Oil leakage estimation device

Abstract

To provide an oil leakage estimation device capable of accurately predicting possibility of occurrence of oil leakage.SOLUTION: An oil leakage estimation device for predicting occurrence of oil leakage from a unit that is formed by joining a plurality of cases to each other and stores oil, the oil leakage estimation device including a calculation device that predicts a possibility of occurrence of oil leakage, wherein the calculation device includes a data-acquisition part that acquires parameters contributing to oil leakage, including a datum related to a temperature change of the unit and a datum related to an external force applied to the unit (step S1), an input-variable-generation part that generates an input-variable based on the parameters acquired by the data-acquisition part (step S2), and an output-variable-generation part that outputs an output-variable indicating the possibility of occurrence of oil leakage when the input-variable is inputted, based on a learned model in which a correspondence relationship between the input-variable and the possibility of oil leakage from the unit is determined in advance (step S4).SELECTED DRAWING: Figure 3

Description

[Technical Field]

[0001] The present invention relates to an oil leakage estimation device that predicts the occurrence of oil leakage from a case or the like that holds oil. [Background technology]

[0002] Patent Document 1 discloses an oil leak detection device for detecting oil leaks in an electric vehicle without using a pressure sensor. The device in Patent Document 1 includes a temperature sensor that is installed vertically at the top of a case that houses a rotating electrical machine and measures the ambient temperature, and an oil temperature sensor that is installed slightly lower than the lower limit line of the oil level in the oil pan when there is no oil leak and acquires the oil temperature. The device in Patent Document 1 compares the ambient temperature detected by the temperature sensor with the temperature detected by the oil temperature sensor, and determines that an oil leak has occurred if the difference between them is equal to or less than a predetermined value.

[0003] Patent Document 2 discloses a determination device capable of determining the possibility of a fuel leak in a fuel supply system for a ship. When both the engine combustion pressure and the engine exhaust gas temperature drop, the device of Patent Document 1 determines the possibility of a fuel leak based on a first state value and a second state value, which are engine parameters. The first state value is a value indicating factors (negative factors) that reduce the possibility of a fuel leak, and the second state value is a value indicating factors (positive factors) that increase the possibility of a fuel leak. As a result, the device of Patent Document 1 determines that the greater the first state value is, the lower the possibility of a fuel leak in the fuel supply system, and determines that the greater the second state value is, the higher the possibility of a fuel leak in the fuel supply system. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Patent Publication No. 2021-69166 [Patent Document 2] Patent Publication No. 2021-120551 Summary of the Invention [Problem to be solved by the invention]

[0005] According to the device of Patent Document 1, when the difference between the detected value of the oil temperature sensor and the detected value of the temperature sensor falls below a predetermined value, it is determined that the oil level has dropped below the lower limit line and the oil temperature sensor is exposed, thereby detecting the occurrence of an oil leak. However, although the device of Patent Document 1 can detect the occurrence of an oil leak, it only detects the occurrence of an oil leak after it actually occurs, so there is a possibility that it is not possible to prevent the oil leak before it actually occurs. Similarly, the device of Patent Document 2 is a device that detects the actual occurrence of a fuel leak, so there is a possibility that it is not possible to take any measures before a fuel leak occurs.

[0006] This invention has been made with an eye on the above-mentioned technical problems, and aims to provide an oil leakage estimation device that can accurately predict the possibility of an oil leakage occurring. [Means for solving the problem]

[0007] In order to achieve the above-mentioned object, the present invention provides an oil leakage estimation device that predicts the occurrence of an oil leakage from a unit formed by joining multiple cases and containing oil, and is equipped with a calculation device that predicts the possibility of the oil leakage from the unit, and the calculation device is characterized by comprising: a data acquisition unit that acquires parameters that contribute to the oil leakage, including data regarding temperature changes of the unit and data regarding external forces applied to the unit; an input variable generation unit that generates input variables based on the parameters acquired by the data acquisition unit; and an output variable generation unit that outputs output variables that indicate the possibility of the oil leakage when the input variables are input, based on a trained model that predetermines the correspondence between the input variables and the possibility of the oil leakage from the unit. [Effects of the Invention]

