CN111865551A - Device and method for coordinated management of multistage system based on fast bus - Google Patents
Device and method for coordinated management of multistage system based on fast bus Download PDFInfo
- Publication number
- CN111865551A CN111865551A CN202010667374.4A CN202010667374A CN111865551A CN 111865551 A CN111865551 A CN 111865551A CN 202010667374 A CN202010667374 A CN 202010667374A CN 111865551 A CN111865551 A CN 111865551A
- Authority
- CN
- China
- Prior art keywords
- plug
- management
- module
- bus
- input
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L5/00—Arrangements affording multiple use of the transmission path
- H04L5/14—Two-way operation using the same type of signal, i.e. duplex
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5005—Allocation of resources, e.g. of the central processing unit [CPU] to service a request
- G06F9/5027—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals
- G06F9/505—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals considering the load
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L12/00—Data switching networks
- H04L12/28—Data switching networks characterised by path configuration, e.g. LAN [Local Area Networks] or WAN [Wide Area Networks]
- H04L12/40—Bus networks
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L12/00—Data switching networks
- H04L12/28—Data switching networks characterised by path configuration, e.g. LAN [Local Area Networks] or WAN [Wide Area Networks]
- H04L12/40—Bus networks
- H04L12/4013—Management of data rate on the bus
Abstract
The invention discloses a device and a management method thereof based on a fast bus and multistage system coordination management, wherein the device is a multi-plug-in interconnection system constructed by adopting a high-speed Ethernet, and comprises a main processor plug-in, a high-speed communication exchange matrix, a communication bus and a slave processor plug-in; the multi-plug-in interconnection system realizes full duplex and high-speed data interaction between any plugs in a high-speed communication switching matrix through an Ethernet bus; the management CPU of the main processor plug-in is connected with the high-speed communication switching matrix through an Ethernet line. The device and the management method thereof solve the problems of low communication rate and large time delay of the management bus of devices such as centralized digital protection, cluster measurement and control and the like, and the problems of complex design of a management service communication interface and unstable management function, and simultaneously can solve the problems of low management efficiency and overlong system starting time caused by a single centralized management mechanism.
Description
Technical Field
The invention relates to a device for coordination management and a management method thereof, in particular to a device for coordination management of a multi-level system based on a fast bus and a management method thereof, belonging to the technical field of power business.
Background
At present, the configuration of secondary equipment of a transformer substation mainly adopts a layered distributed structure, the space is oriented, the functions are independent, the mode has high reliability, the fault of any device cannot influence the normal operation of other equipment, the expandability and the maintainability of the system are good, but the hardware configuration is repeated, the total-station analysis is complex, the information sharing is insufficient, and the coordination and the function optimization of a total-station system layer are lacked.
With the popularization and application of the IEC61850 standard, the application of network sampling makes the total-station information sharing possible, and promotes the appearance of secondary equipment integration devices such as centralized protection, cluster measurement and control and the like. The centralized protection simultaneously collects a plurality of interval network sampling messages through a digital communication plug-in and realizes the protection function of a plurality of intervals in one device through network tripping; the cluster measurement and control system runs 15-20 functions of virtual measurement and control devices in a single device, each virtual measurement and control system completes the functions of a single-interval measurement and control device such as sampling value data receiving and electric quantity calculating, transformer substation time state quantity receiving facing a general object, sending of a remote control GOOSE command, synchronous switching-on operation, spacer layer logic locking and the like, and simultaneously, software and hardware redundancy standby of a total-station measurement and control function is formed with a conventional measurement and control device. Through the condition of the current centralized protection and the cluster measurement and control test point application, the number of transformer substation equipment can be effectively reduced, the number of screen cabinets, cables and occupied area are reduced, and the construction cost of the transformer substation is saved. Meanwhile, the system layout of secondary protection control and automation functions of the intelligent substation is optimized, the online operation time of equipment is prolonged, and the safety and reliability of a power grid are further improved.
Due to the delay uncertainty brought by network sampling and network tripping, the requirement of direct sampling and direct tripping needs to be met, which brings challenges to the overall software and hardware design of the centralized protection device to a certain extent. The digital protection device mostly adopts a single digital communication plug-in to realize the receiving and sending of SV and GOOSE messages, thereby being beneficial to the complete isolation of protection logic operation and communication interface processing and ensuring that the service logic is protected from the influence of network storm; and the other side can flexibly adapt to different requirements of different projects needing to access the acquisition channel by expanding the number of the digital communication plug-ins. The method comprises the steps that networking mode sampling is replaced by point-to-point optical fiber access through direct sampling and direct jumping, protection of 8 lines is completed simultaneously, under the condition that GOOSE/SV is not networked, each line at least needs 1 SV and GOOSE ports, calculation is carried out through the maximum 8 network ports of a single digital plug-in, and the channel access requirement can be met through 2 digital plug-ins in total. For a cluster measurement and control device, a network acquisition and network hopping mode is mostly adopted to realize the access of a process layer network, under the condition of SV and GOOSE networking, a processor on a single plug-in needs to process 20 maximally spaced high-speed SV messages and GOOSE messages in real time, the computing capacity of a single-core processor becomes incapable of coping with the application scene, and the problem is generally solved by improving the performance of the processor at the moment, for example, a 2-core or even 4-core processor is adopted, a plurality of cores adopt a bare-core independent operation mode, the computing capacity of 20 spaced computing quantities is averagely distributed to the plurality of cores, and the expansion of the computing capacity of a single digital communication plug-in is realized.
