METHOD, SYSTEM AND GATEWAY FOR NETWORKING TIME-SENSITIVE FIELDBUSES

Description

Field of the invention

[0001] The invention relates to time-sensitive fieldbus systems, in particular to industrial fieldbus systems for controlling industrial systems. In particular, the invention relates to a method, a system, and a gateway for networking at least two time-sensitive fieldbuses. Background of the invention

[0002] Fieldbuses are used to connect various participant devices (e.g. field devices such as sensors, actuators, measuring probes, final control elements, etc. as well as controllers such as control or automation devices, programmable logic controllers, PLCs, etc.) in a system (e.g. industrial production plant, vehicle, building, etc.) for the purpose of communication. Communication takes place in particular, but not exclusively, between various field devices and a fieldbus controller, which controls and monitors the processes taking place in the system. A participant device can be a field device or a controller and, in particular, has an interface with the fieldbus.

[0003] The messages to be transmitted over a fieldbus (e.g., data blocks, data streams) are at least partially time-sensitive, meaning they have real-time requirements regarding transmission latency. For example, such time-sensitive data requires that a message sent by a subscriber device be reliably received by the fieldbus controller or another field device connected to the fieldbus within a predictable time period.

[0004] The data to be transmitted between fieldbus nodes includes not only value communication, such as measured values, control values, or instructions, but also event communication for the temporal coordination of machines and system components. For example, the coordination and synchronization of movements of machine components is increasingly no longer carried out via mechanical devices (e.g., metal presses with rigidly coupled mechanical insertion and removal devices in the machine cycle, camshafts, gears, etc.), but via digital messages transmitted with real-time requirements via fieldbuses.

[0005] A fieldbus is assigned to a time domain in which all participant devices each have a timer (e.g. a clock). The timers of the participant devices are synchronized with each other within the time domain and thus have a common understanding of time. This means that all participant devices of a fieldbus can follow a defined fieldbus communication cycle and are coordinated with each other to carry out the appropriate actions at the right time. A participant device of the fieldbus acts as the master for the time domain, with the master's timer providing the time reference for the other participant devices in the time domain. The latter participant devices, which derive their time from the master within the time domain, are also referred to as slaves. The role of the time master in the fieldbus can be assumed by a fieldbus controller, but also by another participant device connected to the fieldbus.The fieldbus time master can synchronize its time base with another time base, for example, with the time master of another fieldbus or with a hierarchically superior time base. By synchronizing the masters of multiple fieldbuses with each other, a time domain can also encompass multiple fieldbuses. While synchronizing the time base aligns the "clocks," this does not necessarily mean that two fieldbuses within a time domain have the same communication cycles.

[0006] The first generation of fieldbuses was developed to replace the previously common parallel wiring of machine components using cable harnesses with digital transmission technology. Several, partly proprietary, fieldbus protocols were developed for this purpose. Later generations of fieldbuses are primarily based on the IEEE 802.3 Ethernet standard, with the Ethernet protocols being modified or extended to implement time-sensitive networks (TSNs).

[0007] Standards for real-time Ethernet protocols are being developed, for example, by the IEEE 802.1Q Task Group for TSN networks. A real-time scheduler (time-aware scheduler) is defined, for example, by the IEEE 802.1QBv standard. The scheduler enables communication within a network to be divided into fixed, periodically repeating communication cycles and, within these cycles, to define, reserve, or negotiate fixed time slots for the transmission of time-critical data. This enables clocked end-to-end transmission between two user devices.

[0008] Outside of the time slots designated for real-time transmission, non-time-sensitive data (best-effort data) can be transmitted using the well-known protocols for packet-switched transmission (e.g., Ethernet protocol). The real-time-supporting scheduler can establish guard bands to prevent transmissions of non-time-sensitive data packets from overlapping with the time slots reserved for real-time transmission.

[0009] Document US 2016 / 080533 A1 discloses a gateway that can transmit data between two networks of a vehicle. The first network is a time-sensitive FlexRay network and the second network is a vehicle Ethernet network.

[0010] Furthermore, TSN networks allow the transmission of non-time-sensitive data packets to be interrupted during real-time intervals and then resumed. This enables the coexistence of time-sensitive and non-time-sensitive communication in a TSN-enabled network.

[0011] With the increasing digitalization and networking of systems (e.g., an entire production site), there is a need to enable communication between fieldbuses as well as communication between a fieldbus and devices at a higher control level, process control level, operations control level, and / or corporate level. Networking affects both non-time-sensitive data and time-sensitive data.

