CN112946469B - Monitoring method for self-adaptive dynamic alarm threshold of hydraulic generator - Google Patents
Monitoring method for self-adaptive dynamic alarm threshold of hydraulic generator Download PDFInfo
- Publication number
- CN112946469B CN112946469B CN202011357305.XA CN202011357305A CN112946469B CN 112946469 B CN112946469 B CN 112946469B CN 202011357305 A CN202011357305 A CN 202011357305A CN 112946469 B CN112946469 B CN 112946469B
- Authority
- CN
- China
- Prior art keywords
- monitoring
- data
- index
- monitoring index
- working condition
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000012544 monitoring process Methods 0.000 title claims abstract description 147
- 238000000034 method Methods 0.000 title claims abstract description 25
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims abstract description 24
- 230000011218 segmentation Effects 0.000 claims abstract description 9
- 238000004140 cleaning Methods 0.000 claims abstract description 6
- 238000012216 screening Methods 0.000 claims abstract description 6
- 230000000007 visual effect Effects 0.000 claims description 11
- 230000003044 adaptive effect Effects 0.000 claims description 7
- 238000005259 measurement Methods 0.000 claims description 4
- 230000002159 abnormal effect Effects 0.000 abstract description 6
- 238000011156 evaluation Methods 0.000 abstract description 5
- 238000001514 detection method Methods 0.000 abstract description 3
- 238000004458 analytical method Methods 0.000 description 10
- 230000015556 catabolic process Effects 0.000 description 9
- 238000006731 degradation reaction Methods 0.000 description 9
- 238000004364 calculation method Methods 0.000 description 7
- 238000009825 accumulation Methods 0.000 description 2
- 230000032683 aging Effects 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 2
- 238000007796 conventional method Methods 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 230000005284 excitation Effects 0.000 description 2
- 230000005856 abnormality Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/34—Testing dynamo-electric machines
- G01R31/343—Testing dynamo-electric machines in operation
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Testing And Monitoring For Control Systems (AREA)
Abstract
The invention relates to a monitoring method for a self-adaptive dynamic alarm threshold value of a hydraulic generator, and belongs to the technical field of hydraulic generator state evaluation and detection. The method comprises the following steps: collecting data; acquiring working condition and monitoring index data of the hydraulic generator; data screening; working condition segmentation; the data is divided into a sample data set and a monitoring data set; data cleaning; calculating upper and lower thresholds; calculating an out-of-limit ratio; monitoring and visually displaying. According to the method, the threshold value of the hydraulic generator under different operation conditions, namely the area with normal monitoring indexes of the unit, is rapidly calculated by using state monitoring time sequence data such as the active power, the water head and the vibration swing degree of the hydraulic generator, the area can be adaptively generated according to the operation condition of the unit, a technical reference is provided for monitoring whether each parameter of the hydraulic generator is deteriorated, and the hydraulic generator can timely carry out alarming and reminding when abnormal occurs, so that the expansion of accidents is avoided. The method has the characteristics of simple operation and strong universality, and is easy to popularize and apply.
Description
Technical Field
The invention belongs to the technical field of state evaluation and detection of hydraulic generators, and particularly relates to a monitoring method for a self-adaptive dynamic alarm threshold of a hydraulic generator.
Background
In the state evaluation, trend analysis and fault early warning of the hydraulic generator, an important link is to judge whether the current measured value deviates from the normal value or has a further degradation trend. According to the conventional method, a certain alarm limit value is set for the monitoring index, the change rate of the monitoring index is calculated, and when the monitoring value is higher than the alarm limit value or the change rate exceeds a set value, the abnormal state is judged, and an alarm is output. In the actual running process of the hydroelectric generating set, important monitoring indexes such as host vibration, swing degree and temperature are affected by various factors, the fluctuation trend is presented, some fluctuation is fluctuation caused by normal working conditions, some fluctuation is abnormal fluctuation, and fluctuation areas of the monitoring indexes are different under different working conditions.
When the hydro-generator is in an on-off process and an unstable operation region, various indexes are unstable, the fluctuation range is large, generally speaking, when the active power of the unit is 70% of rated power, various monitoring indexes are relatively stable, and the load section data are selected for evaluation, so that the development change trend of the reaction equipment is more favorably compared. However, even in a stable operation area, each monitoring index of the hydro-generator set is affected by the working conditions such as water head, exciting current and the like, and fluctuates up and down, and the change rate of directly collected data cannot reflect the degradation rate of equipment.
