Control System

Abstract

To provide a control system realizing convenience and security reinforcement compatibly.SOLUTION: A control system comprises a control unit that has a Web server and two or more network ports and an operation unit that has a Web server and a network port. The two or more network ports of the control unit include first and second network ports, and the first network port of the control unit and the network port of the operation unit are connected to a first network segment that is a closed network segment to which no elements other than the control and operation units are connected. The control unit transfers data received from the interface unit to the operation unit and displays a Web-server screen of the operation unit on a Web-server screen of the control unit.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a control system. [Background technology]

[0002] AI technology has made remarkable progress in recent years, and systems that apply this technology are expected to improve business efficiency and accuracy. For example, Patent Document 1 discloses a plant control system that aims to improve the accuracy and speed of plant control by passing I / O (Input / Output) data to a general-purpose PC via a PLC (Programmable Logic Controller), where the data is calculated and the results are output to the PLC. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Laid-Open No. 2009-93375 Summary of the Invention [Problem to be solved by the invention]

[0004] Conventional technology requires rules for exchanging data between a PLC and a general-purpose PC, i.e., dedicated communication protocols and commands, which makes system design cumbersome. Furthermore, processing dedicated protocols and commands requires the PLC's processing power, which can increase the load on the PLC. In light of this, the present invention aims to provide a control system that combines convenience with enhanced security. [Means for solving the problem]

[0005] The control system controls a system to be controlled and comprises a control unit having a web server and two or more network ports, and a calculation unit having a web server and a network port, wherein the two or more network ports of the control unit include a first and a second network port, a first network port of the two or more network ports of the control unit and the network port of the calculation unit are connected to a first network segment which is a closed network segment to which elements other than the control unit and the calculation unit are not connected, a second network port of the two or more network ports of the control unit and a network port of an interface unit are connected to a second network segment different from the first network segment, the interface unit is a client having a web browser which displays a web server screen, and the control unit performs a forwarding process to transfer data received from the interface unit to the calculation unit, The interface unit , a first area for operating the control unit; Web server of the computing unit and a second area displaying the web server screen including Display. [Effects of the Invention]

[0006] A highly secure control system can be provided. [Brief explanation of the drawings]

[0007] [Figure 1] 1 is a configuration example of a control system according to an embodiment of the present invention. [Figure 2] A variation of Figure 1 [Figure 3] Example of operation from the interface to the control unit [Figure 4] Example of operation from the interface unit to the AI ​​calculation unit [Figure 5] Processing flow of operations 120 to 123 received from the interface unit (1) [Figure 6] Interface screen display example (1) [Figure 7]Processing flow of operations 120 to 123 received from the interface part (2) [Figure 8] Interface screen display example (2) [Figure 9] Example of connection configuration for interface, control, and AI calculation units [Figure 10] Example of communication settings for the interface, control, and AI calculation units [Figure 11] Example of a transfer table for the control unit in the communication settings of Figure 10

[0008] Hereinafter, an embodiment of the present invention will be described with reference to the drawings. The following description and drawings are examples for explaining the present invention, and some omissions and simplifications have been made as appropriate for clarity of explanation. The present invention can be implemented in various other forms. Unless otherwise specified, each component may be singular or plural.

[0009] In order to facilitate understanding of the invention, the position, size, shape, range, etc. of each component shown in the drawings may not represent the actual position, size, shape, range, etc. Therefore, the present invention is not necessarily limited to the position, size, shape, range, etc. disclosed in the drawings.

[0010] (One embodiment of the present invention and overall system configuration) (Figure 1) In the control system of the present invention, a control unit 101 that controls a system to be controlled includes a web server 102, and also runs a sequence control process 103 and a forwarding process 104. The control unit 101 also includes two network ports, an Ethernet (registered trademark) port 105 and an Ethernet port 106. The control unit 101 may have two or more network ports. The Ethernet port 105 and the Ethernet port 106 of the control unit 101 are each located in a different network segment.

