阅读说明：本技术 立式混流式水轮发电机转轮贯穿性裂纹早期预警的方法 (Early warning method for vertical mixed-flow hydraulic generator runner penetrating crack ) 是由 朱晓韬 吴永智 刀亚娟 钟新元 吕爱军 戴祥 冷天先 叶超 徐文冰 赵海峰 于 2020-12-14 设计创作，主要内容包括：本发明涉及一种立式混流式水轮发电机转轮贯穿性裂纹早期预警的方法,属于涂料技术领域。该方法包括：监测传感器的安装；采集数据；数据筛选；工况分段；数据分为样本数据集、监测数据集；数据清洗；正常范围上限计算；正常范围上限写入监测数据集；计算越限比；转轮裂纹判定逻辑计算及告警输出。本发明方法利用水轮发电机有功功率、水头、振动摆度等状态监测时间序列数据,快速计算水轮发电机在不同运行工况下的水导摆度、顶盖振动正常范围上限,以及当前的越限比,当转轮发生贯穿性裂纹时,水导摆度、顶盖振动越限比符合判定逻辑,及时输出告警,避免裂纹进一步扩大造成重大事故。本发明方法具有操作简单、通用性强的特点,易于推广应用。(The invention relates to a method for early warning of through cracks of a rotating wheel of a vertical mixed-flow hydraulic generator, belonging to the technical field of coatings. The method comprises the following steps: monitoring the installation of a sensor; collecting data; screening data; segmenting the working condition; the data is divided into a sample data set and a monitoring data set; data cleaning; calculating the upper limit of the normal range; writing the monitoring data set at the upper limit of the normal range; calculating an out-of-limit ratio; and (4) carrying out logic calculation and alarm output on the crack judgment of the rotating wheel. According to the method, the time sequence data are monitored by utilizing states of the active power, the water head, the vibration swing degree and the like of the hydraulic generator, the water guide swing degree, the upper limit of the normal range of the top cover vibration and the current out-of-limit ratio of the hydraulic generator under different operation conditions are rapidly calculated, when the runner has the penetrating cracks, the water guide swing degree and the top cover vibration out-of-limit ratio accord with judgment logic, an alarm is timely output, and major accidents caused by further expansion of the cracks are avoided. The method has the characteristics of simple operation and strong universality, and is easy to popularize and apply.)

1. The method for early warning the penetrability cracks of the vertical mixed-flow hydraulic generator runner is characterized by comprising the following steps of:

(1) a top cover + X-direction horizontal vibration monitoring sensor, a top cover-Y-direction horizontal vibration monitoring sensor and a top cover + X-direction vertical vibration monitoring sensor are arranged at the top cover position of the hydraulic generator; installing a water guide and X-throw monitoring sensor and a water guide-Y-direction throw monitoring sensor at the position of a water guide bearing;

(2) collecting data, comprising: time, generator active power, unit water head, guide vane opening, top cover + X-direction horizontal vibration peak value, top cover-Y-direction horizontal vibration peak value, top cover + X-direction vertical vibration peak value, water guide + X-throw peak value and water guide-Y-direction throw peak value;

(2) and (3) screening data: selecting data with active power more than or equal to 70% of rated power, and deleting the rest data;

(3) working condition segmentation: segmenting data according to two working condition indexes of a water head and a guide vane opening of a unit according to integers, dividing the data into S sections according to the water head, dividing the data into D sections according to the guide vane opening, and making segmented marks;

31) the head is divided into S sections, each section being labeled [1,2,3.. S ]; dividing the guide vane into D sections according to the opening degree of the guide vane, wherein each section is marked as [1,2,3.. D ];

32) any combination of the water head and the opening degree of the guide vane is a working condition;

(4) dividing data into a sample data set and a monitoring data set:

(5) carrying out data cleaning on the sample data set to obtain an effective sample data set;

(6) calculating the effective sample data set to obtain 5 upper limits of the normal range of the monitoring indexes of a top cover + X-direction horizontal vibration peak value, a top cover-Y-direction horizontal vibration peak value, a top cover + X-direction vertical vibration peak value, a water guide + X swing peak value, a water guide + Y swing peak value;

(7) write the upper normal range limit into the monitoring dataset:

the monitoring data set was augmented with 5 columns: comparing the working condition marks with the upper limit of the normal range of water guide + X-direction throw, the upper limit of the normal range of water guide-Y-direction throw, the upper limit of the normal range of top cover + X-direction horizontal vibration, the upper limit of the normal range of top cover-Y-direction horizontal vibration and the upper limit of the normal range of top cover + X-direction vertical vibration, and filling the calculation result obtained in the step (6) into a monitoring data set;

(8) calculating out-of-limit ratios of water guide + X-direction throw, water guide-Y-direction throw, top cover + X-direction horizontal vibration, top cover-Y-direction horizontal vibration and top cover + X-direction vertical vibration;

(9) and (4) carrying out logic calculation and alarm output on the crack judgment of the rotating wheel.

