Intelligent electric control system for stretch bending and straightening

Abstract

An intelligent electric control system for stretching, bending and straightening is characterized by comprising a PLC master controller and an intelligent decision-making device, master control work is divided into two parts, namely a real-time control part and a big data decision-making part, a traditional PLC is used in the implementation part to control the machine row operation function of a whole stretching and straightening line, and the high-performance intelligent decision-making controller is used in the big data decision-making part to control the machine row operation function of the whole stretching and straightening line. And the data are fused and processed to generate a control strategy, the master control PLC is driven to execute control actions, and the intelligent decision-making device performs main logic judgment. The double-control intelligent electric control system enables the whole data processing to be more reasonable through decoupling control and data processing. The big data processing unit for completing unit control by the intelligent decision-making device and the master control PLC basically comprises a bottom driving layer, a data processing layer, a software application layer and a human-computer interface layer.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of steel rolling process, in particular, especially relates to an intelligent electric control system for stretch-bending-straightening. BACKGROUND

[0002] In the continuous processing production line of metal strip (such as stainless steel, silicon steel, aluminum alloy, copper strip, etc.), the stretch-bending-straightening machine is one of the key process equipment. Its main function is to eliminate the shape defects of the strip (such as wave shape, buckling, edge wave, middle wave, etc.) and improve the mechanical properties of the material by applying specific tensile stress and bending deformation (straightening amount). The existing independent or simple linkage control strategy is difficult to maintain the optimal combination of tensile force and straightening amount in real time and accurately. It often leads to problems such as insufficient straightening (shape defects not completely eliminated), over-straightening, unstable tension (causing deviation and risk of strip breakage), etc. The setting and adjustment of tensile force and straightening amount parameters largely depend on the experience of the operator. The setting values may differ between different operators or different shifts, making it difficult to ensure the consistency of product quality and increasing the risk of quality and management cost. SUMMARY

[0003] In summary, the existing stretch-bending straightening machine control system has obvious shortcomings in control precision, dynamic response speed, working condition adaptability, product quality stability and energy efficiency when dealing with complex, dynamic and high requirement strip processing tasks. It is urgent to develop a new type of electric control system that can break through the depth cooperation and joint optimization control of stretching force and straightening amount in the dynamic process, use video sensors to detect the surface quality problems of the plate on the stretch-bending machine line, and through PLC linkage control of the execution equipment, the intelligent decision maker calculates the optimization control instruction based on the machine set parameters through the AI model, and builds a closed-loop automatic quality control and adjustment system, so the perception system is very important in the implementation process, and the data processing work is complex. It is difficult for a master PLC to process sensor data in real time. At present, the types of machine set sensor signal sources have included dozens of vibration, sound, laser, image, electromagnetic and photoelectric sensors, and the number has reached hundreds. The operation capacity of PLC has far failed to meet the needs, especially the visual perception system is more and more applied to monitoring, and PLC has completely failed to complete the processing of video data alone. At this time, a powerful processor is needed to meet the data processing needs of the machine set. The overall efficiency of the stretch-bending straightening process is significantly improved. The present application provides a stretch-bending straightening intelligent electric control system, characterized by including two controllers of PLC master control and intelligent decision maker. That is, the master control work is divided into two parts, real-time control part and big data decision part, the real-time control part still uses the traditional PLC to be responsible for the control function of the machine train, and the big data decision part uses the high-performance intelligent decision maker to generate the control strategy after fusing and processing various data, and drives the master PLC to execute the control action, thereby improving the reliability of the machine train. The main logic of the intelligent decision maker judges, and the intelligent decision maker can also communicate with the server of the factory level to report some information of the machine train to the factory level service system. Through decoupling control and data processing, the whole data processing is more efficient and reasonable, thereby improving the safety and operation precision of the whole machine train. The intelligent decision maker is composed of a bottom drive layer, a data processing layer, a software application layer and a man-machine interface layer, the bottom drive layer is composed of sensors, bottom equipment and a real-time operating system, the data processing layer includes data storage, data processing and data acquisition, the software application layer includes system management services and analysis process data, function management system, algorithm model application and state monitoring algorithm model, and the man-machine interface layer includes system configuration and model arrangement.

[0004] 2. Further, the intelligent electric control system of the stretch-bending straightening, further comprising an intelligent decision maker, which is a functional expansion of the PLC, and the basic unit thereof is an industrial computer, and a Linux system is installed, and the operation capacity is strong, and the main functions include flexible access of the sensor, data acquisition, processing and fusion, and the data sources include video data, audio data, piezoelectric vibration data, optical fiber stress sensor data and the like, and meanwhile, the sensors connected to the original PLC system can be switched to the intelligent decision maker, and the rotation speed and temperature sensors can be switched to the intelligent decision maker, so that the pressure of the PLC can be further reduced.

[0005] 3. Further, the intelligent electric control system of the stretch-bending straightening, wherein the important function of the intelligent decision maker is to generate a control strategy based on various diagnostic models, the video sensor discovers the surface quality problem of the plate, and the strategy is transmitted to the PLC to generate a control action, and the intelligent decision maker calculates the optimized control instruction based on the unit parameters through the AI model.

[0006] 4. Further, the intelligent electric control system of the stretch-bending straightening, wherein the bottom drive layer is the bottom support of the system, and the drive mainly refers to the sensor acquisition card drive, which is used for connecting and analyzing the sensor data, and the system is the operating system and functional support, and contains a database system, a time system and a network system.

[0007] 5. Further, the intelligent electric control system of the stretch-bending straightening, wherein the software application layer contains the entire system service and core model application, and mainly contains the system management service and the algorithm model application.

[0008] According to the process characteristics of actual production, the intelligent electric control system of the stretch-bending straightening can have two capabilities, the basic control capability is completed by the PLC part, and the other part is the intelligent decision maker, and the data processing and analysis capability is completed by the high-speed operation processor, the video sensor discovers the surface quality problem of the plate, and feeds back to the PLC to issue a command to the execution equipment, and according to the detection result, the parameters of the stretch-bending straightening machine are automatically adjusted or subsequent actions (such as marking, speed reduction, sorting, alarm shutdown and the like) are triggered, so as to improve the product quality, reduce the waste products and improve the production efficiency. Meanwhile, in order to consider the cost, the processor cannot be too much, because each processor will increase the overall cost, and will also cause resource waste, and therefore, the structure of one main control and one intelligent decision maker can improve the computing power while reducing the increase of the cost as much as possible. BRIEF DESCRIPTION OF DRAWINGS

[0009] In order to more clearly illustrate the technical scheme in the embodiments of the present application or the prior art, the drawings needed to be used in the following embodiment or prior art description will be briefly introduced, and obviously, the drawings described below are some embodiments of the present application, and those skilled in the art can also obtain other drawings according to these drawings without creating any creative labor.

[0010] Figure 1It is an intelligent electric control system diagram for a stretch-bending and straightening machine;

[0011] Figure 2 It is an intelligent decision maker principle diagram of an intelligent electric control system for a stretch-bending and straightening machine;

[0012] The accompanying drawings Figure 2 Mark 1, bottom driving layer, 2, data processing layer, 3, software application layer, 4, human-computer interface layer, 5, sensor, 6 bottom equipment, 7, real-time operating system, 8, data storage, 9, data processing, 10, data acquisition, 11, function management system, 12, analysis process data, 13, system management service, 14, state monitoring algorithm model, 15, algorithm model application, 16, host computer server, 17, model arrangement, 18, system configuration. DETAILED DESCRIPTION

[0013] The technical solutions in the examples will be described clearly and completely below in combination with the drawings in the examples. Obviously, the described examples are only a part of the examples of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the protection scope of the present application.

[0014] As Figure 2 shown: 1, an intelligent electric control system for a stretch-bending and straightening machine, refers to the internal control system including two controllers of PLC master control and intelligent decision maker. The PLC master control is used for the running control of the machine train, controls the machine train speed, hydraulic pressure, machine set, etc. The existing machine train uses a separate PLC to control the running and speed adjustment of the whole system. The master control PLC itself also has the functions of sensor data acquisition and processing, and makes simple data logic judgment to generate a control strategy according to the sensor data. However, due to the limitation of the algorithm, the master control PLC has a relatively small data operation amount in terms of accessing and processing sensor signals. The present scheme proposes an intelligent electric control system for a stretch-bending and straightening machine which adds an intelligent decision maker. The system is a functional expansion of the PLC, and its basic unit is an industrial computer which is installed with a general Linux system. The characteristics are strong computing power and strong functional expansion capability. The main functions include flexible access of sensors and fusion of data acquisition and processing, discovery of board surface quality problems by video sensors, and feedback to the PLC to issue commands to the execution equipment. Especially, the sensor data which cannot be processed by the master control PLC itself, such as high-frequency vibration data, video data, audio data, etc., are first accessed. At the same time, the original PLC accessed sensor data, such as temperature, speed, etc., are also accessed, to further release the pressure of the PLC and make the PLC better complete the real-time control function. Another important function of the intelligent decision maker is to generate a control strategy based on various diagnostic models and transmit the strategy to the PLC to generate a control action. The intelligent decision maker calculates and optimizes the control instructions based on the machine set parameters through the AI model.

[0015] The intelligent decision maker can also undertake some non-real-time work previously undertaken by PLCs, such as plant server communication, reporting of existing train running status, etc. The intelligent decision maker is composed of a bottom drive layer 1, a data processing layer 2, a software application layer 3, and a man-machine interface layer 4. The bottom drive layer 1 is composed of sensors 5, bottom equipment 6, and a real-time operating system 7. The data processing layer 2 is composed of data storage 8, data processing 9, and data acquisition 10. The software application layer 3 includes a function management system 11, analysis of process data 12, system management services 13, algorithm model application 15, and state monitoring algorithm model 14. The man-machine interface layer 4 includes system configuration 18 and model arrangement 17.

[0016] I. Main features of the PLC-based master control system

[0017] The system has high reliability and strong anti-interference capability. The powerful functions of the microcomputer, electromagnetic interference, power fluctuations, mechanical vibration, temperature and humidity changes in the industrial field can all cause general-purpose microcomputers to malfunction. Traditional relay and contactor control systems have strong interference capability, but they have a large number of mechanical contacts (prone to wear and tear) and short service life, and the system has poor reliability. PLCs use microelectronic technology, and most of the switching actions are completed by non-contact electronic storage devices. Most relays and complex wiring are replaced by software programs, which greatly improves the service life and reliability. In actual use, the average failure-free time of a PLC control system is generally 40-50,000 hours. PLCs have a series of hardware and software anti-interference measures to adapt to various strong interference industrial sites and have fault self-diagnosis capabilities. For example, a general PLC can resist 1000V, 1ms pulse interference, and its working environment temperature is 0-60°C, without forced air cooling. It is small in size, light in weight, low in power consumption, and easy to maintain.

[0018]

[0019] 1. Simple programming, easy to master

[0020] At present, most PLCs still use the "ladder diagram programming method" of relay control. It inherits the clear and intuitive control circuit diagram of traditional control and takes into account the reading habits and programming level of most factory electrical technicians, so it is very easy to accept and master. The symbol and expression method of the programming elements of the ladder diagram language are quite close to the principle diagram of the relay control circuit. Through reading the user manual of the PLC or short-term training, electrical technicians and technical workers can quickly learn to use the ladder diagram to compile control programs. At the same time, functional diagrams, statement tables, and other programming languages are also provided.

[0021] ​PLC in the implementation of the ladder diagram program, with the interpreter to translate it into assembly language and then execute (PLC internal increase in the interpreter). Compared with the direct execution of user programs written in assembly language, the execution of the ladder diagram program is a bit longer, but for most mechanical and electrical control equipment, it is negligible, fully meet the control requirements.

[0022] 2、Reduced the design and construction of the control system workload

[0023] Because of the PLC uses software to replace the relay control system in a large number of intermediate relay, time relay, counter and other devices, the design and installation of control cabinet wiring workload is greatly reduced. At the same time, the user program of PLC can be simulated in the laboratory debugging, more reduce the workload of the field debugging. And, due to the low failure rate of PLC and strong monitoring function, modularization and so on, the maintenance is very convenient.

[0024] 3、Strong, wide adaptability

[0025] Modern PLC not only has the functions of logic operation, timing, counting, sequence control, but also has the functions of digital and analog input and output, power drive, communication, man-machine dialogue, self-checking, recording display. It can control a production machine, a production line, and a production process.

[0026] Disadvantages: low computing power, difficult to meet the requirements of modern industrial data analysis, such as real-time data analysis of sensors, real-time data analysis of images, etc.

[0027] Two, the introduction of intelligent decision maker system

[0028] Intelligent decision maker assists the main control PLC to complete the big data processing unit of the unit control, its basic composition includes: 1, the bottom layer of the drive; 2, data processing layer; 3, software application layer; 4, man-machine interface layer.

[0029] 1、Bottom drive layer

[0030] This part of the function is mainly the bottom support of the system, the drive mainly refers to the sensor acquisition board card drive,

[0031] Used to connect and analyze sensor data; refers to the operating system and basic function support, such as file system, database system, time system, network system, etc.

[0032] 2、Data processing layer

[0033] The data layer includes three main modules of data acquisition, data processing and data storage.

[0034] a) Data acquisition: including the collection of various data sources, such as vibration sensors, video sensors, photoelectric sensors, and master data, including time, environment, internal environment, and operation of the unit;

[0035] b) Data processing: including data processing of individual sensors and data fusion of multiple sensors, such as time-frequency changes of vibration sensors and data alignment of multiple signals.

[0036] c) Data storage: including the storage of four types of data, 1) real-time data stored in memory; 2) short-term time series data stored in a time series database; 3) long-term feature data stored in a local relational database; 4) big data storage stored in local files.

[0037] 3. Software application layer

[0038] The software application layer includes system services and core model applications of the entire system. System services: including various request services for human-computer interaction, such as system login, system status, system configuration, etc.; system function management, such as time management, disk space management, account password management, etc.; model management services, including model addition, deletion, modification, and query functions; log services including log viewing, content downloading, content reading and writing operations, etc.; model scheduling is to execute model services according to model scheduling settings.

[0039] Model application: mainly refers to providing data query API for models during model execution, providing data for models, and saving models to local through model write API while sending control strategies to PLC master system or reporting model results to upper computer server according to rules.

[0040] 4. Human-machine interface layer

[0041] is an interactive tool for field operation and maintenance personnel and model developers. It provides configuration management functions for intelligent decision makers, mainly including configuring communication IP of intelligent decision makers and master PLC, debugging sensor communication, viewing running status of intelligent decision maker system, viewing model execution, uploading models, downloading model results and model logs, etc. It provides model development test environment for model developers, updates models, manually executes models, downloads model running results, logs, etc.

[0042] Figure 1 Features of a stretch-bending and straightening intelligent electric control system:

[0043] 1. High performance, using high-performance X86 structure processor, and adding GPU if necessary, capable of mechanical learning, deep learning, and other high-performance operations, and capable of real-time processing of vibration data, sound data, and video data.

[0044] 2、Functionality strong: the system uses stable industrial-grade Linux operation, support most of the mathematical library, machine learning library, and development of voice, python, c / c++, java, R, go, etc. Can also install a variety of application software, graphical interface, image processing software, Firefox browser.

[0045] 3、External interface can be developed independently, for different needs of the unit, you can customize the development of different external interface board. Currently developed IEPE vibration interface board card, A / D conversion board card, D / A conversion board card, multi-channel sound access, etc. Professional board.

[0046] 4、Communication mode is flexible, all kinds of communication protocol can be realized by software, reduce the cost of hardware development, such as modbus, opcua, can, etc., can be accessed through the TCP interface, reduce the interface type.

[0047] 5、Disadvantages:

[0048] a) poor real-time, due to high CPU usage, easy to cause communication congestion. Impact on the overall real-time.

[0049] b) development workload, all kinds of application without ready-made components need to be developed, in order to adapt to various communication protocols, need to develop a variety of software converter for the system, increase the development workload.

[0050] Through the comparison of the main control PLC and intelligent decision maker function, the following conclusions are obtained:

[0051] 1、Simple PLC control and intelligent decision maker control, both have obvious defects, PLC real-time good, but lack of processing capacity of big data, intelligent decision maker data processing capacity, but poor real-time, development workload of function realization;

[0052] 2、At present, the unit is basically single main control PLC unit, its control ability has basically reached the upper limit, its potential for development is very weak. In order to ensure the safety of the unit, increase another controller is inevitable technical development evolution process.

[0053] 3、If the PLC and intelligent decision maker advantage together, the whole control system will be in performance and function can make up for each other, in the system function performance can be greatly improved;

[0054] 4、Intelligent decision maker not only in the running performance can share the main control pressure, but also in the whole communication network

[0055] Reduce the master control pressure, because of its powerful computing power, originally some field level data processing can also be placed in intelligent decision maker complete, and then push the processed data to the field server, reduce the communication pressure.

[0056] Third, the beneficial technical effects produced by the present application

[0057] 1. A kind of intelligent electric control system of stretch bending correction changes the existing machine group control system prototype, let the function that previous single master control machine group cannot realize, such as image prototype, big data diagnosis etc. It is possible to realize, such as image prototype, big data diagnosis etc.

[0058] 2. An intelligent electric control system of stretch bending correction, liberates the pressure of master control, from image acquisition to actuator action must be completed in extremely short time in millisecond level, can ensure that the defect position reaches adjustable area or execution point before completing decision and output. Let master control focus on real-time control, intelligent decision maker is used to carry out logic operation and big data processing, let control and logic be separated, greatly increase the fault tolerance of machine group, make machine group safer.

[0059] 3, the introduction of intelligent decision maker, can make control more accurate, increase the fine control of machine group, this time choose high-quality light source and equip closed loop brightness control, increase the brightness, uniformity of light source, significantly improve the image quality and detection stability, make the whole system run more reasonably, provide system basis for machine group weight reduction, reduce cost.

[0060] Four, compared with the prior art, the advantages and beneficial effects of the present embodiment are:

[0061] 1. Intelligent decision maker, master control+intelligent decision system, it is a highly integrated, automated closed-loop control system, which significantly improves the intelligent level of the stretch bending production line and the product quality control ability.

[0062] 2. Master control is used for machine train control, and intelligent decision maker is used for logic processing of real-time data of machine train operation, which separates control and data analysis to reduce the burden of master control and increase the execution efficiency of control command.

[0063] 3. The intelligent decision maker provides a guarantee for the safety of the entire machine group with its super data processing and data analysis capabilities.

[0064] 4. Intelligent decision maker is similar to human cerebellum and brain, which makes the entire control system more robust, the system runs more reasonably, and the system is more intelligent.

[0065] 5. As described above, an intelligent electric control system of stretch bending correction breaks the existing machine control structure from the structure, making fine control of the machine group possible.

[0066] It is apparent that the application can be carried out by other embodiments that do not depart from the spirit or essential characteristics thereof. Thus, the embodiments disclosed in this specification are to be considered in all respects as illustrative and not restrictive, and all changes coming within the meaning and equivalency range of the appended claims are intended to be embraced therein.

Claims

1. A smart electric control system for stretch-bending-straightening, characterized in that it comprises two controllers, a PLC master controller and a smart decision maker, which divides the master control into two parts, a real-time control part and a big data decision part, the implementation part uses a traditional PLC to control the running function of the machine train of the stretch-bending-straightening line, and the big data decision part uses a high-performance smart decision maker to generate a control strategy after fusing and processing various data, and drive the master PLC to execute a control action, the smart decision maker main logic judges, and the smart decision maker can also communicate with a plant-level server to report some information of the equipment to the plant-level monitoring system, the smart electric control system decouples control and data processing to make the entire data processing more reasonable. The smart decision maker assists the master PLC to complete the big data processing unit of the stretch-bending-straightening control, wherein the smart decision maker is composed of (1) a bottom drive layer, (2) a data processing layer, (3) a software application layer, and (4) a human-machine interface layer. The bottom drive layer (1) comprises sensors (5), bottom equipment (6), and a real-time operating system (7), the data processing layer (2) comprises data storage (8), data processing (9), and data acquisition (10), the software application layer (3) comprises system management services (13) and analysis process data (12), function management system (11), algorithm model application (15), and state monitoring algorithm model (14), and the human-machine interface layer (4) comprises system configuration (18) and model arrangement (17).

2. The intelligent electric control system for stretch-bending and straightening according to claim 1, characterized in that, Further comprising: The smart decision maker is a functional expansion of the PLC, its basic unit is an industrial computer, and a Linux system is installed, which has strong computing power and mainly includes flexible access of sensors and collection, processing and fusion of data sources. The data sources include video data, audio data, piezoelectric vibration data and optical fiber stress sensing data, and the sensors connected to the original PLC system, such as speed and temperature sensors, are connected to the smart decision maker, which can further reduce the pressure of the PLC.

3. The intelligent electric control system for stretch-bending and straightening according to claim 1, characterized in that: An important function of the smart decision maker is to generate a control strategy based on various diagnostic models, find the surface quality problem of the plate with a video sensor, and pass the strategy to the PLC to generate a control action, and the smart decision maker calculates and optimizes the control instruction based on the unit parameters through an AI model.

4. The intelligent electric control system for stretch-bending and straightening according to claim 1, characterized in that: The bottom drive layer (1) is the bottom support of the system, and the drive mainly refers to the sensor collection board drive for connecting and analyzing sensor data, and the system is an operating system and functional support, including a database system, a time system and a network system.

5. The intelligent electric control system for stretch-bending and straightening according to claim 1, characterized in that: The software application layer (3) includes the entire system service and core model application, mainly including system management services (13) and algorithm model application (15).

Images