Time analysis of transmission of control signals through multi-stage links on industrial networks in order to implement wide control loops with high flexibility.

Master thesis in electrical engineering (control)

Abstract

Temporal analysis of control signal transmission through multi-stage links on industrial networks in order to implement wide control loops with high flexibility

The advancement of technologies related to communication networks in recent decades and Their expansion in the upper layers of the industry, such as the monitoring and management layers, requires that methods for using these networks at the lower levels, i.e. networking devices and sensors, be invented and used, each of which has advantages and disadvantages compared to traditional methods. Profibus mentions the basic information and hierarchical levels of industrial automation and its protocols. In the following, the basic requirements of the design and communication of different parts of Ethernet and Profibus network will be described, and by mentioning the advantages and disadvantages of each, we will show how we can use high-speed but non-real-time networks such as Ethernet in processes that require real-time data, and finally, by combining high-level networks (Ethernet) with lower-level networks (such as Profibus), we will examine the time analysis of the transmission of control signals through multi-stage links.

Due to the widespread use of Profibus network and industrial Ethernet network, in this thesis we will specifically focus on the simultaneous use of these two types of networks for the exchange of control signals and we will study the behavior of time-varying delay, error in sending and other cases in the transmission of signals through both networks and the effect they can have on the performance of the control system with the aim of using the results of the analysis to clearly meet the real time constraints. determined for each control system. We will also seek to provide solutions to help meet these restrictions. 

When digital automation technologies became available in the sixties, they were used to improve and develop industrial automation systems. Concepts such as automatic industries [1] and automatic distributed control systems [2] were introduced in the field of industrial automation and the use of communication networks grew almost significantly. With the expansion of communication networks in industrial automation systems, information gathering and low-level control operations were entrusted to these networks. This development progressed to the point that today in a modern automation system, devices at different levels of the system transmit data through these communication networks. Therefore, efforts were made for international standardization in the field of networks, the important achievement of which was the industrial automation protocol MAP in line with the compatibility of communication systems. The MAP protocol was developed to overcome communication problems between different automation devices and was accepted as an industrial standard for data communication in factories. The performance and reliability of an industrial automation system actually depends on its communication network. In an industrial automation communication network, the improvement of network performance and its reliability and communication standardization are determined according to the size of the system and the increase in the amount of information [1]. In such rooms, there are no more old big panels [3] on which the shape of the process was drawn and equipped with many signal lights. Everything should be searched in computer screens or so-called HMI[4]. But the explorers behind these pages are looking for physical connections between the computer and the process, and with a brief search, they come across the nearby panels where the communication equipment is installed. And looking at the hardware communication equipment of the network at a glance, they realize that the network used is the famous industrial Ethernet network [5] [2].

Today, the Ethernet network has become so popular and common in office applications that many non-expert users are also familiar with its equipment such as hubs, switches, cables, etc.In any case, in the application of HMI, although it is possible in some cases and for some reasons, the above communication can be observed in other ways and through other industrial networks, but in modern systems today, it rarely happens that a network other than industrial Ethernet is used at the HMI level.

To clarify the topic, we will discuss the position of Ethernet and Profibus networks in this automation pyramid:

1-1- The position of Ethernet in the pyramid Automation

The structure of a comprehensive automation system, which includes various control and monitoring equipment, is compared to a pyramidal structure. In this structure, each category of equipment has a special place depending on the type and application. Based on this, different levels are defined for this pyramid and at each level they introduce relevant equipment along with usable industrial networks. The lowest level are sensors and actuators. As its name suggests, it is the surface where sensors and actuators are placed. One of the famous industrial networks used at this level is ASI [6]. The higher level is the field. At this level, equipment such as remote inputs, outputs, registers [7] and other field devices are placed, and the network used by them can be Profibus. When we go higher than the field level, we reach the control level. In this level, PLCs [8], DCS systems [9] and HMIs are placed, in some divisions the control level is divided into two levels, HMI and control; And finally, it is the highest level of management where management information systems such as production, maintenance, repair, sales and purchase systems are placed. In some cases, the information available in the control level cannot be used in raw form for the management level and must be processed. Therefore, the intermediate level between Indo is defined as MES[10]. But what needs to be noticed is that in the above pyramid, the more we approach from the lower level to the higher level, the more information is concentrated. Therefore, to move them, we need higher speed networks [3].

Figure 1?1: Automation pyramid [2]

For example, the information of field equipment that has a lot of dispersion is concentrated in one or more Remote I/O, and the information of several Remote I/O is concentrated in one PLC and the information of several PLCs is concentrated in one HMI system. Perhaps it is because of this concentration that the structure is displayed in a pyramidal form. Another point comes to mind from the concentration of information at higher levels. At these levels, the amount of information has increased and we need higher speed networks to move them. At lower levels, a network like ASI can move information at a maximum speed of 170 Kbps and a network like Profibus [11] can move information at a maximum speed of 12 Mbps. This speed may be slow for data exchange at high levels. Today, industrial Ethernet transfers information at a speed of 100 Mbps or 1 Gbps at high levels such as the Cell Level, and at the Management level, faster Ethernets such as 1 Gbps are usually used. Figure 1-2: Levels of industrial networks [4] The large amount of information at high levels cannot be used in low-level networks that have low speeds. But the question that comes to mind is why fast networks such as Ethernet are not used at low levels such as the field level.

After reading the next parts of this thesis, the dear reader will understand that in Ethernet, there is no certainty for sending data on time and the time of sending data may be different from the previous times. This difference is due to the feature of the access technique in Ethernet, which is called CSMA/CD, where there is a phenomenon of information collision. While networks such as Profibus use Token pass and Master/Slave methods, which, although slower than Ethernet, guarantee the timely arrival of information at a specific time [5]. However, despite all the above issues, today it is possible to connect to Ethernet for many types of devices used in automation, so that in addition to the controller and computer, other equipment such as operator panels (OP, TP) and drives and interfaces between I/O with a network known as Remote I/Os can be connected to the Ethernet network. When Ethernet is used at different levels of automation, as shown in the figure below, it has a great advantage, which is the uniformity of the network and the need to use diverse networks.