2. The method for early warning the through cracks of the vertical mixed-flow hydraulic generator runner according to claim 1, wherein the specific method in the step (4) is as follows:

41) if the runner is inspected in the last year, the last inspection confirms that the runner has no crack and the date is T1, the data in the previous year of T1 is listed as a sample data set, and the data after T1 is taken as a monitoring data set;

42) and if the rotating wheel is not checked in the last year, selecting a time period to be evaluated, and after determining that the time period after the first time T2 of the time period to be evaluated is a monitoring data set, taking the data in the previous year T2 as a sample data set.

3. The method for early warning the through cracks of the vertical mixed-flow hydraulic generator runner according to claim 1, wherein the specific method in the step (5) is as follows: cleaning and explaining the peak value data of the horizontal vibration peak in the top cover and the X direction, and analogizing the other four indexes;

51) calculating a first quartile and a third quartile of a top cover + X-direction horizontal vibration peak value of the sample data set under each working condition; the first quartile is noted as: the peak value of the top cover + X-direction horizontal vibration peak value _ FirstQuartile, and the third quartile is recorded as: the top cover + X-direction horizontal vibration peak value _ ThirdQuartile;

52) calculating an upper limit of the abnormality:

top cap + X-direction horizontal vibration peak value upper limit = top cap + X-direction horizontal vibration peak value _ third quartile +1.5 [ (+ top cap + X-direction horizontal vibration peak value _ third quartile-top cap + X-direction horizontal vibration peak value _ first quartile);

53) calculating an abnormality lower limit:

top cap + X direction horizontal vibration peak value abnormal lower limit = top cap + X direction horizontal vibration peak value _ first quartile-1.5 (top cap + X direction horizontal vibration peak value _ third quartile-top cap + X direction horizontal vibration peak value _ first quartile)

54) And deleting the data which are higher than the abnormal upper limit or lower than the abnormal lower limit in the sample data set to obtain an effective sample data set.

4. The method for early warning the through cracks of the vertical mixed-flow hydraulic generator runner according to claim 1, wherein the specific method in the step (6) is as follows:

61) any combination of the water head and the opening degree of the guide vane is one working condition, and N working conditions are shared; n = S × D;

62) calculating the maximum value and the arithmetic mean of the top cover + X-direction horizontal vibration peak value, the top cover-Y-direction horizontal vibration peak value, the top cover + X-direction vertical vibration peak value, the water guide + X-throw peak value and the water guide-Y-direction throw peak value in the effective sample data set under each working condition;

63) calculating the upper limit of the normal range of the vibration swing under each working condition: upper normal range limit = maximum +0.25 arithmetic mean.

5. The method for early warning the through cracks of the vertical mixed-flow hydraulic generator runner according to claim 1, wherein the specific method in the step (8) is as follows:

out-of-limit ratio = measured value/upper normal range limit 100%.

6. The method for early warning the through cracks of the vertical mixed-flow hydraulic generator runner according to claim 1, wherein the specific method in the step (9) is as follows:

91) the 'water guide + X direction swing degree out-of-limit ratio' is more than or equal to 100% ', and the' water guide + X direction swing degree out-of-limit ratio 'is more than or equal to 100%' meets any one condition;

92) the ' top cover + X-direction horizontal vibration out-of-limit ratio ' ≧ 100% ', the ' top cover-Y-direction horizontal vibration out-of-limit ratio ' ≧ 100% ', and the ' top cover + X-direction vertical vibration out-of-limit ratio ' ≧ 100% ' satisfy any two conditions;

93) when 91) and 92) are met simultaneously, setting the mark of the wheel crack early warning as 1;

94) when the runner crack early warning is more than 5 times per day, the runner of the water turbine is judged to have the through runner crack or drop blocks.

Technical Field

The invention belongs to the technical field of on-line monitoring, state evaluation and fault diagnosis of a hydraulic generator, and particularly relates to a method for early warning of a vertical mixed-flow hydraulic generator runner through crack.

Background

In the online monitoring, state evaluation and fault diagnosis of the hydraulic turbine generator, a crucial task is to judge whether the hydraulic turbine runner works normally and whether cracks, blocks and other defects exist, and the important factor for determining whether the unit needs to be stopped for maintenance is also provided. The runner cracks seriously jeopardize the operation stability and the operation safety of the power station, and the frequent shutdown for welding and repairing the cracks causes great economic loss for the power station.

The traditional method lacks an effective runner defect assessment means, generally adopts the mode of combining minor repair and major repair, namely regular maintenance, and stops the machine to drain water after the unit operates for 1 year or several years, and manually inspects whether the runner has defects. The traditional method is passive, the rotating wheel has serious defects of penetrating cracks, falling blocks and the like generally during regular maintenance, and some power stations even need to carry out C-level or D-level maintenance once every year to check the state of the rotating wheel so as to repair the defects of the rotating wheel cracks and the falling blocks afterwards.

In recent years, the operation of exploring the defect assessment of the rotating wheel is widely developed by collecting the operation data of various working conditions of the unit in the industry, but the technology is a breakthrough. For various vibration swings, complex and variable working condition parameters and mass variable values of the hydraulic generator, relevance analysis is carried out one by one, so that the runner degradation analysis is further carried out, a large amount of manual intervention is required, and the specific fault that the runner cracks cannot be accurately judged through characteristic variables. Therefore, how to overcome the deficiency of the prior art, accurate identification characteristic variable accurately gives the early warning of runner crackle, and online assessment of runner blade crackle state is more and more urgent, and more is the problem that the state assessment technology field of present hydraulic generator needs to solve urgently.