[0008] The oil leak estimation device of the present invention predicts the occurrence of an oil leak at the mating surfaces of multiple cases that make up a unit. The oil leak estimation device acquires parameters that contribute to the oil leak, such as temperature changes in the unit and external forces on the unit. Based on these parameters, the device estimates the possibility of an oil leak in the unit through machine learning. The oil leak estimation device stores a model that has learned the correspondence between input parameters and oil leaks in advance. Therefore, when parameters are input into the oil leak estimation device, output variables corresponding to the parameters are output. In other words, by inputting input variables based on each parameter, the device can accurately predict the occurrence of an oil leak in the unit. As a result, if the possibility of an oil leak is high, the unit can be repaired or replaced in advance, thereby preventing the oil leak from occurring. [Brief explanation of the drawings]

[0009] [Figure 1]1 is an explanatory diagram illustrating the configuration of a unit that houses a power transmission device that is a target for which the occurrence of an oil leak is estimated by an oil leak estimation device according to an embodiment of the present invention; [Figure 2] 1 is a block diagram for explaining the functional configuration of an oil leakage estimation device according to an embodiment of the present invention. [Figure 3] 3 is a flowchart showing an example of control executed by the oil leakage estimation device according to the embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0010] Next, the present invention will be described based on the embodiments shown in the drawings. Note that the embodiments described below are merely examples of specific embodiments of the present invention, and are not intended to limit the present invention.

[0011] The oil leakage estimation device 1 in the embodiment of the present invention performs convolution operations and the like based on a large amount of information (target data) collected using machine learning, and outputs an output value corresponding to the input data. In other words, input variables (features) based on the input data are generated or extracted, and by inputting them into a mapping, the value of an output variable corresponding to the input variable is output. In other words, the oil leakage estimation device 1 in the embodiment of the present invention receives input of multiple parameters that contribute to oil leakage in a target component, and predicts or estimates whether or not there is a high possibility of oil leakage from that component based on the input parameters.

[0012] 1 predicts or estimates oil leakage from the mating surfaces between a first case 2a, a second case 2b, and a third case 2c that constitute a unit 2 that houses a power transmission device in a vehicle. The unit 2 may be the mating surface between a cylinder block and a cylinder head of an engine in a vehicle, or the mating surfaces between a cylinder block, a cylinder head, and a chain cover. The oil leakage estimation device 1 is not limited to units 2 that house such power transmission devices, and may also be used in parts of housings that contain oil and have relatively poor sealing properties.

[0013] Unit 2, which houses the power transmission device, is divided into three cases: first case 2a, second case 2b, and third case 2c, for reasons of assembly and other manufacturing processes. Unit 2 houses components of the power transmission device, such as a driving force source such as an engine, a planetary gear mechanism, bearings, a clutch, and shafts (none of which are shown). Unit 2 also contains oil (not shown) for lubricating and cooling these components. Gaskets 3 are provided on each mating surface between cases 2a, 2b, and 2c to prevent the oil from leaking outside unit 2 from the mating surfaces between cases 2a, 2b, and 2c.

[0014] The gasket 3 is formed along each mating surface of each case 2a, 2b, and 2c. In other words, the gasket 3 is provided around the entire circumference of each mating surface. The gasket 3 may be either a molded gasket that is sandwiched between the mating surfaces of each case 2a, 2b, and 2c and seals the mating surfaces of each case 2a, 2b, and 2c liquid-tightly by assembling the cases 2a, 2b, and 2c together, or a liquid gasket that is uniformly applied to the mating surfaces of each case 2a, 2b, and 2c and dries after the cases 2a, 2b, and 2c are assembled together to seal the mating surfaces of each case 2a, 2b, and 2c liquid-tightly. This prevents oil from leaking outside the unit 2.

