22 to thereby cause the inverter circuit 22 to perform switching. Consequently, the current controller 25 applies, to the coil 21, a voltage for causing a current of a desired current value to flow through the coil 21. The current controller 25 may calculate the voltage value of the voltage to be applied to the coil 21 by performing proportional integral differential (PID) control of the voltage to be applied to the coil 21 on the basis of a deviation between the current command value and the coil actual current value.

A disposition interval of the plurality of coils 21 in the traveling direction of the carrier 17 is denoted by L_coil. It can also be said that L_coilis a distance between the center positions of the coils 21 adjacent to each other in the conveyance path unit 11. L_carrieris the length of the carrier 17 in the traveling direction of the carrier 17. L_coilis shorter than L_carrier. Consequently, each carrier 17 can obtain a thrust by an interaction between magnetic fluxes generated by two or more coils 21.

The length of the permanent magnet 30 in the traveling direction of the carrier 17 is denoted by L_magnet. L_magnetis a length from one end of the permanent magnet 30 to another end of the permanent magnet 30 in the traveling direction of the carrier 17. In a case where N poles and S poles are alternately disposed as illustrated in FIG. 2, L_magnetis the length of the entire permanent magnet 30 including all magnetic poles in the traveling direction of the carrier 17. In a case where there is a space between a magnetic pole and a magnetic pole, L_magnetalso includes the length of the space.

L_magnetis shorter than L_carrier. Since L_magnetis shorter than L_carrier, in a case where two carriers 17 approach each other, a space is provided between the permanent magnet 30 of one carrier 17 and the permanent magnet 30 of the other carrier 17. Since the space is provided between the permanent magnet 30 of one carrier 17 and the permanent magnet 30 of the other carrier 17, it is possible to avoid a state in which the permanent magnet 30 of one carrier 17 and the permanent magnet 30 of the other carrier 17 are present on one coil 21. Since the permanent magnet 30 present on one coil 21 is the permanent magnet 30 of one carrier 17, a calculation of a current command for generating a magnetic flux in one coil 21 can be set as a calculation for one carrier 17. On the other hand, in a case where it is necessary to set the calculation of the current command for generating the magnetic flux in one coil 21 as the calculation for two carriers 17, the calculation of the current command becomes complicated.

According to the first embodiment, since L_magnetis shorter than L_carrier, the conveyance system 1 can set the calculation of the current command for generating the magnetic flux in one coil 21 as the calculation for one carrier 17. Therefore, the conveyance system 1 can prevent the calculation of the current command from becoming complicated.

The conveyance path unit 11 includes a linear scale 26, a processor 28, and a communication slave station 29. The linear scale 26 is a sensing unit that senses a position of the carrier 17 on the conveyance path unit 11. The linear scale 26 is provided on the conveyance path 10 by the plurality of conveyance path units 11 coupled to each other to form the conveyance path 10. The processor 28 is a central processing unit (CPU). The processor 28 may be an arithmetic device, a processing device, a microprocessor, a microcomputer, or a digital signal processor (DSP).

The linear scale 26 includes a plurality of position sensors 27. Each position sensor 27 is a sensor that detects a magnetic field, such as a Hall sensor or a magnetoresistive sensor. Each position sensor 27 detects a magnetic field of the permanent magnet 30 or a magnetic field of the permanent magnet 31. Here, the position sensor 27 is a Hall sensor on which two Hall elements are mounted. An interval between the two Hall elements is an interval corresponding to half of a magnetic pole pitch of the permanent magnet 31. Each Hall element converts a magnetic field into an electrical signal and outputs the electrical signal. The electrical signal output from each Hall element changes with the movement of the carrier 17. The electrical signal output from one Hall element has a waveform of a sin wave. The electrical signal output from the other Hall element has a waveform of a cos wave.

The electrical signal from each position sensor 27 of the linear scale 26 is input to the processor 28. An analog to digital (AD) converter included in the processor 28 detects the sin wave and the cos wave. The processor 28 detects the position of the carrier 17 with respect to the position sensor 27 by calculating arctan on the basis of information on the sin wave and information on the cos wave. Consequently, the processor 28 acquires position sensor information indicating the relative position of the carrier 17 with respect to the position sensor 27. In FIG. 2, a communication line of the electrical signal between each position sensor 27 and the processor 28 is not illustrated.

The communication slave station 29 is a communication slave station on a side of the conveyance path unit 11. The data communication line 15 is connected to the communication slave station 29. In a case where each conveyance path unit 11 and the track controller 20 are connected by the daisy chain connection, the communication slave station 29 is configured to be able to connect two lines constituting the data communication lines 15. For each of the plurality of coils 21 included in the conveyance path unit 11, the communication slave station 29 receives, from the track controller 20, a current command indicating a command value of a current to be caused to flow through the coil 21. The communication slave station 29 transmits the current command to each of the plurality of current controllers 25 of the conveyance path unit 11. Consequently, the conveyance path unit 11 controls the current flowing through each of the plurality of coils 21 in accordance with the current command.

The communication slave station 29 acquires the position sensor information from the processor 28. The communication slave station 29 transmits the acquired position sensor information to the track controller 20.