In addition, various working conditions and monitoring indexes have relevance to different degrees, manual analysis is adopted, the working conditions with strong relevance can be distinguished according to experience, but the relevance analysis is carried out on various monitoring indexes of the hydraulic generator one by one, so that degradation analysis is further carried out, a large amount of manual intervention is required, the universality is poor, and informatization is not facilitated. Therefore, how to overcome the defects of the prior art and truly identify the situation beyond the normal range is a problem to be solved in the technical field of the state evaluation and detection of the current hydraulic generators.
Disclosure of Invention
The invention aims to solve the defects of the prior art and provides a monitoring method for a self-adaptive dynamic alarm threshold value of a hydraulic generator. The method is simple to operate, high in universality and easy to popularize and apply.
In order to achieve the above purpose, the technical scheme adopted by the invention is as follows:
A method for monitoring an adaptive dynamic alert threshold for a hydro-generator, comprising the steps of:
(1) Acquiring system acquisition data, comprising: time, generator active power, unit water head, exciting current and monitoring index data;
(2) Data screening: selecting data with active power more than or equal to 70% of rated power;
(3) Working condition segmentation: segmenting the selected data according to three working condition indexes of active power, a unit water head and exciting current, wherein each working condition index is equally divided into N segments, and marking the segments;
(4) Data is divided into sample data sets, monitoring data sets: after the time period of the monitoring data set is determined, selecting data of one year above the starting time of the monitoring data set as a sample data set;
(5) Performing data cleaning on the sample data set to obtain an effective sample data set;
(6) Calculating the effective sample data set to obtain upper and lower threshold limits:
(7) Writing the upper and lower threshold limits to the monitored dataset:
2 rows of data columns are added to the monitoring data set, and according to the working condition marks, each row of data is filled with the upper limit and the lower limit of the monitoring index threshold;
(8) Calculating an out-of-limit ratio for the monitored dataset:
Monitoring index out-of-limit ratio = monitoring index actual measurement/monitoring index threshold upper limit 100%;
(9) Monitoring and visual display:
91 When the out-of-limit ratio is more than 100%, starting timing;
92 When the out-of-limit ratio is more than 100%, outputting an alarm and continuing to monitor when the timing is more than the set time;
93 When the out-of-limit ratio is less than 100%, or when the out-of-limit ratio is more than 100% and does not exceed the set time, continuing monitoring;
94 A visual presentation).
Further, it is preferable that each of the condition indexes is equally divided into 10 segments.
Further, it is preferable that the specific method of step (5) is:
51 Any one of three working condition indexes of active power, unit water head and exciting current is combined into one working condition, and after each of the three working condition indexes is arranged and combined in N sections, N 3 working conditions are shared;
52 Calculating a first quartile and a third quartile of the sample dataset monitoring index under each working condition; the first quartile is denoted as monitor index_ FirstQuartile and the third quartile is denoted as: monitoring index_ ThirdQuartile;
53 Calculating an outlier upper bound:
monitoring index outlier upper limit = monitoring index_ ThirdQuartile +1.5 (monitoring index_ ThirdQuartile-monitoring index_ FirstQuartile);
54 Calculating an outlier lower bound:
Monitoring index outlier lower bound = monitoring index_ FirstQuartile-1.5 x (monitoring index_ ThirdQuartile-monitoring index_ FirstQuartile)
55 Deleting data in the sample data set above the upper outlier limit or below the lower outlier limit to obtain a valid sample data set.
Further, it is preferable that the specific method of step (6) is:
61 Calculating the maximum value, the minimum value and the arithmetic mean of monitoring indexes in the effective sample data set under each working condition; wherein, the maximum value of the monitoring index is marked as a monitoring index_max, the minimum value is marked as a monitoring index_min, and the arithmetic average number is marked as: monitoring index_mean;
62 Calculating upper and lower threshold limits for each working condition:
Monitoring index threshold upper limit=monitoring index_max+0.25 monitoring index_mean;
Monitor indicator threshold lower limit=monitor indicator_min-0.25 monitor indicator_mean.
According to the method, the historical data of each monitoring index time sequence of the hydraulic generator are used for adaptively calculating the dynamic threshold value of the monitoring index under each working condition, and the method can be used for various data analysis tools or programming tools to calculate the adaptive dynamic threshold value of the hydraulic generator.
