is, for example, a working machine having a plurality of partial movable parts, and the plurality of feature quantities can include one or more values representing the respective movable amounts of the plurality of partial movable parts corresponding to each other on the time axis by coordinate values on a plurality of different coordinate axes. Here, the plurality of partial movable parts correspond to, for example, the boom 11 and the bucket 12. In this case, when the working machine includes three or more partial movable parts (for example, a bucket, an arm, and a boom), the coordinate axes can include one or more values represented by coordinate values on three or more coordinate axes. Further, the one or more values represented by each coordinate value can include values corresponding to the initial value, the maximum value, the minimum value, the average value, or the final value of each coordinate value for each section. Further, the plurality of feature quantities can include a value representing the number of repetitions of the repeated operation of the second movable part occurring for each section.

[0039] The operation classification unit 36 classifies the operations performed by the wheel loader 1 based on the classification results of the operating states of the traveling device 4 (first movable part) and the operating states of the working machine 10 (second movable part). Note that the operation classification unit 36 classifies operations in units of one or more sections. FIG. 23 is a schematic diagram showing an example of operation classification according to an embodiment of the present disclosure. In the example shown in FIG. 23, the types of operations are classified into nine types: "loading", "scooping up", "ground leveling", "sorting", "moving", "stopping", "loading-related operation", "sorting-related operation", and "others". The operation classification flag is information for specifying the type of operation. Note that "loading" is an operation of loading, for example, excavated materials onto the loading platform of a transport vehicle, a hopper, or the like. "Scooping up" is an operation of piling up excavated materials. "Ground leveling" is an operation of leveling the work site. "Sorting" is an operation of, for example, separating excavated materials from non-excavated materials. "Moving" is an operation of moving a predetermined distance or more. "Stopping" is an operation of stopping for a predetermined time or more. "Loading-related operation" is, for example, a loading operation in which a normal single scooping operation is repeated two or more times. "Sorting-related operation" is, for example, a sorting operation in which a larger number of sorting operations than normal are performed. "Others" are operations not classified into the above operations.

[0040] When the transition of the combination of the behavior of the traveling device and the behavior of the work implement in, for example, five consecutive sections becomes the same as the transition shown in FIG. 24 determined in advance, the operation classification unit 36 classifies the five sections as the sections of the loading operation. That is, the operation classification unit 36 determines in advance the combination of the behavior of the traveling device and the behavior of the work implement in one or a plurality of sections corresponding to each type of operation, or the transition of the combination, and when the combination of the behavior of the traveling device and the behavior of the work implement in one or a plurality of sections of the time-series data 24, or the transition of the combination matches, the operation in the section is specified. For example, when there is a transition of a typical combination of the behavior of the traveling device and the behavior of the work implement for each section to be classified, the operation classification unit 36 preferentially selects the section and determines the type. Next, for the sections where the operation has not been classified, a section that is different from the operation classified first and that matches the next typical combination of the behavior of the traveling device and the behavior of the work implement in one or a plurality of sections, or the transition of the combination is searched for, and when there is a match, the type is determined. The operation classification unit 36 classifies the operations in all the target sections by repeating the above processing, for example. As described above, by preferentially determining the types of operations that are easy to determine, highly accurate determination is possible. FIG. 25 is a schematic diagram for explaining an example of the operation classification of the sections (k) to (k + 4) shown in FIGS. 7 and 8. For example, the operation classification unit 36 determines that the transition of the combination of the traveling device behavior determination flag and the work implement behavior determination flag in each section classified according to the classification result of the behavior of the traveling device matches the transition shown in FIG. 24, and classifies the operation in the sections (k) to (k + 4) as the loading operation.

[0041] FIG. 26 is a diagram showing an example of classification results according to an embodiment of the present disclosure. The first operating state classification unit 32, the second operating state classification unit 34, and the work classification unit 36 generate time-series data 24R in which the value of the behavior determination flag of the traveling device, the value of the behavior determination flag of the work machine, and the value of the work classification flag are set for each time-series data (for each sampling), as shown in FIG. 26 for example, and store it in the storage unit 38. Note that the classification result is not limited to the example shown in FIG. 26. For example, the classification result may be represented in a manner such that the value of the behavior determination flag of the traveling device, the value of the behavior determination flag of the work machine, and the value of the work classification flag are recorded in association with information such as the time stamp of the time-series data 24 that specifies each section.