[0015] The oil leakage estimation device 1 estimates that there is a possibility of oil leakage from the power transmission unit 2 configured as described above. The oil leakage estimation device 1 includes a calculation device 4 as a configuration for this purpose.

[0016] The arithmetic device 4 mainly comprises a processor, a communication unit, a storage unit, etc. The arithmetic device 4 is configured to perform calculations in accordance with a predetermined program using data acquired from the outside and pre-stored data, and to output the results of the calculations as control command signals. For example, the arithmetic device 4 executes functions that meet predetermined purposes by having the processor load a program stored in a recording medium into a working area of ​​the storage unit and execute the program, and perform various controls through the execution of the program.

[0017] The processor is, for example, a CPU or a DSP. This processor is configured to control the arithmetic device 4 and perform various information processing operations. The processor may also include a GPU (Graphics Processing Unit) capable of high-speed image processing. The storage unit includes, for example, a RAM and a ROM. As described above, the storage unit has a working area for the processor to execute programs. The storage unit also includes an auxiliary storage unit such as an EPROM or a hard disk drive. This auxiliary storage unit may also include a portable recording medium, i.e., a removable medium. The auxiliary storage unit freely stores various programs, various data, and various tables in the recording medium by reading or writing them. The auxiliary storage unit may also store an operating system. The communication unit is a wireless communication circuit connected to wireless communication to acquire various data and communicate with external communication devices. The arithmetic device 4 is configured to perform machine learning processes using a neural network, such as detection, calculation, and learning, using appropriate elements of the above-mentioned components.

[0018] Next, the functional configuration of the arithmetic device 4 in this embodiment of the present invention will be described with reference to Fig. 2. The arithmetic device 4 estimates or predicts the occurrence of an oil leak from the unit 2 based on the input data as described above. To achieve this, the arithmetic device 4 has a data acquisition unit 5, an input variable generation unit 6, and an output variable generation unit 7.

[0019] The data acquisition unit 5 acquires multiple parameters that contribute to oil leakage from various detectors, etc. The data acquisition unit 5 acquires, for example, time-series data on the temperature of the unit 2, parameters on the gasket 3, and parameters on the external force input to each unit 2. The data acquisition unit 5 acquires data that is the basis for outputting output variables for estimating oil leakage by mapping in machine learning.

[0020] The time-series data related to the temperature of unit 2 is, for example, data such as a map showing the temperature changes occurring over time in each of cases 2a, 2b, and 2c, and the resulting slight shape changes of each of cases 2a, 2b, and 2c over time. The parameters related to gasket 3 are, for example, data such as a map showing the temperature changes occurring over time in gasket 3, and the resulting loads applied to gasket 3 over time. The parameters related to external forces input to each unit 2 are, for example, the cylinder pressure of the internal combustion engine, the intake air volume of the internal combustion engine, the input torque or output torque of the power transmission device, the input rotation speed or output rotation speed of the power transmission device, and the acceleration of the vehicle. Note that if gasket 3 is a molded gasket, the material of the molded gasket may be included as a parameter, and if gasket 3 is a liquid gasket, the application process of the liquid gasket may be included as a parameter.

[0021] The input variable generation unit 6 generates variables based on the various data acquired by the data acquisition unit 5. The input variable generation unit 6 generates input variables by performing multiple processes on the various data and then extracting feature quantities. The input variable generation unit 6 first performs processing to make it easier to acquire and extract more data related to oil leaks from the acquired various data. For example, the input variable generation unit 6 performs data preprocessing on the acquired various data, such as filling in missing values, converting data formats, and normalizing. In other words, the input variable generation unit 6 processes the various data to output highly accurate estimation results.

[0022] The input variable generation unit 6 also extracts feature quantities from the processed data, which is various types of data after preprocessing. Feature quantities are values ​​that quantitatively represent qualitative characteristics related to oil leaks. Feature quantities are extracted by generating a feature map using a learning model. Such feature quantities are extracted, for example, by a convolutional neural network (CNN).