Under the design mode that a main CPU plug-in completes protection, control and other core service logic calculation, HMI (human machine interface) display and external device communication and a secondary processor plug-in completes process layer digital communication interface processing, the distributed calculation processing architecture of the main processor and the secondary processor is bound to organically connect the main processor plug-in and the secondary processor plug-in through a specific backboard physical bus and realize mutual communication to form a whole, and then the coordination work of the device can be realized. The backplane bus needs to bear device management services and data services, and generally, in order to ensure the safety isolation of the data services, the management bus and the data bus are independently designed.
The functions that the management bus needs to undertake include: the method comprises the steps of plug-in management of the whole device, management of a system input/output/parameter database, starting initialization process control of the device, establishment of data interaction relation among functional modules, abnormal monitoring of plug-ins in the running process, debugging agent of the device and the like. In order to simplify the system design and function implementation, the management function of the whole system generally adopts a centralized design and is deployed on a main plug-in management CPU, wherein a core service in the management function is to establish a data interaction relationship between each functional module in the system, so that when the device runs in real time, the functional module for implementing service logic can correctly acquire required input data within an accurate delay requirement according to the self requirement of the service, and after logic calculation, control command output is implemented, or input and use of the next-stage functional module are provided. The data interaction relationship between the functional modules in the device system comprises the functional modules in the plug-ins, the functional modules between different plug-ins, the device and the device functional modules, and the like. The data interaction relation is completed by depending on the information of the input and output signal libraries and the incidence relation between the input and output signals of the system, i.e. the main plug-in management CPU must complete the collection of the input and output signals on each plug-in of the whole device and the collection of the incidence relation between the two.
Specifically, when software is designed, a functional module on a slave plug-in the system actively sends input and output signals and input and output association relation of the module to a management CPU on a master plug-in through a management bus, and the management CPU is responsible for building a library and indexing of all input and output signals in the system. And automatically establishing input and output links according to the incidence relation of the input and output links, arranging and generating an output data set table and an input data set table of each plug-in, and issuing the output data set table and the input data set table to each slave plug-in. In the real-time operation stage, each slave plug-in completes the arrangement and sending of the data of the signal required to be output by the functional module on the plug-in according to the output data set table of the plug-in, and meanwhile, receives the data of the signal required to be input by the functional module on the plug-in and analyzes the data to provide the functional module for use according to the input data set table. In conclusion, the purpose of the system function of the whole device is realized through the data interaction and logic cooperation of the functional modules.
Generally, according to the characteristics of management service and management data and the scale of a service system, no clear requirements are made on performance indexes such as bandwidth and speed of a management bus, and a management bus link generally adopts a low-speed and low-bandwidth physical link design relative to a high-bandwidth and high-speed data bus, but with the increasing integration of devices and the gradual expansion of the scale of the system, for an application scenario that a plurality of slave processing plug-ins are required to be added for centralized protection or digital bus differential protection to expand an access channel, the scale of input and output signal libraries of the system may reach tens of thousands, even 10 tens of thousands of levels. In the process of starting the equipment, the low-speed and low-bandwidth management bus inevitably becomes the performance bottleneck of the device function, so that the starting time of the device is as long as several minutes, the effective running time of the equipment is reduced, and a 'blank period' is brought to the protection and control functions of the primary equipment. Meanwhile, the incoordination between the slow communication interface and the fast processor operation may also cause the loss of messages, which brings hidden troubles to the operation of protection and control equipment.
On the other hand, for the integrated cluster measurement and control or multifunctional protection and measurement device, a multi-core processor is required to be added to improve the application scene of the computing capability of the slave plug-in, the input and output association relationship between the multiple cores on the slave plug-in and the functional modules in the cores, and the arrangement services corresponding to the output data set table and the input data set table of the plug-in are processed by being concentrated on the master plug-in management CPU, so that the communication burden of the management bus and the operation burden of the management CPU are further increased, the concentrated management service mechanism brings serious imbalance of the system load, the efficiency of the device management service is low, the system short board effect is brought, and the integrated cluster measurement and control or multifunctional protection and measurement device cannot adapt to the service requirements of the development trend of a new generation of intelligent substation on.
In 201611198728.5 patent, a multi-board communication system and method using RS485 protocol, the connection between a master control board and multiple slave control boards is established based on RS485 physical links, and since the RS485 bus is half-duplex, when multiple cards initiate communication simultaneously on the bus, collision is inevitably caused. In this case, a flow control or arbitration method needs to be added to the software. In the scheme, a communication switching matrix is added between the master control board and the slave control board, and when the master control board or the slave control board initiates communication, communication path selection is carried out through arbitration.
The thesis "design of station area protection control device based on backplane bus" uses M-LVDS to establish communication bus connections between a monitoring module, a communication module and a plurality of protection modules. The data exchange among the board cards is realized based on a multi-master mode time-sharing multiplexing bus, the design format of a bus link message comprises an arbitration segment and a token segment, and the bus resources are guaranteed to be used by each node on the bus in a time-sharing mode through a software communication protocol.
In the thesis "design of internal communication mechanism of relay protection device based on CAN bus", CAN bus is adopted to establish communication connection between CPU module and IO module, the CAN bus is set to 1Mhz baud rate in the scheme, data transmission of at most 8 bytes per frame is realized by extending frame of CAN2.0b, meanwhile, bus competition problem is solved by setting priority Pri in frame ID, and when the transmission data exceeds 8 bytes, different frames are directly distinguished by frame number. In order to solve the problem of possible message loss, a message timing retransmission mechanism is added in the application layer software design.
