End-to-end data processing methods and systems applied to nuclear power plant DCS systems

Abstract

This application discloses a full-link data processing method and system applied to a nuclear power plant DCS system, relating to the field of nuclear power technology. The method includes: a subscription-publishing core layer subscribing to multi-source data published by a data publisher; a protocol adaptation layer receiving and parsing the multi-source data to generate parsing results, generating priority tags based on the multi-source data types of the parsing results, and generating standard data packets based on the parsing results and priority tags; the subscription-publishing core layer allocating transmission bandwidth based on the priority tags contained in the standard data packets and publishing the standard data packets to the corresponding topics; a secure transport layer subscribing to the standard data packets, encrypting the standard data packets using an end-to-end encryption algorithm, and adding a message authentication code to the encrypted standard data packets to generate encrypted data packets; and a persistence layer receiving the encrypted data packets and performing hierarchical storage based on the priority tags and data types of the encrypted data packets. This application can reduce latency and improve data parsing efficiency in high-concurrency scenarios.

Description

Technical Field

[0001] This application relates to the field of nuclear power technology, and in particular to a full-link data processing method and system for use in nuclear power plant DCS systems. Background Technology

[0002] The Distributed Control System (DCS) is the core hub of nuclear power plant operation, responsible for the real-time acquisition, processing, and transmission of tens of thousands of data points. With the development of intelligent nuclear energy, DCS systems are placing higher demands on data processing, such as: supporting the real-time transmission of hundreds of thousands of data points per second, meeting the high-concurrency, low-latency requirements of millisecond-level response to critical control commands; and being compatible with multiple protocols such as Modbus, OPCUA, and IEC61850, integrating heterogeneous data sources such as sensors, logs, and video streams.

[0003] Currently, nuclear power DCS systems mainly use traditional message middleware such as Kafka and RabbitMQ or customized communication protocols. In high-concurrency scenarios, these systems are prone to insufficient throughput, latency jitter, and low data parsing efficiency, which cannot meet the needs of intelligent nuclear energy. Therefore, there is an urgent need to provide a full-link data processing method that can reduce latency and improve data parsing efficiency. Summary of the Invention

[0004] In view of the above-mentioned defects or deficiencies in the related technologies, the purpose of this application is to provide a full-link data processing method and system for nuclear power plant DCS systems, which can reduce latency and improve data parsing efficiency in high-concurrency scenarios.

[0005] To achieve the above objectives, this application provides the following solution:

[0006] In a first aspect, this application provides a full-link data processing method for a nuclear power plant DCS system. The method includes: a subscription-publishing core layer subscribing to multi-source data published by a data publishing terminal and sending the multi-source data to a protocol adaptation layer; the protocol adaptation layer receiving and parsing the multi-source data to generate a parsing result, generating a priority tag based on the multi-source data type of the parsing result, and generating a standard data packet based on the parsing result and the priority tag; the subscription-publishing core layer allocating transmission bandwidth based on the priority tag contained in the standard data packet and publishing the standard data packet to the corresponding topic; the protocol adaptation layer subscribing to the standard data packet to the corresponding topic, encrypting the standard data packet using an end-to-end encryption algorithm, and adding a message authentication code to the encrypted standard data packet to generate an encrypted data packet; and a persistence layer receiving the encrypted data packet and performing hierarchical storage based on the priority tag and data type of the encrypted data packet.

[0007] Optionally, the priority label includes an urgent label, a real-time label, or a batch label; the subscription and publish core layer allocates transmission bandwidth based on the priority label contained in the standard data packet and publishes the standard data packet to the corresponding topic, including: when the priority of the standard data packet is an urgent label, sending the standard data packet to a high-priority queue under the corresponding topic, and allocating an independent link to the high-priority queue to occupy fixed bandwidth; when the priority of the standard data packet is a real-time label or a batch label, sending the standard data packet to a normal queue under the corresponding topic, and using a time-slice round-robin mechanism to allocate the idle bandwidth of the high-priority queue to the normal queue.

[0008] Optionally, the step of encrypting the standard data packet using an end-to-end encryption algorithm and adding a message authentication code to the encrypted standard data packet to generate an encrypted data packet includes: performing multi-threaded parallel encryption on the standard data packet, and optimizing the encryption algorithm through CPU vectorized instruction set during the encryption process to obtain the encrypted standard data packet; generating a fixed-length message authentication code using a national cryptographic compliant message authentication algorithm, and appending the message authentication code to the end of the encrypted standard data packet to obtain the encrypted data packet.

[0009] Optionally, the hierarchical storage based on the priority label and data type of the encrypted data packet includes: when the priority label of the encrypted data packet is an urgent label, storing the encrypted data packet in a high-speed memory pool built based on non-volatile memory NVDIMM; when the priority of the encrypted data packet is a real-time label or a batch label, storing the encrypted data packet in a distributed file system and attaching a timestamp and metadata label.

[0010] Optionally, the method further includes: the protocol adaptation layer writes a configuration file for the new protocol using PDL language, generates a protocol plugin, and places it in a specified directory; the protocol adaptation layer detects changes in the plugin directory in real time, dynamically loads the newly generated protocol plugin using reflection technology, and registers it with the protocol plugin manager so that the newly generated protocol plugin takes effect immediately; in the case of replacing an old version plugin, the old plugin will be retained in memory until the current data processing is completed, and then safely unloaded by the garbage collection thread.

