CN115242826A

CN115242826A - Real-time transmission and storage method for data of nuclear power plant

Info

Publication number: CN115242826A
Application number: CN202210568516.0A
Authority: CN
Inventors: 方华建; 李青; 罗俊; 肖云龙; 吉艳红; 李敏
Original assignee: China Nuclear Power Operation Technology Corp Ltd
Current assignee: China Nuclear Power Operation Technology Corp Ltd
Priority date: 2022-05-23
Filing date: 2022-05-23
Publication date: 2022-10-25
Anticipated expiration: 2042-05-23
Also published as: CN115242826B

Abstract

The disclosure belongs to the technical field of nuclear power, and particularly relates to a real-time data transmission and storage method for a nuclear power plant. All systems and programs for data acquisition, transmission and distribution and a digital twin system in the method are deployed in a production area, and a distributed message system is used for caching data in the distribution system, so that the problem that when the data volume is too large, the data processing program cannot process all data technically, and data accumulation and transmission link congestion are caused is avoided. In addition, all sensor measuring point data and DCS measuring point data in the time sequence database are classified and stored in a partitioned mode, and storage and use efficiency is improved.

Description

Real-time transmission and storage method for nuclear power plant data

Technical Field

The invention belongs to the technical field of nuclear power, and particularly relates to a real-time data transmission and storage method for a nuclear power plant.

Background

Digitization and intellectualization of nuclear power plants are one of the directions of nuclear power development, and are also important means for improving the operation safety and the economical efficiency of the nuclear power plants. The unit data is an important basis for digitization and intellectualization of the nuclear power plant, after the unit data is obtained in real time, on one hand, data processing and integration can be carried out in a unified nuclear power data standard system framework, a unified data development system and a data service system are created, the efficiency of data supply and demand is improved, and the data demand service is applied.

However, real-time acquisition of crew data involves multiple aspects including: data acquisition, data transmission, data distribution, data storage and the like. At present, in the aspect of nuclear power data acquisition, an acquisition program is a single channel, and if a channel link fails, a data acquisition process cannot be carried out; in the aspect of data transmission and distribution, real machine group data is transmitted in a certain quantity at a certain frequency, and when the data quantity is large, if the data is not transmitted in time, partial data can be lost; in the aspect of data storage, the measuring point data in the PI system of the nuclear power plant are not classified and stored at present, and efficient use of the data is not facilitated. In addition, for a traditional nuclear power plant simulator, the capacity of acquiring unit data in real time is not provided, so that the running state of the unit is difficult to track, and self-optimization and automatic tracking of a simulation model cannot be realized by utilizing the running data of the unit.

Disclosure of Invention

In order to overcome the problems in the related art, a method for transmitting and storing data of a nuclear power plant in real time is provided.

According to an aspect of the embodiments of the present disclosure, a method for real-time transmission and storage of nuclear power plant data is provided, where the method includes:

s1, data acquisition and transmission are respectively in communication connection with a field device sensor and a DCS through a data acquisition interface machine deployed in a power plant production area, one or more acquisition programs are deployed in the data acquisition interface machine to acquire the data of the field device sensor and the data of the DCS in real time, and the acquired data are sent to a distributed message system Kafka;

s2, data distribution, wherein the distributed message system Kafka, the time sequence database and the digital twin system are all deployed in a production area, data acquired by the data acquisition interface machine enter corresponding topics in the distributed message system Kafka and are temporarily cached, a time sequence data writing program of the time sequence database and the digital twin system subscribe the corresponding topics in the distributed message system Kafka respectively, and the time sequence data writing program and the digital twin system consume respectively under the condition that the subscribed data are judged to arrive;

and S3, storing data, namely storing the data subscribed by the time sequence data writing program into a time sequence database for classification and partition storage, and enabling the data subscribed by the digital twin system to enter the digital twin system in real time for further processing and analysis.

In one possible implementation, the field device sensor is provided with two sensor data sending modules for sending the same sensor data at the same time; the DCS is provided with two DCS data sending modules for sending the same DCS data at the same time; the data acquisition interface machine is provided with a sensor data acquisition program and a DCS system data acquisition program, and the method also comprises the following steps:

the sensor data acquisition program receives data from the two sensor data sending modules, compares whether the received data has repeated data, deletes one of the two repeated data and caches the other of the two repeated data under the condition of judging that the two repeated data exist;

the DCS system data acquisition program receives data from the two DCS system data sending modules, compares whether the received data has repeated data or not, deletes one data of the two repeated data and caches the other data of the two repeated data under the condition that the two repeated data are judged to exist.

In one possible implementation, the method further includes:

the sensor data acquisition program compares the data received in the current period with the data received in the previous period of the current period, and judges whether repeated data exist or not;

the DCS system data acquisition program compares the data received in the current period with the data received in the previous period of the current period, and judges whether repeated data exist or not.

In a possible implementation manner, the data sent by the data acquisition interface machine to the distributed message system Kafka at least includes a measurement point code, an acquisition timestamp, data quality, and a measurement point value, where the measurement point code is used to uniquely identify each measurement point of the nuclear power plant, the acquisition timestamp is used to indicate the time when the data acquisition interface machine acquires data, the data quality is used to indicate the quality of the data acquired by the acquisition interface machine, and the measurement point value is used to indicate a value acquired by a measurement point of the nuclear power plant.

In one possible implementation, the method of claim 1, wherein the data sent by the data collection interface machine to the distributed messaging system Kafka includes at least a measurement point code, a collection timestamp, a data quality, and a measurement point value, and the method further comprises:

the sensor data acquisition program judges that the two received data are repeated data under the condition that the measuring point codes, the acquisition timestamps and the measuring point values of the two received data are the same;

and the DCS system data acquisition program judges that the two received data are repeated data under the condition that the measuring point codes, the acquisition timestamps and the measuring point values of the two received data are the same.

In one possible implementation mode, the measuring point code comprises a first part, a second part and a third part, wherein the first part is a nuclear power plant code, the second part is a unit code, and the third part is an original measuring point number of the nuclear power plant.

In one possible implementation, the method further includes:

the time sequence database encodes the received data to obtain encoded data;

the time sequence database stores the coded data in a classified manner according to preset power plant types, unit types and process system types, and the data of the same type are stored in storage areas corresponding to the type.

In one possible implementation, the time-series database encodes the received data, and includes:

under the condition that the type of the time sequence database is an IoTDB database, if the received data is low-frequency measuring point data, coding a double-precision floating point value, a single-precision floating point value and an integer value by adopting a GORILLA coding mode, and coding a Boolean value by adopting a run coding mode.

In one possible implementation, the time-series database encodes the received data, and further includes:

and under the condition that the type of the time sequence database is an IoTDB database, if the measuring point data is high-frequency measuring point data, coding by adopting a PLAIN coding mode.

In one possible implementation, the method further includes: and the data acquisition interface machine transmits the acquired data to the distributed message system in a data packet mode.

The beneficial effect of this disclosure lies in: all systems and programs for data acquisition, transmission and distribution and a digital twin system in the method are deployed in a production area, and a distributed message system is used for caching data in the distribution system, so that the problem that when the data volume is too large, the data processing program cannot process all data technically, and data accumulation and transmission link congestion are caused is avoided. In addition, all sensor measuring point data and DCS measuring point data in the time sequence database are classified and stored in a partitioned mode, and storage and use efficiency is improved.

The real-time data transmission and storage method for the nuclear power plant uses innovative data storage, data transmission and other technologies, and a digital twin system is arranged in a generation area; the digital twin system of the nuclear power plant has the capability of data interaction with an actual unit, so that the self-optimization and automatic tracking of the digital twin system are supported.

Drawings

FIG. 1 is a flow diagram illustrating a method for real-time transmission and storage of nuclear plant data, according to an exemplary embodiment.

FIG. 2 is a DCS data acquisition schematic shown according to an exemplary embodiment.

FIG. 3 is a DCS data acquisition flow diagram shown in accordance with an exemplary embodiment.

FIG. 4 illustrates an example of time series data site encoding in accordance with an exemplary embodiment.

FIG. 5 is an illustration of an IoTDB data model, according to an example embodiment

Detailed Description

The invention is further described in detail below with reference to the drawings and specific embodiments.

FIG. 1 is a flow diagram illustrating a method for real-time transmission and storage of nuclear plant data, according to an exemplary embodiment. As shown in fig. 1, the real-time data transmission, storage and interaction of the digital twin system of the nuclear power plant are mainly carried out through the following steps:

s1, data acquisition and transmission are carried out, wherein the data acquisition interface machine is arranged in a production area of a power plant and is in communication connection with a field device sensor and/or a DCS (Distributed Control System), a sensor data acquisition program and/or a DCS acquisition program are arranged in the data acquisition interface machine so as to acquire sensor data and DCS data in real time, and the acquired data are sent to a Distributed message System Kafka through a data packet format, wherein the sensor data acquisition program and/or the DCS acquisition program carry out data acquisition through double acquisition channels;

the method includes the steps that measuring point parameters of each process system and equipment of an actual unit are obtained through sensors, after DCS process parameters of the actual unit are obtained through a DCS communication protocol, one or more special data acquisition interface machines are arranged in a production area of a power plant, each interface machine is in communication connection with a field equipment sensor and is also in communication connection with the DCS system, one or more acquisition programs (the acquisition programs are divided into a sensor data acquisition program and a DCS acquisition program according to different acquisition objects) are deployed in the interface machines to acquire different equipment sensor data and DCS data respectively, wherein the sensors are acquired through corresponding industrial protocols and interfaces, and for the DCS system, the DCS system data are acquired through corresponding communication protocols and interfaces of the DCS system, for example: for siemens DCS system, UDP protocol is used for DCS data acquisition.

According to the scale of the collected data and the source of the collected data, a plurality of interface machines can be used, for example, a special interface machine for DCS data collection and a special interface machine for sensor data collection. Of course, considering the diversity of the acquisition procedure, if the requirement on performance is not high, the acquisition procedure can be put into an interface machine.

In one possible implementation, the field device sensor is provided with two sensor data sending modules for sending the same sensor data at the same time; the DCS is provided with two DCS data sending modules for sending the same DCS data at the same time; the data acquisition interface machine is provided with a sensor data acquisition program and a DCS system data acquisition program, and the method further comprises the following steps:

the sensor data acquisition program receives data from the two sensor data sending modules, compares whether the received data has repeated data, deletes one data of the two repeated data and caches the other data of the two repeated data under the condition of judging that the two repeated data exist;

The scheme of using double acquisition channels in the method eliminates the influence of single-channel faults, simultaneously receives data in the continuous channels, and does not influence the correct and stable reception of the data due to the fault of any one channel. The fault of any channel is transparent to the user, the real-time data acquisition of the user is not influenced, the effective communication between the systems can be ensured, and the real-time performance of the communication can be ensured.

In one possible implementation, the method further includes:

the DCS system data acquisition program compares the data received in the current period with the data received in the previous period in the current period, and judges whether repeated data exist.

In a possible implementation manner, the data sent by the data acquisition interface machine to the distributed message system Kafka at least includes a measurement point code, an acquisition timestamp, data quality, and a measurement point value, where the measurement point code is used to uniquely identify each measurement point of the nuclear power plant, the acquisition timestamp is used to indicate the time for the data acquisition interface machine to acquire data, the data quality is used to indicate the quality of the data acquired by the acquisition interface machine, and the measurement point value is used to indicate a value acquired by a measurement point of the nuclear power plant, and the method further includes:

under the condition that the measuring point codes, the collecting time stamps and the measuring point values of two received data are the same, the sensor data collecting program judges that the two data are repeated data;

Specifically, the DCS data collection is taken as an example, fig. 2 is a schematic diagram of the DCS data collection shown in an exemplary embodiment, and fig. 3 is a flowchart of the DCS data collection shown in an exemplary embodiment. As shown in fig. 2 and fig. 3, during actual operation, the XU module inside the unit is connected to the application program XUtoUDP, the XUtoUDP is used to receive a message sent by the XU, the XUtoUDP processes the received message and then sends the message to the outside of the unit through UDP in a data packet manner, and the DCS data collection program is used to receive data sent by the XUtoUDP. In order to ensure the correctness and stability of data acquisition, a data acquisition program needs to simultaneously receive point data sent by two XUtoUDP programs, delete one data of the two repeated data and cache the other data of the two repeated data, and when one channel fails, any channel fails, so that the correctness and the stability of the data are not influenced.

In this embodiment, the acquired data may be sent to the outside through a data packet format, for example, the data packet may be a binary data packet format or a JSON format, and the data packet at least includes information of four dimensions, that is, a measurement point code, an acquisition timestamp, data quality, and a measurement point value.

The measuring point codes are measuring point IDs named according to a nuclear power plant measuring point code naming rule, the IDs are unique in the whole power plant, and in order to guarantee uniqueness and distinguishability of the measuring point IDs, in the embodiment, the complete nuclear power industry internet platform time sequence data measuring point codes (measuring point uniform codes) are composed of three parts. The first part is a power plant code consisting of two-bit characters, the second part is a unit code, the third part is a power plant original measuring point number, and the measuring point code is shown in figure 4.

In order to ensure the time precision, measuring a point timestamp as a millisecond timestamp; the data quality is a mark representing the quality of the measured point, generally 0 and 1,0 can represent good value, and 1 represents problem; the point measurement value is the value collected by the sensor when the point is at the corresponding timestamp. In the process, an isolation network gate is not used, and after data are acquired, the data can be directly sent out in a specific data packet through simple processing so as to improve the real-time performance of the data.

And S2, distributing data, wherein the distributed message system Kafka, the data twin system and the time sequence database are deployed in a production area, a time sequence data writing program of the time sequence database and the digital twin system subscribe to corresponding topics in the Kafka respectively, measuring point data are temporarily cached after entering the corresponding topics in the Kafka, and the time sequence data writing program and the digital twin system judge that if data arrive, consumption is carried out in time.

The nuclear power plant is divided into a generation area and a management area, in order to ensure data safety, data in the production area can only flow to the management area in a one-way mode, and data in the management area cannot flow to the production area; in the application, in order to transmit the unit data to the digital twin system in real time, the distributed message system Kafka, the data twin system and the time sequence database are deployed in a production area. To enable bidirectional flow of data therebetween. Meanwhile, in order to avoid that a data distribution program cannot be processed in time due to a large amount of acquisition and transmission, the distributed message system Kafka is used in the embodiment to perform peak clipping and caching on a large amount of sensor data and DCS data, so that the stable data processing process is ensured, and the phenomenon that a large amount of data simultaneously rush into a time sequence database and a digital twin system to cause data accumulation and transmission link congestion is avoided.

Data distribution is carried out in two ways, one way can be directly stored, and the other way can be directly sent to the digital twin system, so that reduction of data real-time performance caused by intermediate storage and access is avoided. The survey point data is temporarily buffered after entering the corresponding Topic in Kafka. On the one hand, the time-series data writer subscribes to the corresponding Topic in Kafka, and if data arrives, the time-series data writer consumes the data in time, processes the data into a format required by the storage of the time-series database, and then writes the data into the time-series database. On the other hand, the digital twin system needs to use real-time data of an actual unit to automatically debug the system state and self-optimize the model, and also subscribes to corresponding Topic in Kafka, so that the real-time data can enter the digital twin system in time and is used for further processing and analysis.

S3, storing data, namely storing the measuring point data subscribed by the time sequence data writing program into a time sequence database for classification and partition storage; the measuring point data subscribed by the digital twin system enters the digital twin system in real time for further processing and analysis;

in step S3, the measuring point data subscribed by the digital twin system can enter the digital twin system in real time and without damage, for automatic debugging of system state and self-optimization of the model, and the measuring point data subscribed by the time sequence data writing program is stored in the time sequence database for classification and partition storage.

In this embodiment S3, the classifying and partition storing in the time sequence database includes the following steps:

s31, encoding the measuring point data to obtain encoded data;

the measuring point data of the power plant is usually in a JSON format, and comprises necessary information such as measuring point codes, collecting timestamps, data quality and measuring point values, and also comprises various symbols such as space characters, brackets, middle brackets and the like. The data amount of the I/O operation can be reduced in both writing data and reading data, thereby improving performance.

The coding modes are different according to different types of the used databases and different sources of the measuring point data.

If the database is an IoTDB database and the measuring point data are low-frequency measuring point measuring values, carrying out data coding on a DOUBLE-precision floating point value (DOUBLE), a single-precision floating point value (FLOAT) and an integer value (INT 32) in a GORILLA coding mode, and carrying out data coding on a Boolean value (BOOLEAN) in a run length coding (RLE) mode; if the database is an IoTDB database, the measuring point data is a high-frequency measuring point, a PLAIN coding mode is adopted for data coding, all the measured value data in one data packet are stored in a binary mode, and the type of the measured value data is a character string value (TEXT).

In the present disclosure, the online measurement data of the nuclear power plant may also be time series data, and the online measurement data may include high frequency data and low frequency data, wherein the high frequency data is generally data generated by a high frequency sensor, the high frequency sensor often generates no less than 100 numbers per second (for example, the high frequency sensor generates thousands of numbers or tens of thousands of numbers per second), the low frequency data is generally data generated by a low frequency sensor, and the low frequency sensor often generates less than 100 numbers per second (for example, the low frequency sensor generates several or several tens of numbers per second).

S32, storing the coded data according to a preset storage area and a preset storage structure, wherein the preset storage area is set according to the fact that the measuring point data of the same power plant, the same unit and the same process system are stored in the same storage area;

in this embodiment, before the time sequence database stores the time sequence data, the time sequence data needs to be classified and stored in a partitioned manner, so as to improve the storage and use efficiency. By classifying and storing the coded data in partitions, the storage and access of the measuring point data among the partitions are not interfered mutually.

When the adopted time sequence databases are different, the specific setting of the preset storage areas is different, and no matter which database is used, the measuring point data of the same power plant, the same unit and the same process system can be stored in the same storage area.

For example, when the time series database is the IoTDB time series data, the setting of the path corresponding to the storage area may be: the IoTDB time sequence data is divided into a plurality of groups, wherein each group is a unit of a power plant, and each group is a unit of a power plant. For example, if the power plant is a three-door nuclear power plant and the unit is a unit No. 1, the storage group corresponding to the data generated by the unit No. 1 should be set as: root.cnnp.zs.01; all data of the measuring point are stored in a storage area corresponding to a CCS system of the three-door nuclear power unit No. 1, and all queries for the measuring point data are limited to the storage area. If the data of the measuring point 10CWS-TE306A comes from the number 1 unit of the three-door nuclear power, the path for storing the time sequence data is as follows: root, CNNP, ZS, 01, ZS _01, u 10CWS-TE306A, wherein CNNP, ZS, 01 is a storage group, 01, u 10CWS-TE306A is a measuring point uniform code, and an IoTDB time sequence data model is shown in FIG. 5.

When the time sequence database is an infiluxdb database, the naming of database should correspond to the power plant, and the measurement element is named in a mode of "power plant code _ unit code", for example: if the measurement data is the time sequence data from the No. 1 unit of the Qinshan factory, the measurement used for storing the time sequence data is named as: q1_ UNIT1, thereby determining the data source, and storing the measuring point data under the same power plant, the same UNIT and the same process system into the same storage area.

The foregoing description of the embodiments of the present disclosure has been presented for purposes of illustration and description, but is not intended to be exhaustive or limited to the embodiments disclosed. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The terms used herein were chosen in order to best explain the principles of the embodiments, the practical application, or technical improvements to the techniques in the marketplace, or to enable others of ordinary skill in the art to understand the embodiments disclosed herein.

Claims

1. A real-time transmission and storage method for nuclear power plant data is characterized by comprising the following steps:

2. The method of claim 1, wherein the field device sensor is provided with two sensor data transmission modules for transmitting the same sensor data simultaneously; the DCS is provided with two DCS data sending modules for sending the same DCS data at the same time; the data acquisition interface machine is provided with a sensor data acquisition program and a DCS system data acquisition program, and the method also comprises the following steps:

3. The method of claim 2, further comprising:

the sensor data acquisition program compares the data received in the current period with the data received in the previous period in the current period, and judges whether repeated data exist or not;

4. The method as claimed in claim 1, wherein the data sent by the data acquisition interface machine to the distributed messaging system Kafka at least includes a measurement point code, an acquisition time stamp, data quality and measurement point values, wherein the measurement point code is used for uniquely identifying each measurement point of the nuclear power plant, the acquisition time stamp is used for representing the time when the data acquisition interface machine acquires the data, the data quality is used for representing the quality of the data acquired by the acquisition interface machine, and the measurement point values are used for representing the values acquired by the measurement points of the nuclear power plant.

5. The method as claimed in claim 2, wherein the data sent by the data collection interface machine to the distributed messaging system Kafka includes at least a station code, a collection timestamp, a data quality, and a station value, and the method further comprises:

6. The method of claim 4, wherein the station code comprises a first portion, a second portion and a third portion, wherein the first portion is a nuclear plant code, the second portion is a unit code, and the third portion is a nuclear plant original station number.

7. The method of claim 1, further comprising:

the time sequence database encodes the received data to obtain encoded data;

8. The method of claim 7, wherein the time series database encodes the received data, comprising:

under the condition that the type of the time sequence database is an IoTDB database, if the received data is low-frequency measuring point data, a GORILLA coding mode is adopted to code a double-precision floating point value, a single-precision floating point value and an integer value, and a run coding mode is adopted to code a Boolean value.

9. The method of claim 7, wherein the time series database encodes the received data, and further comprising:

10. The method of claim 1, further comprising: and the data acquisition interface machine transmits the acquired data to the distributed message system in a data packet mode.