CN113568397A - Self-detection system and self-detection method for turbine monitoring instrument - Google Patents

Abstract

The invention discloses a steam turbine monitoring instrument self-building system and a self-checking method, wherein the system reads monitoring instrument implementation data from a power plant real-time database through a self-checking database server, combines a reasonable range of each monitoring instrument numerical value stored in the self-checking database server and a correlation and persistence data set among each detecting instrument data, performs self-checking on the instruments through calculation of a computer server from the range of the monitoring instruments, the range of the instrument numerical value set right under a specific working condition and the trend change of the instruments, and ensures the normal operation of the instruments, thereby improving the accuracy and the comprehensiveness of the instrument checking and increasing the safety and the reliability of the operation of a steam turbine of a power plant.

Description

Self-detection system and self-detection method for turbine monitoring instrument

Technical Field

The invention relates to a monitoring instrument self-checking system and a monitoring instrument self-checking method, in particular to a power plant steam turbine monitoring instrument self-checking system and a power plant steam turbine monitoring instrument self-checking method.

Background

The steam turbine is one of three big core devices of power plant, because the system is complicated, spare part is numerous, and the control system of power plant is because of reasons such as degree of automation high grade, consequently install measuring point that quantity is very much often, for example a 1000MW four-cylinder four-steam-exhaust steam turbine unit, just install 450 a plurality of monitoring instrument, these instruments mainly are used for monitoring equipment and circulating medium's temperature, flow, pressure, vibration, rotational speed etc. signal, are the basis of guaranteeing that the steam turbine is in controllable safe high-efficient state operation.

Similar to the safety check of the automobile during each starting, the steam turbine needs to check whether the monitoring instrument works normally during each starting. The next operation can be carried out only if the instrument works normally, otherwise, the unit is possibly in a dangerous operation state because the instrument cannot effectively monitor the equipment state.

The current power plant operating personnel is when the steam turbine starts at every turn to the inspection current situation of monitoring instrument, has relevant operating personnel to inspect the instrument of unit at power plant's centralized control room, and this operation all adopts artifical inspection mode at present, generally is through maintaining the unit for a period under certain rotational speed and the load state, looks over whether the instrument indicating value is in reasonable range for carry out the normality judgement of instrument state. However, in the current inspection, although some modern meters have a self-inspection function, the basis for performing self-inspection is mostly only to the strength or the existence of the self-signal, and only an alarm signal can be sent out when the self-signal is weak or disappears, and the correctness of the signal indication value and the rationality of the signal change trend cannot be judged, or cannot be judged (because the judgment is carried out based on the characteristics of the physical quantity monitored by the meter strictly, the monitoring meter cannot know the characteristics before leaving the factory); meanwhile, due to the fact that the number of the monitoring instruments of the power plant is large, operators cannot inspect all the instruments one by one, and therefore the inspection range can only cover a few important instruments. The instrument inspection work is required to be carried out when the unit is started every time, and when the unit is started in a warm state, the inspection work takes about 10 minutes due to short shutdown time of the unit; however, for cold start with longer downtime, especially after overhaul inspection of the unit, it usually takes 30 minutes or more because the meter is not used for a long time and special attention needs to be paid to whether the meter is normal.

In conclusion, the problems that the workload is large, the manual inspection efficiency is low, time is wasted, the inspection coverage is small and omission is prone to occur in the conventional instrument inspection work of the power plant can be known.

Disclosure of Invention

The invention aims at the situation, and particularly provides a method and a system for self-checking of a turbine monitoring instrument, which can greatly reduce the workload of operators and improve the accuracy and comprehensiveness of instrument checking:

a turbine monitoring instrument self-detection system comprises a monitoring instrument, a DCS system, a power plant real-time database, a self-detection database server, a calculation server, a data display server and a man-machine interaction terminal, wherein the monitoring instrument is distributed on a turbine to detect turbine operation parameters everywhere, the DCS system collects values of the monitoring instruments and transmits the values to the power plant real-time database for storage, the self-detection data server calls corresponding real-time monitoring data of the monitoring instruments from the power plant real-time database, the self-detection data server also stores preset reasonable ranges of the values of the monitoring instruments and correlation and persistence data sets of data pieces of the monitoring instruments, the calculation server analyzes and judges states of the monitoring instruments and sends analysis results to the data display server, and the data display server transmits the analysis result to the human-computer interaction terminal for browsing.

