Feature prediction system, feature prediction device, feature prediction method, and program

Abstract

The present invention provides a feature prediction system, feature prediction device, feature prediction method, and program that can predict other unknown feature quantities when any two of the feature quantities of a rubber composition—viscoelasticity, surface roughness, and ice friction characteristics—are known. [Solution] The prediction system 100 comprises a prediction device 10, an information terminal 20, and a database 30. The control unit 11 of the prediction device 10 predicts a feature quantity representing another feature based on two known index values ​​representing two of the features of the rubber to be predicted, namely viscoelasticity, surface roughness, and ice friction characteristics. The control unit 21 of the information terminal 20 displays the predicted value along with the index values ​​used to predict the predicted value on the display unit 23.

Description

Technical Field

[0001] The present disclosure relates to a feature prediction system, a feature prediction device, a feature prediction method, and a program for predicting feature amounts of a rubber composition.

Background Art

[0002] Conventionally, an ice friction characteristic prediction method for predicting the ice friction characteristic (ice friction coefficient) of a rubber block on ice has been known (see Patent Document 1). This prediction method is based on the premise that ice melts due to frictional heat generated when the rubber block slides on ice, and includes the adhesion friction coefficient in the adhesion region directly contacting the ice, the lubrication friction coefficient in the lubrication region contacting the ice through a water film formed by melted water, the average value of the friction coefficients in each of the adhesion region and the lubrication region, and the ratio (adhesion rate) of the adhesion region to the contact area of the tire. An ice friction coefficient indicating the ice friction characteristic is calculated using a calculation formula represented by these factors.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] Conventional prediction methods require prior determination of the adhesive friction coefficient, the lubrication friction coefficient, the friction coefficient for each region, and the ratio in order to calculate the friction coefficient of a rubber block on ice. Therefore, for example, when calculating the friction coefficient of rubber blocks on ice that have different contact surface shapes and viscoelastic properties, it is necessary to prior determination of the adhesive friction coefficient, the lubrication friction coefficient, the friction coefficient for each region, and the ratio for each rubber block, which places a heavy burden on the calculation process. Furthermore, although the friction coefficient of a rubber block on ice is greatly influenced by the surface roughness of its contact surface and the viscoelasticity of the rubber block, conventional prediction methods do not take these factors into consideration, and therefore cannot predict the friction coefficient on ice with high accuracy.

[0005] Furthermore, conventional prediction methods cannot predict other unknown features when any two of the features of viscoelasticity, surface roughness, and ice friction coefficient are known.

[0006] The purpose of this disclosure is to provide a feature prediction system, feature prediction device, feature prediction method, and program that can predict other unknown feature quantities when any two of the feature quantities among the viscoelasticity, surface roughness, and ice friction characteristics of a rubber composition are known.

[0007] Another object of this disclosure is to provide a feature prediction system, feature prediction device, feature prediction method, and program that can predict the coefficient of friction on ice, which indicates the ice friction characteristics of a rubber composition, with light load and high accuracy. [Means for solving the problem]

[0008] A feature prediction system relating to one aspect of this disclosure includes: an input receiving unit that receives input of two index values ​​representing two of the following features: the viscoelasticity of a rubber composition, the surface roughness of the contact portion of the rubber composition that contacts an ice surface, and the ice friction characteristics of the rubber composition; a prediction unit that predicts a feature representing the other feature based on the two index values ​​received by the input receiving unit; and an output processing unit that outputs the predicted value predicted by the prediction unit to a predetermined output destination. [Effects of the Invention]

[0009] According to this disclosure, if any two of the features of the rubber composition—viscoelasticity, surface roughness, and ice friction characteristics—are known, it is possible to predict the other unknown features.

[0010] Furthermore, according to this disclosure, it is possible to predict the coefficient of friction on ice, which indicates the frictional properties of the rubber composition on ice, with light load and high accuracy. [Brief explanation of the drawing]

[0011] [Figure 1] Figure 1 shows the configuration of the feature prediction system according to the embodiment of this disclosure. [Figure 2] Figure 2 is a block diagram showing the configuration of the prediction device included in the feature prediction system. [Figure 3] Figure 3 is a block diagram showing the configuration of the information terminals included in the feature prediction system. [Figure 4] Figure 4 shows an example of a prediction item input screen displayed on the information terminal used by the user. [Figure 5] Figure 5 shows an example of an explanatory variable input screen displayed on an information terminal used by a user. [Figure 6] Figure 6 shows another example of an explanatory variable input screen displayed on an information terminal used by a user. [Figure 7] Figure 7 shows another example of an explanatory variable input screen displayed on an information terminal used by a user. [Figure 8]FIG. 8 is a diagram showing an example of a prediction result display screen displayed on an information terminal used by a user. [Figure 9] FIG. 9 is a diagram showing another example of a prediction result display screen displayed on an information terminal used by a user. [Figure 10] FIG. 10 is a flowchart showing an example of a procedure of prediction processing executed by a control unit of a prediction device.

Embodiments for Carrying Out the Invention

[0012] Hereinafter, embodiments of the present disclosure will be described with reference to the accompanying drawings. Note that the following embodiments are an example of embodying the present disclosure and do not limit the technical scope of the present disclosure.

[0013] [Configuration of Feature Quantity Prediction System 100] A feature quantity prediction system 100 (hereinafter simply referred to as prediction system 100) according to an embodiment of the present disclosure is a system that predicts a feature quantity indicating any one of the viscoelasticity of a rubber composition, the surface roughness of a contact portion where the rubber composition contacts an ice surface, and the ice friction characteristics of the rubber composition. As shown in FIG. 1, the prediction system 100 includes a prediction device 10 (an example of the feature quantity prediction device of the present disclosure), an information terminal 20, and a database 30. These are connected to each other via a network N1 so as to be communicable. The network N1 is, for example, a wired communication network connected by a LAN or the like, or a wireless communication network such as a dedicated line or a public line. Note that the prediction system 100 is an example of the feature quantity prediction system of the present disclosure. Further, the prediction device 10 is an example of the feature quantity prediction device of the present disclosure.