[0023] The output variable generation unit 7 generates output variables by performing predetermined processing on the input variables that have been input. The output variable generation unit 7 pre-stores mapping data for clarifying the specific correspondence between input and output. In other words, the mapping data is data that indicates the specific correspondence between input data and output data, and is used to clarify that relationship. The output variable generation unit 7 uses the mapping data to train a model as data that indicates the specific correspondence (correlation) between various data indicating the usage status of the unit 2, such as the possibility of an oil leak occurring and the period until the oil leak occurs. The output variable generation unit 7 uses a model that has learned the correspondence regarding the occurrence of an oil leak in this way to output an output variable that estimates or predicts an oil leak in the unit 2. The output variable output at this time is a value that indicates a symptom of an oil leak in the unit 2.

[0024] In the mapping data, for example, the unit 2 or each case 2a, 2b, 2c and gasket 3 are stored as correspondences, such as the rate of temperature rise, the amount of change (rate of change) in shape caused by the continuation of a high temperature state for a predetermined period of time or more, and the number of times this has been repeated, as well as the corresponding location and probability of oil leakage at the mating surface.

[0025] The time-series data for the temperature of unit 2 is a parameter obtained by observing, in advance through experiments, simulations, etc., the time-series temperature change of unit 2 in response to external factors such as outside temperature and the operating state of the drive source. For example, high outside temperature, engine heat due to high engine operating conditions, and heat generated by friction in gears, clutches, etc. can cause unit 2 as a whole to reach a relatively high temperature. Such a high temperature can cause expansion of unit 2 depending on the linear expansion coefficient of the metal material that makes up the power transmission device. Furthermore, the gasket 3 that seals the mating surfaces of each case 2a, 2b, and 2c can also reach a high temperature, which can cause a load.

[0026] Such repeated changes between high temperature and room temperature (low temperature) conditions result in repeated expansion and contraction of the cases 2a, 2b, and 2c and load on the gasket 3. Furthermore, due to aging and other factors, the possibility of oil leakage from the mating surfaces of the cases 2a, 2b, and 2c increases. In order to estimate or predict conditions that increase the possibility of such oil leakage, maps or the like obtained by conducting experiments or simulations over time are stored in the memory unit. In other words, correlations between temperature changes in the unit 2 caused by multiple factors and the increase in the possibility of oil leakage from the cases 2a, 2b, and 2c and the gasket 3 are stored in advance. Output variables are generated based on these correlations.

[0027] Similarly, the correspondence between the materials of the unit 2 and the gasket 3, or the above-mentioned external force on the unit, and oil leakage, is also stored. As a result, when an input variable is input to the output variable generation unit, an output variable (output value) that is associated with the input variable (input value) by mapping data that predefines the correspondence (mapping) is output.

[0028] For example, a fully connected forward propagation neural network is used for the process of outputting output variables based on input variables. That is, the input layer, intermediate layer (hidden layer), and output layer are configured to obtain a desired output result using weights optimized for the input variables. For example, the hidden layer includes a convolution layer, a normalization layer, and an activation function (activation layer), which perform processes such as linear transformation, convolution processing, and pooling processing.

[0029] Next, a description will be given of the processing executed by the oil leakage estimation device 1 in the embodiment of the present invention. As an example of this processing, a flowchart is shown in Figure 3, which shows the processing executed when parameters related to oil leakage are input to the calculation device 4 and a predetermined process is performed to obtain the possibility of oil leakage from the unit 2.

[0030] As shown in FIG. 3, first, in step S1, various data are input to the arithmetic unit 4. The various data input to the arithmetic unit 4 are parameters that contribute to oil leakage at the mating surfaces of the cases 2a, 2b, and 2c. For example, time-series data related to the temperature of the unit 2, parameters related to the gasket 3, and parameters related to external forces input to each unit 2 are acquired. Such parameters are acquired from sensors or detectors that detect temperature, pressure, torque, etc. provided in the power transmission device, a calculator that summarizes the correlations between detected values, and the manufacturing specifications of the unit 2. In this way, in step S1, parameters that contribute to oil leakage in the unit 2 are acquired from various sensors, etc.

