Processing system and processing method for unvulcanized rubber
The system addresses the challenge of achieving target specifications for unvulcanized rubber by using a predictive model and control device to adjust processing conditions, enhancing productivity and stability.
Patent Information
- Application Number
- JP2025077868
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2025-05-08
- Publication Date
- 2026-11-19
AI Technical Summary
Existing methods fail to achieve strip-shaped unvulcanized rubber with target specifications at the end of the processing line with high productivity due to variations in shape and mass caused by shrinkage during conveyance.
A system using sensors to acquire intermediate specifications, a predictive model constructed via machine learning, and a control device to adjust processing conditions, such as conveying speed, to minimize the difference between predicted final specifications and target specifications.
The system effectively produces unvulcanized rubber meeting target specifications at the final point with increased productivity by predicting and controlling processing conditions, reducing the need for manual adjustments and improving stability.
Smart Images

Figure 2026182000000001_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to an unvulcanized rubber processing system and processing method, and more specifically, to an unvulcanized rubber processing system and processing method that can obtain the strip-shaped unvulcanized rubber discharged from a processing apparatus at the end point of the conveyed processing line to meet target specifications with higher productivity.
Background Art
[0002] At manufacturing sites for rubber products such as tires, there are processing lines where strip-shaped unvulcanized rubber is discharged from processing apparatuses such as extruders and rolling mills, and conveyed by conveying devices such as conveyor belts. The strip-shaped unvulcanized rubber shrinks during the process of being conveyed from the processing apparatus to the end point of the processing line, and it is arranged that unvulcanized rubber meeting target specifications can be obtained at this end point.
[0003] A method has been proposed in which the width of the extruded rubber immediately after being extruded from a rubber extruder is measured, and the peripheral speeds of the first drive roller and the second drive roller in the first roller conveyor are adjusted based on the deviation between the measured value and a set value (see, for example, Patent Document 1). By adjusting the peripheral speeds, the shrinkage of the extruded rubber is promoted to achieve early shape stabilization. In addition, in this method, the mass per unit length of the extruded rubber is measured by a weigher disposed between the first roller conveyor and the second roller conveyor, and the peripheral speed of the drive roller in the second roller conveyor is adjusted based on the deviation between the measured value and a set value. In this invention, the peripheral speed of the drive roller is controlled so that the measured value of the extruded rubber at an intermediate point of the processing line falls within a predetermined range of a preset set value. Since the shape and mass of the extruded rubber are not equal at the intermediate point and the end point of the processing line, simply making the measured value at the intermediate point of the processing line approximate to the set value is not sufficient to obtain the extruded rubber with desired specifications at the end point. Therefore, there is room for improvement in order to obtain the strip-shaped unvulcanized rubber discharged from the processing apparatus to meet target specifications with higher productivity at the end point of the conveyed processing line.
Prior Art Documents
Patent Documents
[0004] [Patent Document 1] Japanese Patent Publication No. 2018-183916 [Overview of the Initiative] [Problems that the invention aims to solve]
[0005] The object of the present invention is to provide an unvulcanized rubber processing system and processing method that can obtain strip-shaped unvulcanized rubber discharged from a processing device into a target specification with greater productivity at the final point of the conveyed processing line. [Means for solving the problem]
[0006] The present invention provides an unvulcanized rubber processing system for achieving the above objectives, comprising: a processing device for discharging strip-shaped unvulcanized rubber; a conveying device for transporting the strip-shaped unvulcanized rubber toward the final point of the processing line by the processing device, wherein the unvulcanized rubber processing system for obtaining unvulcanized rubber of target specifications at the final point, further comprising: a sensor for acquiring the specifications of the strip-shaped unvulcanized rubber at intermediate points between the processing device and the final point as intermediate specifications; and a control device for controlling the processing conditions on the processing line, wherein the control device includes at least the transport speed of the conveying device, the processing conditions, and the intermediate specifications associated with the strip-shaped unvulcanized rubber at the final point. Multiple final specifications for unvulcanized rubber are input to a predictive model constructed by machine learning using a dataset of accumulated data as training data. When the strip-shaped unvulcanized rubber discharged from the processing apparatus is transported toward the final destination, the control device performs data processing to predict the final specifications by inputting the processing conditions and the intermediate specifications acquired by the sensors into the predictive model, and data processing to identify the processing conditions that minimize the difference between the final specifications predicted by this data processing and the target specifications as appropriate processing conditions, thereby controlling the processing conditions so that they become the appropriate processing conditions.
