Method and system for manufacturing compounded rubber

Abstract

To provide a manufacturing method and manufacturing system for a kneaded rubber, by using an open-structured roll kneader to efficiently manufacture final kneaded rubber with a desired quality for each kneaded product specification.SOLUTION: There is provided a manufacturing method of a kneaded rubber, comprising: detecting a condition of a kneaded material R circulating through the circulation path using a sensor unit 9 for each specification of the kneaded material R; inputting detected data to a calculation device 10 along with kneading conditions at a time when the data is detected; classifying the input detected data into normal data indicating a good kneading state and abnormal data indicating an unsatisfactory kneading state based on a preset criteria in the calculation device 10; calculating kneading conditions that can realize normal data by the calculation device 10 by performing machine learning using the kneading conditions input to the calculation device 10 together with normal data and the kneading conditions input to the calculation device 10 together with abnormal data as learning data; and performing kneading by applying these calculated kneading conditions to a next kneaded material R of same specifications.SELECTED DRAWING: Figure 2

Description

Technical Field

[0004] , , , , ,

[0001] The present invention relates to a method and a manufacturing system for producing kneaded rubber, and more particularly to a method and a manufacturing system for efficiently producing a final kneaded rubber of desired quality for each specification of a kneaded product using an open-structured roll kneader.

Background Art

[0002] Various rubber products such as tires and conveyor belts are manufactured by vulcanizing unvulcanized rubber. This unvulcanized rubber is, for example, a final kneaded rubber produced by sufficiently kneading a kneaded product of a primary kneaded rubber produced by kneading a raw rubber and a non-vulcanized compounding agent and a vulcanizing compounding agent. As an apparatus for producing the final kneaded rubber, various open-structured roll kneaders have been proposed (see, for example, Patent Document 1).

[0003] In an open-structured roll kneader, a kneaded product of a primary kneaded rubber and a vulcanizing compounding agent is formed in a loop shape and circulated through a circulation path, and is sufficiently kneaded by repeatedly passing between a pair of kneading rolls installed in the circulation path to produce a final kneaded rubber. In Patent Document 1, it is proposed to knead by adjusting the clearance (roll gap) between a pair of kneading rolls, the bank amount of the kneaded product on the pair of kneading rolls, the temperature of the kneaded product, and the like. However, the conditions at the manufacturing site (environment and equipment) are not always constant and have some variations, and there are also some variations even when the primary kneaded rubber and the vulcanizing compounding agent are set to the same specifications. Therefore, it is assumed that even if the parameters proposed in Patent Document 1 are simply adjusted, a final kneaded rubber of desired quality cannot be efficiently produced. Therefore, there is room for improvement in efficiently producing a final kneaded rubber of desired quality for each specification of a kneaded product using an open-structured roll kneader.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

[0005] The object of the present invention is to provide a method and system for manufacturing compounded rubber that can efficiently produce final compounded rubber of desired quality according to the specifications of the compounded material, using an open-structure roll kneader. [Means for solving the problem]

[0006] In order to achieve the above objective, a method for producing compounded rubber involves forming a kneaded mixture of raw rubber and a non-vulcanizing compound into a loop shape, and then circulating this mixture through a circulation path of an open-structured roll kneader equipped with a pair of kneading rolls to produce the final compounded rubber, For each specification of the aforementioned kneaded material, the circulation path is circulated. doing The sensor unit detects the state of the mixed material, and the detected data, along with the mixing conditions at the time of detection, is input to the calculation unit. The calculation unit, based on a preset criterion, distinguishes the input detected data into normal data indicating a good mixing state and abnormal data indicating a poor mixing state. By using machine learning with the mixing conditions input to the calculation unit along with the normal data and the mixing conditions input to the calculation unit along with the abnormal data as training data, the calculation unit calculates the mixing conditions that can achieve the normal data, and applies these calculated conditions when mixing the next batch of mixed material of the same specifications. The detection data used includes the bank amount of the kneaded material remaining on the pair of kneading rolls, whether or not there is damage to the loop shape of the kneaded material in the circulation path, whether or not there are cracks or chips in the kneaded material, and whether or not there are any raised lumps. It is characterized by the following:

[0007] The present invention relates to a rubber compounding system, which includes an open-structure roll kneader having a circulation path in which a mixture of primary rubber, obtained by mixing raw rubber with a non-vulcanizing compound, and a vulcanizing compound is formed in a loop shape and circulated for mixing, and a pair of kneading rolls are installed in this circulation path. The system further relates to a system for manufacturing compounded rubber that circulates the circulation path according to the specifications of the mixture. doing A sensor unit that detects the state of the kneaded material, and the roll kneader that detects the detection data when the detection data is detected. in of kneading The system includes a calculation unit to which conditions are input, and the calculation unit distinguishes the input detection data into normal data indicating a good mixing state and abnormal data indicating a poor mixing state based on a preset criterion, and the normal data is input to the calculation unit together with the conditions. kneading The conditions and the abnormal data are input to the calculation device. kneading The conditions are used as training data for machine learning, thereby enabling the realization of the normal data. kneading The conditions are configured to be calculated by the calculation device, and these calculated conditions kneading The conditions apply when mixing the aforementioned mixture of the same specifications which will be mixed next. The detection data used includes the bank amount of the kneaded material remaining on the pair of kneading rolls, whether or not the loop shape of the kneaded material in the circulation path is damaged, whether or not there are cracks or chips in the kneaded material, and whether or not there are any raised lumps. It is characterized by the following: [Effects of the Invention]

[0008] According to the present invention, by using machine learning to apply the mixing conditions input to the calculation device along with the normal data, and the mixing conditions input to the calculation device along with the abnormal data, as training data for each specification of the compound, it becomes possible to accurately calculate the mixing conditions that can achieve the normal data. Therefore, by applying these calculated mixing conditions to the next compound of the same specifications to be compounded, the compound will be compounded in a state equivalent to that of the normal data. Thus, it is advantageous to efficiently manufacture the final compounded rubber of the desired quality for each specification of the compound. [Brief explanation of the drawing]

[0009] [Figure 1] This is an explanatory diagram illustrating an embodiment of a compound rubber manufacturing system in a side view. [Figure 2] Figure 1 is an explanatory diagram illustrating the process of mixing the ingredients using the manufacturing system shown in Figure 1. [Figure 3] Figure 2 is an explanatory diagram illustrating the area around the pair of kneading rolls in a front view. [Figure 4] Figure 3 is an explanatory diagram illustrating the state in which the Frenda is in operation. [Figure 5] This is an explanatory diagram illustrating the state in which the final mixed rubber produced is being discharged from the roll mixer shown in Figure 2. [Figure 6] Figure 2 is an explanatory diagram illustrating a state in which the loop shape of the kneaded material has broken. [Modes for carrying out the invention]

[0010] The method for producing compounded rubber and the production system of the present invention will be described below based on the embodiments shown in the figures.

[0011] In the embodiment of the compounded rubber manufacturing system 1 illustrated in Figures 1 to 4, a vulcanizing compound M3 is added to a primary compounded rubber R1, which is made by mixing raw rubber M1 with a non-vulcanizing compound M2. The mixture R (M1, M2, M3) of the primary compounded rubber R1 and the vulcanizing compound M3 is thoroughly mixed to produce a final unvulcanized compounded rubber Rf of the desired quality. This desired quality is a state in which the vulcanizing compound M3 (and non-vulcanizing compound M2) are evenly dispersed and the target viscosity is achieved. However, if the target viscosity is achieved, it can be assumed that M2 and M3 are generally evenly dispersed, so the desired quality can also be replaced with the target viscosity. The raw rubber M1, non-vulcanizing compound M2, and vulcanizing compound M3 that constitute the final compounded rubber Rf (mixture R) are each mixed using predetermined amounts.

[0012] The raw rubber M1 is selected from natural rubber, various synthetic rubbers, etc., and one or more types are used depending on the purpose. The non-vulcanizing compounding agent M2 is appropriately selected from various known compounding agents such as carbon black and silica, depending on the purpose. The vulcanizing compounding agent M3 is appropriately selected from various known compounding agents such as sulfur, vulcanization activators and vulcanization accelerators, depending on the purpose.

[0013] The primary kneaded rubber R1 obtained by kneading the raw rubber M1 and the non-vulcanized compounding agent M2 may be kneaded by this roll kneader 2 or by other kneaders (for example, a Banbury mixer, etc.). In order to prevent the powdery vulcanizing compounding agent M3 from scattering, for example, a masterbatch produced by previously kneading the raw rubber M1 and the vulcanizing compounding agent M3 is used.