The invention is based on the following principle:
When the unit operates normally, the monitoring index of the unit has certain fluctuation, and when the unit is abnormal, the monitoring index has trend fluctuation. With the conventional method, it is difficult to distinguish whether the fluctuation of the monitoring index is caused by the change of the working condition or the abnormality exists in the equipment.
(1) The self-adaptive threshold calculated by the method is generated in historical data of the unit, is dynamic according to working condition changes, reflects the centralized trend of the monitoring index, and is more accurate along with accumulation of the historical data.
(2) When the equipment operates normally, the monitoring index fluctuates in a certain interval, and the threshold value reflects the normal upper limit and the normal lower limit of the interval. And the calculation of the threshold value is derived from statistical methods and expertise.
(3) The computer system can generate certain outlier data in links of data acquisition, storage and the like, and can influence a calculation result.
(4) If fluctuation is caused by working condition change, the calculated out-of-limit ratio is lower than 100%, and if fluctuation is caused by equipment degradation, the calculated out-of-limit ratio is higher than 100%. The threshold ratio is higher than 100% of the actual reaction equipment degradation.
Compared with the prior art, the invention has the beneficial effects that:
1. The current monitoring system and the online monitoring system of the hydroelectric generating set monitor important indexes, set alarm limit values for the indexes, set limit values are higher in order to avoid false alarms, and serious faults possibly occur when the set reaches the high alarm limit values. The threshold value set by the invention is a dynamic limit value, and the unit is different in each index fluctuation range under different working conditions, and the degradation condition of the monitoring index is truly reflected when the threshold value is exceeded.
2. The data stored in the data center of each hydropower station come from each system, and in the links of data acquisition, collection and the like, abnormal data or noise can appear, and the threshold value is calculated by using all data and can be influenced by abnormal values.
3. The invention corrects the upper and lower limits of the threshold value, leaves a certain margin, and avoids false alarms to a greater extent.
4. The hydroelectric generating set changes the state of the set along with the prolonging of the operation period, belongs to normal aging, and the influence of the normal aging of the set is not considered in the original technical means. The invention adopts self-adaption, if the power plant operates to consider that certain degradation trends are normal degradation, the power plant does not need to be processed, and the threshold range can be automatically corrected along with data accumulation, so that human intervention is not needed.
5. Various working conditions and monitoring indexes have relevance to different degrees, the working conditions with strong relevance can be distinguished according to experience by adopting manual analysis in the prior art, but the relevance analysis is carried out on various monitoring indexes of the hydraulic generator one by one, so that degradation analysis is further carried out, a large amount of manual intervention is required, the universality is poor, and informatization is not facilitated. According to the invention, no manual relevance test is needed, all vibration swing degree monitoring indexes are subjected to working condition segmentation by using active power, a unit water head and exciting current, and the temperature quantity and the environmental temperature are good in universality.
6. In the prior art, the line crossing condition and the change rate of the monitoring index are calculated independently, and the standardization processing is not performed, so that the follow-up calculation and the visual display are not facilitated. After the invention is standardized, a plurality of correlative monitoring indexes can be displayed on the same interface.
Drawings
FIG. 1 is a flow chart of a method of monitoring an adaptive dynamic alert threshold for a hydro-generator of the present invention;
FIG. 2 is a visual display of measured values and threshold values;
Fig. 3 is a visual display of a plurality of monitor indicator out-of-limit ratios.
Detailed Description
The present invention will be described in further detail with reference to examples.
It will be appreciated by those skilled in the art that the following examples are illustrative of the present invention and should not be construed as limiting the scope of the invention. The specific techniques or conditions are not identified in the examples and are performed according to techniques or conditions described in the literature in this field or according to the product specifications. The materials or equipment used are conventional products available from commercial sources, not identified to the manufacturer.
