Information processing device

Abstract

We provide technology that proactively alerts you to oil leaks from the breather. [Solution] The information processing device according to the present disclosure is an information processing device comprising a control unit and a storage unit that stores a predictive model trained by machine learning, wherein the control unit is configured to acquire first time-series data of the rotational speed of a gear immersed in oil and second time-series data of the oil temperature from a sensor, acquire observed atmospheric pressure, acquire the degree of oil deterioration, input the acquired first time-series data, second time-series data, observed atmospheric pressure, and degree of oil deterioration into the predictive model, and execute calculation processing of the predictive model to acquire a result from the predictive model that predicts signs of oil leakage from the breather, and output a notification to the user if the acquired prediction result indicates signs of oil leakage.

Description

Technical Field

[0001] This disclosure relates to an information processing apparatus.

Background Art

[0002] Patent Document 1 proposes a control device that suppresses oil leakage from a breather even when the defoaming performance deteriorates due to the amount of oil deterioration.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] One object of this disclosure is to provide a technique for notifying oil leakage from a breather in advance.

Means for Solving the Problems

[0005] The information processing apparatus according to this disclosure is an information processing apparatus including a control unit and a storage unit that stores a prediction model trained by machine learning. The control unit acquires first time-series data of the rotation speed of a gear immersed in oil and second time-series data of the oil temperature from a sensor, acquires an observed value of atmospheric pressure, acquires the degree of oil deterioration, inputs the acquired first time-series data, the second time-series data, the observed value of atmospheric pressure, and the degree of oil deterioration into the prediction model, executes arithmetic processing of the prediction model, acquires a result of predicting an oil leakage sign from the breather from the prediction model, and outputs a notification to the user when the acquired prediction result indicates a sign of oil leakage.

Effects of the Invention

[0006] According to this disclosure, it is possible to proactively notify the customer of oil leaks from the breather. [Brief explanation of the drawing]

[0007] [Figure 1] Figure 1 schematically illustrates an example of a scenario in which this disclosure applies. [Figure 2] Figure 2 schematically shows an example of the hardware configuration of the estimation device of this disclosure. [Figure 3] Figure 3 shows an example of the processing procedure of the estimation apparatus of this disclosure. [Modes for carrying out the invention]

[0008] Furthermore, as another form of the information processing device relating to the above embodiment, one aspect of this disclosure may be an information processing method that implements all or part of the above components, a program, or a machine-readable storage medium that stores such a program. A machine-readable storage medium is a medium that stores information such as programs by electrical, magnetic, optical, mechanical, or chemical action.

[0009] [1. Application Examples] Figure 1 schematically shows an example of a scenario to which this disclosure applies. The estimation device 1 according to this embodiment is an example of an information processing device. The estimation device 1 is configured to estimate whether or not there are signs of oil leakage from the breather 23 provided in the housing 21. When estimating, the estimation device 1 uses a first time system Input data 200 is obtained, which includes column data 210, second time series data 220, measured atmospheric pressure 230, and oil degradation degree 240. Estimation device 1 inputs the input data 200 into prediction model 100 and estimates whether or not there are signs of oil leakage. If the estimation result indicates signs of oil leakage from breather 23, estimation device 1 notifies the user.

[0010] (vehicle) The vehicle may be of any type (number of wheels, power source 20, size, etc.) as long as it contains at least an estimation device 1, a sensor S, a power source 20, a housing 21, gears 22, a breather 23, and oil 24. The vehicle may be selected from, for example, two-wheeled vehicles, three-wheeled vehicles, four-wheeled vehicles, etc. The power source 20 may be selected from, for example, electricity, fuel, etc. If the vehicle is an automobile, the size of the vehicle may be selected from large, medium, semi-medium, regular, large special, small special, etc. If the vehicle is a two-wheeled vehicle, the size of the vehicle may be selected from large, regular, etc. The user may be any person involved with the vehicle. Typically, the user may be the driver.

[0011] The power source 20 transmits power to the drive wheels 25 via one or more gears, including a gear 22, and a shaft. The power source 20 may be located inside or outside the housing 21. The housing 21 contains the gear 22 and oil 24. The gear 22 is immersed in the oil 24 for lubrication and to prevent wear. The housing 21 also includes a breather 23. The breather 23 is a device that regulates the pressure inside the housing 21. The breather 23 is usually located on top of the housing 21. The sensors S may include, for example, an oil temperature sensor OS, a rotation speed sensor RS, and other sensors.