Disclosure of Invention

The invention aims to solve the defects of the prior art and provides a method for early warning of the through cracks of a rotating wheel of a vertical mixed-flow hydraulic generator. The method is simple to operate, high in universality and easy to popularize and apply.

In order to achieve the purpose, the technical scheme adopted by the invention is as follows:

the method for early warning the penetrating cracks of the vertical mixed-flow hydraulic generator runner comprises the following steps:

(1) a top cover + X-direction horizontal vibration monitoring sensor, a top cover-Y-direction horizontal vibration monitoring sensor and a top cover + X-direction vertical vibration monitoring sensor are arranged at the top cover position of the hydraulic generator; installing a water guide and X-throw monitoring sensor and a water guide-Y-direction throw monitoring sensor at the position of a water guide bearing;

(2) collecting data, comprising: time, generator active power, unit water head, guide vane opening, top cover + X-direction horizontal vibration peak value, top cover-Y-direction horizontal vibration peak value, top cover + X-direction vertical vibration peak value, water guide + X-throw peak value and water guide-Y-direction throw peak value;

(2) and (3) screening data: selecting data with active power more than or equal to 70% of rated power, and deleting the rest data;

(3) working condition segmentation: segmenting data according to two working condition indexes of a water head and a guide vane opening of a unit according to integers, dividing the data into S sections according to the water head, dividing the data into D sections according to the guide vane opening, and making segmented marks;

31) the head is divided into S sections, each section being labeled [1,2,3.. S ]; the guide vane opening is divided into D sections, and each section is marked as [1,2,3.. D ].

32) Any combination of the water head and the opening degree of the guide vane is a working condition; such as: if the [ water head and guide vane opening ] [1,5] is a working condition, a total of N working conditions (N ═ S × D) exist;

(4) dividing data into a sample data set and a monitoring data set:

(5) carrying out data cleaning on the sample data set to obtain an effective sample data set;

(6) calculating the effective sample data set to obtain 5 upper limits of the normal range of the monitoring indexes of a top cover + X-direction horizontal vibration peak value, a top cover-Y-direction horizontal vibration peak value, a top cover + X-direction vertical vibration peak value, a water guide + X swing peak value, a water guide + Y swing peak value;

(7) write the upper normal range limit into the monitoring dataset:

the monitoring data set was augmented with 5 columns: comparing the working condition marks with the upper limit of the normal range of water guide + X-direction throw, the upper limit of the normal range of water guide-Y-direction throw, the upper limit of the normal range of top cover + X-direction horizontal vibration, the upper limit of the normal range of top cover-Y-direction horizontal vibration and the upper limit of the normal range of top cover + X-direction vertical vibration, and filling the calculation result obtained in the step (6) into a monitoring data set;

(8) calculating out-of-limit ratios of water guide + X-direction throw, water guide-Y-direction throw, top cover + X-direction horizontal vibration, top cover-Y-direction horizontal vibration and top cover + X-direction vertical vibration;

(9) and (4) carrying out logic calculation and alarm output on the crack judgment of the rotating wheel.

Further, it is preferable that the specific method of step (4) is:

41) if the runner is inspected in the last year, the last inspection confirms that the runner has no crack and the date is T1, the data in the previous year of T1 is listed as a sample data set, and the data after T1 is taken as a monitoring data set;

42) if the rotating wheel is not checked in the last year, selecting a time period to be evaluated, and determining that the time period after the first time T2 of the time period to be evaluated is a monitoring data set, and the data in the previous year T2 is a sample data set; the time period to be evaluated is selected according to the self requirement, and the invention is not particularly limited to this.

The method is not limited to data of one year for the sample data set, and the longer the data time is, the more accurate the result is, but the problem of overlarge calculated amount is brought; the data may be 2-3 months, but the final result may be biased because the data of the sample data set is not comprehensive, preferably 8-16 months of data as the sample data set.

Further, it is preferable that the specific method of step (5) is: cleaning and explaining the peak value data of the horizontal vibration peak in the top cover and the X direction, and analogizing the other four indexes;

51) calculating a first quartile and a third quartile of a top cover + X-direction horizontal vibration peak value of the sample data set under each working condition; the first quartile is noted as: the peak value of the top cover + X-direction horizontal vibration peak value _ First Quartile, and the third Quartile is recorded as: the top cover + X-direction horizontal vibration peak value _ ThirdQuartile;

52) calculating an upper limit of the abnormality:

the upper limit of the abnormal peak value of the top cover + X-direction horizontal vibration peak value is equal to the top cover + X-direction horizontal vibration peak value _ third quartile +1.5 (the top cover + X-direction horizontal vibration peak value _ third quartile-top cover + X-direction horizontal vibration peak value _ first quartile);

53) calculating an abnormality lower limit:

top cover + X direction horizontal vibration peak value abnormal lower limit ═ top cover + X direction horizontal vibration peak value _ first quartile-1.5 × (top cover + X direction horizontal vibration peak value _ third quartile-top cover + X direction horizontal vibration peak value _ first quartile)

54) And deleting the data which are higher than the abnormal upper limit or lower than the abnormal lower limit in the sample data set to obtain an effective sample data set.

