ELECTROSTATIC SEPARATOR

Abstract

Electrostatic separator comprising: an electrification device (2) for moving a separation target material (11) containing a plurality of types of separation target particles (11a, 11b) to electrify each separation target particle (11a, 11b), the plurality of types of separation target particles (11a, 11b) having different electrification properties; an electric field generator (4, 5, 6) for applying the electric field to each separation target particle (11a, 11b) electrified by the electrification device (2) and causing each separation target particle (11a, 11b) to fall into a position corresponding to an electrified state of each separation target particle (11a, 11b); a collecting container (7) for collecting each separation target particle (11a, 11b) that has fallen from the electric field generator (4, 5, 6); a separating element (8, 9) which is movable in a predetermined direction in the collecting container (7), wherein the separating element (8, 9) divides the collecting container (7) into a plurality of collecting chambers (7a, 7b, 7c); and a controller (10) for controlling a position of the separation element (8, 9) in the predetermined direction, wherein the controller (10) samples a part of the separation target material (11) and measures a composition ratio between the plurality of types of separation target particles (11a, 11b), analyzes a change tendency of the composition ratio from a previously measured composition ratio and a currently measured composition ratio and controls the position of the separating element (8, 9) on the basis of an analysis result of the change tendency, wherein the controller (10) comprises: a measuring unit (20) for taking samples from a part of the separation target material (11) and measuring the composition ratio, a time series data generation unit (32) for arranging the previously measured composition ratio and the composition ratio currently measured by the measuring unit (20) in the order of measurement to generate time series data on the composition ratio, a time series data analysis unit (33) for analyzing the change tendency of the composition ratio from the time series data and estimating a current composition ratio estimate based on the analysis result of the change tendency, and a position control unit (40) for controlling the position of the separating element (8, 9) in the predetermined direction based on the current composition ratio estimated by the time series data analysis unit (33).

Description

TECHNICAL FIELD

[0001] The present invention relates to an electrostatic separation apparatus and an electrostatic separation method. STATE OF THE ART

[0002] Many plastics (resins) are used in the casings of household appliances such as air conditioners, refrigerators, and washing machines. When a used household appliance is disposed of, the casing is removed from the discarded appliance and shredded in a shredder. Since a group of plastic particles discharged from the shredder contains a variety of plastic particle types, plastic recycling involves separating the various plastic particle types contained in the group of plastic particles and collecting them by type.

[0003] For example, the plastic particle group includes a variety of plastic particle types, such as polypropylene resin particles (hereinafter referred to as "PP particles"), acrylonitrile-butadiene-styrene copolymer resin particles (hereinafter referred to as "ABS particles"), and polystyrene resin particles (hereinafter referred to as "PS particles"). In a recycling process, the PP particles with a low specific gravity are first separated and collected from the plastic particle group by a gravity separation method, and the ABS particles and PS particles are then separated and collected by an electrostatic separation method.

[0004] The electrostatic separation process takes advantage of the fact that the electrified state (such as the polarity of static electricity, the amount of charge) of the plastic particles, which are triboelectrically charged by moving a variety of plastic particle types with different electrification properties, and the force exerted by an electric field vary depending on the type of plastic particle.

[0005] In this electrostatic separation process, the electrified states of the various types of plastic particles change depending on the mixing ratio between the plastic particles (this mixing ratio is sometimes referred to as the "composition ratio" and may be referred to as the "composition ratio" below), and the optimal conditions during electrostatic separation also change accordingly. For example, if the composition ratio between ABS particles and PS particles contained in the group of plastic particles is 58:42, the electrostatic separation conditions are optimized according to this composition ratio.

[0006] Japanese Patent Application Laid-Open No. 2001-129435 (PTL 1) discloses an electrostatic separation device that has a plurality of collection chambers defined by a plurality of separation plates arranged in a collection unit, and that separates a plurality of types of separation target particles (plastic particles). When the plurality of types of separation target particles have not been accurately separated, this electrostatic separation device adjusts the collection conditions by moving the positions of the separation plates of the collection unit (see PTL 1).

[0007] Japanese Patent Application Laid-Open No. JP 2011-115753 A (PTL 2) discloses an electrostatic separation device and an electrostatic separation method that deflects electrostatically charged materials due to their static charge in free fall from an electrostatic field and separates them into collection units defined by separation plates. When the plurality of types of separation target particles have not been accurately separated, this electrostatic separation device adjusts the collection conditions based on the weight ratio of the materials by moving the positions of the separation plates of the collection unit (see PTL 2).

[0008] Japanese Patent Application Laid-Open No. JP 2018-065123 A (PTL 3) discloses an electrostatic separation device and an electrostatic separation method that deflects electrostatically charged materials due to their static charge in free fall from an electrostatic field and separates them in collection units defined by separation plates. If the various types of separation target particles have not been accurately separated, this electrostatic separation device adjusts the collection conditions, which are detected by sensors above the collection unit, by moving the positions of the separation plates of the collection unit (see PTL 3).

[0009] PCT patent application publication WO 2012 / 101874 A1 (PTC 4) discloses a separation process that identifies a plurality of types of separation target particles (plastic particles) using light (see PTC 4). LITERATURE LISTPATENTLITERATURE PTL 1: Japanese Patent Application Laid-Open No. JP 2001-129 435 A PTL 2: Japanese Patent Application Laid-Open No. JP 2011-115 753 A PTL 3: Japanese Patent Application Laid-Open No. JP 2018-065 123 A PTL 4: PCT patent application publication WO 2012 / 101874 A1 SUMMARY OF THE INVENTION TECHNICAL PROBLEM

[0010] It is difficult to inspect the entire separation target material by the electrostatic separation method, so sampling is usually performed. Even in the electrostatic separation device disclosed in Japanese Patent Application Laid-Open No. 2001-129435, the separation target material is sampled before electrostatic separation of the separation target material. However, the separation and collection of separation target particles may be performed with insufficient accuracy because the measured value of the composition ratio by sampling has a sampling error of ± a few percentage points.