[0011] Secondly, this application provides a nuclear power plant DCS system, which includes: a publish-subscribe core layer, a protocol adaptation layer, a secure transmission layer, and a persistence layer; wherein,

[0012] The subscription and publish core layer is used to subscribe to multi-source data published by the data publisher and send the multi-source data to the protocol adaptation layer; receive standard data packets sent by the protocol adaptation layer, allocate transmission bandwidth based on the priority tags contained in the standard data packets, and publish the standard data packets to the corresponding topics.

[0013] The protocol adaptation layer is used to: receive and parse the multi-source data to generate a parsing result; generate a priority label based on the multi-source data type of the parsing result; generate the standard data packet based on the parsing result and the priority label; and send the standard data packet to the subscription and publication core layer.

[0014] The secure transport layer is used to subscribe to the standard data packets for the corresponding topic, encrypt the standard data packets using the national cryptographic SM4 / SM9 algorithm, add a message authentication code to the encrypted standard data packets to generate encrypted data packets, and send the encrypted data packets to the persistence layer.

[0015] The persistence layer is used to receive the encrypted data packets and perform hierarchical storage based on the priority tags and data types of the encrypted data packets.

[0016] Optionally, the subscription and publish core layer is specifically used to: when the standard data packet contains an urgent priority label, send the standard data packet to a high-priority queue under the corresponding topic, and allocate an independent link to the high-priority queue to occupy fixed bandwidth using a time-slice round-robin mechanism; when the standard data packet contains a real-time priority label or a batch priority label, send the standard data packet to a normal queue under the corresponding topic, and allocate the idle bandwidth of the high-priority queue to the normal queue.

[0017] Optionally, the protocol adaptation layer is also used to: write configuration files for new protocols using PDL language, generate protocol plugins and place them in a specified directory; detect changes in the plugin directory in real time, dynamically load newly generated protocol plugins using reflection technology and register them to the protocol plugin manager so that the newly generated protocol plugins take effect immediately; and when replacing old versions of plugins, the old plugins will be retained in memory until the current data processing is completed, and then safely unloaded by the garbage collection thread.

[0018] Optionally, the secure transport layer is specifically used to perform multi-threaded parallel encryption on the standard data packet, and optimize the encryption algorithm through CPU vectorized instruction set during the encryption process to obtain the encrypted standard data packet; generate a fixed-length message authentication code using a national cryptographic compliant message authentication algorithm, and append the message authentication code to the end of the encrypted standard data packet to obtain the encrypted data packet.

[0019] Optionally, the persistence layer is specifically used to store the encrypted data packet in a high-speed memory pool built based on non-volatile memory NVDIMM when the priority tag of the encrypted data packet is an urgent tag; and to store the encrypted data packet in a distributed file system and attach a timestamp and metadata tag when the priority tag contained in the encrypted data packet is a real-time tag or a batch tag.

[0020] According to the specific embodiments provided in this application, the following technical effects are disclosed:

[0021] This application provides a full-link data processing method and system for nuclear power plant DCS systems. By constructing a hierarchical data bus architecture and clarifying the independent roles of the subscription and publication core layer, protocol adaptation layer, secure transmission layer, and persistence layer, it can overcome the limitations of functional coupling in the traditional integrated architecture of nuclear power DCS data bus. That is, when it is necessary to add a nuclear power-specific protocol to parse multi-source data, only the protocol adaptation layer needs to be adjusted independently to improve data parsing efficiency, without having to rebuild the entire data bus architecture.

[0022] By parsing multi-source data and extracting multi-source data types through the protocol adaptation layer, raw data from different sources and protocols, such as sensors, controllers, and log systems, can be uniformly converted into standard data packets that can be recognized by the subscription and publish core layer. This enables seamless access to multiple heterogeneous data without modifying the data source, thereby improving the system's ability to integrate data from multiple devices.

[0023] By generating priority tags through the protocol adaptation layer, the subscription and publish core layer can dynamically allocate topic paths and loads, which can solve the problems of emergency data being blocked and latency jitter exceeding the standard in high-concurrency scenarios of traditional nuclear power DCS systems, and effectively meet the system's requirements for multi-protocol compatibility and real-time response.

[0024] By using a secure transport layer to encrypt standard data packets end-to-end, data transmission efficiency and security can be effectively balanced.

[0025] By using a persistent layer to store encrypted data packets in a hierarchical manner, the shortcomings of traditional nuclear power DCS systems, such as high cost of full memory storage and large latency of full disk storage, can be solved. This can effectively meet the dual needs of real-time access to high-priority data and low-cost management of massive historical data. Attached Figure Description

[0026] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0027] Figure 1 A structural diagram of a nuclear power plant DCS system provided in one embodiment of this application;

[0028] Figure 2 A hierarchical diagram of the protocol adaptation layer provided in an embodiment of this application;

[0029] Figure 3 This is a flowchart illustrating a full-link data processing method for a nuclear power plant DCS system, provided as an embodiment of this application. Detailed Implementation

[0030] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0031] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0032] In one exemplary embodiment, this application provides a nuclear power plant DCS system, such as Figure 1 As shown, the nuclear power plant DCS system includes a publish-subscribe core layer, a protocol adaptation layer, a secure transport layer, and a persistence layer; among which,

[0033] The core layer of subscription and publication is used to subscribe to multi-source data published by the data publisher and send multi-source data to the protocol adaptation layer; receive standard data packets sent by the protocol adaptation layer, allocate transmission bandwidth based on the priority tags contained in the standard data packets, and publish standard data packets to the corresponding topics.