Further, it is preferable that the specific method of step (6) is:

61) any combination of the water head and the opening degree of the guide vane is one working condition, and N working conditions are shared; n ═ S × D;

62) calculating the maximum value and the arithmetic mean of the top cover + X-direction horizontal vibration peak value, the top cover-Y-direction horizontal vibration peak value, the top cover + X-direction vertical vibration peak value, the water guide + X-throw peak value and the water guide-Y-direction throw peak value in the effective sample data set under each working condition;

63) calculating the upper limit of the normal range of the vibration swing under each working condition: upper normal range limit is the maximum +0.25 arithmetic mean.

Further, it is preferable that the specific method of the step (8) is:

out-of-limit ratio-measured value/upper normal range limit 100%.

Further, it is preferable that the specific method of step (9) is:

91) the 'water guide + X direction swing degree out-of-limit ratio' is more than or equal to 100% ', and the' water guide + X direction swing degree out-of-limit ratio 'is more than or equal to 100%' meets any one condition;

92) the ' top cover + X-direction horizontal vibration out-of-limit ratio ' ≧ 100% ', the ' top cover-Y-direction horizontal vibration out-of-limit ratio ' ≧ 100% ', and the ' top cover + X-direction vertical vibration out-of-limit ratio ' ≧ 100% ' satisfy any two conditions;

93) when 91) and 92) are met simultaneously, setting the mark of the wheel crack early warning as 1;

94) when the runner crack early warning is more than 5 times per day, the runner of the water turbine is judged to have the through runner crack or drop blocks.

The method uses the working condition of the water wheel generator, the water guide swing degree and the historical data of the vibration time sequence of the top cover, adaptively calculates the upper limit of the normal range of the vibration swing degree under each working condition, compares the evaluation data with the upper limit of the normal range, and gives an alarm when the judgment logic is met. The method can be used for various data analysis tools or programming tools, and the turbine runner penetrating crack can be detected without disassembling the machine set.

The invention is based on the following principle:

(1) the upper limit of the normal range related to the method is obtained by processing historical data of the unit, extracting effective sample data and calculating, and is dynamic according to working condition changes, and the upper limit of the normal range reflects the concentration trend of the vibration swing of the healthy unit.

(2) The computer system can generate certain outlier data in the links of data acquisition, storage and the like, and can influence the calculation result.

(3) If the fluctuation is caused by the change of the working condition, the out-of-limit ratio is lower than 100%, and if the fluctuation is caused by the deterioration of the equipment, the out-of-limit ratio is higher than 100%. The out-of-limit ratio is higher than 100% of the actual reaction equipment deterioration condition.

(4) And logically judging the cracks of the runner, namely adopting an out-of-limit ratio as input, simultaneously exceeding the water guide swing degree and the top cover vibration out-of-limit ratio and outputting a runner crack alarm.

(5) The water guide swing degree and the top cover vibration are out-of-limit logic combination, the water guide swing degree is out-of-limit or one, and the top cover vibration is selected two, so that the verification is more in line with the rule of equipment, the alarm is accurate, and the false alarm is not easy to occur.

Compared with the prior art, the invention has the beneficial effects that:

1. in the prior art, manual detection methods are mostly adopted, visual observation and flaw detection are carried out on the rotating wheel part, and the detection can only be carried out after the unit is stopped, so that the generating benefit of the unit is influenced.

2. At present, a monitoring system and an on-line monitoring system of the water turbine generator set monitor important indexes, a higher alarm limit value is set for avoiding false alarm, the limit value is a static value, and serious faults may occur during alarm. The reference value of the out-of-limit ratio is a dynamic value, the out-of-limit ratio truly reflects the health state of the vibration swing, and early warning can be performed once the deterioration trend deviating from the health state occurs.

3. When the unit normally operates, certain fluctuation exists in the water guide swing degree and the top cover vibration of the unit along with load adjustment and working condition change; in case of abnormality, the water swing guide degree and the vibration of the top cover also fluctuate. With the conventional method, it is difficult to distinguish whether the vibration swing fluctuation is caused by the change of working conditions or the abnormality of the equipment, and the method can identify the water guide swing and the vibration rise of the top cover caused by the equipment problem.

4. The data stored in each hydropower station data center come from each system, some abnormal data or noise points may appear in links such as data acquisition and collection, a normal range is calculated by using the full data, and the influence of abnormal values may be caused.

5. For various vibration swings, complex and variable working condition parameters and mass variable values of the hydraulic generator, relevance analysis is carried out one by one, so that the runner degradation analysis is further carried out, a large amount of manual intervention is required, and the specific fault that the runner cracks cannot be accurately judged through characteristic variables. The invention does not need a large amount of manual intervention, and can give early warning of the cracks of the runner by eliminating other factors such as the looseness of the clearance of the bearing bush and the like according to the variation trend of certain characteristic variables.