The communication slave station 29 performs, for example, fixed-cycle communication in which a current command is received and position sensor information is transmitted in a fixed-cycle. The communication slave station 29 may aperiodically perform the reception of the current command and the transmission of the position sensor information instead of the fixed-cycle communication.

As described above, the conveyance path unit 11 mainly has a function of performing energization control of the coils 21 and a function of acquiring position sensor information. Each of the plurality of conveyance path units 11 constituting the conveyance path 10 similarly performs energization control of the coils 21 and similarly acquires position sensor information.

Next, a configuration of the track controller 20 will be described. FIG. 3 is a diagram illustrating an exemplary configuration of the track controller 20 included in the conveyance system 1 according to the first embodiment. The track controller 20 includes a carrier position controller 41, a current command generator 42, a position information generator 43, a communication slave station 44, and a communication master station 45.

The communication slave station 44 is a communication slave station on a track controller 20 side. The communication slave station 44 receives a position command from the motion controller 19. In the first embodiment, the communication slave station 44 receives position commands for the carriers 17A, 17B, and 17C of the conveyance path units 11. The communication slave station 44 receives a position command for each carrier 17 of the conveyance path units 11, and outputs the received position command to the carrier position controller 41.

The communication master station 45 is a communication master station on the track controller 20 side. The communication master station 45 receives the position sensor information from the communication slave station 29 of each conveyance path unit 11. That is, the communication master station 45 receives the position sensor information acquired by the processor 28 of each conveyance path unit 11. The communication master station 45 outputs the received position sensor information to the position information generator 43.

The position information generator 43 acquires the position sensor information from each conveyance path unit 11 and calculates the position of each carrier 17 on the basis of the acquired position sensor information. The position information generator 43 generates position information indicating the actual position of each carrier 17 on the conveyance path 10. In the first embodiment, the position information generator 43 generates the position information of the carriers 17A, 17B, and 17C indicating the actual positions of the carriers 17A, 17B, and 17C on the conveyance path 10. The position information generator 43 generates the position information of each carrier 17 of the conveyance system 1, and outputs the generated position information to the carrier position controller 41 and the current command generator 42.

The carrier position controller 41 acquires the position command of each carrier 17 and the position information of each carrier 17. The carrier position controller 41 generates a thrust command of each carrier 17 on the basis of a difference between the position command and the position information. In the first embodiment, the carrier position controller 41 generates the thrust command of the carrier 17A on the basis of a difference between the position command of the carrier 17A and the position information of the carrier 17A. The carrier position controller 41 generates a thrust command of the carrier 17B on the basis of a difference between the position command of the carrier 17B and the position information of the carrier 17B. The carrier position controller 41 generates a thrust command of the carrier 17C on the basis of a difference between the position command of the carrier 17C and the position information of the carrier 17C. The carrier position controller 41 outputs the generated thrust commands to the current command generator 42.

The current command generator 42 acquires the thrust command of each carrier 17 and the position information of each carrier 17. The current command generator 42 generates current commands for controlling a current flowing through the plurality of coils 21 on the basis of the thrust commands and the position information. In each control cycle when a current command is generated in the track controller 20, the current command generator 42 generates current commands for performing current control of all of the plurality of coils 21 of each conveyance path unit 11. Here, regarding the “control cycle”, one control cycle is from the generation of the position information of the carrier 17 by the position information generator 43 to the generation of the current commands for the coils 21 of each conveyance path unit 11 by the current command generator 42. That is, the current command generator 42 of the first embodiment generates current commands for all of the plurality of coils 21 of each conveyance path unit 11 in each control cycle when a current command is generated. The current command generator 42 outputs the generated current commands to the communication master station 45. The communication master station 45 transmits the current commands to the communication slave station 29 of each conveyance path unit 11. In the following description, a group of current commands, each of the current commands being a current command for one of the plurality of coils 21 included in the conveyance path unit 11, is referred to as a bundle of current commands. The communication master station 45 transmits the bundle of current commands to the communication slave station 29 of each conveyance path unit 11.

The current commands generated by the current command generator 42 include a current command in a case where the current flowing through the coil 21 is set to zero. Regarding the expression “generates current commands for performing current control of all of the plurality of coils 21 of each conveyance path unit 11”, a case is included where a current command value for at least one of the plurality of coils 21 is set to zero.

Next, details of processes by the current command generator 42 will be described. FIG. 4 is a flowchart illustrating a processing procedure by the current command generator 42 of the conveyance system 1 according to the first embodiment. FIG. 4 illustrates a procedure of processes executed by the current command generator 42 for each control cycle.

In step S1, the current command generator 42 calculates a bundle of current commands I_condAon the basis of the thrust command of the carrier 17A and the position information of the carrier 17A. The bundle of current commands I_condAis a group of current commands for generating a thrust for moving the carrier 17A. The current command generator 42 generates the bundle of current commands I_condAin which current commands for all the coils 21 of each conveyance path unit 11 are bundled.

In step S2, the current command generator 42 calculates a bundle of current commands I_condBon the basis of the thrust command of the carrier 17B and the position information of the carrier 17B. The bundle of current commands I_condBis a group of current commands for generating a thrust for moving the carrier 17B. The current command generator 42 generates the bundle of current commands I_cmdBin which current commands for all the coils 21 of each conveyance path unit 11 are bundled.

In step S3, the current command generator 42 calculates a bundle of current commands I_condCon the basis of the thrust command of the carrier 17C and the position information of the carrier 17C. The bundle of current commands I_condCis a group of current commands for generating a thrust for moving the carrier 17C. The current command generator 42 generates the bundle of current commands I_condCin which current commands for all the coils 21 of each conveyance path unit 11 are bundled.

In step S4, the current command generator 42 calculates a current command for each coil 21 of each conveyance path unit 11 by using the bundles of current commands I_cmdA, I_cmdB, and I_cmdCcalculated in steps S1 to S3. The current command generator 42 adds up current commands for the same coil 21 among the bundles of current commands I_cmdA, I_cmdBand I_cmdCto calculate a current command I_totfor each of the coils 21 of each conveyance path unit 11. As described above, the current command generator 42 obtains the current commands for respective carriers 17 for each of the coils 21, and adds up the current commands for respective carriers 17 for each of the coils 21 to generate the current command for each coil 21.

The current command generator 42 generates a bundle of current commands I_cmdin which current commands I_totfor all the coils 21 of each conveyance path unit 11 are bundled. The bundle of current commands I_cmdis a group of current commands for generating thrusts for moving all the carriers 17A, 17B, and 17C of the conveyance system 1. The current command generator 42 outputs the generated bundle of current commands I_cmd. Thus, the current command generator 42 ends the processes in accordance with the procedure illustrated in FIG. 4. The current command generator 42 repeats the processes in accordance with the procedure illustrated in FIG. 4 for each control cycle.

Here, an example of a method for calculating the bundles of current commands I_cmdA, I_cmdBand I_cmdCby the current command generator 42 will be described. In the following description, a current command for each of the coils 21 of each conveyance path unit 11 for generating the thrust for moving the carrier 17A is expressed as “I_cmdA_ Aa” or the like. “Aa” in “I_cmdA_Aa” indicates that it is the current command for the coil 21a of the conveyance path unit 11A. For each coil 21 other than the coil 21a of the conveyance path unit 11A, the current command for each of the coils 21 is expressed in the same manner as in the case of the coil 21a of the conveyance path unit 11A.

FIG. 5 is a diagram illustrating an example of a relationship between a position of the carrier 17 and a thrust constant of the coil 21 in the first embodiment. FIG. 5 illustrates a graph indicating a relationship between a distance x based on the center position of the coil 21 in the traveling direction of the carrier 17 and a thrust constant k(x) of the coil 21. The thrust constant k(x) indicates a ratio of a thrust received by the carrier 17 to a current flowing through the coil 21. In FIG. 5, the vertical axis represents the thrust constant k(x), and the horizontal axis represents the distance x. In FIG. 5, the unit of the thrust constant k(x) is N/A. In FIG. 5, the unit of the distance x is any unit. In the following description, the center position of the coil 21 is the center position of the coil 21 in the traveling direction of the carrier 17. The center position of the carrier 17 is the center position of the carrier 17 in the traveling direction of the carrier 17.

FIG. 5 illustrates, as a reference, the carrier 17 when a front end of the carrier 17 in the traveling direction of the carrier 17 coincides with the center position of the coil 21. As the center position of the carrier 17 approaches the center position of the coil 21, a larger thrust can be generated with a smaller current by a relationship between the position of the coil 21 and the phase of the permanent magnet 30. The magnitude of the thrust varies depending on the relationship between the position of the coil 21 and the phase of the permanent magnet 30.

When the distance x from the center position of the coil 21 to the center position of the carrier 17 is longer than L_carrier/2 which is a length from the front end of the carrier 17 to the center position of the carrier 17, the thrust constant becomes zero. That is, when the distance x is longer than L_carrier/2, even if a current is caused to flow through the coil 21, the thrust that can be applied to the carrier 17 is not generated. Note that the graph illustrated in FIG. 5 is an example. The relationship between the position of the carrier 17 and the thrust constant of the coil 21 varies depending on how the permanent magnet 30 is disposed.

In the following description, the thrust constant k(x) for each of the coils 21 of each conveyance path unit 11 is expressed as “kAa(x)” or the like. “Aa” in “kAa(x)” indicates that it is the thrust constant k(x) of the coil 21a of the conveyance path unit 11A. For each coil 21 other than the coil 21a of the conveyance path unit 11A, the thrust constant k(x) for each of the coils 21 is expressed in the same manner as in the case of the coil 21a of the conveyance path unit 11A.

In addition, in the following description, the center position of each of the coils 21 of each conveyance path unit 11 is expressed as “pAa” or the like. “Aa” in “pAa” indicates that it is the center position of the coil 21a of the conveyance path unit 11A. For each coil 21 other than the coil 21a of the conveyance path unit 11A, the center position for each of the coils 21 is expressed in the same manner as in the case of the coil 21a of the conveyance path unit 11A.

Current commands I_cmdA_Aa, I_cmdA_Ab, . . . , I_cmdA_Hh, and I_cmdA_Hi for generating the thrust for moving the carrier 17A, each for one of the coils 21 of each conveyance path unit 11, are expressed by the following formulas. Note that τA represents a thrust command τ of the carrier 17A. xA is a distance x from the center position of the coil 21 to the center position of the carrier 17A, and represents the actual position of the carrier 17A with respect to the coil 21.