[0034] The protocol adaptation layer is used to receive and parse multi-source data to generate parsing results; generate priority tags based on the multi-source data types of the parsing results; generate standard data packets based on the parsing results and priority tags; and send standard data packets to the subscription and publish core layer.

[0035] The secure transport layer is used to subscribe to standard data packets for the corresponding topic, encrypt the standard data packets using an end-to-end encryption algorithm, add a message authentication code to the encrypted standard data packets to generate encrypted data packets, and send the encrypted data packets to the persistence layer.

[0036] The persistence layer is used to receive encrypted data packets and store them hierarchically based on the priority tags and data types of the encrypted data packets.

[0037] Optionally, the subscription and publish core layer is specifically used to: when the standard data packet contains an urgent priority label, send the standard data packet to the high priority queue under the corresponding topic, and use a time-slice round-robin mechanism to allocate an independent link to the high priority queue to occupy fixed bandwidth; when the standard data packet contains a real-time priority label or a batch priority label, send the standard data packet to the ordinary queue under the corresponding topic, and allocate the idle bandwidth of the high priority queue to the ordinary queue.

[0038] Optionally, the protocol adaptation layer is also used to: write configuration files for new protocols using the PDL language, generate protocol plugins and place them in a specified directory; detect changes in the plugin directory in real time, dynamically load newly generated protocol plugins using reflection technology and register them to the protocol plugin manager so that the newly generated protocol plugins take effect immediately; and when replacing old versions of plugins, the old plugins will be retained in memory until the current data processing is completed, and then safely unloaded by the garbage collection thread.

[0039] Optionally, the secure transport layer is specifically used to perform multi-threaded parallel encryption on standard data packets, and optimize the encryption algorithm through CPU vectorized instruction set during the encryption process to obtain encrypted standard data packets; generate a fixed-length message authentication code using a national cryptographic compliant message authentication algorithm, and append the message authentication code to the end of the encrypted standard data packets to obtain encrypted data packets.

[0040] Optionally, the persistence layer is specifically used to store the encrypted data packet in a high-speed memory pool built based on non-volatile memory NVDIMM when the priority label of the encrypted data packet is an urgent label; and to store the encrypted data packet in a distributed file system with a timestamp and metadata label when the priority label of the encrypted data packet is a real-time label or a batch label.

[0041] Combination Figure 1Understandably, the data publishing end in this application embodiment includes sensors, controllers, and log sources, etc. The subscription-publishing core layer is built based on an improved subscription-publishing model, receiving multi-source data in the form of subscriptions. This multi-source data may include control commands, device status, logs, etc. After processing, the multi-source data is published to the protocol adaptation layer. Dynamic topic routing and load balancing algorithms based on priority tags and data types are used to improve throughput. The protocol adaptation layer supports multi-source data access and has a built-in protocol parsing engine for nuclear power-specific protocols. The protocol parsing engine parses the multi-source data to extract data types and provides priority tags to the subscription-publishing core layer based on these data types. Protocol standardization tools are used to convert the parsed multi-source data into standard data packets required by the receiver. The secure transmission layer integrates the national cryptographic algorithms SM4 / SM9 to encrypt the published data and reduce the performance overhead of encryption calculations. The persistence layer is responsible for persistently storing the parsed subscription data. Through a collaborative design of hierarchical storage and a rolling time window mechanism, it balances the high real-time requirements of data with the need for massive historical data storage.

[0042] Furthermore, the publish-subscribe core layer receives multi-source data from the data publisher in the form of subscriptions and publishes standard data to the secure transmission layer. After subscribing to the data, the protocol adaptation layer parses the data and assigns three priority levels—urgent, real-time, and batch—based on the data type. Different data transmission strategies are then adopted to maximize system throughput. These data transmission strategies include: for urgent data, a priority queue algorithm is used to exclusively utilize the FPGA's high-speed link for transmission; for real-time data, time-slice round-robin scheduling is used to dynamically allocate bandwidth for transmission; and for batch data, asynchronous thread fragmentation is used to avoid blocking the real-time transmission channel and reducing latency. The three priority levels—urgent, real-time, and batch—are shown in Table 1.

[0043] Table 1

[0044]

[0045] Urgent data is handled using a priority queue algorithm, allowing it to directly preempt current transmission channel resources. A high-priority queue and a regular queue are set up in the publish-subscribe core layer. The high-priority queue has a pre-allocated independent link with fixed bandwidth, allowing only urgent data to be inserted. The regular queue handles real-time and batch data transmission; when the high-priority queue is idle, the regular queue can utilize its bandwidth.

[0046] Real-time data is distributed through time-slice round-robin to ensure fairness among multiple data streams. The base duration of each time slice is specified. The number of real-time data streams N is dynamically adjusted according to the following formula (1):

[0047] (1)

[0048] When N>10, the starter time slices are grouped and rotated, with each group containing 5 streams. The weight W of each data stream is determined by the following equation (2):

[0049] (2)

[0050] Among them, coefficient C i Equipment importance is categorized into levels 1 to 10. F i This refers to the data update frequency.