[0012] (Oil leak from the breather) Oil 24 may leak from the breather 23 to the outside of the housing 21 for various reasons. When oil 24 leaks, problems such as insufficient lubrication due to a decrease in oil volume and contamination of parts may occur. Therefore, it is desirable for the user to take measures such as changing the oil before oil 24 leaks. In order for the user to know whether or not there are signs of oil leakage, the estimation device 1 may output the results of the prediction model 100 that predicts whether or not there are signs of oil leakage.

[0013] (Predictive model) The prediction model 100 is configured to predict signs of an oil leak from given input data 200. The signs of an oil leak may be information indicating whether or not an oil leak will occur within a predetermined time. The predetermined time may be defined as appropriate. The configuration of the prediction model 100 is not particularly limited, as long as such prediction processing can be performed, and may be determined as appropriate depending on the embodiment. In one example, the prediction model 100 may consist of at least one of a rule-based model and a trained machine learning model (trained model).

[0014] A rule-based model is configured to derive an inference result (in this embodiment, an estimate of oil leaks) from a given input according to rules. The rules may be set as appropriate. A machine learning model is configured to have one or more computational parameters that can be adjusted by machine learning. One or more computational parameters are used for the calculation of the desired inference (in this embodiment, the estimation of oil leaks). The machine learning model may consist of, for example, a neural network, a regression model, a decision tree model, a support vector machine, or other functional equations (computational models). The machine learning method may be selected as appropriate depending on the machine learning model adopted (e.g., backpropagation).

[0015] Machine learning involves adjusting (optimizing) the values ​​of computational parameters using training samples. Typically, supervised learning uses multiple training datasets, each consisting of a combination of input samples (training samples) and output samples (teacher signals, labels). A trained model may be generated by performing training. For example, the input sample may be a sample of input data 200 (first time series data 210, second time series data 220, atmospheric pressure 230, oil degradation degree 240, etc.), and the output sample may be a sample of output data (whether or not an oil leak occurred within a predetermined time). In supervised learning, the values ​​of the computational parameters of the machine learning model may be adjusted so that the output obtained from the machine learning model fits the corresponding output sample when an input sample is provided. However, the method for generating a trained model is not limited to this example and may be modified as appropriate depending on the embodiment.

[0016] In one example, the prediction model 100 may include a neural network. The structure of the neural network is not particularly limited and may be determined as appropriate depending on the embodiment. The structure of the neural network may be specified, for example, by the number of layers from the input layer to the output layer, the type of each layer, the number of nodes (neurons) included in each layer, and the connection relationships between the nodes in each layer. In one example, the neural network may include any mechanism such as a recursive structure, a self-attention mechanism, or an autoregressive model. The neural network may also include any layer such as a fully connected layer, a convolutional layer, a pooling layer, an inverse convolutional layer, an unpooling layer, a normalization layer, a dropout layer, or an LSTM (Long short-term memory). The neural network may include any type of model such as a diffusion model, a Transformer model, or a generative model. The weights of the connections between each node included in the neural network and the thresholds of each node are examples of computational parameters.

[0017] (Input data) The input data 200 may include various data that may influence the events causing oil leakage from the breather 23. For example, events that may cause oil leakage include a large amount of oil 24 being stirred up, a large amount of oil 24 being agitated, a decrease in the viscosity of oil 24, vaporization of oil 24, deterioration of oil 24, and a large pressure difference between the inside and outside of the housing 21.

[0018] In one example, the input data 200 may include first time-series data 210 of the rotational speed of the gear 22 immersed in the oil 24. The first time-series data 210 is defined as data indicating the change in the rotational speed over a predetermined time interval. The first time-series data 210 may be a set of data indicating time and the rotational speed at that time at a plurality of time points within a predetermined time interval. The number and format of the data may not be particularly limited and may be defined as appropriate. The predetermined time interval may also be defined as appropriate. The rotational speed may be defined as the number of rotations of the gear per unit time. The rotational speed may be obtained from, for example, a rotational speed sensor RS. The longer the state where the rotational speed of the gear 22 continues to be high, the greater the amount of oil 24 scraped up to the upper part, and the higher the probability that the oil 24 leaks from the breather 23. Therefore, by including the first time-series data 210 in the input data 200, the prediction model 100 can improve the prediction accuracy of oil leakage due to a large amount of oil 24 being scraped up to the upper part.