The main problems of the scheme of '201611198728.5 a multi-board communication system and method using RS485 protocol' are that the RS485 bus is limited by the physical layer protocol, the speed is relatively low, the highest rate is only about 10Mbps, and data exchange between slave boards must pass through the master control board, which not only reduces the communication rate, but also increases the data processing load of the master control board. And the bus transmission mechanism is undefined, cannot meet the active real-time data transmission requirement of the slave plate, and has low interaction efficiency. The problem of communication collision of a plurality of plug-in units on the bus is solved by using a special arbitration module, extra software and hardware design is added, the device cost is improved, and the device cannot adapt to an application scene with high bus bandwidth.
Although the communication bandwidth is improved in the scheme of the thesis 'design of the station domain protection control device based on the backplane bus' compared with the RS485 bus, a bus arbitration mechanism is lacked, and communication must be performed through a main control board card. Still using half-duplex communication mode, arbitration and token words need to be added to the software communication protocol to solve the bus time-division multiplexing problem. If a full duplex scheme is adopted, more physical channels are occupied. When the number of the plug-ins on the bus is large, the communication delay is increased due to the mode of occupying the bus in a time-sharing mode. The bus transmission mechanism is undefined and cannot meet the requirement of active real-time data transmission of the slave board.
The scheme of the thesis 'design of internal communication mechanism of relay protection device based on CAN bus' has certain adaptability to the system with small communication data volume of IO module type modules and balanced flow load on the whole system management bus. However, in the case of a large number of input and output signals of the digital plug-in the relay protection control equipment, and a plurality of digital plug-ins are arranged in the device, the bandwidth of 1Mhz at most still becomes a bottleneck of communication transmission. Meanwhile, the data lengths of different services on the management bus are different, a CAN frame transmits 8 bytes of data at most, and a software driver needs to increase frame splitting during sending and frame grouping during receiving, so that great limitation is brought to the design of application software.
Disclosure of Invention
The purpose of the invention is as follows: the first purpose of the present invention is to provide a device based on fast bus and multi-level system coordination management, which solves the problems of complex design of management service communication interfaces and unstable management functions caused by low communication rate and large time delay of device management buses such as centralized digital protection, cluster measurement and control, etc., and can solve the problems of low management efficiency and long system startup time caused by a single centralized management mechanism.
The technical scheme is as follows: the invention relates to a device for coordinated management of a multi-level system based on a fast bus, which comprises:
the device is a multi-plug-in interconnection system constructed by adopting high-speed Ethernet, and comprises a main processor plug-in, a high-speed communication switching matrix, M communication buses and a plurality of slave processor plugs-in;
the multi-plug-in interconnection system realizes full duplex and high-speed data interaction between any plugs in a high-speed communication switching matrix through an Ethernet bus;
the management CPU of the main processor plug-in is connected with the high-speed communication switching matrix through an Ethernet circuit;
the M slave processor plug-ins are connected with the high-speed communication switching matrix through M Ethernet lines and backboard terminals;
The M +1 Ethernet lines comprise transceiving links to realize full-duplex communication;
the high-speed communication exchange matrix is positioned in the master processor plug-in or the slave processor plug-in, and the communication exchange function is realized through programmable logic hardware;
the communication bus adopts an Ethernet protocol;
the communication matrix can support Ethernet communication under different rates through configuration;
the high-speed communication switching matrix can support standard Ethernet interaction modes such as message unicast, multicast and broadcast through configuration.
The high-speed communication switching matrix interface can realize data exchange among any ports, and can carry out priority data exchange according to priority through priority definition.
The data exchange between the interface ports of the high-speed communication exchange matrix can provide accurate data exchange delay.
The management service of the whole system adopts a multi-level management mechanism, and the management function of the whole system is completed by the coordination of the master processor plug-in and each slave processor plug-in.
A management CPU on the main processor plug-in deploys a device-level main management module which is responsible for device-level management services;
each slave processor plug-in deploys a sub-management module which is responsible for management services at the plug-in level.
The slave processor plug-in may be a single core processor or a multi-core processor. When the core is a multi-core processor, a plurality of cores adopt a bare-core independent operation mode, and the sub-management module is deployed on the core 1.
The management service of the whole device mainly comprises the establishment of data interaction relation among all functional modules in the system.
Each functional module relationship includes 4 types: (1) functional modules in single-core processor plug-ins, (2) functional modules in single-core processor plug-ins, (3) functional modules of different cores of multi-core processor plug-ins, and (4) functional modules of different plug-ins.
And the sub-management module automatically completes the establishment of the data interaction relation of the types (1), (2) and (3).
And the main management module automatically completes the establishment of the data interaction relation of the type (4).
Establishing a data interaction relationship means that an input signal of a certain functional module in a system is linked to an output signal of another functional module during initialization, so that the data linkage relationship between the two functional modules is realized. When the device runs, the system transmits the output data of the latter functional module to the input of the former functional module in real time according to the link relation.
The invention also provides a method for multi-level system coordination management based on the fast bus, which comprises the following steps:
the high-speed communications switching matrix is implemented by programmable logic.
The format of the data interaction message between the boards is self-defined, and the message comprises information exchanged by a source port domain, a target port domain, a priority control domain, a sending time domain and a receiving time domain.
The high-speed communication switching matrix module internally comprises m paths of switching modules for parallel processing. The exchange module comprises a configuration bus interface, a receiving engine, a receiving cache and control thereof, a receiving exchange bus interface, a sending engine, a sending arbitration module and a sending exchange bus interface.