[0051] Rotation press W i Bandwidth is allocated proportionally to ensure priority transmission for critical equipment (such as reactor pressure sensors). If a data stream is not fully transmitted in a given session, the remaining data will be used for transmission. D 剩余 It is included in the next time slice, and its weight is dynamically increased by the following formula (3):

[0052] (3)

[0053] Batch-level data triggers asynchronous fragmented transmission, reducing the impact of large data blocks (such as logs) on real-time performance. Large data blocks are split into fixed-size fragments of 1MB each, with each fragment having a sequence number and checksum. Transmission is performed via an asynchronous thread pool, with each thread bound to an independent memory buffer to avoid lock contention.

[0054] The protocol adaptation layer and the publish-subscribe layer work together to parse the subscribed multi-source data and obtain the data type, data content, checksum, etc. The data type serves as the basis for the publish-subscribe layer to determine the priority. Before the data is published, different types of multi-source data are standardized so that the secure transmission layer can process them uniformly.

[0055] like Figure 2 As shown, the protocol adaptation layer of this application provides a standardized software development kit (SDK), allowing users to quickly define and extend nuclear power industry-specific protocol parsing rules, thus solving the problem of traditional DCS system protocol adaptation relying on customized development. The core components of the SDK include:

[0056] Protocol Description Language (PDL): Based on YAML syntax, users can declaratively configure and define the protocol structure (such as protocol fields, data types, and validation rules).

[0057] Custom code generation tool: The protocol configuration file is converted into high-performance parsing code by the compiler, and interface code for reading parsed data within modules is provided to ensure parsing efficiency. After receiving the subscribed multi-source raw data, the plugin matching engine of the protocol adaptation layer is invoked based on the preamble or feature fields of the data stream. The matching engine quickly searches the protocol plugin directory through a hash table and loads the corresponding protocol description file (PDL configuration file). Then, the data is extracted byte by byte according to the field offsets, lengths, and data types defined in the PDL, and the parsed data types are submitted to the subscription and publication layer as the basis for priority determination.

[0058] Protocol verification toolchain: This toolchain converts the parsed data into a binary data stream conforming to the target device's protocol (e.g., Modbus RTU frames, nuclear power proprietary protocol messages) using protocol standardization tools. First, based on the target device's protocol type, the corresponding PDL configuration file is loaded, and the fields are filled in reverse according to the rules defined in the PDL. Then, it is converted to binary data according to the byte order (big-endian / little-endian) specified in the PDL. Finally, a checksum is calculated on the encapsulated binary data using a checksum algorithm defined in the PDL (e.g., CRC32) and appended to the end of the message.

[0059] Furthermore, to achieve zero-downtime upgrades of protocol plugins, a dynamic loading and hot-swapping mechanism is provided. This involves real-time detection of changes to the protocol plugin directory (such as additions, deletions, and updates) through operating system-level file monitoring, and dynamically loading plugin interfaces using reflection technology and registering them with the protocol plugin manager to ensure that new protocols take effect immediately. During plugin replacement, the old version of the plugin remains in memory until the current data processing is complete, after which it is safely unloaded by the garbage collection thread to prevent data parsing interruptions. This design significantly improves the availability and maintainability of nuclear power DCS systems.

[0060] The secure transport layer subscribes to standard data packets for the corresponding topic and employs end-to-end encryption to perform multi-threaded encryption processing on the standard data packets. This application's secure transport layer integrates the national cryptographic algorithms SM4 / SM9, achieving a synergistic improvement in security and performance through multi-threaded parallel encryption and deep optimization using SIMD instructions. Specifically, it includes:

[0061] Multi-threaded parallel encryption: Utilizing a multi-core CPU architecture to achieve parallel processing of encryption tasks, thereby improving throughput. The data to be encrypted is divided into segments of a fixed block size (e.g., 16KB), and each segment is assigned to an independent thread for processing. The dynamic thread pool adjusts the number of active threads according to the system load, calculated using the following formula (4):

[0062] (4)

[0063] in, Number of CPU cores This represents the total amount of data to be encrypted. Simultaneously, thread-local storage is used to pre-allocate the encryption context, avoiding the overhead of repeated initialization.

[0064] SIMD Instruction Acceleration: Round function operations for SM4 / SM9 are accelerated through CPU vectorized instruction sets. The S-box and linear transformation table are preloaded into the CPU cache, and 32-byte parallel table lookups are implemented using the AVX-512 instruction set. Large integer multiplication and modular reduction operations are vectorized based on ARM Neon or Intel AVX2 instruction sets. Simultaneously, scalar multiplication results with fixed base points are pre-computed during the key generation stage, reducing real-time computation.

[0065] It's important to note that in the Chinese cryptographic algorithms SM4 and SM9, the S-box (Substitution Box) and the linear transformation table are core components used to implement non-linear substitution and linear diffusion, ensuring the algorithm's security. Specifically, the S-box is a 16×16 static lookup table used to map 8-bit input data to another 8-bit data, achieving non-linear substitution. In SM4, the linear transformation table refers to the linear transformation L, used to perform diffusion operations on the S-box output, typically achieved through table lookup or shift-and-XOR operations. In SM9, the linear transformation table is used for large number operations (such as elliptic curve multiplication), referring to a pre-computed base-point multiplication result table used to accelerate scalar multiplication.