[0011] AI calculation unit 107, which performs machine learning under the control of control unit 101, is equipped with web server 108, and runs learning processing 109 and inference processing 110. AI calculation unit 107 is equipped with Ethernet port 111 that provides a one-to-one Ethernet connection to Ethernet port 105 of control unit 101, thereby forming a closed network segment to which no elements other than control unit 101 and AI calculation unit 107 are connected. This network segment is referred to as a first network segment.

[0012] Interface unit 112 includes web browser 113 that displays a web server screen. Interface unit 112 also includes Ethernet port 114. Interface unit 112 is, for example, a device that performs wireless communication, such as a tablet, and when control unit 101 connects Ethernet port 106 to wireless access point 115, it is wirelessly connected to Ethernet port 114 of interface unit 112. As a result, Ethernet port 106 and Ethernet port 114 are connected to a second network segment that is different from the first network segment. This establishes network configurations that are independent of each other, thereby ensuring security.

[0013] Conventionally, when the two networks are configured independently, data communication received from interface unit 112 cannot be sent directly to AI calculation unit 107 via control unit 101. However, control unit 101 of the present invention operates forwarding process 104 for transferring data received from interface unit 112 to AI calculation unit 107, allowing data received from interface unit 112 to reach AI calculation unit 107 located in a different network segment. Note that port forwarding or the like is used as a method for implementing forwarding process 104.

[0014] Although not shown, the control unit 101 has a memory and a processor connected to the memory and the Ethernet ports 105 and 106 in addition to the Ethernet ports 105 and 106. The processor executes a program stored in the memory to realize the web server 102, and the processor also executes sequence control processing 103 and forwarding processing 104.

[0015] The above-mentioned Ethernet port 105 is an example of a first network port, and the Ethernet port 106 is an example of a second network port. The control unit 101 is, for example, a control device such as a PLC (for example, a device that controls a system to be controlled, such as industrial equipment or other equipment, at a defined control period).

[0016] In addition to the Ethernet port 111, the AI ​​calculation unit 107 has a memory and a processor connected to the memory and the Ethernet port 111. The processor executes a program stored in the memory to realize a web server 108, and the processor also executes a learning process 109 and an inference process 110.

[0017] In the learning process 109, a machine learning model such as a neural network is trained. The machine learning model receives, for example, image data captured by a camera as input, and outputs data representing the classification result of the image data (for example, data representing the label of an object shown in the image represented by the image data, or data representing the presence or absence of an abnormality in the object). In the inference process 110, inference is performed using the trained model. For example, the AI ​​calculation unit 107 is an inspection device that receives input of captured image data of an inspection target and estimates the presence or absence of an abnormality in the inspection target from the image represented by the captured image data. Furthermore, the inspection target and the control target may be different or the same.

[0018] The first and second network segments may be different segments of the same network, or may be different networks. A network or a network segment may be an example of a communication medium through which data travels.

[0019] (Figure 2) Although FIG. 1 shows a wireless connection, the present invention can achieve the same effect even if the Ethernet port 106 of the control unit 101 and the Ethernet port 114 of the interface unit 112 are directly connected by wire, as shown in FIG. 2.

[0020] (Fig. 3, Fig. 4) An embodiment using the control system of the present invention will be described. The control system of the present invention is used, for example, in an inspection device system that performs everything from AI learning to inference using on-site equipment (edge) without using a PC. Such an inspection device includes a sensor unit that detects the passage of an inspected product; a camera unit that releases the shutter in response to a signal received from the control unit and stores the captured image in the AI ​​calculation unit; a control unit that detects the passage of the inspected product by the sensor and releases the shutter of the camera unit, and receives the inspection results from the AI ​​calculation unit and removes defective products from the line; an AI calculation unit that captures the camera image of the inspected product and uses the learning results input into a trained model to determine whether the appearance of the inspected product is good or bad, and reports the result to the control unit; and an interface unit that sets parameters for the control unit and sets parameters, commands, and displays the results of AI learning in the AI ​​calculation unit.

[0021] 3 shows an example of operations from the interface unit 112 to the control unit 101 when the control system of the present invention is used in the edge AI inspection device system. Operations from the interface unit 112 to the control unit 101 can be broadly classified into two: parameter settings 120 for the control unit 101 and commands 121 to the control unit 101. The parameter settings 120 for the control unit 101 include settings such as the time and encoder related to adjusting the timing from sensor input to output of the image capture signal. The commands 121 to the control unit 101 include taking a test image, etc.

[0022] 4 shows an example of operations from interface unit 112 to AI calculation unit 107 when the control system of the present invention is used in the edge AI inspection device system. Operations from interface unit 112 to AI calculation unit 107 can also be broadly classified into two: parameter settings 122 for AI calculation unit 107 and commands 123 to AI calculation unit 107. Parameter settings 122 for AI calculation unit 107 include adjustments to the position related to adjustment of the camera's crop angle of view and the size of the captured image. Commands 123 to AI calculation unit 107 include commands to start learning, stop learning, start inference, stop inference, etc.

[0023] By having the control unit 101 operate the forwarding process 104, not only can the parameter setting 120 of the control unit 101 and the command 121 to the control unit 101 as shown in Figure 3 reach the AI ​​calculation unit 107 via communication received from the interface unit 112, but also the parameter setting 122 of the AI ​​calculation unit 107 and the command 123 to the AI ​​calculation unit 107 as shown in Figure 4 can reach the AI ​​calculation unit 107 via communication received from the interface unit 112.

[0024] (Figure 5) 5 shows the flow of processing of operations 120 to 123 received from the interface unit 112 when the forwarding process 104 of the control unit 101 is realized by port forwarding. The connection configuration of the interface unit 112, control unit 101, and AI calculation unit 107 at this time is shown in FIG.

[0025] The control unit 101 has a forwarding table for distinguishing between packets to be forwarded to the AI ​​calculation unit 107 and packets to be processed by the control unit 101 itself. An example of the forwarding table is shown in FIG. 11. The forwarding process 104 of the control unit 101 executes forwarding processing based on this forwarding table. The forwarding table in FIG. 11 is used when the IP addresses of the interface unit 112, control unit 101, and AI calculation unit 107 are set as shown in FIG. 10.

[0026] Of the communications received from the interface unit 112, parameter setting 120 and command 121 to be processed by the control unit 101 are sent to the Ethernet port 106 of the control unit 101, with the destination being the IP address: 192.168.0.1 and port number: 8080 (see FIG. 10 ), and the command is sent to the Ethernet port 106, and the control unit 101 then performs the parameter setting process and the process instructed by command 121.

[0027] On the other hand, parameter setting 122 and command 123 to be processed by AI calculation unit 107 match the rules set in the forwarding table (see FIG. 11) by specifying IP address 192.168.0.1 and port number 80 (see FIG. 10) of Ethernet port 106 of control unit 101 as their destination. As a result, the command sent to Ethernet port 106 of control unit 101 is forwarded from Ethernet port 106 to Ethernet port 105 by port forwarding, and then transmitted from Ethernet port 105 to Ethernet port 111 of AI calculation unit 107. In this way, the command is forwarded from control unit 101 to AI calculation unit 107, and parameter setting 122 and command 123 can be processed by AI calculation unit 107.

[0028] When the forwarding process 104 of the control unit 101 is realized by port forwarding, the user's operation commanded from the interface unit 112 does not need to be sent from the control unit 101 to the AI ​​calculation unit 107 in a bucket brigade manner, thereby reducing the processing load on the control unit 101.

[0029] (Figure 6) 6 is an example of a screen display of the interface unit 112. The web browser 113 of the interface unit 112 displays the web server 102 of the control unit 101. The display of the web server 102 of the control unit 101 is made up of two areas: an operation area 130 for the control unit, and a display area 131 for the web server 108 of the AI ​​calculation unit 107.

[0030] Operation area 130 for control unit 101 consists of two areas: parameter setting area 132 for setting parameters 120 for control unit 101, and command operation area 133 for issuing commands 121 to control unit 101. Display area 131 of Web server 108 for AI calculation unit 107 consists of three areas: parameter setting area 134 for setting parameters 122 for AI calculation unit 107, command operation area 135 for issuing commands 123 to AI calculation unit 107, and display area 136 for displaying captured images of the inspected product, etc.

[0031] In this way, in the target system, the user can operate two devices, the control unit 101 and the AI ​​calculation unit 107, from a single user interface. Furthermore, because the user can operate the web server screens of the control unit 101 and the AI ​​calculation unit 107 on a single screen, there is no need for the user to be aware of whether the target of the operation is the control unit 101 or the AI ​​calculation unit 107, and an interface with a sense of system unity can be provided while strengthening security. In addition, the processing load on the control unit 101 is reduced.

[0032] (Figure 7) As a variation of the processing flow shown in Figure 5, an example of a flow is shown in which the control unit 101 realizes the forwarding process 104, in which the control unit 101 transfers data received from the interface unit 112 to the AI ​​calculation unit 107, by combining port forwarding and dedicated communication.

[0033] First, when the forwarding process 104 of the control unit 101 is realized by a combination of port forwarding and dedicated communication commands, the processing of operations 120 to 123 received from the interface unit 112 will be explained below. When the forwarding process 104 is realized by a combination of port forwarding and dedicated communication commands, some of the user operations sent from the interface unit 112 are delivered from the control unit 101 to the AI ​​calculation unit 107 using a bucket brigade method. This means that the transmission of command 123 is realized by a dedicated communication command.

[0034] By sending some of the user operations from the control unit 101 to the AI ​​calculation unit 107 using dedicated communication commands rather than forwarding them, it becomes possible for the control unit 101 to recognize the target user operation. For example, if the control unit 101 wants to know whether the AI ​​calculation unit 107 is currently learning, performing inference, or what it is currently doing, by converting the corresponding command from the user into a dedicated communication command, the control unit 101 will first receive a learning start command issued by the interface unit 112 and then issue a learning start command to the AI ​​calculation unit 107, allowing the control unit 101 to know that the calculation unit 107 is currently learning.

[0035] When parameter setting 122 is sent from interface unit 112, the destination port number is set to 80, which matches the transfer rules (see Figure 11), and parameter setting 122 is sent from Ethernet port 114 of interface unit 112 to Ethernet port 106 of control unit 101.The parameter setting 122 is then forwarded from Ethernet port 106 of control unit 101 to Ethernet port 105 by port forwarding, and then sent from Ethernet port 105 to Ethernet port 111 of AI calculation unit 107, whereby parameter setting 122 is processed in AI calculation unit 107.

[0036] On the other hand, when command 123 is issued from Ethernet port 114 of interface unit 112, its destination port number is 8080, so command 123 is received at Ethernet port 106 of control unit 101. In control unit 101, process 140 determines the content of command 123 in forwarding process 104 and creates communication data to be sent to AI calculation unit 107. The created communication data is sent from Ethernet port 105 of control unit 101 to Ethernet port 111 of AI calculation unit 107, causing command 123 to be processed by AI calculation unit 107. In this way, a similar effect can be achieved by combining port forwarding and dedicated communication commands.

[0037] (Figure 8) Figure 8 is a modified example of the screen display example of the Web server 102 of the interface unit 112 shown in Figure 6. The Web browser 113 of the interface unit 112 displays the Web server 102 of the control unit 101, and the display of the Web server 102 of the control unit 101 is made up of an operation area 150 for the control unit 101, an operation area 151 for the AI ​​calculation unit 107, and a display area 152 for the Web server 108 of the AI ​​calculation unit 107.

[0038] Operation area 150 is made up of parameter setting area 153 for setting parameters 120 for control unit 101, and command operation area 154 for issuing commands 121 to the control unit. The reason operation area 151 is located on Web server 102 of control unit 101 is so that control unit 101 can receive commands to AI calculation unit 107 and create communication data for AI calculation unit 107.

[0039] The display area 152 of the Web server 108 of the AI ​​calculation unit 107 is composed of a parameter setting area 155 for setting parameters 122 of the AI ​​calculation unit 107, and a display area 156 for displaying captured images of the inspected product, etc.

[0040] According to the embodiment of the present invention described above, there is no need for a router or the like for distributing communications, which was previously necessary when securing independent routes. Furthermore, there is no communication discrepancy between the control unit 101 and the AI ​​calculation unit 107 due to an abnormality in the communication command.

[0041] Although the present invention has been described assuming an Ethernet port as an example of a network port, the network port does not have to be limited to an Ethernet port as long as similar effects can be expected.

[0042] According to the embodiment of the present invention described above, the following advantageous effects are achieved.

[0043] (1) The control system controls a control target system and includes a control unit 101 having a web server and two or more network ports, and an AI calculation unit 107 having a network port for the web server. The two or more network ports of the control unit 101 include a first and a second network port. The first network port of the control unit 101 and the network port of the AI ​​calculation unit 107 are connected to a first network segment, which is a closed network segment to which elements other than the control unit 101 and the AI ​​calculation unit 107 are not connected. The second network port of the control unit 101 and the network port of the interface unit 112 are connected to a second network segment different from the first network segment. The interface unit 112 is a client having a web browser that displays a web server screen. The control unit 101 performs a forwarding process to transfer data received from the interface unit 112 to the AI ​​calculation unit 107, and displays the web server screen of the AI ​​calculation unit 107 on the web server screen of the control unit 101. This configuration provides a control system with high security.

[0044] (2) The control unit 101 transfers the data received from the interface unit 112 to the AI ​​calculation unit 107 by port forwarding. In this way, the data received from the interface unit 112 can reach the AI ​​calculation unit 107 located in a different network segment.

[0045] (3) The control unit 101 combines port forwarding and a dedicated communication command to transfer data received from the interface unit 112 to the AI ​​calculation unit 107. In this way, even in other embodiments, the data received from the interface unit 112 can be delivered to the AI ​​calculation unit 107 located in a different network segment.

[0046] (4) The data transmitted from the interface unit 112 to the control unit 101 by the control system is either a parameter setting or a command for the control unit 101, and the data transmitted from the interface unit 112 to the calculation unit 107 via the control unit 101 by forwarding processing is either a parameter setting or a command for the calculation unit 107. In this way, the AI ​​calculation unit 107 can execute parameter setting and command processing.

[0047] (5) The control system includes a sensor that detects the passage of the inspected product and a camera that photographs the inspected product in response to a photographing signal from the control unit, and the control unit outputs a photographing signal to the camera when it receives input from the sensor that the passage of the inspected product has been detected, and the calculation unit performs model training that takes an image of the inspected product as input and outputs a judgment result of whether the inspected product is good or defective, and performs inference that judges whether the inspected product is good or defective by inputting the image of the inspected product photographed by the camera into the trained model. In this way, it can be used as an inspection device system that performs everything from AI learning to inference using on-site equipment (edge).

[0048] (6) The parameter settings transmitted from the interface unit 112 to the control unit 101 are the time and encoder settings related to timing adjustment from input from the sensor to output of the imaging signal. In this way, the settings of the inspection device system can be realized.

[0049] (7) The parameter settings transmitted from the interface unit 112 to the calculation unit 107 via the control unit 101 are the position related to the adjustment of the cropping angle of the camera and the size of the captured image. In this way, the settings of the inspection device system can be realized.

[0050] (8) The commands transmitted from the interface unit 112 to the calculation unit 107 via the control unit 101 are commands to start learning, stop learning, start inference, and stop inference. In this way, commands can be executed in the inspection device system.