Example 1
A method for monitoring an adaptive dynamic alert threshold for a hydro-generator, comprising the steps of:
(1) Acquiring system acquisition data, comprising: time, generator active power, unit water head, exciting current and monitoring index data;
(2) Data screening: selecting data with active power more than or equal to 70% of rated power;
(3) Working condition segmentation: segmenting the selected data according to three working condition indexes of active power, a unit water head and exciting current, wherein each working condition index is equally divided into N segments, and marking the segments;
(4) Data is divided into sample data sets, monitoring data sets: after the time period of the monitoring data set is determined, selecting data of one year above the starting time of the monitoring data set as a sample data set;
(5) Performing data cleaning on the sample data set to obtain an effective sample data set;
(6) Calculating the effective sample data set to obtain upper and lower threshold limits:
(7) Writing the upper and lower threshold limits to the monitored dataset:
2 rows of data columns are added to the monitoring data set, and according to the working condition marks, each row of data is filled with the upper limit and the lower limit of the monitoring index threshold;
(8) Calculating an out-of-limit ratio for the monitored dataset:
Monitoring index out-of-limit ratio = monitoring index actual measurement/monitoring index threshold upper limit 100%;
(9) Monitoring and visual display:
91 When the out-of-limit ratio is more than 100%, starting timing;
92 When the out-of-limit ratio is more than 100%, outputting an alarm and continuing to monitor when the timing is more than the set time;
93 When the out-of-limit ratio is less than 100%, or when the out-of-limit ratio is more than 100% and does not exceed the set time, continuing monitoring;
94 A visual presentation).
Example 2
A method for monitoring an adaptive dynamic alert threshold for a hydro-generator, comprising the steps of:
(1) Acquiring computer system acquisition data, comprising: time, generator active power, unit water head, exciting current and monitoring index data, and the time is required to be accurate to minutes or more.
(2) Data screening: and selecting data with active power being more than or equal to 70% of rated power, and deleting the rest data.
(3) Working condition segmentation: the data are segmented according to three working condition indexes of active power, a water head of a unit and exciting current, each working condition index is divided into 10 segments averagely, and segmented marks are marked. If the active power data column is newly added, writing a segmentation mark, such as [1, 2. ], 10]; adding a unit head segment data column, and writing segment marks such as [1, 2.. The number is 10]; the field current segment data column is added and the segment flag is written, for example [1, 2..10 ]. Each segment label may use letters or numbers, but is not repeated.
(4) Data is divided into sample data sets, monitoring data sets: the sample data set is used to calculate a threshold value and the monitoring data set is used for state assessment or trend analysis. And after the time period of the monitoring data set is determined, selecting the data of one year from the starting time of the monitoring data set as a sample data set.
(5) Data cleaning is carried out on the sample data set, and an effective sample data set is obtained:
51 Any one of three working condition indexes of active power, unit water head and exciting current is combined into a working condition, and after each working condition index is arranged and combined in 10 sections, 1000 working conditions are combined, such as the combination of active power, unit water head and exciting current: [50,10,90] is a working condition.
52 Calculating the first quartile of the sample dataset monitoring indicator for each condition (noted: monitor index_ FirstQuartile), the third quartile (noted: monitor index_ ThirdQuartile);
53 Calculating an outlier upper bound:
monitoring index outlier upper limit = monitoring index_ ThirdQuartile +1.5 (monitoring index_ ThirdQuartile-monitoring index_ FirstQuartile)
54 Calculating an outlier lower bound:
Monitoring index outlier lower bound = monitoring index_ FirstQuartile-1.5 x (monitoring index_ ThirdQuartile-monitoring index_ FirstQuartile)
55 Deleting data in the sample data set above the upper outlier limit or below the lower outlier limit to obtain a valid sample data set.
(6) Calculating the effective sample data set to obtain upper and lower threshold limits:
61 Calculating the maximum value of the monitoring index in the effective sample data set under each working condition (marked as: monitor index_max), minimum value (noted as: monitor index_min), arithmetic mean (noted as: monitoring index_mean);
62 Calculating upper and lower threshold limits for each working condition:
Monitoring index threshold upper limit=monitoring index_max+0.25 monitoring index_mean;
monitoring index threshold lower limit = monitoring index_min-0.25 monitoring index_mean;
(7) Writing the upper and lower threshold limits to the monitored dataset:
The monitoring data set is added with 2 data columns, and each row of data is filled with upper and lower threshold limits according to the working condition marks.
(8) Calculating an out-of-limit ratio for the monitored dataset:
Monitoring index out-of-limit ratio = monitoring index actual measurement/monitoring index threshold upper limit 100%.
(9) Monitoring and visually displaying.
91 When the out-of-limit ratio is greater than 100%, start timing.
92 When the out-of-limit ratio is greater than 100%, the timing is greater than the set time, an alarm is output and monitoring is continued.