Further, the self-checking data server can only read relevant data from the real-time database of the power plant based on the secondary development port and cannot write in the real-time database of the power plant.

Further, the self-checking data server stores the implementation data of the monitoring instrument since the last starting time of the steam turbine, and the data storage leveling rate is within the range of 0.5s-2 min.

Furthermore, the reasonable range of the numerical values of the monitoring instruments comprises the range of the measuring ranges of the monitoring instruments and the reasonable clustering range of each monitoring instrument under the specific working condition of the steam turbine.

A steam turbine monitoring instrument self-detection method includes that a database collects values of all monitoring instruments of a steam turbine, a calculator is used for checking the values of the monitoring instruments in at least the following two aspects, and whether the collected values of all the monitoring instruments are in a reasonable value cluster range under a specific working condition that the rotating speed of the steam turbine is constant is checked; and when the rotating speed of the steam turbine changes, checking whether the correlation of the change trend among the data of each monitoring instrument is reasonable.

Furthermore, the check of the numerical value of the monitoring instrument through the calculator also comprises the check of whether the acquired numerical value of each monitoring instrument exceeds the range of the measuring range; the range of the meter value of each monitor is set manually by an operator, and if the monitor exceeds the range calculator, the name, the time point and the value of the monitor are recorded.

Furthermore, the specific working condition is a warm-up working condition or a low-load working condition, the steam turbine stays for a certain time under the working condition, and the numerical value clusters of the monitoring instruments under the working conditions are manually set by workers.

Furthermore, the relevance of the change trend of the monitoring instrument is determined by a data analysis program according to a reasonable operation data set by a worker, and the relevance principle and the principal component analysis are utilized to automatically learn and conclude the relevance of every two data of the monitoring instruments.

By adopting the system and the method, the measuring range of the monitoring instrument and the numerical cluster detection of the monitoring instrument under the working condition of a certain specific rotating speed or load of the steam turbine can be realized through a computer, and whether the trend change of each instrument meets the requirement can be checked through software learning, so that the accuracy and the comprehensiveness of the instrument check are improved, and the safety and the reliability of the operation of the steam turbine of the power plant are improved.

Drawings

FIG. 1 is a schematic diagram of the system of the present invention;

FIG. 2 is a diagram illustrating the variation of the rotational speed of the steam turbine during several start-up and stop processes within a time period;

FIG. 3 is a graph showing the temperature change of the steam in front of the high-pressure main valve, which is matched with the rotation speed change in FIG. 2;

FIG. 4 is a graph of wall temperature changes in the high pressure cylinder that match the speed changes in FIG. 2;

FIG. 5 is a graph of the main steam pressure variation matched to the speed variation of FIG. 2;

FIG. 6 is a graph of medium pressure superheater outlet pressure variation matching the speed variation of FIG. 2;

FIG. 7 is a graph of the main steam flow rate variation matched to the speed variation of FIG. 2.

Detailed Description

The turbine self-detection system shown in fig. 1 includes monitoring instruments, a DCS system, a power plant real-time database, a self-detection database server, a calculation server, a data display server and a display, wherein the monitoring instruments are distributed on the turbine and collect data of the turbine such as temperature, pressure, flow and rotation speed, the DCS system can collect values of the monitoring instruments for internal systems of the power plant and transmit the values to the power plant real-time database for storage, the self-detection data server can retrieve real-time monitoring data of the corresponding monitoring instruments from the power plant real-time database, the self-detection data server stores preset reasonable numerical ranges of the monitoring instruments (the ranges include ranges of the monitoring instruments and numerical value cluster ranges of the monitoring instruments under specific working conditions of the turbine) and data correlation and persistence data sets of the monitoring instruments, and the calculation server stores the data in the self-detection data server and retrieves the real-time numerical values from the power plant real-time database And analyzing and judging the state of the instrument and sending an analysis result to a data display server, and transmitting the analysis result to a human-computer interaction terminal for browsing by the data display server.