[0014] In this embodiment, a configuration in which the database 30 is connected to the network N1 is illustrated. However, for example, the database 30 may be provided in the prediction device 10 or the information terminal 20. Further, in this embodiment, a configuration in which the information terminal 20 is connected to the network N1 is illustrated. However, each component and various functions included in the information terminal 20 may be mounted on the prediction device 10.

[0015] The rubber composition is a rubber-like elastic body obtained by vulcanizing a polymer composition composed of a plurality of materials such as polymers and additives, and specifically, it is a rubber material used for tire products such as pneumatic tires mounted on vehicles such as automobiles. The prediction result (predicted value) by the prediction system 100 is used for the development of the tire product.

[0016] In this embodiment, as an example of the rubber composition, a rubber material used for the tire product is exemplified. However, for example, the rubber composition may be a rubber material used not only for the tire product but also for industrial rubber products such as vibration-proof rubber.

[0017] <​​​​​​​​​​​​​​​​Furthermore, if the prediction device 10 has known index values ​​for the ice friction characteristics and viscoelasticity of the rubber to be predicted, it predicts a feature quantity indicating the surface roughness of the rubber to be predicted based on these index values.

[0022] Furthermore, if the prediction device 10 has known index values ​​for the surface roughness and the ice friction coefficient of the rubber to be predicted, it predicts the characteristic quantity indicating the viscoelasticity of the rubber to be predicted based on these index values.

[0023] Here, the index values ​​for viscoelasticity are specifically the loss tangent tanδ and the complex shear modulus G. * , is the glass transition temperature Tg. Note that, for example, the storage shear modulus G′ and the loss shear modulus G″ may be used as indicators of viscoelasticity. The indicators of viscoelasticity can be measured using, for example, a viscoelasticity measuring device from the "Iplexer®" series manufactured by NETZSCH. Note that the indicators of viscoelasticity can also be measured using conventionally known measuring or analytical devices other than those exemplified above.

[0024] Specifically, the index value indicating the surface roughness is the arithmetic mean height Sa per unit area at the contact surface where the rubber to be predicted comes into contact with the ice surface. The arithmetic mean height Sa represents the average value of the height difference from the average surface and is a parameter defined by the International Organization for Standardization (ISO 25178). Alternatively, other index values ​​for surface roughness may be used, such as the double-root mean square height Sq, maximum height Sz, maximum peak height Sp, and maximum valley depth Sv, as defined by the International Organization for Standardization. Furthermore, the number of recesses and the number of protrusions within a unit area may also be used as index values ​​for surface roughness. These index values ​​can be obtained by analyzing magnified images of the contact surface of the rubber to be predicted based on the well-known atomic force microscopy (AFM) method.

[0025] The coefficient of friction on ice, which indicates the ice friction characteristics of the rubber subject to prediction, can be measured, for example, using a dynamic friction tester, a friction characteristic measuring device manufactured by Nippon Sangyo Co., Ltd. It should be noted that the coefficient of friction on ice can also be measured using conventionally known friction characteristic measuring devices other than the example device described above.

[0026] In the prediction device 10 of this embodiment, if any two of the following are known: the index value of viscoelasticity, the index value indicating surface roughness, and the index value of ice friction characteristics (coefficient of friction on ice), it is possible to predict the index value of other unknown characteristics based on those two index values.

[0027] The prediction device 10 is an information processing device capable of performing various calculations, and is, for example, a server computer, cloud server, or personal computer connected to network N1. The prediction device 10 is not limited to a single computer; it may be a computer system in which multiple computers work together, or a cloud computing system. Furthermore, the various processes performed by the prediction device 10 may be distributed and executed by one or more processors. The prediction device 10 has a program or computer software installed for operating the prediction system 100.

[0028] The information terminal 20 is an information processing device or terminal device used by the user. The information terminal 20 is a so-called desktop computer, laptop computer, or portable device such as a smartphone or tablet. The user inputs various information about the rubber to be predicted from the information terminal 20. The information terminal 20 also displays the prediction results transmitted from the prediction device 10 on its display screen. Therefore, the information terminal 20 has programs or computer software installed that cooperate with the prediction system 100 to transmit the various information to the prediction device 10 and to display the prediction results on its display screen.

[0029] Database 30 is a collection of data stored on a storage medium based on a predetermined data management method, containing various data handled by the prediction system 100. Database 30 can be implemented in various forms, such as a storage device, information processing device, cloud server, or data server, all of which are connected to the network N1 for communication. Database 30 includes training datasets for generating prediction models 123, 124, and 125 (see Figure 2) used in prediction processing by the prediction device 10. Database 30 stores training datasets corresponding to each of the prediction models 123, 124, and 125.

[0030] The aforementioned training dataset is information used to generate predictive models 123, 124, and 125 by machine learning. Specifically, the training dataset includes material information for each of the numerous sample rubbers Tk (k=1,2,···,n) consisting of numerous rubber compositions with different compositional structures, and an index value indicating the viscoelasticity, an index value indicating the surface roughness, and an ice friction coefficient, which is an index value indicating the ice friction characteristics, for each sample rubber Tk.

[0031] The machine learning method may be supervised learning, unsupervised learning, or reinforcement learning. Furthermore, deep learning, which learns to extract features themselves, may also be employed. When supervised learning is employed, the training dataset includes so-called training data, where input values ​​are associated with the correct output values ​​corresponding to those input values.

[0032] The aforementioned sample rubber Tk includes rubber products manufactured to date (such as tire products and industrial rubber products), prototype rubber products manufactured during research for the development of the aforementioned rubber products, or test pieces made from the aforementioned rubber composition manufactured during the aforementioned research. A dataset containing the material information and index values ​​indicating each of the aforementioned characteristics (viscoelasticity, surface roughness, and ice friction characteristics) for each of the numerous sample rubber Tk is stored in the database 30 as the learning dataset. Note that the aforementioned sample rubber Tk is an example of several other rubber compositions in this disclosure.