[0042] The detailed analysis unit 37 statistically analyzes the time-series data 24 in units of sections, for example, based on the operation of the operator. FIGS. 27 to 31 are diagrams showing examples of the results of the detailed analysis by the detailed analysis unit 37. FIG. 27 is a pie chart showing the ratio of work for a specific period, a specific work machine, and a specific operator. FIG. 28 is a histogram showing the frequency of the boom angle at the time of soil discharge in the loading work for a specific period, a specific work machine, and a specific operator. The detailed analysis unit 37 calculates the frequency of the boom angle, for example, for each section where the work classification result is "loading work" and the behavior determination flag of the work machine shown in FIG. 17 is classified as "50", "51", or "52" (soil discharge). FIG. 29 is a scatter diagram showing the relationship between the fuel consumption [L / h] and the work amount [ton / h] in the loading work for a specific period, a specific work machine, and a specific operator. The scatter diagram in FIG. 29 includes a regression line. FIG. 30 is a histogram showing the frequency of the cycle time in the loading work for a specific period, a specific work machine, and a specific operator. FIG. 31 is a histogram showing the frequency of the cycle time in the loading work for a specific period, a specific work machine, and two different operators, which can be compared between the operators.

[0043] The output unit 39 outputs the analysis result of the detailed analysis unit 37 to a display device or the like, for example.

[0044] (Operation example of the analysis device) FIG. 32 is a flowchart showing an operation example of an analysis device according to an embodiment of the present disclosure. When the operator specifies an analysis target and instructs the execution of the analysis, first, the acquisition unit 31 acquires a file of the time-series data 24 (step S1). Next, the first operation state classification unit 32 classifies the behavior of the traveling device based on the vehicle speed and the gear position for all target times of the time-series data 24 (step S2). Next, the sectioning unit 33 sections the time axis of the time-series data 24 into a plurality of sections based on the classification result of the behavior of the traveling device (step S3). Next, the second operation state classification unit 34 calculates a plurality of feature amounts for each section based on the boom angle and the bucket angle (step S4). Next, the second operation state classification unit 34 classifies the behavior of the work machine for each section using the learned machine learning model 35 (step S5). For the learning of the machine learning model, teacher data may be prepared, and for example, in the process of step S5, machine learning may be performed to generate the learned machine learning model 35, and then classification processing may be performed, or for example, the learned machine learning model 35 may be generated in advance before starting the processing shown in FIG. 32.

[0045] Next, the work classification unit 36 classifies the work for each one or more sections based on the classification result of the behavior of the traveling device and the classification result of the behavior of the work machine (step S6). Next, the detailed analysis unit 37 statistically analyzes the time-series data, for example, in units of sections, and the output unit 39 outputs the analysis result (step S7).

[0046] (Function and Effect) According to the analysis device, analysis method, program, learned machine learning model, and machine learning method of the present disclosure, the operating state of the work machine 10 (second movable part) is classified for each section obtained by sectioning the time-series data 24 based on the classification result of the traveling device 4 (first movable part). Therefore, the operating state of the work machine 10 (second movable part) can be efficiently classified. Accordingly, based on the time-series data 24, the work of the wheel loader 1 (work machine) having the traveling device 4 (first movable part) and the work machine 10 (second movable part) can be efficiently classified.

[0047] The embodiments of the present invention have been described above with reference to the drawings. However, the specific configuration is not limited to the above embodiments, and also includes design changes and the like within the scope not departing from the gist of the present invention. In addition, part or all of the programs executed by the computer in the above embodiments can be distributed via a computer-readable recording medium or a communication line.

[0048] (Supplementary Note) The analysis apparatus, analysis method, program, learned machine learning model, and machine learning method according to the present disclosure are understood as follows, for example.