6. The method is successfully verified on the early warning of the two-time penetrating cracks of the turning wheel blade of the unit No. 3 of the Ruilijiang power station and the turning wheel blade of the unit No. 1 of the Bay power station, and is simple to operate, strong in universality and easy to popularize and apply. The method is used for monitoring, the Bay No. 1 machine gives an early warning in 11 months in 2019, and a penetrating crack of 360mm at the maximum is found in unit overhaul inspection in 12 months in 2019. The Bay No. 1 machine warns at the bottom of 7 months in 2020, and the maximum 570mm penetrating crack is found by overhaul and inspection in 12 months in 2020. And after the runner of the Ruiljiang No. 3 machine 2019 in 4 months falls off, data are exported for verification, and cracks can be early warned in the beginning of 4 months. The early warning accuracy rate is 100%.

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 invention only and should not be taken as limiting the scope of the invention. The examples do not specify particular techniques or conditions, and are performed according to the techniques or conditions described in the literature in the art or according to the product specifications. The materials or equipment used are not indicated by manufacturers, and all are conventional products available by purchase.

Example 1

The method for early warning the penetrating cracks of the vertical mixed-flow hydraulic generator runner comprises the following steps:

(1) a top cover + X-direction horizontal vibration monitoring sensor, a top cover-Y-direction horizontal vibration monitoring sensor and a top cover + X-direction vertical vibration monitoring sensor are arranged at the top cover position of the hydraulic generator; installing a water guide and X-throw monitoring sensor and a water guide-Y-direction throw monitoring sensor at the position of a water guide bearing;

(2) collecting data, comprising: time, generator active power, unit water head, guide vane opening, top cover + X-direction horizontal vibration peak value, top cover-Y-direction horizontal vibration peak value, top cover + X-direction vertical vibration peak value, water guide + X-throw peak value and water guide-Y-direction throw peak value;

(2) and (3) screening data: selecting data with active power more than or equal to 70% of rated power, and deleting the rest data;

(3) working condition segmentation: segmenting data according to two working condition indexes of a water head and a guide vane opening of a unit according to integers, dividing the data into S sections according to the water head, dividing the data into D sections according to the guide vane opening, and making segmented marks;

31) the head is divided into S sections, each section being labeled [1,2,3.. S ]; the guide vane opening is divided into D sections, and each section is marked as [1,2,3.. D ].

32) Any combination of water head and guide vane opening degree is a working condition, such as: if the [ water head and guide vane opening ] [1,5] is a working condition, a total of N working conditions (N ═ S × D) exist;

(4) dividing data into a sample data set and a monitoring data set:

41) if the runner is inspected in the last year, the last inspection confirms that the runner has no crack and the date is T1, the data in the previous year of T1 is listed as a sample data set, and the data after T1 is taken as a monitoring data set;

42) if the rotating wheel is not checked in the last year, selecting a time period to be evaluated, and determining that the time period after the first time T2 of the time period to be evaluated is a monitoring data set, and the data in the previous year T2 is a sample data set; the time period to be evaluated is selected according to the self requirement, and the invention is not particularly limited to this;

(5) carrying out data cleaning on the sample data set to obtain an effective sample data set;

(6) calculating the effective sample data set to obtain 5 upper limits of the normal range of the monitoring indexes of a top cover + X-direction horizontal vibration peak value, a top cover-Y-direction horizontal vibration peak value, a top cover + X-direction vertical vibration peak value, a water guide + X swing peak value, a water guide + Y swing peak value;

(7) write the upper normal range limit into the monitoring dataset:

the monitoring data set was augmented with 5 columns: comparing the working condition marks with the upper limit of the normal range of water guide + X-direction throw, the upper limit of the normal range of water guide-Y-direction throw, the upper limit of the normal range of top cover + X-direction horizontal vibration, the upper limit of the normal range of top cover-Y-direction horizontal vibration and the upper limit of the normal range of top cover + X-direction vertical vibration, and filling the calculation result obtained in the step (6) into a monitoring data set;

(8) calculating out-of-limit ratios of water guide + X-direction throw, water guide-Y-direction throw, top cover + X-direction horizontal vibration, top cover-Y-direction horizontal vibration and top cover + X-direction vertical vibration;

(9) and (4) carrying out logic calculation and alarm output on the crack judgment of the rotating wheel.

Example 2

A method for early warning of the penetrability cracks of a vertical mixed-flow hydraulic generator runner comprises the following steps:

(1) a top cover + X-direction horizontal vibration monitoring sensor, a top cover-Y-direction horizontal vibration monitoring sensor and a top cover + X-direction vertical vibration monitoring sensor are arranged at the top cover position of the hydraulic generator; and a water guide + X-throw monitoring sensor and a water guide-Y-direction throw monitoring sensor are arranged at the position of the water guide bearing.

(2) Acquiring system acquisition data, comprising: time, generator active power, unit water head, guide vane opening, top cover + X-direction horizontal vibration peak value, top cover-Y-direction horizontal vibration peak value, top cover + X-direction vertical vibration peak value, water guide + X-throw peak value and water guide-Y-direction throw peak value;

(2) and (3) screening data: selecting data with active power more than or equal to 70% of rated power, and deleting the rest data;

(3) working condition segmentation: segmenting data according to two working condition indexes of a water head and a guide vane opening of a unit according to integers, dividing the data into S sections according to the water head, dividing the data into D sections according to the guide vane opening, and making segmented marks;

31) the head is divided into S sections, each section being labeled [1,2,3.. S ]; the guide vane opening is divided into D sections, and each section is marked as [1,2,3.. D ].