[0033] The index values ​​indicating the viscoelasticity of the sample rubber Tk are, specifically, the loss tangent tanδ and the complex shear modulus G. * The glass transition temperature is Tg. Note that, for example, the storage shear modulus G′ and the loss shear modulus G″ may be used as indicators of viscoelasticity.

[0034] The index value indicating the surface roughness of the sample rubber Tk is specifically the arithmetic mean height Sa per unit area at the contact surface where the sample rubber Tk contacts the ice surface. Alternatively, the index value for surface roughness may be the double-root mean square height Sq, maximum height Sz, maximum peak height Sp, maximum valley depth Sv, etc., as defined by the International Organization for Standardization. Furthermore, the index value for surface roughness may also be the number of recesses or protrusions within a unit area.

[0035] The material information includes, for example, material identification information indicating each material such as polymers and additives that constitute the sample rubber Tk, and blending information indicating the blending ratio of each material. Note that the material information does not necessarily have to be included in the training dataset, nor does it have to be used in the prediction process by the prediction models 123, 124, and 125.

[0036] [Prediction device 10] The configuration of the prediction device 10 will be described below with reference to Figure 2. Here, Figure 2 is a block diagram showing the configuration of the prediction device 10.

[0037] The prediction device 10 is for realizing the prediction system 100 of this embodiment, and as shown in Figure 2, it comprises a control unit 11, a storage unit 12, a communication unit 13, a display unit 14, and an operation unit 15.

[0038] The communication unit 13 is a communication interface that connects the prediction device 10 to the network N1 and performs data communication with each device connected to the network N1 according to a predetermined communication protocol. Specifically, the communication unit 13 performs data communication with the information terminal 20 and the database 30 through the network N1.

[0039] The storage unit 12 is a non-volatile storage medium such as an HDD, SSD, or flash memory that stores various types of information and data. The storage unit 12 stores the control program 121 and the prediction models 123, 124, and 125. Each of the prediction models 123, 124, and 125 may be implemented as an electronic circuit equipped with a memory that stores each prediction model 123, 124, and 125.

[0040] The control program 121 may be non-temporarily recorded on a computer-readable recording medium such as a CD or DVD, and may be read from the recording medium by a reading device (not shown), such as a CD drive or DVD drive, electrically connected to the prediction device 10, and copied to the storage unit 12. Alternatively, the control program 121 may be read from external storage connected to the network N1, input via the communication unit 13, and copied to the storage unit 12.

[0041] The control program 121 is a program that uses the prediction models 123, 124, and 125 to cause the control unit 11 to execute the prediction process described later (see Figure 10).

[0042] Each of the prediction models 123, 124, and 125 is a trained model used in the prediction process described later (see Figure 10). They predict feature quantities that represent specific characteristics of the rubber (rubber composition) that is the target of prediction, and output the predicted values ​​from the output unit. Note that prediction models 123, 124, and 125 may also include a well-known function (prediction function) that calculates predicted values ​​(output data) for input values ​​(input data).

[0043] Specifically, when the prediction model 123 receives index values ​​representing the viscoelasticity and surface roughness of the rubber to be predicted, it predicts the ice friction coefficient (a feature quantity of ice friction characteristics) representing the ice friction characteristics of the rubber to be predicted based on these index values ​​and outputs the predicted value. In this case, the input values ​​(explanatory variables) for the prediction model 123 are the index values ​​representing the viscoelasticity and surface roughness of the rubber to be predicted, and the output value (dependent variable) is the ice friction coefficient of the rubber to be predicted.

[0044] When the prediction model 124 receives index values ​​representing the ice friction characteristics and viscoelasticity of the rubber to be predicted, it predicts a feature quantity representing the surface roughness of the rubber to be predicted based on these index values ​​and outputs the predicted value. In this case, the input values ​​(explanatory variables) for the prediction model 124 are the index values ​​representing the ice friction characteristics and viscoelasticity of the rubber to be predicted, and the output value (dependent variable) is a feature quantity representing the surface roughness of the rubber to be predicted.

[0045] When the prediction model 124 receives index values ​​representing the surface roughness and ice friction coefficient of the rubber to be predicted, it predicts the viscoelasticity feature quantity of the rubber to be predicted based on these index values ​​and outputs the predicted value. In this case, the input values ​​(explanatory variables) for the prediction model 125 are the index values ​​representing the surface roughness and ice friction coefficient of the rubber to be predicted, respectively, and the output value (dependent variable) is the viscoelasticity feature quantity of the rubber to be predicted.

[0046] Each of the prediction models 123, 124, and 125 is trained and generated by machine learning performed by the control unit 11 based on the training dataset stored in the database 30 and a predetermined algorithm. Alternatively, each of the prediction models 123, 124, and 125 may be trained and generated by a control unit other than the control unit 11, then transferred from an external source and stored in the memory unit 12.

[0047] Each of the prediction models 123, 124, and 125 can be constructed using various methods. For example, suitable algorithms for constructing prediction models 123, 124, and 125 include multiple regression, generalized linear regression, principal component regression, ridge regression, lasso regression, kernel regression, random forest regression, Gaussian process regression, multilayer neural networks, clustering, support vector machines, and RBF networks defined by radial basis functions. Furthermore, each of the prediction models 123, 124, and 125 may be constructed using deep learning with multiple hidden layers in the neural network. Note that each of the prediction models 123, 124, and 125 may be machine-trained using one of the above algorithms, or may be machine-trained using multiple algorithms.

[0048] Predictive models 123, 124, and 125 can be constructed using commercially available computer software (for example, MATLAB® from The MathWorks, or modeFRONTIER from ESTECO).