$I_{cmdA}_Aa = kAa (xA - pAa) \times τ A / {{kAa (xA - pAa)}^{2} + {kAb (xA - pAb)}^{2} + \dots + {kHh (xA - pHh)}^{2} + {kHi (xA - pHi)}^{2}}$

$I_{cmdA}_Ab = kAb (xA - pAb) \times τ A / {{kAa (xA - pAa)}^{2} + {kAb (xA - pAb)}^{2} + \dots + {kHh (xA - pHh)}^{2} + {kHi (xA - pHi)}^{2}}$

$\dots$

$I_{cmdA}_Hh = kHh (xA - pHh) \times τ A / {{kAa (xA - pAa)}^{2} + {kAb (xA - pAb)}^{2} + \dots + {kHh (xA - pHh)}^{2} + {kHi (xA - pHi)}^{2}}$

$I_{cmdA}_Hi = kHi (xA - pHi) \times τ A / {{kAa (xA - pAa)}^{2} + {kAb (xA - pAb)}^{2} + \dots + {kHh (xA - pHh)}^{2} + {kHi (xA - pHi)}^{2}}$

In general, in a case where the thrust command τ and the thrust constant k(x) are given, there are an infinite number of sets of current commands of each coil 21 that realize the given thrust command T. A set of the current commands I_cmdA_Aa, I_cmdA_Ab, . . . , I_cmdA_Hh, and I_cmdA_Hi obtained by the above formula is a set that minimizes the sum of squares of the current flowing through the coils 21. That is, according to the above formula, the current commands I_cmdA_Aa, I_cmdA_Ab, . . . , I_cmdA_Hh, and I_cmdA_Hi that can minimize the copper loss of the coils 21 are obtained.

The current command generator 42 bundles the current commands I_cmdA_Aa, I_cmdA_Ab, . . . , I_cmdA_Hh, and I_cmdA_Hi to thereby generate the bundle of current commands I_cmdAfor the carrier 17A.

Also for the bundle of current commands I_cmdBfor the carrier 17B and the bundle of current commands I_cmdCfor the carrier 17C, the current command generator 42 performs calculation similarly to the case of the bundle of current commands I_cmdAfor the carrier 17A. In the procedure illustrated in FIG. 4, the calculations are performed in the order of the bundle of current commands I_cmdA, the bundle of current commands I_cmdB, and the bundle of current commands I_cmdC, but the bundles of current commands I_cmdA, I_cmdB, and I_cmdCare calculated in any order.

Next, an example of a method for calculating the current command I_totfor each of the coils 21 of each conveyance path unit 11 for generating thrusts for moving the carriers 17A, 17B, and 17C will be described. The current command generator 42 calculates the current command I_totfor each of the coils 21 of each conveyance path unit 11 by using the bundles of current commands I_cmdA, I_cmdB, and I_cmdC. In the following description, the current command I_totfor each of the coils 21 of each conveyance path unit 11 is expressed as “I_tot_Aa” or the like. “Aa” in “I_tot_Aa” indicates that it is the current command I_totfor the coil 21a of the conveyance path unit 11A. For each coil 21 other than the coil 21a of the conveyance path unit 11A, the current command I_totfor each of the coils 21 is expressed in the same manner as in the case of the coil 21a of the conveyance path unit 11A.

The current commands I_tot,_Aa, I_tot_Ab, . . . , I_tot_Hh, and I_tot_Hi, each for one of the coils 21 of each conveyance path unit 11, are expressed by the following formulas.

$I_{tot}_Aa = I_{cmdA}_Aa + I_{cmdB}_Aa + I_{cmdC}_Aa$

$I_{tot}_Ab = I_{cmdA}_Ab + I_{cmdB}_Ab + I_{cmdC}_Ab$

$\dots$

$I_{tot}_Hh = I_{cmdA}_Hh + I_{cmdB}_Hh + I_{cmdC}_Hh$

$I_{tot}_Hi = I_{cmdA}_Hi + I_{cmdB}_Hi + I_{cmdC}_Hi$

FIG. 6 is a diagram illustrating examples of calculation results of current commands by the conveyance system 1 according to the first embodiment. FIG. 6 illustrates examples of calculation results of the current commands I_tot_Aa, I_tot_Ab, . . . , I_tot_Hh, and I_tot_Hi by the above-described calculation method.

As illustrated in FIG. 5, when the distance x from the center position of the coil 21 to the center position of the carrier 17 is in the vicinity of L_carrier/2, the thrust constant k(x) becomes nearly zero, and the current flowing through the coil 21 hardly contributes to the generation of the thrust of the carrier 17. Therefore, for example, in a case where I_cmdA_Aa is a value other than zero, each of I_cmdA_Aa and I_cmdA_Aa becomes nearly zero, and I_tot_Aa=I_cmdA_Aa substantially holds. As described above, it can be said that the number of carriers 17 on which each coil 21 can apply the thrust is substantially one, and one coil 21 does not simultaneously apply the thrust on two or more carriers 17.

In addition, when the conveyance system 1 is in the state illustrated in FIG. 1, there is no presence of the coil 21 of which the distance x is shorter than L_carrier/2 in the conveyance path unit 11B, for example. With the coil 21d which is one of the plurality of coils 21 provided in the conveyance path unit 11B as an example, a current command I_tot_Bd of the coil 21d is calculated as follows.

$I_{tot}_Bd = I_{cmdA}_Bd + I_{cmdB}_Bd + I_{cmdC}_Bd = 0 \times τ A / {{kAa (xA - pAa)}^{2} + {kAb (xA - pAb)}^{2} + \dots + {kHh (xA - pHh)}^{2} + {kHi (xA - pHi)}^{2}} + 0 \times τ B / {{kAa (xA - pAa)}^{2} + {kAb (xA - pAb)}^{2} + \dots + {kHh (xA - pHh)}^{2} + {kHi (xA - pHi)}^{2}} + 0 \times τ C / {{kAa (xA - pAa)}^{2} + {kAb (xA - pAb)}^{2} + \dots + {kHh (xA - pHh)}^{2} + {kHi (xA - pHi)}^{2}} = 0$

As described above, in a case where there is no presence of the coil 21 of which the distance x is shorter than L_carrier/2, the current command value for the coil 21 is calculated to be zero. That is, the current command generator 42 generates a current command in which the current command value is set to zero for the coil 21 at a position other than the position where the thrust can be applied to the carrier 17.

As described above, the current command generator 42 generates the current commands I_tot_Aa, I_tot_Ab, . . . , I_tot_Hh, and I_tot_Hi for all of the coils 21a, 21b, . . . , and 21i of the conveyance path units 11A, 11B, . . . , and 11H. The current command generator 42 generates current commands for performing current control of all of the plurality of coils 21 of each conveyance path unit 11 in all control cycles when controlling the plurality of conveyance path units 11.

The current command generator 42 generates the bundle of current commands I_cmdin which the current commands I_tot_Aa, I_tot_Ab, . . . , I_tot_Hh, and I_tot_Hi are bundled. The current command generator 42 outputs the generated bundle of current commands I_cmdto the communication master station 45. For example, the communication master station 45 transmits the current commands I_tot_Aa, I_tot_Ab, . . . , and I_tot_Ai out of the bundle of current commands I_cto the communication slave station 29 of the conveyance path unit 11A. The communication master station 45 transmits the current commands I_tot_Ha, I_totHb, . . . , and I_tot_Hi out of the bundle of current commands I_cmdto the communication slave station 29 of the conveyance path unit 11H. As described above, the communication master station 45 transmits the current commands to the communication slave station 29 of each conveyance path unit 11.

According to the first embodiment, the current command generator 42 of the conveyance system 1 generates current commands for performing current control of all of the plurality of coils 21 of each conveyance path unit 11 in each control cycle when controlling the plurality of conveyance path units 11. The conveyance system 1 sends current commands not only for the coil 21 at a position where the thrust can be applied to the carrier 17 but also for all the coils 21 including the coil 21 at a position other than the position where the thrust can be applied to the carrier 17. Since the conveyance system 1 does not require a switch that performs switching between supply of the current to the coils 21 and interruption of the current flowing to the coils 21, a circuit configuration can be simplified. In addition, since the conveyance system 1 does not need to generate and output an opening/closing command for the switch separately from the current commands, it is possible to perform control by a simple program.

In a case where the coil 21 to be supplied with the current is selected from the plurality of coils 21 of each conveyance path unit 11, the current command is calculated only for the selected coil 21. On the other hand, in the case of the first embodiment, since the current command generator 42 uniformly generates the current commands for all the coils 21 of each conveyance path unit 11, a procedure of selecting the coil 21 and calculating the current command only for the selected coil 21 is unnecessary. Consequently, the conveyance system 1 can control each conveyance path unit 11 by a simple procedure.

The current command generator 42 generates a current command in which the current command value is set to zero for the coil 21 at a position other than the position where the thrust can be applied to the carrier 17. Even in a case where an induced current can be generated in the coil 21 at a position other than the position where the thrust can be applied to the carrier 17, the conveyance system 1 can perform adjustment to make the current zero by canceling the induced current. In the conveyance system 1, it is unnecessary to perform proof by cases for processes depending on the presence or absence of the carrier 17 at the position where the thrust can be applied thereto, and the procedure of the processes can be simplified.

Thus, the conveyance system 1 achieves an effect that a circuit configuration can be simplified and control can be performed by simple processes.

Second Embodiment

In the first embodiment, the current command generator 42 obtains the current commands for respective carriers 17 for each of the coils 21, and adds up the current commands for respective carriers 17 for each of the coils 21 to generate the current command for each coil 21. In a second embodiment, the current command generator 42 selects, for each coil 21, one carrier 17 closest to the coil 21 among the plurality of carriers 17, and obtains a current command that applies a thrust to the selected carrier 17, thereby generating a current command for each coil 21. In the second embodiment, the same components as those in the first embodiment are denoted by the same reference numerals as those therein, and configurations different from those in the first embodiment will mainly be described.