[0012] Although most of the real-time fieldbus protocols in use today are based on Ethernet, despite this commonality, there are technical differences that make different fieldbus types incompatible or not fully compatible with each other. It may also happen that different generations of fieldbuses are used, for example, when parts of a plant are added, replaced, or modernized, whereby fieldbuses of different generations are not necessarily fully compatible. Efforts are currently underway to specify universally applicable real-time protocols that will replace the many proprietary and mutually incompatible protocols for time-sensitive fieldbuses. However, their use will require the replacement or upgrade of existing fieldbus systems, which is often not economically viable.

[0013] Existing solutions for time-sensitive communication between two or more fieldbuses assume that all connected fieldbuses have a uniform understanding of time, while also synchronizing all time offsets, durations, and / or start times of the fieldbus communication cycles. However, fieldbuses are often extensively configured at the beginning of a production process so that the machines and / or system components controlled by the fieldbus are optimally coordinated. However, adapting the cycle times of an existing fieldbus to newly added system components with additional fieldbuses would require reconfiguration of the existing processes controlled by the fieldbus.

[0014] There is therefore a need to network time-sensitive fieldbuses that are not fully compatible with each other. Summary of the invention

[0015] The object is achieved in particular by a method for networking a first time-sensitive fieldbus with a second time-sensitive fieldbus according to claim 1, by a system for networking at least two time-sensitive fieldbuses according to claim 9 and by a gateway for time synchronization and for networking at least two time-sensitive fieldbuses according to claim 13.

[0016] To solve this problem, a gateway is used that mediates communication between time-sensitive fieldbuses. Because the gateway supports the respective time domain of all connected fieldbuses, the fieldbuses can communicate with each other largely without synchronizing the time, and especially the times of the communication cycles. This applies to both time-sensitive and non-time-sensitive data.

[0017] One aspect of the present invention relates to a method for networking a first time-sensitive fieldbus with a second time-sensitive fieldbus. The first time-sensitive fieldbus comprises a first subscriber device and has its own first time domain. The second time-sensitive fieldbus comprises a second subscriber device and has its own second time domain. The first and second fieldbuses are connected to one another by means of a gateway for data transmission. The method comprises defining the first time domain in the first fieldbus by the first subscriber device of the first fieldbus and defining the second time domain in the second fieldbus by the second subscriber device of the second fieldbus, transmitting time-sensitive data in defined first time slots within the first fieldbus, and transmitting time-sensitive data in defined second time slots within the second fieldbus.The method further comprises synchronizing the gateway as a slave with respect to the first time domain of the first fieldbus and additionally with respect to the second time domain of the second fieldbus, wherein the gateway supports at least two time domains simultaneously. By supporting both time domains, the gateway is configured to transmit messages or data, such as data packets, between the fieldbuses while adhering to the communication cycles and time slots for time-sensitive data defined on the fieldbuses, without the respective time domains of the fieldbuses being synchronized with respect to absolute time and the cycle times of the communication cycles having to be adjusted.

[0018] In one embodiment, the method further comprises the steps of transmitting non-time-sensitive data from the first fieldbus to the gateway at a time outside the first time slots, buffering the non-time-sensitive data in the gateway, and transmitting the non-time-sensitive data received from the first fieldbus from the gateway to the second fieldbus at a time outside the second time slots. In this way, non-time-sensitive data can be transmitted while observing the first time slots of the first fieldbus and the second time slots of the second fieldbus defined on the respective fieldbuses.

[0019] In practice, it is possible that a time offset occurs between the first and second time domains. This time offset arises due to the different definitions of time in the first and second time domains, as well as from the difference between the start times of the communication cycles of the first and second fieldbuses. Furthermore, the time domains of the fieldbuses can have different frequencies. In particular, the clocks of the participant devices have specific clock frequencies that are frequency-synchronized within a time domain, but this is not the case between the time domains. Even if the clock frequencies in the time domains of the first and second fieldbuses are nominally the same, the local oscillators for the clocks can have slight differences, leading to frequency differences between the clocks.In the first and second time domains, different time periods can also be defined for the communication cycles, within which the first and second time slots for transmitting the time-sensitive data are periodically repeated. Furthermore, there can be temporal differences between the first and second fieldbuses, e.g., with respect to the first and second time slots for transmitting time-sensitive data.

[0020] In one embodiment, the gateway comprises a first and a second timer, and the method further comprises the steps of synchronizing the first timer as a slave with the first time domain of the first fieldbus and synchronizing the second timer as a slave with the second time domain of the second fieldbus. As a result, the gateway "knows" the time defined in the first and second time domains and can, for example, coordinate the transmission of non-time-sensitive data such that the time slots for transmitting time-sensitive data are not affected.

