Method for evaluating kneading state and kneader

Abstract

The purpose of the present invention is to provide a kneading state evaluation method capable of ascertaining a change in physical properties of a kneading material in a sealed kneader having a pair of rotors, and a kneader. This method is for evaluating the kneading state in a sealed kneader 1 comprising a pair of rotors that are coupled by a pair of gears and rotated by an electric motor. In the evaluation method, the temperature of a kneading material, the pressure of a pressure lid in the kneader, the effective value of an AC current supplied to the electric motor, the DC current value supplied to the electric motor, the power consumed by the electric motor, the load factor of the electric motor, the output torque of the electric motor, the number of rotations per minute of the rotor, or the angular velocity of the rotor is repeatedly sampled and acquired. The moving average of the predetermined freely selected time span of the acquired input value, the variance of the freely selected time span, the standard deviation of the freely selected time span, the deviation between the input value and the moving average of the freely selected time span, or values processed from these, or a combination of these, is used as an evaluation index.

Description

Method for evaluating kneading state and kneading machine

[0001] The present invention relates to a method for evaluating the mixing state when mixing materials such as rubber, plastics, and ceramics using a mixer, for example, in a batch system, and to the mixer.

[0002] Conventionally, a closed-type kneader is known as an apparatus for kneading various materials (see, for example, Patent Document 1). In this closed-type kneader, after the materials are charged into a kneading tank, the materials are kneaded by rotating two kneading rotors. Generally, as the kneading rotors of a closed-type kneader, there are known intermeshing rotors in which the two kneading rotors rotate so as to intermesh with each other, and tangential (non-intermeshing) rotors.

[0003] An example of an internal mixer equipped with tangential rotors is shown in Fig. 15. Fig. 15 is a schematic plan view of the mixer. In Fig. 15, an internal mixer 21 has a kneading tank 22, two kneading rotors 23A and 23B arranged side by side in the kneading tank 22, bearings 26A and 26B that rotatably support rotor shafts 25A and 25B of the kneading rotors 23A and 23B, and a pair of gears 27A and 27B. Spiral blades 24a and 24b are formed on the outer peripheries of the kneading rotors 23A and 23B, respectively.

[0004] 15, one of the rotor shafts 25A and 25B (for example, 25A) is connected to a driving means such as a motor, and the other rotor shaft (for example, 25B) is connected to the other rotor shaft via a pair of gears 27A and 27B. By driving the rotor shaft 25A to rotate with the driving means, the kneading rotors 23A and 23B rotate and the material to be kneaded is kneaded. In this case, the kneading rotor 23A connected to the driving means corresponds to the driving rotor, and the kneading rotor 23B corresponds to the driven rotor.

[0005] In a tangential rotor closed-type mixer such as that shown in Fig. 15, a speed difference of about 15% to 25% is generally provided between the drive rotor and the driven rotor by varying the gear ratio of a pair of gears. By rotating these mixer rotors at different speeds, the phases of the drive rotor and the driven rotor change, which is said to result in even mixing.

[0006] Japanese Patent Application Laid-Open No. 9-313916 International Publication No. 2021 / 033390

[0007] Conventionally, in a tangential rotor hermetic mixer, the end of mixing has been determined using indicators such as (1) mixing time, (2) temperature of the material to be mixed, (3) power consumption, or a combination of these. For example, the temperature of the material to be mixed increases as mixing progresses. When the temperature of the material to be mixed is used as an indicator, the timing to end mixing is determined, for example, when the temperature reaches a predetermined temperature. However, the above indicators that have been used conventionally mainly indicate the amount of energy input during mixing, and it is difficult to grasp changes in the physical properties of the material to be mixed during mixing.

[0008] Recently, a system has been proposed for determining the state of an object to be stirred, which includes an acquisition unit that acquires waveform data indicating a waveform related to a current supplied to a drive unit of an agitator having a mechanism for agitating the object and a drive unit for driving the mechanism, and a determination unit that determines the state of the object to be stirred based on changes resulting from a component of a specific direction of a force applied to the drive unit, which are obtained from the waveform data (see Patent Document 2). In Patent Document 2, the current supplied to the drive unit is an AC current having a reference frequency, and this technology uses waveform data of the AC current, i.e., the instantaneous value of the AC current. Specifically, the instantaneous waveform of the AC current, which is a sine wave, is input, and the state of the object to be stirred is determined by focusing on the output component of the reference frequency and other components. Thus, a new method for understanding changes in the physical properties of a material being mixed during mixing is needed.

[0009] The present invention has been made in consideration of the above circumstances, and aims to provide a method for evaluating the mixing state in an internal mixer with a pair of rotors, which can grasp changes in the physical properties of the material to be mixed, and a mixer.

[0010] In one form, the method for evaluating the kneading state of the present invention is a method for evaluating the kneading state in a kneading machine having a pair of rotors connected by a pair of gears and rotated by the drive of an electric motor, and is characterized in that the evaluation method repeatedly samples and acquires the temperature of the material to be kneaded, the pressure of the pressure lid of the kneading machine, the effective value of the AC current supplied to the electric motor, the DC current value supplied to the electric motor, the power value consumed by the electric motor, the load factor of the electric motor, the output torque of the electric motor, the number of rotations per minute of the rotor, or the angular velocity of the rotor, and uses, as evaluation indexes, a moving average of the acquired input values ​​for a predetermined arbitrary time width, a variance of an arbitrary time width, a standard deviation of an arbitrary time width, the deviation between the input value and the moving average of an arbitrary time width, or processed values ​​of these, or a combination of these.

[0011] Here, a series of data in which the input values ​​obtained by the repeated sampling are recorded in the order of sampling time along with the sampling times is called time series data. In particular, the input values ​​obtained by sampling and recorded directly are called original time series data. The evaluation method includes not only a mode in which accumulated data is analyzed and evaluated later, but also a mode in which the kneading machine is monitored in real time. As mentioned above, the phrase "using as an evaluation index" does not only refer to the direct use of the above-mentioned items as evaluation indexes, but also encompasses the use of processed values ​​of the above-mentioned items (e.g., a value obtained by first-order differentiation of a moving average of a predetermined arbitrary time width, a value obtained by second-order or higher order differentiation, a value processed by a low-pass filter, and a combination thereof) as evaluation indexes. Time series data including the above-mentioned processed values ​​is called derived time series data.

[0012] In the kneader, the pair of rotors are tangential rotors, the gears have different integer numbers of teeth, and the pair of rotors rotate at different speeds. The evaluation method is characterized in that the arbitrary time width is set to the period immediately preceding the current time of the input value until the pair of rotors return to the same phase, and the evaluation index is set to the moving average of the period, the variance of the period, the standard deviation of the period, the deviation between the input value and the moving average of the period, or a combination of these.

[0013] In the case where the rotation speed of the pair of rotors in the kneader changes during the course of the kneading, the arbitrary time width is changed in accordance with the change in rotation speed, and a moving average of the changed time width, a variance of the changed time width, a standard deviation of the changed time width, a deviation between the input value and the moving average of the changed time width, or a combination thereof is calculated. For example, the changed time width may be a period from when the rotation speed is changed until one of the pair of rotors rotates a certain number of times.