FIG. 7 is a diagram illustrating an exemplary configuration of the conveyance path unit 11 included in the conveyance system 1 according to the second embodiment. FIG. 7 illustrates a part of the configuration of the conveyance path unit 11 and two carriers 17A and 17B. FIG. 7 illustrates a case where two carriers 17A and 17B are present on the conveyance path unit 11A which is one of the plurality of conveyance path units 11 of the conveyance system 1. In the conveyance system 1, in addition to the case where two carriers 17 are present on one conveyance path unit 11 as illustrated in FIG. 7, there are a case where one carrier 17 is present on one conveyance path unit 11 and a case where no carrier 17 is present on one conveyance path unit 11.

In the state illustrated in FIG. 7, the carrier 17 located at the closest position from the center position of the coil 21a in the traveling direction of the carrier 17 is the carrier 17A. With the use of TA which is the thrust command τ for the carrier 17A and xA which is the distance x for the carrier 17A, the current command generator 42 calculates the current command I_tot_Aa for the coil 21a of the conveyance path unit 11A in accordance with the following formula.

$I_{tot}_Aa = I_{cmdA}_Aa = kAa (xA - pAa) \times τ A / {{kAa (xA - pAa)}^{2} + {kAb (xA - pAb)}^{2} + \dots + {kHh (xA - pHh)}^{2} + {kHi (xA - pHi)}^{2}}$

In a case different from the case illustrated in FIG. 7, if the carrier 17 is present at a position far away from the center position of the coil 21a, kAa(xA−pAa)=0 holds by the relationship illustrated in FIG. 5. In that case, the coil 21a does not generate the thrust that can be applied to the carrier 17.

In the state illustrated in FIG. 7, for example, the carrier 17 located at the closest position from the center position of the coil 21e in the traveling direction of the carrier 17 is the carrier 17B. With the use of τB which is the thrust command τ for the carrier 17B and xB which is the distance x for the carrier 17B, the current command generator 42 calculates the current command I_tot_Ae for the coil 21e of the conveyance path unit 11A in accordance with the following formula.

$I_{tot}_Ae = I_{cmdA}_Ae = kAe (xB - pAe) \times τ B / {{kAa (xB - pAa)}^{2} + {kAb (xB - pAb)}^{2} + \dots + {kHh (xB - pHh)}^{2} + {kHi (xB - pHi)}^{2}}$

In the state illustrated in FIG. 7, for example, the carrier 17 located at the closest position from the center position of the coil 21d in the traveling direction of the carrier 17 is the carrier 17A. With the use of TA which is the thrust command τ for the carrier 17A and xA which is the distance x for the carrier 17A, the current command generator 42 calculates the current command I_tot_Ad for the coil 21d of the conveyance path unit 11A in accordance with the following formula.

$I_{tot}_Ad = I_{cmdA}_Ad = kAd (xA - pAd) \times τ A / {{kAa (xA - pAa)}^{2} + {kAb (xA - pAb)}^{2} + \dots + {kHh (xA - pHh)}^{2} + {kHi (xA - pHi)}^{2}}$

In the state illustrated in FIG. 7, a part of the carrier 17B is also present on the coil 21d. However, since the distance x from the center position of the coil 21d to the center position of the carrier 17 is longer than the length from the front end of the carrier 17B to the center position of the carrier 17B, the thrust constant of the thrust that can be applied by the coil 21d to the carrier 17B becomes nearly zero in accordance with the relationship illustrated in FIG. 5. Since a magnetic flux generated by the coil 21d hardly contributes to the movement of the carrier 17B, the current command I_cmdA_Ad based on τA and xA for the carrier 17A can be used as the current command I_tot_Ad for the coil 21d as it is.

FIG. 8 is a diagram illustrating an example of a thrust command used to generate a current command for each coil 21 in the conveyance system 1 according to the second embodiment. FIG. 8 illustrates, for each of the coils 21a to 21i in the state illustrated in FIG. 7, the carrier 17 located closest to the center position of the coil 21 and the thrust command τ used for generating the current command. As for the coils 21a to 21d, the carrier 17 located closest to the center position of the coil 21 is the carrier 17A. The thrust command τ used for generating the current command for each of the coils 21a to 21d is TA. As for the coils 21e to 21i, the carrier 17 located closest to the center position of the coil 21 is the carrier 17B. The thrust command τ used for generating the current command for each of the coils 21e to 21i is TB.

As described above, the current command generator 42 selects, for each coil 21, one carrier 17 closest to the coil 21 among the plurality of carriers 17, and obtains a current command that applies a thrust to the selected carrier 17, thereby generating a current command for each coil 21. The current command generator 42 also calculates a current command for each coil 21 of each of the conveyance path units 11B to 11H similarly to the case of the current command for each coil 21 of the conveyance path unit 11A.

The current command generator 42 generates the bundle of current commands I_cmdin which current commands for the coils 21 of each conveyance path unit 11 are bundled. The current command generator 42 outputs the generated bundle of current commands I_cmdto the communication master station 45. For example, the communication master station 45 transmits the current commands I_tot_Aa, I_tot_Ab, . . . , and I_tot_Ai out of the bundle of current commands I_cmdto the communication slave station 29 of the conveyance path unit 11A. The communication master station 45 transmits the current commands I_tot_Ha, I_tot_Hb, . . . , and I_tot_Hi out of the bundle of current commands I_cmdto the communication slave station 29 of the conveyance path unit 11H. As described above, the communication master station 45 transmits the current commands to the communication slave station 29 of each conveyance path unit 11.