[0049] The display unit 14 is a display device such as a liquid crystal display or an organic EL display that displays various types of information. The operation unit 15 is an input device such as a mouse, keyboard, or touch panel that accepts input from the operator.

[0050] The control unit 11 controls the operation of each part of the prediction device 10. The control unit 11 includes control devices such as a CPU, ROM, and RAM. The CPU is a processor that performs various arithmetic operations. The ROM is a non-volatile storage medium in which control programs for causing the CPU to perform various arithmetic operations are pre-stored. The RAM is a volatile or non-volatile storage medium that stores various information and is used as a temporary storage memory (work area) for the various arithmetic operations performed by the CPU. The control unit 11 controls the prediction device 10 by having the CPU execute various control programs pre-stored in the ROM or storage unit 12.

[0051] As shown in Figure 2, the control unit 11 includes various processing units such as a data acquisition processing unit 110, a prediction processing unit 111 (an example of a prediction unit in this disclosure), an output processing unit 112 (an example of an output processing unit in this disclosure), and a model learning unit 113. The control unit 11 functions as one of these various processing units when the CPU executes various calculation processes according to the control program. The control unit 11 or the CPU is an example of a computer or processor that executes the control program.

[0052] Furthermore, some or all of the processing units included in the control unit 11 may be composed of electronic circuits. Also, the control program may be a program that causes multiple processors to function as the various processing units.

[0053] The data acquisition processing unit 110 performs the process of acquiring input information to be used as input values ​​(explanatory variables) for the prediction models 123, 124, and 125 used in the prediction processing unit 111.

[0054] When the prediction processing (see Figure 10) for predicting the friction coefficient on ice using the prediction model 123 is performed by the prediction processing unit 111, the input information consists of index values ​​representing the viscoelasticity and surface roughness of the rubber to be predicted.

[0055] Furthermore, when the prediction processing unit 111 performs a prediction process (see Figure 10) to predict the surface roughness feature quantities using the prediction model 124, the input information consists of index values ​​representing the ice friction characteristics and viscoelasticity of the rubber subject to prediction.

[0056] Furthermore, when the prediction processing unit 111 performs a prediction process (see Figure 10) to predict the viscoelasticity using the prediction model 125, the input information consists of index values ​​representing the surface roughness and the ice friction coefficient of the rubber to be predicted, respectively.

[0057] The aforementioned input information is input by the user at the information terminal 20 when the prediction device 10 is to predict the predicted value of a feature quantity that represents a specific characteristic of the rubber being predicted. For example, when the input information input by the user at the information terminal 20 is transferred from the information terminal 20 to the prediction device 10 via the network N1, the data acquisition processing unit 110 acquires that input information.

[0058] The data acquisition processing unit 110 may, for example, access a storage device (external storage device or storage unit 22) indicated by address information input from the information terminal 20, read the input information stored in the storage device, and store it in the storage unit 12 to acquire the input information of the rubber to be predicted.

[0059] Furthermore, the input information regarding the rubber to be predicted may be directly input to the prediction device 10 via the operation unit 15 of the prediction device 10.

[0060] The prediction processing unit 111 performs a process to predict a feature quantity representing the other feature based on two known index values ​​representing two of the features of the rubber to be predicted, namely the viscoelasticity, surface roughness, and ice friction characteristics.

[0061] Specifically, when the data acquisition processing unit 110 acquires an index value indicating the viscoelasticity of the rubber to be predicted (an example of a first index value in this disclosure) and an index value indicating the surface roughness of the rubber to be predicted (an example of a second index value in this disclosure), the prediction processing unit 111 performs a prediction process (ice friction coefficient prediction process) to predict the ice friction coefficient indicating the ice friction characteristics of the rubber to be predicted based on these index values. More specifically, the prediction processing unit 111 inputs the acquired index values ​​indicating the viscoelasticity and surface roughness of the rubber to be predicted, respectively, into the input unit of the prediction model 123, causes the prediction model 123 to predict the ice friction coefficient of the rubber to be predicted, and outputs the predicted value (prediction result) of the ice friction coefficient from the output unit of the prediction model 123.

[0062] Furthermore, when the data acquisition processing unit 110 acquires the ice friction coefficient, which is an index value indicating the ice friction characteristics of the rubber to be predicted, and the index value indicating the viscoelasticity of the rubber to be predicted, the prediction processing unit 111 performs a prediction process (surface roughness prediction process) to predict a feature quantity indicating the surface roughness of the rubber to be predicted based on these index values. Specifically, the prediction processing unit 111 inputs the acquired index values ​​indicating the ice friction characteristics and viscoelasticity of the rubber to be predicted, respectively, into the input unit of the prediction model 124, causes the prediction model 124 to predict a feature quantity indicating the surface roughness of the rubber to be predicted, and outputs the predicted value (prediction result) of the feature quantity indicating the surface roughness from the output unit of the prediction model 124.

[0063] Furthermore, when the prediction processing unit 111 obtains an index value indicating the surface roughness of the rubber to be predicted, and the ice friction coefficient, which is an index value indicating the ice friction characteristics, the prediction processing unit 111 performs a prediction process (viscoelasticity prediction process) to predict the viscoelasticity feature quantity of the rubber to be predicted based on these index values. Specifically, the prediction processing unit 111 inputs the obtained index values ​​indicating the surface roughness and ice friction characteristics of the rubber to be predicted, respectively, into the input unit of the prediction model 125, causes the prediction model 125 to predict the viscoelasticity feature quantity of the rubber to be predicted, and outputs the predicted value (prediction result) of the viscoelasticity feature quantity from the output unit of the prediction model 125.

[0064] In this embodiment, when the information terminal 20 receives identification information (prediction target identification information) indicating the feature to be predicted (prediction feature), the prediction processing unit 111 selects a prediction model corresponding to the received prediction target identification information and uses the selected prediction model to predict the feature quantity of the prediction feature.

