Please elaborate on the working principle of GPS time synchronization in power, post and telecommunications, communication and other networks. thank you

The equipment is equipped with a GPS clock chip, which synchronously receives GPS clock signals (receiving tick at regular intervals). The device uses GPS as the main clock source. At other times, the system is calibrated based on this clock source. For more information, please refer to the following:

GPS clock system

GPS clock system is a high-tech product, which is used to time computers and control equipment in automation system. GPS digital products obtain standard time signals from GPS satellites, and transmit these information to the equipment in the automation system (computer, protection device, fault recorder, event sequence recorder, safety automatic device and remote RTU) through various interface types, thus realizing the time synchronization of the whole system. With the rapid development of computer and network communication technology, the era of digitalization and networking of thermal automation system in thermal power plants has arrived. On the one hand, it provides a better platform for data exchange, analysis and application between various control and information systems, on the other hand, it also puts forward higher requirements for the accuracy of various real-time and historical data timestamps. Unify the clocks of all systems in the factory with cheap GPS clock.

Junfeng Ye Wei Company Clock System Picture

, has become a standard practice adopted in the design of thermal power plants. The master clocks of distributed control system (DCS), auxiliary system programmable controller (PLC), plant-level monitoring information system (SIS) and power plant management information system (MIS) in power plants obtain the standard TOD (year, month, day, minute and second) time through appropriate GPS clock signal interfaces, and then according to their respective clock synchronization mechanisms, the slave clock deviation in the system is limited to a small enough range, thus realizing the whole system. 1.GPS Clock and Output 1. 1 GPS Clock Global Positioning System (GPS) consists of a group of satellites launched by the US Department of Defense in 1978. * * * There are 24 satellites operating in six geocentric orbital planes. According to time and place, the number of satellites visible on the earth has been 4 to 160. A GPS clock is a receiving device, which receives low-power radio signals transmitted by GPS satellites and calculates GPS time. In order to obtain accurate GPS time, the GPS clock must first receive signals from at least four GPS satellites and calculate their three-dimensional positions. After obtaining the specific position, the GPS clock can guarantee the accuracy of the clock as long as it receives 1 GPS satellite signals. As the standard clock of thermal power plant, our basic requirements for GPS clock are: at least 8 satellites can be tracked at the same time, the cold and hot start-up time is as short as possible, spare batteries are equipped, and the clock output signal is high-precision and flexible. 1.2 GPS clock signal output At present, there are mainly three kinds of GPS clock output signals used in power plants:1.2.1PPS/1ppm. The time signal in this format outputs a pulse every second or minute. Obviously, the clock pulse output does not contain specific time information. 1.2.2 IRIG-B output IRIG (American shooting range instrument group) * * There are several coding standards (IRIG standard 200-98). Among them, IRIG-B coding is the most widely used in clock synchronization applications, including bc level shift (DC code), 1kHz sinusoidal carrier amplitude modulation (AC code) and other formats. IRIG-B signal is output one frame per second (1fps), and each frame is one second long. A frame * * * has 100 symbols (100pps), and the width of each symbol is 10ms. Binary 0, 1 and position mark (p) are represented by symbols with different positive pulse widths, as shown in figure 1.2.2-65442. For understanding, figure 1.2.2-2 shows an output example of IRIG-B time frame. Among them, seconds, minutes, hours and days (days since June 65438 +654381October +0) are represented by BCD codes, while the control function (CF) and standard binary seconds (SBS) are filled with a series of binary zeros (CF and SBS are optional, but not used in this example). 1.2.3 RS-232/RS-422/RS-485 output This clock output sends a string of date and time messages in ASCII code through the EIA standard serial interface, and outputs them once per second. Parity check, clock status, diagnostic information, etc. Can be inserted into a time message. At present, there is no standard format for this output. The following figure takes 17 byte transmission standard time as an example: the application of 1.3 GPS clock in power automation system. There are many systems or devices that need to be synchronized with GPS clock in power automation system, such as DCS, PLC, NCS, SIS, MIS, RTU, fault recorder, microcomputer protection device, etc. To determine the GPS clock, we should pay attention to the following points: (1) These systems belong to thermal control, electrical and system specialty. If it is decided to realize time synchronization by GPS clock provided by DCS manufacturers (the current common practice), professional cooperation should be carried out before DCS contract negotiation to determine the requirements of clock signal interface. (Generally, GPS clocks can be configured with different numbers and types of output modules. If the relevant requirements cannot be determined in advance, there should be room for adjustment in the corresponding contract terms. (2) Whether each system * * * uses a set of GPS clock equipment should be comprehensively considered according to the difficulty of system clock interface coordination and the geographical location of the system. If there are great differences in the interface type or accuracy requirements of GPS clock signals between different professions, you can configure your own GPS clock, which can reduce the mutual constraints between professions and make the clock synchronization scheme of each system easier to realize. In addition, when the systems are far away (for example, the chemical water treatment workshop and desulfurization workshop are far