KR20070036760A - Ai plc controller base on u-network - Google Patents

Abstract

The present invention is a patent for PLC among industrial automation devices. Although the method of operation between PLC and PLC and the solution of process error was the form of the operator's experience or the temporary pause recovery of simple process error, PLC's artificial intelligence is equipped with the artificial intelligence that can be connected between process. New concept that enables process error self-learning and correction of errors caused by linking process between processes by installing neural network engine in PLC to enable quick recovery method and self-check in case of the next similar or repetitive error It is devised about PLC design and operation method of PLC.

The operation of the existing PLC has been mixed and used like a separate controller, and has been configured to centrally manage by adding a network module, but in the present invention, by adding the network function to the PLC itself, it reduces the cost of unnecessary FA controller During programming and work, process changes can be carried out immediately at the central control center or at any place under the connected network. It is characterized by the active control of the process by cross-verifying the malfunction status between processes during the PLC control period.The malfunction and device status information generated when an error occurs between processes is learned, and the next similar or repetitive process is In the event of an error, the process status can be quickly and accurately determined by learned content. And recovery.

PLC, Ubiquitous, Process Control

Description

Intelligent PLC Controller Using Neural Networks {AI PLC Controller Base on U-Network.}

  1 is a network configuration diagram of a process controller of a new concept

  2 is a diagram illustrating an existing FA configuration and operation

  3 is an intelligent PLC configuration of the present invention

  4 is a process flow diagram according to the present invention.

The present invention relates to a PLC controller for controlling a process among the process automation equipment, an FA controller and a central management system for controlling and controlling the process. The general PLC controller is called Programmable Logic Controller and has a role to automate the process by supplying power to each Mechanic component in the automation process. General process automation consists of Mechanic part, PLC for controlling it, and FA controller that can be connected by extension of PLC to the central control management system.

       A general FA controller is composed of mechanical components, a PLC, and a FA controller as shown in FIG. 2 (FIG. 2A), a network configuration for central management (FIG. 2B), and a central control system (FIG. 2-B). It is composed of

       Independent process control means that each process controller of the semi-finished product is manufactured in the process A, the first process is processed in the process B, and the process is produced as a finished product in the process C as shown in FIG.

    Common process problems occur as follows.

When a problem occurs at A, which is the top-level process, the entire process B and C are urgently stopped in order to identify the type of error in the A process where the problem occurs and recover it. Fortunately, if the current process controller PLC is equipped with an additional module and can transmit the status information to the central center, the process A can only be stopped when the central center is in HALT state. If it is idle or idle for a while, the post process does not need to be stopped. In addition, even in the HALT state, if stopped immediately, the B and C processes should continue to operate. However, a typical control system is inefficient in productivity because it stops the entire process, recovers from error, and restarts it.

 The bigger problem is when it occurs in process B. In the event of product defects in Process B, the existing control system will actually try to stop the entire process once and identify the problems in Process B. If a problem occurs in the actual process B, stopping the entire process is justified, but in some cases, the defective part may occur in the process B due to an error that the process A did not detect.

In the general process control part, each process is not performed independently but is closely linked between front and rear processes to form a whole process. Therefore, in case of malfunction between individual processes, the countermeasures are different from case to case. Status information between individual processes must be operated with status information or operation information between the central system and the associated process controllers to prevent deterioration in operating rates and productivity.

The present invention is to solve the above problems, the entire network configuration is not a subordinate structure of the Server Client, but the entire network can be operated by synopsis based on the ubiquitous basis, and there is an object of the PLC configuration capable of such control.

  In order to form an organic network, a separate NetWork Module is mounted on the PLC controller, so that the general mechanical driver is not only a simple ON / OFF contact state but also a practical operating state (for example, RPM and an operating power supply voltage). Each operation information and status information) are transmitted to the central controller and the upper PLC controller in real time, and the interworking NetWork configuration can also implement synoptic information transfer and central control. The main feature of the present invention is that the local controller must implement the server by transmitting the status information while performing each server function, the Client Server Network Module must be mounted in the PLC controller and performs the basic MPU for the unique role of the PLC. There is a characteristic that the network is organically constructed under the specialized PLC controller as described above.

The present invention relates to a configuration of an intelligent PLC controller for interlocking control of an FA controller and an interlocking process control method composed of an intelligent PLC.

 Interlocked process control is to intelligently control the inter-process workflow by connecting each process organically when a problem occurs between the 1 process and the 2 process 3 process in the [Fig. The intelligent control means that when the manufacturing process problem occurs in the preliminary process 2, the state of the first process, the first process, and the third process, the third process is normal when the problem is actually a process 2. In this case, step 3 quickly stops the process, and the first step, which is a preliminary step, slows down the processing speed slightly and waits for the recovery of the second step while the line of one step is normally operated until the second step is normally operated. However, the fact that there was an operational problem in the second process does not mean that this is a problem of the current process. If an unexpected defective product comes out in the first process, the aftermath is a problem in the second process, the next process.

Therefore, process 2 checks the status of process 1 and its own process, which is the previous process, and if it is a problem of process 1, process 3, which is a post process, processes the work quickly and completes work on the remaining process line, I have to wait. Although these partial process controllers operate under the control and command of the central system, the work efficiency can be improved by transmitting organic state information for each independent process.

 In order to network each process controller as above, each controller must operate in a separate Server / Client Mode. However, when the TCP / IP stack and the Server / Client module are installed in a general SOC, the efficiency of the logic controller processor is very low. Therefore, for PLC control, the main function, TCP / IP Stack must be composed of internal memory and FPGA module for client / server implementation for the built-in SOC and programmable logic.

 [Figure 3] shows the configuration of the Logic Controller as above and has organic connectivity using its own AI NetWork.

Implemented as above, if there is an organic cross-checking function between the processes and a malfunction recovery function, and the central status monitoring function, the operation rate is improved because the priority and uptime adjustment are performed in case of problems in each process. It can be kept at maximum, and these process controllers have the advantage of reducing cost and return time and being very easy to manage.

Claims (5)

PLC controller for self-learning and storing device and status information. Each PLC controller is equipped with a neural network to increase the process utilization rate by intelligently judging the control of the after-process according to the kind of the malfunction in case of the after-process malfunction. Even if a process error is not listed in the work schedule, the PLC controller itself has a learning function, so it can be quickly recovered at the next process error. PLC controller performs cross check between PLCs for error verification The existing PLC controller notifies the central management system when a malfunction occurs and is controlled by the central management system or controls the process independently. However, in the present invention, since each process undergoes cross-validation in the form of a cross-check of the post-process after notification when a problem occurs, a more accurate process problem can be identified. In addition, this type of control has a form of transmission after pre-measurement to the central control center, thereby enabling faster process control. How to learn when an interprocess malfunction occurs. In the process control method, the occurrence of error is not the order of recovery and restart. Instead, when the process error occurs, each device receives the status information as an input variable and stores the status information such as the problem location equipment as a digital value. City reference or method you can use. In case of malfunction between processes, each controller checks the status information of each device for error recovery. In the event of a process failure, a PLC controller or a similar controller acting as a PLC controller between the processes not only transmits the error status to the central control center, but also monitors the status of the controllers and components between the unexpected linked processes. How to check at the same time. How to predict malfunction by learning status information. A PLC controller or a controller with similar functions checks not only its own status information but also the status information of surrounding processes.It also learns the status of malfunctions in the process and learns the status information of the components of surrounding process controllers by neural network engine. It is a controller that can be prevented before malfunction occurs in the whole process if the status value of the component of the current process controller or the operation equipment of the peripheral process is abnormal.

Images