[0065] For example, when the information terminal 20 receives identification information indicating the ice friction characteristics as the target of prediction, the prediction processing unit 111 selects a prediction model 123 from a plurality of prediction models that corresponds to the received identification information, and uses that prediction model 123 to predict the ice friction characteristic coefficient of the rubber to be predicted.

[0066] Furthermore, when the information terminal 20 receives identification information indicating the surface roughness as the target of prediction, the prediction processing unit 111 selects a prediction model 124 from a plurality of prediction models that corresponds to the received identification information, and uses that prediction model 124 to predict the feature quantities of the surface roughness.

[0067] Furthermore, when the information terminal 20 receives identification information indicating viscoelasticity as the target of prediction, the prediction processing unit 111 selects a prediction model 124 from a plurality of prediction models that corresponds to the received identification information, and uses that prediction model 125 to predict the viscoelastic feature quantities.

[0068] The predicted values ​​predicted by the prediction processing unit 111 are transferred to the information terminal 20 via the communication unit 13 in order to be displayed on the display unit 23 of the information terminal 20.

[0069] The output processing unit 112 performs the process of outputting the predicted value (prediction result) of the prediction process performed in the prediction device 10 to a predetermined output destination. In this embodiment, one example of the output destination is the information terminal 20, and another example is the display unit 14 provided in the prediction device 10.

[0070] The output processing unit 112 outputs the input information used to predict the predicted value, along with the predicted value, to the information terminal 20.

[0071] In this embodiment, the output processing unit 112 outputs the predicted value and the input information used to predict the predicted value to the information terminal 20, causing the display unit 23 of the information terminal 20 to display the predicted value along with each index value included in the input information.

[0072] The model learning unit 113 constructs (generates) trained prediction models 123, 124, and 125 by training a pre-trained prediction model using machine learning based on the training dataset in the pre-prepared database 30 and the predetermined algorithm described above.

[0073] [Information Terminal 20] The configuration of the information terminal 20 will be described below with reference to Figure 3. Here, Figure 3 is a block diagram showing the configuration of the information terminal 20.

[0074] As shown in Figure 3, the information terminal 20 includes a control unit 21, a storage unit 22, a display unit 23, an operation unit 24, a communication unit 25, and the like. The information terminal 20 is an information processing device or terminal device used by the user. The user inputs various information about the rubber to be predicted, which is necessary for the prediction processing by the prediction device 10, from the information terminal 20.

[0075] The communication unit 25 is a communication interface for connecting the information terminal 20 to the network N1 and for performing data communication with the prediction device 10 via the network N1 in accordance with a predetermined communication protocol.

[0076] The memory unit 22 is a non-volatile storage medium such as flash memory that stores various types of information. The memory unit 22 stores control programs 221 that cause the control unit 21 to execute various calculation processes, as well as data used in various calculation processes.

[0077] The display unit 23 is a display device such as a liquid crystal display or an organic EL display that displays various types of information. The operation unit 24 is an input device such as a mouse, keyboard, or touch panel that accepts input operations from the user.

[0078] In this embodiment, the predicted values ​​predicted by the prediction processing unit 111 of the prediction device 10 and the input information used in the prediction processing of the prediction processing unit 111 are displayed on the display unit 23. Specifically, a list including the predicted values ​​and the input information is displayed on the prediction result display screen 232 (see Figure 8) displayed on the display unit 23.

[0079] The control unit 21 controls the operation of each part of the information terminal 20. The control unit 21 includes control devices such as a CPU, ROM, and RAM. The CPU is a processor that performs various arithmetic operations. The ROM is a non-volatile storage medium in which control programs for causing the CPU to perform various arithmetic operations are pre-stored. The RAM is a volatile or non-volatile storage medium that stores various information and is used as a temporary storage memory (work area) for the various arithmetic operations performed by the CPU. The control unit 21 controls the information terminal 20 by executing various control programs pre-stored in the ROM or storage unit 22 using the CPU.

[0080] As shown in Figure 3, the control unit 21 includes various processing units such as an input receiving unit 211 (an example of an input receiving unit in this disclosure), a display processing unit 212 (an example of a display processing unit in this disclosure), and an index value changing unit 213 (an example of an index value changing unit in this disclosure). The control unit 21 functions as the various processing units by having the CPU execute various calculation processes in accordance with the control program. The control unit 21 or the CPU is an example of a computer or processor that executes the control program.

[0081] Furthermore, some or all of the processing units included in the control unit 21 may be composed of electronic circuits. Also, the control program may be a program that causes multiple processors to function as the various processing units.

[0082] The input receiving unit 211 processes input of an index value indicating a predetermined characteristic of the rubber to be predicted. Specifically, when the feature to be predicted (hereinafter referred to as the predicted feature) is input from the predicted feature input screen 230 (see Figure 4), the input receiving unit 211 then displays an explanatory variable input screen 231 (see Figure 5) on the display unit 23 for inputting explanatory variables to be used in the prediction process of the input predicted feature. When a predetermined numerical value is input as the index value to the explanatory variable input screen 231 by the operation unit 24, the input receiving unit 211 acquires the input index value. The acquired index value is then transmitted to the prediction device 10 by the communication unit 25.

[0083] Figure 4 shows the prediction feature input screen 230 displayed on the display unit 23 of the information terminal 20. The prediction feature input screen 230 is a user interface for inputting the characteristics (predicted features) of the rubber to be predicted to be predicted by the prediction processing unit 111.

[0084] As shown in Figure 4, the prediction feature input screen 230 includes an input box 40 for inputting the prediction feature to be predicted. Any feature can be selected from the pull-down menu 40A of the input box 40. In Figure 4, the pull-down key of the input box 40 is pressed to open the pull-down menu 40A, and the state in which the ice friction characteristic is selected as the prediction feature in the pull-down menu 40A is shown. When the ice friction characteristic is selected in the pull-down menu 40A, the ice friction characteristic is input as the prediction feature in the input box 40. After the ice friction characteristic is input in the input box 40, if the next key 32 is pressed thereafter, the screen displayed on the display unit 23 transitions to the explanatory variable input screen 231 (see Figure 5).