[0066] The persistent layer, based on the encrypted data obtained from the aforementioned secure transmission layer, employs a tiered storage and rolling time window mechanism to balance the high real-time requirements of nuclear power DCS data with the need for massive historical data storage, thus resolving the inherent trade-off between performance and capacity in traditional persistent storage solutions. Specifically:

[0067] Data classification and storage rules: High-priority data such as control commands and device alarms are stored in a high-speed memory pool based on non-volatile NVDIMM memory, supporting microsecond-level read and write operations, and employing dual-copy redundancy to prevent data loss. Non-critical data such as operation logs and historical records are written to the distributed file system via an asynchronous disk flushing mechanism, with timestamps and metadata tags attached for easy retrospective analysis. Simultaneously, the write thread only operates on the current buffer, while a background thread asynchronously flushes the full buffer to disk, avoiding I / O blocking during real-time writing.

[0068] Time window rolling mechanism: The time window is the core time unit for data segmentation, storage, and management. Essentially, it divides a continuous data stream into logical segments of fixed or dynamic duration to achieve efficient batch processing and resource scheduling. This application dynamically adjusts the rolling window size based on system load factors such as CPU utilization and disk IOPS. The weighted calculation is performed based on the following formula (5):

[0069] (5)

[0070] in, This indicates the baseline window size; in this embodiment, it can be set to 1 minute by default. This indicates the current CPU utilization rate (0%~100%). This indicates the current disk IOPS (times / second). This indicates the CPU utilization threshold; in this embodiment, it can be set to 80% by default. This represents the disk IOPS threshold; in this embodiment, it can be set to 10,000 by default. and This represents the weighting coefficient, reflecting the balanced impact of CPU and disk.

[0071] When the load exceeds a threshold, the window is reduced to decrease the amount of data flushed to disk in a single operation. Data within the window is divided into blocks in chronological order, and an index file is generated to record the start and end times and storage locations of the blocks, accelerating historical data retrieval.

[0072] This application achieves efficient, secure, and reliable processing of data collected by reactor pressure sensors from acquisition, parsing, transmission, encryption to storage through the aforementioned multi-layer collaborative processing flow, effectively meeting the needs of nuclear power plant distributed control systems for high real-time performance, multi-protocol compatibility, high security, and massive data management.

[0073] Specifically, this application constructs a layered data bus architecture, clearly defining the independent roles of the subscription and publication core layer, protocol adaptation layer, secure transmission layer, and persistence layer. This can overcome the limitations of functional coupling in the traditional nuclear power DCS data bus integrated architecture. That is, when a new nuclear power-specific protocol is needed to parse multi-source data, only the protocol adaptation layer needs to be adjusted independently to improve data parsing efficiency, without having to rebuild the entire data bus architecture.

[0074] By parsing multi-source data and extracting multi-source data types through the protocol adaptation layer, raw data from different sources and protocols (including nuclear power-specific protocols) such as sensors, controllers, and log systems can be uniformly converted into standard data packets that can be recognized by the subscription and publication core layer. Without modifying the data source, seamless access to multiple heterogeneous data can be achieved, improving the system's ability to integrate data from multiple devices.

[0075] By generating priority tags through the protocol adaptation layer, the subscription and publish core layer can dynamically allocate topic paths and loads, which can solve the problems of emergency data being blocked and latency jitter exceeding the standard in high-concurrency scenarios of traditional nuclear power DCS systems, and effectively meet the system's requirements for multi-protocol compatibility and real-time response.

[0076] By using a secure transport layer to encrypt standard data packets end-to-end, data transmission efficiency and security can be effectively balanced.

[0077] By using a persistent layer to store encrypted data packets in a hierarchical manner, the shortcomings of traditional nuclear power DCS systems, such as high cost of full memory storage and large latency of full disk storage, can be solved. This can effectively meet the dual needs of real-time access to high-priority data and low-cost management of massive historical data.

[0078] Furthermore, this application generates emergency-level tags, real-time-level tags, and batch-level tags through a protocol adaptation layer. Combined with dynamic topic routing and load balancing in the subscription and publish core layer, it solves the problems of emergency data being blocked and latency jitter exceeding standards in high-concurrency scenarios in existing technologies. Specifically, emergency-level data that is critical to nuclear safety and corresponds to the emergency-level tag can be prioritized to preempt transmission resources, achieving end-to-end transmission and effectively reducing latency. Real-time-level data that corresponds to the real-time-level tag ensures transmission stability through dynamic bandwidth allocation. Batch-level data that corresponds to the batch-level tag does not occupy critical transmission resources, which can effectively meet the system's requirements for security priority and real-time response.

[0079] This application integrates the national cryptographic SM4 / SM9 algorithm into the secure transport layer to achieve end-to-end encryption of all data, preventing the leakage or tampering of sensitive nuclear power data and complying with regulations such as the "Nuclear Industry Data Security Specification". At the same time, through multi-threaded parallel encryption and SIMD instruction acceleration optimization, the performance loss of the encryption process can be reduced, ensuring data transmission security while maintaining the throughput of the data bus.

[0080] This application addresses the pain points of high cost and high latency of full-scale memory storage and full-scale disk storage in existing technologies by using a tiered storage and dynamic time window rolling mechanism in the persistence layer. Specifically, urgent data is stored in a high-speed NVDIMM memory pool, supporting microsecond-level read / write and dual-copy redundancy to ensure real-time access to core data and prevent data loss; real-time data and batch data are asynchronously flushed to a distributed file system, and the dynamic adjustment of the time window reduces disk I / O pressure, thus meeting the dual needs of real-time access to urgent data and low-cost management of massive historical data.

[0081] This application improves the stability of system operation by implementing a collaborative operation across the entire chain of data reception, protocol parsing, dynamic routing, security encryption, and hierarchical storage, as well as seamless data flow at each level based on priority tags.

[0082] Based on the same inventive concept, this application also provides a full-link data processing method for a nuclear power plant DCS system. The solution provided by this method is similar to the implementation scheme described in the above system. Therefore, the specific limitations of one or more embodiments of the full-link data processing method for a nuclear power plant DCS system provided below can be found in the limitations of the nuclear power plant DCS system described above, and will not be repeated here.

[0083] In one exemplary embodiment, this application provides a full-link data processing method applied to a nuclear power plant DCS system, such as... Figure 3 As shown, the method includes the following steps S301 to S305, wherein:

[0084] Step S301: Subscribe to the multi-source data published by the core layer's data publisher and send the multi-source data to the protocol adaptation layer.

[0085] Step S302: The protocol adaptation layer receives and parses multi-source data to generate a parsing result, generates a priority label based on the multi-source data type of the parsing result, and generates a standard data packet based on the parsing result and the priority label.

[0086] Step S303: The subscription and publication core layer allocates transmission bandwidth based on the priority tags contained in the standard data packets and publishes standard data packets to the corresponding topics;

[0087] Step S304: The secure transport layer subscribes to standard data packets from the corresponding topic, encrypts the standard data packets using an end-to-end encryption algorithm, and adds a message authentication code to the encrypted standard data packets to generate encrypted data packets.

[0088] In step S305, the persistence layer receives the encrypted data packet and performs hierarchical storage based on the priority label and data type of the encrypted data packet.

[0089] As an optional implementation, the priority label includes an emergency label, a real-time label, or a batch label; step S303 may include: when the priority of the standard data packet is an emergency label, sending the standard data packet to the high-priority queue under the corresponding topic, and allocating an independent link to the high-priority queue to occupy fixed bandwidth; when the priority of the standard data packet is a real-time label or a batch label, sending the standard data packet to the ordinary queue under the corresponding topic, and using a time-slice round-robin mechanism to allocate the idle bandwidth of the high-priority queue to the ordinary queue.

[0090] As an optional implementation, step S304 may include: using the national cryptographic SM4 / SM9 algorithm to perform multi-threaded parallel encryption on the standard data packet, and optimizing the encryption algorithm through the CPU vectorized instruction set during the encryption process to obtain the encrypted standard data packet; using a national cryptographic compliant message authentication algorithm to generate a fixed-length message authentication code, and appending the message authentication code to the end of the encrypted standard data packet to obtain the encrypted data packet.

[0091] As an optional implementation, step S305 may include: storing the encrypted data packet in a high-speed memory pool built based on non-volatile memory NVDIMM when the priority tag of the encrypted data packet is an urgent tag; storing the encrypted data packet in a distributed file system and attaching a timestamp and metadata tag when the priority tag contained in the encrypted data packet is a real-time tag or a batch tag.

[0092] As an optional implementation, the above method may further include: the protocol adaptation layer writes a configuration file for the new protocol using PDL language, generates a protocol plugin, and places it in a specified directory; the protocol adaptation layer detects changes in the plugin directory in real time, dynamically loads the newly generated protocol plugin using reflection technology, and registers it with the protocol plugin manager so that the newly generated protocol plugin takes effect immediately; in the case of replacing an old version plugin, the old plugin will be retained in memory until the current data processing is completed, and then safely unloaded by the garbage collection thread.

[0093] For example, let's take the multi-source data collected by the reactor pressure sensor of Nuclear Power Unit 1 as an example to explain the above method. That is:

[0094] The reactor pressure sensor, acting as a data publisher, transmits pressure measurements collected by the sensor to the data bus via the ModbusTCP protocol. The publish-subscribe core layer of this application subscribes to multi-source data from the reactor pressure sensor in real time, and then forwards the acquired raw binary data stream to the protocol adaptation layer for parsing. The protocol adaptation layer has a built-in parsing engine for nuclear power-specific protocols. After recognizing the preamble of the raw binary data stream, the parsing engine calls the matching parsing plugin and extracts the device identifier "NP1_RPV_Sensor_001", the data type "Reactor_Pressure", the specific data value of the pressure measurement, and the timestamp information byte-by-byte according to the Protocol Description Language (PDL) configuration file, and automatically generates a real-time priority tag accordingly. Subsequently, the parsing engine encapsulates the parsing result into a standard data packet with a header structure and sends it back to the publish-subscribe core layer.

[0095] After receiving standard data packets, the core layer of the subscription and publish layer triggers a dynamic topic routing algorithm based on its real-time priority label. This dynamic topic routing algorithm calculates transmission weights based on data characteristics. For example, if the importance coefficient C of the reactor pressure sensor... i =8, update frequency F i =10Hz, substitute into the weighting formula W i =0.6 C i +0.4 log(F) i+1) The calculated value is 5.216. Based on this weight value and the sum of the weights of other data flows within the current link, the system allocates transmission bandwidth proportionally and routes standard data packets to all subscribers of the topic " / NP1 / Reactor / Pressure / Value". Simultaneously, high-priority, urgent data can preempt the transmission channel using a priority queue mechanism, ensuring that its end-to-end latency does not exceed 10 milliseconds.

[0096] After leaving the publish-subscribe core layer, standard data packets enter the secure transport layer for encryption. The secure transport layer fragments the data packets into 16KB blocks and performs multi-threaded encryption tasks. The encryption task utilizes the S-box and linear transformation table preloaded into the CPU cache, performs parallel table lookups using the AVX-512 vectorized instruction set, executes the national standard SM4 algorithm for encryption, and appends a message authentication code to form encrypted data packets, ensuring data integrity and tamper-proof performance.

[0097] The encrypted data packet is ultimately delivered to the persistence layer for storage decision-making. Based on the real-time tag and data type in the encrypted data packet header, the persistence layer invokes a tiered storage strategy: first, the decrypted plaintext data is synchronously written to a high-speed NVDIMM-based memory pool for microsecond-level real-time access by the main control room monitoring system; simultaneously, the encrypted data packet is sent to an asynchronous write buffer. The persistence layer synchronously initiates a time window rolling mechanism, dynamically adjusting the disk flushing window based on the system's real-time load. In this embodiment, if CPU utilization (75%) and disk IOPS (9500) are both below the set threshold, the window can maintain a baseline value of 60 seconds. When the window expires, the data in the buffer is batch compressed into Parquet format and written to the distributed file system. Simultaneously, a time range index is generated and stored in the database, significantly reducing disk I / O pressure and ensuring efficient retrieval of historical data.

[0098] By implementing steps S301 to S305 above, a hierarchical data bus architecture can be constructed, clearly defining the independent roles of the subscription and publication core layer, protocol adaptation layer, secure transmission layer, and persistence layer. This can overcome the limitations of functional coupling in the traditional nuclear power DCS integrated data bus architecture. That is, when a new nuclear power-specific protocol is needed to parse multi-source data, only the protocol adaptation layer needs to be adjusted independently to improve data parsing efficiency, without having to rebuild the entire data bus architecture.

[0099] By parsing multi-source data and extracting multi-source data types through the protocol adaptation layer, raw data from different sources and protocols, such as sensors, controllers, and log systems, can be uniformly converted into standard data packets that can be recognized by the subscription and publish core layer. This enables seamless access to multiple heterogeneous data without modifying the data source, thereby improving the system's ability to integrate data from multiple devices.

[0100] By generating priority tags through the protocol adaptation layer, the subscription and publish core layer can dynamically allocate topic paths and loads, which can solve the problems of emergency data being blocked and latency jitter exceeding the standard in high-concurrency scenarios of traditional nuclear power DCS systems, and effectively meet the system's requirements for multi-protocol compatibility and real-time response.

[0101] By using a secure transport layer to encrypt standard data packets end-to-end, data transmission efficiency and security can be effectively balanced.

[0102] By using a persistent layer to store encrypted data packets in a hierarchical manner, the shortcomings of traditional nuclear power DCS systems, such as high cost of full memory storage and large latency of full disk storage, can be solved. This can effectively meet the dual needs of real-time access to high-priority data and low-cost management of massive historical data.

[0103] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0104] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).

[0105] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0106] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0107] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A full-link data processing method applied to a DCS system of a nuclear power plant, characterized in that, The full-link data processing method applied to the DCS system of a nuclear power plant comprises the following steps: The subscription and publication core layer subscribes to multi-source data published by a data publication end of the subscription and publication core layer, and sends the multi-source data to the protocol adaptation layer; The protocol adaptation layer receives and analyzes the multi-source data to generate an analysis result, generates a priority label based on the type of the multi-source data of the analysis result, and generates a standard data packet based on the analysis result and the priority label; The subscription and publication core layer allocates transmission bandwidth based on the priority label contained in the standard data packet, and publishes the standard data packet to a corresponding topic; The secure transmission layer subscribes to the standard data packet to a corresponding topic, encrypts the standard data packet by using an end-to-end encryption algorithm, and adds a message authentication code to the encrypted standard data packet to generate an encrypted data packet; The persistence layer receives the encrypted data packet, and performs hierarchical storage based on the priority label and the data type of the encrypted data packet.

2. The full-link data processing method applied to a DCS system of a nuclear power plant according to claim 1, characterized in that, The priority label comprises an emergency level label, a real-time level label, or a batch level label; The subscription and publication core layer allocates transmission bandwidth based on the priority label contained in the standard data packet, and publishes the standard data packet to a corresponding topic, which comprises the following steps: In the case that the priority label contained in the standard data packet is an emergency level label, the standard data packet is sent to a high-priority queue under the corresponding topic, and an independent link is allocated to the high-priority queue to occupy a fixed bandwidth; In the case that the priority label contained in the standard data packet is a real-time level label or a batch level label, the standard data packet is sent to a normal queue under the corresponding topic, and a time slice rotation mechanism is used to allocate the idle bandwidth of the high-priority queue to the normal queue.

3. The full-link data processing method applied to a DCS system of a nuclear power plant according to claim 1, characterized in that, The encrypted data packet is generated by encrypting the standard data packet by using an end-to-end encryption algorithm, and adding a message authentication code to the encrypted standard data packet, which comprises the following steps: The standard data packet is encrypted by using a multi-thread parallel encryption algorithm of the national standard SM4 / SM9 algorithm, and the encryption algorithm is optimized by using a CPU vector instruction set during the encryption process to obtain the encrypted standard data packet; A message authentication code of a fixed length is generated by using a message authentication algorithm in line with the national standard, and the message authentication code is attached to the tail of the encrypted standard data packet to obtain the encrypted data packet.

4. The full-link data processing method applied to a DCS system of a nuclear power plant according to claim 1, characterized in that, The encrypted data packet is stored based on the priority label and the data type of the encrypted data packet, which comprises the following steps: In the case that the priority label of the encrypted data packet is an emergency level label, the encrypted data packet is stored in a high-speed memory pool constructed based on a non-volatile memory NVDIMM; In the case that the priority label contained in the encrypted data packet is a real-time level label or a batch level label, the encrypted data packet is stored in a distributed file system, and a timestamp and a metadata label are attached.

5. The full-link data processing method applied to a DCS system of a nuclear power plant according to claim 1, characterized in that, The method further comprises the following steps: The protocol adaptation layer writes a configuration file of a new protocol by using a PDL language, generates a protocol plug-in, and places the protocol plug-in in a specified directory; The protocol adaptation layer detects changes in the plug-in directory in real time, dynamically loads the newly generated protocol plug-in by using a reflection technology, and registers the protocol plug-in to a protocol plug-in manager, so that the newly generated protocol plug-in takes effect immediately; In the case of replacing the old version plug-in, the old plug-in is kept in the memory until the current data processing is completed, and then unloaded by the garbage collection thread safely.

6. A nuclear power plant DCS system characterized by, The nuclear power plant DCS system comprises a subscription publishing core layer, a protocol adaptation layer, a secure transmission layer and a persistence layer, wherein The subscription publishing core layer is configured to subscribe to multi-source data published by a data publishing end, and send the multi-source data to the protocol adaptation layer; receive a standard data packet sent by the protocol adaptation layer, allocate a transmission bandwidth based on a priority label contained in the standard data packet, and publish the standard data packet to a corresponding topic; receive a standard data packet sent by the protocol adaptation layer, allocate a transmission bandwidth based on a priority label contained in the standard data packet, and publish the standard data packet to a corresponding topic; The protocol adaptation layer is configured to receive and analyze the multi-source data to generate an analysis result, generate a priority label based on a multi-source data type of the analysis result, generate the standard data packet based on the analysis result and the priority label, and send the standard data packet to the subscription publishing core layer; The secure transmission layer is configured to subscribe to the standard data packet to a corresponding topic, encrypt the standard data packet using an end-to-end encryption algorithm, add a message authentication code to the encrypted standard data packet to generate an encrypted data packet, and send the encrypted data packet to the persistence layer; The persistence layer is configured to receive the encrypted data packet, and perform hierarchical storage based on a priority label and a data type of the encrypted data packet.

7. The nuclear power plant DCS system in accordance with claim 6, wherein, The subscription publishing core layer is specifically configured to, In the case that the priority label contained in the standard data packet is an emergency level label, the standard data packet is sent to a high priority queue under the corresponding topic, and a time slice rotation mechanism is used to allocate an independent link to the high priority queue to occupy a fixed bandwidth; In the case that the priority label contained in the standard data packet is a real-time level label or a batch level label, the standard data packet is sent to a normal queue under the corresponding topic, and the normal queue is allocated with idle bandwidth of the high priority queue.

8. The nuclear power plant DCS system in accordance with claim 6, wherein, The protocol adaptation layer is further configured to, Write a configuration file of a new protocol through a PDL language, generate a protocol plug-in and place it in a specified directory; Detect changes in the plug-in directory in real time, dynamically load the newly generated protocol plug-in using reflection technology and register it to a protocol plug-in manager, so that the newly generated protocol plug-in takes effect immediately; In the case of replacing the old version plug-in, the old plug-in is kept in the memory until the current data processing is completed, and then unloaded by the garbage collection thread safely.

9. The nuclear power plant DCS system in accordance with claim 6, wherein, The secure transmission layer is specifically configured to, Use the national SM4 / SM9 algorithm to perform multi-thread parallel encryption on the standard data packet, and optimize the encryption algorithm through CPU vectorization instruction set during the encryption process to obtain the encrypted standard data packet; Use a national message authentication algorithm to generate a message authentication code of a fixed length, and attach the message authentication code to the tail of the encrypted standard data packet to obtain the encrypted data packet.

10. The nuclear power plant DCS system in accordance with claim 6, wherein, The persistence layer is specifically configured to, In the case that the priority label of the encrypted data packet is an emergency level label, the encrypted data packet is stored in a high-speed memory pool constructed based on a non-volatile memory NVDIMM; In the case that the priority level of the encrypted data packet is real-time or batch, the encrypted data packet is stored in the distributed file system and a timestamp and metadata tag are attached.

Images