[0014] This manufacturing system 1 includes an open-structured roll kneader 2, a non-contact sensor unit 9 (9a, 9b), and an arithmetic unit 10. A display 11 is connected to the arithmetic unit 10 via wire or wirelessly.

[0015] As illustrated in FIG. 2, in the roll kneader 2, the kneaded material R (M1, M2, M3) is formed in a loop shape. As the roll kneader 2, various known open-structured ones can be used. This roll kneader 2 includes a pair of opposing kneading rolls 2a, 2b, a reiterative conveyor 5a, a transfer conveyor 5b, a supply conveyor 6, a blender 7, and a control device 8. The arrows D, W, H in the figure indicate the front-rear direction, width direction, and height direction of the roll kneader 2 (kneading rolls 2a, 2b), respectively.

[0016] The pair of kneading rolls 2a, 2b are rotated by hydraulic or electric drive motors 4a, 4b provided separately, respectively. The rotation speeds can be adjusted independently and can rotate forward and backward. An actuator 3 such as a hydraulic cylinder is connected to one of the kneading rolls 2a, and the gap (roll gap) between the kneading rolls 2a, 2b can be adjusted by moving this kneading roll 2a closer to and away from the other kneading roll 2b.

[0017] The kneading conveyor 5a extends horizontally from a position above the pair of kneading rolls 2a and 2b, and the extending direction changes obliquely downward midway. The transfer conveyor 5b extends horizontally from a position below the pair of kneading rolls 2a and 2b and is rotatable in the vertical direction midway. The lower end portion extending obliquely downward of the kneading conveyor 5a is located around the rotation center of the transfer conveyor 5b. As illustrated in FIG. 2, when the transfer conveyor 5b is rotated upward midway, a circulation path through which the loop-shaped kneaded material R circulates is formed by the kneading conveyor 5a and the transfer conveyor 5b.

[0018] The supply conveyor 6 is disposed at a position above the pair of kneading rolls 2a and 2b. The sheet-shaped masterbatch (vulcanization compounding agent M3) conveyed by the supply conveyor 6 is introduced between the pair of kneading rolls 2a and 2b. Known conveyor devices can be used for the kneading conveyor 5a, the transfer conveyor 5b, and the supply conveyor 6 respectively. The kneaded material R is placed on the upper surface of the conveyor belt provided in each of them and conveyed.

[0019] The blender 7 is disposed vertically between the tip of the kneading conveyor 5a and the pair of kneading rolls 2a and 2b. The blender 7 changes the widthwise position of the kneaded material R introduced from the tip of the kneading conveyor 5a onto the pair of kneading rolls 2a and 2b.

[0020] In this embodiment, as illustrated in FIGS. 3 and 4, a pair of belt conveyor-type blenders 7 are installed spaced apart in the width direction. Each blender 7 is slidable in the width direction. Since the kneaded material R passes from above downward through the gap between the blenders 7, by moving each blender 7 in the widthwise position to change the widthwise position of the gap between the blenders 7, the widthwise position of the kneaded material R introduced into the pair of kneading rolls 2a and 2b changes. The blender 7 is not limited to the type shown in this embodiment, and various known types can be used.

[0021] The control device 8 controls the operation of the components of the roll kneader 2. Specifically, the operation of the pair of kneading rolls 2a and 2b (drive motors 4a and 4b), actuator 3, kneading conveyor 5a, transfer conveyor 5b, supply conveyor 6, and blender 7 is controlled by the control device 8. A known computer is used as the control device 8. The control device 8 receives data such as the drive torque of the drive motors 4a and 4b, the roll gap of the kneading rolls 2a and 2b, the roll rotation speed, the conveying speed of the kneading conveyor 5a and transfer conveyor 5b, the temperature of the kneaded material R, and data (instruction data) from the calculation device 10, which will be described later. Based on the input data, the control device 8 controls the operation of each component.

[0022] The sensor unit 9 detects the state of the kneaded material R (final kneaded rubber Rf) circulating in the circulation path according to the specifications of the kneaded material R. As the sensor unit 9, known non-contact sensors such as digital cameras, laser sensors, and infrared sensors can be used. The sensor unit 9 is installed in a location where it can detect the kneaded material R within the detection range of the circulation path, and the number of units installed is not particularly limited. In this embodiment, the sensor unit 9a is positioned to detect the kneaded material R accumulating on a pair of kneading rolls 2a and 2b, and the sensor unit 9b is positioned to detect the kneaded material R being fed from the tip of the kneading conveyor 5a toward the pair of kneading rolls 2a and 2b.