[0085] Figure 5 shows the explanatory variable input screen 231 displayed on the display unit 23 of the information terminal 20. The explanatory variable input screen 231 is a user interface for inputting index values ​​of features other than the predicted features entered in the input frame 40 (see Figure 4). In Figure 4, if the ice friction characteristics are entered as the predicted features, the explanatory variable input screen 231 for inputting the index value of the surface roughness of the rubber to be predicted and the index value of the viscoelasticity of the rubber to be predicted is displayed on the display unit 23, as shown in Figure 5.

[0086] As shown in Figure 5, the explanatory variable input screen 231 includes an input box 41 for inputting the index value of the surface roughness and input boxes 42 to 44 for inputting the index value of the viscoelasticity. Input box 42 is an input box for inputting the value of the loss tangent tanδ, which is one of the properties of viscoelasticity. Input box 43 is for inputting the complex shear modulus G, which is one of the properties of viscoelasticity. * This is an input box for entering a numerical value. Input box 44 is an input box for entering a numerical value for the glass transition temperature Tg, which is one of the viscoelastic properties. The user can enter any numerical value into input boxes 41 to 44 by operating the control unit 24.

[0087] Furthermore, in the prediction feature input screen 230 of Figure 4, when the surface roughness is entered as the prediction feature into the input frame 40 and the next key 32 is pressed, an explanatory variable input screen 231A (see Figure 6) is displayed on the display unit 23, as shown in Figure 6. This screen includes an input frame 45 for entering the coefficient of friction on ice, which is an index value of the friction characteristics on ice, and input frames 42 to 44 for entering the index value of viscoelasticity.

[0088] Furthermore, in the prediction feature input screen 230 of Figure 4, when the viscoelastic modulus is entered as the prediction feature into the input frame 40 and the next key 32 is pressed, an explanatory variable input screen 231B (see Figure 7) is displayed on the display unit 23, as shown in Figure 7. This screen includes an input frame 45 for entering the coefficient of friction on ice, which is an index value of the friction characteristics on ice, and an input frame 41 for entering an index value of the surface roughness.

[0089] In the explanatory variable input screens 231, 231A, and 231B, after the index values ​​as explanatory variables are entered into each of the input fields 41 to 45, when the registration key 33 is pressed, the index values ​​entered into each input field are stored in the storage unit 22, and each entered index value and the identification information of the prediction feature to be predicted are transmitted to the prediction device 10 by the communication unit 25. Each index value transmitted to the prediction device 10 is used for prediction processing by the prediction processing unit 111.

[0090] When the display processing unit 212 receives the predicted value transmitted from the prediction device 10, it performs a process to display a prediction result display screen 232 (see Figure 8) on the display unit 23, which includes the predicted value and the index value used in the prediction process. As shown in Figure 8, the prediction result display screen 232 includes a display frame 232A for displaying the predicted value, a display frame 232B for displaying the index value used in the prediction process by the prediction processing unit 111, and a re-prediction key 34. The display processing unit 212 may also display identification information of the rubber to be predicted on the prediction result display screen 232. In Figure 8, the predicted value of the ice friction coefficient, which is a prediction feature, is "0.165", the index value of the surface roughness (arithmetic mean height Sa) is "1.25", the index value of the viscoelastic loss tangent tanδ is "0.28", and the complex shear modulus G of the viscoelastic material is also shown. * An example of the prediction result display screen 232 is shown, where the index value of is "2.6" and the index value of the viscoelastic glass transition temperature Tg is "-75".

[0091] The indicator value modification unit 213 modifies the numerical values ​​of each indicator value used in the prediction process by the prediction processing unit 111. For example, when the re-prediction key 34 provided on the prediction result display screen 232 is pressed, the display processing unit 212 modifies the indicator value input screen 233 (an example of a modified value input screen in this disclosure) on the display unit 23, as shown in Figure 9. Here, Figure 9 is a user interface for inputting the modified numerical values ​​of the indicator values ​​used in the prediction process by the prediction device 10. As shown in Figure 9, the indicator value input screen 233 displays a display frame 232A for displaying the predicted values, a display frame 232B for displaying the indicator values ​​used in the prediction process, and a modified value input frame 46.

[0092] The change value input field 46 is an input field for entering numerical values ​​to change the input values ​​of the index values ​​of each feature used as explanatory variables in the prediction process. The user can input the changed numerical values ​​(change values) of any feature of their choice into the change value input field 46 via the operation unit 15. The user may change all the index values ​​displayed in the display field 232B to any numerical value, or they may change one or more of the index values ​​to any numerical value.

[0093] When a numerical value is entered into the change value input box 46, and then the re-prediction execution key 35 provided on the change index value input screen 233 is pressed, the index value change unit 213 transmits the entered numerical value to the prediction device 10 in order to set the change value entered into the change value input box 46 as an explanatory variable to be used for re-prediction by the prediction processing unit 111, causing the control unit 11 of the prediction device 10 to change the explanatory variable to be used during re-prediction.

[0094] When the control unit 11 of the prediction device 10 receives the changed value, it sets the changed value as an explanatory variable to be used for the prediction processing again by the prediction processing unit 111, and inputs it into the input section of the prediction model used by the prediction processing unit 111. The prediction processing unit 111 executes the prediction processing again using the changed value and transmits the predicted value and each index value used for the prediction back to the information terminal 20. Note that for index values ​​of features that were not entered in the changed value input frame 46, the index values ​​are not changed, and the prediction processing is executed with the original input values. Then, the prediction result display screen 232 (see Figure 8), which includes the re-predicted predicted value and the index values ​​used for the re-prediction, is displayed on the display unit 23 by the display processing unit 212.

[0095] [Predictive processing] The feature prediction method of this disclosure will be described below, along with an example of the procedure for the prediction process performed in the prediction system 100, with reference to the flowchart in Figure 10.