[0014] In another embodiment, the method for evaluating the kneading state of the present invention is a method for evaluating the kneading state in a kneading machine having a pair of rotors connected by a pair of gears and rotated by the drive of an electric motor, wherein the evaluation method repeatedly samples and acquires the temperature of the material to be kneaded, the pressure of the pressure lid of the kneading machine, the effective value of the AC current supplied to the electric motor, the DC current value supplied to the electric motor, the power value consumed by the electric motor, the load factor of the electric motor, the output torque of the electric motor, the number of rotations per minute of the rotor, or the angular velocity of the rotor, and uses, as evaluation indexes, a moving average of the acquired input values ​​for a predetermined arbitrary number of rotations of one of the rotors (e.g., the drive side rotor), the variance for an arbitrary number of rotations, the standard deviation for an arbitrary number of rotations, the deviation between the input value and the moving average for an arbitrary number of rotations, or processed values ​​of these, or a combination of these.

[0015] The end timing may be determined based on whether the variance, the standard deviation, or the deviation is equal to or less than a threshold value. The end timing may also be determined based on whether the value of the input value at the current time falls within a range of 80% to 120% of the moving average for a predetermined period of time.

[0016] The kneader is characterized in that it is a kneader for kneading non-Newtonian fluids, such as rubber, plastics, ceramics, silicones, and chewing gum compositions.

[0017] In one form, the kneader of the present invention is a kneader having a pair of rotors connected by a pair of gears and rotated by the drive of an electric motor, and is characterized in that the kneader has an acquisition unit that repeatedly samples and acquires the temperature of the material to be kneaded, the pressure of the pressure lid in the kneader, the effective value of the AC current supplied to the electric motor, the DC current value supplied to the electric motor, the power value consumed by the electric motor, the load factor of the electric motor, the output torque of the electric motor, the number of rotations per minute of the rotor, or the angular velocity of the rotor, and a calculation unit that calculates a moving average of a predetermined arbitrary time width, a variance of an arbitrary time width, a standard deviation of an arbitrary time width, a deviation between the input value and the moving average of an arbitrary time width, or processed values ​​of these, or a combination of these, of the input values ​​acquired by the acquisition unit.

[0018] In the kneader, the pair of rotors are tangential rotors, the gears have different integer numbers of teeth, and the pair of tangential rotors rotate at different speeds, and the calculation unit uses the period immediately preceding the current time of the input value until the pair of rotors return to the same phase as the arbitrary time width, and calculates a moving average of the period, a variance of the period, a standard deviation of the period, a deviation between the input value and the moving average of the period, or a combination of these.

[0019] When the rotation speed of the pair of rotors in the kneader changes midway, the calculation unit changes the arbitrary time width in accordance with the change in rotation speed, and calculates a moving average of the changed time width, a variance of the changed time width, a standard deviation of the changed time width, a deviation between the input value and the moving average of the changed time width, or a combination of these.

[0020] In another embodiment, the kneader of the present invention is a kneader having a pair of rotors connected by a pair of gears and rotated by the drive of an electric motor, the kneader having an acquisition unit that repeatedly samples and acquires the temperature of the material to be kneaded, the pressure of the pressure lid in the kneader, the effective value of the AC current supplied to the electric motor, the DC current value supplied to the electric motor, the power value consumed by the electric motor, the load factor of the electric motor, the output torque of the electric motor, the number of rotations per minute of the rotor, or the angular velocity of the rotor, and a calculation unit that calculates a moving average of the input values ​​acquired by the acquisition unit for a predetermined number of rotations of one of the rotors, a variance for the arbitrary number of rotations, a standard deviation for the arbitrary number of rotations, a deviation between the input value and the moving average for the arbitrary number of rotations, or values ​​obtained by processing these, or a combination of these.

[0021] The present invention may further comprise a determination unit that determines the timing of termination based on whether the variance, the standard deviation, or the deviation calculated by the calculation unit is equal to or less than a determination threshold.The present invention may also comprise a determination unit that determines the timing of termination based on whether the value of the input value at the current time falls within a range of 80% to 120% of the moving average calculated by the calculation unit for a predetermined period of time.

[0022] A method for evaluating a mixing state according to one embodiment of the present invention is a method for evaluating a mixing state in a mixer having a pair of rotors connected by a pair of gears and driven by an electric motor. In this evaluation method, the temperature of the material to be mixed, the pressure of a pressure lid of the mixer, the effective value of the AC current supplied to the motor, the DC current supplied to the motor, the power consumed by the motor, the load factor of the motor, the output torque of the motor, the number of rotations per minute of the rotor, or the angular velocity of the rotor are repeatedly sampled and acquired, and the moving average of the acquired input values ​​over a predetermined arbitrary time interval, the variance over an arbitrary time interval, the standard deviation over an arbitrary time interval, the deviation between the input value and the moving average over an arbitrary time interval, or a processed value thereof, or a combination thereof, is used as an evaluation index, thereby enabling accurate understanding of changes in physical properties such as the viscosity, dispersion, and uniformity of the material to be mixed during mixing. For example, by accurately understanding the timing of the end of a batch process, it is possible to achieve effects such as improved stability of physical properties between batches, improved hourly productivity, and reduced energy consumption by preventing excessive energy input.

[0023] In particular, in a kneading machine in which the pair of rotors are tangential rotors and the pair of tangential rotors are configured to rotate at different speeds, by using the moving average of the period until the pair of rotors return to the same phase immediately before the current time of the input value, the variance of the period, the standard deviation of the period, the deviation between the input value and the moving average of the period, or a combination of these as an evaluation index, it becomes easier to more accurately grasp the state of each cycle in which the pair of rotors return to the same phase, which ultimately leads to a better understanding of the changes in the physical properties of the material being kneaded during kneading.

[0024] Another embodiment of the method for evaluating the kneading state of the present invention involves repeatedly sampling and acquiring each parameter, and using as evaluation indices the moving average of the acquired input values ​​for a predetermined number of rotations of one of the rotors, the variance for the predetermined number of rotations, the standard deviation for the predetermined number of rotations, the deviation between the input value and the moving average for the predetermined number of rotations, or values ​​obtained by processing these, or a combination of these, to accurately grasp changes in physical properties such as the viscosity, dispersion, and uniformity of the material being kneaded, and by using the number of rotations as the axis (reference), it becomes easier to accurately grasp the state of each cycle in which the pair of rotors return to the same phase, which ultimately leads to a good understanding of changes in the physical properties of the material being kneaded. It can also deal with changes in the rotor rotation speed.

[0025] The timing to end kneading is determined based on whether the variance, standard deviation, or deviation is below the judgment threshold, i.e., whether the variation in input values ​​(such as temperature and power values) is small, so that the timing to end kneading can be determined quickly and accurately.

[0026] FIG. 1 is an explanatory diagram showing a schematic configuration of the entire kneader according to the present invention. FIG. 2 is a diagram for explaining the driving of the kneading rotor of the kneader of FIG. 1. FIG. 3 is a diagram showing an example of the number of teeth of a pair of gears. FIG. 4 is a flowchart for explaining an example of processing executed by a control unit. FIG. 5 is an image diagram of calculation of a simple moving average SMA. FIG. 6 is a graph showing the temporal fluctuation of each parameter in the prior art. FIG. 7 is a graph showing the temporal fluctuation of each parameter in the present invention. FIG. 8 is a graph showing the temporal fluctuation of each parameter in the present invention. FIG. 9 is a graph showing the temporal fluctuation of power values ​​after processing by another moving average. FIG. 10 is a graph showing the temporal fluctuation of power values ​​after processing by a low-pass filter. FIG. 11 is a graph in which each smoothing process is superimposed. FIG. 12 is a graph showing the temporal fluctuation of each parameter in the present invention and the prior art. FIG. 13 is a graph showing an example of obtaining extreme values ​​(maximum values ​​and minimum values). FIG. 14 is a graph showing an example of obtaining an inflection point. FIG. 15 is a diagram for explaining the driving of the kneading rotor of a conventional kneader.