32) Any combination of water head and guide vane opening degree is a working condition, such as: if [1,5] is one operating mode, there are N operating modes in total (N ═ S × D).

(4) Dividing data into a sample data set and a monitoring data set:

41) if the runner is inspected in the last year, the last inspection confirms that the runner has no crack and the date is T1, the data in the previous year of T1 is listed as a sample data set, and the data after T1 is taken as a monitoring data set;

42) if the rotating wheel is not checked in the last year, after the time period after T2 is determined to be the monitoring data set, the data column in the previous year before T2 is the sample data set.

(5) Carrying out data cleaning on the sample data set to obtain an effective sample data set;

cleaning and explaining the peak value data of the horizontal vibration peak in the top cover and the X direction, and analogizing the other four indexes;

51) the first quartile and the third quartile of the top cover + X-direction horizontal vibration peak value in the sample data set under each working condition (each combination of S and D) are calculated. The first quartile is noted as: the peak value of the top cover + X-direction horizontal vibration peak value _ FirstQuartile, and the third quartile is recorded as: the top cover + X-direction horizontal vibration peak value _ ThirdQuartile;

52) calculating an upper limit of the abnormality:

the upper limit of the abnormal peak value of the top cover + X-direction horizontal vibration peak value is equal to the top cover + X-direction horizontal vibration peak value _ third quartile +1.5 (the top cover + X-direction horizontal vibration peak value _ third quartile-top cover + X-direction horizontal vibration peak value _ first quartile);

53) calculating an abnormality lower limit:

top cover + X direction horizontal vibration peak value abnormal lower limit ═ top cover + X direction horizontal vibration peak value _ first quartile-1.5 × (top cover + X direction horizontal vibration peak value _ third quartile-top cover + X direction horizontal vibration peak value _ first quartile)

54) And deleting the data which are higher than the abnormal upper limit or lower than the abnormal lower limit in the sample data set to obtain an effective sample data set.

(6) Calculating the effective sample data set to obtain 5 upper limits of the normal range of the monitoring indexes of a top cover + X-direction horizontal vibration peak value, a top cover-Y-direction horizontal vibration peak value, a top cover + X-direction vertical vibration peak value, a water guide + X swing peak value, a water guide + Y swing peak value; such as:

61) calculating the maximum value and the arithmetic mean of the effective sample data set and the vibration swing under each working condition; such as:

the maximum value of the top cover plus the X-direction horizontal vibration peak value is calculated as 'top cover plus X-direction horizontal vibration peak value mark _ max';

the top cover + X-direction horizontal vibration peak-to-peak arithmetic mean is calculated as "top cover + X-direction horizontal vibration peak-to-peak mark _ mean".

And so on.

62) Calculating the upper limit of the normal range of the vibration swing under each working condition: upper normal range limit is the maximum +0.25 arithmetic mean.

(7) Write the upper normal range limit into the monitoring dataset:

the monitoring data set was augmented with 5 columns: comparing the working condition marks with the upper limit of the normal range of water guide + X-direction throw, the upper limit of the normal range of water guide-Y-direction throw, the upper limit of the normal range of top cover + X-direction horizontal vibration, the upper limit of the normal range of top cover-Y-direction horizontal vibration and the upper limit of the normal range of top cover + X-direction vertical vibration, and filling the calculation result obtained in the step (6) into a monitoring data set;

(8) calculating out-of-limit ratios of water guide + X-direction throw, water guide-Y-direction throw, top cover + X-direction horizontal vibration, top cover-Y-direction horizontal vibration and top cover + X-direction vertical vibration:

the 'water conduction + X yaw out-of-limit ratio' ('water conduction + X yaw peak value'/'water conduction + X yaw normal range upper limit'. 100%;

the water guide-Y direction swing out-of-limit ratio '(' water guide-Y direction swing peak value '/' water guide-Y direction swing normal range upper limit '(' 100%);

the ' top cover + X-direction horizontal vibration out-of-limit ratio ' ═ top cover + X-direction horizontal vibration '/' top cover + X-direction horizontal vibration normal range upper limit ' × 100%;

the ' top cover-Y direction horizontal vibration out-of-limit ratio ' ═ top cover-Y direction horizontal vibration '/' top cover-Y direction horizontal vibration normal range upper limit ' × 100%.

The ' top cover + X direction vertical vibration out-of-limit ratio ' ═ top cover + X direction vertical vibration '/' top cover + X direction vertical vibration normal range upper limit ' × 100%;

(9) and (3) carrying out logic calculation and alarm output of the runner crack judgment:

91) the 'water guide + X direction swing degree out-of-limit ratio' ≧ 100% ', and the' water guide + X direction swing degree out-of-limit ratio '≧ 100%' satisfy any one of the conditions.

92) The ' top cover + X direction horizontal vibration out-of-limit ratio ' ≧ 100% ', the ' top cover-Y direction horizontal vibration out-of-limit ratio ' ≧ 100% ', and the ' top cover + X direction vertical vibration out-of-limit ratio ' ≧ 100% ' satisfy arbitrary two of them condition.