The configuration bus interface realizes the functions of Ethernet communication rate of the module, on-line configuration of the communication mode and the like.
And the receiving engine finishes the receiving of the external exchange Ethernet message, records the receiving timestamp record and stores the receiving timestamp record in a receiving cache.
And after the receiving exchange bus interface judges that the message to be exchanged exists, the sending request is sent to the target port according to the port description.
And after receiving the sending confirmation command, the receiving exchange bus interface transmits the message to be exchanged and repeats the process until the message to be exchanged is completely sent.
And after the sending exchange bus interface judges that the message to be exchanged exists, carrying out bus arbitration. After the arbitration is finished, a sending confirmation command is replied to the source port to be sent which is allowed by the arbitration, and the message to be exchanged is started to be sent in a butt joint mode. The process is repeated until all the messages to be exchanged are completely sent.
After the sending engine is started, the message to be exchanged is converted into an Ethernet message, and a sending timestamp and verification information are added to the tail of the message.
The bus arbitration module arbitrates the transmission sequence according to the priority, and if the priority is the same, the bus arbitration module decides the transmission sequence of the messages to be exchanged according to the serial number sequence of the source ports.
The multi-stage system coordination management method is characterized in that a main management module on a main plug-in and a sub management module on each slave plug-in are matched to complete the management service of the system.
And the master management module is responsible for starting flow control in the device and enabling the initialization flow of the slave plug-ins according to the serial number of the slave plug-ins.
And each functional module on the slave plug-in sends the input and output information of the module and the input and output associated information between the functional modules to the master management module in the initialization process.
And the sub-management module on the slave plug-in is responsible for the sending processing of the registration information. And the input and output information of the functional module is sent to the main management module, and an input and output signal library of the plug-in module is established locally. And screening the input and output associated information between the internal function modules of the plug-in, establishing an associated information table locally, and only sending the management information table between the plug-ins to the main management module.
The main management module receives the registration information sent by the plug-in sub management module, and establishes an input information base and an output information base of the whole system for the input information and the output information; and establishing an association information table between the plug-ins of the whole system for the input and output association information tables.
The sub-management module obtains the address information of the signals to be correlated by searching the local input and output signal library according to the local input and output correlation information table, and completes the link operation of input and output locally.
And the main management module obtains the address information of the signals needing to be correlated by searching an information base of the whole system according to the input and output correlation information table among the plug-ins of the whole system, and arranges the address information to obtain an input and output link relation mapping table.
And the master management module generates an output link mapping table and an input link mapping table of each plug-in according to the dimension of the plug-in of the link relation mapping table, and transmits the link mapping tables to the corresponding slave management modules on the plug-ins.
And the sub-management module receives the input and output link mapping table issued by the main management module to complete the link operation.
Has the advantages that: compared with the prior art, the invention has the following remarkable advantages:
(1) physical communication of a plurality of processors is established by adopting a high-speed Ethernet bus, full duplex communication is realized, a clock channel is not needed, and signal links of a back plate are reduced; the main control board only needs to design one standard Ethernet channel, thereby simplifying the design of system hardware;
(2) the highest speed of the Ethernet bus can reach Gps, the data communication capacity of high bandwidth solves the transmission bottleneck problem of massive management data in a multi-plug-in and large-scale system, greatly reduces the transmission delay of the data, and simultaneously makes the management bus expand and apply related services possible;
(3) The communication exchange matrix is realized by independent programmable logic, the communication exchange processing of bus data is completed by independent logic devices, is not limited by the number of channels of a physical exchange chip, and can freely realize the expansion of the number of slave plug-ins;
(4) the communication exchange matrix supports message transmission with priority and can ensure the non-delay exchange of core data; the method supports two transmission modes of unicast and multicast, and the broadcast mode can be that data is transmitted between a plurality of plug-ins at the same time, thereby greatly improving the overall efficiency of data transmission in the system;
(5) the maximum link transmission message length of the Ethernet is 1518 bytes, meet the flexible variable length requirement of the management service message, has simplified the design of the message receiving and sending communication driving interface, has improved the reliability of the software system;
(6) the master-slave multi-stage cooperation finishes the mechanism of system management, and the interactive information between the plug-in module is processed on the spot, so that the information of all the plug-ins in the device is prevented from being collected in a unified way in the master plug-in, the burst intensive transmission of data on a bus is reduced, the starting time of the device is shortened, and the on-line running time of equipment is improved;
(7) the sub-management module completes the calculation of the related management service on the plug-in unit on site, and the main management module completes the calculation of the related management service at the device level, so that the calculation load balance of the device management service is realized, the problem that the main processor receives and loses the message due to the imbalance of the receiving and sending processing loads of the main processor and the slave processor is avoided, and the reliability of the device management service is improved.
Drawings
FIG. 1 is a schematic diagram of a high speed switching matrix architecture according to the present invention;
FIG. 2 is a schematic diagram of the internal module components of the high-speed communication switching matrix according to the present invention;
FIG. 3 is a schematic diagram of a transceiving process of a high speed communication switch matrix according to the present invention;
FIG. 4 is a schematic diagram of a device management service multi-level system coordination management mechanism according to the present invention;
FIG. 5 is a flow diagram of a sub-management module implementing in-place management logic in accordance with the present invention;
FIG. 6 is a logic diagram of the main (sub) management module implementing the data link function according to the present invention.
Detailed Description
The technical solution of the present invention is further illustrated by the following examples.