[0049] (Supplementary Note 1) An acquisition unit that acquires time-series data representing at least the operating state of the first movable part and the operating state of the second movable part of a working machine having the first movable part and the second movable part; A first operating state classification unit that classifies the operating state of the first movable part based on the time-series data; A sectioning unit that divides the time axis of the time-series data into a plurality of sections based on the classification result of the operating state of the first movable part; A second operating state classification unit that classifies the operating state of the second movable part for each of the sections based on the time-series data; An operation classification unit that classifies the operation by the working machine based on the classification result of the operating state of the first movable part and the classification result of the operating state of the second movable part An analysis apparatus comprising:

[0050] (Supplementary Note 2) The operation classification unit classifies the operation in units of one or more of the sections The analysis apparatus according to Supplementary Note 1.

[0051] (Supplementary Note 3) The second operating state classification unit: Inputs input information including a plurality of predetermined feature amounts related to the operating state of the second movable part for each of the sections, Is a machine learning model that outputs output information representing the result of classifying the operating state of the second movable part, Using a learned machine learning model that has been machine-learned using the input information labeled with the type of operating state of the second movable part, classify the operating state of the second movable part The analysis device according to appendix 1 or appendix 2.

[0052] (Appendix 4) The first movable part is a traveling part, The second movable part is a working machine The analysis device according to any one of appendices 1 to 3.

[0053] (Appendix 5) A detailed analysis unit that statistically analyzes the time series data in units of the intervals The analysis device according to any one of appendices 1 to 4, comprising the detailed analysis unit.

[0054] (Appendix 6) A step of obtaining time series data representing at least the operating state of the first movable part and the operating state of the second movable part of a working machine having a first movable part and a second movable part; A step of classifying the operating state of the first movable part based on the time series data; A step of dividing the time axis of the time series data into a plurality of intervals based on the classification result of the operating state of the first movable part; A step of classifying the operating state of the second movable part for each interval based on the time series data; A step of classifying the work by the working machine based on the classification result of the operating state of the first movable part and the classification result of the operating state of the second movable part An analysis method including the above steps.

[0055] (Appendix 7) A step of obtaining time series data representing at least the operating state of the first movable part and the operating state of the second movable part of a working machine having a first movable part and a second movable part; A step of classifying the operating state of the first movable part based on the time series data; A step of dividing the time axis of the time series data into a plurality of sections based on the classification result of the operating state of the first movable part; A step of classifying the operating state of the second movable part for each of the sections based on the time series data; A step of classifying the work by the working machine based on the classification result of the operating state of the first movable part and the classification result of the operating state of the second movable part A program for causing a computer to execute.

[0056] (Appendix 8) Inputting predetermined input information based on at least the time series data representing the operating state of the first movable part and the operating state of the second movable part of a working machine having a first movable part and a second movable part, A machine learning model that outputs output information representing the result of classifying the operating state of the second movable part, The input information includes a plurality of predetermined feature amounts related to the operating state of the second movable part for each of the sections obtained by dividing the time axis of the time series data into a plurality of sections based on the classification result of the operating state of the first movable part, Machine learning has been performed using the input information in which the types of the operating states of the second movable part are labeled Trained machine learning model.

[0057] (Appendix 9) The second movable part is a working machine having a plurality of partial movable parts, The plurality of feature amounts include one or more values representing each movable amount of the plurality of partial movable parts corresponding to each other on the time axis by coordinate values on different coordinate axes The trained machine learning model according to Appendix 8.

[0058] (Appendix 10) The one or more values represented by the respective coordinate values include values corresponding to the initial value, maximum value, minimum value, average value, or final value of each coordinate value for each of the sections The trained machine learning model according to Appendix 9.

[0059] (Appendix 11) The plurality of feature amounts include a value representing the number of times of the repeated operation of the second movable part that occurred for each of the intervals. The learned machine learning model according to any one of Appendices 8 to 10.

[0060] (Appendix 12) A machine learning method of a machine learning model that inputs predetermined input information based on time-series data at least representing the operating state of the first movable part and the operating state of the second movable part of a working machine having a first movable part and a second movable part, and outputs output information representing a result of classifying the operating state of the second movable part, The input information includes a plurality of predetermined feature amounts related to the operating state of the second movable part for each of the intervals obtained by dividing the time axis of the time-series data into a plurality of intervals based on the classification result of the operating state of the first movable part. The machine learning model is machine-learned using the input information with the types of the operating states of the second movable part labeled. Machine learning method.

Description of Signs

[0061] 1 Wheel loader, 24 Time-series data, 30 Analysis device, 31 Acquisition unit, 32 First operating state classification unit, 33 Division unit, 34 Second operating state classification unit, 35 Learned machine learning model, 36 Work classification unit, 37 Detailed analysis unit

Claims

1. An acquisition unit that acquires time-series data representing at least the operating state of the first movable part and the operating state of the second movable part of a working machine having the first movable part and the second movable part; A first operating state classification unit that classifies the operating state of the first movable part based on the time-series data; A sectioning unit that divides the time axis of the time-series data into a plurality of sections based on the classification result of the operating state of the first movable part; A second operating state classification unit that classifies the operating state of the second movable part for each section based on the time-series data; A work classification unit that classifies the work performed by the working machine based on the classification result of the operating state of the first movable part and the classification result of the operating state of the second movable part An analysis device comprising.

2. The work classification unit classifies the work in units of one or more of the sections The analysis device according to claim 1.

3. The second operating state classification unit is Inputting input information including a plurality of predetermined feature quantities related to the operating state of the second movable part for each section, A machine learning model that outputs output information representing the result of classifying the operating state of the second movable part, Using a learned machine learning model learned using the input information with the types of operating states of the second movable part labeled, Classify the operating state of the second movable part The analysis device according to claim 2.

4. The first movable part is a traveling part, The second movable part is a working machine The analysis device according to claim 3.

5. A detailed analysis unit that statistically analyzes the time-series data in units of the sections The analysis device according to claim 4, comprising.

6. A step of acquiring time-series data representing at least the operating state of the first movable part and the operating state of the second movable part of a working machine having the first movable part and the second movable part; A step of classifying the operating state of the first movable part based on the time-series data; A step of dividing the time axis of the time-series data into a plurality of sections based on the classification result of the operating state of the first movable part; A step of classifying the operating state of the second movable part for each section based on the time-series data; A step of classifying the work performed by the working machine based on the classification result of the operating state of the first movable part and the classification result of the operating state of the second movable part An analysis method including.

7. A step of acquiring time-series data representing at least the operating state of the first movable part and the operating state of the second movable part of a working machine having the first movable part and the second movable part; A step of classifying the operating state of the first movable part based on the time series data; A step of dividing the time axis of the time series data into a plurality of sections based on the classification result of the operating state of the first movable part; A step of classifying the operating state of the second movable part for each of the sections based on the time series data; A step of classifying the work by the working machine based on the classification result of the operating state of the first movable part and the classification result of the operating state of the second movable part A program for causing a computer to execute.

8. Inputting predetermined input information based on time series data representing at least the operating state of the first movable part and the operating state of the second movable part of a working machine having a first movable part and a second movable part, A machine learning model that outputs output information representing the result of classifying the operating state of the second movable part, The input information includes a plurality of predetermined feature amounts related to the operating state of the second movable part for each of the sections obtained by dividing the time axis of the time series data into a plurality of sections based on the classification result of the operating state of the first movable part, Machine learning is performed using the input information in which the types of the operating states of the second movable part are labeled A trained machine learning model.

9. The second movable part is a working machine having a plurality of partial movable parts, The plurality of feature amounts include one or more values representing the respective movable amounts of the plurality of partial movable parts corresponding to each other on the time axis by coordinate values on different coordinate axes The trained machine learning model according to claim 8.

10. The one or more values represented by the respective coordinate values include values corresponding to the initial value, maximum value, minimum value, average value, or final value of each coordinate value for each of the sections The trained machine learning model according to claim 9.

11. The plurality of feature amounts include a value representing the number of repetitions of the repetitive operation of the second movable part occurring for each of the sections The trained machine learning model according to claim 10.

12. A machine learning method of a machine learning model that inputs predetermined input information based on time series data representing at least the operating state of the first movable part and the operating state of the second movable part of a working machine having a first movable part and a second movable part, and outputs output information representing the result of classifying the operating state of the second movable part, The input information includes a plurality of predetermined feature amounts related to the operating state of the second movable part for each of the sections obtained by dividing the time axis of the time series data into a plurality of sections based on the classification result of the operating state of the first movable part, A machine learning method of performing machine learning on the machine learning model by using the input information labeled with the type of the operating state of the second movable part. Machine learning method.

Images