93) 91), 92) are satisfied simultaneously, and the mark of wheel crack warning is set as 1.

94) When the runner crack early warning is more than 5 times per day, the runner of the water turbine is judged to have the through runner crack or drop blocks.

Examples of the applications

The monitoring index of the turbine runner of the model 1 of the Bay is monitored and evaluated by the embodiment.

The information system collects active power of the No. 1 unit, a water head of the No. 1 unit, guide vane opening of the No. 1 unit, water guide of the No. 1 unit and X-direction swing peak value, water guide-Y-direction swing peak value of the No. 1 unit, a top cover of the No. 1 unit and X-direction horizontal vibration peak value, a top cover of the No. 1 unit and Y-direction horizontal vibration peak value, and time sequence data of the top cover of the No. 1 unit and X-direction vertical vibration peak value.

The rated active power of the unit is 700MW, under the condition of non-stop inspection, the crack condition of the runner from 11 months in 2019 to 11 months in 2020 is monitored and evaluated, and the steps are as follows:

(1) a top cover + X-direction horizontal vibration monitoring sensor, a top cover-Y-direction horizontal vibration monitoring sensor and a top cover + X-direction vertical vibration monitoring sensor are arranged at the top cover position of the hydraulic generator; and a water guide + X-throw monitoring sensor and a water guide-Y-direction throw monitoring sensor are arranged at the position of the water guide bearing.

(2) Acquiring data from 10 months in 2018 to 11 months in 2020, comprising the following steps: active power of the No. 1 unit, a water head of the No. 1 unit, guide vane opening of the No. 1 unit, water guide of the No. 1 unit and X-direction swing peak value, water guide of the No. 1 unit, Y-direction swing peak value, a top cover of the No. 1 unit, X-direction horizontal vibration peak value, top cover of the No. 1 unit, Y-direction horizontal vibration peak value, and top cover of the No. 1 unit and X-direction vertical vibration peak value are shown in a table 1 (partially displayed).

When the fluctuation range of the data is required to be more than 5%, the data acquisition equipment acquires and stores the data, the time coordinates are consistent, and the time is accurate to minutes or more.

TABLE 1

(2) And (3) screening data: selecting data (70% rated active power) with active power of the No. 1 unit being more than or equal to 490MW, and deleting the rest data;

(3) working condition segmentation: the data is segmented according to two working condition indexes of a water head and a guide vane opening degree of a unit:

31) dividing the water head of the No. 1 unit into 20 sections, adding 1 row to be named as the water head _ BIN of the No. 1 unit, and marking the ranges of the upper limit and the lower limit of the table 2 as [1,2,3.. 20 ];

TABLE 2

No. 1 unit water head _ BIN	Lower limit of	Upper limit of
			1	≥156.625	<160.621075
2	≥160.621075	<164.61715
			3	≥164.61715	<168.613225
4	≥168.613225	<172.6093
			5	≥172.6093	<176.605375
6	≥176.605375	<180.60145
			7	≥180.60145	<184.597525
8	≥184.597525	<188.5936
			9	≥188.5936	<192.589675
10	≥192.589675	<196.58575
			11	≥196.58575	<200.581825
12	≥200.581825	<204.5779
			13	≥204.5779	<208.573975
14	≥208.573975	<212.57005
			15	≥212.57005	<216.566125
16	≥216.566125	<220.5622
			17	≥220.5622	<224.558275
18	≥224.558275	<228.55435
			19	≥228.55435	<232.550425
20	≥232.550425	<＝236.5465

32) The number 1 unit guide vane opening degree is divided into 20 sections, 1 row is added to be named as the number 1 unit guide vane opening degree _ BIN, and the upper limit range and the lower limit range of the number 1 unit guide vane opening degree _ BIN are marked as [1,2,3.. 20] respectively according to the table 3;

TABLE 3

No. 1 unit guide vane opening _ BIN	Lower limit of	Upper limit of
			1	≥38.77721	<41.6637235
2	≥41.6637235	<44.550237
			3	≥44.550237	<47.4367505
4	≥47.4367505	<50.323264
			5	≥50.323264	<53.2097775
6	≥53.2097775	<56.096291
			7	≥56.096291	<58.9828045
8	≥58.9828045	<61.869318
			9	≥61.869318	<64.7558315
10	≥64.7558315	<67.642345
			11	≥67.642345	<70.5288585
12	≥70.5288585	<73.415372
			13	≥73.415372	<76.3018855
14	≥76.3018855	<79.188399
			15	≥79.188399	<82.0749125
16	≥82.0749125	<84.961426
			17	≥84.961426	<87.8479395
18	≥87.8479395	<90.734453
			19	≥90.734453	<93.6209665
20	≥93.6209665	<＝96.50748

34) Any combination of No. 1 unit water head _ BIN and No. 1 unit guide vane opening degree _ BIN is a working condition, such as: the [ No. 1 unit water head _ BIN and No. 1 unit guide vane opening degree _ BI N ] ═ 1,5] are one working condition, and there are 400 working conditions in total.

(4) Dividing data into a sample data set and a monitoring data set: determining that the data set is a monitoring data set after 11 months in 2019, and using the data set from 11 months in 2018 to 10 months in 2019 as a sample data set;

(5) carrying out data cleaning on the sample data set to obtain an effective sample data set;

51) calculating a first quartile and a third quartile of 5 monitoring indexes of a top cover + X-direction horizontal vibration peak value, a top cover-Y-direction horizontal vibration peak value, a top cover + X-direction vertical vibration peak value, a water guide + X-throw peak value, a water guide + Y-direction throw peak value under each working condition; the first quartile is recorded as the vibration swing degree _ FirstQuartile, and the third quartile is recorded as: oscillating throw _ third quartile;

52) calculating an upper limit of the abnormality:

the upper limit of the abnormal vibration swing is vibration swing _ third quartile +1.5 (vibration swing _ third quartile-vibration swing _ first quartile);

53) calculating an abnormality lower limit:

the lower limit of the oscillation swing anomaly is oscillation swing _ FirstQuartile-1.5 × (oscillation swing _ third quartile-oscillation swing _ FirstQuartile).

After calculation of the upper and lower abnormality limits, the results are shown in Table 4 (partially shown).

TABLE 4

54) And deleting the data which are higher than the abnormal upper limit or lower than the abnormal lower limit in the sample data set to obtain an effective sample data set.

(6) Calculating the effective sample data set to obtain the upper limit of the normal range of the top cover + X-direction horizontal vibration peak value, the top cover-Y-direction horizontal vibration peak value, the top cover + X-direction vertical vibration peak value, the water guide + X swing peak value, the water guide + Y swing peak value;

61) and calculating the maximum value and the arithmetic mean of the effective sample data set and the vibration swing under each working condition.

63) Calculating the upper limit of the normal range of the vibration swing under each working condition: upper normal range limit is the maximum +0.25 arithmetic mean. The calculation results are shown in Table 5 (partially shown).

TABLE 5

(7) Write the upper normal range limit into the monitoring dataset:

the monitoring data set was augmented with 5 columns: comparing the working condition marks with the upper limit of the normal range of water guide + X-direction throw, the upper limit of the normal range of water guide-Y-direction throw, the upper limit of the normal range of top cover + X-direction horizontal vibration, the upper limit of the normal range of top cover-Y-direction horizontal vibration and the upper limit of the normal range of top cover + X-direction vertical vibration, and filling the calculation result obtained in the step (6) into a monitoring data set;

(8) calculating out-of-limit ratios of water guide + X-direction throw, water guide-Y-direction throw, top cover + X-direction horizontal vibration, top cover-Y-direction horizontal vibration and top cover + X-direction vertical vibration:

the 'water conduction + X yaw out-of-limit ratio' ('water conduction + X yaw peak value'/'water conduction + X yaw normal range upper limit'. 100%;

the water guide-Y direction swing out-of-limit ratio '(' water guide-Y direction swing peak value '/' water guide-Y direction swing normal range upper limit '(' 100%);

the ' top cover + X-direction horizontal vibration out-of-limit ratio ' ═ top cover + X-direction horizontal vibration '/' top cover + X-direction horizontal vibration normal range upper limit ' × 100%;

the ' top cover-Y direction horizontal vibration out-of-limit ratio ' ═ top cover-Y direction horizontal vibration '/' top cover-Y direction horizontal vibration normal range upper limit ' × 100%.

The ' top cover + X direction vertical vibration out-of-limit ratio ' ═ top cover + X direction vertical vibration '/' top cover + X direction vertical vibration normal range upper limit ' × 100%;

the calculation results are shown in Table 6 (partially shown).

TABLE 6

(9) And (3) carrying out logic calculation and alarm output of the runner crack judgment:

91) the water guide and X direction swing degree out-of-limit ratio of No. 1 unit is more than or equal to 100 percent, and the water guide and X direction swing degree out-of-limit ratio of No. 1 unit is more than or equal to 100 percent meet any one condition.

92) The ' number 1 unit top cover + X direction horizontal vibration limit-exceeding ratio ' ≧ 100% ', the ' number 1 unit top cover-Y direction horizontal vibration limit-exceeding ratio ' ≧ 100% ', the ' number 1 unit top cover + X direction vertical vibration limit-exceeding ratio ' ≧ 100% ' satisfies arbitrary two of them condition.

93) 91), 92) are satisfied simultaneously, and the mark of wheel crack warning is set as 1.

94) When the runner crack early warning is more than 5 times per day, the runner of the water turbine is judged to have the through runner crack or drop blocks. Data with the 'runner crack early warning' value of 1 are screened out and shown in a table 7 (part of the data is displayed), early warning is found in 11 months and 17 days in 2019, runner penetrating cracks are found in overhaul and inspection of the unit in 2020 and 12 months, and welding treatment is carried out. And early warning for the second time in 7 months in 2020, and checking the unit in 12 months in 2020 to confirm that the unit has the penetrating crack.

TABLE 7

The foregoing shows and describes the general principles, essential 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, which are described in the specification and illustrated only to illustrate the principle of the present invention, but that various changes and modifications may be made therein without departing from the spirit and scope of the present invention, which fall within the scope of the invention as claimed. The scope of the invention is defined by the appended claims and equivalents thereof.