[0021] In one embodiment, initially only the frequency of the first timer of the gateway is set to the frequency of the second time domain of the second timer of the gateway, and the first timer of the gateway is set as the master for the first fieldbus in order to set the frequency of the first fieldbus to the frequency of the second time domain. In this way, synchronization of the clock frequencies in the time domains of the first and second fieldbuses can take place. Based on such minimal synchronization, the first and second fieldbuses or the first and second time domains can have differently defined times, but the time offset is time-invariant, i.e., constant over time. In particular, by only synchronizing the clock frequency of the timers in the first and second fieldbuses, a hard time jump in the first fieldbus is avoided, which could occur if the first and second fieldbus times were aligned.The frequency adjustment can be carried out over a period of time that does not affect the operation of the processes controlled in the first fieldbus.

[0022] In one embodiment, after frequency synchronization has been completed, the gateway determines the cycle time duration of the first time domain and the cycle time duration of the second time domain at a specific reference time. The latter reference time can be determined with respect to the time valid in the first and / or second time domain.

[0023] Furthermore, the gateway can determine a time offset between the first time domain and the second time domain at the reference time. The time offset includes the differences in the definition of time in the first and second time domains as well as the difference between the start times of the communication cycles of the first and second fieldbuses. The gateway can additionally communicate the cycle time duration of the first time domain to the second fieldbus and the cycle time duration of the second time domain to the first fieldbus. In this way, for example, the first participant device in the first time domain or the second participant device in the second time domain is able to determine and continue the cycle time duration and the time offset in the other fieldbus with respect to the reference time.

[0024] In one embodiment, the first subscriber device defines a first time slot in the first fieldbus for communicating time-critical data, and the second subscriber device defines a second time slot in the second fieldbus for communicating time-critical data. The time slots of the first fieldbus and the second fieldbus overlap in time. The first and second time slots may be identical (i.e., completely overlap), or they may at least have a temporal intersection in order to establish time-critical communication between the first fieldbus and the second fieldbus.

[0025] In one embodiment, time-critical data is passed through the gateway from the second fieldbus to the first fieldbus and / or vice versa during the temporally overlapping first and second time slots. This enables clocked end-to-end transmission between two subscriber devices across fieldbus boundaries.

[0026] A further aspect of the invention relates to a system for networking at least two time-sensitive fieldbuses, comprising a first time-sensitive fieldbus having a first subscriber device and its own first time domain, wherein the first time domain includes first time slots for transmitting time-sensitive data. The system further comprises a second time-sensitive fieldbus having a second subscriber device and its own second time domain, wherein the second time domain includes second time slots for transmitting time-sensitive data.The system further has at least one gateway that connects the first and second fieldbuses for data transmission, wherein the gateway has a first timer that is configured to synchronize as a slave with respect to the first time domain of the first fieldbus and the gateway has a second timer that is configured to synchronize as a slave with respect to the second time domain of the second fieldbus. By supporting both time domains, the gateway is configured to transmit messages or data, such as data packets, between the fieldbuses while observing the communication cycles and time slots defined for time-sensitive data on the fieldbuses, without the respective time domains of the fieldbuses being synchronized with respect to absolute time and without having to match the cycle times of the communication cycles.

[0027] In one embodiment, the gateway may have a buffer for buffering non-time-sensitive data of the first fieldbus transmitted from the first fieldbus and received by the gateway at a time outside of the first time slots, and for transmitting the buffered data from the gateway to the second fieldbus outside of the second time slots. This allows the transmission of non-time-sensitive data to occur at times when no time-sensitive data is being transmitted on either fieldbus.

[0028] The gateway can further be configured to receive information about the first and second time slots of the first and second fieldbuses and to take this information into account when transmitting the non-time-critical data. This prevents the non-time-critical data from being transmitted during a time slot of the second fieldbus reserved for time-critical transmission.

[0029] The subscriber device of the first fieldbus can be configured to define the first time slots, and the subscriber device of the second time slot can be configured to define second time slots, such that the first and second time slots overlap in time. The gateway can also be configured to forward time-critical data from the second fieldbus to the first fieldbus, or vice versa, during the overlapping first and second time slots of the first and second fieldbus.