93 If the out-of-limit ratio is less than 100%, or if the out-of-limit ratio is greater than 100% and not more than the set time, continuing the monitoring.
94 A visual presentation).
Application instance
The present example monitors trend analysis and status of rack vibration on machine number 1. The machine 1 collects time series data of the upper frame plus the horizontal vibration peak value of the X direction, the water head of the machine set, the active power and the exciting current. The rated active power of the unit is 700MW, the active power fluctuates between 0MW and 719MW during normal operation, the water head fluctuates between 156 m and 235m, and the exciting current fluctuates between 1656A and 3264A. The method evaluates the vibration condition of the frame on 1 month to 10 months in 2020, and the implementation steps of the self-adaptive dynamic threshold are as follows:
Step (1), acquiring computer system acquisition data, comprising: time, active power of the No. 1 unit, exciting current of the No. 1 unit, water head of the No. 1 unit, and monitoring data of horizontal vibration peak and peak of the X-direction of the upper frame of the No. 1 unit are shown in a table 1.
When the fluctuation range of the data is required to be more than 5%, the data acquisition equipment needs to acquire and store the data, the time coordinates are consistent, and the time is accurate to be more than minutes.
TABLE 1
Step (2), data screening: selecting data (70% rated active power) with active power of No. 1 unit not less than 490MW, and deleting the rest data;
Step (3) working condition segmentation: the data are segmented according to three working condition indexes of active power, a water head of a unit and exciting current, each working condition index is divided into 10 segments averagely, and segmented marks are marked.
After segmentation, a column of active power_bin named as "No. 1 unit active power_bin" is added, the data is segmented and marked, and a computer randomly generates a non-repeated numerical mark to fill the column, in this case [1,2,3, 4..10 ]. Such as: and when the active power is 490MW, the active power_BIN of the No. 1 unit is marked as 1.
After the water head range of the unit is 156-235 m, a row of unit water head_BIN named as No. 1 is added for sectionally marking the data, and a computer randomly generates a non-repeated numerical mark to fill the row, and the scheme is filled in [1,2,3,4.. 10]. Such as: when the unit water head is 158m, the unit water head_BIN number 1 is marked as 1.
After the excitation current is in the area 1656-3264A, a column named as 'No. 1 unit excitation current_BIN' is added for sectionally marking the data, and a computer randomly generates a non-repeated numerical mark to fill the column, and in the scheme, the column is filled with [1,2,3,4....10]. Such as: when the exciting current is 1700A, the exciting current_BIN of the No. 1 unit is marked as 1.
All data are marked with conditions as shown in table 2:
TABLE 2
Step (4) dividing the data into sample data sets and monitoring data sets: the monitoring period starting time 2020 is 1 month and 1 day as nodes, and the data are divided into 2 data sets: a sample data set 1 month and 1 day before 2020 for calculating a threshold value; the monitoring data set is used for state monitoring or trend analysis after 1 month and 1 day in 2020.
Step (5), data cleaning is carried out on the sample data set, and an effective sample data set is obtained:
51 Any combination of the condition flags calculated in step (3) is defined as a condition, such as: the active power_bin=1 of the No. 1 unit, the water head_bin=1 of the No. 1 unit and the exciting current_bin=3 of the No. 1 unit are one working condition, and 1000 working conditions are all used.
52 Calculating the first quartile of the frame +X-direction horizontal vibration peak value on the No. 1 unit (recorded as: "machine frame on machine set 1+x-direction horizontal vibration peak value_ FirstQuartile"), third quartile (noted as: "machine frame on No. 1 set+X-direction horizontal vibration peak value_ ThirdQuartile");
53 Upper outlier limit calculation:
Upper limit of +X-direction outlier= "upper rack of No. 1+X-direction horizontal vibration peak value_ ThirdQuartile" +1.5X "upper rack of No. 1+X-direction horizontal vibration peak value_ ThirdQuartile" - "upper rack of No. 1+X-direction horizontal vibration peak value_ FirstQuartile"
54 Outlier lower limit calculation:
+X-direction outlier lower limit= "No. 1 upper rack+X-direction horizontal vibration peak value_ FirstQuartile" -1.5X "(" No. 1 upper rack+X-direction horizontal vibration peak value_ ThirdQuartile "-" No. 1 upper rack+X-direction horizontal vibration peak value_ FirstQuartile ")
After calculation, the data table is added with two columns of outlier upper limit and outlier lower limit, as in table 3:
TABLE 3 Table 3
55 Deleting data in the sample data set above the upper outlier limit or below the lower outlier limit to obtain a valid sample data set.