[0023] One of the sensor units 9a primarily detects the bank volume B of the kneaded material R that remains on the pair of kneading rolls 2a and 2b. The bank volume B is defined, for example, as the volume of the kneaded material R that is above the tangent line connecting the upper ends of the pair of kneading rolls 2a and 2b (a straight line shown as a dashed line in Figure 2). This bank volume B is calculated by the arithmetic unit 10 through approximate calculation based on the detection data (image data) from the sensor unit (camera device) 9a.

[0024] The other sensor unit 9b detects whether there is an abnormal condition in the kneaded material R circulating in the circulation path (such as the presence or absence of breakage, cracks or chips, or raised lumps). It is advisable to identify in advance the locations in the circulation path where abnormal conditions of the kneaded material R are likely to occur, and to install the sensor unit 9b in a position where it can detect the kneaded material R in those identified locations.

[0025] The calculation unit 10 receives the detection data from the sensor unit 9 along with the mixing conditions at the time the detection data was received. A known computer is used as the calculation unit 10. These mixing conditions include at least three types: the roll caps of the mixing rolls 2a and 2b, the roll rotation speed, and the conveying speed of the mixing conveyor 5a and the transfer conveyor 5b (i.e., the circulation speed of the mixed material R in the circulation path). In addition, for example, the ambient temperature when the mixed material R is being mixed can also be included. The mixing conditions input to the calculation unit 10 along with the detection data from the sensor unit 9 are selected to have a significant impact on the quality of the final mixed rubber Rf produced.

[0026] The display 11 shows various data input to the arithmetic unit 10 and data processed by the arithmetic unit 10 (calculation results). For example, detection data (image data) from the sensor unit (camera device) 9 is displayed on the display 11 in real time. The display 11 is not mandatory and can be provided at the discretion of the system.

[0027] To efficiently produce the final compounded rubber Rf of the desired quality for each compound specification using this manufacturing system 1, it is necessary to calculate the optimal mixing conditions and apply those calculated optimal mixing conditions to the compound R. Therefore, an example of the procedure for calculating the optimal mixing conditions will be explained below.

[0028] First, using this manufacturing system 1, a large number of compounding materials R of the same specifications (multiple batches) are mixed to produce a large number of final compounded rubber Rf of the desired quality. More specifically, as illustrated in Figure 1, the vulcanizing compounding agent M3 is fed as a master batch onto a pair of mixing rolls 2a and 2b via a supply conveyor 6. In this roll mixer 2, the primary compounded rubber R1 circulates through an annular circulation path formed by a re-mixing conveyor 5a and a transfer conveyor 5b.

[0029] Next, as illustrated in Figure 2, the compound R, in which the vulcanizing compound M3 is mixed with the primary compounded rubber R1, repeatedly circulates through the circulation path. In the circulation path, the compound R is compressed each time it passes between a pair of mixing rolls 2a and 2b that rotate in opposite directions, applying a suitable shear force and being mixed. The compound R accumulates on the pair of mixing rolls 2a and 2b, creating a predetermined bank volume B. If the bank volume B is excessive, a compound R in which the vulcanizing compound M3 is sufficiently dispersed cannot be obtained. Therefore, in order to reduce the bank volume B, the blender 7 is operated in a predetermined pattern, as illustrated in Figures 3 and 4, to change the widthwise position of the compound R being fed into the pair of mixing rolls 2a and 2b (leveling the height of the accumulated compound R). As the compound R repeatedly circulates through the circulation path and is mixed, the vulcanizing compound M3 is evenly dispersed, and the final compounded rubber Rf of the desired quality (target viscosity) is produced.

[0030] Once the final mixed rubber Rf is produced, the transfer conveyor 5b is rotated downwards midway through the process, as illustrated in Figure 5, to eliminate the circulation path. As a result, the final mixed rubber Rf is discharged from the roll kneader 2 by the transfer conveyor 5b and transported to the next process. The mixing conditions when mixing one batch of the mixed material R are kept constant from start to finish.

[0031] In this way, when mixing a mixture R of the same specifications, the sensor unit 9 detects the state of the mixture R circulating in the circulation path, and inputs the detected data, along with the mixing conditions at the time the data was detected, to the calculation unit 10. The calculation unit 10 distinguishes the input detected data into normal data, which indicates a good mixing state, and abnormal data, which indicates a poor mixing state, based on pre-set criteria.

[0032] More specifically, the computing unit 10 is input with pre-set criteria for distinguishing the detection data from the sensor unit 9 into normal data indicating a good mixing state and abnormal data indicating a poor mixing state. For example, for the detection data from sensor unit 9a (bank amount B), the acceptable range of bank amount B in which the vulcanizing compounding agent M3 is sufficiently dispersed is known empirically, so the detection data is distinguished into normal data and abnormal data based on this acceptable range. For the detection data from sensor unit 9b (presence or absence of fractures, cracks or chips, and raised lumps in the mixed material R circulating in the circulation path), if there are no fractures, cracks or chips in the mixed material R, it is classified as normal data, and if there are, it is classified as abnormal data. For raised lumps in the mixed material R, the range of sizes in which these lumps adversely affect the mixing state is known empirically, so the detection data is distinguished into normal data and abnormal data based on this size range.

[0033] When detection data from the sensor unit 9 is input to the arithmetic unit 10, the detection data is classified into normal data and abnormal data based on a preset criterion. Both the normal data and the abnormal data are linked to the input kneading conditions and stored in the arithmetic unit 10.

[0034] In this way, a large number of kneading conditions input to the arithmetic unit 10 along with normal data, and kneading conditions input to the arithmetic unit 10 along with abnormal data, are stored and used as training data for machine learning. The arithmetic unit 10 uses this training data to perform machine learning and calculates kneading conditions that can produce normal data. That is, it calculates kneading conditions such that when the kneaded mixture R is detected by the sensor unit 9, the detected data is distinguished from normal data. As a machine learning method, various known methods such as deep learning using neural networks can be employed. When the mixture R is kneaded with the calculated kneading conditions, the detected data detected by the sensor unit 9 is classified as normal data, so these calculated kneading conditions can be said to be optimal kneading conditions.

[0035] As described above, after calculating the optimal mixing conditions, these calculated mixing conditions are applied when mixing the next batch of compound R with the same specifications. That is, when mixing the next batch of compound R as illustrated in Figures 2 to 4, the mixing conditions calculated by the calculation unit 10 are input to the control device 8, and the control device 8 applies the calculated mixing conditions to perform the mixing. By applying these mixing conditions, the compound R is mixed in a state equivalent to that of normal data. Therefore, in the circulation path where the compound R is mixed, the bank amount B does not become excessive, and there is no occurrence of loop-shaped breakage, cracks, or chips, and the mixing is performed in a normal state. Therefore, according to this embodiment, it is advantageous to efficiently manufacture the final compounded rubber Rf of the desired quality for each specification of compound R. When mixing each compound R, there are actually slight variations in the manufacturing site (environment and equipment), and even if the compound R is set to the same specifications, there will still be some variation. However, by mixing the compound R as in this embodiment, it is possible to efficiently produce the final compounded rubber Rf of the desired quality for each specification of compound R.

[0036] Furthermore, the detection data from each roll mixer 2 can be used as training data to calculate the optimal mixing conditions, and these calculated mixing conditions can be applied when mixing the next compound R using that roll mixer 2. In addition, if roll mixers 2 have equivalent specifications, the detection data (training data) can be shared and used. That is, the detection data from multiple roll mixers 2 with equivalent specifications can be used as training data to calculate the optimal mixing conditions, and these calculated mixing conditions can be applied when mixing the next compound R using these roll mixers 2.

[0037] While a large number of training data points is preferable, if the data is acquired very long ago, it may not adequately reflect the current state of the roll mixer 2. Therefore, it is advisable to use the training data for a predetermined number of recently mixed materials R, and update the training data each time the next batch of materials R is mixed. For example, this predetermined number should be around 30 to 50. This will be even more advantageous in calculating the optimal training conditions for the current roll mixer 2.

[0038] To efficiently obtain the final kneaded rubber Rf of the desired quality, the bank volume B of the kneaded material R and whether or not the loop shape of the kneaded material R is damaged in the circulation path have a very significant impact. Therefore, it is preferable to use at least these two types of detection data from the sensor unit 9.

[0039] To correct abnormalities in the compounding material R (such as breakage, cracks, defects, or bulging lumps) in the bank amount B of the compounding material R and in the circulation path, it is particularly effective to change the roll cap between the pair of compounding rolls 2a and 2b, the roll rotation speed of the pair of compounding rolls 2a and 2b, and the circulation speed of the compounding material R in the circulation path. Therefore, it is preferable to include at least these three types of compounding conditions in the compounding conditions used in this method of manufacturing compounded rubber.

[0040] The operation of the Frenda 7 is also beneficial in correcting abnormalities in the bank amount B of the compound R and in the circulation path of the compound R. Therefore, it is more preferable to include the operating conditions of the Frenda 7 in the compounding conditions used in this method of manufacturing compounded rubber.

[0041] If there is a large difference in ambient temperature between different batches of compound R during mixing, it will affect the bank volume B of the compound R and the occurrence of abnormalities in the compound R in the circulation path. Therefore, it is preferable that the mixing conditions used in this compound rubber manufacturing method include the ambient temperature during mixing of each compound R, and that the calculation device 10 calculates the mixing conditions that will produce normal data according to the ambient temperature when mixing the next batch of compound R with the same specifications.

[0042] More specifically, the ambient temperature during the mixing of each compound R is classified into several levels, and the calculation device 10 calculates the mixing conditions that can achieve normal data for each classified level. Then, when mixing the same compound R again, the ambient temperature level at that time is input to the calculation device 10, and the calculation device 10 calculates (extracts) the mixing conditions that can achieve normal data at that ambient temperature level, and these calculated mixing conditions are applied to the mixing of that compound R.

[0043] This disclosure encompasses the following inventions. Invention 1: In a method for producing compounded rubber, in which a primary compounded rubber, in which a non-vulcanizing compound is mixed with raw rubber, is formed into a loop shape from a mixture of the primary compounded rubber and a vulcanizing compound, and this mixture is circulated and mixed in a circulation path of an open-structured roll mixer equipped with a pair of mixing rolls to produce the final compounded rubber, A method for manufacturing kneaded rubber, comprising: detecting the state of the kneaded material circulating in the circulation path using a sensor unit for each specification of the kneaded material; inputting the detected data, along with the kneading conditions at the time the detected data was received, into a calculation unit; the calculation unit distinguishing the input detected data into normal data indicating a good kneading state and abnormal data indicating a poor kneading state based on a preset standard; using the kneading conditions input to the calculation unit along with the normal data and the kneading conditions input to the calculation unit along with the abnormal data as learning data through machine learning; calculating the kneading conditions that can achieve the normal data; and applying these calculated kneading conditions when kneading the next kneaded material of the same specification. Invention 2: A method for manufacturing kneaded rubber according to Invention 1, wherein the kneading conditions for a predetermined number of kneaded materials most recently are used as learning data, and the learning data is updated each time the next kneaded material of the same specifications is kneaded. Invention 3: A method for manufacturing kneaded rubber according to Invention 1 or 2, using at least two types of detection data: the bank amount of the kneaded material remaining on the pair of kneading rolls, and whether or not the loop shape of the kneaded material in the circulation path is damaged. Invention 4: A method for manufacturing kneaded rubber according to any one of Inventions 1 to 3, wherein the kneading conditions are at least three types: the roll cap between the pair of kneading rolls, the roll rotation speed of the pair of kneading rolls, and the circulation speed of the kneaded material. Invention 5: A method for manufacturing kneaded rubber according to any one of Inventions 1 to 4, wherein the kneading conditions include operating conditions for a frenda that changes the widthwise input position of the kneaded material to the pair of kneading rolls. Invention 6: A method for manufacturing kneaded rubber according to any one of Inventions 1 to 5, wherein the kneading conditions include the ambient temperature when the kneaded material is being kneaded, and the kneading conditions that enable the realization of the normal data are calculated by the calculation device according to the ambient temperature when the next kneaded material is being kneaded. Invention 7: In a rubber compounding system equipped with an open-structure roll kneader having a circulation path in which a mixture of primary compounded rubber (raw rubber mixed with a non-vulcanizing compound) and a vulcanizing compound is formed in a loop shape and circulated for mixing, a pair of kneading rolls is installed in this circulation path, For each specification of the compound, a sensor unit detects the state of the compound circulating in the circulation path, and the detection data is used by the roll kneader when the detection data is detected. in of kneading It has a calculation unit that receives input along with conditions, In the aforementioned calculation device, based on a preset criterion, the input detection data is distinguished into normal data indicating a good mixing state and abnormal data indicating a poor mixing state, and the normal data is input to the calculation device along with the abnormal data. kneading The conditions and the abnormal data are input to the calculation device. kneading The conditions are used as training data for machine learning, thereby enabling the realization of the normal data. kneading The conditions are configured to be calculated by the calculation device, and these calculated conditions kneading A rubber compounding system in which conditions are applied when compounding the compound of the same specifications to be compounded next. [Explanation of symbols]

[0044] 1. Manufacturing System 2-roll kneader 2a, 2b Mixing rolls 3 Actuators 4a, 4b drive motors 5a Mixing conveyor 5b Transfer conveyor 6. Supply conveyor 7 Brenda 8 Control device 9(9a, 9b) Sensor section 10 Arithmetic unit 11 displays R1 Primary compounding rubber R Mixed ingredients Rf Final Mixed Rubber M1 Raw Rubber M2 Non-vulcanizing compounding agent M3 vulcanizing compound

Claims

1. In a method for producing compounded rubber, in which a primary compounded rubber, in which a non-vulcanizing compound is mixed with raw rubber, is formed into a loop shape from a mixture of the primary compounded rubber and a vulcanizing compound, and this mixture is circulated and mixed in a circulation path of an open-structured roll mixer equipped with a pair of mixing rolls to produce the final compounded rubber, For each specification of the compound, the sensor unit detects the state of the compound circulating in the circulation path, and inputs the detected data, along with the mixing conditions at the time of detection, into a calculation unit. The calculation unit, based on a preset criterion, distinguishes the input detected data into normal data indicating a good mixing state and abnormal data indicating a poor mixing state. By using machine learning with the mixing conditions input to the calculation unit along with the normal data and the mixing conditions input to the calculation unit along with the abnormal data as training data, the calculation unit calculates the mixing conditions that can achieve the normal data, and applies these calculated mixing conditions when mixing the next compound of the same specification. A method for manufacturing kneaded rubber, using the bank amount of the kneaded material remaining on the pair of kneading rolls, whether or not there is damage to the loop shape of the kneaded material in the circulation path, whether or not there are cracks or chips in the kneaded material, and whether or not there are raised lumps, as the detection data.

2. A method for manufacturing kneaded rubber according to claim 1, wherein the kneading conditions for a predetermined number of kneaded materials most recently are used as learning data, and the learning data is updated each time the next kneaded material of the same specifications is kneaded.

3. The method for manufacturing kneaded rubber according to claim 1 or 2, wherein the kneading conditions consist of at least three types: the roll cap between the pair of kneading rolls, the roll rotation speed of the pair of kneading rolls, and the circulation speed of the kneaded material.

4. The method for manufacturing kneaded rubber according to claim 1 or 2, wherein the kneading conditions include operating conditions for a frenda that changes the widthwise input position of the kneaded material to the pair of kneading rolls.

5. The method for manufacturing kneaded rubber according to claim 1 or 2, wherein the kneading conditions include the ambient temperature when the kneaded material is being kneaded, and the calculation device calculates the kneading conditions that can achieve the normal data according to the ambient temperature when the next kneaded material is being kneaded.

6. In a rubber compounding system equipped with an open-structure roll kneader having a circulation path in which a mixture of primary compounded rubber (raw rubber mixed with a non-vulcanizing compound) and a vulcanizing compound is formed in a loop shape and circulated for mixing, a pair of kneading rolls is installed in this circulation path, For each specification of the compound, the system includes a sensor unit that detects the state of the compound circulating in the circulation path, and a calculation unit that receives the detected data along with the mixing conditions in the roll kneader at the time the data was detected. In the aforementioned computing device, based on a preset criterion, the input detection data is distinguished into normal data indicating a good mixing state and abnormal data indicating a poor mixing state. The mixing conditions input to the computing device along with the normal data, and the mixing conditions input to the computing device along with the abnormal data, are machine-learned as training data, so that the computing device calculates the mixing conditions that can achieve the normal data. These calculated mixing conditions are then applied when mixing the next batch of the same material. A rubber compounding manufacturing system in which the detection data includes the bank amount of the compounded material accumulating on the pair of compounding rolls, whether or not there is damage to the loop shape of the compounded material in the circulation path, whether or not there are cracks or chips in the compounded material, and whether or not there are any raised lumps.

Images