away from the centralized control building), in order to reduce the electromagnetic interference caused by the long-distance transmission of clock signals, a GPS clock can also be set on site. A separate GPS clock is also beneficial to reduce the impact of clock failure. (3)IRIG-B code has high reliability and standard interface. If the clock synchronization interface is optional, you can use it first. However, it should be noted that IRIG-B is only a general term for class B codes, and it is divided into many types (such as IRIG-B000) according to whether the codes are modulated or not, and whether there are CF and SBS. Therefore, the clock receiver should be equipped with a corresponding decoding card, otherwise accurate clock synchronization cannot be achieved. (4) 1PPS/ 1PPM pulse does not transmit TOD information, but it has high synchronization accuracy and is often used for clock synchronization of SOE module. Although RS-232 time output is widely used, due to the lack of standard format, it is necessary to pay special attention to confirm whether the clock message format of the sender and receiver of the clock signal can reach an agreement. (5) Although the control and information systems in the thermal power plant are interconnected, the clock synchronization protocols of each system may be different, so it is still necessary to access the GPS clock signal separately. Even for DCS and public DCS connected by bridge, if the clock synchronization signal is delayed in the network, it should be considered to synchronize with GPS clock respectively. Second, Siemens TELEPERMXP clock synchronization method Here, take Siemens TXP system as an example to see how DCS and clock are synchronized. TXP power plant bus is an Ethernet based on CSMA/CD, and there are two master clocks on the bus: real-time transmitter (RTT) and a communication/clock card of AS620 and CP 1430. Under normal circumstances, RTT is the master clock of TXP system. When it dies for about 40 seconds, CP 1430 as the backup clock will be automatically replaced (in fact, two pieces can be configured on ES680). See figure 2- 1. RTT can run freely or synchronize with external GPS clock through TTY interface (20mA current loop). Synchronization with GPS clock includes serial message (32 bytes long, 9600 baud, 1 start bit, 8 data bits and 2 stop bits) and second/minute pulse. RTT generates and sends the master clock synchronization message at the network layer and sends it to the power plant bus every 10 second. RTT sending time information can wait for 1ms at most. If the message can't be sent to the bus within 1ms, cancel the sending of this message; If the message sending process is interrupted, a message at the current time will be generated immediately. The clock message has a multicast address and a special frame header. The date is from 1984.0 1 to the number of days of the day, and the time is from 00: 00: 00,000 h of the day to the current ms value. The resolution is10 ms. OM650 obtains the time message from the power plant bus. In OM650, Unix functions are used to transmit time to SU, OT, etc. On the terminal bus. Generally, PU is used as a time server, and other OM650 devices log in as inbound customers. The AP of AS620 automatically realizes clock synchronization with CP 1430 by calling the "synchronization" function block after startup. Then CP 1430 synchronizes with AP every 6s. The accuracy of TXP clock is as follows: From the above-mentioned TXP clock synchronization mode and clock accuracy, it can be seen that every incoming clock in TXP system adopts master-slave hierarchical synchronization mode, that is, the lower clock is synchronized with the upper clock, and the higher the accuracy. Third, the clock and clock synchronization error 3. 1 clock error As we all know, computer clocks generally use quartz crystal oscillators. The crystal oscillator continuously generates clock pulses with a certain frequency, and the counter accumulates these pulses to get the time value. Because the pulse of the clock oscillator is influenced by many unstable factors such as ambient temperature, load sharing capacitance, excitation level, crystal aging and so on, the clock itself inevitably has errors. For example, the hourly error of a clock with an accuracy of 20ppm is (/kloc-0 /× 60×××1000 ms )× (20/10.6) = 72 ms, and the accumulated error in one day can reach1.73s; If the working environment temperature is changed from the rated 25℃ to 45℃, an extra error of 25ppm will be added. It can be seen that if the clocks in DCS are not synchronized regularly, the error after a period of free operation can reach an unbearable level in system application. With the development of crystal oscillator manufacturing technology, there are various crystal oscillators with high stability to choose from in applications that need high-precision clocks, such as TCXO (Temperature Compensated Crystal Oscillator), VCXO (Voltage Controlled Crystal Oscillator) and OCXO (oven controlled crystal oscillator). 3.2 Clock Synchronization Error If the clock synchronization method similar to TXP is analyzed, it is not difficult to find that the DCS absolute timing error generated in the clock top-down synchronization process can be composed of the following three parts: 3.2. The error between the1GPS clock and UTC (Coordinated Universal Time) launched by the satellite is determined by the accuracy of the GPS clock. For the output of 1PPS, the accuracy is generally between tens of ns and 1μs with the pulse leading edge as a point edge; Taking the serial output of IRIG-B code and RS-232 as an example, based on the deviation between the leading edge of 1PPS or the beginning of reference symbol and the leading edge, the synchronization accuracy of the ground clock products of National Time Service Center of Chinese Academy of Sciences is 0.3μs and 0.2ms respectively. 3.2.2 Synchronization error between DCS master clock and GPS clock. The master clock and GPS clock on DCS network are synchronized through "hard wiring". Usually, the hardware of standard time coding and GPS clock output is received by the clock synchronization card in DCS station. For example, if the transmission delay of ASCII code bytes output by RS-232 is compensated at the receiving end, or if IRIG-B code is encoded by symbol carrier period counting or high-frequency pin-phase decoding card, the synchronization accuracy between the master clock and the GPS clock can reach high accuracy. 3.2.3 Synchronization error of master and slave clocks of DCS stations. The master and slave clocks of DCS stations are synchronized through the network, and there is a time delay in the sending, propagation and processing of clock messages. As shown in the following figure: (1) When the master clock generates and sends the time message, the kernel protocol processing, the call overhead of the operating system to the synchronization request, the time to send the time message to the network communication interface, etc. (2) Before the time message goes online, it must wait for the network to be idle (for Ethernet) and then resend it to prevent conflicts; (3) After surfing the Internet, the time message needs to be transmitted from the master clock end to the slave clock end through the DCS network medium within a certain period of time (the propagation speed of electromagnetic wave in optical fiber is 2/3 of the speed of light, and for DCS LAN, the propagation delay is several hundred ns, which can be ignored); (4) After confirming that it is a time message from the network communication interface of the clock terminal, it takes time to accept the message, record the arrival time of the message, send an interrupt request, calculate and correct the slave clock, etc. These delays will more or less lead to time synchronization errors between the master clock and the slave clock of DCS and between the slave clocks. Of course, different DCS network types, different clock communication protocols and different synchronization algorithms will make the synchronization accuracy of network time service different. The above analysis is only based on general principles. In fact, with people's unremitting research on network clock synchronization technology, various complex but efficient and high-precision clock synchronization protocols and algorithms have appeared and been applied in practice. For example, Network Time Protocol (NTP), which is widely used on the Internet, has been able to provide the timing accuracy of 1 ms on DCS local area network (such as GE's IC distributed control system), while Standard Precise Time Protocol (PTP) based on IEEE 1588 can synchronize the master and slave clocks on Ethernet in real time. Fourth, clock accuracy and SOE design Although the scanning rate of general switches in DCS has reached 1 ms, in order to meet the requirements of SOE resolution ≤ 1 ms, people have been using this design method for a long time, that is, all SOE points are placed under one controller and the switch signals triggered by events are hardwired into SOE modules. The reason is that there are some errors in the clocks of different controllers. In this regard, Siemens described the engineering practice of decentralized configuration of FUN B modules in its TXP system, but it was unable to realize SOE analysis of asynchronous clocks by 1ms SOE, and even the SOE event recording order was reversed due to the clock difference of nearly 100 ms ... So, how to solve the problem without excessively reducing SOE resolution, Meet the requirements of the project for the decentralized design of SOE (for example, after the public DCS is set, the SOE of the unit should be separated from the SOE of the public system, or the trip signals of MFT and ETS that want to enter the controller can be used for SOE without output and then returned to the SOE module)? Through the analysis of DCS products, it is not difficult to find that the usual method is to synchronize the clock of controller or SOE module directly with the external GPS clock signal. For example, in ABB Symphony, the timekeeping master module (INTKM0 1) of SOEServerNode (generally located on the public DCS network) accepts IRIG-B time coding, and the RS-485 clock synchronization signal generated by it is linked to the SOE time synchronization module (LPD250A) of each controller (HCU), and its onboard hardware timer clock can be externally connected with 65430. For another example, the distributed processing unit (DPU 4E) of Max 1000+Plus can be synchronized with IRIG-B, so that the DI point of DPU can be used as SOE at the same time. Due to the adoption of 1PPM or RS-485 "external synchronization" and IRIG-B hard-line clock, the problem that the accuracy of DCS clock is still poor after network synchronization is avoided, and the control clocks can be synchronized with each other. It can be seen that the design scheme of SOE should be determined by combining the characteristics of DCS in engineering design. The switching scanning rate of 1ms or the clock relative error of the controller (or SOE module) of 1ms cannot be equal to the SOE resolution of 1ms, thus simply dispersing SOE points in the whole system. At the same time, it should be noted that although the resolution of SOE point "dispersion" is lower than that of "concentration", as long as the relative error of the clock is small (for example, the difference from 1ms is one order of magnitude), it can fully meet the actual needs of power plant accident analysis. V. Conclusion 5. 1 At present, the control system of thermal power plant is no longer an independent information island, and a large number of real-time data need to be time stamped in different places, and then sent to SIS and MIS for various applications. Therefore, the clock synchronization schemes of various systems and the clock synchronization accuracy to be achieved should be carefully considered in the design. 5.2 In the design of DCS, we should not only pay attention to the absolute timing accuracy of master and slave clocks, but also pay attention to the relative error between clocks. Because if we want to design SOE points in a decentralized way without excessively reducing the event resolution, the key is to make the deviation of each clock as small as possible. 5.3 There is every reason to believe that with the continuous development of network clock synchronization technology, it will become very common to synchronize the clocks of high-precision systems through the network. In the future, the timing accuracy of power plant systems will be greatly improved, and applications based on high-precision clocks, such as SOE point decentralized design, will continue to appear.