[0019] In one example, the input data 200 may include second time-series data 220 of the temperature of the oil 24. The second time-series data 220 is defined as data indicating the change in the temperature of the oil 24 over a predetermined time interval. The second time-series data 220 may be a set of data indicating time and the temperature of the oil 24 at that time at a plurality of time points within a predetermined time interval. The number and format of the data may not be particularly limited and may be defined as appropriate. The predetermined time interval may also be defined as appropriate. The temperature may be obtained from, for example, an oil temperature sensor OS. When the temperature of the oil 24 rises, events such as, for example, the viscosity of the oil 24 decreasing or the oil 24 vaporizing may occur. The oil 24 with reduced viscosity or vaporized oil 24 is likely to leak through the gaps of the breather 23. Therefore, by including the second time-series data 220 in the input data 200, the prediction model 100 can improve the prediction accuracy of oil leakage due to the decrease in the viscosity of the oil 24 or the vaporization of the oil 24.

[0020] In one example, the input data 200 may include the observed value of the atmospheric pressure 230. The atmospheric pressure 230 may be obtained from the atmospheric pressure sensor of the vehicle or from an external server or the like. When the atmospheric pressure 230 is low, a pressure difference may occur between the inside and the outside of the housing 21. When a pressure difference occurs between the inside and the outside of the housing 21, the breather 23 is burdened and oil leakage is likely to occur. Therefore, by including the atmospheric pressure 230 in the input data 200, the prediction model 100 can improve the prediction accuracy of oil leakage due to a large pressure difference between the inside and the outside of the housing 21.

[0021] In one example, the input data 200 may include the oil degradation degree 240. The oil degradation degree 240 may be an index indicating the degree of degradation of the oil 24. The index indicating the degree of degradation may be defined as appropriate. For example, when the oil 24 degrades, the defoaming performance decreases, so the number of bubbles on the oil surface of the oil 24 increases. By increasing the number of bubbles on the oil surface, the height of the oil surface of the oil 24 increases. Therefore, the index indicating the degree of degradation may be the height of the oil surface of the oil 24. That is, the oil degradation degree 240 may be the height of the oil surface of the oil 24. The height of the oil surface of the oil 24 may be measured by, for example, an oil level sensor. When the oil 24 degrades, it is likely to leak from the gap of the breather 23 because the viscosity of the oil 24 decreases or the oil 24 vaporizes. Therefore, by including the oil degradation degree 240 in the input data 200, the prediction model 100 can improve the prediction accuracy of oil leakage due to the degradation of the oil 24.

[0022] (Oil Leakage Prediction Result and Output) The estimation device 1 may output any information regarding the prediction result of whether or not there are signs of oil leakage from the breather 23. For example, if the estimation device 1 determines that there are signs of oil leakage, it may output information indicating that fact. If the estimation device 1 determines that there are no signs of oil leakage, it may omit outputting any information. The content to be output and the output destination are not particularly limited and may be appropriately selected depending on the embodiment. For example, the output content may include recording to a log file, notifying the user, etc. Notifying the user may involve displaying on the display that there are signs of oil leakage.

[0023] [2 Example Configurations] Figure 2 schematically shows an example of the hardware configuration of the estimation device 1 of this disclosure. As shown in Figure 2, the estimation device 1 according to this embodiment is a computer in which a control unit 11, a storage unit 12, a communication interface 13, an external interface 14, and an output device 15 are electrically connected. The control unit 11 includes a CPU (Central Processing Unit), RAM (Random Access Memory), ROM (Read Only Memory), etc., and is configured to perform arbitrary information processing. The storage unit 12 may be composed of, for example, a hard disk drive, a solid-state drive, etc. In this embodiment, the storage unit 12 stores the program 81 and the model data 110. The program 81 is a program that causes the information processing device to execute the information processing according to this embodiment. The program 81 includes a series of instructions for the information processing. The model data 110 may consist of various data (such as calculation parameters) for executing the prediction model 100.

[0024] The communication interface 13 is configured to perform wired or wireless data communication over a network. The communication interface 13 may consist of, for example, a wired LAN (Local Area Network) module, a wireless LAN module, etc. In this embodiment, the estimation device 1 may use the communication interface 13 to perform data communication over a network with another computer (for example, an external server). The external interface 14 may be, for example, a USB (Universal Serial Bus) port, a dedicated port, a wireless communication port, etc., and is configured to connect to an external device by wired or wireless connection. In this embodiment, the estimation device 1 may be connected to the oil temperature sensor OS and the rotation speed sensor RS via the external interface 14. The output device 15 is a device for outputting, for example, a display, a speaker, etc.