[0011] Therefore, a main object of the present invention is to provide an electrostatic separation apparatus and an electrostatic separation method that can improve the accuracy of separation and collection by electrostatic separation. SOLUTION TO THE PROBLEM

[0012] An electrostatic separation device of the present invention includes an electrification device, an electric field generator, a collection container, a separation element, and a controller. The electrification device moves a separation target material containing a plurality of separation target particles to electrify each separation target particle, the plurality of separation target particles having different electrification properties. The electric field generator applies an electric field to each separation target particle electrified by the electrification device and causes each separation target particle to fall to a position corresponding to an electrified state of each separation target particle. The collection container collects all the separation target particles that have fallen from the electric field generator.The separating element is movable in a predetermined direction within the collection container and divides the collection container into a plurality of collection chambers. The controller samples a portion of the separation target material and measures a composition ratio between the plurality of types of separation target particles, analyzes a change trend of the composition ratio from the previously measured composition ratio and the composition ratio measured this time, and controls a position of the separating element in the predetermined direction based on the analysis result of the change trend. ADVANTAGEOUS EFFECTS OF THE INVENTION

[0013] The electrostatic separation device analyzes the change trend of the composition ratio from the previously measured composition ratio and the composition ratio measured this time and controls the position of the separation element based on the analysis result. This allows the position of the separation element to be controlled based on the composition ratio, removing any measurement error caused by the sampling of the separation target material. Accordingly, the electrostatic separation device can improve the accuracy of separation and collection by electrostatic separation. BRIEF DESCRIPTION OF THE DRAWINGS Fig. 1 shows a configuration of an electrostatic separator according to Embodiment 1. Fig. 2 is a block diagram showing a configuration of a Fig. 1 shows the control. Fig. 3 shows the analysis results of a Fig. 2 shown time series data analysis unit. Fig. 4 is a flowchart showing an example method of a Fig. 2 shows the composition ratio data analysis unit. Fig. Figure 5 shows exemplary case distributions of plastic particles. Fig. Figure 6A shows the relationship between a composition ratio and a peak position of a Fig. 5 shown curved line Ca. Fig. Figure 6B shows the relationship between a composition ratio and a peak width of the Fig. 5 shown curved line Ca. Fig. Figure 7 is a diagram illustrating the operation of a Fig. 2 shown position calculation unit. Fig. 8 shows a configuration of a Fig. 2 shown drive. Fig. 9 is a block diagram showing a configuration of a processing circuit used in the Fig. 2 shown control is included. Fig. 10 is a block diagram showing a configuration of a controller in Embodiment 2. Fig. 11 is a diagram illustrating a process by a Fig. Composition ratio prediction unit shown in Figure 10. Fig. 12 is a flowchart showing an example method of a Fig. 10 shows the composition ratio data analysis unit. DESCRIPTION OF EMBODIMENTS Embodiment 1

[0014] Fig. 1 shows a configuration of an electrostatic separator according to Embodiment 1. Referring to Fig. 1, the electrostatic separator comprises a charging unit 1, an electrification drum 2, a vibration feeder 3, electrodes 4, 5, a DC power supply 6, a collecting container 7, separating plates 8, 9 and a controller 10.

[0015] The electrostatic separation device electrostatically separates a plastic particle group 11 (separation target material) containing a plurality of types (e.g., two types) of plastic particles (separation target particles) 11a, 11b with different electrification properties into plastic particles 11a and plastic particles 11b. Hereinafter, the plastic particles 11a are ABS particles, and the plastic particles 11b are PS particles. The plastic particles 11a, 11b are obtained, for example, by crushing a housing of a household appliance with a crusher and drying the crushed housing, and forming it into a square of approximately 10 mm.

[0016] The charging unit 1 comprises a container 1a and a charge supply device 1b. The container 1a is loaded with a group of plastic particles 11 that has been dried by a dryer (not shown). The container 1a supplies the charge supply device 1b with the group of plastic particles 11 in a predetermined amount per unit time. The charge supply device 1b feeds the group of plastic particles 11 from the container 1a to the electrification drum 2.

[0017] The electrification drum 2 rotates to move the plastic particle group 11. In the electrification drum 2, the different types of plastic particles 11a, 11b contained in the plastic particle group 11 rub against each other and become electrified. Each of the charged plastic particles 11a, 11b has a polarity (positive or negative) and a charge quantity corresponding to a triboelectric series. In this example, the plastic particles 11a, which are the ABS particles, are positively charged, and the plastic particles 11b, which are the PS particles, are negatively charged.

[0018] The electrically charged plastic particles 11a, 11b are fed to the rear edge region of an upper surface of the vibration feeding device 3.

[0019] Positively charged plastic particles 11a and negatively charged plastic particles 11b attract each other through electrostatic force to become paired. The vibration feeder 3 releases the pairing of the plastic particles 11a, 11b through vertical ejection vibrations, which push the plastic particles 11a, 11b forward while moving up and down, causing the plastic particles 11a, 11b to migrate in the X direction in the figure and then fall from the front edge of the vibration feeder 3. The electrification drum 2 and the vibration feeder 3 constitute an electrification device that electrifies each of the plastic particles 11a, 11b and then causes each electrified plastic particle to fall.

[0020] The electrodes 4, 5 are a first electrode and a second electrode, respectively, and each plate is shaped as a flat plate. The electrodes 4, 5 are arranged in the X direction in the figure and face each other perpendicular to a path through which the plastic particles 11a, 11b fall. A ground voltage GND is applied to the electrode 4. The DC power supply 6 applies a prescribed DC voltage between the electrodes 4 and 5 to generate an electrostatic field between the electrodes 4 and 5.

[0021] When plastic particles 11a, 11b separated by the vibration feeder 3 fall between the electrodes 4, 5, each plastic particle falls while being attracted to one of the two electrodes 4, 5 by an electrostatic force corresponding to its electrified state (polarity, amount of charge). In other words, each plastic particle falls to a different position while traversing a parabolic trajectory corresponding to its electrified state. In this example, the plastic particles 11a are positively charged and therefore fall toward the side of electrode 4. In contrast, the plastic particles 11b are negatively charged and therefore fall toward the side of electrode 5.

[0022] The electrodes 4, 5 and the DC power supply 6 form an electric field generator that applies an electrostatic field to each of the charged plastic particles and causes each plastic particle to fall into a position corresponding to the electrified state of the respective plastic particle.

[0023] Located below the electrodes 4, 5 is the collection container 7, which collects the plastic particles 11a, 11b that have fallen from the vibration feeder 3 between the electrodes 4, 5. The collection container 7 has the shape of a rectangular parallelepiped and is open at the top. The opening of the collection container 7 is in the shape of a rectangle, the long sides of which are aligned in the X direction in the figure.

[0024] Each of the partition plates 8, 9 is also referred to as a partition element. In the collection container 7, each partition plate is arranged parallel to the YZ plane and movable in the X direction in the figure. The position of each partition plate 8, 9 in the X direction is controlled by the controller 10. The partition plate 8 is arranged on the electrode 4 side, and the partition plate 9 is arranged on the electrode 5 side. The collection container 7 is divided by the partition plates 8, 9 into a collection chamber 7a on the electrode 4 side, a collection chamber 7b on the electrode 5 side, and a collection chamber 7c between the collection chambers 7a, 7b.