[0030] A further aspect relates to a gateway for networking at least two time-sensitive fieldbuses with different time domains. The gateway has a first timer that is configured to synchronize itself as a slave with respect to the first time domain of the first fieldbus, and a second timer that is configured to synchronize itself as a slave with respect to the second time domain of the second fieldbus. By supporting both time domains, the gateway is configured to transmit messages or data, such as data packets, between the fieldbuses while adhering to the communication cycles and time slots for time-sensitive data defined on the fieldbuses, without the respective time domains of the fieldbuses being synchronized with respect to absolute time and the cycle times of the communication cycles having to be adjusted.

[0031] In one embodiment, the gateway is configured to transmit non-time-critical data between the first fieldbus and the second fieldbus at times or time periods outside of time slots for time-critical data of the first fieldbus or second fieldbus, respectively, wherein the gateway has a buffer for buffering received non-time-critical data of one fieldbus before sending it to the other fieldbus.

[0032] The gateway may further be configured to receive information about time slots of the first fieldbus and information about time slots of the second fieldbus and to take them into account when sending data.

[0033] Furthermore, the gateway can be configured to forward time-critical data from the first fieldbus to the second fieldbus or vice versa during overlapping time slots.

[0034] In one embodiment, the gateway is configured to take into account the information about the time slot of the first fieldbus and the information about the time slot of the second fieldbus when forwarding time-critical data.

[0035] Further embodiments of the invention emerge from the subclaims. Short description of the drawings

[0036] In the following detailed description, aspects of the invention are described with reference to the drawings, wherein Fig. 1 shows a block diagram of a system with a first time-sensitive fieldbus and with a second time-sensitive fieldbus; Fig. 2 schematically shows a timing diagram of a first and second fieldbus and the transmission of non-time-sensitive data through the gateway; Fig. 3 schematically shows a detailed timing diagram of the cycle times defined for a first and second fieldbus; and Fig. 4 schematically shows a timing diagram of a first and second fieldbus and the transmission of time-sensitive data through the gateway. Detailed description

[0037] The following detailed description explains exemplary embodiments of the invention with reference to the above-mentioned figures. In this case, identical reference symbols in different figures describe the same device, the same method step, the same time unit, etc.

[0038] Fig. 1shows a block diagram of an exemplary system 1 with a first time-sensitive fieldbus 10 and a second time-sensitive fieldbus 20. The first fieldbus 10 and the second fieldbus 20 comprise, by way of example, the participant devices 12, 14a-c, and 22, 24a-c, respectively, wherein the participant devices 12, 22 are each controllers of the fieldbus, for example an automation device, a programmable logic controller, PLC, a node or another industrial controller, and the participant devices 14a-c, and 24a-c are field devices, for example I / O modules for sensors and / or actuators, which can measure or influence variables of a process automated by the fieldbus. The subscriber devices 12, 14a-c, or 22, 24a-c are communicatively connected to each other in the respective fieldbus 10, 20 via an interface, whereby the definition of the interface includes an interface protocol.

[0039] The subscriber devices 12, 14a-c of the first fieldbus 10 belong to a first time domain 16, whereby all subscriber devices 12, 14a-c have the same understanding of absolute time. For this purpose, each subscriber device 12, 14a-c has a timer which is Fig. 1 schematically symbolized by the face of a clock. Correspondingly, the subscriber devices 22, 24a-c of the second fieldbus 20 belong to a second time domain 26. In the fieldbuses 10 and 20, the fieldbus lines are connected to the field devices in a star configuration, starting from the controller 12, 22. In general, fieldbuses can also be formed by other topologies, e.g., tree topology, bus topology, or ring topology. It is also possible for the subscriber devices 12, 24a-c, or 22, 24a-c of a fieldbus 10, 20 to be at least partially interconnected by a wireless network.

[0040] Each participant device 12, 24a-c, or 22, 24a-c sets the time for the respective fieldbus 10 or 20 and is therefore referred to as the master. For the first fieldbus 10, for example, the controller 12 is the master, which in Fig. 1 is shown as a clock face with solid lines. In the second fieldbus 20, for example, the subscriber device 24a is the fieldbus master. The clock of a master 12 or 24a specifies the time in the fieldbus, with the other subscriber devices of a fieldbus, i.e. subscriber devices 14a-c in the first fieldbus or subscriber devices 22, 24b-c, each deriving their time from the master. This occurs, for example, when the subscribers within a fieldbus exchange messages at regular or irregular intervals according to the Precision Time Protocol in accordance with the IEEE 1588 standard, and in this way achieve and maintain a common understanding of time.

[0041] Within the first fieldbus 10 and the second fieldbus 20, both non-time-sensitive and time-sensitive data can be transmitted, as will be described in more detail below. For example, data transmission in both fieldbuses can be divided into transmission cycles according to the IEEE 802.1Q standards, with first time slots defined in the first fieldbus 10 and second time slots defined in the second fieldbus 20, in which a clocked end-to-end transmission between two subscriber devices of the same fieldbus is possible.

[0042] System 1 further comprises a gateway 30, via which both fieldbuses 10, 20 are communicatively connected to each other. Fig. 1that the fieldbuses 10, 20 are connected to each other via their controllers 12 and 22, but this is not mandatory. Rather, the gateway 30 has an interface to each of the fieldbuses 10 and 20, so that it represents a subscriber device with respect to the respective fieldbus 10, 20. The gateway 30 supports at least two time domains 32 and 34. Additional time domains can be supported depending on the number of fieldbuses to be connected. In particular, the gateway includes timers 36 and 38, which can be configured for the first 16 and the second 26 time domains, respectively.

[0043] The gateway 30 can connect the first fieldbus 10 and the second fieldbus 20 for data transmission. For this purpose, the gateway 30 maintains a memory 35 for buffering or temporarily storing data. For example, non-time-sensitive data can be stored in the memory 35. However, the gateway 30 can also transmit a clocked end-to-end transmission between the first and second fieldbuses 10, 20 during defined time slots without delay.

[0044] Fig. 2 shows an exemplary time diagram of the first fieldbus 10 and the second fieldbus 20. Here, the first fieldbus 10 has communication or transmission cycles with periodically recurring cycle intervals TC1, which are determined in terms of position and duration according to the Fig. 2 upper timeline. Accordingly, the second fieldbus has 20 communication cycles with periodically recurring cycle intervals TC2, which are defined in terms of position and duration according to the Fig. 2lower timeline.

[0045] In the fieldbuses 10, 20, first time slots 50 and second time slots 52 for transmitting time-sensitive data within the respective fieldbus 10, 20 can be defined, negotiated, or determined. For this purpose, the real-time scheduler (time-aware scheduler) defined in the IEEE 802.1QBv standard can be used, for example. For example, the first time slots 50 and the second time slots 52 coincide with the beginning of the corresponding communication cycles. In general, the time slots 50, 52 for time-sensitive transmission can be located at any position within a transmission cycle. It is also possible to define several separate time slots 50, 52 for time-sensitive communication per transmission cycle.

[0046] Fig. 2also shows time periods 54 and 56, which belong to the respective transmission cycles that lie outside the time slots 50 and 52 reserved for the transmission of time-sensitive data. In these intervals 54 and 56, non-time-sensitive data can be transmitted between fieldbuses 10 and 20, for example, according to the Ethernet protocol, for which the CSMA / CD (Carrier Sense Multiple Access with Collision Detection) access procedure applies.

[0047] In one embodiment, the timers 36 and 38 of the gateway 30 are first synchronized as a slave with respect to the time domains 16, 26 of the fieldbuses 10 and 20. Since the gateway 30 has an interface as a subscriber device to both the first 10 and the second fieldbus 20, the gateway 30 also knows the corresponding cycle times TC1 and TC2, the timing of the communication cycles, and the time slots 50 and 52 in which time-sensitive communication is carried out in the respective fieldbuses 10, 20.

[0048] Accordingly, the intervals 54 and 56 to which non-time-sensitive data is transmitted are also known.

[0049] A transmission of non-time-sensitive data between the first fieldbus 10 and the second fieldbus 20 can be achieved by transmitting non-time-sensitive data 60 from the first fieldbus 10 to the gateway 30 during a first non-time-sensitive interval 54, as symbolized by the arrow 60. This data 60 can be temporarily stored in the buffer 35 of the gateway 30, as symbolized by step 62. Finally, the temporarily stored data can be transmitted from the gateway 30 to the second fieldbus 20 during intervals 56 outside the second time slots 52, as symbolized by the arrow 64. For example, in order to transmit the data 64 to the second fieldbus using the CSMA / CD method or another contention method, the gateway 30 must ensure that the second fieldbus 20 is free for transmission during the time slots 56.

[0050] Fig. 3schematically shows a detailed timing diagram of the cycle times defined for the first and second fieldbuses 10, 20 according to some aspects of the invention. In one example, the clock generators in the subscriber devices of the fieldbuses 10 and 20 may have different clock frequencies f1 and f2, respectively. Fig. 3 The inverse of the clock frequencies 1 / f1 and 1 / f2, namely the clock cycle duration, is represented by a black bar. Even if both clock frequencies f1 and f2 nominally (i.e., according to a specification) have the same frequencies, the clocks of time domains 16 and 26 of the two fieldbuses run at different speeds if the actual values ​​of the clock frequencies f1 and f2 differ slightly.