[0007] The present invention relates to a method for processing unvulcanized rubber, which involves discharging a strip of unvulcanized rubber from a processing device, transporting the strip of unvulcanized rubber by a transport device toward the final point of the processing line of the processing device, and obtaining unvulcanized rubber of a target specification at the final point. The method is characterized in that the specifications of the strip of unvulcanized rubber at intermediate points between the processing device and the final point are defined as intermediate specifications, and a prediction model is constructed by machine learning using a plurality of accumulated datasets as training data, which is input into a control device. When the strip of unvulcanized rubber discharged from the processing device is transported toward the final point, the control device predicts the final specifications by inputting the processing conditions and the intermediate specifications obtained by sensors at the intermediate points into the prediction model, identifies the processing conditions that minimize the difference between the predicted final specifications and the target specifications as appropriate processing conditions, and controls the processing conditions so that they become the appropriate processing conditions. [Effects of the Invention]
[0008] According to the present invention, the final specifications of the strip-shaped unvulcanized rubber at the final point are predicted by a prediction model constructed using machine learning, using the intermediate specifications of the strip-shaped unvulcanized rubber at an intermediate point between the processing apparatus and the final point. The processing conditions are then controlled to minimize the difference between the predicted final specifications and the target specifications. Therefore, it is advantageous to obtain unvulcanized rubber with the target specifications at the final point of the processing line with greater productivity. [Brief explanation of the drawing]
[0009] [Figure 1] This is an explanatory diagram illustrating an example of a processing system for unvulcanized rubber. [Figure 2] This flowchart illustrates the steps involved in the preparation process for processing unvulcanized rubber. [Figure 3]This flowchart illustrates the steps involved in the processing of unvulcanized rubber. [Figure 4] This is an explanatory diagram illustrating a dataset. [Figure 5] This is an explanatory diagram illustrating the changes over time in the final specifications of unvulcanized rubber. [Modes for carrying out the invention]
[0010] The processing system for unvulcanized rubber and the processing method of the present invention will be described below based on the embodiments shown in the figures.
[0011] The embodiment of processing system 1 illustrated in Figure 1 comprises a processing line 2 and a control device 3. The processing line 2 includes a processing device 4 (extruder or rolling device), a conveying device 5 (upstream conveyor 5a, downstream conveyor 5b, 5c), a sensor 6, a cooling device 7, and a confirmation sensor 8. The cooling device 7 and confirmation sensor 8 can be optionally provided. In Figure 1, one dashed line indicates the final point A of the processing line 2, and the other dashed line indicates an intermediate point B between the processing device 4 and the final point A. In this processing line 2, strip-shaped unvulcanized rubber R discharged from the processing device 4 is conveyed toward the final point A by the conveying device 5. Note that the strip-shaped unvulcanized rubber R also includes so-called strip material (long, thin unvulcanized rubber).
[0012] This processing system 1 is used to obtain strip-shaped unvulcanized rubber R (hereinafter referred to as "unvulcanized rubber R") discharged from the processing device 4 to meet target specifications at the final location A. These target specifications include one of the following: width, thickness, cross-sectional area, or weight per unit length of the unvulcanized rubber R. Therefore, it is possible to achieve target specifications for only one, two, three, or all four of these specifications.
[0013] The unvulcanized rubber R, transported to the final point A, is either transported to the next process or temporarily stored in a stock area before being transported to the next process. In the next process, this unvulcanized rubber R is processed as needed and used as a component of rubber products. Examples of rubber products include tires, conveyor belts, and hoses.
[0014] The control device 3 controls the movement of the components of the processing system 1. In this embodiment, the processing device 4 is controlled by a control device separate from the control device 3, but it is also possible to configure the control device 3 to control the processing device 4. Of the processing conditions in the processing line 2, at least the transport speed of the transport device 5 is controlled by the control device 3.
[0015] The control device 3 receives and stores various types of data, and performs data processing using this data. The control device 3 can use various known computers. The control device 3 has an arithmetic unit (CPU) 11, a main memory unit (memory) 12, and an auxiliary storage unit (HDD, etc.) 13. The auxiliary storage unit 13 stores the dataset 9 and prediction model 10, which will be described later.
[0016] The processing device 4 is either an extruder or a rolling mill, and various known types can be used. In this embodiment, an extruder is used as the processing device 4. In the processing line 2, unvulcanized rubber R is discharged from the processing device 4. In the case of an extruder 4, unvulcanized rubber R with a predetermined cross-sectional shape is extruded. In the case of a rolling mill 4, unvulcanized rubber R with a predetermined thickness is rolled out.