[0025] Each plastic particle that has fallen from the vibration feeder 3 between the electrodes 4, 5 is collected by any of the three collection chambers 7a to 7c according to its electrified state. Since in this example, the plastic particles 11a are positively charged, they are collected in the collection chamber 7a. Since the plastic particles 11b, on the other hand, are negatively charged, they are collected in the collection chamber 7b. Insufficiently charged plastic particles 11a, 11b are collected in the collection chamber 7c.

[0026] The plastic particles 11a collected in the collection chamber 7a, the plastic particles 11b collected in the collection chamber 7b and the plastic particles 11a, 11b collected in the collection chamber 7c are conveyed by a conveying device (not shown) and are located in separate containers.

[0027] The controller 10 samples a portion of the plastic particle group 11 at a predetermined frequency, measures a composition ratio between the plastic particles 11a, 11b (a composition ratio of the separation target particles), and arranges the previously measured composition ratio and the composition ratio measured this time in the order of measurement, thereby generating time series data on the composition ratio. Subsequently, the controller 10 analyzes the time series data to determine a change trend of the composition ratio from which a measurement error is removed. The controller 10 then estimates the final composition ratio based on the change trend and controls the positions of the separation plates 8, 9 in the X direction based on the estimated final composition ratio.

[0028] Fig. 2 is a block diagram showing a configuration of the controller 10. Referring to Fig. 2, the controller 10 includes a composition ratio detection unit 20, a composition ratio data analysis unit 30, and a position control unit 40. The composition ratio detection unit 20 (measuring unit) includes a sampler 21 and a composition ratio measuring device 22.

[0029] The electrostatic separation device processes a huge number of plastic particles per day, making it difficult to sample all of them. Therefore, the sampler 21 takes samples from a portion of the plastic particle group 11 from a predetermined location on the loading unit 1 or the vibration feeder 3. The weight of the plastic particle group 11 from which a sample is taken at a time is, for example, approximately 70 g. Sampling can be performed at any frequency, for example, once a day or twice a week.

[0030] The composition ratio measuring device 22 analyzes the composition of the plastic particle group 11 sampled by the sampler 21 according to a predetermined method and calculates a composition ratio between the plastic particles 11a, 11b based on the analysis result.

[0031] One example of a method for analyzing the composition is the near-infrared analysis method. In the near-infrared analysis method, a plastic particle is irradiated with near-infrared rays, and the spectrum (characteristics of the wavelengths contained) of a reflected wave is analyzed. Different types of plastic particles have different spectral properties, and the type of plastic particle is identified based on the spectral properties. The composition ratio among the plurality of plastic particle types is determined as the ratio between the amounts of the identified plurality of plastic particle types.

[0032] The measured value of the composition ratio obtained by random inspection using such sampling has a sampling error of ± several percentage points. Even if the ratio of plastic particles 11a determined by the composition ratio detection unit 20 is, for example, 50%, the actual value may be 45% or 55%. In particular, if the positions of the separation plates 8, 9 are controlled using the measured value obtained by random inspection without any change, an error occurs in controlling the positions of the separation plates 8, 9, thereby reducing the accuracy of separating the plastic particles 11a, 11b.

[0033] Therefore, in the electrostatic separator according to Embodiment 1, the composition ratio data analysis unit generates time series data on composition ratios measured by the composition ratio measuring device 22 from the past to the present, and analyzes the time series data, thereby estimating the latest composition ratio from which a measurement error is removed.

[0034] The composition ratio data analysis unit 30 includes a composition ratio storage unit 31, a time series data generation unit 32, a time series data analysis unit 33, and a composition ratio estimated value output unit 34. A composition ratio database is stored in the composition ratio storage unit 31. The composition ratio database contains data on a large number of composition ratios previously measured by the composition ratio measuring device 22, arranged in the order of measurement. Information on the date and time of measurement is added to the data on each composition ratio.

[0035] The time series data generation unit 32 reads the composition ratio database from the composition ratio storage unit 31 and arranges the past composition ratio included in the read composition ratio database and the composition ratio measured this time by the composition ratio measuring device 22 in the order of measurement, thereby generating time series data on composition ratios.

[0036] For example, if the composition ratios previously recorded for 300 days are stored in the composition ratio storage unit 31, and the composition ratios for 10 days are re-measured this time by the composition ratio measuring device 22, the time series data generation unit 32 arranges the composition ratios for 310 days in date order, thus generating time series data. The generated time series data is stored in the composition ratio storage unit 31 as a new composition ratio database and is also supplied to the time series data analysis unit 33.

[0037] The time series data analysis unit 33 analyzes a change trend of the composition ratio from the time series data of the composition ratios generated by the time series data generation unit 32, and estimates the latest composition ratio from which a measurement error is removed based on an analysis result of the change trend. The time series data analysis method is, for example, a method using a state-space model. The state-space model is a method for estimating, for data with a measurement error included in a measured value, a value (hereinafter referred to as a state value) from which the measurement error is removed. The state-space model is a well-known technique, and a method for estimating a state value will be briefly described.

[0038] In the state-space model, filtering and smoothing are performed to estimate a state value. Filtering is a process of estimating a state value α given time series data from a time t1 to a time tN. n at a time tn based on the observed values ​​Y t ={y1, ..., y n-1} up to a mean time t(n-1). Where N is an integer not less than 2, n is an integer not less than 1 and less than N, and a time t is, for example, a date.

[0039] In embodiment 1, a Kalman filter is used for filtering. The Kalman filter assumes that the errors contained in the observed values ​​are normally distributed and predicts a state value α n at time tn using the observed values ​​Y t ={y1, ..., y (n-1)} from time t1 to time t(n-1). Such a prediction is called one-step prediction. One-step prediction is performed sequentially from time t1 to time tN, and at each time point, a state value is estimated. Smoothing is a process of estimating the state values ​​α1, ..., α N all time points using all observed values ​​Y t ={y1, ..., y N} that are given.

[0040] Fig. 3 shows the analysis results of the time series data analysis unit 33. Referring to Fig. 3, the horizontal axis represents a date (time t), and the vertical axis represents a composition ratio (%). The composition ratio is a ratio (%) of plastic particles 11a to plastic particles 11a, 11b contained in the sampled plastic particle group 11. A mark ◯ represents a measured value of a composition ratio. A smooth curve is a curved line representing the state values ​​α1, ..., α N all times t1 to tN obtained by smoothing. The state value α N at time tN is an estimated value for the latest composition ratio.