[0027] The kneader used in the method for evaluating the kneaded state of the present invention is a closed-type kneader for kneading non-Newtonian fluids including rubber, plastic, etc. Fig. 1 is an explanatory diagram showing a schematic overall configuration of the kneader according to the present invention, and mainly shows a schematic cross-sectional view of the kneading tank located at the lower end of the kneader.

[0028] As shown in Figure 1, the closed-type mixer 1 mainly comprises a mixer mechanism including a mixer tank 2 and an electric motor 8 (see Figure 2), a pressurizing mechanism including a pressurizing lid 9a that applies pressure to the material to be mixed introduced into the mixer tank, and a control unit 12 that controls the mixing of the material to be mixed. The evaluation method and mixer of the present invention are characterized in that they repeatedly sample and acquire mixer parameters detected by a sensor or the like, and use calculated values ​​obtained by performing predetermined arithmetic processing on the acquired input values ​​as evaluation indices. Specifically, parameters indicating the dispersion of the input values, such as the mean value, variance, and standard deviation, over a predetermined arbitrary time span or arbitrary number of rotor rotations are used as evaluation indices.

[0029] The kneading tank 2 has an inner peripheral surface shaped like two roughly C-shaped partial peripheral surfaces connected together facing each other, and contains two adjacent and communicating rotor chambers 2A, 2B. A mountain-shaped, rising edge wall 2C is formed at the boundary between the inner peripheral surfaces of the rotor chambers 2A, 2B at the inner bottom of the kneading tank 2. Both axial ends of each rotor chamber 2A, 2B are closed by respective tank end walls (not shown). The cross-sectional shape of the rotor chambers 2A, 2B is constant in the axial direction.

[0030] The closed-type mixer 1 has a temperature sensor 10 that detects the temperature inside the mixer tank during mixing. The temperature sensor 10 is arranged with its detection tip 10a protruding from the upper surface of the rear wall portion 2C, and is capable of detecting the temperature of the material being mixed that comes into contact with the detection tip 10a. A well-known temperature sensor is used as the temperature sensor 10, for example, a thermocouple temperature detector that measures temperature by housing a thermocouple element in a protective tube. This thermocouple temperature detector may be, for example, a grounded type in which the tip of the thermocouple element is welded to the tip of the protective tube and the temperature of the material being mixed is detected on the outer surface of the protective tube, or an ungrounded type in which the thermocouple element is insulated from the protective tube and the temperature inside the protective tube is detected.

[0031] The temperature sensor provided in the internal mixer 1 is not limited to the sensor configuration or arrangement shown in Fig. 1 as long as it detects the temperature of the material to be mixed during mixing. Temperature changes in the material to be mixed detected by the temperature sensor will be described later.

[0032] As shown in FIG. 1 , kneading rotors 3A and 3B for kneading the material are rotatably disposed within the rotor chambers 2A and 2B, respectively, at a distance from the inner circumferential surfaces of the rotor chambers 2A and 2B. Each kneading rotor 3A and 3B includes a plurality of blades 4a and 4b (two blades in FIG. 1 ). The blades 4a and 4b each have a mountain-shaped cross section extending from their starting end to their terminal end, with a land portion at the apex. The land portion rotates while maintaining a predetermined distance from the inner circumferential surface. The rotation of the kneading rotors 3A and 3B changes the shape of the kneading space in the rotor chambers 2A and 2B. The kneading rotors 3A and 3B rotate in opposite directions, with the blade rotating downward on the side where the rotor chambers 2A and 2B communicate.

[0033] An opening is provided above the kneading tank 2 for introducing the material to be kneaded. The pressurizing lid 9a can be moved up and down by a cylinder device or the like, and the material to be kneaded is introduced through the opening while the pressurizing lid 9a is raised. The pressurizing lid 9a is then lowered by the rod 9b, and the two kneading rotors 3A, 3B are rotated while pressurizing the material to be kneaded. In this case, the spiral blades 4a, 4b knead the material to be kneaded by flowing it in complex directions, including not only the rotational direction of the rotors but also the axial direction.

[0034] The kneader according to the present invention is not limited to the configuration shown in Fig. 1. The closed kneader 1 in Fig. 1 is of a type in which, after kneading, the kneading tank 2 is inverted and the material to be kneaded is taken out from an opening, but the kneader may be of a type in which, after kneading, the material to be kneaded is taken out from the bottom of the kneading tank. Furthermore, the closed kneader 1 in Fig. 1 is of a type called a tangential type in which the rotating areas of the kneading rotors 3A and 3B do not overlap, but it may also be of a type called an intermeshing type in which the kneading rotors 3A and 3B rotate at the same speed and their rotating areas overlap.

[0035] FIG. 2 shows a schematic plan view of an internal mixer. As shown in FIG. 2, the internal mixer 1 includes the above-described mixing tank 2, mixing rotors 3A and 3B, bearings 6A and 6B that rotatably support rotor shafts 5A and 5B, and a pair of gears 7A and 7B. Helical blades 4a and 4b are formed on the outer peripheries of the mixing rotors 3A and 3B. For example, in the mixing rotor 3A, the blades 4a and 4b have their respective starting ends at positions on both axial ends of the mixing rotor 3A that are 180° out of phase with each other in the circumferential direction, and extend from these starting ends in a helical direction around the outer periphery of the mixing rotor 3A. The configuration of the blades in the mixing rotor is not limited to this. In FIG. 2, the blades 4a and 4b have different lengths in the helical direction, with the blade 4a being a long blade and the blade 4b being a short blade. The number of blades is not limited to two, and three, four, six, etc. may also be used. In this case, for example, the positions of the starting ends between the blades (phase in the circumferential direction) are set appropriately depending on the number of blades.

[0036] In the closed-type kneader 1, the rotor shafts 5A and 5B of the two kneading rotors 3A and 3B are arranged in parallel. The rotor shaft 5A is connected to the output shaft 5A' of the electric motor 8 via a coupling 17. Note that the coupling 17 may be omitted and the rotor shafts 5A and 5A' may be configured as an integral unit. On the other hand, the rotor shaft 5B is connected to the rotor shaft 5A via a pair of gears 7A and 7B. The electric motor 8 has a circuit section 8a and a motor section 8b. The circuit section 8a generates electric power based on a control signal and supplies the generated electric power to the motor section 8b. The electric motor 8 has a power sensor 11 that detects the electric power supplied to the motor section 8b. In the present invention, the electric motor may be an AC motor driven by an AC power source or a DC motor driven by a DC power source.

[0037] The electric motor 8 may be equipped with a speed reducer, and may be configured to reduce the rotational force generated by the drive source before outputting it. The pair of gears 7A, 7B is not limited to gears provided externally to the electric motor 8 as shown in Fig. 2, but may be gears built into the electric motor 8 or the speed reducer. The gear configuration is not limited to spur gears, and may be helical gears, etc. In the kneader, each kneading rotor may be connected to each gear via a coupling.

[0038] In the configuration shown in Fig. 2, driving the electric motor 8 rotates the rotor shafts 5A and 5B, thereby rotating the kneading rotors 3A and 3B, and kneading is performed. In Fig. 2, the closed-type kneader 1 has a configuration in which the pair of kneading rotors 3A and 3B rotate at different speeds. The number of teeth of the pair of gears 7A and 7B is not particularly limited, but a combination of different numbers of teeth that are not prime integers is more effective because it shortens the time until the phases are restored to the same (see Fig. 3, for example). In Fig. 2, the kneading rotor 3A corresponds to the drive rotor, and the kneading rotor 3B corresponds to the driven rotor.

[0039] Furthermore, in a configuration in which the pair of kneading rotors 3A, 3B rotate at different speeds, the combination of the number of teeth is not particularly limited, but for example, it is preferable that the number of teeth of the gear with the larger number of teeth is 10% to 50% larger than the number of teeth of the other gear. Furthermore, if the number of teeth of the pair of gears is such that the high-speed rotor and the low-speed rotor (driven rotor) return to the same phase within 10 rotations of the high-speed rotor (drive rotor), then by using this period as the time for moving averages, the convergence of variance and standard deviation can be determined more clearly.

[0040] Returning to Fig. 1, the control unit 12 is mainly composed of a microcomputer including a well-known CPU, ROM, RAM, etc. The sensors 10, 11 provided in the closed-type kneader 1 and detecting kneading parameters are connected to the control unit 12. During kneading, the kneading parameters, such as the temperature of the material to be kneaded, the pressure of the pressurizing lid of the kneader, the effective value of the AC current supplied to the motor, the DC current value supplied to the motor, the power consumed by the motor, the load factor of the motor, or the output torque of the motor, are repeatedly sampled and acquired by the control unit 12 and stored. The control unit 12 also has various calculation functions.

[0041] For example, the temperature of the material to be kneaded is detected by a temperature sensor 10 as shown in Fig. 1, and the detection signal is acquired by the control unit 12. The pressure of the pressurizing lid 9a in the closed-type kneader 1 is a pressure load due to the weight pressure of the pressurizing lid 9a that suppresses the material to be kneaded from above the kneading tank 2, and is detected by an arbitrary sensor, and the detection signal is acquired by the control unit 12.

[0042] If the motor is an AC motor, the effective value of the AC current supplied to the motor is calculated based on the detection signal of the current sensor. For example, the effective value of the AC current can be calculated by dividing the maximum instantaneous value by √2. If the motor is a DC motor, the value of the DC current supplied to the motor is detected based on the detection signal of the current sensor.

[0043] The power value consumed by the electric motor is detected by a power sensor 11 provided on the electric motor 8 as shown in Fig. 2, for example, and the detection signal is acquired by the control unit. For example, if the electric motor is an AC motor, the power value consumed by the electric motor is active power P, which is expressed by the following equation (1): Active power P = V I cos θ (1) In the above equation (1), V is the effective value of the voltage applied to the electric motor, I is the effective value of the AC current supplied to the electric motor, θ is the phase difference between the voltage and the current, and cos θ is the power factor. Furthermore, if the electric motor is a DC motor, the power value consumed by the electric motor is expressed as the product of the voltage V applied to the electric motor and the DC current I supplied to the electric motor.

[0044] The load factor of the motor is expressed by the following formula (2) based on the power consumed by the motor and the rated value of the motor: Load factor (%) = [power consumed by motor (W) / rated value of motor (W)] × 100 (2)

[0045] The output torque of the electric motor is detected, for example, by a torque sensor provided in the electric motor. Alternatively, in a means that does not rely on a torque sensor, the output torque is calculated, for example, based on the power value consumed by the electric motor and the rotation speed or angular velocity of the electric motor. Well-known sensors can be used for each sensor for acquiring the above-mentioned kneading parameters.

[0046] Although FIG. 1 shows a configuration in which the temperature of the material to be kneaded and the value of the power consumed by the electric motor are acquired as kneading parameters, the present invention is not limited to this.

[0047] As shown in Fig. 1, the control unit 12 includes an acquisition unit 13 that repeatedly samples and acquires at least one of the kneading parameters described above. Also, in Fig. 1, the control unit 12 includes a calculation unit 14 that calculates, as evaluation indices, a moving average of a predetermined arbitrary time width, a variance of the arbitrary time width, a standard deviation of the arbitrary time width, a deviation between the input value and the moving average of the arbitrary time width, or a processed value of any of these, or a combination of these, for the input value acquired by the acquisition unit 13. For example, the calculation unit 14 calculates, for the input value, (A) a moving average of a predetermined arbitrary time width immediately before the current time, (B) a variance of a predetermined arbitrary time width immediately before the current time, (C) a standard deviation of a predetermined arbitrary time width immediately before the current time, (D) a deviation between the input value and the moving average of a predetermined arbitrary time width immediately before the current time, or a processed value of (A) to (D), or a combination of (A) to (D) and the processed values ​​thereof.

[0048] Furthermore, when the kneader is configured such that a pair of tangential rotors rotate at different speeds, it is preferable that the calculation unit 14 calculates, for the input values, (A1) a moving average of the period immediately preceding the current time until the kneading rotors 3A and 3B return to the same phase, (B1) a variance of the period immediately preceding the current time until the kneading rotors 3A and 3B return to the same phase, (C1) a standard deviation of the period immediately preceding the current time until the kneading rotors 3A and 3B return to the same phase, (D1) a deviation between the input value and the moving average of the period immediately preceding the current time until the kneading rotors 3A and 3B return to the same phase, or a value obtained by processing (A1) to (D1), or a combination of (A1) to (D1) and the values ​​obtained by processing these. 7, the period until the kneading rotors 3A and 3B return to the same phase is used as the predetermined arbitrary time width, so that the state of each cycle from a certain phase until the kneading rotors 3A and 3B return to the same phase can be accurately grasped, and as a result, it becomes easier to grasp the change in the physical properties of the material to be kneaded. This makes it possible, for example, to accurately determine the timing when kneading is to be completed.

[0049] The moving average can be calculated using a known moving average calculation method such as a simple moving average (SMA), a weighted moving average (WMA), an exponential moving average (EMA), or a centered moving average (CMA).

[0050] As an example, Figure 3 shows a pair of gears in which the drive gear has 25 teeth and the driven gear has 30 teeth. In this case, the drive rotor connected to the drive gear rotates six times (the driven rotor rotates five times), and the drive rotor and the driven rotor return to the same phase. Figure 3 shows a view of the pair of rotor shafts from the gear side, showing the state in which the pair of gears mesh. For convenience, in Figure 3, circled numbers indicate the positions of the valleys of the driven gear where the teeth of the drive gear mesh.

[0051] 1, the control unit 12 is configured to be able to evaluate changes in the kneading state based on the evaluation index calculated by the calculation unit 14. Specifically, the control unit 12 has a determination unit 15 that determines the timing to end the kneading based on the evaluation index calculated by the calculation unit, and a notification unit 16 that notifies the determination unit 15 that it is time to end the kneading when determined by the determination unit 15.

[0052] The determination unit 15 determines the end timing based on the moving average, variance, standard deviation, and deviation calculated by the calculation unit 14. The method for this determination is not particularly limited, and the method described in FIG. 4 or the like can be used.