Step (6) calculating the effective sample data set to obtain upper and lower threshold limits:
61 Under each working condition, the maximum value of the 'frame on the No.1 unit plus X-direction horizontal vibration peak value' (recorded as: "machine frame on machine set No. 1+x horizontal vibration peak value_max"), minimum value (noted as: "machine frame on machine set No. 1+x-direction horizontal vibration peak value_min"), arithmetic mean (noted as: "frame on unit 1+horizontal vibration peak-to-peak value of X-direction_mean");
62 Calculating upper and lower threshold limits for each working condition:
the upper frame of the No. 1 machine set+X-direction horizontal vibration threshold upper limit "=" the upper frame of the No. 1 machine set+X-direction horizontal vibration peak value_max "+0.25X" "theupper frame of the No. 1 machine set+X-direction horizontal vibration peak value_mean";
The lower limit of the horizontal vibration threshold of the upper rack of the No. 1 machine set and the horizontal vibration threshold of the X direction is = "the upper rack of the No. 1 machine set and the horizontal vibration peak value of the X direction is =" 0.25X "-the upper rack of the No. 1 machine set and the horizontal vibration peak value of the X direction is @ mean";
TABLE 4 Table 4
Step (7) writing the upper and lower threshold limits into the monitored data set:
the monitoring dataset was increased by 2 columns of data, and each column of data was filled with the upper and lower threshold limits, as shown in table 5, against the operating mode markers.
TABLE 5
/>
Step (8) monitoring data set data and calculating out-of-limit ratio:
The ratio of the out-of-limit of the horizontal vibration of the upper machine frame of the No. 1 plus the horizontal vibration of the X direction is = "the peak value of the horizontal vibration of the upper machine frame of the No. 1 plus the horizontal vibration of the X direction"/"the upper machine frame of the No. 1 plus the upper limit of the horizontal vibration threshold of the X direction" x100%. The calculation results are shown in Table 6.
TABLE 6
And (9) monitoring and visually displaying.
91 When the out-of-limit ratio is more than 100%, starting timing
92 When the out-of-limit ratio is greater than 100%, the timing is greater than 1 minute, an alarm of 'the upper machine frame of the No. 1 machine set plus the out-of-limit trend of the X-direction horizontal vibration peak value' is output, and the monitoring is continued;
93 When the out-of-limit ratio is less than 100%, or when the out-of-limit ratio is more than 100% and does not exceed the set time, continuing to monitor
94 A visual presentation).
The foregoing has shown and described the basic principles, principal features and advantages of the invention. It will be understood by those skilled in the art that the present invention is not limited to the embodiments described above, and that the above embodiments and descriptions are merely illustrative of the principles of the present invention, and various changes and modifications may be made without departing from the spirit and scope of the invention, which is defined in the appended claims. The scope of the invention is defined by the appended claims and equivalents thereof.
