JPH08248085A - Load cut-off test analyzing system - Google Patents

Abstract

PURPOSE: To shorten the analyzing time of a trend graph, to reduce the analyzing error and to provide measured data having no personal difference. CONSTITUTION: A sensor 10 detects the varied value of a generator, processes it by a load cut-off testing unit 20, and stores it in a memory recorder 30. The data is stored in a floppy disk 40, and input to a personal computer 50. A trend graph 61 is formed in the computer 50, and the graph 61 is data analyzed. A test result table 62 in which the graph 61 and the data analyzed result are described is printed by a printer 60.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a load shedding test analysis system, and more particularly, to a load shedding test analysis system which can shorten time, reduce errors, and obtain measurement data without individual differences.

[0002]

2. Description of the Related Art When a load is cut off due to a power system accident during operation of a generator, a load cut-off test of a water turbine generator at a hydroelectric power plant shows the rotation speed of the water turbine, the generator voltage, and the water pressure in the waterway. This is done for the purpose of confirming that the fluctuation value does not exceed the guaranteed value and that the turbine and generator can be safely transferred to no-load operation. The breaking load is usually 1/4 of the maximum output, 2 / The value is increased to 4, 3/4, 4/4, and if the fluctuation value exceeds the guaranteed value, adjustment is required.

In the same test, in addition to the above fluctuation values, transient states such as the servo motor stroke are measured and recorded by a recorder and its peripheral equipment, and the fluctuation values are read from the recording to calculate the fluctuation rate and the like. Up until now, many devices such as electromagnetic oscilloscopes and dynamic strain measurement devices have been placed, wired and connected, and their transportation,
A lot of labor was required for setting work, and skill and labor were also required for actual test measurement, record organization, and report writing.

FIG. 12 shows a conventional load cutoff test system, which has a sensor 10, a load cutoff tester 20, a memory recorder 30, and a printer 60. The sensor 10 includes a generator voltage detector 11 that detects the voltage of a generator (not shown) and a generator current detector 12 that detects the current of the generator.
And a pressure converter 13 that outputs an iron pipe (water channel) water pressure signal
And a displacement converter 14 that outputs an opening (hereinafter also referred to as a stroke as a synonym) signal such as a guide vane / deflector / nozzle / pressure suppressor. The load cutoff test device 20 receives a voltage signal from the generator voltage detection unit 11 and outputs a rotation speed (frequency) signal and a DC voltage signal of a water turbine, and a voltage deviation detection unit / frequency deviation detection unit 21a, 21b, and a generator. Dynamic strain measuring device 2 that amplifies and outputs the output signals of the current detection unit 12, the pressure converter 13, and the displacement converter 14 as a dynamic strain value.
2 The signal output from the load interruption test apparatus 20 is stored in the memory recorder 30, and the printer 60 prints out the trend graph 61 based on the stored signal.

FIG. 13 shows a trend graph printed out in this manner. Although this trend graph is that of the present invention, it is conventionally printed out without any display in the graph.

FIG. 14 shows the first closing time T ₁ and the equivalent closing time T _c which are particularly important as an example in the guide vane opening of the trend graph of FIG. A field engineer analyzes this trend graph and adjusts the fluctuation value so that it does not exceed the guaranteed value, and prepares for load shedding.

[0007]

However, according to the conventional load shedding test system, the reading of the obtained trend graph causes, for example, individual differences a and b, and the first closing time is equivalent to T _1a and T _1b . There is an inconvenience that the closing time becomes T _ca and T _cb , which causes a difference. In addition, since a ruler is used to write an auxiliary line and the trend graph is read, it takes time, and a reading difference occurs depending on the thickness of the auxiliary line or the printed line.

SUMMARY OF THE INVENTION Therefore, an object of the present invention is to provide a load shedding test analysis system capable of shortening the analysis time of a trend graph, reducing errors, and obtaining measurement data having no individual difference.

[0009]

In order to achieve the above object, the present invention achieves the above-mentioned objects such as the rotational speed of a water turbine, a turbine, etc., the generator voltage, etc. even if the load supplied to the generator is cut off. In a load interruption test analysis system for confirming that the variable value can safely shift to no-load operation without exceeding the guaranteed value, a detecting means for detecting the variable value and a change according to the passage of time of the variable value are provided. There is provided a load interruption test analysis system including signal processing analysis means for creating a trend graph to be represented and for analyzing data of the trend graph, and output means for outputting the trend graph and the result of the data analysis.

[0010]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 shows an embodiment of the present invention, and the same parts as those in FIG. In this load shedding test analysis system,
The output of the memory recorder 30 is stored in a floppy disk 40 and input to a personal computer (hereinafter referred to as computer) 50. The computer 50 stores an input interface 51 for inputting data, a CPU 52 for signal processing, a ROM 53 storing a program for processing data, a program for creating a test report, etc., and input data, signal processing results, etc. RAM 54
And an output interface 55 for outputting trend graph creation data, trend graph reading values (measured values), and analysis values. A printer 60 on the computer 50
Is connected to the printer 60, and the printer 60 displays the trend graph 61 and the trend graph 61 according to the output data.
The test result table 62 including the read value (measured value) of is printed out.

FIG. 2 shows an analysis of a stationary time such as a guide vane stroke and a closing time as an example of a signal processing flow of the computer 50, and each of them will be described below. (1) Waveform smoothing The waveform captured as measurement data contains noise in addition to the regular waveform. When this waveform is directly processed as data, noise peaks are often set as maximum values or inflection points (points at which the waveform bends), and the measurement results lack accuracy. In order to prevent this, as shown in FIG. 3, smoothing (moving average) is performed before measurement to reduce the influence of noise. Of the data for the smoothing width, the maximum and minimum data are deleted and the rest is averaged to obtain the data value at the center position of the width. By deleting the maximum and minimum values, it is possible to reduce the effect of noise-containing waveforms. Further, regarding the relationship between the smoothing width and the influence degree of noise, as shown in FIG. 4, as the smoothing width is widened, the influence degree of protruding data such as noise becomes smaller. However, if the width is widened, the shape is different from the source waveform (original waveform), so the width cannot be expanded more than necessary. In determining the smoothing width, the number of data is set by referring to the actual test data. For the current waveform, this smoothing process is not performed because the breaking point is detected. (2) Detection of load cutoff point Basically, the point where the current value becomes zero is the cutoff point. However, since the current signal is AC, 0 is repeated before the cutoff. Moreover, even if noise is included after the cutoff, it does not become 0 completely. For this reason, as shown in FIG. 5, when all the data of the sampling number set by setting the threshold value which is regarded as the current value 0 falls within the range of the threshold value, the first data position is set as the cutoff point. (3) Measurement before shutoff and maximum value of each item For preshutdown values such as voltage, rotation speed, iron pipe water pressure, etc.
The data subjected to the [waveform smoothing] processing is sampled as data [value before interruption] 0.5 seconds before the interruption. Also, the maximum data from the time after the interruption until the stability signal is turned on is [maximum value]. The volatility is

[Equation 1] Is calculated as (4) Inflection point and rate of change of waveform When measuring immobile time or closing time, the position of the inflection point must be detected as data. The rate of change of the waveform is used to detect the inflection point as data from such a waveform.

FIG. 6 shows a method of calculating the rate of change, where a1, a2 and a3 are three data sampled at regular intervals. At this time, the change rate of a2 is
[Change rate of a2 = (y2-y1) / K]. Also,
The range of the rate of change is the interval between the three data points sampled. When this width is widened, the slope of the waveform is measured in a wide range, and an inflection point can be obtained even with a waveform having a gentle change. The rate of change width is adjusted by self-tuning according to the slope of the waveform.

FIG. 7 shows the change rate of the movement of the guide vane as an example. In obtaining τ · T1 · T2, the maximum point and the minimum point of the rate of change are set as the inflection points of the waveform, which are the maximum point and the minimum point, respectively. Further, in actual use, in order to distinguish from an inflection point due to noise, each inflection point is given a condition and one that satisfies the condition is defined as an inflection point.

The movement of each inflection point and the conditions for establishment thereof are as follows. An inflection point [used to determine the end point of τ and the position of the reference line of Tc. ] S and E are the inflection point start point and inflection point, respectively, which are within the 0 threshold value. An inflection point is set when the following four conditions are met at the same time. Conditions a. The minimum value is large. I. The stroke at that time is 75% or more of the maximum value. C. The stroke before S is horizontal (within the set threshold). D. The stroke after E descends to the right. Inflection point [Used to determine the end point of T1 and the position of the TC reference line. ] S and E are the inflection point start point and end point, respectively, which are within the 0 threshold value. An inflection point is set when the following four conditions are met at the same time. Conditions a. The maximum value is large. I. Is located after. C. The stroke is less than the stroke. D. The stroke after S descends to the right. Inflection point [Used to determine the end point of T2. ] S and E are the inflection point start point and end point, respectively, which are within the 0 threshold value. An inflection point is set when the following three conditions are met at the same time. Conditions a. The maximum value is large. I. Is located after. C. The stroke is less than the stroke. D. The stroke after S descends to the right. (5) Immobility time (τ) Immobility time is the time from when the load is cut off until the guide vanes and deflectors start moving. Two methods were examined for measuring the immobility time.

The first proposal is a method of treating a point where the stroke changes (end point of the immobile time) as a point where the stroke has decreased to 95% of the stroke before the change. However, this method cannot raise the judgment of change from 95% or more because it is necessary to distinguish it from noise. Therefore, the point of 95% in the case of a waveform having a gentle slope as shown in FIG. This result was rejected because the test results showed that it was far from the end of time.

The second method is a method of utilizing the rate of change. FIG. 9 shows the relationship between the immobile time and the rate of change. This Figure 9
The rate of change at [Time 39] is 20 before and after that time.
There was a time difference of (half of rate of change width 40) [Time 19]
And the difference between the slopes of [time 59]. This means that "the rate of change is 0 if the rate of change is in a straight line," in other words, "the end point of the straight line [the end of the immobility time] is smaller than the point where the rate of change deviates from 0. It's only half the point behind. " However, in determining whether or not the rate of change is 0, it is necessary to set a 0 threshold value so as to have a range in 0 in order to prevent erroneous determination due to the influence of noise or the like.

When an experiment was carried out using actual test data by this method, the result was also good, so the immobility time was measured here by the second alternative. (6) First Closing Time (T1) T1 is the time taken for the first linear closing movement of the guide vane deflector.

FIG. 10 shows the relationship between the first closing time and the change rate, but the change rate at the inflection point becomes maximum when the straight line portion ends and the next new change (straight line or curved line) is reached. And T1 can be obtained by measuring the time from the end point of the immovable time previously obtained to the inflection point. (7) Second closing time (T2) T2 is the time taken by the guide vane deflector to perform a linear closing operation after T1. Based on this, in the present embodiment, T2 is defined as follows. A. The time at which the end point (inflection point) of T1 is the start point and the inflection point is the end point is measured. However, if any of the following conditions is satisfied, it is considered that T2 does not exist, and calculation and display are not performed. I. The stroke becomes 0 near the inflection point E. (Only T1 was closed.) C. The inclination of the stroke became horizontal between the inflection points S and E. (The stroke is constant or changes to the opening direction.) (8) Closing time (Ts) The time from the end of the immobility time of the guide vane deflector nozzle to the fully closed position of the stroke is measured. If the stroke is not fully closed, calculation / display will not be performed. (9) Equivalent Closing Time (Tc) FIG. 11 shows the relationship between the rate of change and TC. Reference point A
Represents the inflection point E, and is a position advanced from the starting point of the straight line portion of T1 by the number of data of half the change rate width. One reference point B, which represents the inflection point S, is a position that has returned from the end point of the straight line of T1 by the number of data that is half the change rate width. Let C be the point of the opening on the straight line connecting the two reference points before interruption, and D be the point at which the other opening becomes 0. TC is the time measured between C and D. (10) Suppressor opening time (To) Measure the time between two points starting from the end of the fixed time and ending at the point where the stroke is maximum and the rate of change is zero. By adding a condition that the change rate is 0 to the end point, pulse-like noise (the change rate does not become 0 even if the stroke is large) is distinguished from the end point.

In this way, the trend graph shown in FIG. 13 is obtained, and the test result table in which the data in the trend graph is described is obtained. Although the hydroelectric power generation equipment has been described in the above embodiments, it can be used for thermal power generation equipment, nuclear power generation equipment, and the like. Table 1 shows a test result table.

[Table 1]

[0020]

As is apparent from the above description, the system of the present invention has the following effects as compared with the conventional system. (1) The error in the measurement signal data processing process is reduced. Since the computer analyzes the measurement data, the error caused when reading the trend graph can be eliminated. (2) Even beginners can obtain the same measurement data as experienced users. In the conventional system, when the guide vane opening does not move linearly, the measurer has difficulty in drawing an auxiliary line for obtaining the closing times Tc, T1, and T2. In the system of the present invention, since the auxiliary line is drawn according to a certain law to determine the closing time, there is no room for the individual difference of the measurer, and even a beginner can obtain the same measurement data as an expert. (3) The analysis time can be shortened because the analysis is performed inside the computer without the need for the engineer to perform the analysis.

[Brief description of drawings]

FIG. 1 is a block diagram showing an embodiment of the present invention.

FIG. 2 is a flowchart showing the operation of the present invention.

FIG. 3 is an explanatory diagram showing smoothing according to the present invention.

FIG. 4 is an explanatory diagram showing the degree of influence of smoothing width and noise of the present invention.

FIG. 5 is an explanatory diagram showing detection of a load cutoff point according to the present invention.

FIG. 6 is an explanatory diagram showing calculation of a change rate according to the present invention.

FIG. 7 is an explanatory diagram showing a guide vane opening degree and a rate of change according to the present invention.

FIG. 8 is an explanatory diagram showing a fixed time end point according to the first plan.

FIG. 9 is an explanatory diagram showing the immobility time and the rate of change of the present invention.

FIG. 10 is an explanatory view showing the relationship between T1 and T2 of the present invention and the rate of change.

FIG. 11 is an explanatory diagram showing measurement of an equivalent closing time according to the present invention.

FIG. 12 is a block diagram showing a conventional load shedding test system.

FIG. 13 is an explanatory diagram showing a trend graph by a conventional system.

FIG. 14 is an explanatory diagram showing a reading error of a trend graph.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 10 sensor 11 generator voltage detection unit 12 generator current detection unit 13 pressure converter 14 displacement converter 20 load interruption tester 21 voltage / deviation detection unit 22 dynamic strain measurement device 30 memory recorder 40 floppy disk 50 personal computer 60 printer 61 Trend Graph 62 Test Report

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Hiroshi Yamada 5-18 Shimizu, Toyama City (72) Inventor Shoichiro Takada 2994-3 Hamakurosaki, Toyama City

Claims (1)

[Claims] 1. Even if a load is cut off in a load to which the generator is supplying electric power, fluctuations in the rotational speed of the water turbine, turbine, etc., generator voltage, etc. do not exceed the guaranteed values and are safe for no-load operation. In a load shedding test analysis system for confirming that a transition can be made, a detecting means for detecting the fluctuation value, and a trend graph representing a change in the fluctuation value according to the passage of time are created, and data analysis of the trend graph. A load cutoff test analysis system, comprising: a signal processing analysis unit that performs the above; and an output unit that outputs the trend graph and the result of the data analysis.