[0051] Furthermore, a time offset 58 can occur between the start times of the communication cycles. Such a time offset 58 arises, on the one hand, due to the different definition of time in the two fieldbuses 10, 20, and, on the other hand, from the difference between the start times of the communication cycles of the first 10 and second fieldbus 20, even if both fieldbuses had the same understanding of time. Furthermore, the communication cycles can have different durations TC1 and TC2. Therefore, in general, the time offset 58 can only be defined with reference to an (absolute) reference time TRef. The absolute reference time TRef can be expressed accordingly both in the time of the first fieldbus 10 and in the time of the second fieldbus 20. If the time durations TC1 and TC2 are in a rational relationship to one another, a certain time offset 58 returns after a predictable number of cycles of the first or second fieldbus.of the second fieldbus. For example, if the cycle time TC1 in the first fieldbus is 50 ms and the cycle time TC2 in the second fieldbus is 25 ms, the time offset 58 can be zero after a first cycle in the second fieldbus, 25 ms after a second cycle in the second fieldbus compared to the first fieldbus, and then zero again, and so on. However, if the time offset 58 is known at a time TRef, it is possible to determine the time offset 58 in advance for all communication cycles of the first and second fieldbus, even if the cycle times TC1 and TC2 are not in a rational relationship to one another.

[0052] Furthermore, different durations TS1 and TS2 of the first and second time slots 50 and 52 for time-sensitive communication can occur on fieldbuses 10 and 20, respectively, since these can be agreed upon independently for both fieldbuses 10 and 20. Furthermore, the first and second time slots 50 and 52 can exhibit temporal differences TD in their start times. These can generally only be predictably determined with reference to an absolute reference time TRef, since the start times of time slots 50 and 52 for time-sensitive communication can vary with the cycle times in the individual fieldbuses.

[0053] According to one embodiment, as described above, the first timer 36 of the gateway 30 is configured as a slave of the first fieldbus 10, and the second timer 38 of the gateway 30 is configured as a slave of the second fieldbus 20. To simplify communication between the first and second fieldbuses 10, 20, it may now be useful to synchronize only the frequencies f1 and f2 of the first time domain 16 and the second time domain 26. By synchronizing the timer clock frequencies in this way, the time offset between the times of both time domains 16, 26 remains stable, thereby simplifying the determination of times in the gateway 30 or in the subscriber devices 12, 14a-c or 22, 24a-c of the first and second fieldbuses with respect to the other time domain 16, 26. In one embodiment, the clocks of both time domains 16, 26 have the same nominal frequency.Frequency synchronization can now be achieved by designating the first timer 36 of the gateway 30 as the master for the first fieldbus 10, so that the gateway 30 can determine the timing in the first fieldbus. Furthermore, the frequency f1 of the first timer 36 of the gateway 30 can be synchronized to the frequency f2 of the second time domain 26. Synchronizing only the frequencies f1 and f2 in the time domains 16, 26 can be useful to avoid time jumps in the participant devices in the first fieldbus 10. The frequencies can be synchronized during ongoing operation of the first fieldbus 10 using an adaptation period that is selected to be long enough so that the processes of the first fieldbus 10 can run smoothly.

[0054] After frequency synchronization has been achieved, the gateway 30 can determine the cycle time durations TC1 and TC2 of the first and second time domains 16, 26 at a reference time TRef. Furthermore, the gateway 30 can determine the time offset 58 between the first and second time domains 16, 26 at the reference time TRef. The gateway 30 can further communicate the cycle time duration TC1 of the first time domain 16 to the second fieldbus 20 and the cycle time duration TC2 of the second time domain 26 to the first fieldbus 10. This can be done with reference to the reference time TRef. In this way, specific subscriber devices of the first or second fieldbus 10, 20 can predictably determine the communication cycles of the other fieldbus 20, 10.

[0055] Fig. 4shows a schematic timing diagram of a first fieldbus 10 and a second fieldbus 20 during a transmission of time-sensitive data. Since the cycle times TC1 and TC2 of the first and second time domains 16, 26 are known at a reference time TRef in both fieldbuses, the subscriber device 14c of the first fieldbus 10, for example, can transmit time-sensitive data to the subscriber device 24b of the second fieldbus 20. This can be achieved by the subscriber devices 14c and 24b in the respective fieldbus 10, 20 defining time slots 50, 52 for the communication of time-critical data, which at least partially overlap. For example, the subscriber devices 14c and 24b, provided they are configured to do so (e.g., as schedulers according to the IEEE802.1QBv standard), can exchange possible time intervals for time-sensitive communication as part of negotiations.For example, subscriber device 14c could send a request to subscriber device 24b to query possible time intervals and then determine a common time interval. It is also possible for subscriber device 14c and / or subscriber device 24b to have another configured subscriber device of the first or second fieldbus 10, 20 determine the first or second time slots 50, 52 for communicating time-critical data.