[0017] The conveying device 5 conveys the unvulcanized rubber R discharged from the processing device 4 toward the final point A. As the conveying device 5, various known conveyor devices can be used. In this embodiment, it includes an upstream conveyor 5a disposed immediately after the processing device 4, and two downstream conveyors 5b and 5c disposed on the downstream side thereof. The lower end position (around the lower end position) of the downstream conveyor 5c disposed at the most downstream side serves as the final point A on the processing line 2. The number of conveyors 5a, 5b, 5c arranged in series is not limited to three, and may be two, or four or more. The conveying device 5 may also be configured as a single conveyor extending from immediately after the processing device 4 to the final point A.
[0018] The cooling device 7 is disposed at a position midway along the processing line 2. In this embodiment, the cooling device 7 is disposed on the downstream conveyor 5b between the upstream conveyor 5a and the most downstream downstream conveyor 5c. Accordingly, the unvulcanized rubber R is conveyed from the processing device 4 to the cooling device 7 by the upstream conveyor 5a, and the unvulcanized rubber R is conveyed from the cooling device 7 to the final point A by the downstream conveyor 5c.
[0019] This cooling device 7 is of a type that injects and sprays cooling water onto the unvulcanized rubber R from nozzles 7a. Since the unvulcanized rubber R discharged from the processing device 4 is at a high temperature of, for example, about 100°C, it is cooled by the cooling device 7 while being conveyed along the processing line 2. In this embodiment, the conveyed unvulcanized rubber R is cooled by the cooling water injected and sprayed from the nozzles 7a. As for other types of the cooling device 7, a cooling tank in which cooling water is stored or the like may also be used. At the final point A, the unvulcanized rubber R reaches, for example, approximately normal temperature (ambient temperature). When the processing device 4 is a rolling device, a cooling drum, for example, is used as the cooling device 7. The unvulcanized rubber R discharged from the rolling device is cooled while in contact with a plurality of cooling drums through which cooling water flows, and is conveyed to the downstream side.
[0020] The sensor 6 is disposed at an intermediate point B, and acquires the specification of the unvulcanized rubber R at the intermediate point B as an intermediate specification. The type of the specification of the unvulcanized rubber R acquired by the sensor 6 corresponds to the type of the target specification, and includes any one of the width dimension, thickness dimension, cross-sectional area, and weight per unit length of the unvulcanized rubber R. Therefore, only any one, only any two, only any three, or all four of these specifications are acquired by the sensor 6 as the intermediate specification. As the sensor 6 for acquiring the width dimension, thickness dimension, and cross-sectional area of the unvulcanized rubber R, various non-contact sensors such as a laser sensor, an ultrasonic sensor, and a profile sensor can be used. As the sensor 6 for acquiring the weight per unit length of the unvulcanized rubber R, a known metering roller incorporated in the conveying device 5 (5a) or the like can be used. Data acquired by the sensor 6 is sequentially input to the control device 3.
[0021] The sensor 6 is preferably disposed at a position as close to the processing device 4 as possible. Therefore, as in this embodiment, it is more preferable to dispose the sensor 6 in the conveying range of the upstream-side conveying conveyor 5a among the upstream-side conveying conveyor 5a and the downstream-side conveying conveyors 5b, 5c. In this embodiment, the processing line 2 includes the upstream-side conveying conveyor 5a that conveys the unvulcanized rubber R from the processing device 4 to a downstream position including the intermediate point B, and the downstream-side conveying conveyors 5b, 5c that are disposed downstream of the upstream-side conveying conveyor 5a and convey the unvulcanized rubber R to the final point A.
[0022] The confirmation sensor 8 is disposed at the final point A, and acquires the specification of the unvulcanized rubber R at the final point A as a final specification. The specification of the unvulcanized rubber R acquired by the confirmation sensor 8 is the same as the intermediate specification, and the confirmation sensor 8 can use the same type of sensor as the sensor 6. Data acquired by the confirmation sensor 8 is sequentially input to the control device 3. The final specification of the unvulcanized rubber R may be acquired in the conveying range of the most downstream downstream-side conveying conveyor 5c, or may be acquired using necessary equipment (such as a sensor) on an adjacent measurement table, a take-up reel or the like after the unvulcanized rubber R is conveyed by the downstream-side conveying conveyor 5c. That is, the final specification of the unvulcanized rubber R may be acquired in the vicinity of the final point A.
[0023] The processing conditions controlled by the control device 3 include at least the conveying speed of the conveying device 5. However, in this embodiment, only the conveying speed of the upstream conveying conveyor 5a is included, and the conveying speeds of the other conveying devices 5 (downstream conveying conveyors 5b and 5c) are not included. The conveying speeds of the downstream conveying conveyors 5b and 5c are set to a predetermined value (constant value). The conveying speeds of the upstream conveying conveyor 5a and the downstream conveying conveyors 5b and 5c can also be included in the processing conditions controlled by the control device 3. The control device 3 controls the conveying speed of the conveying device 5 by controlling the rotational speed of the drive motors of the conveying device 5, thereby adjusting the conveying speed to the desired speed. In this embodiment, the upstream conveying conveyor 5a, the downstream conveying conveyor 5b, and the downstream conveying conveyor 5c are each equipped with separate drive motors. When two conveyor belts are arranged in series, or when four or more conveyor belts are arranged in series, for example, the conveying speed of the upstreammost conveyor belt 5a should be controlled by the control device 3 under specific processing conditions. The intermediate specifications of the unvulcanized rubber R being conveyed by the upstreammost conveyor belt 5a are then measured by the sensor 6.
[0024] More specifically, the discharge rate of unvulcanized rubber R from the processing device 4 is set to a predetermined constant value. The high-temperature unvulcanized rubber R discharged from the processing device 4 shrinks mainly due to cooling during the process of being transported by the conveying device 5. The degree of this shrinkage varies depending on differences in the specifications of the unvulcanized rubber R, differences in the lots of unvulcanized rubber R even if they are the same specifications, and differences in the ambient temperature of the processing line 2. Furthermore, the degree of shrinkage also differs due to variations in the residual stress of the unvulcanized rubber R after passing through the processing device 4. As a result, variations occur in the final specifications of the unvulcanized rubber R at the final point A.
[0025] If there is a speed difference between the discharge speed of the unvulcanized rubber R from the processing device 4 and the conveying speed of the conveying device 5, deformation occurs in the unvulcanized rubber R. The greater the conveying speed relative to the discharge speed, the greater the tensile force acting on the unvulcanized rubber R, causing it to stretch and deform. Therefore, in this processing system 1, the system takes into account the aforementioned degree of shrinkage and stretch deformation of the unvulcanized rubber R and controls the system so that the final specifications of the unvulcanized rubber R at the final point A match the target specifications.
[0026] Therefore, it is desirable to select a parameter that has a greater influence on the deformation of the unvulcanized rubber R as a processing condition controlled by the control device 3. The temperature of the unvulcanized rubber R being transported is higher on the upstream conveyor 5a than on the downstream conveyors 5b and 5c, so the unvulcanized rubber R is more easily deformed on the upstream conveyor 5a. Therefore, changing the transport speed of the upstream conveyor 5a makes it easier to change the specifications of the unvulcanized rubber R compared to changing the transport speed of the downstream conveyors 5b and 5c. In other words, changing the transport speed of the downstream conveyors 5b and 5c does not easily change the specifications of the unvulcanized rubber R. Therefore, in order to bring the unvulcanized rubber R to the target specifications, it is effective to change the transport speed of the upstream conveyor 5a as a processing condition. For this reason, it is preferable to include the transport speed of the upstream conveyor 5a as a processing condition.
[0027] In addition, for example, the ambient temperature of processing line 2 and / or the temperature of the cooling water in the cooling device 7 that cools the unvulcanized rubber R in processing line 2 can also be controlled by the control device 3 as processing conditions. When the control device 3 controls the ambient temperature, the ambient temperature is adjusted to the desired temperature by controlling the operation of the air conditioning equipment for the area where processing line 2 is installed. When the control device 3 controls the cooling water temperature, the cooling water is adjusted to the desired temperature by controlling the operation of the temperature control device for the cooling water. Furthermore, these processing conditions can also include the control parameters of processing device 4 (such as discharge speed and discharge temperature). However, since the control parameters of processing device 4 are generally set within an appropriate range according to the specifications of the unvulcanized rubber R, it is better to set them to a constant value rather than sequentially controlling and changing them.
[0028] Next, an example of the procedure for an embodiment of the processing method for unvulcanized rubber will be described.
[0029] In this processing method, the preparation process S100, illustrated in Figure 2, and the processing process S200, illustrated in Figure 3, are executed in order. The contents of each step of the preparation process S100 (S110, S120) and each step of the processing process S200 (S210~S280) will be explained below.
[0030] In step S110, illustrated in Figure 2, a dataset 9, illustrated in Figure 4, is created. The created dataset 9 is stored in the auxiliary storage unit 13 of the control device 3. The dataset 9 contains multiple final specifications associated with processing conditions and intermediate specifications.
[0031] The dataset 9 illustrated in Figure 4 will be used as training data in step S120, which will be described later. The dataset 9 contains, for each sample (1, 2, ..., m) shown in the leftmost column of the table, the final specifications (cross-sectional area at the final point A) associated with the processing conditions (transport speed of the upstream transport conveyor 5a, ambient temperature) and intermediate specifications (cross-sectional area at intermediate point B).
[0032] The processing conditions can be arbitrarily selected, but a type that has a high impact on the deformation of the unvulcanized rubber R is selected. In this embodiment, as described above, only the upstream conveyor 5a is included among the conveying devices 5. Therefore, the conveying speeds of the conveying devices 5 other than the upstream conveyor 5a (downstream conveyor 5b, 5c) are set to a predetermined value (constant value). Even if there is a difference in the conveying speed between the upstream conveyor 5a and the downstream conveyor 5b, the difference will not be very large. In addition, the unvulcanized rubber R is passed between the upstream conveyor 5a and the downstream conveyor 5b with slack, and this slack acts as a buffer to absorb the difference in the conveying speeds of the two. Therefore, problems such as the unvulcanized rubber R becoming unable to be conveyed or undergoing excessive tensile deformation due to the difference in conveying speeds of the two are avoided.
[0033] Dataset 9 is created by accumulating intermediate and final specification data obtained using sensors 6 and verification sensors 8, etc., when processing unvulcanized rubber R on processing line 2, under different desired processing conditions. Dataset 9 may include samples where the final specification does not meet the target specification. A rubber product manufacturer will have accumulated a vast amount of data, such as data obtained during the manufacturing and development process, and the results of computer simulations. Therefore, Dataset 9 can be created using this vast amount of data.
[0034] Dataset 9 should be created for each specification of the unvulcanized rubber R (differences in rubber type, type of compounding agent, or differences in the amount of rubber and compounding agents used). Then, in step S120 described later, the appropriate dataset 9 is selected according to the specifications of the unvulcanized rubber R to be processed. Note that even if the specifications of the unvulcanized rubber R are different, they can be treated as the same specification if the degree of deformation on processing line 2 can be evaluated as equivalent.
[0035] In step S120, data processing is performed to construct a prediction model 10 by performing machine learning using the dataset 9 as training data via the control device 3. The constructed prediction model 10 is stored in the auxiliary storage unit 13 of the control device 3. This prediction model 10 is a type of computer program that predicts the final specifications based on the input processing conditions and intermediate specifications. The prediction model 10 is constructed using supervised machine learning with the processing conditions and intermediate specifications of the dataset 9 as explanatory variables and the final specifications as the dependent variable. Various known models can be used for the prediction model 10, such as regression models based on multiple regression analysis, ensemble models based on decision trees, and models based on neural networks. If there are no changes to the settings in the processing system 1, steps S110 and S120 can be omitted.
[0036] In the processing step S200 illustrated in Figure 3, the unvulcanized rubber R is actually discharged from the processing device 4 and transported to the final point A of the processing line 2. For processing of unvulcanized rubber R in the same lot, S210 is performed once initially, and S220 to S270 are repeated at a set predetermined time cycle and performed at least once.
[0037] In step S210, the target specifications are input to the control device 3 via an input device such as a keyboard and stored in the auxiliary storage unit 13 of the control device 3. Note that the target specifications only need to be entered once for processing unvulcanized rubber R in the same lot. If the target specifications remain unchanged for subsequent lots, they can be used as is.
[0038] In step S220, the control device 3 performs data processing to read the processing conditions stored in the auxiliary storage unit 13. Once the target specifications are entered and the processing conditions are read, processing of the unvulcanized rubber R is started using these processing conditions. If no processing conditions are stored in the auxiliary storage unit 13, processing conditions corresponding to the target specifications are entered into the control device 3 via an input device such as a keyboard.
[0039] In step S230, intermediate specifications are obtained by sensor 6. The unvulcanized rubber R immediately after being discharged from the processing device 4 may have an unstable shape. Therefore, it is preferable to obtain intermediate specifications by sensor 6 at a position where the unvulcanized rubber R discharged from the processing device 4 is stable.
[0040] In step S240, the input processing conditions and intermediate specifications obtained by the sensor 6 are input into the prediction model 10, and the control device 3 performs data processing to predict the final specifications. The predicted final specifications are stored in the auxiliary storage unit 13 of the control device 3.
[0041] In prediction model 10, processing conditions are treated as disturbances, and the degree of change from intermediate specifications to final specifications due to these disturbances is analyzed. Therefore, this prediction model 10 is capable of accurately predicting the final specifications at the final point A when unvulcanized rubber R is processed under the input processing conditions.
[0042] In step S250, data processing is performed to identify appropriate processing conditions that minimize the difference between the final specifications predicted by the control device 3 and the target specifications. Appropriate processing conditions are those that allow the final specifications of the unvulcanized rubber R processed under those conditions to be within the acceptable range of the target specifications. In step S250, various known optimization algorithms and inverse analysis of the prediction model 10 (black box optimization) can be used.
[0043] One example of a method for inversely analyzing the prediction model 10 is mathematical optimization (mathematical programming). In mathematical optimization, the inverse function of the prediction model 10 is used to fix the final specifications to the target specifications, and the intermediate specifications and the transport speed of the upstream transport conveyor 5a among the processing conditions are used as variables, while the remaining parameters after removing the transport speed of the upstream transport conveyor 5a from the processing conditions are used as constraints. The solution of this mathematical optimization becomes the appropriate processing conditions that bring the final specifications within the acceptable range of the target specifications. If the transport speed of the upstream transport conveyor 5a and the ambient temperature are used as processing conditions, mathematical optimization will use the ambient temperature as a constraint, and the transport speed of the upstream transport conveyor 5a will be obtained as the solution.
[0044] In step S260, the appropriate processing conditions identified by the control device 3 are set as new processing conditions. In this embodiment, a more appropriate conveying speed is set relative to the current conveying speed of the upstream conveying conveyor 5a. In step S270, the control device 3 performs data processing to determine whether or not the processing of one lot of unvulcanized rubber R has been completed.
[0045] In step S270, data processing is performed to control the appropriate processing conditions set in step S260 to become new processing conditions until it is determined that the processing of one lot of unvulcanized rubber R is completed, and the unvulcanized rubber R is processed under the new processing conditions. Therefore, the current conveying speed of the upstream conveyor 5a is controlled to become a more appropriate conveying speed than the set one.
[0046] If the time cycle for repeating steps S220 to S270 is too short, the processing conditions will be changed frequently. However, if the processing conditions are changed frequently, the shape of the unvulcanized rubber R may not be stable, so the appropriate time cycle should be set by conducting tests beforehand. If the time cycle for repeating steps S220 to S270 is set to be long, the control to change to new processing conditions may only be performed once or a few times. When it is determined that the processing of one lot of unvulcanized rubber R is complete, proceed to step S280.
[0047] In step S280, data processing is performed to store the appropriate processing conditions set by the control device 3 in the auxiliary storage unit 13. The appropriate processing conditions stored in the auxiliary storage unit 13 are used as the initial processing conditions (processing conditions read out in step S220) when processing the next lot of unvulcanized rubber R.
[0048] According to this embodiment, the final specifications of the unvulcanized rubber R at the final point A are predicted by a predictive model 10 built using machine learning, using the intermediate specifications of the unvulcanized rubber R at intermediate point B between the processing device 4 and the final point A. Then, the control device 3 controls the processing conditions to achieve appropriate processing conditions that minimize the difference between the predicted final specifications and the target specifications. This reduces the length over which unvulcanized rubber R that does not reflect appropriate processing conditions is transported to the final point A, compared to the case where the specifications of the unvulcanized rubber R at the final point A of the processing line 2 are obtained and appropriate processing conditions are fed back. Therefore, it is advantageous to obtain unvulcanized rubber R that meets the target specifications more productively at the final point A of the processing line 2. Since the amount of unvulcanized rubber R that does not satisfy the target specifications can be reduced, it is possible to eliminate or reduce the frequency of tasks such as returning such unvulcanized rubber R to the processing device 4.
[0049] Furthermore, according to this embodiment, appropriate processing conditions adapted to the degree of shrinkage, which varies in various ways due to disturbances such as differences in raw material lots and seasonal changes, can be automatically set by data processing by the control device 3. This eliminates the need for human judgment in controlling processing conditions on the processing line 2, which is advantageous for stably obtaining unvulcanized rubber R of the target specifications without being affected by variations in the skill level of the workers.
[0050] Figure 5 illustrates the change over time in the final specifications of each lot when multiple lots of unvulcanized rubber R with the same specifications are processed on processing line 2. In Figure 5, the horizontal axis from time t0 to time t2 represents the processing time of the first lot, and from time t3 to time t5 represents the processing time of the second lot. The time between time t2 and time t3 is the preparation time between lots.
[0051] In the first lot, the above-described step 200 is performed, and the processing conditions are controlled by the control device 3 to appropriate processing conditions, thereby bringing the final specifications of the unvulcanized rubber R to the target specifications. In the second lot, by using the appropriate processing conditions at the end of processing in the first lot as the initial processing conditions, it becomes easier to bring the final specifications of the unvulcanized rubber R to the target specifications from the initial stage. When the preparation time between lots (time t2 to time t3) is short and multiple lots of unvulcanized rubber R are processed continuously, using the appropriate processing conditions at the end of processing in the immediately preceding lot as the initial processing conditions for the immediately following lot is particularly effective in bringing the final specifications of the unvulcanized rubber R to the target specifications from the initial stage of the immediately following rod.
[0052] As in this embodiment, by obtaining the intermediate specifications of the unvulcanized rubber R being transported by the upstream transport conveyor 5a, which is located closer to the processing device 4, using the sensor 6, and performing step S200, the length over which the unvulcanized rubber R, which does not reflect the appropriate processing conditions, is transported to the final point A can be further reduced. For this reason, it is desirable to basically place the sensor 6 as close to the processing device 4 as possible to obtain the intermediate specifications of the unvulcanized rubber R. However, as mentioned above, the shape of the unvulcanized rubber R immediately after discharge from the processing device 4 may not be stable, so it is best to place the sensor 6 so that the intermediate specifications can be measured when the shape of the unvulcanized rubber R is stable and as close to the processing device 4 as possible.
[0053] Dataset 9 can also include processing conditions (appropriate processing conditions), intermediate specifications, and final specifications obtained during the process of performing step S200. In other words, the results obtained by performing the processing method of the embodiment can be added to the number of samples in Dataset 9 and updated.
[0054] The present invention is not limited to any particular embodiment, and various modifications and changes are possible within the scope of the gist of the invention.
[0055] This disclosure encompasses the following inventions: Invention 1: An unvulcanized rubber processing system comprising a processing device for discharging strip-shaped unvulcanized rubber and a conveying device for transporting the strip-shaped unvulcanized rubber toward the final point of the processing line by the processing device, wherein unvulcanized rubber conforming to a target specification is obtained at the final point, The system includes a sensor that acquires the specifications of the strip-shaped unvulcanized rubber at an intermediate point between the processing apparatus and the final point as an intermediate specification, and a control device that controls the processing conditions in the processing line, The control device is input to a predictive model constructed by machine learning using a dataset of which is a collection of processing conditions, including at least the transport speed of the transport device, and the final specifications of the strip-shaped unvulcanized rubber at the final point, associated with the intermediate specifications. An unvulcanized rubber processing system in which, when the strip-shaped unvulcanized rubber discharged from the processing apparatus is transported toward the final destination, the control device performs data processing to predict the final specifications by inputting the processing conditions and the intermediate specifications acquired by the sensor into the prediction model, and data processing to identify the processing conditions that minimize the difference between the final specifications predicted by this data processing and the target specifications as appropriate processing conditions, thereby controlling the processing conditions so that they become the appropriate processing conditions. Invention 2: The system includes an upstream conveyor that transports the strip-shaped unvulcanized rubber from the processing apparatus to a downstream position including the intermediate point, and a downstream conveyor positioned downstream of the upstream conveyor that transports the strip-shaped unvulcanized rubber to the final point. The processing system for unvulcanized rubber according to Invention 1, wherein the conveying device is the upstream conveying conveyor, and the conveying speed of the downstream conveying conveyor is set to a predetermined value and is not included in the processing conditions. Invention 3: The processing system for unvulcanized rubber according to Invention 1 or 2, wherein the processing conditions include the ambient temperature of the processing line and / or the temperature of the cooling water used to cool the strip of unvulcanized rubber in the processing line. Invention 4: The aforementioned target specification is an unvulcanized rubber processing system according to any one of Inventions 1 to 3, which includes the width dimension, thickness dimension, cross-sectional area, or weight per unit length of the strip-shaped unvulcanized rubber. Invention 5: An unvulcanized rubber processing system according to any one of Inventions 1 to 4, wherein the processing of the strip-shaped unvulcanized rubber by the processing line is performed lot by lot, and the appropriate processing conditions at the time the processing of the immediately preceding lot is completed are used as the initial processing conditions for the immediately following lot. Invention 6: In a method for processing unvulcanized rubber, in which strip-shaped unvulcanized rubber is discharged from a processing device, and the strip-shaped unvulcanized rubber is transported by a conveying device toward the final point of the processing line by the processing device, and unvulcanized rubber conforming to a target specification is obtained at the final point, The specifications of the strip-shaped unvulcanized rubber at intermediate points from the processing apparatus to the final point are taken as intermediate specifications, and the processing conditions in the processing line, including at least the conveying speed of the conveying apparatus, and the final specifications of the strip-shaped unvulcanized rubber at the final point, associated with the intermediate specifications, are used to construct a predictive model using machine learning with multiple accumulated datasets as training data, and input into the control device. A method for processing unvulcanized rubber, wherein when the strip-shaped unvulcanized rubber discharged from the processing apparatus is transported toward the final destination, the control device predicts the final specifications by inputting the processing conditions and the intermediate specifications acquired by a sensor at an intermediate point into the prediction model, identifies the processing conditions that minimize the difference between the predicted final specifications and the target specifications as appropriate processing conditions, and controls the processing conditions so that they become the appropriate processing conditions. [Explanation of symbols]
[0056] 1. Processing System 2 Processing lines 3. Control device 4 Processing equipment 5. Conveying device 5a Upstream conveyor 5b, 5c Downstream conveyor 6 sensors 7 Cooling device 7a Nozzle 8. Confirmation sensor 9 datasets 10 Predictive Models 11. Arithmetic Processing Unit 12 Main memory 13 Auxiliary storage A Final point B Halfway point
Claims
1. An unvulcanized rubber processing system comprising a processing device for discharging strip-shaped unvulcanized rubber and a conveying device for transporting the strip-shaped unvulcanized rubber toward the final point of the processing line by the processing device, wherein unvulcanized rubber conforming to a target specification is obtained at the final point, The system includes a sensor that acquires the specifications of the strip-shaped unvulcanized rubber at an intermediate point between the processing apparatus and the final point as an intermediate specification, and a control device that controls the processing conditions in the processing line, The control device is input to a predictive model constructed by machine learning using a dataset of which is a collection of processing conditions, including at least the transport speed of the transport device, and the final specifications of the strip-shaped unvulcanized rubber at the final point, associated with the intermediate specifications. An unvulcanized rubber processing system in which, when the strip-shaped unvulcanized rubber discharged from the processing apparatus is transported toward the final destination, the control device performs data processing to predict the final specifications by inputting the processing conditions and the intermediate specifications acquired by the sensor into the prediction model, and data processing to identify the processing conditions that minimize the difference between the final specifications predicted by this data processing and the target specifications as appropriate processing conditions, thereby controlling the processing conditions so that they become the appropriate processing conditions.
2. The system includes an upstream conveyor that transports the strip-shaped unvulcanized rubber from the processing apparatus to a downstream position including the intermediate point, and a downstream conveyor positioned downstream of the upstream conveyor that transports the strip-shaped unvulcanized rubber to the final point. The processing system for unvulcanized rubber according to claim 1, wherein the conveying device is the upstream conveying conveyor, and the conveying speed of the downstream conveying conveyor is set to a predetermined value and is not included in the processing conditions.
3. The processing system for unvulcanized rubber according to claim 1 or 2, wherein the processing conditions include the ambient temperature of the processing line and / or the temperature of the cooling water used to cool the strip of unvulcanized rubber in the processing line.
4. The processing system for unvulcanized rubber according to claim 1 or 2, wherein the target specification includes any of the width dimension, thickness dimension, cross-sectional area, or weight per unit length of the strip-shaped unvulcanized rubber.
5. The unvulcanized rubber processing system according to claim 1 or 2, wherein the processing of the strip-shaped unvulcanized rubber by the processing line is performed lot by lot, and the appropriate processing conditions at the time the processing of the immediately preceding lot is completed are used as the initial processing conditions for the immediately following lot.
6. In a method for processing unvulcanized rubber, in which strip-shaped unvulcanized rubber is discharged from a processing device, and the strip-shaped unvulcanized rubber is transported by a conveying device toward the final point of the processing line by the processing device, and unvulcanized rubber conforming to a target specification is obtained at the final point, The specifications of the strip-shaped unvulcanized rubber at intermediate points from the processing apparatus to the final point are taken as intermediate specifications, and the processing conditions in the processing line, including at least the conveying speed of the conveying apparatus, and the final specifications of the strip-shaped unvulcanized rubber at the final point, associated with the intermediate specifications, are used to construct a predictive model using machine learning with multiple accumulated datasets as training data, and input into the control device. A method for processing unvulcanized rubber, wherein when the strip-shaped unvulcanized rubber discharged from the processing apparatus is transported toward the final destination, the control device predicts the final specifications by inputting the processing conditions and the intermediate specifications acquired by a sensor at an intermediate point into the prediction model, identifies the processing conditions that minimize the difference between the predicted final specifications and the target specifications as appropriate processing conditions, and controls the processing conditions so that they become the appropriate processing conditions.
Citation Information
Patent Citations
Apparatus for manufacturing extruded rubber and method for conveying extruded rubber
JP2018183916A