[0051] The present invention is not limited to the above-described embodiments, and various modifications and combinations of other configurations are possible without departing from the spirit of the present invention. Furthermore, the present invention is not limited to those having all of the configurations described in the above-described embodiments, and includes those in which some of the configurations are omitted. [Explanation of symbols]

[0052] 101: Control unit 102: Web server of control unit 101 103: Sequence control process of the control unit 101 104: Forwarding process of the control unit 101 105: Ethernet port of control unit 101 106: Ethernet port of control unit 101 107:AI calculation section 108: Web server of AI calculation unit 107 109: Learning process of AI calculation unit 107 110: Inference processing of AI calculation unit 107 111: Ethernet port of AI calculation unit 107 112: Interface section 113: Web browser of interface unit 112 114: Ethernet port of interface section 112 115: Wireless access point 120: Parameter setting for control unit 101 121: Command to control unit 101 122: Parameter setting for AI calculation unit 107 123: Instruction to AI calculation unit 107 130: Operation area for the control unit of display example 1 131: Display area of ​​the Web server 108 of the AI ​​calculation unit 107 in display example 1 132: Parameter setting area for setting the parameter 120 in display example 1 133: Command operation area for executing command 121 in display example 1 134: Parameter setting area for setting parameter 122 in display example 1 135: Command operation area for executing command 123 in display example 1 136: Display area for displaying the photographed image of the inspected product in display example 1 140: Processing for creating communication data for the AI ​​calculation unit 107 150: Operation area for control unit in display example 2 151: Command operation area for executing command 123 in display example 2 152: Display area of ​​the Web server 108 of the AI ​​calculation unit 107 in display example 2 153: Parameter setting area for setting parameter 120 in display example 2 154: Command operation area for executing command 121 in display example 2 155: Parameter setting area for setting parameter 122 in display example 2 156: Display area for displaying the photographed image of the inspected product in display example 2

Claims

1. a control unit that controls a system to be controlled and has a web server and two or more network ports; a computing unit having a web server and a network port, the two or more network ports of the control unit include a first and a second network port; a first network port of the two or more network ports of the control unit and the network port of the calculation unit are connected to a first network segment which is a closed network segment to which elements other than the control unit and the calculation unit are not connected; a second network port of the two or more network ports of the control unit and a network port of the interface unit are connected to a second network segment different from the first network segment; the interface unit is a client having a web browser that displays a web server screen, the control unit performs a forwarding process of transferring data received from the interface unit to the calculation unit; The interface unit displays the web server screen including a first area for operating the control unit and a second area for displaying the web server of the calculation unit. Control system.

2. 10. The control system of claim 1, The control unit transfers the data received from the interface unit to the processing unit by port forwarding. Control system.

3. 10. The control system of claim 1, The control unit combines the data received from the interface unit with a port forward command and a dedicated communication command, and transfers the data to the calculation unit. Control system.

4. 10. The control system of claim 1, the data transmitted from the interface unit to the control unit in response to the user's operation in the first area is either a parameter setting or a command to the control unit; The data transmitted from the interface unit to the calculation unit through the control unit by the forwarding process in response to the user's operation in the second area is either a parameter setting or a command to the calculation unit. Control system.

5. 5. The control system of claim 4, a sensor that detects the passage of an inspected product; a camera that photographs the product to be inspected in response to a photographing signal from the control unit; Equipped with the control unit outputs a photographing signal to the camera when receiving an input from the sensor that the passage of the inspected product has been detected; The calculation unit Learning a model that takes an image of the product to be inspected as input and outputs a judgment result as to whether the product is good or bad; An inference in which an image of the inspected product taken by the camera is input into a trained model to determine whether the inspected product is a good product or a defective product. Do Control system.

6. 6. The control system of claim 5, The parameter settings transmitted from the interface unit to the control unit include the time and encoder settings related to timing adjustment from input from the sensor to output of the image capture signal. Control system.

7. 6. The control system of claim 5, The parameter settings transmitted from the interface unit to the calculation unit through the control unit are the position related to the adjustment of the cropping angle of the camera and the size of the captured image. Control system.

8. 6. The control system of claim 5, The commands transmitted from the interface unit to the calculation unit through the control unit are commands to start learning, stop learning, start inference, and stop inference. Control system.

Images