As shown in fig. 1-2, the present invention provides a device for coordinated management of a multi-level system based on a fast bus:
the bus system of the device comprises a main control board, a high-speed communication exchange module positioned on the main control board and 3 slave boards.
The master control board communicates with the high-speed communication exchange module through 1 path of gigabit Ethernet, and the 3 slave boards communicate with the high-speed communication exchange module through 3 paths of gigabit Ethernet.
The high-speed communication exchange module has 5-path gigabit Ethernet as external port and 1 path reserved as expansion interface.
The data exchange communication channel occupies 10 paths of differential channels of the backboard, and full-duplex communication is realized.
The highest data communication exchange rate bit among any board cards is 1 Gbps.
The high-speed communication switching module is realized by FPGA.
The data communication channel of the master plate and the slave plate can realize rate configuration and channel expansion enabling through configuration.
The management service of the device adopts a multi-level management mechanism, and a master processor plug-in and a plurality of slave processor plugs coordinate to complete the management function of the whole system.
The main processor plug-in manages the CPU and disposes the main management module of the device level, is responsible for the management business of the whole apparatus, mainly include: (1) an input/output signal library (2) for all the cards in the management apparatus manages the signal link relationship between the functional modules in the cards in the management apparatus.
The system comprises a plurality of slave processor plugins, wherein each plugin is provided with a plugin level sub-management module which is responsible for management services in the plugins and mainly comprises: (1) managing input and output signal libraries of all functional modules in the plug-in, and (2) managing signal link relations among the functional modules in the plug-in.
The functional modules of the device need to establish an input/output signal link relationship between the two through the device management module, so that real-time data interaction between the functional modules is realized, and the service logic of the whole device is realized through the cooperative work of a plurality of functional modules in the device.
The number of the slave processor plug-ins can be expanded according to the requirement of the system for accessing the service channel, and the master management module realizes the communication with the slave management modules on the slave plug-ins through the high-speed communication switching module.
The CPU architecture on the slave plug-in can be expanded according to the requirement of the system on the business computing capacity, and the slave CPU can be a single core, double cores or multi-core.
When the CPU is multi-core from the plug-in, the sub-management module is deployed on the core 1, and the inside establishes communication between the core 1 and other cores through a high-speed shared memory channel.
As shown in fig. 4, the input/output signal link relationship between functional modules in the device is divided into 4 types according to the plug-in address relationship between input signal and output signal: (1) the method comprises the following steps of (1) linking relations among single-core plug-in function modules, (2) linking relations among function modules in a multi-core plug-in core, (3) linking relations among function modules among multi-core plug-in cores, (4) linking relations among function modules among plug-ins.
And the sub-management module on each plug-in completes the management of the data association relations of the types (1), (2) and (3), obtains a link relation mapping table of input and output signals in the plug-in through local arrangement, and establishes a link relation between the input and output signals.
And (4) finishing the management of the data association relation of the type (4), arranging and obtaining a link relation mapping table of input and output signals between the device plug-in modules on the main plug-in module, and establishing a link relation between the device plug-in modules and the main plug-in module.
The invention also provides an embodiment of a multi-level system coordination management method based on the fast bus, which comprises the following steps:
the high-speed communication exchange module comprises the following components:
the module is internally provided with 5 independent switching modules, and comprises a configuration bus interface, a receiving engine, a receiving cache and control thereof, a receiving switching bus interface, a sending engine, a sending arbitration module and a sending switching bus interface.
The receiving engine mainly completes the functions of receiving and checking the external exchange Ethernet message, recording a receiving timestamp according to a system clock, receiving and caching the message and the like. The number of the message receiving caches is 32, and the size of the message receiving caches is 2 KB.
As shown in fig. 3, the module implements the high-speed switching steps as follows:
and after judging that the message to be exchanged exists by the receiving exchange bus interface, sending a sending request to the target port according to the port description, and if receiving a sending confirmation command, transmitting the message to be exchanged. The process is repeated until all the messages to be exchanged are sent.
And after the sending exchange bus interface judges that the message to be exchanged exists, carrying out bus arbitration.
The bus arbitration module arbitrates the sending sequence according to the priority, and if the priority is the same, the bus arbitration module decides the transmission sequence of the messages to be exchanged according to the sequence of the source ports. The message priority is divided into 4 levels, the 4 level is the highest priority, and the port priority is 0, 1, 2 and 3.
After the arbitration is finished, a sending confirmation command is replied to the source port to be sent which is allowed by the arbitration, and the message to be exchanged is sent. The process is repeated until all the messages to be exchanged are completely sent.
After the sending engine is started, the message to be exchanged is converted into an Ethernet message, and a sending timestamp and CRC32 check information are added to the tail of the message.
The multi-stage system coordination management method divides the input and output signal incidence relation of the function modules in the whole system into two types of plug-in and plug-in, the signal incidence relation in the plug-in is processed locally by the sub management module on the slave plug-in, and the signal management relation between the plug-ins is processed by the main management module on the main processor plug-in.
The sub-management module is deployed on each slave plug-in, and optionally, if the slave plug-in is a multi-core CPU, the sub-management module is deployed on the core 1.
The sub-management module is responsible for the transceiving function of the management bus communication interface of the plug-in and the main management module, and the registration information of all the functional modules on the plug-in is forwarded to the main management module by the sub-management module; meanwhile, the management command issued by the main management module is forwarded to the functional module by the sub-management module.