[0053] The notification unit 16 has a function of notifying the completion of kneading. The notification means is not particularly limited, and one or a combination of the following means can be used: displaying the completion to the operator on a monitor, notifying by sound or voice, notifying by communication to the outside, notifying by a lamp display, etc.

[0054] Next, an example of a processing procedure during kneading executed by the control unit is shown in Fig. 4. In the flowchart of Fig. 4, the processing from start to end is repeated at predetermined time intervals. Note that, although the above-mentioned kneading parameters can be used as appropriate, the following explanation will be given of an example using power values.

[0055] First, the power value detected by the power sensor is acquired by the acquisition unit at predetermined sampling intervals (e.g., every 0.1 to 1 second) (step S11). The acquired power value is stored in the storage unit each time it is acquired.

[0056] In step S12, the calculation unit calculates, for example, a moving average of power values ​​for n seconds immediately preceding the current time based on the acquired power values. FIG. 5 shows an image of calculating a simple moving average (SMA) as an example of this moving average. In FIG. 5, in a configuration in which power values ​​are acquired every 0.2 seconds, for example, the calculation unit calculates a simple moving average for 5 seconds immediately preceding the current time (5-second simple moving average). Note that the simple moving average is calculated in a manner that includes the input value for the current time. For example, SMA1 is calculated at current time 1, SMA2 is calculated at current time 2 (1 second after current time 1), and SMA3 is calculated at current time 3 (1 second after current time 2).

[0057] In step S13, the calculation unit calculates the standard deviation SD of the power values ​​for, for example, the last n seconds from the current time based on the acquired power values. Note that the last n seconds from the current time is considered to be the same as in FIG. 5 .

[0058] In step S14, the determination unit determines the timing to end the kneading. The determination of the timing to end the kneading is performed, for example, based on the calculated standard deviation SD. Specifically, it is determined whether the standard deviation SD is equal to or less than the determination threshold. If the standard deviation SD is not equal to or less than the determination threshold, it is determined that it is not the time to end the kneading (step S14: No), and the process ends. On the other hand, if the standard deviation SD is equal to or less than the determination threshold, it is determined that it is the time to end the kneading (step S14: Yes). Here, the determination threshold is a value at which the variation in power value is considered to have converged, and is set appropriately through experiments, etc. Furthermore, the determination threshold may be set to a different value depending on the kneading parameters used.

[0059] If it is determined in step S14 that it is time to end the kneading, for example, the supply of driving power to the motor unit is stopped. Also, if necessary, the notification unit notifies the operator that the kneading has ended (step S15).

[0060] As described above, the processing procedure in FIG. 4 combines a moving average with a standard deviation over the same period. The method for determining the end timing is not limited to the above method. For example, the variance of the power values ​​over the last n seconds from the current time may be used, and the variance may be compared with a judgment threshold. If the variance is equal to or less than the judgment threshold, it may be determined that the end timing has arrived. The variance is calculated as the square of the standard deviation. The moving average MA itself may also be used as the judgment index. In this case, the end timing may be determined based on whether the input value at the current time falls within a range of 80% to 120% (preferably 90% to 110%) of the moving average for a predetermined period of time.

[0061] In addition to the level of the kneading parameters, the above judgment may be made based on the level of the moving average MA of the calculated kneading parameters, the first derivative of the moving average MA and its change, the maximum and minimum of the moving average MA, the standard deviation SD, or a combination of these.

[0062] Next, the conventional technology and the present invention will be compared using specific graphs in Figures 6 to 8. First, Figure 6 illustrates the conventional technology. Figure 6 also illustrates the temporal fluctuations of various parameters when a rubber composition is mixed as a mixing material using a tangential-type internal mixer having a pair of gears, with the drive gear having 25 teeth m and the driven gear having 30 teeth n. In this example, mixing is performed with the drive rotor rotating at 36 rpm and the driven rotor rotating at 30 rpm, and the drive rotor and the driven rotor return to the same phase every 10 seconds. Here, the temperature in the figure indicates the temperature of the mixing material (continuous value) detected by the temperature sensor, the power indicates the power consumed by the electric motor (specifically, the active power of an AC electric motor) detected by the power sensor, and the torque indicates the torque of the electric motor (continuous value) detected by the torque sensor.

[0063] In Figure 6, the kneading parameters were monitored using a conventional method, and it can be seen that the values ​​of each kneading parameter fluctuate in a discrete zigzag pattern. In particular, the power fluctuates significantly in a zigzag pattern. From such a graph, it is relatively difficult to grasp the changes in the state of the kneaded material over time.

[0064] In contrast, Figures 7 and 8 show graphs to which the method of the present invention is applied. Specifically, these graphs visualize the calculated simple moving average (SMA) of power values ​​over a predetermined time interval and the standard deviation (2σ, -2σ) of power values ​​over the same time interval, in addition to the temporal fluctuations of the original kneading parameters. Note that the time interval for calculating the simple moving average SMA and the like is different between Figures 7 and 8. That is, in Figure 7, the time interval for calculating the simple moving average SMA and the standard deviation is set to 10 seconds, while in Figure 8, it is set to 17 seconds. In Figure 7, the time interval for the calculation is the period until the pair of rotors return to the same phase. On the other hand, in Figure 8, the time interval is set arbitrarily.

[0065] In FIG. 7, the simple moving average W ＳＭＡ was calculated as a simple moving average of the most recent 10 seconds from the current time. When the number of samples for the most recent 10 seconds is n, it was calculated using the following formula (3).

[0066] In addition, the standard deviation W of the power values ​​sampled n times in the last 10 seconds of the current time is ＳＤ was calculated using the following formula (4).

[0067]

[0068] As shown in Figure 7, the standard deviations (power (SD; 2δ), power (SD; -2δ)) can be plotted on both the positive and negative sides of the 10-second simple moving average (power (SMA)) of power. Figure 7 shows that the bandwidths of the positive and negative standard deviations converge significantly around the time period just before 15:08. This indicates that the variation in the power consumed by the motor converges significantly from this time period, indicating that the progress of kneading has led to the overall state of the kneaded material becoming homogenous, for example, by the progress of mixing and dispersion. This state change can be used as a direct indicator of the end of kneading, or as an indicator of indirect end timing, such as by the passage of a certain amount of time after this state change or the input of a certain amount of energy. Therefore, by determining when the standard deviation is below a threshold, the convergence of the bandwidths shown in Figure 7 can be determined, which ultimately leads to the determination of the end timing. In Figure 7, the fluctuating power values ​​are smoothed using a simple moving average, and the variation (standard deviation) is further visualized.

[0069] In addition, in Fig. 8, the time width for calculating the simple moving average SMA and standard deviation is set to 17 seconds. As shown in Fig. 8, the bandwidth of the standard deviation converges toward the end of the kneading process, and even in this case, the end timing can be determined by comparing the standard deviation with a determination threshold. Note that the behavior of the convergence of the standard deviation is gentler than in Fig. 7.

[0070] 7 and 8, it can be said that it is easier to grasp the state change by setting the time period until the pair of rotors return to the same phase as the time width for calculating the moving average, standard deviation, etc. By utilizing the characteristic cycle in such a configuration in which the rotors rotate at different speeds in the calculation process, it becomes easier to grasp the change in the physical properties of the kneaded material, and ultimately the end timing can be accurately determined.