[0025] Furthermore, regarding the specific hardware configuration of the information processing device, components can be omitted, replaced, and added as appropriate depending on the embodiment. For example, the control unit 11 may include multiple hardware processors. Hardware processors include microprocessors, FPGAs (field-programmable gate arrays), DSPs (digital signal processors), and GPs. It may be composed of a U (Graphics Processing Unit), an ASIC (Application Specific Integrated Circuit), etc.

[0026] [3 Examples of operation] Figure 3 shows an example of the processing procedure of the estimation device 1 of this disclosure. The following processing procedure is an example of an information processing method performed by a computer. Furthermore, the following processing procedure may be performed at any time. For example, the control unit 11 may perform the following series of processes at predetermined time intervals.

[0027] In step S101, the control unit 11 acquires input data 200 to be input to the prediction model 100. The input data 200 may include first time series data 210, second time series data 220, atmospheric pressure 230, and oil degradation degree 240. The first time series data 210 and second time series data 220 may be acquired from the rotation speed sensor RS and the oil temperature sensor OS. The atmospheric pressure 230 may be acquired from the atmospheric pressure sensor of the device itself, or from an external server or the like. The oil degradation degree 240 may be a measured value of the oil level height of the oil 24. The oil level height of the oil 24 may be acquired, for example, from an oil level sensor. Once the input data 200 is acquired, the control unit 11 proceeds to step S102.

[0028] In step S102, the control unit 11 performs calculation processing on the prediction model 100. At this time, the control unit 11 may acquire model data 110 from the storage unit 12. If the prediction model 100 is a machine learning model, the model data 110 may be calculation parameters that have been adjusted through training. The control unit 11 may set the values ​​of the acquired calculation parameters in the prediction model 100 and perform calculation processing.

[0029] In step S103, the control unit 11 may determine whether the execution result of the prediction model 100 indicates an oil leak. If there is an oil leak, the control unit 11 may proceed to step S104 and notify the user accordingly. If there is no oil leak in step S103, or if step S104 is completed, the control unit 11 terminates this processing procedure.

[0030] [Features] In this embodiment, the prediction model 100 takes first time-series data 210, second time-series data 220, atmospheric pressure 230, and oil degradation degree 240 as input data 200 and predicts oil leakage from the breather 23. This data can influence the occurrence of events such as the oil 24 being violently agitated, vaporized, and degraded, as well as an increase in the pressure difference between the inside and outside of the housing 21. Therefore, by using input data 200 that takes these events that can induce oil leakage into account, the prediction model 100 can predict the presence or absence of signs of oil leakage with high accuracy.

[0031] [4. Variant] While embodiments of this disclosure have been described in detail above, the above description is merely illustrative in all respects of this disclosure. Various improvements or modifications can be made without departing from the scope of this disclosure. Needless to say, this is possible. The processes and means described in this disclosure can be freely combined and implemented, as long as no technical inconsistencies arise.

[0032] In one example, the input data 200 may further include a third time series of vehicle acceleration data. The third time series data may be defined as data showing the change in vehicle acceleration over a predetermined time interval. The third time series data may be a set of data showing the time and acceleration at multiple points in time within the predetermined time interval. The number and format of the data are not particularly limited and may be defined as appropriate. The predetermined time interval may also be defined as appropriate. The acceleration may be obtained, for example, from an acceleration sensor or by calculating the derivative of the velocity. When acceleration is applied, the oil level of the oil 24 may tilt. In this case, the oil 24 is more likely to reach the breather 23, increasing the possibility of oil leakage. Therefore, by including the third time series data in the input data 200, the prediction model 100 can be expected to improve the accuracy of predicting oil leakage due to the tilting of the oil level of the oil 24. [Explanation of symbols]

[0033] 1. Estimation device, 11. Control unit, 12. Memory unit, 13. Communication interface, 14. External interface, 15. Output device, 81. Program, 100... Predictive model, 200... Input data, S-Sensor

Claims

[Claim 1] An information processing device comprising a control unit and a storage unit for storing a predictive model trained by machine learning, The control unit, To acquire first time-series data of the rotational speed of the gear immersed in oil and second time-series data of the oil temperature from a sensor. Obtaining atmospheric pressure observations, To obtain the degree of oil degradation, The acquired first time series data, the second time series data, the observed atmospheric pressure, and the degree of oil degradation are input into the prediction model, and the calculation process of the prediction model is executed to obtain the result of predicting signs of oil leakage from the breather from the prediction model, and If the obtained prediction results indicate signs of an oil leak, a notification will be issued to the user. Configured to perform, Information processing device.

Images