As shown in fig. 5, when receiving the registration input and output information of the functional module, the sub-management module forwards the registration input and output information to the main management module, and then inserts the local signal input and output library;
when the sub-management module receives the signal associated information of the functional module, the sub-management module performs multi-stage processing according to the type of the associated information, and the specific logic is as follows:
the input and output signal associated information registration interfaces are divided into two categories: the input of the module is related to the output of other modules, and the output of the module is related to the input of other modules. For the former interface, the sub-management module searches in a local output signal library through the name of the associated output signal, if the name can be found, the input and output associated information belongs to the plug-in, and is inserted into a local information association table; otherwise, the information is sent to the main management module.
For the latter interface, the sub-management module searches in a local input signal library through the name of the associated input signal, if the name is found, the input and output associated information belongs to the plug-in and is inserted into a local information associated table; otherwise, the information is sent to the main management module.
By the method, the input and output signal library and the input and output signal association table in the plug-in are established in the sub management module. And establishing an input and output signal association table between the input and output signal libraries and the plug-in the whole system at the main management module. The managed data in the system is divided into two levels of plug-in level and system level according to a multi-level management strategy, and is distributed on each slave plug-in and the master plug-in.
Furthermore, the main management module and the sub management module respectively and independently complete the management service function according to the local managed data, namely, the actual link operation between the function modules in the system is established according to the input and output signal association table, and the main management module and the sub management module adopt the same logic, so the management modules are collectively called below.
Specifically, as shown in fig. 6, the management module establishes an input/output link relationship according to the following steps:
step 1: the content of the registered input and output signal messages of the functional module comprises the data attributes of input and output variables, such as memory addresses, data types, plug-in addresses and the like, and also comprises the alias of the signal. Optionally, if the slave plug-in is a multi-core CPU, the slave plug-in further includes a core serial number.
Step 2: and the content of the register input and output association information message of the functional module comprises an input variable alias and an output variable alias.
And step 3: the management module takes out 1 piece of record information from the input and output association information table, and searches in the input and output signal library through the alias of the input and output signal respectively to obtain the data attributes of the input variable and the output variable, such as the plug-in address (core serial number), the memory address, the data type and the like.
And 4, establishing an input-output mapping relation table of the functional modules in the system by taking the output plug-in address (core serial number) and the input plug-in address (core serial number) as indexes, and obtaining a data interaction relation table between the system plug-ins (cores) by taking the output of one plug-in (core) to the input of the other plug-in (core) as dimension sorting. For example: the output of all the functional modules on the plug-in (core) 1 to the input of all the functional modules on the plug-in (core) 2 is 1 sheet of table, and the output of all the functional modules on the plug-in (core) 1 to the input of all the functional modules on the plug-in (core) 3 is another 1 sheet of table.
And 5, respectively sorting each mapping table by taking the output angle and the input angle as dimensions to obtain a plug-in (core) output mapping table and a plug-in (core) input mapping table. For example: the insert (core) 1 includes: an output mapping table to plug-in (core) 2 and an output mapping table to plug-in (core) 3.
And 6, the management module issues the output mapping table of the plug-in (core) to the data interaction module actually completed on the plug-in (core), and through the steps, the module establishes the input and output link relation among all the functional modules in the system.
When the device runs, each slave plug-in (core) data interaction module sends data according to the output mapping table, and the corresponding plug-in (core) data interaction module receives data according to the input mapping table, so that real-time data interaction between any functional modules in the system is realized.
Claims (11)
1. A device based on fast bus, multistage system coordination management which characterized in that: the device is a multi-plug-in interconnection system constructed by adopting high-speed Ethernet, and comprises a main processor plug-in, a high-speed communication exchange matrix, a communication bus and a slave processor plug-in;
the multi-plug-in interconnection system realizes full duplex and high-speed data interaction between any plugs in a high-speed communication switching matrix through an Ethernet bus;
the management CPU of the main processor plug-in is connected with the high-speed communication switching matrix through an Ethernet circuit;
the M slave processor plug-ins are connected with the high-speed communication switching matrix through M Ethernet lines and backboard terminals;
the M +1 Ethernet lines comprise transceiving links to realize full-duplex communication;
the high-speed communication exchange matrix is positioned in the master processor plug-in or the slave processor plug-in, and realizes the communication exchange function through programmable logic hardware;
The communication bus adopts an Ethernet protocol.
2. The apparatus for fast bus based, multi-level system coordination management according to claim 1, wherein: the high-speed communication exchange communication matrix supports Ethernet communication at different rates; the high-speed communication switching matrix supports message unicast, multicast and broadcast, the high-speed communication switching matrix is realized through programmable logic, the format of the data interaction messages among the board cards is self-defined, and the messages comprise source port domain, target port domain, priority control domain, sending time domain and receiving time domain exchange information.
3. The apparatus for fast bus based, multi-level system coordination management according to claim 1, wherein: the high-speed communication switching matrix interface realizes data exchange between any ports, and performs priority data exchange according to priority through priority definition; data exchange among the interface ports of the high-speed communication exchange matrix provides data exchange delay;
the high-speed communication switching matrix module internally comprises m paths of switching modules for parallel processing, and the switching module internally comprises a configuration bus interface, a receiving engine, a receiving cache and control thereof, a receiving switching bus interface, a sending engine, a sending arbitration module and a sending switching bus interface;
The configuration bus interface realizes the on-line configuration function of the Ethernet communication rate and the communication mode of the module.
4. The apparatus for fast bus based, multi-level system coordination management according to claim 1, wherein: the management service in the multi-plug-in interconnection system adopts a multi-level management mechanism, and a master processor plug-in and each slave processor plug-in coordinate to complete the management function of the multi-plug-in interconnection system.