[0041] Again referring to Fig. 2, the composition ratio estimation output unit 34 outputs the state value α Nat time tN to the position control unit 40 as a composition ratio estimate. For example, if the measured value of the composition ratio at time tN obtained by the composition ratio acquisition unit 20 is 45% and the estimated composition value at time tN obtained by the time series data analysis unit 33 is 42%, the composition ratio estimate output unit 34 does not output a measured ratio of 45%, but outputs an estimated value of 42%.

[0042] Fig. 4 is a flowchart showing an example process of the Fig. 2. A series of processings shown in this flowchart are repeatedly performed for each prescribed cycle.

[0043] With reference to Fig. 4, when the composition ratio acquisition unit 20 receives a measured value of a new composition ratio, the time series data generation unit 32 arranges the previously measured composition ratio and the composition ratio measured this time in order of date in step S1, thus generating time series data on the composition ratio. In step S2, the time series data analysis unit 33 analyzes a change trend of the composition ratio in the time series data. In step S3, the time series data analysis unit 33 extracts a smooth curve ( Fig. 3). In step S4, the composition ratio estimated value output unit 34 outputs the latest composition ratio estimated value.

[0044] Again referring to Fig. 2, the position control unit 40 moves each of the partition plates 8, 9 ( Fig. 1) to an optimal position in the X direction based on the latest composition ratio estimate output from the composition ratio data analysis unit 30. The position control unit 40 includes a case distribution storage unit 41, a case distribution prediction unit 42, a purity setting unit 43, a position calculation unit 44, a position output unit 45, and a drive unit 46.

[0045] The relationship between the composition ratio between plastic particles 11a, 11b and the falling distribution of the plastic particles 11a, 11b will now be described. Fig. 5 shows the case distributions of the plastic particles 11a, 11b when the composition ratio between the plastic particles 11a, 11b has a certain value.

[0046] In Fig. 5, the horizontal axis represents a position x (mm) in the X direction ( Fig. 1), and the vertical axis represents a weight ratio (wt%) between the fallen plastic particles 11a, 11b. The position x directly below the front edge of the vibration feeder 3 is taken as the origin (x = 0) in the X direction, and the falling distribution of the individual plastic particles 11a, 11b is assumed to be a normal distribution. The curved lines Ca and Cb represent the falling distributions of the plastic particles 11a and 11b, respectively.

[0047] When the composition ratio between the plastic particles 11a, 11b changes, the number of individual plastic particles 11a, 11b changes, and the electrification amount of all the plastic particles 11a and the electrification amount of all the plastic particles 11b changes. Accordingly, the peak position and peak width of each of the curved lines Ca, Cb change. For each composition ratio (e.g., 90%, 50%, 10%), the peak position and peak width of each of the curved lines Ca, Cb are experimentally determined in advance. Therein, the composition ratio between the plastic particles 11a, 11b is represented as the ratio (%) of the weight of the plastic particles 11a to the weight of the plastic particles 11a, 11b.

[0048] Fig. 6A, Fig. 6B each show the relationship between a pre-experimental determined composition ratio (%) and the curved line Ca. Fig. 6A shows the relationship between a composition ratio (%) and a peak position (mm) of the curved line Ca, and Fig. Figure 6B shows the relationship between a composition ratio (%) and a peak width (mm) of the curved line Ca. The peak position (mm) indicates the distance from the origin (x = 0).

[0049] As can be seen from the Fig. 6A and Fig. 6B, the peak position (mm) decreases proportionally to the composition ratio (%), and the peak width (mm) increases proportionally to the composition ratio (%). Since for each of the Fig. When performing regression analysis, three experimental values ​​are connected by a single regression line, and the regression line is represented by a linear regression equation. Therefore, even if the composition ratio (%) is an arbitrary value, a peak position (mm) and a peak width (mm) of the curved line Ca can be obtained based on the composition ratio (%) and two regression equations.

[0050] Even if the composition ratio (%) is an arbitrary value, a peak position (mm) and a peak width (mm) of the curved line Cb can be obtained based on the composition ratio (%) and the other two regression equations. Thus, the case distributions of the plastic particles 11a, 11b can be predicted for each composition ratio (%).

[0051] As in Fig. As shown in Figure 2, a database is stored in the case distribution storage unit 41. The database shows the relationship between the composition ratio between plastic particles 11a, 11b and the case distributions of the plastic particles 11a, 11b. For example, two regression equations for determining a peak position (mm) and a peak width (mm) of the curved line Ca from an arbitrary composition ratio (%), and two other regression equations for determining a peak position (mm) and a peak width (mm) of the curved line Cb from an arbitrary composition ratio (%) are stored in the database.

[0052] The case distribution prediction unit 42 predicts the curved lines Ca, Cb (ie, the case distributions of the plastic particles 11a, 11b) based on the latest composition ratio estimate value (ie, an arbitrary composition ratio) provided from the composition ratio estimate value output unit 34 and the database stored in the case distribution storage unit 41.

[0053] The purity setting unit 43 sets a target value Pat for the purity Pa of the plastic particles 11a collected in the collection chamber 7a and a target value Pbt for the purity Pb ​​of the plastic particles 11b collected in the collection chamber 7b.

[0054] The purity Pa refers to the ratio of plastic particles 11a to plastic particles 11a, 11b collected in the collection chamber 7a. The purity Pb ​​refers to a ratio of plastic particles 11b to the plastic particles 11a, 11b collected in the collection chamber 7b. The target purity value Pat and the target purity value Pbt can be the same or different.

[0055] The position calculation unit 44 determines a position x1 of the partition plate 8 ( Fig. 1), at which a collection rate Ra of the plastic particles 11a is maximized, and a position x2 of the partition plate 9 ( Fig. 1), at which the collection rate Rb of the plastic particles 11b is maximized, under the condition that the purity target values ​​Pat, Pbt set in advance by the purity setting unit 43 are achieved.

[0056] The collection rate Ra is a ratio between the amount of plastic particles 11a collected in the collection chamber 7a and the total amount of plastic particles 11a, 11b fed to the electrostatic separator. The collection rate Rb is the ratio between the amount of plastic particles 11b collected in the collection chamber 7b and the total amount of plastic particles 11a, 11b fed to the electrostatic separator.

[0057] Fig. Fig. 7 is a diagram illustrating an operation of the position calculation unit 44 obtained by adding partition plates 8, 9 to Fig. 5 is obtained. The partition plates 8, 9 are arranged at the positions x=x1 and x=x2, respectively, where x1 > 0 and x2 > 0. The area to the left of the partition plate 8 corresponds to the collection chamber 7a, the area to the right of the partition plate 9 corresponds to the collection chamber 7b, and the area between the partition plates 8, 9 corresponds to the collection chamber 7c.

[0058] The falling distributions of the plastic particles 11a, 11b are represented by Gaussian functions with a falling position x as a variable and are denoted by fa(x) and fb(x), respectively. Assuming that Qa and Qb are the quantities of the plastic particles 11a and 11b collected in the collection chamber 7a, respectively, Qa and Qb are values ​​obtained by integrating the Gaussian functions fa(x), fb(x) from -∞ to x 1 and are expressed by the following equations (1), (2). Qa=∫−∞x1fa(x)dx Qb=∫−∞x1fb(x)dx

[0059] Since the purity Pa of the plastic particles 11a is a ratio of the plastic particles 11a to the plastic particles 11a, 11b collected in the collection chamber 7a, 11b, Pa = Qa / (Qa + Qb). Since the collection rate Ra of the plastic particles 11a is a ratio between the amount of plastic particles 11a collected in the collection chamber 7a and the total amount of plastic particles 11a, 11b supplied to the electrostatic separator, the collection rate Ra is expressed by the following equation (3). Ra=Qa / {∫−∞∞fa(x)dx+∫−∞∞fb(x)dx}

[0060] Assuming that the total amount of plastic particles 11a, 11b fed to the electrostatic separator, that is, the denominator of equation (3), is 1, Ra = Qa. The purity Pb ​​and the collection rate Rb of the plastic particles 11b are determined by the same method as the purity Pa and the collection rate Ra of the plastic particles 11a.

[0061] Next, the relationship between the position of the partition plate 8 and the purity Pa and the collection rate Ra of the plastic particles 11a is described. First, the case where the position of the partition plate 8 was shifted from x=x1 to x=x1+Δx, where Δx > 0, is described.

[0062] In other words, if the position of the partition plate 8 in Fig. 7 is shifted to the right, the area of ​​the falling distribution of the plastic particles 11a in the collection chamber 7a increases, which increases the collection rate Ra of the plastic particles 11a. On the other hand, the area of ​​the falling distribution of the plastic particles 11b in the collection chamber 7a increases, which decreases the purity Pa of the plastic particles 11a.

[0063] Next, the case is described where the position of the partition plate 8 has been shifted from x=x1 to x=x1-Δx. In other words, when the position of the partition plate 8 in Fig. 7 is shifted to the left, the area of ​​the falling distribution of the plastic particles 11a in the collection chamber 7a decreases, thereby decreasing the collection rate Ra of the plastic particles 11a. On the other hand, the area of ​​the falling distribution of the plastic particles 11b in the collection chamber 7a decreases, thereby increasing the purity Pa of the plastic particles 11a.

[0064] In other words, there is a trade-off between the purity Pa and the collection rate Ra, where the purity Pa decreases when the position of the partition plate 8 is moved to increase the collection rate Ra, and the collection rate Ra decreases when the position of the partition plate 8 is moved to increase the purity Pa.

[0065] The position calculation unit 44 ( Fig. 2) determines the position x1 of the partition plate 8 such that the purity Pa is not less than the target purity Pat when the purity setting unit 43 sets the target purity Pat to, for example, 99%. In this case, the determined position x1 of the partition plate 8 is a position at which the maximum collection rate Ra is achieved under the condition that the purity Pa is not less than the target purity Pat. Since the relationship between the position x2 of the partition plate 9 and the purity Pb ​​and the collection rate Rb of the plastic particles 11b is the same as the relationship between the position x1 of the partition plate 8 and the purity Pa and the collection rate Ra of the plastic particles 11a, the description will not be repeated.

[0066] The position output unit 45 outputs the positions x1, x2 of the separating plates 8, 9 determined by the position calculation unit 44. The drive unit 46 moves the separating plate 8 to the position x1 output by the position output unit 45 and moves the separating plate 9 to the position x2 output by the position output unit 45.

[0067] Fig. Figure 8 shows a configuration of the drive unit 46, showing the collecting container 7 from above. As in Fig. As shown in Figure 8, the drive unit 46 comprises rails 51, 52, slides 53, 54, ball screws 55, 56, and motors 57, 58. The rails 51, 52 are aligned in the X direction in the figure and arranged parallel to each other at a predetermined distance. The collection container 7 is arranged between the rails 51, 52, viewed from above.

[0068] Each of the sliders 53, 54 is arranged across the rails 51, 52 through a band-shaped hole formed in the side walls of the collecting container 7 and is connected to the rails 51, 52 by a sliding member (not shown). Each of the sliders 53, 54 is movable in the X direction along the rails 51, 52, and their movements in the Y and Z directions are restricted by a stopper (not shown). The partition plate 8 is arranged parallel to the YZ plane in the figure and is fixed to the slider 53. The partition plate 9 is arranged parallel to the YZ plane in the figure and is fixed to the slider 54.

[0069] A threaded hole extending in the X direction passes through the rear end (the upper end in the figure) of the slider 53, and the ball screw 55 is screwed into the threaded hole. The motor 57 is connected to one end of the ball screw 55. The motor 57 rotates the ball screw 55 so that the partition plate 8 is located at the position x1 output by the position output unit 45.

[0070] A threaded hole extending in the X direction passes through the front end (the lower end in the figure) of the slider 54, and the ball screw 56 is screwed into the threaded hole. The motor 58 is connected to one end of the ball screw 56. The motor 58 rotates the ball screw 56 so that the partition plate 9 is located at the position x2 output by the position output unit 45.

[0071] Fig. Fig. 9 is a block diagram showing a configuration of a processing circuit 60 included in the controller 10 ( Fig. 2). In the controller 10, a part other than a mechanism unit can be implemented by the processing circuit 60. The processing circuit 60 includes at least one processor 61 and at least one memory 62. The processing circuit 60 may include at least one piece of special hardware 63 along with or instead of the processor 61 and the memory 62.

[0072] If the processing circuit 60 includes a processor 61 and a memory 62, each function of the controller 10 is implemented by software, firmware, or a combination of software and firmware. At least one of the software and firmware is described as a program. The program is stored in the memory 62. The processor 61 reads the program stored in the memory 62 and executes it to implement each function of the controller 10.

[0073] The processor 61 is also referred to as a central processing unit (CPU), processing unit, arithmetic unit, microprocessor, microcomputer, or digital signal processor (DSP). The memory 62 is configured, for example, by a non-volatile or volatile semiconductor memory such as a random access memory (RAM), a read-only memory (ROM), a flash memory, an erasable programmable read-only memory (EPROM), or an electrically erasable programmable read-only memory (EEPROM).

[0074] When the processing circuit 60 includes dedicated hardware 63, the processing circuit 60 is implemented, for example, by a single circuit, a complex circuit, a programmed processor, a parallel programmed processor, an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or a combination thereof.

[0075] Each function of controller 10 can be implemented by processing circuitry 60. Alternatively, the functions of controller 10 can be implemented collectively by processing circuitry 60. Some of the functions of controller 10 can be implemented by dedicated hardware 63, and the other functions can be implemented by software or firmware. In this way, processing circuitry 60 implements the functions of controller 10 through hardware 63, software, firmware, or a combination thereof.

[0076] Next, an operation of the electrostatic separator is described. Referring again to Fig. 1, a plastic particle group 11, which has been taken out of the crusher (not shown) and dried by the dryer (not shown), is fed to the loading unit 1 and fed by the loading unit 1 in a predetermined amount per unit time to the electrification drum 2. The plurality of plastic particle types 11a, 11b contained in the plastic particle group 11 are agitated in the electrification drum 2, so that the plastic particles 11a, 11b are positively and negatively triboelectrically charged, respectively.

[0077] The electrified plastic particles 11a, 11b are fed to the vibratory feeder 3, conveyed in the X direction on the vibratory feeder 3 while accompanied by vertical ejection vibrations, and caused to fall one by one between the electrodes 4, 5. A high DC voltage is applied between electrode 4 and electrode 5 by the DC power supply 6 to generate an electric field. The path through which each of the plastic particles 11a, 11b falls changes depending on the electric field and the electrified state (polarity, amount of charge) of each of the plastic particles 11a, 11b.

[0078] The collection container 7 is located below the electrodes 4, 5 and divided into three collection chambers 7a to 7c by two partition walls 8, 9. The positively charged plastic particles 11a fall while being attracted by the electrode 4 and are collected primarily in the collection chamber 7a. The negatively charged plastic particles 11b fall while being attracted by the electrode 5 and are collected primarily in the collection chamber 7b. Insufficiently electrified plastic particles 11a and 11b are collected in the collection chamber 7c between the collection chambers 7a and 7b.

[0079] The plastic particles 11 collected in the collection chamber 7a are recycled as the same plastic type, and the plastic particles 11b collected in the collection chamber 7b are recycled as the same plastic type. The plastic particles 11a, 11b collected in the collection chamber 7c are mixed with the newly produced plastic particles 11a, 11b from the shredder (not shown) and fed back to the loading unit 1.

[0080] In parallel to the above process, a part of the plastic particle group 11 is sampled by the sampler 21 ( Fig. 2) are sampled at a predetermined frequency at predetermined locations on the loading unit 1 and the vibration feeder 3. The composition ratio between the plastic particles 11a, 11b contained in the sampled plastic particle group 11 is measured by the composition ratio measuring device 22.

[0081] The previously measured composition ratio stored in the composition ratio storage unit 31 and the composition ratio measured this time by the composition ratio measuring device 22 are arranged by the time series data generation unit 32 in order of measurement date to generate time series data on the composition ratio. The generated time series data is analyzed by the time series data analysis unit 33 to determine a change trend of the composition ratio from which a measurement error is removed. Based on the change trend, the current composition ratio is estimated.

[0082] A database is stored in the case distribution storage unit 41. The database shows the relationship between the composition ratio between plastic particles 11a, 11b and the case distributions of the plastic particles 11a, 11b. Based on the latest estimated value of the composition ratio and the database stored in the case distribution storage unit 41, the case distribution of each of the plastic particles 11a, 11b is predicted by the case distribution prediction unit 42.

[0083] Based on the predicted drop distributions, the position calculation unit 44 determines the positions x1, x2 of the separation plates 8, 9 such that the purity target values ​​Pat, Pbt of the collected plastic particles 11a, 11b are achieved. Then, the determined positions x1, x2 are transmitted to the drive unit 46 ( Fig. 2 and Fig. 8) and the drive unit 46 moves the separating plates 8, 9 to the respective positions x1, x2.

[0084] In Embodiment 1, the time series data on the measured composition ratio is analyzed to estimate the latest composition ratio from which a measurement error is removed. Based on the estimated latest composition ratio, the positions of the separation plates 8, 9 are controlled as described above. The separation accuracy can thus be improved.

[0085] Embodiment 1 described the case where the plastic particles 11a, 11b are ABS particles and PS particles, respectively, but the present invention is not limited to this. Two types of plastic particles with different electrification properties can be appropriately selected as the plastic particles 11a, 11b. The plastics can be, for example, PP, polyethylene (PE), polyethylene terephthalate (PET), and polyvinyl chloride (PVC), in addition to ABS and PS.

[0086] In Embodiment 1, the collection container 7 is divided into three collection chambers 7a, 7b, 7c by two partition walls 8, 9, but the present invention is not limited thereto. The collection box 7 can be divided into two collection chambers 7a, 7b by a partition plate.

[0087] In Embodiment 1, two types of plastic particles 11a, 11b are separated, but three or more types of plastic particles can also be separated. In this case, the collection container 7 is divided into four or more collection chambers by three or more partition walls.

[0088] Embodiment 1 described the case where a plurality of types of plastic particles are separated, but the present invention is not limited to this. Multiple types of separation target particles with different electrification properties can also be separated. A separation target particle may be made of a material other than plastic, as long as it can be electrified. Embodiment 2

[0089] In Embodiment 1, a portion of the plastic particle group 11 is sampled, the composition ratio between the plastic particles 11a, 11b is measured, the latest composition ratio estimate is calculated based on the value measured in the past and the value measured at this time, and the separation plates 8, 9 are moved to the positions corresponding to the composition ratio estimate. Therefore, if the composition ratio between the plastic particles 11a, 11b has changed before the next composition ratio estimate is determined and the separation plates 8, 9 are moved, the falling distributions of the plastic particles 11a, 11b may change according to the changes in the composition ratio, thereby reducing the purities Pa, Pb and the collection rates Ra, Rb. Embodiment 2 aims to solve this problem.

[0090] An electrostatic separator according to Embodiment 2 includes a controller 70 instead of the controller 10 in the configuration of the electrostatic separator according to Embodiment 1, which is shown in Fig. 1 is shown.