According to the second embodiment, the current command generator 42 of the conveyance system 1 generates current commands for performing current control of all of the plurality of coils 21 of each conveyance path unit 11 in each control cycle when controlling the plurality of conveyance path units 11. Consequently, the conveyance system 1 achieves an effect that a circuit configuration can be simplified and control can be performed by simple processes.

Furthermore, the current command generator 42 selects, for each coil 21, one carrier 17 closest to the coil 21 among the plurality of carriers 17, and obtains a current command that applies a thrust to the selected carrier 17. Since the current command generator 42 calculates the current command corresponding to the selected carrier 17 for each coil 21, the amount of calculation can be reduced as compared with the case of calculating the current command corresponding to each of the plurality of carriers 17 for each coil 21. Consequently, the conveyance system 1 can reduce the amount of calculation for control.

Third Embodiment

In a third embodiment, an example will be described in which a thrust command is corrected on the basis of a thrust command correction value and machine learning is applied to calculation of the thrust command correction value. In the third embodiment, the same components as those in the first or second embodiment are denoted by the same reference numerals as those therein, and configurations different from those in the first or second embodiment will mainly be described.

In each conveyance path unit 11, the coils 21 are arranged at regular intervals, but continuity of the arrangement of the coils 21 is interrupted at a connection portion between the conveyance path units 11 in the conveyance path 10. Therefore, at the connection portion between the conveyance path units 11, a cogging torque different from a cogging torque generated between the coils 21 in the conveyance path unit 11 is generated, and the cogging torque fluctuates. Since the alignment of the conveyance path units 11 is often performed by a user and a misalignment between the conveyance path units 11 is likely to occur, it is difficult to assume a correction value for correcting the cogging torque at the connection portion between the conveyance path units 11 in advance. In a case where the correction value is obtained after the assembly of the conveyance path 10 is actually completed, it is necessary to accurately measure the cogging torque at the connection portion between the conveyance path units 11, so that the number of man-hours in the assembly of the conveyance path 10 increases.

In the third embodiment, by calculating the thrust command correction value calculated by machine learning to correct the thrust command, the cogging torque can be corrected with high accuracy, and the number of man-hours in assembling the conveyance path 10 can be reduced. In the third embodiment, the thrust command correction value is a correction value used for correction of the thrust command.

FIG. 9 is a diagram illustrating an exemplary configuration of a track controller 50 included in the conveyance system 1 according to the third embodiment. The track controller 50 includes the current command generator 42, the position information generator 43, the communication slave station 44, the communication master station 45, a carrier position controller 51, a learning device 52, and a learned model storage unit 53.

The carrier position controller 51 acquires a position command of each carrier 17 and position information of each carrier 17. The carrier position controller 51 generates a thrust command of each carrier 17 on the basis of a difference between the position command and the position information. In addition, the carrier position controller 51 acquires a learned model from the learned model storage unit 53, and obtains the thrust command correction value on the basis of the learned model and the position information of each carrier 17. The carrier position controller 51 corrects the thrust command by using the thrust command correction value and outputs the corrected thrust command to the current command generator 42.

The learning device 52 acquires the position information and the thrust command correction value of each carrier 17. The learning device 52 learns a thrust command correction value that enables highly accurate correction of the cogging torque. The learning device 52 outputs a learned model which is a result of the learning. The learned model storage unit 53 stores the learned model.

FIG. 10 is a diagram illustrating an exemplary configuration of the learning device 52 included in the conveyance system 1 according to the third embodiment. The learning device 52 includes a data acquisition unit 61 and a model generation unit 62. The data acquisition unit 61 acquires learning data and creates a data set obtained by combining the learning data. The learning data is position information and a thrust command correction value of each carrier 17. That is, the data acquisition unit 61 acquires learning data including the position information and the thrust command correction value. The data acquisition unit 61 acquires the position information from the position information generator 43. The data acquisition unit 61 acquires the thrust command correction value from the carrier position controller 51.

The model generation unit 62 generates a learned model by using the learning data. The model generation unit 62 generates a learned model to be used for inference of a thrust command correction value from position information on the basis of the learning data. The model generation unit 62 outputs the generated learned model. The learned model is stored in the learned model storage unit 53. Note that the model generation unit 62 may read a learned model that has been already generated from the learned model storage unit 53, and update the learned model by relearning in accordance with the learning data.

As a learning algorithm used by the model generation unit 62, a known algorithm such as supervised learning, unsupervised learning, or reinforcement learning can be used. As an example, a case will be described where reinforcement learning is applied to a learning algorithm used by the model generation unit 62. Reinforcement learning is learning in which a subject of action as an agent in an environment observes the current state and decides what action to take. The agent gets a reward from the environment by selecting an action, and learns a policy with which a maximum reward is obtained through a series of actions. As representative methods of reinforcement learning, Q-learning, TD-learning, and the like are known. For example, in a case of Q-learning, an action-value table which is a general update formula for an action-value function Q(s,a) is expressed by the following formula (1). The action-value function Q(s,a) indicates an action value Q which is a value of an action of selecting an action “a” under an environment “s”.