5. The apparatus for fast bus based, multi-level system coordination management according to claim 1, wherein: the main processor plug-in unit manages a CPU and deploys a device-level main management module, and is responsible for device-level management services; and each slave processor plug-in deploys a sub-management module which is responsible for management services at the plug-in level.
6. The apparatus for fast bus based, multi-level system coordination management according to claim 1, wherein: the slave processor plug-in is a single-core processor or a multi-core processor, each core of the multi-core processor adopts a bare core independent operation mode, and the sub-management module is deployed on the core 1.
7. The apparatus for fast bus based, multi-level system coordination management according to claim 1, wherein: the management service of the device comprises the establishment of data interaction relations among all functional modules in the system, wherein the data interaction relations among all the functional modules comprise functional modules in a single-core processor plug-in, functional modules in a single core of a multi-core processor plug-in, functional modules in different cores of the multi-core processor plug-in and functional modules in different plug-ins.
8. The apparatus for fast bus based, multi-level system coordination management according to claim 7, wherein: the sub-management module automatically completes the establishment of data interaction relations among the functional modules in the single-core processor plug-in, among the functional modules in the single core of the multi-core processor plug-in and among the functional modules of different cores of the multi-core processor plug-in; the establishment of the data interaction relation among the different plug-in functional modules is automatically completed by the main management module.
9. The apparatus for fast bus based, multi-level system coordination management according to claim 7, wherein: the data interaction relation is established by linking an input signal of a certain functional module in the system to an output signal of another functional module during initialization, so that the data link relation of the two functional modules is realized; when the device runs, the system transmits the output data of the latter functional module to the input of the former functional module in real time according to the link relation.
10. A method for coordinating and managing a multi-level system based on a fast bus is characterized in that high-speed data interaction between any plug-in units is realized through an Ethernet bus, and comprises the following steps:
(1) if the message to be received exists, receiving the Ethernet message to be exchanged according to the standard protocol, carrying out necessary message verification, removing control information, and entering the step (2), otherwise, waiting;
(2) After attaching receiving time stamp and check information at the tail of the message to be exchanged, framing again;
(3) the recombined messages are sorted according to priority and time and then are placed into a receiving cache;
(4) sending an exchange request and waiting for an arbitration result;
(5) performing request arbitration according to the target port state of the exchange message and the request priority of each exchange message;
(6) after the transmission permission is waited, the message to be exchanged is filled to the transmission cache of the target port through the high-speed switching matrix route;
(7) sending a message to be exchanged, adding control information, and additionally sending a timestamp and necessary check information at the tail of the message, wherein the specific steps are as follows:
(a) the receiving engine completes the receiving of the external exchange Ethernet message, records the receiving timestamp record and stores the receiving timestamp record into the receiving cache;
(b) after receiving the switching bus interface and judging that the message to be switched exists, the switching bus interface makes a sending request to a target port according to the port description;
(c) after receiving the sending confirmation command, the receiving exchange bus interface transmits the message to be exchanged and repeats the process until the message to be exchanged is completely sent;
(d) after the sending exchange bus interface judges that the message to be exchanged exists, carrying out bus arbitration, after the arbitration is finished, replying a sending confirmation command to the source port to be sent, which is allowed by the arbitration, starting to butt joint and send the message to be exchanged, and repeating the step until all the messages to be exchanged are completely sent;
(e) After the sending engine is started, the message to be exchanged is converted into an Ethernet message, and sending time stamp and verification information are added to the tail of the message;
(f) the bus arbitration module arbitrates the transmission sequence according to the priority, and if the priority is the same, the bus arbitration module decides the transmission sequence of the messages to be exchanged according to the serial number sequence of the source ports.
11. The method for multi-level system coordination management based on fast bus according to claim 10, further comprising a multi-level management mechanism adopted by the device management service, wherein the management function of the whole system is completed by the coordination of the master management module on the master processor and the sub-management modules on each slave processor plug-in, comprising the following steps:
(1) the master management module is responsible for controlling the starting processes of all the slave plug-ins in the device and enabling the initialization processes of the slave plug-ins according to serial numbers of the slave plug-ins;
(2) each functional module on the slave plug-in sends the input and output information of the module and the input and output associated information between the functional modules to the master management module in the initialization process;
(3) the slave plug-in sub-management module takes charge of the actual sending processing of the registration information, sends the input and output information of the functional module to the master management module, and establishes an input and output signal library of the plug-in module locally; for the input and output associated information, screening the input and output associated information between the internal function modules of the plug-in, establishing an associated information table locally, and only sending the management information table between the plug-ins to the main management module;
(4) The sub-management module obtains the address information of the signals to be correlated by searching a local input signal library and a local output signal library according to the local input and output correlation information table, and completes the link operation of input and output locally;