[0031] In step S2, input variables are generated based on the various data acquired in step S1. In step S2, the various data are first preprocessed to unify the data format and improve the quality and quantity of the data. This improves the accuracy of estimating the occurrence of an oil leak. Then, features are extracted from the processed data, which is data that has been preprocessed in this way. In other words, the processed data is input into a neural network and features are extracted to generate input variables.

[0032] In step S3, the input variables generated in step S2 are input to the neural network. In step S3, the input variables are input to the neural network having a model that has undergone supervised learning. The model has previously learned mapping data that clarifies the correspondence between input and output. For example, if data regarding the temperature of unit 2 has a large contribution to oil leakage from unit 2, the weight of that data is increased. Alternatively, if data regarding the engine's internal cylinder pressure has a small contribution to oil leakage, the weight of that data is decreased. In this way, optimized weights for each input variable are preset in the neural network.

[0033] In step S4, an output variable is output. In step S4, an output variable indicating the possibility of an oil leak occurring in unit 2 is output. Parameters output as output variables include, for example, the probability of an oil leak occurring, the location where the oil leak will occur, the area where the oil leak will occur, and the remaining lifespan (remaining lifespan) which is the period until an oil leak occurs in unit 2. In this way, the processing in step S4 is completed by outputting an output variable related to the prediction of an oil leak occurring in unit 2. In other words, the possibility of an oil leak occurring in unit 2 is output.

[0034] As described above, the oil leakage estimation device 1 in this embodiment of the present invention estimates the possibility of an oil leakage in the unit 2 through machine learning based on parameters that contribute to an oil leakage. Furthermore, the calculation device 4 includes a supervised learning neural network that preliminarily determines the correlation between each parameter and an oil leakage. Therefore, by inputting input variables based on each parameter into the calculation device 4, it is possible to accurately predict the occurrence of an oil leakage in the unit 2. As a result, if there is a high possibility of an oil leakage in the unit 2, it becomes possible to repair or replace the unit 2 in advance, thereby preventing an oil leakage from occurring.

[0035] Furthermore, by using the material of unit 2 and the material or application process of gasket 3 as parameters when generating input variables, it is possible to take into account the rate at which durability decreases due to differences in material and application process, thereby enabling more accurate estimation of oil leakage from unit 2. Furthermore, external forces input to the unit include the cylinder pressure of the internal combustion engine, the intake air volume of the internal combustion engine, the input torque or output torque of the power transmission device, the input rotation speed or output rotation speed of the power transmission device, and the acceleration of the vehicle. This makes it possible to take into account the magnitude of the force that deforms each mating surface between each of cases 2a, 2b, and 2c, enabling more accurate estimation of oil leakage from unit 2. [Explanation of symbols]

[0036] 1 Oil leak estimation device 2 units 2a Case 1 2b Second Case 2c Case 3 3 gaskets 4 Arithmetic unit 5 Data Acquisition Section 6 Input variable generation section 7 Output variable generation section

Claims

[Claim 1] An oil leakage estimation device that predicts the occurrence of oil leakage from a unit that is formed by joining a plurality of cases together and that contains oil, a computing device for predicting the possibility of oil leakage from the unit; The computing device a data acquisition unit that acquires parameters that contribute to the oil leakage, including data related to temperature changes of the unit and data related to external forces acting on the unit; an input variable generation unit that generates input variables based on the parameters acquired by the data acquisition unit; and an output variable generation unit that outputs an output variable indicating the possibility of oil leakage occurring when the input variable is input, based on a trained model that defines a correspondence relationship between the input variable and the possibility of oil leakage from the unit in advance. An oil leakage estimation device characterized by:

Images