Claims (2)
1. A method for monitoring an adaptive dynamic alert threshold for a hydro-generator, comprising the steps of:
(1) Acquiring system acquisition data, comprising: time, generator active power, unit water head, exciting current and monitoring index data;
(2) Data screening: selecting data with active power more than or equal to 70% of rated power;
(3) Working condition segmentation: segmenting the selected data according to three working condition indexes of active power, a unit water head and exciting current, wherein each working condition index is equally divided into N segments, and marking the segments;
(4) Data is divided into sample data sets, monitoring data sets: after the time period of the monitoring data set is determined, selecting data of one year above the starting time of the monitoring data set as a sample data set;
(5) Performing data cleaning on the sample data set to obtain an effective sample data set;
51 Any one of three working condition indexes of active power, unit water head and exciting current is combined into one working condition, and after each of the three working condition indexes is arranged and combined in N sections, N 3 working conditions are shared;
52 Calculating a first quartile and a third quartile of the sample dataset monitoring index under each working condition; the first quartile is denoted as monitor index_ FirstQuartile and the third quartile is denoted as: monitoring index_ ThirdQuartile;
53 Calculating an outlier upper bound:
monitoring index outlier upper limit = monitoring index_ ThirdQuartile +1.5 (monitoring index_ ThirdQuartile-monitoring index_ FirstQuartile);
54 Calculating an outlier lower bound:
Monitor index outlier lower limit = monitor index_ FirstQuartile-1.5 (monitor index_ ThirdQuartile-monitor index_ FirstQuartile);
55 Deleting data in the sample data set which is higher than the upper limit of the outlier or lower than the lower limit of the outlier to obtain a valid sample data set;
(6) Calculating an effective sample data set to obtain upper and lower threshold limits;
61 Calculating the maximum value, the minimum value and the arithmetic mean of monitoring indexes in the effective sample data set under each working condition; wherein, the maximum value of the monitoring index is marked as a monitoring index_max, the minimum value is marked as a monitoring index_min, and the arithmetic average number is marked as: monitoring index_mean;
62 Calculating upper and lower threshold limits for each working condition:
monitoring index threshold upper limit=monitoring index_max+0.25 monitoring index_mean;
monitoring index threshold lower limit = monitoring index_min-0.25 monitoring index_mean;
(7) Writing the upper and lower threshold limits to the monitored dataset:
2 rows of data columns are added to the monitoring data set, and according to the working condition marks, each row of data is filled with the upper limit and the lower limit of the monitoring index threshold;
(8) Calculating an out-of-limit ratio for the monitored dataset:
monitoring index out-of-limit ratio = monitoring index actual measurement/monitoring index threshold upper limit 100%;
(9) Monitoring and visual display:
91 When the out-of-limit ratio is more than 100%, starting timing;
92 When the out-of-limit ratio is more than 100%, outputting an alarm and continuing to monitor when the timing is more than the set time;
93 When the out-of-limit ratio is less than 100%, or when the out-of-limit ratio is more than 100% and does not exceed the set time, continuing monitoring;
94 A visual presentation).
2. The method for monitoring an adaptive dynamic alert threshold for a hydro-generator according to claim 1 wherein each condition indicator is divided equally into 10 segments.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011357305.XA CN112946469B (en) | 2020-11-26 | 2020-11-26 | Monitoring method for self-adaptive dynamic alarm threshold of hydraulic generator |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011357305.XA CN112946469B (en) | 2020-11-26 | 2020-11-26 | Monitoring method for self-adaptive dynamic alarm threshold of hydraulic generator |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112946469A CN112946469A (en) | 2021-06-11 |
| CN112946469B true CN112946469B (en) | 2024-06-18 |
Family
ID=76234708
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011357305.XA Active CN112946469B (en) | 2020-11-26 | 2020-11-26 | Monitoring method for self-adaptive dynamic alarm threshold of hydraulic generator |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112946469B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114091792B (en) * | 2022-01-21 | 2022-06-03 | 华电电力科学研究院有限公司 | Hydro-generator degradation early warning method, equipment and medium based on stable working conditions |
| CN116111727B (en) * | 2023-04-13 | 2023-06-30 | 盛锋电力科技有限公司 | Comprehensive distribution box abnormity monitoring method based on dynamic temperature threshold |
| CN117034109A (en) * | 2023-08-03 | 2023-11-10 | 中国人民解放军95616部队保障部 | Engine oil abrasive grain analysis method and system based on segmentation threshold and computer readable storage medium |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101289990A (en) * | 2008-06-17 | 2008-10-22 | 四川中鼎自动控制有限公司 | Hydro-turbo generator set vibration protection accomplishing method |