[0056] In the present example according to Fig. 4Corresponding subscriber devices of the first and second fieldbuses have agreed on common overlap areas OL1 and OL2 in time slots 50 and 52, respectively, and exchanged this information via the gateway 30, so that this information is also known to the gateway 30. It is also possible for the corresponding time slots 50 and 52 for time-sensitive communication to be shifted relative to the cycle interval of the respective first and second fieldbuses in the respective cycles, so that an overlap area OL1 and OL2 can be created in each cycle.

[0057] In one embodiment, it is not necessary to create an overlap area OL1, OL2 for time-sensitive communication in each cycle TC1 or TC2 of the first or second fieldbus 10, 20. For example, it may be sufficient to define an overlap area during every second, third, etc. cycle of the first or second fieldbus 10, 20.

[0058] If the first and second time slots 50 and 52, respectively, reserved for time-sensitive communication within the fieldbuses are known during the communication cycles of the first and second fieldbuses 10, 20, the overlap areas OL1 and OL2 for time-sensitive communication between the fieldbuses can be determined or calculated based on rules. Then, it is sufficient to determine a single pair of overlapping time slots 50 and 52 at a reference time TRef.

[0059] If the cycle times TC1 and TC2 are in a rational relationship to each other, regularly recurring overlap areas OL1 and OL2 result. In this case, the negotiation and / or determination of time slots 50 and 52 is simplified, so that in these cases an overlap area OL1, OL2 recurs regularly and therefore does not need to be negotiated or determined individually.

[0060] In one embodiment, the gateway 30 is further configured to transmit time-critical data 70, 72 during the temporally overlapping time slots, e.g., OL1 or OL2. In one embodiment, this is achieved by the gateway 30 determining the time slots for time-sensitive transmission 50 and 52 determined between the first and second fieldbuses 10, 20, and in particular the respective overlap intervals OL1, OL2, and passing the signals between the first and second fieldbuses during these intervals.

[0061] Using the described method, System 1, and Gateway 30, it is possible to network time-sensitive fieldbuses 10, 20 without having to align the fieldbus times and communication cycles of the fieldbuses. This allows existing and partially incompatible fieldbuses, in particular, to be configured for the exchange of both non-time-sensitive and time-sensitive data.

Claims

1. Method for networking a first time-sensitive field bus (10) with a second time-sensitive field bus (20), wherein the first time-sensitive field bus (10) comprises a first subscriber device (12, 14) and has its own first time domain (16), and wherein the second time-sensitive field bus (20) comprises a second subscriber device (22, 24) and has its own second time domain (26), wherein the first and the second field bus (10, 20) are connected to one another by means of a gateway (30) for data transmission, wherein the method has the following steps: - defining the first time domain (16) in the first field bus (10) by the first subscriber device (12, 14) of the first field bus (10); - defining the second time domain (26) in the second field bus (20) by the second subscriber device (12, 24) of the second field bus (20); - transmitting time-sensitive data in defined first time slots (50) within the first field bus (10); - transmitting time-sensitive data in defined second time slots (52) within the second field bus (20); and - synchronizing the gateway (30) as a slave with respect to the first time domain (16) of the first field bus (10) and additionally with respect to the second time domain (26) of the second field bus (20), wherein the gateway (30) supports at least two time domains (32, 34) at the same time.

2. Method according to Claim 1, furthermore having the following steps: - transmitting (60) non-time-sensitive data at a time (54) outside the first time slots (50) from the first field bus (10) to the gateway (30); - buffering (62) the non-time-sensitive data in the gateway (30); and - transmitting (64) the non-time-sensitive data received from the first field bus (10) from the gateway (30) to the second field bus (20) at a time (56) outside the second time slots (52).

3. Method according to either of Claims 1 or 2, wherein the following is provided between the first time domain (16) and the second time domain (26): a. a time offset (58); and / or b. different frequencies (f1, f2); and / or c. different cycle time durations (TC1, TC2); and / or d. different durations (TS1, TS2) of the first and second time slots (50, 52); and / or e. time differences (TD) between first and second time slots (50, 52).