[0096] Each step in the prediction process is executed by the control unit 11 of the prediction device 10 or the control unit 21 of the information terminal 20. One or more steps in the prediction process may be omitted as appropriate, and the execution order of each step may differ to the extent that similar effects are produced.

[0097] First, in step S11, the control unit 11 determines whether the prediction start conditions have been met. When the control unit 11 receives the identification information of the prediction features necessary for prediction processing (prediction target identification information) and the index values ​​of other features other than the prediction features from the information terminal 20, it determines that the prediction start conditions have been met.

[0098] In the next step S12, the control unit 11 selects a prediction model corresponding to the received prediction target identification information.

[0099] In the next step, S13, the control unit 11 performs a process to predict the feature quantities of the predicted features using the selected prediction model. Specifically, it inputs the index values ​​of other features sent from the information terminal 20 to the input unit of the selected prediction model and outputs the feature quantities of the predicted features from the output unit of the prediction model. Note that step S13 is an example of the prediction step of this disclosure.

[0100] In step S14, the control unit 11 outputs the predicted values ​​(prediction results) obtained by the prediction process to the information terminal 20 in order to display the predicted values ​​on the information terminal 20. When the control unit 21 of the information terminal 20 receives the predicted values ​​from the prediction device 10, it displays the predicted values ​​on the prediction result display screen 232. At this time, not only the predicted values ​​but also the index values ​​of other features used in the prediction process are displayed on the prediction result display screen 232. Step S14 is an example of an output step in this disclosure.

[0101] In the next step S15, the control unit 11 determines whether or not there has been a change in the indicator value used for the prediction process. Specifically, the control unit 21 of the information terminal 20 determines whether or not the re-prediction key 34 has been pressed. If the re-prediction key 34 is pressed, the control unit 21 displays the change value input box 46 on the prediction result display screen 232 and waits until the user enters the change value. Then, when the change value is entered on the prediction result display screen 232 and the re-prediction execution key 35 is pressed, the entered change value is sent from the information terminal 20 to the prediction device 10 along with the execution instruction. Upon receiving the execution instruction and the change value, the control unit 11 of the prediction device 10 determines that there has been a change in the indicator value.

[0102] If the aforementioned indicator value is not changed, and a prediction termination instruction is entered (Yes in S17), the series of prediction processes will end.

[0103] If it is determined in step S15 that there has been a change in the indicator value, in the next step S16, the input information entered into the prediction model is updated with the changed value. Then, using the updated input information, the processes from step S13 onward are executed again.

[0104] As described above, in the prediction system 100 of this embodiment, the feature quantities of the prediction feature are predicted using index values ​​representing two of the feature quantities of the viscoelasticity, surface roughness, and ice friction characteristics of the rubber to be predicted, which are input from the information terminal 20, and a prediction model corresponding to the other feature, the prediction feature. In other words, according to the prediction system 100 of this embodiment, if any two of the feature quantities of the viscoelasticity, surface roughness, and ice friction characteristics of the rubber to be predicted are known, it becomes possible to predict the other unknown feature quantity.

[0105] Furthermore, in this embodiment, the coefficient of ice friction, which indicates the ice friction characteristics of the rubber to be predicted, is predicted based on the first index value indicating viscoelasticity and the second index value indicating surface roughness. This makes it possible to predict the coefficient of ice friction of the rubber to be predicted with minimal effort, without requiring computationally intensive processing. In addition, since the coefficient of ice friction of the rubber to be predicted is greatly influenced by the surface roughness of its contact surface and the viscoelasticity of the rubber block, predicting the coefficient of ice friction using these index values ​​improves the prediction accuracy of the predicted value of the coefficient of ice friction. In other words, the prediction system 100 makes it possible to predict the coefficient of ice friction of the rubber to be predicted with higher accuracy.

[0106] Furthermore, by changing the aforementioned indicator values ​​displayed on the prediction result display screen 232, the input information entered into the prediction model during re-prediction can be easily updated, thus improving user convenience and operability.

[0107] Furthermore, in the above-described embodiment, a prediction system 100 was illustrated in which a prediction processing unit 111 is provided in the prediction device 10 and a display processing unit 212 is provided in the information terminal 20. However, this disclosure is not limited to this configuration. For example, the configuration of the information terminal 20 may be provided in the prediction device 10.

[0108] [Notes on the invention] The embodiments of this disclosure described above include the following disclosure items. Note that the configurations and processing functions of the following disclosure items can be selected and combined as desired.

[0109] (Disclosure Item 1) An input receiving unit that receives input of two index values ​​representing two of the following characteristics of the rubber composition: the viscoelasticity of the rubber composition, the surface roughness of the contact portion of the rubber composition that contacts the ice surface, and the friction characteristics of the rubber composition on ice. A prediction unit predicts a feature quantity that indicates other features based on the two index values ​​received by the input receiving unit, An output processing unit that outputs the predicted value predicted by the prediction unit to a predetermined output destination, A feature prediction system equipped with the following features.

[0110] (Disclosure Item 2) The feature prediction system described in Disclosure 1, wherein the output processing unit outputs the predicted value and the two index values ​​received by the input receiving unit.

[0111] (Disclosure Item 3) The feature quantity prediction system according to Disclosure 2, wherein the output processing unit outputs the predicted value and the two index values ​​received by the input receiving unit to a predetermined output destination and displays them on a predetermined display unit provided at the predetermined output destination.

[0112] (Disclosure Item 4) The feature quantity prediction system according to any one of disclosures 1 to 3, wherein the prediction unit predicts the coefficient of friction on ice, which indicates the frictional properties of the rubber composition, based on a first index value indicating viscoelasticity and a second index value indicating surface roughness.

[0113] (Disclosure Item 5) The system further includes an index value changing unit that changes at least one of the two index values ​​used in the prediction by the prediction unit, The feature prediction system according to any one of disclosure items 1 to 4, wherein the prediction unit re-predicts the feature quantities that represent the other features based on the modified index values ​​changed by the index value modification unit.