[0091] Fig. 10 is a block diagram showing a configuration of the controller 70 in Embodiment 2. Fig. 10 is used for comparison with the one described in embodiment 1 Fig. 2.

[0092] With reference to Fig. 10, the controller 70 includes a composition ratio data analysis unit 71 instead of the composition ratio data analysis unit 30 of the controller 10 in Embodiment 1. The composition ratio data analysis unit 71 is obtained by adding a composition ratio prediction unit 72 to the composition ratio data analysis unit 30 and replacing the composition ratio estimated value output unit 34 with a composition ratio prediction value output unit 73.

[0093] The composition ratio prediction unit 72 generates a prediction curve (a change tendency of the composition ratio in a prediction period) based on the smooth curve (the change tendency of the composition ratio during the period from the past to the present time) determined by the time series data analysis unit 33 and external factors that affect the changes in the composition ratio during the prediction period from a present time to a future time, and predicts a composition ratio during the prediction period based on the prediction curve.

[0094] In particular, the types and quantities of used household appliances vary depending on the season (spring, summer, autumn, winter). For example, used air conditioners tend to increase in the summer. Since the types and quantities of used household appliances vary depending on the season, the composition ratio between the plastic particles 11a, 11b fed into the electrostatic separation device also varies.

[0095] The composition ratio prediction unit 72 extracts in advance the external factors related to the changes in the composition ratio (%) between the plastic particles 11a, 11b based on the smooth curve generated by the time series data analysis unit 33, and generates a prediction model.

[0096] Conceivable examples of external factors influencing the changes in the composition ratio (%) between plastic particles 11a, 11b are (a) changes in the number of household appliances brought to recycling facilities, (b) changes in the number of household appliances dismantled at recycling facilities, which depend on the types of household appliances, and (c) season and weather conditions.

[0097] The composition ratio prediction unit 72 analyzes the correlation between various external factors such as (a) to (c) described above and the slope of the smooth curve, and extracts the external factors that have a certain correlation with the slope of the smooth curve. The composition ratio prediction unit 72 then generates a prediction model based on the extracted external factors and determines, based on the prediction model, whether the composition ratio (%) between the plastic particles 11a, 11b shows an increasing trend, a decreasing trend, or no increase or decrease during the prediction period.

[0098] As an example, consider a case where a certain correlation is found between the past smooth curve and the external factor (b) described above when constructing a forecast model for a certain period in August. In this case, if it is found that the number of a certain type of household appliances being dismantled tends to increase during a certain period in August each year, it is determined from the correlation that the composition ratio (%) between plastic particles 11a, 11b also tends to increase during this period.

[0099] The composition ratio prediction unit 72 then generates a prediction curve representing changes in the composition ratio from a current time tN (corresponding to the time tN in Fig. 3) until a future time tM shows how in Fig. 11, based on the generated prediction model and the smooth curve. Here, M is an integer greater than N. Time tM is desirably a time at which a measured value of a new composition ratio should be added next.

[0100] The composition ratio prediction unit 72 then generates a predicted value of the composition ratio (%) between the plastic particles 11a, 11b based on the prediction curve. The predicted value of the composition ratio (%) is desirably an average of the estimated value of the composition ratio (%) at time tN and the predicted value of the composition ratio (%) at time tM.

[0101] Again referring to Fig. 10, the composition ratio prediction value output unit 73 outputs the composition ratio prediction value (%) generated by the composition ratio prediction unit 72 to the position control unit 40. The position control unit 40 controls the respective positions of the partition plates 8, 9 based on the composition ratio prediction value (%) instead of a composition ratio estimate value (%).

[0102] Fig. 12 is a flowchart showing an example method of the Fig. 10 shows the composition ratio data analysis unit 71. Fig. 12 is used for comparison with Fig. 4, which is described in Embodiment 1.

[0103] With reference to Fig. 12, steps S11 to S13 are the same as steps S1 to S3 of Fig.4. In step S11, the time series data generation unit 32 generates time series data on the composition ratio. In step S12, the time series data analysis unit 33 analyzes a change trend in the composition ratio of the time series data. In step S13, the time series data analysis unit 33 extracts a smooth curve.

[0104] Then, in the composition ratio data analysis unit 71 in Embodiment 2, in step S14, the composition ratio prediction unit 72 generates a prediction model based on the smooth curve and the external factor that affects the change tendency of the composition ratios. In step S15, the composition ratio prediction unit 72 generates a prediction curve based on the smooth curve and the prediction model, and predicts a future composition ratio based on the prediction curve. In step S16, the composition ratio prediction value output unit 73 outputs a predicted value of the future composition ratio.

[0105] Since the partition plates 8, 9 are moved after predicting a future composition ratio in Embodiment 2 as described above, each of the partition plates 8, 9 can be moved to an optimal position even if the composition ratio (%) between the plastic particles 11a, 11b has varied before each of the partition plates 8, 9 is moved next time.

[0106] It is to be understood that the embodiments disclosed herein are illustrative and not restrictive in any respect. Therefore, the technical scope of the present invention is intended to be defined by the claims, not only by the embodiments described above, and to include all modifications and variations that come within the meaning and scope of the claims. EXPLANATION OF REFERENCE SYMBOLS 1 loading unit 1a container 1b Charge feeding device 2 electrification drums 3 Vibration feeding device 4, 5 Electrode 6 DC power supply 7 collection containers 7a-7c Collection chamber 8, 9 dividing plate 10, 70 Control 11 Plastic particle group 11a, 11b Plastic particles 20 Composition ratio detection unit 21 samplers 22 Composition ratio measuring device 30, 71 Composition ratio data analysis unit 31 Composition ratio storage unit 32 Time series data generation unit 33 Time Series Data Analysis Unit 34 Output unit for composition ratio estimates 40 Position control unit 41 Case distribution storage unit 42 Case Distribution Prediction Unit 43 Purity specification unit 44 Position calculation unit 45 Position output unit 46 Drive unit 51, 52 rail 53, 54 slider 55, 56 Ball screw 57, 58 Engine 60 processing circuit 61 processor 62 storage 63 Hardware 72 Composition ratio prediction unit 73 Output unit for composition ratio prediction values.