The process of collecting the numerical values of the monitoring instruments through the DCS and storing the numerical values into the real-time database of the power plant is the inherent system flow of the power plant, the real-time database of the power plant is used for storing the data of the monitoring instruments and also storing other important data of the power plant, and the real-time database of the power plant can only be read in consideration of the safety requirements of the power plant on data management, so that the self-checking data server is based on a secondary development interface and retrieves the monitoring data of the monitoring instruments by asking for the data reading permission from the real-time database of the power plant.

When the self-detection server reads data, the real-time data of the monitoring instrument from the latest starting moment is stored in the power plant real-time database, the data storage frequency is in the range of 0.5-2 min, the frequency cannot be too fast so as to avoid huge pressure on storage and calculation resources, and the frequency is not suggested to be too slow so as to avoid losing some important change characteristics; in addition, the server stores information such as preset range values of each meter, preset reasonable range values of meter indication values when the turbine speed is maintained at a certain speed or load working condition state (the running state can be a certain speed and load maintaining state when the meters are manually checked in the background art), preset correlation characteristic data sets among meter data, and the like, wherein correlation refers to the degree of correlation between two variables, and correlation analysis refers to analysis of two or more variable elements with correlation, so that the degree of correlation closeness between the two variable factors is measured. Correlation characteristics, such as positive correlation degree, negative correlation degree or uncorrelated characteristics, between every two monitoring instrument data can be obtained by observing a large amount of operation data of the unit shown in fig. 2-7 for correlation analysis.

The system is an important link for checking the numerical value of the monitoring instrument through a calculator, and the specific method is to ensure the accuracy of the instrument detection through three aspects of checking:

1. and checking whether the acquired meter values of each monitor exceed the range.

The specific method is to compare the indicating value of each detecting instrument to be checked with the measuring range value of each instrument pre-stored in the self-detecting server. The range of the measurement range can be specified and modified in the self-test server by the operator. The program can judge whether the indication value of the monitoring instrument is in the range of the range in real time, if the indication value exceeds the range, data information such as a time point, the size of the data value, the name of the instrument and the like in the process of exceeding the range can be recorded, and the data information is recorded as a log document and then fed back to the front-end display server to remind an operator. This check process is performed throughout the operation of the steam turbine.

2. And checking whether the acquired meter values of each monitor are in a reasonable numerical value cluster range under the specific working condition that the rotating speed of the steam turbine is constant.

The steam turbine is maintained under specific operating mode, and the steam turbine stops at a certain specific rotational speed and load this moment, carries out the inspection of each monitoring instrument numerical value cluster scope this moment, then does not carry out the inspection of monitoring instrument numerical value cluster scope under other circumstances.

The inspection is a static inspection, and only the accuracy and the reasonableness of the data of the monitoring instrument under certain specific working conditions are judged. The specific working condition may be a certain fixed rotation speed working condition of the turbine, such as a warming-up working condition of the turbine at 600r/min or 1800r/min, or a certain low load working condition of the turbine (once the turbine is loaded, the rotation speed is constant), such as a working condition with 5% initial load, under which the turbine stays for a certain time. Ideally, the above-mentioned working conditions can select the medium-low speed warm-up working condition and the working condition with initial load of the steam turbine, because the steam turbine itself also stays under these working conditions when operating, therefore, when selecting the working conditions, the inherent working conditions set in the operation process can be selected.

The reasonable value range of the monitoring instrument under the specific working condition can be preset in the self-checking database server by an operator, and can also be a value range which is automatically learned and defined by a program according to a running data set which is determined to be reasonable by the operator. The value range is based on statistical results and will be more reasonable as the number of operational data sets increases.

The specific judgment process is described as follows, assuming that the reasonable range of the indication value of a certain instrument under a certain specific working condition is [ R1, R2], when the indication value A of the monitoring instrument belongs to [ R1, R2], feeding back a signal of 'reasonable indication value of the instrument'; when A < R1, feeding back to a front-end signal that the instrument indication value is lower than the lower limit reasonable value; when A > R2, the signal is fed back to the front end that the instrument value is higher than the upper limit reasonable value.

As shown in fig. 2 to 7, with reference to the change of the rotational speed of the steam turbine in fig. 2, after the rotational speed is constant, the verticality of each monitoring instrument tends to be balanced and is jumping in a certain range, the cluster check is performed to check whether the numerical range of each monitoring instrument is reasonable in this state, and as long as the detected numerical value of the monitoring instrument is within the cluster range, it indicates that the monitoring instrument is working normally.

3. And checking whether the correlation of the variation trend among the data of the monitoring instruments is reasonable.

When the rotating speed of the steam turbine changes, the numerical values of the monitoring instruments also change in a trend manner, so that the checking process is only carried out on the change of the rotating speed of the steam turbine.

This check is to judge the reasonableness of the dynamic change characteristic of the monitoring meter. When the instrument is manually checked, the reasonability of the indication value of the monitoring instrument under certain specific working conditions is mainly judged, and the dynamic change characteristic of the unit in the process of reaching the specific working conditions from a static state cannot be identified and evaluated, so that the response hysteresis of the instrument cannot be effectively identified, however, if the monitoring instrument cannot timely and quickly capture the change characteristic data of the unit, and operation accidents may be caused for power plant equipment with fast response control requirements.

For the situation, the data analysis program can automatically learn and summarize pairwise correlation among the data of the monitoring instruments according to the operation data set which is determined to be reasonable by the operator (for example, a plurality of groups of unit operation data sets which are determined to be reasonable are input into the self-checking data server by the operator in advance), and the correlation is used as a reference data set. And when the deviation degree between the newly acquired correlation data and the reference data set exceeds C (threshold value), the reasonability of the dynamic response characteristic of the meter can be considered to be in a problem, and prompt information is sent to the front end.

When the criterion is used for judgment, a specific judgment process is described as follows, the calculation server gives a plurality of sets of reasonable unit operation data sets stored in the self-inspection data server by an operator to a data analysis program in advance, and after program analysis, pairwise correlation data sets of data of each monitoring instrument are summarized and recorded as [ Z ] (reference range) and stored in the self-inspection data server. When the program analyzes the data of the monitoring instrument in the process of the last startup, a correlation data set among the instruments can be obtained and is marked as [ Y ]. When the correlation characteristic value [ Yn ] between the indicating value of a certain monitoring instrument n and the indicating values of other instruments exceeds the range of the reference [ Zn ] and reaches the threshold value C, the reasonability of the dynamic response characteristic of the instrument is considered to be in a problem, and prompt information is sent to a data display server to promote operators to further investigate reasons. It should be noted that the correlation data set [ Z ] is based on statistical results, and the data set will tend to be more reasonable with the increase of reasonable operation data sets.

The trend monitoring basis is proposed according to the recurrence performance of each monitoring instrument in fig. 2 to fig. 7, and it can be seen through observing fig. 2 to fig. 7 that the trend of each monitoring instrument changes along with the change of the rotating speed of the steam turbine, so that the starting point of the trend change is judged according to the rotating speed of the steam turbine as the reference in the checking process, and the correlation check is not performed on the rotating speed monitoring instrument of the steam turbine, not only because the rotating speed monitoring instrument of the steam turbine is used as the reference, but also because the accuracy and the reliability of the rotating speed monitoring instrument of the steam turbine are high.

The above three inspections are not required to be all carried out, because if the trend inspection and the cluster inspection under the specific working condition are both in a normal state, the monitoring instrument is always in a reasonable range, and the trend inspection and the cluster inspection under the specific working condition cover the whole operation process of the steam turbine, the range inspection can be selectively adopted or cancelled as an integrity inspection means. It should be noted that a perfect situation is that the cluster range detection can be performed under the condition that the rotating speed of the working state of the steam turbine is constant, and under the condition, the range detection is not required to be adopted at all, because the steam turbine is in constant rotating speed operation under the warm-up working condition or the low-load working condition except the above, the steam turbine is also in constant rotating speed in normal operation, and the load is changed along with the requirement of the generated energy, the detection of the cluster range is set according to the working condition of the steam turbine, the working condition state of the load change in the normal operation of the steam turbine is more complicated, the set cluster range quantity is too large, the larger data storage requirement can be generated on the system, and the logging by an operator is not facilitated, therefore, the self-detection purpose of the monitoring instrument can be achieved only by checking the cluster range of a plurality of specific and necessary working conditions, and the cluster range detection under all the working conditions is not required, under other working conditions (such as normal operation, the rotation speed of the steam turbine is not changed and the load is changed according to the change of the generated energy), the normal operation of each monitoring instrument and the protection of the steam turbine set need to be ensured through the process range inspection.

Claims (8)

1. The utility model provides a steam turbine monitoring instrument self-detecting system, including monitoring instrument, DCS system, the real-time database of power plant, its characterized in that: also comprises a self-checking database server, a calculation server, a data display server and a man-machine interaction terminal, the monitoring instruments are distributed on each part of the steam turbine to detect the operating parameters of the steam turbine, the DCS system collects the numerical values of the monitoring instruments and transmits the numerical values to the real-time database of the power plant for storage, the self-checking data server calls corresponding monitoring instrument real-time monitoring data from the power plant real-time database, the self-detection data server is also stored with preset reasonable range of each monitoring instrument value and correlation and continuity data set among each monitoring instrument data, the computing server analyzes and judges the state of the instrument according to the data stored in the self-checking data server and the real-time data called from the power plant real-time database, and sends the analysis result to the data display server, and the data display server transmits the analysis result to the human-computer interaction terminal for browsing.

2. The turbine monitoring instrument self-test system of claim 1, wherein: the self-checking data server can only read related data from the real-time database of the power plant based on the secondary development port and cannot write in the real-time database of the power plant.

3. The turbine monitoring instrument self-test system according to claim 1 or 2, wherein: the self-checking data server stores the implementation data of the monitoring instrument since the last starting moment of the steam turbine, and the data storage average rate is in the range of 0.5s-2 min.

4. The turbine monitoring instrument self-test system of claim 1, wherein: the reasonable range of the numerical values of the monitoring instruments comprises the range of the measuring range of each monitoring instrument and the reasonable clustering range of each monitoring instrument under the specific working condition of the steam turbine.

5. A self-detection method of a turbine monitoring instrument is characterized by comprising the following steps: the method comprises the following steps that a database collects values of monitoring instruments of a steam turbine, the values of the monitoring instruments are checked through a calculator, and whether the collected values of the monitoring instruments are in a reasonable value cluster range under a specific working condition that the rotating speed of the steam turbine is constant is checked; and when the rotating speed of the steam turbine changes, checking whether the correlation of the change trend among the data of each monitoring instrument is reasonable.

6. The turbine monitoring instrument self-test method according to claim 5, wherein: the checking of the numerical value of the monitoring instrument through the calculator also comprises checking whether the collected numerical value of each monitoring instrument exceeds the range of measuring range; the range of the meter value of each monitor is set manually by an operator, and if the monitor exceeds the range calculator, the name, the time point and the value of the monitor are recorded.

7. The turbine monitoring instrument self-test method according to claim 5, wherein: the specific working condition is a warm-up working condition or a low-load working condition, the turbine stays for a certain time under the working condition, and numerical value clusters of the monitoring instruments under the working conditions are manually set by workers.

8. The turbine monitoring instrument self-test method according to claim 5, wherein: the correlation of the change trend of the monitoring instrument is determined by a data analysis program according to a reasonable operation data set by a worker, and the correlation principle and the principal component analysis are utilized to automatically learn and conclude the pairwise correlation among the data of each monitoring instrument.

Images