[0071] On the other hand, in conventional methods, for example, the end temperature of kneading is set to a predetermined temperature (e.g., 120°C), and kneading is terminated when the temperature of the material to be kneaded reaches that temperature. Since the temperature of the material to be kneaded varies depending on the heat capacity of the raw materials, the amount of energy input, and the amount of energy lost through heat removal and heat dissipation, the determination of the end of kneading based on the temperature level can be said to be a judgment criterion based on the integration of the amount of energy input from the start of kneading. In contrast, the above-mentioned method of the present invention acquires kneading parameters such as the power value of the electric motor in real time, and performs kneading while determining the trend of change in the kneading parameters, the convergence of variation, etc. using the moving average and standard deviation over the same period, thereby providing a means for understanding changes in the physical properties of the material to be kneaded during kneading, and therefore kneading can be terminated quickly and at an appropriate time, enabling efficient kneading processing.

[0072] 7 and 8 show the results using the power consumption value of the electric motor as the kneading parameter, but it is also possible to grasp the change in the physical properties of the kneaded material using other kneading parameters. For example, the effective value of the AC current supplied to the electric motor, the DC current value supplied to the electric motor, the load factor of the electric motor, and the output torque of the electric motor are parameters closely related to the power consumption value of the electric motor, and similar results can be obtained.

[0073] Although a simple moving average is used as the smoothing process in the above, a moving average calculated by other moving average methods may also be used. The results of each process are shown below using a power chart.

[0074] First, the power chart in Figure 9(a) shows a simple moving average (SMA), while the other Figures 9(b) to 9(d) show a comparison with the simple moving average. In each power chart shown in Figure 9, the horizontal axis represents the rotation speed of the drive rotor. In this case, the drive rotor rotates six times (the driven rotor rotates five times) until the drive rotor and the driven rotor return to the same phase. The time span for calculating the moving average is the period until the pair of rotors return to the same phase, and the moving average is calculated over the time it takes for the drive rotor to rotate six times.

[0075] The weighted moving average (WMA) in Figure 9(b) is a linear weighting method that gradually reduces the weight from the present to the past, making it easier to respond to fluctuations. As shown in Figure 9(b), the WMA has better tracking ability for raw time series data than the SMA. The WMA is calculated using the following equations (5) and (6).

[0076] The exponential moving average (EMA) in Figure 9(c) is a method of assigning weights that decrease exponentially. As shown in Figure 9(c), the EMA has better tracking ability for raw time series data than the SMA. However, the amplitude is relatively large. For example, the exponential moving average S at time t ｔ is calculated by the following formula (7): The initial value is calculated as a simple moving average of the previous raw time series data, and the EMA is calculated recursively.

[0077] The centered moving average (CMA) in Figure 9(d) is a method of taking the average of N / 2 data points in the past and N / 2 data points in the future. As shown in Figure 9(d), the CMA has better tracking ability for raw time series data than the SMA.

[0078] Fig. 10 shows an example of smoothing processing using a low-pass filter. For example, high-frequency components are removed by performing a discrete Fourier transform (see Fig. 10(a)). As shown in Fig. 10(b), good tracking is shown in the middle of the kneading processing, but fitting is not possible immediately after the start or end of the kneading processing.

[0079] Figure 11 shows a chart overlaid with the results of each smoothing process. As shown in Figure 11, the smoothing process method has different levels of tracking to fluctuations, with EMA, WMA, and SMA (CMA) showing the strongest tracking in this order. Table 1 shows the results of evaluating these methods using mean square error.

[0080]

[0081] As shown in Table 1, although vibrations remain in the EMA, it fits the raw time series data well with good tracking. Also, CMA is considered suitable for predicting the final torque (power), final temperature, and amplitude after removing the vibrations.

[0082] Next, another form of evaluation method using the moving average MA will be described. A conventional method for determining changes in the kneading state is, for example, a method of visually evaluating the trend of changes in a power chart. For example, the time from adding compounding ingredients such as carbon black to integrating them is widely recognized and emphasized in the rubber industry as the Black Incorporation Time (BIT). However, as shown in FIG. 12( a), an actual power curve is a discrete zigzag broken line, and the point within the range indicated by the cloud in the figure that is considered to be a peak (maximum or minimum) depends on the subjective judgment of the individual. Therefore, when attempting to evaluate an actual power curve using conventional techniques, the power value changes significantly over a short period of time, making it difficult to clearly grasp the position of the peak or the trend of change.

[0083] The new evaluation method of the present invention can solve the above-mentioned problems. Figure 12(b) shows an example of applying the present invention, in which a moving average MA is used to identify a local maximum.

[0084] The specific procedure is as follows: (1) As described above, the moving average MA for a recent arbitrary time span is calculated. (2) The trend is grasped by the first derivative of the moving average MA, the slope of the moving average MA, or the distribution of the kneading parameters relative to the moving average MA. (3) The local maximum and minimum are identified by comparing the first and second derivatives of the moving average MA, or consecutive values ​​of the moving average MA. For example, a point where the first derivative of the moving average MA changes from positive to negative, or a point where the first derivative is zero and the second derivative is negative, can be identified as a local maximum of the original kneading parameters. Also, a point where the first derivative of the moving average MA changes from negative to positive, or a point where the first derivative is zero and the second derivative is positive, can be identified as a local minimum. The means for identifying local maximums and minimums using the moving average MA may include methods using the first derivative, higher-order derivatives (second derivative or higher), or means for identifying by distribution or comparison of current and previous values.

[0085] Furthermore, (4) the variance and standard deviation (SD) may be calculated from the distribution of the mixing parameters for the target period for which the moving average MA was calculated, and the changes in these may be observed. In Fig. 12(b), the moving average MA and the standard deviation SD are displayed and visualized together with the mixing parameters.

[0086] FIG. 13 shows a specific method for obtaining BIT and Black Wetting Time (BWT) as extreme values. BIT is the point at which the torque reaches its maximum value when carbon is embedded in the rubber, the rubbers adhere to each other, and the rubber becomes a large rubber mass and integrates. BWT is the point at which the torque reaches its minimum value when the finely divided polymer is coated with carbon black, making it slippery. FIG. 13(a) shows a method for obtaining extreme values ​​by differentiating a function obtained by applying a low-pass filter to EMA. Applying a low-pass filter after obtaining EMA can improve smoothing (see Table 1). Furthermore, because a low-pass filter is obtained as a differentiable function, it is suitable for differentiation to obtain extreme values. FIG. 13(b) shows a method for obtaining extreme values ​​by applying the SciPy find_peaks function to CMA. For example, when obtaining a local maximum value, find_peaks considers a point higher than both neighboring points as the local maximum value.

[0087] 12 and 13, smoothing using the moving average MA and further combining it with other processes can identify local maximums and minimums. It is also believed that identifying the local maximum can lead to identifying the BIT, and identifying the local minimum can lead to identifying the BWT. It is also possible to evaluate the elapsed time from the start of kneading to reaching the local maximum, and the elapsed time from reaching the local maximum to the convergence of the standard deviation SD.

[0088] As another form of evaluation method using the moving average MA, it is also possible to grasp the point (inflection point) at which the slope of the kneading parameter changes at the end of the kneading process. After the inflection point, the rubber becomes soft due to the influence of temperature, and further kneading cannot be effective. Therefore, by grasping the inflection point, the end of rubber kneading can be determined. For example, the inflection point can be obtained by second-order differentiation of a function obtained by applying a low-pass filter to the EMA. For example, in Figure 14, it can be determined that a specific inflection point among the circles shown on the graph indicates the end point.

[0089] FIG. 14 shows 12 inflection points. Between the extreme values ​​(maximum and minimum values) shown in FIG. 13(a), there are always an odd number of inflection points, including one in the case where there is only one, and the odd-numbered inflection point after the third maximum value in FIG. 13(a) is a candidate for the end point. Which maximum value to focus on is arbitrarily set depending on the blend and kneading pattern to be kneaded. For this determination, a threshold value is set to determine the vicinity of 0 (zero) for the second-order differentiated value, and when the second-order differentiated value after the odd-numbered inflection point after the maximum value exceeds the threshold on the positive side near 0 (zero), it can be determined that the inflection point immediately before that represents the end point. For example, in FIG. 14, it can be determined that the inflection point shown around 100 rotations on the horizontal axis represents the end point.

[0090] From the time the material is added until the completion of mixing, it passes through the stages of "pulverization and adhesion," "integration," and "mixing and homogenization," and finally a homogenized mixed material is obtained. As described above, for example, by smoothing using a moving average MA, it is possible to grasp the maximum and minimum values, inflection points, etc., and accurately grasp the state of each stage. As a result, it is believed that stable mixing can be performed even with different batches and conditions.

[0091] In this specification, population variance and population standard deviation are given as examples of application of the technology, but the present invention is not limited to these, and sample variance, sample standard deviation, sample unbiased variance, and sample unbiased standard deviation may also be used.

[0092] According to the present invention described above, for example, the following effects can be obtained: The trend of the blending parameters can be clearly determined from the slope of the moving average MA. The maximum and minimum values ​​of the blending parameters can be clearly determined. There is no influence from human subjectivity. The distribution of the blending parameters relative to the moving average MA can be determined as the standard deviation SD, and the magnitude of the fluctuation range can be determined. The convergence point and convergence trend of the standard deviation SD can be used to determine changes in the blending state. The bias in the distribution of the blending parameters relative to the moving average MA can be used to predict future trends in changes in the blending parameters. For example, in FIG. 12(b), it can be seen that, after the convergence point (see FIG. 7), the plotted power points are concentrated below the moving average MA, with almost no distribution above the moving average MA. Therefore, by understanding this bias in the distribution relative to the moving average, it is possible to predict the trend in the change in the moving average, that is, the moving average MA will continue to decline.

[0093] Trends can also be determined using short-term and long-term moving averages. Each moving average can be calculated using the method described above. For example, the long-term moving average is calculated over an interval that is an integer multiple (e.g., 2 to 5 times) of the interval over which the short-term moving average is calculated. If the short-term moving average is larger than the long-term moving average, it can be determined that the trend is rising, and if the long-term moving average is larger than the short-term moving average, it can be determined that the trend is falling.

[0094] The specific configuration of the evaluation method and kneader of the present invention is not limited to the configuration shown in the above figures, and can be modified as appropriate.

[0095] While the examples shown in the above figures illustrate the case where the rotational speed of the drive rotor is constant, the present invention can also be applied, for example, when the rotational speed (= r) of the drive rotor changes during the process (e.g., from r1 to r2). In this case, the time required for one rotation of the drive rotor also changes as the rotational speed changes. Therefore, even after the rotational speed changes, the time span for calculation may be changed to correspond to the period after the change so that the moving average can continue to be calculated over the same number of rotations of the drive rotor. For example, before the change in rotational speed, the moving average was calculated over 10 seconds (a period during which the drive rotor rotates six times). After the rotational speed is reduced to two-thirds, the moving average may be calculated over 15 seconds (a period during which the drive rotor rotates six times), thereby changing the time span for calculation. This makes it easier to continuously and accurately grasp the kneading state even when the rotational speed of the pair of rotors changes during the process.

[0096] In addition, in response to changes in the rotation speed, time series data with the sampling time as an index can be converted into rotation speed series data or rotation angle series data with the rotation speed, rotation angle, or angular distance as an index, and used. The time series data obtained by sampling includes the sampling time and the rotor rotation speed (r / min) or angular velocity (rad / s or ° / s). ｎ and sampling time t ｎ＋１ By integrating this displacement angle, the number of rotations (r or rotations) or the rotation angle or angular distance (rad or °) at each sampling time when the rotor has rotated from the initial state can be calculated, and new data can be derived and added to the time series data.

[0097] From this derived time series data, it is possible to use rotation number series data or rotation angle series data in which the rotation number, rotation angle, or angular distance is used as an index instead of the sampling time. Here, the series that serves as an index will be called an index series.

[0098] For example, by using rotation speed series data with the rotation speed as an index, it is possible to find the moving average and standard deviation of values ​​in a series other than the index series over an arbitrary rotation speed period, such as six rotations of the drive rotor. In this way, by deriving rotation speed series data or rotation angle series data with the rotation speed, rotation angle, or angular distance as an index from time series data, it is possible to obtain the same effect as changing the time width for calculating the moving average based on changes in the rotation speed.

[0099] Whether it is time series data, or rotation speed series data or rotation angle series data in which the rotation speed or rotation angle is used as an index, the original data can be resampled and used so that the index series values ​​are at any regular interval. For example, in the case of rotation speed series data in which the index is the rotation speed, resampling can be performed at regular intervals of 1 / 6 rotation. In this case, for series other than the index series, values ​​corresponding to the index series can be obtained from the original rotation speed series data using linear interpolation, first-order spline interpolation, second-order spline interpolation, or third-order spline interpolation, and new rotation speed series data can be derived.

[0100] The resampling of the index sequence values ​​to a fixed interval is performed before performing a discrete Fourier transform to evaluate the periodic fluctuations of any data sequence or the spectrum of the signal, or before applying a low-pass filter.

[0101] FIG. 13( a) shows an example in which, after conversion to rotation speed series data, an exponential moving average (EMA) is calculated, and the rotation speed values, which are the index series, are resampled at regular intervals, and then a low-pass filter is applied, with the horizontal axis representing the rotation speed.

[0102] Although the above describes a method for evaluating kneading in real time, the present invention may also be applied to the evaluation by analyzing accumulated data later. Examples of a method for analyzing later include a method for calculating a moving average MA and a standard deviation SD based on the kneading parameters collected and recorded, and visualizing and displaying the convergence timing by overlaying the kneading parameters with the moving average MA and the standard deviation SD on a graph (see Figures 7 and 8), and a method for comparing two different batches by displaying the same data.

[0103] The calculation of the moving average MA and standard deviation SD, the conversion to rotation speed series data or rotation angle series data, the application of resampling and low-pass filters, and other conversions and derivations of data series and processing based on data series are performed by a conversion routine that inputs one or more data series and conversion parameters and outputs one or more derived data series.

[0104] In addition, multiple conversion routines are chained and executed in a predefined order depending on the purpose to obtain the desired data series. This chain of multiple predefined conversion routines is called a pipeline. In the example of Figure 13(a), the pipeline is configured as follows: A rotation speed series (unit: r) from the initial state is derived from the sampling time series and rotation speed (r / min) series of the original time series data. Using the rotation speed series as an index, an exponentially smoothed moving average (EMA) of the power series over six rotations is calculated to derive the power EMA series. The rotation speed series and EMA series are input and smoothed into a curve by applying a low-pass filter. The function of the smoothed curve is input and extreme values ​​are found (Figure 13(a)). The function of the smoothed curve is input and inflection points are found (Figure 14).

[0105] While multiple conversion routines are applied in a single pipeline, using multiple pipelines in combination can be used to understand the kneading state and changes in physical properties. In the pipeline applied to Figure 12(b), it can be seen that the standard deviation of the power values ​​converges after 15:08. During the same period, the step can be determined to be complete when the function of the smoothed curve of the power values ​​EMA passes its maximum value and the first derivative remains negative, as in the example of Figure 13(a) described above, and when the second derivative of the processed function passes the last maximum value and, after an odd-numbered inflection point, exceeds a positive threshold near zero and changes sign, as in the example of Figure 14.

[0106] The kneading state evaluation method and kneading machine of the present invention can grasp changes in the physical properties of the material to be kneaded using various kneading parameters in a closed-type kneading machine with a pair of rotors, and can therefore be widely used in the kneading of rubber, plastics, ceramics, etc. This is expected to contribute to improved productivity and stabilization of quality, such as by enabling the kneading end point to be grasped with higher accuracy, thereby shortening the kneading time.

[0107] REFERENCE SIGNS LIST 1 Closed type kneader 2 Kneading tank 2A, 2B Rotor chamber 2C Ceiling wall portion 3A, 3B Kneading rotor 4a, 4b Blade 5A, 5B Rotor shaft 6A, 6B Bearing 7A, 7B Gear 7a Mounting hole 7b Key groove 8 Electric motor 8a Circuit portion 8b Motor portion 9a Pressurizing lid 9b Rod 10 Temperature sensor 10a Detection end 11 Power sensor 12 Control portion 13 Acquisition portion 14 Calculation portion 15 Determination portion 16 Notification portion 17 Coupling

Claims

1. A method for evaluating a mixing state in a kneading machine having a pair of rotors connected by a pair of gears and rotated by the drive of an electric motor, the evaluation method comprising repeatedly sampling and acquiring the temperature of the material to be mixed, the pressure of a pressure lid in the kneading machine, the effective value of the AC current supplied to the electric motor, the DC current value supplied to the electric motor, the power value consumed by the electric motor, the load factor of the electric motor, the output torque of the electric motor, the number of rotations per minute of the rotor, or the angular velocity of the rotor, and using, as an evaluation index, a moving average of a predetermined arbitrary time width, a variance of an arbitrary time width, a standard deviation of an arbitrary time width, a deviation between the input value and the moving average of an arbitrary time width, or processed values ​​of these, or a combination of these.

2. The method for evaluating the kneading state described in claim 1, characterized in that in the kneading machine, the pair of rotors are tangential rotors, the gears have different integer numbers of teeth, and the pair of tangential rotors rotate at different speeds, and the evaluation method uses, as the arbitrary time width, the period immediately preceding the current time of the input value until the pair of rotors return to the same phase, and uses as the evaluation index a moving average of the period, a variance of the period, a standard deviation of the period, a deviation between the input value and the moving average of the period, or a combination of these.

3. A method for evaluating the kneading state described in claim 1, characterized in that when the rotation speed of the pair of rotors in the kneader changes during the process, the arbitrary time width is changed in accordance with the change in rotation speed, and a moving average of the changed time width, a variance of the changed time width, a standard deviation of the changed time width, a deviation between the input value and the moving average of the changed time width, or a combination of these is calculated.

4. A method for evaluating a mixing state in a kneading machine having a pair of rotors connected by a pair of gears and rotated by the drive of an electric motor, the evaluation method comprising repeatedly sampling and acquiring the temperature of the material to be mixed, the pressure of a pressure lid in the kneading machine, the effective value of the AC current supplied to the electric motor, the DC current value supplied to the electric motor, the power value consumed by the electric motor, the load factor of the electric motor, the output torque of the electric motor, the number of rotations per minute of the rotor, or the angular velocity of the rotor, and using as an evaluation index a moving average of the acquired input values ​​for a predetermined arbitrary number of rotations of one of the rotors, a variance for an arbitrary number of rotations, a standard deviation for an arbitrary number of rotations, a deviation between the input value and the moving average for an arbitrary number of rotations, or processed values ​​of these, or a combination of these.

5. The method for evaluating the kneaded state according to claim 1, wherein the end timing is determined based on whether the variance, the standard deviation, or the deviation is equal to or smaller than a determination threshold value.

6. The method for evaluating a kneaded state according to claim 1, wherein the kneader is a kneader for kneading a non-Newtonian fluid.

7. A method for evaluating the kneading state according to claim 1, characterized in that a combination of a moving average of the period and a standard deviation of the period is used as an evaluation index, and the end timing is determined based on the standard deviation being equal to or less than a judgment threshold value.

8. A kneading machine having a pair of rotors connected by a pair of gears and rotated by the drive of an electric motor, the kneading machine characterized in having an acquisition unit that acquires by repeatedly sampling the temperature of the material to be kneaded, the pressure of a pressure lid in the kneading machine, the effective value of the AC current supplied to the electric motor, the DC current value supplied to the electric motor, the power value consumed by the electric motor, the load factor of the electric motor, the output torque of the electric motor, the number of rotations per minute of the rotor, or the angular velocity of the rotor, and a calculation unit that calculates a moving average of a predetermined arbitrary time width, a variance of an arbitrary time width, a standard deviation of an arbitrary time width, a deviation between the input value and the moving average of an arbitrary time width, or a value obtained by processing these, or a combination of these, of the input values ​​acquired by the acquisition unit.

9. The kneader according to claim 8, wherein the pair of rotors are tangential rotors, the gears have different integer numbers of teeth, and the pair of tangential rotors rotate at different speeds, and the calculation unit uses the period immediately preceding the current time of the input value until the pair of rotors return to the same phase as the arbitrary time width, and calculates a moving average of the period, a variance of the period, a standard deviation of the period, a deviation between the input value and the moving average of the period, or a combination of these.

10. A kneader as described in claim 8, characterized in that when the rotation speed of the pair of rotors in the kneader changes during the process, the calculation unit changes the arbitrary time width in accordance with the change in rotation speed, and calculates a moving average of the changed time width, a variance of the changed time width, a standard deviation of the changed time width, a deviation between the input value and the moving average of the changed time width, or a combination of these.

11. A kneading machine having a pair of rotors connected by a pair of gears and rotated by the drive of an electric motor, the kneading machine having an acquisition unit that repeatedly samples and acquires the temperature of the material to be kneaded, the pressure of a pressure lid in the kneading machine, the effective value of the AC current supplied to the electric motor, the DC current value supplied to the electric motor, the power value consumed by the electric motor, the load factor of the electric motor, the output torque of the electric motor, the number of rotations per minute of the rotor, or the angular velocity of the rotor, and a calculation unit that calculates a moving average of the input values ​​acquired by the acquisition unit for a predetermined arbitrary number of rotations of one of the rotors, a variance for an arbitrary number of rotations, a standard deviation for an arbitrary number of rotations, a deviation between the input value and the moving average for an arbitrary number of rotations, or values ​​obtained by processing these, or a combination of these.

12. The kneading machine according to claim 8, further comprising a judgment section for judging the end timing based on whether the variance, the standard deviation, or the deviation calculated by the calculation section is equal to or smaller than a judgment threshold value.

Images