4. Method according to one of Claims 1 to 3, wherein the gateway (30) has a first timer (36) and a second timer (38), and wherein the method furthermore has the following steps: synchronizing the first timer (36) as a slave with the first time domain (16) of the first field bus (10); and synchronizing the second timer (38) as a slave with the second time domain (26) of the second field bus (20).

5. Method according to Claim 4, wherein the method furthermore has the following steps: initially exclusively synchronizing the frequency (f1) of the first timer (36) of the gateway (30) with the frequency (f2) of the second time domain (26) of the second timer (38) of the gateway (30); and setting the first timer (36) of the gateway (30) as a master for the first field bus (10) in order to set the frequency (f1) of the first field bus (10) with the frequency (f2) of the second time domain (26).

6. Method according to Claim 5, wherein the method furthermore has the following steps: determining the cycle time duration (TC1) of the first time domain (16) and the cycle time duration (TC2) of the second time domain (26) by the gateway (30) at a reference time (TRef); and determining a time offset (58) between the first time domain (16) and the second time domain (26) by the gateway (30) at the reference time (TRef); and communicating the cycle time duration (TC1) of the first time domain (16) to the second field bus (20) and communicating the cycle time duration (TC2) of the second time domain (26) to the first field bus (10) and communicating the time offset (58) with respect to the reference time (TRef) to both field buses (10, 20).

7. Method according to Claim 6, wherein the method furthermore has the following steps: defining in the first field bus (10) the first time slot (50) for the communication of time-critical data by the first subscriber device (12, 14) of the first field bus (10); and defining in the second field bus (20) the second time slot (52) for the communication of time-critical data by the second subscriber device (22, 24) of the second field bus (20); wherein the definition is carried out in such a way that the time slots (50, 52) of the first field bus (10) and of the second field bus (20) overlap in time (OL1, OL2).

8. Method according to Claim 7, wherein the method furthermore has the following step: passing time-critical data (70, 72) from the second field bus (20) to the first field bus (10) or vice versa during the time slots (OL1, OL2) which overlap in time.

9. Gateway (30) for networking at least two time-sensitive field buses (10, 20) with different time domains (16, 26), having: a. a first timer (36) which is set up to synchronize itself as a slave with respect to the first time domain (16) of the first field bus (10); and b. a second timer (38) which is set up to synchronize itself as a slave with respect to the second time domain (26) of the second field bus (20).

10. Gateway according to Claim 9, wherein the gateway (30) is set up to transmit non-time-critical data between the first field bus (10) and the second field bus (20) at times outside time slots (50, 52) for time-critical data of the first field bus (10) or second field bus (20), wherein the gateway (30) has a buffer for buffering received non-time-critical data of a field bus (10, 20) before sending to the other field bus (20, 10).

11. Gateway according to either of Claims 9 or 10, wherein the gateway (30) is furthermore set up to pass time-critical data from the second field bus (20) to the first field bus (10) or vice versa during time slots (50, 52) which overlap in time (OL1, OL2).

12. System (1) for networking at least two time-sensitive field buses (10, 20), having: a. a first time-sensitive field bus (10), having a first subscriber device (12, 14) and its own first time domain (16), wherein the first time domain (16) contains first time slots (50) for the transmission of time-sensitive data; b. a second time-sensitive field bus (20), having a second subscriber device (22, 24) and its own second time domain (26), wherein the second time domain (26) contains second time slots (52) for the transmission of time-sensitive data; and c. at least one gateway (30) according to one of Claims 9 to 11, which connects the first and the second field bus (10, 20) to one another for data transmission.

13. System according to Claim 12, wherein the gateway (30) has a buffer (35) for buffering (62) non-time-sensitive data of the first field bus (10) transmitted (60) at a time (54) outside the first time slots (50) from the first field bus (10) and received from the gateway (30), and for transmitting (64) the buffered data from the gateway (30) to the second field bus (20) outside the second time slots (52).

14. System according to either of Claims 12 or 13, wherein the gateway (30) is set up to receive information about the first and second time slots (50, 52) of the first and the second field bus (10, 20) and to take it into account when sending the non-time-critical data.

15. System according to one of Claims 12 to 14, wherein the subscriber device (12, 14) of the first field bus (10) is set up to define the first time slots (50) in such a way, and the subscriber device (22, 24) of the second field bus (20) is set up to define the second time slots (52) in such a way that the first and second time slots (50, 52) overlap in time; and wherein the gateway (30) is furthermore set up to pass time-critical data from the second field bus (20) to the first field bus (10) or vice versa during the overlapping first and second time slots (50, 52) of the first and second field bus (10, 20).