[0114] (Disclosure Item 6) The output processing unit displays a modified value input screen on the predetermined display unit, which includes a display frame for the predicted value, display frames for each of the two indicator values, and an input frame for inputting the modified value of at least one of the two indicator values. The feature quantity prediction system according to disclosure item 5, wherein the index value changing unit changes at least one of the two index values ​​used in the prediction by the prediction unit to a numerical value entered into the input frame via a predetermined operation unit.

[0115] (Disclosure Item 7) The feature prediction system according to any one of disclosures 1 to 6, wherein the prediction unit predicts feature quantities indicating the other characteristics of the rubber composition based on a prediction model learned on learning data including index values ​​indicating the viscoelasticity, surface roughness, and ice friction characteristics of each of a plurality of other rubber compositions, and the two index values ​​received by the input receiving unit.

[0116] (Disclosure Item 8) A prediction unit predicts a feature quantity representing the other feature based on two index values ​​representing two of the following features: the viscoelasticity of the rubber composition, the surface roughness of the contact portion where the rubber composition contacts the ice surface, and the ice friction characteristics of the rubber composition. An output processing unit that outputs the predicted value predicted by the prediction unit, A feature prediction device equipped with the following features.

[0117] (Disclosure Item 9) A prediction step in which a feature quantity representing the other feature is predicted based on two index values ​​representing two of the following features: the viscoelasticity of the rubber composition, the surface roughness of the contact portion where the rubber composition contacts the ice surface, and the ice friction characteristics of the rubber composition. An output step which outputs the predicted value predicted by the prediction step, A feature prediction method performed by one or more processors.

[0118] (Disclosure Item 10) A prediction step in which a feature quantity representing the other feature is predicted based on two index values ​​representing two of the following features: the viscoelasticity of the rubber composition, the surface roughness of the contact portion where the rubber composition contacts the ice surface, and the ice friction characteristics of the rubber composition. An output step which outputs the predicted value predicted by the prediction step, A program for causing one or more processors to execute, or a non-temporary computer-readable storage medium in which such program is stored. [Explanation of symbols]

[0119] 100: Feature prediction system 10: Prediction device 11: Control Unit 20: Information terminal 21: Control Unit 30: Database 40-45: Input fields 46: Input field for changed value 110: Data acquisition processing unit 111: Prediction Processing Unit 112: Output Processing Unit 113: Model Learning Department 123-125: Predictive Models 211: Input Reception Section 212: Display Processing Unit 213: Indicator value change section 230: Prediction Feature Input Screen 231: Explanatory variable input screen 231A: Explanatory variable input screen 231B: Explanatory variable input screen 232: Prediction result display screen

Claims

1. An input receiving unit that receives input of two index values ​​representing two of the following characteristics of the rubber composition: the viscoelasticity of the rubber composition, the surface roughness of the contact portion of the rubber composition that contacts the ice surface, and the friction characteristics of the rubber composition on ice. A prediction unit predicts a feature quantity that indicates other features based on the two index values ​​received by the input receiving unit, An output processing unit that outputs the predicted value predicted by the prediction unit to a predetermined output destination, A feature prediction system equipped with the following features.

2. The feature quantity prediction system according to claim 1, wherein the output processing unit outputs the predicted value and the two index values ​​received by the input receiving unit.

3. The feature quantity prediction system according to claim 2, wherein the output processing unit outputs the predicted value and the two index values ​​received by the input receiving unit to a predetermined output destination and displays them on a predetermined display unit provided at the predetermined output destination.

4. The feature quantity prediction system according to claim 1 or 2, wherein the prediction unit predicts the coefficient of friction on ice, which indicates the frictional properties of the rubber composition, based on a first index value indicating viscoelasticity and a second index value indicating surface roughness.

5. The system further includes an index value changing unit that changes at least one of the two index values ​​used in the prediction by the prediction unit, The feature quantity prediction system according to claim 3, wherein the prediction unit re-predicts the feature quantity that indicates the other features based on the modified index value changed by the index value changing unit.

6. The output processing unit displays a modified value input screen on the predetermined display unit, which includes a display frame for the predicted value, display frames for each of the two indicator values, and an input frame for inputting the modified value of at least one of the two indicator values. The feature quantity prediction system according to claim 5, wherein the index value changing unit changes at least one of the two index values ​​used in the prediction by the prediction unit to a numerical value entered into the input frame via a predetermined operation unit.

7. The feature quantity prediction system according to claim 1 or 2, wherein the prediction unit predicts feature quantities indicating the other characteristics of the rubber composition based on a prediction model learned based on learning data including index values ​​indicating the viscoelasticity, surface roughness, and ice friction characteristics of each of a plurality of other rubber compositions, and the two index values ​​received by the input receiving unit.

8. A prediction unit predicts a feature quantity representing the other feature based on two index values ​​representing two of the following features: the viscoelasticity of the rubber composition, the surface roughness of the contact portion where the rubber composition contacts the ice surface, and the ice friction characteristics of the rubber composition. An output processing unit that outputs the predicted value predicted by the prediction unit, A feature prediction device equipped with the following features.

9. A prediction step in which a feature quantity representing the other feature is predicted based on two index values ​​representing two of the following features: the viscoelasticity of the rubber composition, the surface roughness of the contact portion of the rubber composition that contacts the ice surface, and the ice friction characteristics of the rubber composition. An output step which outputs the predicted value predicted by the prediction step, A feature prediction method performed by one or more processors.

10. A prediction step in which a feature quantity representing the other feature is predicted based on two index values ​​representing two of the following features: the viscoelasticity of the rubber composition, the surface roughness of the contact portion of the rubber composition that contacts the ice surface, and the ice friction characteristics of the rubber composition. An output step which outputs the predicted value predicted by the prediction step, A program that causes one or more processors to run.

Images