Claims

[1] Electrostatic separator comprising: an electrification device (2) for moving a separation target material (11) containing a plurality of types of separation target particles (11a, 11b) to electrify each separation target particle (11a, 11b), the plurality of types of separation target particles (11a, 11b) having different electrification properties; an electric field generator (4, 5, 6) for applying the electric field to each separation target particle (11a, 11b) electrified by the electrification device (2) and causing each separation target particle (11a, 11b) to fall into a position corresponding to an electrified state of each separation target particle (11a, 11b); a collecting container (7) for collecting each separation target particle (11a, 11b) that has fallen from the electric field generator (4, 5, 6); a separating element (8, 9) which is movable in a predetermined direction in the collecting container (7), wherein the separating element (8, 9) divides the collecting container (7) into a plurality of collecting chambers (7a, 7b, 7c); and a controller (10) for controlling a position of the separation element (8, 9) in the predetermined direction, wherein the controller (10) samples a part of the separation target material (11) and measures a composition ratio between the plurality of types of separation target particles (11a, 11b), analyzes a change tendency of the composition ratio from a previously measured composition ratio and a currently measured composition ratio and controls the position of the separating element (8, 9) on the basis of an analysis result of the change tendency, wherein the controller (10) comprises: a measuring unit (20) for taking samples from a part of the separation target material (11) and measuring the composition ratio, a time series data generation unit (32) for arranging the previously measured composition ratio and the composition ratio currently measured by the measuring unit (20) in the order of measurement to generate time series data on the composition ratio, a time series data analysis unit (33) for analyzing the change tendency of the composition ratio from the time series data and estimating a current composition ratio estimate based on the analysis result of the change tendency, and a position control unit (40) for controlling the position of the separating element (8, 9) in the predetermined direction based on the current composition ratio estimated by the time series data analysis unit (33). [2] The electrostatic separator according to claim 1, wherein the time series data analysis unit (33) extracts a smooth curve showing the change tendency of the composition ratio of the time series data by a state space model, and estimates the current composition ratio estimated value based on the extracted smooth curve. [3] An electrostatic separator according to claim 1 or 2, wherein the position control unit (40) comprises: a case distribution storage unit (41) for storing a relationship between a plurality of composition ratios and Fall distributions, wherein the fall distributions are distributions of the plurality of types of separation target particles (11a, 11b) falling into the collecting container (7), a case distribution prediction unit (42) for predicting the case distributions of the plurality of types of separation target particles (11a, 11b) into the collection container (7) on the basis of the current composition ratio estimate value estimated by the time series data analysis unit (33) and a storage content of the case distribution storage unit (41), a position calculation unit (44) for determining the position of the separation element (8, 9) in the predetermined direction based on a result of the prediction by the case distribution prediction unit (42) and purity target values ​​in the plurality of collection chambers (7a, 7b, 7c) of the plurality of types of separation target particles (11a, 11b) collected in the plurality of collection chambers (7a, 7b, 7c), and a drive unit (46) for moving the separating element (8, 9) to the position determined by the position calculation unit (44). [4] Electrostatic separator comprising: an electrification device (2) for moving a separation target material (11) containing a plurality of types of separation target particles (11a, 11b) to electrify each separation target particle (11a, 11b), the plurality of types of separation target particles (11a, 11b) having different electrification properties; an electric field generator (4, 5, 6) for applying the electric field to each separation target particle (11a, 11b) electrified by the electrification device (2) and causing each separation target particle (11a, 11b) to fall into a position corresponding to an electrified state of each separation target particle (11a, 11b); a collecting container (7) for collecting each separation target particle (11a, 11b) that has fallen from the electric field generator (4, 5, 6); a separating element (8, 9) which is movable in a predetermined direction in the collecting container (7), wherein the separating element (8, 9) divides the collecting container (7) into a plurality of collecting chambers (7a, 7b, 7c); and a controller (70) for controlling a position of the separating element (8, 9) in the predetermined direction, wherein the controller (70) sampling a portion of the separation target material (11) and measuring a composition ratio between the plurality of types of separation target particles (11a, 11b), a change trend of the composition ratio from a previously measured composition ratio and currently measured composition ratio is analyzed and controls the position of the separating element (8, 9) on the basis of an analysis result of the change tendency, wherein the control (70) comprises: a measuring unit (20) for taking samples from a part of the separation target material and measuring the composition ratio, a time series data generation unit (32) for arranging the previously measured composition ratio and the composition ratio currently measured by the measuring unit in the order of measurement to generate time series data on the composition ratio, a time series data analysis unit (33) for analyzing the change tendency of the composition ratio from the time series data, a composition ratio prediction unit (72) for predicting the change tendency of the composition ratio during a prediction period from a current time to a future time based on the analysis result of the change tendency and an external factor that influences the changes in the composition ratio during the prediction period, and predicting the composition ratio during the prediction period based on a prediction result of the change tendency, and a position control unit (40) for controlling the position of the separating element (8, 9) in the predetermined direction based on the composition ratio predicted by the composition ratio prediction unit (72). [5] Electrostatic separator according to claim 4, wherein the time series data analysis unit (33) extracts a smooth curve showing the change tendency of the composition ratio from the time series data through a state space model, and the composition ratio prediction unit (72) generates a prediction curve showing the change tendency of the composition ratio during the prediction period based on the smooth curve and the external factor, and predicts the composition ratio during the prediction period based on the generated prediction curve. [6] An electrostatic separator according to claim 4 or 5, wherein the position control unit (40) comprises: a case distribution storage unit (41) for storing a relationship between a plurality of composition ratios and Fall distributions, wherein the fall distributions are distributions of the plurality of types of separation target particles (11a, 11b) falling into the collecting container (7), a falling distribution prediction unit (42) for predicting the falling distributions of the plurality of types of separation target particles (11a, 11b) into the collecting container (7) on the basis of the composition ratio predicted by the composition ratio prediction unit (72) and a storage content of the falling distribution storage unit (41), a position calculation unit (44) for determining the position of the separation element (8, 9) in the predetermined direction based on a result of the prediction by the case distribution prediction unit (42) and purity target values ​​of the plurality of types of separation target particles (11a, 11b) collected in the plurality of collection chambers (7a, 7b, 7c), and a drive unit (46) for moving the separating element (8, 9) to the position determined by the position calculation unit (44). [7] An electrostatic separator according to any one of claims 1 to 6, wherein the electric field generator (4, 5, 6) comprises: a first electrode (4) and a second electrode (5) arranged in the predetermined direction and facing each other, and a DC power supply (6) for applying a DC voltage between the first electrode (4) and the second electrode (5) to generate the electric field between the first electrode (4) and the second electrode (5), wherein the electrification device (2) causes each electrified separation target particle (11a, 11b) to fall between the first electrode (4) and the second electrode (5), and the collecting container (7) collects each separation target particle (11a, 11b), which has fallen from the electrification device (2) between the first electrode (4) and the second electrode (5).

Images