Family Cites Families (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR100317842B1 (en) * | 1999-12-16 | 2001-12-24 | 임정규 | A real-time efficiency monitoring system of hydro-turbine generator and pump motor |
| CN103779876B (en) * | 2014-01-15 | 2015-10-21 | 国家电网公司 | A kind of power system dispatching method out-of-limit based on the elimination section of sensitivity analysis |
| CN104131950B (en) * | 2014-07-24 | 2017-02-01 | 重庆大学 | Partitioning determination method for threshold value of temperature characteristic quantity of wind generating set |
| CN104748839B (en) * | 2015-04-02 | 2017-06-20 | 贵州电力试验研究院 | Vibration of hydrogenerator set state region monitoring method based on real time on-line monitoring |
| CN105628421B (en) * | 2015-12-25 | 2018-06-08 | 南京南瑞集团公司 | A kind of out-of-limit monitoring and pre-alarming method of Hydropower Unit divided working status runout |
| CN105678025A (en) * | 2016-02-29 | 2016-06-15 | 华能澜沧江水电股份有限公司小湾水电厂 | Water-turbine running optimizing method and system based on dynamic stress test and stability test |
| WO2017181322A1 (en) * | 2016-04-18 | 2017-10-26 | Abb Schweiz Ag | A method, system and apparatus for operating a hydraulic turbine |
| CN109670400B (en) * | 2018-11-13 | 2021-10-22 | 国网浙江省电力有限公司紧水滩水力发电厂 | Method for evaluating stability state of hydroelectric generating set in starting process |
| CN110989548B (en) * | 2019-11-01 | 2022-11-01 | 华能澜沧江水电股份有限公司 | Method for judging abnormal closed-loop regulation function of active power of single machine of hydraulic generator |
| CN111092442A (en) * | 2019-12-19 | 2020-05-01 | 国网浙江省电力有限公司紧水滩水力发电厂 | Hydroelectric generating set multi-dimensional vibration region fine division method based on decision tree model |
| CN111539553B (en) * | 2020-03-31 | 2023-10-24 | 华北电力大学 | Wind turbine generator fault early warning method based on SVR algorithm and off-peak degree |
| CN111537257B (en) * | 2020-05-30 | 2022-05-17 | 华能澜沧江水电股份有限公司 | Method for online detection of abnormality of air cooler of hydraulic generator |
| CN111931849B (en) * | 2020-08-11 | 2023-11-17 | 北京中水科水电科技开发有限公司 | Hydropower unit operation data trend early warning method |
-
2020
- 2020-11-26 CN CN202011357305.XA patent/CN112946469B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101289990A (en) * | 2008-06-17 | 2008-10-22 | 四川中鼎自动控制有限公司 | Hydro-turbo generator set vibration protection accomplishing method |
Non-Patent Citations (1)
| Title |
|---|
| 水轮发电机组稳定性参数统计特性与监测报警阈值研究;张飞等;水力发电学报;20131031;第32卷(第5期);269-272 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112946469A (en) | 2021-06-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112946469B (en) | Monitoring method for self-adaptive dynamic alarm threshold of hydraulic generator | |
| JP2009075081A (en) | Fleet anomaly detection method | |
| JP2009076056A (en) | Anomaly aggregation method | |
| CN106092190A (en) | Pump-storage generator stable sexual state deterioration method for early warning and system | |
| CN116660672B (en) | Power grid equipment fault diagnosis method and system based on big data | |
| CN104793605A (en) | Method for judging equipment faults by means of normal distribution | |
| CN103912448A (en) | Method for monitoring power characteristics of units of regional wind farms | |
| CN117251812A (en) | High-voltage power line operation fault detection method based on big data analysis | |
| CN110647093A (en) | Intelligent monitoring system and monitoring method for power system based on big data analysis | |
| CN113221455A (en) | Equipment health state detection method and device | |
| CN114813124B (en) | Bearing fault monitoring method and device | |
| CN109840601B (en) | Operation management method combining wind power plant equipment monitoring and production management | |
| CN106443503A (en) | Method and system for detecting quality of power supply air-conditioning capable of detecting quality of power supply | |
| CN110687851A (en) | Terminal operation monitoring system and method | |
| CN113759785A (en) | Method for realizing equipment fault early warning based on big data analysis technology | |
| CN112395550B (en) | Rotary machine fault early warning method based on visual characteristic parameter matrix | |
| CN111191950B (en) | Method and device for analyzing abnormal oil temperature of gearbox of wind turbine generator | |
| CN114063582A (en) | Method and device for monitoring a product test process | |
| JP2018036970A (en) | Facility management system, facility management method, and program | |
| CN112731022A (en) | Photovoltaic inverter fault detection method, device and medium | |
| JP2010262630A (en) | Device and method for monitoring industrial process | |
| CN115600879A (en) | Circuit breaker abnormity early warning method, system and related device | |
| CN112796920B (en) | Early warning method for vertical mixed-flow hydraulic generator runner penetrating crack | |
| CN114264902A (en) | Method and system for monitoring working state of lightning protection box, electronic equipment and storage medium | |
| CN115982665B (en) | Quality anomaly auditing method and system for water turbine measurement data |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant |