1. A key phase-based equal phase processing method for a hydro-turbo generator set runout waveform is characterized by comprising the following steps:

1) installing a key phase sensor in the + X position of a main shaft of a generator, connecting the key phase sensor with various waveform data acquisition systems of the existing hydroelectric generating set, simultaneously arranging a concave or convex point on the main shaft of the generator opposite to the key phase sensor, taking the time point of the concave or convex point sweeping the key phase sensor when the generator rotates for one circle as the phase starting point of synchronous acquisition of various waveform data, recording the phase starting point as 0 degrees, recording the starting time of various waveform data as 0ms, marking a first key phase point, and marking a key phase point on the waveform data after every one circle of rotation, then:

the key phase point offset time is the current waveform data time-waveform data start time;

2) the data acquisition of the swing degree and the vibration waveform is continuous periodic cosine voltage or current signals, discrete sampling is carried out on the waveform signals by a pulse sampling function with the time of each rotation of the generator being T, the number of sampling points per week being N and the period being T, a swing degree or vibration waveform data sequence is generated for calculation, wherein the acquisition time interval of each swing degree or vibration waveform data point is as follows: T/(N-1), such that each acquired swing or vibration waveform data point has the same phase interval: 360 DEG/N, then generating a data packet by the acquired 1-8-week swing or vibration waveform data and uploading the data packet to a data server;

3) the server decompresses the data packet in the step 2), and the data packet is corrected according to the following steps:

31) correcting the number of key phases: the time deviation of the key phase offset time collected every revolution of the generator and the initial 0 point is as follows: t [ n ], n is 1, 2, 3, … … 8, and the number of bond phases is: n +1, and the deficiency is corrected to n + 1;

32) calculating the actual waveform data point of each rotation of the generator: the time for acquiring the eight-cycle waveform data is Ts, the number of actual waveform data points is M-Nxn, N is 1, 2, 3, … … 8, N is the number of sampling points per week, and the number of actual waveform data points N [ i ] of each rotation of the generator is as follows:

taking an integer;

33) and (3) waveform data correction: and (3) performing equal phase processing on the actual waveform data points N [ i ]:

when N [ i ] is more than N, the following redundant (N [ i ] -N) data are discarded;

and when the N [ i ] is less than N, interpolating the N [ i ] according to the N-point waveform data by using a linear interpolation method, so that the fixed sampling point number of each rotation of the generator is the waveform data sampling point number N in the step 2).

Background

When the water turbine generator set runs, main components of the water turbine generator set can generate vibration and swing of different degrees in the axial direction and the radial direction due to manufacturing and mounting processes, mechanical, hydraulic, electromagnetic and other reasons in the running process. The corresponding quantity of vibration and swing amplitude sensors are required to be installed according to the requirements of industry specifications so as to measure the vibration and swing amplitude, and the key phase sensor is required to be installed so as to measure the rotation period of the hydraulic generator. The vibration and swing amplitude sensor collects continuous periodic signals by taking the time of one rotation of the hydraulic generator as a period, and the key phase sensor collects pulse signals by taking one rotation of the hydraulic generator as a period.

The phase is an important parameter for analyzing the space axis, the oscillation amplitude, the oscillation influence quantity of the oscillation degree, the oscillation waveform frequency spectrum, the dynamic balance test, the counterweight test and the like of the hydraulic generator. At present, when the actual sampling is carried out on the oscillation waveform data by using the on-line monitoring system of the hydroelectric generating set, the method is generally as follows: the method comprises the steps of continuously collecting a vibration waveform signal by taking the time of one circle of rotation of a generator as a period, marking waveform key phase points in an equal period mode, generating a data packet in a mode of taking 8, 16 or 32 circles, and uploading the data packet to a server for data analysis. The waveform data acquisition based on the method is used for analyzing the hydraulic generator, and has the following problems:

(1) the initial phase of the waveform of the initial point of each rotation of the collected data packet is uncertain, although the average value and the peak value result of the vibration and the swing are not influenced, when the data related to the phase, such as the eccentric angle, the centrifugal force, the counterweight angle, the turning amount, the influence amount and the like of the hydraulic generator are analyzed, the data calculation result of each rotation will generate an error of 10-20%, and the accumulated error will gradually increase along with the increase of the number of the rotation cycles.

(2) Because the unit is a dynamic process in the actual operation, the rotating speed is not a constant value but a dynamic value, and the difference exists in the actual time of each rotation, if the key phase signal is marked as the time of one week according to the fixed rotating speed, the uploaded key phase signal offset time and the actual time have errors, the waveform calculation analysis result is influenced, sampling at the variable rotating speed causes the actual points of the waveform collected every week to be inconsistent, interference is brought to subsequent analysis, the condition index of the unit equipment is judged by mistake by operating personnel, the calculation data error greatly increases the mode adjustment workload during maintenance, prolongs the maintenance time, and influences the smooth operation of the maintenance plan. Therefore, there is a need for improvements in the prior art.

Disclosure of Invention

In order to effectively solve the problems in the prior art, the invention provides a method for acquiring waveform data based on a key phase sensor and processing the waveform data according to the offset time of the key phase and the number of the key phases, so that each acquired vibration and swing waveform data packet has the same initial phase, and each data packet has the same data sequence and the same phase interval. When the processed data is used for analyzing the running state index of the hydraulic generator, the accuracy can be greatly improved, and therefore technical support is provided for guiding the actual running and maintenance arrangement of the hydraulic generator set.

The invention is realized by the following technical scheme: a key phase-based equal phase processing method for a hydro-turbo generator set runout waveform is characterized by comprising the following steps:

1) installing a key phase sensor in the + X position of a main shaft of a generator, connecting the key phase sensor with various waveform data acquisition systems of the existing hydroelectric generating set, simultaneously arranging a concave or convex point on the main shaft of the generator opposite to the key phase sensor, taking the time point of the concave or convex point sweeping the key phase sensor when the generator rotates for one circle as the phase starting point of synchronous acquisition of various waveform data, recording the phase starting point as 0 degrees, recording the starting time of various waveform data as 0ms, marking a first key phase point, and marking a key phase point on the waveform data after every one circle of rotation, then:

the key phase point offset time is the current waveform data time-waveform data start time;

2) the data acquisition of the swing degree and the vibration waveform is continuous periodic cosine voltage or current signals, discrete sampling is carried out on the waveform signals by a pulse sampling function with the time of each rotation of the generator being T, the number of sampling points per week being N and the period being T, a swing degree or vibration waveform data sequence is generated for calculation, wherein the acquisition time interval of each swing degree or vibration waveform data point is as follows: T/(N-1), such that each acquired swing or vibration waveform data point has the same phase interval: 360 DEG/N, then generating a data packet by the acquired 1-8-week swing or vibration waveform data and uploading the data packet to a data server;

3) the server decompresses the data packet in the step 2), and the data packet is corrected according to the following steps:

31) correcting the number of key phases: the time deviation of the key phase offset time collected every revolution of the generator and the initial 0 point is as follows: t [ n ], n is 1, 2, 3, … … 8, and the number of bond phases is: n +1, and the deficiency is corrected to n + 1;

32) calculating the actual waveform data point of each rotation of the generator: the time for acquiring the eight-cycle waveform data is Ts, the number of actual waveform data points is M-Nxn, N is 1, 2, 3, … … 8, N is the number of sampling points per week, and the number of actual waveform data points N [ i ] of each rotation of the generator is as follows:

taking an integer;

33) and (3) waveform data correction: and (3) performing equal phase processing on the actual waveform data points N [ i ]:

when N [ i ] is more than N, the following redundant (N [ i ] -N) data are discarded;

and when the N [ i ] is less than N, interpolating the N [ i ] according to the N-point waveform data by using a linear interpolation method, so that the fixed sampling point number of each rotation of the generator is the waveform data sampling point number N in the step 2).

The data processed by the invention should have an error within 1% of the specified error when compared with the actually measured data.

The invention has the following advantages and effects: the method has important functions of analyzing the running state of the hydraulic generator and guiding the maintenance work arrangement of the hydraulic generator based on the vibration and the swing waveform data of the hydraulic generator. The traditional data acquisition and processing method causes great deviation to a calculation result, and actual operation and maintenance arrangement are easily misled. The invention improves data acquisition and data processing on the basis of the original method, realizes that the initial phase of each acquired waveform data is consistent, the waveform data sequence length is equal, the corresponding phase of each data point is the same, meanwhile, based on the data acquired by the invention, the error is within 1% compared with the actual measurement after calculation, maintenance personnel can be guided to accurately judge whether the unit operation index is good, the calculated data is used for guiding maintenance arrangement, a maintenance scheme is formulated in advance, the maintenance period is effectively shortened, the maintenance cost is saved, and the method has important significance for analyzing the hydraulic generator and guiding the maintenance work of the hydraulic generator and has good popularization value.

Drawings

FIG. 1 is a schematic diagram of a data acquisition mode before modification;

FIG. 2 is a schematic diagram of a data acquisition mode after modification;

FIG. 3 is a schematic view of a key phase sensor installation;

FIG. 4 is a graph of waveform data acquisition based on the method of the present invention;

FIG. 5 is a waveform of data acquisition prior to modification in an embodiment;

FIG. 6 is a waveform of data acquisition after modification of the embodiment.

In fig. 3, 1 is a generator main shaft, 2 is a concave point provided on the generator main shaft 1, and 3 is a key phase sensor mounted in the + X direction of the generator main shaft 1.

Detailed Description

The present invention will be further described with reference to the following examples.

Example 1

In this embodiment, the following acquisition and processing are performed on the swing waveform data of the upper guide bearing of the power plant No. 6 unit in the + X direction:

1) as shown in fig. 3, a key phase sensor 3 is installed in the + X direction of a generator main shaft 1, the key phase sensor 3 is connected with a swing waveform data acquisition system of the existing hydroelectric generating set, a concave point 2 is arranged on the generator main shaft 1 opposite to the key phase sensor 3, the time point of the concave point 2 sweeping the key phase sensor 3 when the generator rotates for one circle is used as the phase starting point of synchronous acquisition of the swing waveform data, the phase starting point is marked as 0 °, the starting time of the swing waveform data is marked as 0ms, a first key phase point is marked, and then, each circle of rotation is carried out, a key phase point is marked on the swing waveform data, then:

the key phase point offset time is equal to the current swing waveform data time-swing waveform data starting time;

2) the swing waveform data is collected as a continuous periodic approximate cosine voltage or current signal, discrete sampling is carried out on the waveform signal by a pulse sampling function with the period of T480 ms and the number of sampling points per week of the generator being 256 and the period of T480 ms, and a swing waveform data sequence (as shown in figure 4) is generated, wherein the collection time interval of each swing waveform data point is as follows: 480/(256-1) ≈ 1.882ms, with the same phase spacing for each acquired slew waveform data point: 360 DEG/256 is approximately equal to 1.406 DEG, and then the acquired 1-8-week swing waveform data is generated into a data packet and uploaded to a data server;

3) the server decompresses the data packet in the step 2), and the data packet is corrected according to the following steps:

31) correcting the number of key phases: the time deviation of the key phase offset time collected every revolution of the generator and the initial 0 point is as follows: t [ n ], n is 1, 2, 3, … … 8, and the time deviations of the key phase offset time of the first to eighth weeks and the initial 0 point actually collected at this time are as follows: t [1] ═ 480ms, T [2] ═ 960ms, T [3] ═ 1440ms, T [4] ═ 1921ms, T [5] ═ 2401ms, T [6] ═ 2882ms, T [7] ═ 3362ms, T [8] ═ 3842ms, and the number of bond phases: n +1 ═ 9;

32) calculating the actual swing waveform data point of each rotation of the generator: the time for actually acquiring the eight-cycle swing waveform data is Ts 3842ms, and the number of the actual waveform data points is M256 × 8 2048, so that the number of the actual waveform data points N [ i ] of each cycle of the generator is as follows:

taking an integer;

the data points of each week are calculated according to the formula as follows: n [1] ═ 255, N [2] ═ 255, N [3] ═ 255, N [4] ═ 256, N [5] ═ 255, N [6] ═ 256, N [7] ═ 255, and N [8] ═ 255

33) And (3) waveform data correction: and (3) carrying out equal phase processing on the actual swing degree waveform data points N [ i ]:

when N [ i ] is more than 256, the following redundant (N [ i ] -256) data are discarded;

and when the N [ i ] is less than 256, interpolating the N [ i ] according to the swing waveform data of 256 points by using a conventional linear interpolation method, so that the fixed sampling point number of each rotation of the generator is the waveform data sampling point number N of the step 2) to be 256 points.

By the method, the initial phases of the acquired waveform data are the same and are all 0, the time of each rotation of the generator set is the same as the actual time and is 480ms, the number of swing waveform data points is equal and is 256, and the data phase of each corresponding point is the same;

and (3) obtaining complete and correct swing waveform data of the generator per rotation cycle through the steps 1) -3) for calculating and analyzing the operation indexes of the hydraulic generator, so that the calculation result precision can be greatly improved, and the calculation error is reduced.

Taking the first cycle swing waveform data (see table 1) collected in the embodiment 1 as an example, the corresponding upper guide bearing swing eccentric angle error value is calculated according to the conventional method to illustrate the real effectiveness of the method, and the method specifically comprises the following steps:

1) because the swing-degree waveform data collected and decompressed in the first week of the embodiment 1 is 255 points, according to the step 33 in the embodiment 1), when the data points are less than 256, 255 points need to be interpolated according to 256 points by using a conventional linear interpolation algorithm, and after interpolation, swing-degree waveform data obtained on the 256 th point is 153.8, which is shown in table 1;

TABLE 1

2) Calculating the average value of the throw of the upper guide bearing conventionally

The average value of the swing waveform of the upper guide bearing is 0, the characteristic of a cosine function is met, and the average value of equal-interval sampling within one circle is 0;

3) calculating the peak-to-peak value (maximum-average value) of the swing waveform of the upper guide bearing conventionally:

the average value and the peak value of the swing waveform data are equal to those of the original swing waveform data, and the swing waveform data are consistent with the swing waveform average value and the swing waveform peak value before and after the processing described in the background technology of the invention without influence;

4) calculating the component value of each point location swing degree waveform value in the X, Y direction according to the following formula:

5) conventionally, the swing influence quantity of the leading bearing in the X, Y direction is calculated:

6) conventionally obtaining an upper guide bearing swing eccentric angle:

7) carry out actual measurement when the unit overhauls, it is this to go up guide bearing throw off-centre angle to obtain: 32.1 ° and the error is calculated according to the following deviation calculation formula:

whereinRepresents the calculation result value, X represents the actual measurement value, and Y represents the measurement range; the collected first cycle throw waveform data is calculated, the error between the throw eccentric angle of the upper guide bearing and actual measurement is 0.86%, the deviation is within 1%, and the deviation belongs to a controllable range.

Comparative example

Directly collecting the swing waveform data of the upper guide bearing rotating for one circle in the + X direction based on the prior art is shown in FIG. 5, and the waveform data points are 256 points (see Table 2):

table 2:

the swing eccentricity angle of the upper guide bearing calculated by a conventional method is 67.76 degrees, the eccentricity angle of the upper guide bearing is actually measured to be 32.1 degrees during unit maintenance, and the calculation is carried out according to the following deviation calculation formula:

the error is 10.41%, which far exceeds the specification of 1%.

As demonstrated by the above example 1 and comparative example: the waveform sampling data of the hydraulic generator processed by the method is used for calculation and analysis, the error of the calculation result is 0.86 percent and is within the specified error range of 1 percent, while the error of the comparative example is 10.41 percent and far exceeds the specification of 1 percent. Thus, the method is effective, accurate and reliable.