$\begin{matrix} Formula 1 &  \\ Q (s_{t}, a_{t}) \leftarrow Q (s_{t}, a_{t}) + α (r_{t + 1} + γ \max_{a} Q (s_{t + 1}, a_{t}) - Q (s_{t}, a_{t})) & (1) \end{matrix}$

In the formula (1), “st” represents an environment at time “t”. “a_t” represents an action at time “t”. The action “at” changes the environment to “s_t−1”. “r_t+1” represents a reward given by the change in the environment. “γ” represents a discount rate. “α” represents a learning coefficient. In the third embodiment, the position information is the environment “s_t”. The thrust command correction value is the action “a_t”.

The update formula expressed by the formula (1) increases an action value Q if an action value of best action “a” at time “t+1” is larger than an action value Q of action “a” executed at time “t”, and decreases the action value Q if opposite is true. In other words, the action-value function Q(s,a) is updated so that the action value Q of action “a” at time “t” approaches a best action value at time “t+1”. Consequently, a best action value in a certain environment is sequentially propagated to action values in the previous environments.

The model generation unit 62 includes a reward calculation unit 63 and a function update unit 64. The reward calculation unit 63 calculates a reward on the basis of the data set. The function update unit 64 updates a function for determining a thrust command correction value in accordance with the reward calculated by the reward calculation unit 63.

Specifically, the reward calculation unit 63 calculates a reward “r” on the basis of variability of the speed of the carrier 17. The variability of the speed of the carrier 17 is obtained on the basis of, for example, the position information of the carrier 17. For example, in a case where the variability of the speed of the carrier 17 becomes small, the reward calculation unit 63 increases the reward “r”. The reward calculation unit 63 increases the reward “r” by giving “1” which is a value of the reward. Note that the value of the reward is not limited to “1”. On the other hand, in a case where the variability of the speed of the carrier 17 becomes large, the reward calculation unit 63 decreases the reward “r”. The reward calculation unit 63 decreases the reward “r” by giving “−1” which is a value of the reward. Note that the value of the reward is not limited to “−1”.

The function update unit 64 updates a function which is a model for determining a thrust command correction value in accordance with the reward calculated by the reward calculation unit 63. The function can be updated in accordance with the data set, for example, by updating the action-value table. The action value table is a data set in which any action and an action value thereof are stored in association with each other in a form of a table. For example, in the case of Q learning, an action value function Q(s_t,a_t) expressed by the above formula (1) is used as a function for determining the thrust command correction value.

FIG. 11 is a flowchart illustrating a processing procedure of the learning device 52 included in the conveyance system 1 according to the third embodiment. A reinforcement learning method of updating the action-value function Q(s,a) will be described with reference to the flowchart of FIG. 11.

In step S11, the learning device 52 acquires, by the data acquisition unit 61, position information and a thrust command correction value of each carrier 17. That is, the learning device 52 acquires learning data. The data acquisition unit 61 outputs a data set collectively including the learning data to the model generation unit 62.

In step S12, the learning device 52 calculates a reward by the reward calculation unit 63. The reward calculation unit 63 calculates a reward for a combination of the position information of each carrier 17 and the thrust command correction value for each carrier 17. The reward calculation unit 63 increases or decreases the reward on the basis of the variability of the speed of the carrier 17.

In step S13, the learning device 52 updates the action-value function by the function update unit 64. The function update unit 64 updates the action-value function Q(s,a) on the basis of the reward calculated in step S12. The learning device 52 updates the action-value function Q(s_t,a_t) stored in the learned model storage unit 53.

In step S14, the learning device 52 determines, by the function update unit 64, whether the action-value function Q(s,a) has converged. The function update unit 64 determines that the action-value function Q(s,a) has converged when the action-value function Q(s,a) is no longer updated in step S13.

If the function update unit 64 determines that the action-value function Q(s,a) has not converged (step S14, No), the learning device 52 returns the procedure to step S11. On the other hand, if the function update unit 64 determines that the action-value function Q(s,a) has converged (step S14, Yes), the learning device 52 ends the processes in accordance with the procedure illustrated in FIG. 11. Note that the learning device 52 may continue learning by returning the procedure from step S13 to step S11 without making the determination in step S14. The learned model storage unit 53 stores a learned model which is the generated action-value function Q(s,a).

In the third embodiment, the case has been described where reinforcement learning is applied to the learning algorithm used by the learning device 52, but learning other than reinforcement learning may be applied to the learning algorithm. The learning device 52 may execute machine learning by using a known learning algorithm other than reinforcement learning, for example, a learning algorithm such as deep learning, a neural network, genetic programming, inductive logic programming, or a support vector machine.

The learning device 52 illustrated in FIGS. 9 and 10 is a device built in the track controller 50. The learning device 52 may be a device outside the track controller 50. The learning device 52 which is a device outside the track controller 50 constitutes the conveyance system 1. The learning device 52 may be a device connectable to the track controller 50 via a network. The learning device 52 may be a device present on a cloud server.

The learning device 52 may learn a relationship between the position information and the thrust command correction value in accordance with a data set created for a plurality of conveyance systems 1. The learning device 52 may acquire learning data from a plurality of conveyance systems