(5) the main management module receives the registration information sent by the plug-in sub management module, and establishes an input information base and an output information base of the whole system for the input information and the output information; for the input and output associated information tables, establishing an associated information table between the plug-ins of the whole system;
(6) the main management module obtains address information of signals needing to be correlated by searching an information base of the whole system according to an input and output correlation information table among the plug-ins of the whole system, and arranges the address information to obtain an input and output link relation mapping table;
(7) the master management module generates an output link mapping table and an input link mapping table of each plug-in according to the dimension of the plug-in of the link relation mapping table, and transmits the link mapping tables to the corresponding slave management modules on the plug-ins;
(8) and the sub-management module receives the input and output link mapping table issued by the main management module to complete the link operation.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010667374.4A CN111865551B (en) | 2020-07-13 | 2020-07-13 | Device and method for coordinated management of multistage system based on fast bus |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010667374.4A CN111865551B (en) | 2020-07-13 | 2020-07-13 | Device and method for coordinated management of multistage system based on fast bus |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111865551A true CN111865551A (en) | 2020-10-30 |
| CN111865551B CN111865551B (en) | 2022-11-11 |
Family
ID=72983245
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010667374.4A Active CN111865551B (en) | 2020-07-13 | 2020-07-13 | Device and method for coordinated management of multistage system based on fast bus |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111865551B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114745466A (en) * | 2022-04-01 | 2022-07-12 | 惠州市德赛西威智能交通技术研究院有限公司 | Linux system based mobile phone interconnection plug-in design system and method |
| CN114896187A (en) * | 2021-12-24 | 2022-08-12 | 上海御渡半导体科技有限公司 | ATE tester frame type equipment and method for uniformly collecting serial ports thereof |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20040085910A1 (en) * | 2002-11-01 | 2004-05-06 | Zarlink Semiconductor V.N. Inc. | Media access control device for high efficiency ethernet backplane |
| US20100322265A1 (en) * | 2009-06-22 | 2010-12-23 | Sandhya Gopinath | Systems and methods for receive and transmission queue processing in a multi-core architecture |
| CN102780789A (en) * | 2011-05-11 | 2012-11-14 | 孙然 | Design scheme for high-bandwidth network processor data bus interface module |
| CN103488605A (en) * | 2013-09-24 | 2014-01-01 | 许继集团有限公司 | Bus architecture for multiprocessor parallel communication |
| CN103500150A (en) * | 2013-09-24 | 2014-01-08 | 许继集团有限公司 | Bus framework for multiple processors to process applications concurrently |
-
2020
- 2020-07-13 CN CN202010667374.4A patent/CN111865551B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20040085910A1 (en) * | 2002-11-01 | 2004-05-06 | Zarlink Semiconductor V.N. Inc. | Media access control device for high efficiency ethernet backplane |
| US20100322265A1 (en) * | 2009-06-22 | 2010-12-23 | Sandhya Gopinath | Systems and methods for receive and transmission queue processing in a multi-core architecture |
| CN102780789A (en) * | 2011-05-11 | 2012-11-14 | 孙然 | Design scheme for high-bandwidth network processor data bus interface module |
| CN103488605A (en) * | 2013-09-24 | 2014-01-01 | 许继集团有限公司 | Bus architecture for multiprocessor parallel communication |
| CN103500150A (en) * | 2013-09-24 | 2014-01-08 | 许继集团有限公司 | Bus framework for multiple processors to process applications concurrently |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114896187A (en) * | 2021-12-24 | 2022-08-12 | 上海御渡半导体科技有限公司 | ATE tester frame type equipment and method for uniformly collecting serial ports thereof |
| CN114745466A (en) * | 2022-04-01 | 2022-07-12 | 惠州市德赛西威智能交通技术研究院有限公司 | Linux system based mobile phone interconnection plug-in design system and method |
| CN114745466B (en) * | 2022-04-01 | 2024-03-08 | 惠州市德赛西威智能交通技术研究院有限公司 | Mobile phone interconnection plug-in design system and method based on Linux system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111865551B (en) | 2022-11-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111865551B (en) | Device and method for coordinated management of multistage system based on fast bus | |
| CN104618207B (en) | The FC-AE-1553 network systems and exchange method of a kind of heterogeneous | |
| CN106776436A (en) | A kind of high-speed serial bus structure and its communication means suitable for multiple spot interconnection | |
| CN106549847A (en) | A kind of novel integrated modularity avionics system framework | |
| CN202077061U (en) | Multi-port and multi-network protocol converter | |
| CN101394301A (en) | Micro general hardware platform system, apparatus for telecommunication and computer, and communication method thereof | |
| CN106873541A (en) | A kind of EtherCAT Distributed Servo kinetic control systems based on DSP | |
| CN115022109A (en) | Data management method for intelligent edge all-in-one machine and station cloud system | |
| CN210693970U (en) | Multi-bus testing device | |
| CN101917318A (en) | High-low speed bus systems and connection device thereof | |
| CN109413094A (en) | Aerospace data Transmission system based on HS-ADIX network | |
| CN1964286B (en) | A master control device with double CPU | |
| US7454513B2 (en) | Interface link layer device for long delay connections | |
| CN206575442U (en) | A kind of novel integrated modularization avionics system | |
| CN202014269U (en) | Remote data acquisition and transmission unit of aviation control equipment | |
| CN102081383B (en) | Device and method for secure network protocol for field control | |
| CN110069430B (en) | Data acquisition system of chrysanthemum chain structure and self-adaptive transmission method | |
| CN110676818B (en) | In-situ bus protection method and system based on star network topology | |
| CN114466040A (en) | Spacecraft integrated biplane network system | |
| CN116955239B (en) | Open modularized instrument framework applied to measuring and automation fields | |
| WO2014008869A1 (en) | Access device and system, and data sending method | |
| CN2884710Y (en) | Exchanging distance net bridge appliance with virtual net function for providing multiple interfaces | |
| CN114567899B (en) | Terminal detection device and signal processing method based on VPX architecture | |
| CN219247867U (en) | Centralized management circuit of optical fiber transceiver based on singlechip | |
| CN220273788U (en) | AFDX bus exchange equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |