Estimation device, estimation method, and program

Abstract

An estimation device according to the present invention estimates the degree of damage to an engine mounted on a work machine, and comprises: a first acquisition unit that acquires the cumulative amount of fuel consumption of the engine; a second acquisition unit that acquires the accumulated amount of errors which may cause failure of the engine; and an estimation unit that estimates the degree of damage to a component constituting the engine on the basis of the amount of fuel consumption acquired by the first acquisition unit, the accumulated amount of errors acquired by the second acquisition unit, and a prescribed evaluation function.

Description

Estimation device, estimation method, and program

[0001] This disclosure relates to an estimation device, an estimation method, and a program. This application claims priority to Japanese Patent Application Nos. 2024-108084, 2024-108221, and 2024-108314, filed in Japan on July 4, 2024, the contents of which are incorporated herein by reference.

[0002] Work machines undergo periodic maintenance. Various technologies for notifying users when maintenance is due have been proposed. For example, a system has been proposed that collects information related to the operating time of a work machine and prompts the user to perform maintenance when the accumulated operating time reaches a predetermined time (Patent Document 1).

[0003] Japanese Patent Application Laid-Open No. 2003-119831

[0004] However, it is difficult to quantitatively measure the extent of damage to internal engine components. As a result, in practice, there are cases where work machines with little damage are subjected to maintenance (e.g., overhaul), and conversely, work machines with heavy damage are not maintained, making it difficult to maintain each work machine efficiently. An object of the present disclosure is to provide technology that enables efficient maintenance of each work machine.

[0005] According to a first aspect of the present disclosure, the estimation device is an estimation device that estimates the degree of damage to an engine mounted on a work machine, and includes an acquisition unit that acquires information on each part around the engine when it is in use, a first estimation unit that estimates the degree of damage to each of the parts based on the information on use acquired by the acquisition unit and an evaluation function corresponding to each of the parts, a second estimation unit that estimates the degree of damage to the engine based on the degree of damage to each of the parts estimated by the first estimation unit, and an output unit that outputs the degree of damage to the engine estimated by the second estimation unit.

[0006] According to the above aspect, maintenance can be performed efficiently on each work machine.

[0007] 1 is a diagram showing an example of a network configuration of a damage degree estimation system according to an embodiment. FIG. 1 is a block diagram showing the hardware configuration of a computer device of each device provided in the damage degree estimation system 1. FIG. 2 is a diagram showing an example of the functional configuration of the damage degree estimation system. FIG. 3 is a diagram showing an example of damage degree estimation performed by a damage degree estimation server. FIG. 4 is a flowchart showing trend data transmission processing performed by a work machine. FIG. 5 is a flowchart showing total damage degree calculation processing performed by a damage degree estimation server. FIG. 6 is a diagram showing an example of an estimation result of the total damage degree. FIG. 7 is a diagram showing an example of a core part. FIG. 8 is a diagram showing an example of a fault code table. FIG. 9 is a diagram showing an example of a calculation result relating to the damage degree of a core part. FIG. 10 is a diagram showing an example of a main metal included in an engine. FIG. 11 is a diagram showing an example of an imaging result relating to exposure of a copper layer. FIG. 12 is a diagram showing an example of a factor causing an increase in wear of the main metal. FIG. 13 is a diagram showing an example of a calculation result 1400 relating to the damage degree of the main metal.

[0008] <Embodiment> <Damage Degree Estimation System 1> Figure 1 is a diagram showing an example of the network configuration of a damage degree estimation system according to an embodiment. As shown in Figure 1, the damage degree estimation system 1 includes a damage degree estimation server 100, a plurality of work machines 110, and a user terminal 120. The damage degree estimation system 1 is installed at a work site where the work machines 110 work, such as a construction site or a quarry. The damage degree estimation system 1 estimates the damage degree of the engine for each of a plurality of work machines 110 managed by a user.

[0009] In the damage degree estimation system 1 shown in Figure 1, a damage degree estimation server 100, a plurality of work machines 110, and a user terminal 120 are communicatively connected via a network 140. Each device is a computer device equipped with a CPU (Central Processing Unit), ROM (Read Only Memory), RAM (Random Access Memory), a communication unit, etc.

[0010] The work machine 110 is, for example, a hydraulic excavator, bulldozer, dump truck, wheel loader, or other type of heavy machinery. The work machine 110 includes an engine 111, various sensors that detect the operation of the engine 111, an engine ECU (Electronic Control Unit) 112, and a vehicle control ECU 113. The engine 111 is, for example, a six-cylinder gasoline engine. The sensors output sensor values ​​to the engine ECU 112. The engine ECU 112 calculates the sensor values ​​to obtain calculated values. The engine ECU 112 outputs the calculated values ​​to the vehicle control ECU 113. The vehicle control ECU 113 further calculates maximum, minimum, average, and other values ​​from the calculated values ​​obtained from the engine ECU 112 at regular intervals to obtain trend data. The vehicle control ECU 113 transmits the trend data to the damage degree estimation server 100 at regular intervals (for example, every 20 hours).

[0011] The damage degree estimation server 100 is an example of an estimation device. The damage degree estimation server 100 estimates the degree of damage to the engine 111 mounted on each work machine 110. The damage degree estimation server 100 is equipped with a trend data DB (database) 101 and an estimation result DB 102. The trend data DB 101 stores (updates) trend data received from the work machine 110 as needed. The damage degree estimation server 100 estimates the degree of damage to the engine 111 for each work machine 110 based on the trend data stored in the trend data DB 101. The damage degree of the engine 111 is the degree of deterioration of the engine 111, and can be expressed as a numerical value between 0 and 1, ranging from a state in which no failure is expected (a state without deterioration) to a state in which a failure may occur (a state with deterioration), for example. The estimation result DB 102 stores estimation results relating to the damage degree of the engine 111 estimated for each work machine 110.

[0012] The user terminal 120 is a terminal device operated by the manager (user) of the work machine 110. The user terminal 120 is a computer device such as a personal computer equipped with a display. The user terminal 120 is located, for example, in an administrative office that manages the work machine 110. The user terminal 120 may also be a portable terminal device such as a smartphone or tablet terminal. The user terminal 120 is capable of displaying the estimation results related to the degree of damage to the engine 111 estimated by the damage degree estimation server 100. Based on the estimation results, the user specifies the work machine 110 to be subjected to maintenance from among the multiple work machines 110. The estimation results are also output to the work machine 110. The work machine 110 can also display the estimation results on an on-board monitor equipped on the work machine 110.

[0013] 2 is a block diagram showing the hardware configuration of the computer device of each device provided in the damage degree estimation system 1. A computer 200 is implemented in each of the damage degree estimation server 100, the work machine 110 (body ECU), and the user terminal 120.

[0014] The computer 200 includes a processor 201, a main memory 202, a storage 203, and an interface 204. The processor 201 reads various programs stored in the storage 203, loads them into the main memory 202, and executes processing in accordance with the programs. The processor 201 also allocates a storage area in the main memory 202 in accordance with the programs. Examples of the processor 201 include a CPU (Central Processing Unit), a GPU (Graphic Processing Unit), and a microprocessor.

[0015] Storage 203 includes, for example, a magnetic disk, a magneto-optical disk, an optical disk, or a semiconductor memory. Storage 203 may be an internal medium directly connected to the bus of computer 200, or an external medium connected to computer 200 via interface 204 or a communication line. Furthermore, when this program is distributed to computer 200 via a communication line, computer 200 that receives the program may load the program into main memory 202 and execute the process.

[0016] If the apparatuses in which computer 200 is implemented are a work machine 110 and a user terminal 120, input devices and output devices are connected to computer 200 via interface 204. Input devices include, for example, a touch panel and operation buttons. In the case of a work machine 110, input devices further include, for example, a camera and a GNSS (Global Navigation Satellite System) unit. Output devices include, for example, a display and a speaker. In the case of a user terminal 120, output devices may further include, for example, a printer.

[0017] <<Functional Configuration of Damage Degree Estimation System 1>> Figure 3 is a diagram showing an example of the functional configuration of the damage degree estimation system. As shown in Figure 3, the damage degree estimation server 100 includes functional units, namely, an acquisition unit 301, an estimation unit 302, and an output unit 303. The acquisition unit 301 includes functional units, namely, a first acquisition unit 301a and a second acquisition unit 301b. The estimation unit 302 includes functional units, namely, a first estimation unit 302a and a second estimation unit 302b. Each functional unit is realized by the processor 201 of the damage degree estimation server 100. In other words, each functional unit is realized by the processor 201 of the damage degree estimation server 100 executing a predetermined program (damage degree estimation program) stored in the storage 203.

[0018] The work machine 110 periodically detects trend data, which is information on the use of each part around the engine. The parts around the engine include the engine 111 and the devices peripheral to the engine 111. The trend data includes, for example, the engine 111 rotation speed, engine oil pressure, blow-by pressure, coolant temperature, exhaust gas temperature, turbo rotation speed, fuel consumption, and fault codes (described below). The work machine 110 periodically transmits the detected trend data to the damage extent estimation server 100.

[0019] When the damage degree estimation server 100 receives trend data from the work machine 110, it stores the trend data in the trend data DB 101. The acquisition unit 301 acquires, from the trend data DB 101, trend data corresponding to each part.

[0020] The first estimation unit 302a estimates the damage level of each part based on the trend data acquired by the acquisition unit 301 and an evaluation function corresponding to each part. Here, the evaluation function will be described.

[0021] <Evaluation Functions> Fig. 4 is a diagram showing an example of damage degree estimation performed by the damage degree estimation server 100. As shown in Fig. 4, the damage degree estimation server 100 estimates the damage degree of each part using a plurality of evaluation functions 400 (400a to 400e). Specifically, the evaluation functions 400 include, for example, a core component damage degree calculation evaluation function 400a, a main metal damage degree calculation evaluation function 400b, an exhaust manifold thermal damage degree calculation evaluation function 400c, a turbo damage degree calculation evaluation function 400d, and a load factor evaluation function 400e.

[0022] A general method can be adopted for the exhaust manifold thermal damage degree calculation evaluation function 400c, the turbo damage degree calculation evaluation function 400d, and the load factor evaluation function 400e. Each of the evaluation functions 400a to 400e will be briefly explained below.

[0023] <<Core Component Damage Degree Calculation Evaluation Function 400a>> The core component damage degree calculation evaluation function 400a, details of which will be described later, is a function capable of estimating the degree of damage to core components (e.g., cylinder heads and other major components that constitute the engine) within the engine 111. The acquisition unit 301 acquires, as trend data, the accumulated amount of errors that may be the cause of failure (details of which will be described later) from the trend data DB 101. The first estimation unit 302a estimates the degree of damage to core components of the engine 111 based on the accumulated amount of errors acquired by the acquisition unit 301 and the core component damage degree calculation evaluation function 400a. The first estimation unit 302a estimates (calculates) the degree of damage to the core components by normalizing the value to a value between 0 (a state in which failure is unlikely) and 1 (a state in which failure may occur).

[0024] <<Main Metal Damage Level Calculation Evaluation Function 400b>> The main metal damage level calculation evaluation function 400b, details of which will be described later, is a function capable of estimating the wear state of the main metal that constitutes the engine 111. The acquisition unit 301 acquires, as trend data, an accumulated value of damage that becomes a cause of failure (details of which will be described later) from the trend data DB 101. The first estimation unit 302a estimates the wear state of the main metal that constitutes the engine 111 based on the accumulated value of damage acquired by the acquisition unit 301 and the main metal damage level calculation evaluation function 400b. The first estimation unit 302a estimates (calculates) the wear state of the main metal by normalizing the value to a value between 0 and 1.

[0025] <<Exhaust Manifold Thermal Damage Calculation Evaluation Function 400c>> The exhaust manifold thermal damage calculation evaluation function 400c is a function that converts exhaust gas temperature changes in the engine's exhaust manifold (a manifold that combines multiple exhaust flow paths into one) into stress values ​​and calculates (estimates) the cumulative damage level using the rainflow method. The acquisition unit 301 acquires the exhaust gas temperature as trend data from the trend data DB 101. In this embodiment, the vehicle control ECU 113 calculates the damage level over a 20-hour period, and the acquisition unit 301 acquires the damage level every 20 hours as trend data. The engine ECU 112 may perform the calculation instead of the vehicle control ECU 113. The first estimation unit 302a estimates the damage level of the exhaust manifold based on the exhaust gas temperature acquired by the acquisition unit 301 and the exhaust manifold thermal damage calculation evaluation function 400c. The first estimation unit 302a normalizes the damage level to a value between 0 and 1 and estimates (calculates) the damage level.

[0026] <<Turbo Damage Degree Calculation Evaluation Function 400d>> The turbo damage degree calculation evaluation function 400d is a function that converts fluctuations in turbo rotation speed into stress values ​​applied to the turbine or compressor and calculates (estimates) the cumulative damage degree using the rainflow method. The acquisition unit 301 acquires the turbo rotation speed as trend data from the trend data DB 101. In this embodiment, the vehicle body control ECU 113 calculates the damage degree over a 20-hour period, and the acquisition unit 301 acquires the damage degree for each 20-hour period as trend data. The engine ECU 112 may perform the calculation instead of the vehicle body control ECU 113. The first estimation unit 302a estimates the damage degree related to the turbocharger and its surroundings based on the turbo rotation speed acquired by the acquisition unit 301 and the turbo damage degree calculation evaluation function 400d. The first estimation unit 302a normalizes the damage degree to a value between 0 and 1 to estimate (calculate) the damage degree.

[0027] <<Load Factor Evaluation Function 400e>> The load factor evaluation function 400e is a function that can calculate (estimate) the cumulative damage level based on the extent to which the load due to fuel consumption was compared with the design assumption (for example, "cumulative fuel consumption" / "estimated cumulative fuel consumption"). The acquisition unit 301 acquires fuel consumption as trend data from the trend data DB 101. The first estimation unit 302a estimates the damage level related to fuel consumption based on the fuel consumption acquired by the acquisition unit 301 and the load factor evaluation function 400e. The first estimation unit 302a normalizes the damage level to a value between 0 and 1 and estimates (calculates) the damage level.

[0028] <<Second estimating unit 302b>> The second estimating unit 302b estimates the damage level of the engine 111 (total damage level 410 in FIG. 4 ) based on the damage level of each part estimated by the first estimating unit 302a. Specifically, the second estimating unit 302b estimates the highest damage level among the damage levels of each part estimated by the first estimating unit 302a through normalization as the total damage level 410.

[0029] For example, the damage degrees estimated by the first estimating unit 302a using each of the evaluation functions 400a to 400e are assumed to be the following values: Core part: "0.5" Main metal: "0.8" Exhaust manifold: "0.6" Turbo damage degree: "0.3" Load factor: "0.2" In this case, the maximum value among these values ​​is "0.8" for the main metal. Therefore, the second estimating unit 302b estimates "0.8" as the total damage degree 410 for that work machine 110. The estimating unit 302 estimates the total damage degree 410 for each work machine 110.

[0030] The output unit 303 outputs the total damage degree 410 of each work machine 110 estimated by the second estimating unit 302b to the estimation result DB 102. Specifically, the output unit 303 outputs the total damage degree 410 for each work machine 110 to the estimation result DB 102.

[0031] The damage degree estimation server 100 transmits the total damage degree 410 stored in the estimation result DB 102 to the user terminal 120 and the work machine 110. The timing of the transmission may be when a transmission request is received from the user terminal 120, or may be a preset timing such as a specified time.

[0032] The user terminal 120 is capable of displaying the total damage degree 410 on the display in response to a display instruction from the user. The user terminal 120 is capable of displaying the total damage degree 410 of each work machine 110 in a list format. This allows the user to easily understand which work machine 110 should be given priority for maintenance.

[0033] The work machine 110 receives the total damage degree 410 of its own vehicle from the damage degree estimation server 100. The work machine 110 is capable of displaying the total damage degree 410 of its own vehicle on an on-board monitor in response to a display instruction from the user.

[0034] <<Method>> Next, the processing performed by the work machine 110 and damage degree estimation server 100 according to this embodiment will be described. Fig. 5 is a flowchart showing the trend data transmission processing performed by the work machine 110. In Fig. 5, the work machine 110 acquires sensor values ​​from various sensors while the engine 111 is operating (step S501). The work machine 110 then creates trend data (step S502). Thereafter, the work machine 110 transmits the trend data to the damage degree estimation server 100 (step S503), and the series of processing steps ends.

[0035] 6 is a flowchart showing the overall damage degree calculation process performed by the damage degree estimation server 100. In FIG. 6, the damage degree estimation server 100 acquires trend data corresponding to each part from the trend data DB 101 (step S601). Then, the damage degree estimation server 100 uses the acquired trend data to calculate each evaluation function 400 (400a to 400e) (step S602).

[0036] Next, the damage degree estimation server 100 normalizes the calculation results and estimates the damage degree of each part (step S603). Then, the damage degree estimation server 100 estimates the highest damage degree among the normalized damage degrees of each part as the total damage degree 410 (step S604). Next, the damage degree estimation server 100 stores the total damage degree 410 in the estimation result DB 102 (step S605). Furthermore, the damage degree estimation server 100 transmits the total damage degree 410 stored in the estimation result DB 102 to the user terminal 120 and the work machine 110 (step S606), and the series of processes ends.

[0037] <<Example of Estimation Results of Total Damage Degree>> Figure 7 is a diagram showing an example of the estimation results of the total damage degree. The estimation result 700 shown in Figure 7 shows the total operating time on the horizontal axis and the total damage degree on the vertical axis. It is assumed that when the total operating time reaches a predetermined time (for example, 8,000 hours), maintenance (overhaul) is performed compulsorily. In the figure, each plot indicates the total operating time and the total damage degree of each work machine 110. Each plot 710 corresponds to a work machine 110.

[0038] The total damage degree can be displayed on the display of the user terminal 120. Therefore, the user can prioritize maintenance of work machines 110 with a high total damage degree among work machines 110 with a total operating time of 8000 hours. For example, the multiple (three) plots shown in the plot group 711 all indicate work machines 110 with a total operating time close to 8000 hours. Of the multiple plots shown in the plot group 711, plot 711a has the highest total damage degree. Therefore, the user can prioritize maintenance of the work machine 110 corresponding to plot 711a.

[0039] The output unit 303 (FIG. 3) outputs the total damage level estimated by the second estimation unit 302b and a predetermined threshold value in a viewable manner. The predetermined threshold value includes a value indicating that maintenance is not required (first threshold value) and a value indicating that maintenance is required immediately (second threshold value). The threshold value is a value that can be set arbitrarily.

[0040] As shown in the figure, the estimation result 700 includes a first threshold 701 (e.g., 0.3) and a second threshold 702 (e.g., 0.8). For example, if the total damage level is equal to or less than the first threshold 701, it indicates that the damage level to the engine is low. On the other hand, if the total damage level is equal to or greater than the second threshold 702, it is estimated that the damage level to the engine is high, i.e., it is estimated that maintenance is required, for example.

[0041] In addition, in the drawing, the work machine 110 indicated by the plot 712 has a total damage level of 1.0. Therefore, the plot 712 indicates that maintenance is required immediately. This allows the user to prioritize maintenance of the work machine 110 corresponding to the plot 712.

[0042] <<First Action and Effect>> As explained above, the damage degree estimation server 100 according to this embodiment estimates the degree of damage to each part based on information (trend data) about when each part is in use and the evaluation function 400 corresponding to each part, and estimates the damage degree of the engine (total damage degree 410) based on the estimated damage degree of each part. This makes it possible to quantitatively measure the degree of damage. For example, if there is even one part among the parts with a high degree of damage, that work machine 110 can be given priority as the target for maintenance. Specifically, a work machine 110 with high damage can be targeted for maintenance, and conversely, a work machine 110 with low damage can be excluded from the target for maintenance. Therefore, the damage degree estimation server 100 according to this embodiment can efficiently maintain each work machine.

[0043] Furthermore, the damage degree estimation server 100 normalizes and estimates the damage degree of each part, and estimates the highest damage degree among the normalized estimated damage degrees of each part as the total damage degree 410. This allows the damage degree of each part to be expressed as a numerical value between 0 and 1, making it easy to compare the damage degrees of each part and improving the estimation accuracy of the total damage degree 410.

[0044] Furthermore, the damage degree estimation server 100 outputs the estimated total damage degree 410 and a predetermined threshold value in a viewable manner, thereby enabling the user to easily determine whether or not maintenance (overhaul) is required for each work machine 110.

[0045] <<Estimation of Damage Level of Core Components>> Next, estimation of damage level of core components will be described. First, core components will be described. FIG. 8 is a diagram showing an example of a core component. As shown in FIG. 8, the engine 111 includes core components 810 (810a to 810e) and other internal components 820 (820a, 820b).

[0046] The core parts 810 include the following parts: a cylinder block 810a, a cylinder head 810b, a crankshaft 810c, a connecting rod 810d, and a camshaft 810e.

[0047] The other internal components 820 include, for example, the following components: a valve 820a; a piston 820b;

[0048] The core component 810 will be described in detail below. The pair of cylinder block 810a and cylinder head 810b form the framework of the engine 111. The connecting rod 810d is connected to the piston 820b and crankshaft 810c and converts the reciprocating motion of the piston 820b into the rotational motion of the crankshaft 810c. The crankshaft 810c is connected to the connecting rod 810d and is a rotating shaft that transmits the rotational motion converted by the connecting rod 810d. The camshaft 810e is a shaft that manages the opening and closing operation of the valve 820a, which draws the air-fuel mixture into the combustion chamber and exhausts it from the combustion chamber after combustion.

[0049] The degree of damage to the core part 810 depends on the cumulative fuel consumption (hereinafter referred to as "cumulative fuel consumption") and also on the accumulated amount of errors that cause failures in the engine 111. The accumulated amount of errors includes an accumulated amount related to "wear" of the engine 111 and an accumulated amount related to "heat" of the engine 111.

[0050] As the accumulated amount of errors related to "wear" of engine 111 increases, the degree of damage to, for example, crankshaft 810c, connecting rod 810d, and camshaft 810e tends to increase. Also, as the accumulated amount of errors related to "heat" of engine 111 increases, the degree of damage to, for example, cylinder block 810a and cylinder head 810b tends to increase. When a sensor detects an error, a fault code corresponding to the error is issued and recorded.

[0051] <Fault Code Table> FIG. 9 is a diagram showing an example of a fault code. As shown in FIG. 9, the fault code table 900 includes the following items: "Damage Type," "Fault Code," "Alarm Activation Condition," "Sensor," "Engine Internal State at Time of Alarm Activation," and "Damaged Core Component." The "Damage Type" indicates the type of cause of damage to the engine 111 (wear or heat). The "Fault Code" indicates a specific detection target for wear damage or heat damage. The "Alarm Activation Condition Item" indicates the conditions for detecting an abnormality for each fault code. The "Sensor" indicates a sensor for detecting the detection target indicated by the fault code. The "Engine Internal State at Time of Alarm Activation" indicates the damage state of the engine 111 when the alarm activation condition is met. The "Damaged Core Component" indicates a component of the core component 810 that will be damaged when the alarm activation condition is met. The "Damaged Core Component" corresponds to the "Damage Type" (wear or heat).

[0052] To explain the fault code table 900 in more detail, the "damage type" indicates either "wear" or "heat." The "fault codes" that belong to the "damage type" of "wear" are "engine overspeed," "engine oil pressure drop," and "blow-by pressure increase." When an error is detected, a fault code corresponding to the error is recorded.

[0053] <<Engine Overspeed>> The condition for an "engine overspeed" alarm is that the vehicle is traveling with the rotation signal of the transmission input shaft exceeding a certain value. "Sensor" refers to a rotation sensor. That is, the rotation speed of the input shaft is measured by the rotation sensor. The internal state of the engine when an "engine overspeed" alarm is issued indicates that the engine is rotating at or above the allowable speed, which accelerates wear of the internal parts, including core parts 810 (crankshaft 810c, connecting rod 810d, and camshaft 810e).

[0054] <<Low Engine Oil Pressure>> The condition for issuing a "Low Engine Oil Pressure" alert is when the engine oil pressure (oil pressure) drops below the operating range for a certain period of time or longer while the engine 111 is running at a certain RPM or higher. "Sensor" refers to an oil pressure sensor. That is, the engine oil pressure is measured by the oil pressure sensor. When an alert for "Low Engine Oil Pressure" is issued, the internal state of the engine is that the engine oil pressure is low, which indicates that insufficient oil lubrication could cause the engine 111 (crankshaft 810c, connecting rod 810d, and camshaft 810e) to seize.

[0055] <<Blow-by Pressure Rise>> "Blow-by pressure" refers to the pressure of the combustion gas (exhaust) and unburned mixture gas that have become high-pressure during the combustion stroke and leak into the crankcase through the gap between the piston rings (compression rings) when the pressure exceeds the sealing capacity of the piston rings. The condition for issuing a "blow-by pressure rise" alarm is that the blow-by pressure value exceeds a certain value. "Sensor" refers to a pressure sensor installed downstream of the crankcase. In other words, the blow-by pressure is measured by this pressure sensor. When an alarm for "blow-by pressure rise" is issued, the internal condition of the engine indicates that wear inside engine 111 has progressed, and that continuing operation in this state may result in damage to core parts 810 (crankshaft 810c, connecting rod 810d, and camshaft 810e) due to wear.

[0056] <<Overheat>> In the fault code table 900, the "fault codes" belonging to the "damage type" of "heat" are "overheat" and "exhaust temperature rise." The conditions for an "overheat" alarm to be issued are that the engine 111 operates at a certain number of revolutions per minute or more for a certain period of time or more, and the coolant temperature of the engine 111 exceeds the operating range. "Sensor" indicates a water temperature sensor that detects the temperature of the coolant. In other words, the coolant temperature is measured by the water temperature sensor. The internal state of the engine when an "overheat" alarm is issued indicates that the engine 111 cannot be cooled, and that the core parts 810 (cylinder block 810a and cylinder head 810b) may be damaged due to thermal fatigue.

[0057] <<Rise in Exhaust Temperature>> The condition for issuing an "Exhaust Temperature Rise" alarm is that the upper limit value of the exhaust temperature is detected for a certain period of time or longer. "Sensor" refers to a temperature sensor. In other words, the exhaust temperature is measured by the temperature sensor. When an "Exhaust Temperature Rise" alarm is issued, the internal condition of the engine indicates that the core parts 810 (cylinder block 810a and cylinder head 810b) are at risk of being damaged due to thermal fatigue.

[0058] <Logistic Regression Model> The first estimation unit 302a estimates the damage level of the core part 810 using a logistic regression model (supervised learning) as the core part damage level calculation evaluation function 400a. The logistic regression model is a statistical model that predicts a target variable with a value between 0 and 1 from multiple explanatory variables. The logistic regression model of this embodiment calculates the damage level of the core part 810 using the cumulative fuel consumption and the number of times a report was issued for each fault code (explanatory variables) as the amount of damage. The logistic regression model learns the contribution of damage to the core part 810 (partial regression coefficient β), which is a coefficient related to the number of times a report was issued for each fault code. An example of a calculation formula for the logistic regression model is shown in Equation (1).

[0059]

[0060] When generating the logistic regression model, for example, an investigation report prepared by a work staff member when overhauling (disassembling) the engine 111 is used. The investigation report includes, for example, the cumulative fuel consumption of the engine 111, the number of times each fault code has been reported, and the damage level (for example, a three-level value between 0 and 1). In generating the logistic regression model, the explanatory variables are, for example, the cumulative fuel consumption of the engine 111 and the number of times each fault code has been reported. Furthermore, the objective variable is, for example, the damage level (for example, a three-level value between 0 and 1). Using the values ​​of the variables described in the investigation report, partial regression coefficients β (β1, β2, β3) are learned by a maximum likelihood method or the like. Note that in Equation (1), there are three variables: x1, x2, and x3. For example, x1 corresponds to the cumulative fuel consumption, x2 corresponds to wear damage, and x3 corresponds to thermal damage.

[0061] 3, a functional configuration of the damage degree estimation server 100 related to estimating the damage degree of the core part 810 will be described. The first acquisition unit 301a acquires the cumulative fuel consumption amount of the engine 111 from the trend data DB 101.

[0062] The second acquisition unit 301b acquires the accumulated amount of errors that are factors that cause failures in the engine 111. The accumulated amount of errors is the number of times that the alarm condition is met (the cumulative number of alarms). Specifically, the second acquisition unit 301b acquires the accumulated amount of errors related to "wear" of the engine 111 and the accumulated amount of errors related to "heat" of the engine 111 from the trend data stored in the trend data DB 101.

[0063] The accumulated amount of errors related to "wear" of the engine 111 is, for example, as follows: - As an accumulated amount related to the rotation speed of the engine 111, the accumulated number of times that an engine overspeed alarm has been issued. The "engine overspeed" fault code is issued when the rotation signal of the input shaft of the transmission is equal to or greater than a certain value. - As an accumulated amount related to the engine oil pressure, the accumulated number of times that an engine oil pressure drop has been issued. The "engine oil pressure drop" fault code is issued when the engine oil pressure (hydraulic pressure) drops below the operating range for a certain period of time or more when the rotation speed of the engine 111 is equal to or greater than a certain number of rotations. - As an accumulated amount related to the blow-by pressure, the accumulated number of times that a blow-by pressure increase has been issued. The "blow-by pressure increase" fault code is issued when the blow-by pressure value exceeds a certain value. Note that the accumulated amount of errors related to "wear" includes the above three accumulated numbers of times that an alarm has been issued, but it is sufficient to include at least one accumulated number of times that an alarm has been issued.

[0064] The accumulated amount of errors related to "heat" of the engine 111 is, for example, as follows: - The accumulated number of times an overheat alarm has been issued, as an accumulated amount related to the engine 111 coolant temperature. The "overheat" fault code is issued when the engine 111 operates at a speed of more than a certain number of revolutions for more than a certain period of time, and the engine 111 coolant exceeds the operating range. - The accumulated number of times an exhaust temperature rise alarm has been issued, as an accumulated amount related to the exhaust temperature. The "exhaust temperature rise" fault code is issued when the upper limit value of the engine 111 exhaust temperature is detected for more than a certain period of time. Note that the accumulated amount of errors related to "heat" includes the above two accumulated numbers of times an alarm has been issued, but it is sufficient to include at least one accumulated number of times an alarm has been issued.

[0065] Note that errors that cause engine 111 failure may also be caused by the external environment, such as dust in the air getting into engine 111. For this reason, if the work machine 110 is configured to be able to detect errors related to the external environment, the accumulated amount of errors may include the accumulated amount of errors related to the external environment. In this case, the second acquisition unit 301b may acquire the accumulated amount of errors related to the external environment from the trend data stored in trend data DB 101.

[0066] The estimation unit 302 (first estimation unit 302a) estimates the damage level of the core part 810 based on the cumulative fuel consumption amount acquired by the first acquisition unit 301a, the accumulated amount acquired by the second acquisition unit 301b, and the core part damage level calculation evaluation function 400a. Specifically, the first estimation unit 302a estimates the damage level of the core part 810 based on the fuel consumption amount, the accumulated amount of errors related to "wear," the accumulated amount of errors related to "heat," and a logistic regression model.

[0067] The first estimation unit 302a estimates a normalized damage level (a value between 0 and 1) by inputting the cumulative fuel consumption of the engine 111 and the cumulative number of times each fault code has been reported, both obtained from the trend data, into a trained logistic regression model. The first estimation unit 302a estimates a rating based on the result of comparing the damage level (a value between 0 and 1) with a predetermined threshold. The rating is expressed, for example, in three levels, "A," "B," and "C," in descending order of quality. Specifically, for example, the first estimation unit 302a estimates the rating as "C" if the damage level (a value between 0 and 1) is equal to or greater than the predetermined threshold, and estimates the rating as "A" or "B" if the damage level is less than the predetermined threshold.

[0068] The contribution of an event related to the explanatory variable to the damage can be calculated by multiplying the value x of the explanatory variable by the partial regression coefficient β. Therefore, the first estimation unit 302a can estimate which event is causing the damage by calculating and comparing the contribution of each explanatory variable.

[0069] <<Example of Calculation Results Related to Damage Degree of Core Component 810>> FIG. 10 is a diagram showing an example of a calculation result 1000 related to the damage degree of the core component 810. In the calculation result 1000 shown in FIG. 10 , the horizontal axis represents the total operating time, and the vertical axis represents the damage degree PC of the core component 810. When the damage degree PC of the core component 810 is equal to or greater than a first threshold value 1011 (e.g., 0.65), this indicates that the rating is "C." Specifically, when the damage degree PC is equal to or greater than the first threshold value 1011, this indicates, for example, that there are multiple components in the core component 810 that need to be replaced. When the rating is "C," the damage degree estimation server 100 outputs a value of "1" indicating that the core component 810 should be replaced (a value indicating a rating of "C") as the estimation result of the damage degree of the core component 810.

[0070] Furthermore, when the damage level pc is less than the first threshold value 1011 and equal to or greater than the second threshold value 1012, the rating is "B." Specifically, when the damage level pc is less than the first threshold value 1011 and equal to or greater than the second threshold value 1012 (e.g., 0.5), this indicates, for example, that one of the core parts 810 needs to be replaced.

[0071] Furthermore, the damage level pc being less than the second threshold value 1012 indicates that the rating is "A." Specifically, the damage level pc being less than the second threshold value 1012 indicates, for example, that there are no parts among the core parts 810 that need to be replaced.

[0072] 10 is not limited to 0.65 and may be set to another value. Similarly, the second threshold 1012 is not limited to 0.5 and may be set to another value.

[0073] In the illustration, each line 1001 (1001a, 1001b, 1001c) indicates the total operating time of three work machines 110 and the degree of damage to the core parts 810, respectively. In the illustration, it is estimated that line 1001a will exceed the first threshold value 1011 the earliest. In other words, it is estimated that for the work machine 110 corresponding to line 1001a, the number of parts requiring replacement among the core parts 810 will be multiple the earliest. Also in the illustration, it is estimated that line 1001c will exceed the first threshold value 1011 the latest. In other words, it is estimated that for the work machine 110 corresponding to line 1001c, the number of parts requiring replacement among the core parts 810 will be multiple the latest.

[0074] <<Second Action and Effect>> As described above, the damage degree estimation server 100 according to this embodiment estimates the damage degree of the core part 810 that constitutes the engine 111 based on the cumulative fuel consumption, the accumulated amount of fault codes that are causes of faults in the engine 111, and the predetermined evaluation function 400. This makes it possible to accurately estimate the damage degree of the core part 810. Therefore, the estimation accuracy of the total damage degree 410 can be improved, allowing efficient maintenance of each work machine.

[0075] Furthermore, the damage degree estimation server 100 estimates the damage degree of the core parts 810 based on the cumulative fuel consumption, the accumulated amount of errors related to "wear" of the engine 111, and a predetermined evaluation function 400. This makes it possible to accurately estimate the damage degree of core parts 810 that are susceptible to wear damage (crankshaft 810c, connecting rod 810d, and camshaft 810e) among the core parts 810.

[0076] Furthermore, the damage degree estimation server 100 estimates the degree of damage to the core parts 810 based on the cumulative fuel consumption, the accumulated amount of errors related to "heat" of the engine 111, and a predetermined evaluation function 400. This makes it possible to accurately estimate the degree of damage to core parts 810 (cylinder block 810a and cylinder head 810b) that are susceptible to heat damage.

[0077] In this embodiment, the accumulated amount of errors related to "wear" includes at least one of an accumulated amount related to the speed of engine 111, an accumulated amount related to the oil pressure of engine oil, and an accumulated amount related to blow-by pressure. This makes it possible to easily accumulate factors of wear damage to core part 810 (crankshaft 810c, connecting rod 810d, and camshaft 810e).

[0078] In this embodiment, the accumulated amount of errors related to "heat" includes at least one of an accumulated amount related to the coolant temperature of the engine 111 and an accumulated amount related to the exhaust temperature. This makes it possible to easily accumulate factors that cause thermal damage to the core component 810 (cylinder block 810a and cylinder head 810b).

[0079] In this embodiment, the predetermined evaluation function 400 is a logistic regression model that has learned a partial regression coefficient β that indicates the contribution of damage to the core part 810. The damage degree estimation server 100 normalizes and estimates the damage degree of the core part 810 by inputting the cumulative fuel consumption amount and the accumulated number of failure codes that are the causes of failures into the logistic regression model. This makes it possible to estimate the damage degree of the core part 810 simply and accurately.

[0080] <<Modification Related to Damage Level of Core Component 810>> In the above explanation, an example has been described in which the damage level of the core component 810 is estimated based on the accumulated amount of wear-related errors and the accumulated amount of heat-related errors. Below, an example will be described in which the damage level of the core component 810 is estimated based on either the accumulated amount of wear-related errors or the accumulated amount of heat-related errors. Note that below, an example will be described in which the damage level of the core component 810 is estimated based on the accumulated amount of wear-related errors.

[0081] In generating a logistic regression model according to the modified example, the explanatory variables are the cumulative fuel consumption of the engine 111 and the number of times a wear damage-related fault code has been issued. The objective variable is, for example, the damage level (e.g., a value between 0 and 1). The partial regression coefficients β (β1, β2) are learned by a maximum likelihood method or the like using the values ​​of the variables listed in the investigation report. Note that in equation (1), there are two variables, x1 and x2. For example, x1 corresponds to the cumulative fuel consumption, and x2 corresponds to the wear damage.

[0082] The second acquisition unit 301b acquires the accumulated amount of errors related to wear of the engine 111 from the trend data stored in the trend data DB 101. The first estimation unit 302a estimates the degree of damage to the core parts 810 based on the cumulative fuel consumption, the accumulated amount of errors related to wear, and a logistic regression model. This makes it possible to accurately estimate the degree of damage to the core parts 810 (crankshaft 810c, connecting rod 810d, and camshaft 810e), which increases in accordance with the accumulated amount of errors related to "wear."

[0083] Although not explained further, it is also possible to estimate the degree of damage to the core part 810 based on the amount of accumulated heat-related errors. This makes it possible to accurately estimate the degree of damage to the core part 810 (cylinder block 810a and cylinder head 810b), which increases in accordance with the amount of accumulated heat-related errors.

[0084] <<Estimation of Wear of Main Metal>> Next, estimation of the wear (damage) of the main metal will be described. First, the main metal will be described. FIG. 11 is a diagram showing an example of a main metal included in an engine 111. As shown in FIG. 11, the engine 111 includes a six-cylinder engine. Each of the six cylinders 1100 includes a piston 820b, a connecting rod 810d, a crankshaft 810c, and a main metal 1110. The main metal 1110 is a bearing portion of the crankshaft 810c and is disposed on both sides of each cylinder 1100. Therefore, the total number of main metals 1110 (1110a to 1110g) is "7", which is the number of cylinders "6" + "1". The main metal 1110 has a copper layer, and the surface of the copper layer is coated.

[0085] When the engine 111 is used, the main metal 1110 is worn away. Specifically, the coating on the surface of the main metal 1110 is removed by the wear, thereby exposing the copper layer.

[0086] <<Imaging Results Regarding Exposed Copper Layer>> Fig. 12 is a diagram showing an example of the exposed state of the copper layer. Exposure examples 1200 (1200a, 1200b, 1200c) shown in Fig. 12 each show the state in which seven main metals 1110 (1110a to 1110g) have been removed and arranged. Exposure example 1200a shows zero exposed copper layer locations. Exposure example 1200b shows four exposed copper layer locations. Exposure example 1200c shows seven exposed copper layer locations.

[0087] In this embodiment, the more main metals with exposed copper layers, the greater the degree of damage, and the number of exposed metals is used as an index of the degree of damage. For example, if six or more of the seven main metals 1110 have exposed copper layers, it can be determined that the engine 111 is severely damaged and requires an overhaul. On the other hand, if the number of main metals 1110 with exposed copper layers is less than six (five or less), it can be determined that the engine 111 is not severely damaged, and therefore does not require an overhaul.

[0088] The degree of wear (wear state) of the main metal 1110 depends on the cumulative fuel consumption amount as well as the cumulative value of damage that can cause a malfunction (wear-increasing factors) of the engine 111. Factors that cause wear-increasing factors include the engine operating at a predetermined rotation speed or higher, or the boost pressure being lower than a predetermined value. Here, the wear-increasing factors will be described with reference to FIG. 13 .

[0089] 13 is a diagram showing an example of wear increase factors of the main metal 1110. Wear increase factors 1300 shown in FIG. 13 include items such as "measurement target," "set value," "data," and "whether or not to add to cumulative value."

[0090] The "measurement target" indicates factors that increase the degree of wear of the main metal 1110. Specifically, the "measurement target" includes the following items: engine speed, boost pressure, blow-by pressure, engine oil pressure (normal), engine oil pressure (low idle), and engine oil pressure (high idle).

[0091] The engine speed indicates the average rotation speed of the engine 111 during one day of use. The boost pressure is the pressure of compressed air (supercharging) forcibly sent to the internal combustion engine by the supercharger. As mentioned above, the blow-by pressure is the pressure of gas such as combustion gas leaking into the crankcase. The engine oil pressure (normal) indicates the oil pressure of the engine oil circulating within the engine 111 under normal conditions. The engine oil pressure (low idle) indicates the oil pressure of the engine oil circulating within the engine 111 when the engine 111 is used with the rotation speed set to a low rotation speed. The engine oil pressure (high idle) indicates the oil pressure of the engine oil circulating within the engine 111 when the engine 111 is used with the rotation speed set to a high rotation speed.

[0092] The "set value" indicates an upper limit or lower limit value that will not cause damage to the main metal 1110. In other words, exceeding the upper limit or falling below the lower limit will cause damage to the main metal 1110. The "data" includes a "raw value" that is an actually detected measured value, and a "difference" that indicates the difference between the "set value" and the "raw value." In other words, "raw value" - "set value" = "difference." The presence or absence of "addition to cumulative value" indicates the presence or absence of addition to the cumulative value depending on whether the value exceeds the upper limit of the "set value" or falls below the lower limit of the "set value."

[0093] Each of the "measurement objects" shown in the figure will be explained. For "engine RPM," if the set value is, for example, "2500 (upper limit)" and the "raw value" is, for example, "2600," the "difference" will be "+100." In this way, when the difference is a positive value, "whether to add to cumulative value" is set to "yes." That is, in this case, the difference "+100" is added to the cumulative value related to "engine RPM."

[0094] For "boost pressure," the set value is set to, for example, the lower limit. If the "raw value" falls below the lower limit, for example, the "difference" will be a negative value. In this way, when the difference is a negative value, "whether to add to cumulative value" is set to "yes." That is, in this case, the difference is added to the cumulative value related to "boost pressure." Note that when adding to the cumulative value, if the difference is a negative value, it is multiplied by "-1" to process it as a positive value.

[0095] For "blow-by pressure," the set value is, for example, the upper limit. If the "raw value" is, for example, below the upper limit, the "difference" becomes a negative value. In this way, when the difference is a negative value, "whether to add to cumulative value" becomes "no." In other words, in this case, the difference is not added to the cumulative value for "blow-by pressure."

[0096] For "Engine oil pressure (normal)", the set value is, for example, the lower limit. If the "raw value" exceeds the lower limit, for example, the "difference" becomes a positive value. In this way, when the difference is a positive value, "whether to add to cumulative value" indicates "no". In other words, in this case, the difference is not added to the cumulative value for "Engine oil pressure (normal)".

[0097] For "engine oil pressure (low idling)", the set value is set to, for example, the lower limit. If the "raw value" is below the lower limit, the "difference" will be a negative value. In this way, when the difference is a negative value, "whether to add to cumulative value" is set to "yes". That is, in this case, the difference is added to the cumulative value for "engine oil pressure (low idling)". Note that when adding to the cumulative value, if the difference is a negative value, it is multiplied by "-1" to process it as a positive value.

[0098] For "engine oil pressure (high idle)", the set value is, for example, the lower limit. If the "raw value" is below the lower limit, the "difference" becomes a negative value. In this way, when the raw value is a negative value, "addition to cumulative value" becomes "yes". In other words, in this case, the difference is added to the cumulative value for "engine oil pressure (high idle)".

[0099] <<Logistic Regression Model for Main Metal Damage Degree Computation Evaluation Function 400b>> The first estimation unit 302a uses a logistic regression model (supervised learning) as the main metal damage degree computation evaluation function 400b to estimate the wear degree of the main metal 1110. The logistic regression model is a model that learns the contribution degree (partial regression coefficient β) of wear to the main metal 1110. The contribution degree (partial regression coefficient β) is a value based on the cumulative fuel consumption amount and the cumulative value of damage that becomes a cause of failure.

[0100] When generating the logistic regression model, for example, an investigation report prepared by a worker when overhauling (disassembling) the engine 111 is used. The investigation report includes, for example, the cumulative fuel consumption of the engine 111, the cumulative damage value, and the wear level (e.g., the number of exposed copper layers: a seven-level value obtained by dividing 0 to 1 into seven). In generating the logistic regression model, the explanatory variables are, for example, the cumulative fuel consumption of the engine 111 and the cumulative damage value. Furthermore, the objective variable is, for example, the wear level (e.g., the number of exposed copper layers: a seven-level value obtained by dividing 0 to 1 into seven). Using the values ​​of the variables described in the investigation report, partial regression coefficients β (β1, β2, β3, β4) are learned by a maximum likelihood method or the like. In Equation (1) for the main metal damage level calculation evaluation function 400b, x1 corresponds to the cumulative fuel consumption. x2 corresponds to the cumulative value related to the engine rotation speed and the cumulative value related to the boost pressure. x3 corresponds to the cumulative value related to the blow-by pressure. x4 corresponds to the cumulative value for engine oil pressure.

[0101] 3, a functional configuration of the damage degree estimation server 100 related to estimation of the wear degree of the main metal 1110 will be described. The first acquisition unit 301a acquires the cumulative fuel consumption amount of the engine 111 from the trend data DB 101.

[0102] The second acquisition unit 301b acquires the cumulative value of damage that is a cause of a failure of the engine 111. Specifically, the second acquisition unit 301b acquires the cumulative value of damage from trend data stored in the trend data DB 101. The cumulative value of damage is, for example, as follows: -Cumulative value related to engine speed -Cumulative value related to boost pressure -Cumulative value related to blow-by pressure -Cumulative value related to engine oil pressure (normal, low idle, high idle). Note that the cumulative value of damage includes the above four cumulative values, but it is sufficient to include at least one cumulative value.

[0103] The estimation unit 302 (first estimation unit 302a) estimates the degree of wear of the main metal 1110 based on the cumulative fuel consumption acquired by the first acquisition unit 301a, the cumulative values ​​acquired by the second acquisition unit 301b, and the main metal damage degree calculation evaluation function 400b (logistic regression model).

[0104] The first estimation unit 302a estimates a normalized damage level (a value between 0 and 1) by inputting the cumulative fuel consumption amount of the engine 111 and each cumulative value obtained from the trend data into a trained logistic regression model. For example, if the damage level (a value between 0 and 1) is equal to or greater than a predetermined threshold, the first estimation unit 302a estimates the rating as "C," and if it is less than the predetermined threshold, the first estimation unit 302a estimates the rating as "A" or "B."

[0105] The contribution of an event related to the explanatory variable to the damage can be calculated by multiplying the value x of the explanatory variable by the partial regression coefficient β. Therefore, the first estimation unit 302a can estimate which event is causing the damage by calculating and comparing the contribution of each explanatory variable.

[0106] <<Example of Calculation Results Related to Damage Degree of Main Metal 1110>> Fig. 14 is a diagram showing an example of a calculation result 1400 related to the damage degree of the main metal 1110. In the calculation result 1400 shown in Fig. 14, the horizontal axis indicates the total operation time, and the vertical axis indicates the damage degree pm of the main metal 1110. When the damage degree pm of the main metal 1110 is equal to or greater than a threshold value 1410 (e.g., 0.9), this indicates that the rating is "C." Specifically, when the damage degree pm is equal to or greater than a first threshold value 1411, this indicates, for example, that a large number of exposed copper layers of the main metal 1110 are present. When the rating is "C," the damage degree estimation server 100 outputs a value of "1" (a value indicating a rating of "C") indicating that the main metal 1110 should be replaced as the estimated result of the wear degree of the main metal 1110.

[0107] Furthermore, the damage level pm being less than the first threshold value 1411 indicates that the rating is "A" or "B." Specifically, the damage level pm being less than the first threshold value 1411 indicates, for example, that the number of exposed copper layers of the main metal 1110 is small.

[0108] 14 is not limited to 0.9 and may be set to another value. Similarly, the second threshold 1412 is not limited to 0.8 and may be set to another value.

[0109] In the illustration, each line 1401 (1401a, 1401b, 1401c) indicates the total operating time and the degree of damage (degree of wear) of the three work machines 110, respectively. In the illustration, line 1401a is estimated to exceed the first threshold value 1411 the earliest. That is, for the work machine 110 corresponding to line 1401a, it is estimated that the number of exposed copper layers of the main metal 1110 will reach six the earliest, for example. Also in the illustration, line 1401c is estimated to exceed the first threshold value 1411 the latest. That is, for the work machine 110 corresponding to line 1401c, it is estimated that the number of exposed copper layers of the main metal 1110 will reach six the latest, for example.

[0110] 13, the factors shown as "measurement targets" are those that have a relatively large influence on the estimation of the degree of damage to the main metal 1110. In addition to these factors, other factors (such as exhaust temperature, engine oil temperature, and coolant temperature) may be included as long as they have a relatively large influence.

[0111] <<Third Action and Effect>> As described above, the damage degree estimation server 100 according to this embodiment estimates the wear state of the main metal 1110 based on the cumulative fuel consumption, the cumulative value of damage that is a cause of failure of the engine 111, and a predetermined evaluation function. This makes it possible to accurately estimate the wear state of the main metal 1110. Therefore, the estimation accuracy of the total damage degree 410 can be improved, allowing efficient maintenance of each work machine.

[0112] In this embodiment, the cumulative damage value includes at least one of a cumulative value related to the rotation speed of the engine 111, a cumulative value related to the boost pressure, a cumulative value related to the blow-by pressure, and a cumulative value related to the engine oil pressure. This makes it possible to easily accumulate the causes of wear damage to the main metal 1110.

[0113] Furthermore, in this embodiment, the damage degree estimation server 100 estimates the damage degree of the main metal 1110 as the wear state of the main metal 1110 based on the cumulative fuel consumption amount, the cumulative value of damage, and a predetermined evaluation function 400. This allows the wear state of the main metal 1110 to be estimated with high accuracy.

[0114] In this embodiment, the predetermined evaluation function 400 is a logistic regression model that has learned a partial regression coefficient β that indicates the contribution of damage to the main metal 1110. The damage degree estimation server 100 inputs the cumulative fuel consumption amount and the cumulative value of damage that is a cause of failure into the logistic regression model, thereby normalizing and estimating the wear state of the main metal 1110. This makes it possible to estimate the wear state of the main metal 1110 simply and accurately.

[0115] <<Modification for Estimating the Wear State of the Main Metal 1110>> In the above description, an example has been described in which the number of exposed copper layers of the main metal 1110 is estimated as an estimate of the wear state of the main metal 1110. Below, an example will be described in which the amount of wear of the main metal 1110 is estimated as an estimate of the wear state of the main metal 1110.

[0116] Of the various parts of the main metal 1110, if the amount of wear of the main metal 1110 exceeds a limit value, it is estimated that the damage to the engine 111 is significant. In other words, an overhaul is required. On the other hand, if the amount of wear of the main metal 1110 is less than the limit value, it is estimated that the damage to the engine 111 is not significant. In other words, an overhaul is not required. The parts where the amount of wear is measured are one or more specific points on the main metal 1110. In this embodiment, two points are used. If two or more points are used, the amount of wear is taken as the average value of the points. The amount of wear may be measured at three or more points.

[0117] The degree of damage to the main metal 1110 depends on the cumulative fuel consumption amount and, similarly to the above-described embodiment, also depends on the cumulative value of wear-increasing factors (for example, the engine operating at a predetermined rotation speed or higher, or the boost pressure operating at a predetermined value or lower). Note that in a modified example, the wear-increasing factors may be factors different from those in the embodiment.

[0118] In a modified example, when generating the logistic regression model, for example, an investigation report prepared by a worker when performing maintenance (disassembly) on the engine 111 is used. The investigation report includes, for example, the cumulative fuel consumption of the engine 111, the cumulative value of damage, and the "amount of wear." In generating the logistic regression model, the explanatory variables are, for example, the cumulative fuel consumption of the engine 111 and the cumulative value of damage. The objective variable is, for example, the "amount of wear." The partial regression coefficients β (β1, β2, β3, β4) are learned by a maximum likelihood method or the like using the values ​​of the variables described in the investigation report.

[0119] The first estimation unit 302a estimates the damage level (a value between 0 and 1) by inputting the cumulative fuel consumption amount of the engine 111 and each cumulative value obtained from the trend data into a trained logistic regression model. For example, if the damage level (a value between 0 and 1) is equal to or greater than a predetermined threshold, the first estimation unit 302a estimates the rating as "C," and if it is less than the predetermined threshold, the first estimation unit 302a estimates the rating as "A" or "B."

[0120] For example, if the amount of wear of the main metal 1110 is equal to or greater than a threshold value, this indicates a rating of "C." Specifically, if the damage level is equal to or greater than a threshold value (e.g., 0.9), this indicates, for example, that the wear state of the main metal 1110 is equal to or greater than a limit value. If the rating is "C," the damage level estimation server 100 outputs a value of "1" (a value indicating a rating of "C") indicating that replacement is required as the calculation result of the main metal damage level calculation evaluation function 400b. Furthermore, if the damage level is less than the threshold value, this indicates, for example, that the wear state of the main metal 1110 has not reached a limit value.

[0121] The damage degree estimation server 100 according to the modified example estimates the amount of wear of the main metal 1110 based on the cumulative fuel consumption amount and the cumulative damage value, as the wear state of the main metal 1110. This makes it possible to accurately estimate the wear state of the main metal 1110.

[0122] Other Embodiments One embodiment has been described in detail above with reference to the drawings, but the specific configuration is not limited to that described above, and various design modifications and the like are possible.

[0123] In the damage degree estimation server 100 and work machine 110 according to the embodiment described above, the programs are each stored in the storage 203, but this is not limiting. For example, the programs may be distributed to the damage degree estimation server 100 and work machine 110 via a communication line. In this case, the damage degree estimation server 100 and work machine 110 that receive the programs each load the programs into the main memory 202 and execute the above processing.

[0124] Furthermore, each of the programs may be designed to realize some of the above-described functions. For example, each of the programs may realize the above-described functions in combination with other programs stored in the storage 203 or in combination with other programs installed in other devices.

[0125] Furthermore, the damage degree estimation server 100 and the work machine 110 may each be equipped with a programmable logic device (PLD) in addition to or instead of the above configuration. Examples of PLDs include programmable array logic (PAL), generic array logic (GAL), complex programmable logic device (CPLD), and field programmable gate array (FPGA). In this case, some of the functions realized by the processor 201 may be realized by the PLD.

[0126] Furthermore, the damage degree estimation server 100 and the work machine 110 may each be equipped with multiple processors 201, or may be configured from multiple computers.

[0127] According to the above aspect, maintenance can be performed efficiently on each work machine.

[0128] 1...Damage degree estimation system, 100...Damage degree estimation server, 101...Trend data DB, 102...Estimation result DB, 110...Construction machine, 111...Engine, 120...User terminal, 301...Acquisition unit, 301a...First acquisition unit, 301b...Second acquisition unit, 302...Estimation unit, 302a...First estimation unit, 302b...Second estimation unit, 303...Output unit, 810...Core part, 1110...Main metal

Claims

1. An estimation device for estimating the degree of damage to an engine mounted on a work machine, comprising: an acquisition unit that acquires information on each part around the engine when it is in use; a first estimation unit that estimates the degree of damage to each of the parts based on the information on use acquired by the acquisition unit and an evaluation function corresponding to each of the parts; a second estimation unit that estimates the degree of damage to the engine based on the degree of damage to each of the parts estimated by the first estimation unit; and an output unit that outputs the degree of damage to the engine estimated by the second estimation unit.

2. The estimation device described in claim 1, wherein the first estimation unit normalizes and estimates the degree of damage to each of the parts, and the second estimation unit estimates the highest degree of damage among the degrees of damage to each of the parts normalized and estimated by the first estimation unit as the degree of damage to the engine.

3. The estimation device according to claim 1 or 2, wherein the output unit outputs the degree of damage to the engine estimated by the second estimation unit and a predetermined threshold value in a viewable manner.

4. An estimation method in which an estimation device for estimating the degree of damage to an engine mounted on a work machine executes processing including: an acquisition step of acquiring information on each part around the engine when it is in use; a first estimation step of estimating the degree of damage to each of the parts based on the information on use acquired in the acquisition step and an evaluation function corresponding to each of the parts; a second estimation step of estimating the degree of damage to the engine based on the degree of damage to each of the parts estimated in the first estimation step; and an output step of outputting the degree of damage to the engine estimated in the second estimation step.

5. A program that causes a computer of an estimation device that estimates the degree of damage to an engine mounted on a work machine to function as: an acquisition unit that acquires information on each part around the engine when it is in use; a first estimation unit that estimates the degree of damage to each of the parts based on the information on use acquired by the acquisition unit and an evaluation function corresponding to each of the parts; a second estimation unit that estimates the degree of damage to the engine based on the degree of damage to each of the parts estimated by the first estimation unit; and an output unit that outputs the degree of damage to the engine estimated by the second estimation unit.

6. An estimation device for estimating the degree of damage to an engine mounted on a work machine, comprising: a first acquisition unit that acquires the cumulative fuel consumption of the engine; a second acquisition unit that acquires an accumulated amount of errors that may cause a failure of the engine; and an estimation unit that estimates the degree of damage to parts that make up the engine based on the fuel consumption acquired by the first acquisition unit, the accumulated amount of errors acquired by the second acquisition unit, and a predetermined evaluation function.

7. The estimation device described in claim 6, wherein the second acquisition unit acquires an accumulated amount of errors related to wear of the engine, and the estimation unit estimates the degree of damage to the part based on the fuel consumption, the accumulated amount of errors related to wear, and the predetermined evaluation function.

8. The estimation device described in claim 6 or 7, wherein the second acquisition unit acquires an accumulated amount of heat-related errors of the engine, and the estimation unit estimates the degree of damage to the part based on the fuel consumption, the accumulated amount of heat-related errors, and the predetermined evaluation function.

9. The estimation device according to claim 7, wherein the accumulated amount of wear-related errors includes at least one of an accumulated amount related to the rotation speed of the engine, an accumulated amount related to engine oil pressure, and an accumulated amount related to blow-by pressure.

10. The estimation device according to claim 8, wherein the accumulated amount of heat-related errors includes at least one of an accumulated amount related to the engine coolant temperature and an accumulated amount related to the exhaust temperature.

11. The estimation device described in claim 6 or 7, wherein the predetermined evaluation function is a logistic regression model that has learned multiple regression coefficients that indicate the contribution of damage to the part, and the estimation unit normalizes and estimates the damage level of the part by inputting the fuel consumption amount acquired by the first acquisition unit and the accumulated amount of error detected by the second acquisition unit into the logistic regression model.

12. An estimation method in which an estimation device that estimates the degree of damage to an engine mounted on a work machine executes processing including: a first acquisition step of acquiring the cumulative fuel consumption of the engine; a second acquisition step of acquiring an accumulated amount of errors that may be a cause of engine failure; and an estimation step of estimating the degree of damage to parts that make up the engine based on the fuel consumption acquired in the first acquisition step, the accumulated amount of errors acquired in the second acquisition step, and a predetermined evaluation function.

13. A program that causes a computer of an estimation device that estimates the degree of damage to an engine mounted on a work machine to function as: a first acquisition unit that acquires the cumulative fuel consumption of the engine; a second acquisition unit that acquires the accumulated amount of errors that may cause the engine to fail; and an estimation unit that estimates the degree of damage to parts that make up the engine based on the fuel consumption acquired by the first acquisition unit, the accumulated amount of errors acquired by the second acquisition unit, and a predetermined evaluation function.

14. An estimation device for estimating the degree of damage to an engine mounted on a work machine, comprising: a first acquisition unit that acquires the cumulative fuel consumption of the engine; a second acquisition unit that acquires the cumulative value of damage that is a cause of engine failure; and an estimation unit that estimates the wear state of a main metal that constitutes the engine based on the fuel consumption acquired by the first acquisition unit, the cumulative value of damage acquired by the second acquisition unit, and a predetermined evaluation function.

15. The estimation device according to claim 14, wherein the cumulative damage value includes at least one of a cumulative value related to the engine rotation speed, a cumulative value related to boost pressure, a cumulative value related to blow-by pressure, and a cumulative value related to engine oil pressure.

16. An estimation device according to claim 14 or 15, wherein the estimation unit estimates the number of exposed copper layers of the main metal as the wear state of the main metal based on the fuel consumption amount, the cumulative value of the damage, and the predetermined evaluation function.

17. An estimation device according to claim 14 or 15, wherein the estimation unit estimates the amount of wear of the main metal as the wear state of the main metal based on the amount of fuel consumed, the cumulative value of the damage, and the predetermined evaluation function.

18. An estimation device as described in claim 14 or 15, wherein the predetermined evaluation function is a logistic regression model that has learned multiple regression coefficients that indicate the contribution of damage to the main metal, and the estimation unit estimates the degree of damage to the main metal by inputting the fuel consumption acquired by the first acquisition unit and the cumulative value of the damage detected by the second acquisition unit into the logistic regression model.

19. An estimation method in which an estimation device for estimating the degree of damage to an engine mounted on a work machine executes processing including: a first acquisition step of acquiring the cumulative fuel consumption of the engine; a second acquisition step of acquiring a cumulative value of damage that may be a cause of engine failure; and an estimation step of estimating the wear state of a main metal that constitutes the engine based on the fuel consumption acquired in the first acquisition step, the cumulative value of damage acquired in the second acquisition step, and a predetermined evaluation function.

20. A program that causes a computer of an estimation device that estimates the degree of damage to an engine mounted on a work machine to function as: a first acquisition unit that acquires the cumulative fuel consumption of the engine; a second acquisition unit that acquires the cumulative value of damage that is a cause of engine failure; and an estimation unit that estimates the wear state of the main metal that constitutes the engine based on the fuel consumption acquired by the first acquisition unit, the cumulative value of damage acquired by the second acquisition unit, and a predetermined evaluation function.

Images