JP2702206B2 - Operating status monitor for hydraulic machinery - Google Patents

Description

Description: TECHNICAL FIELD The present invention relates to an operating state monitoring device for a hydraulic machine.

[Conventional technology]

The permissible monitoring value in the conventional hydraulic machine operation monitoring device and its operating condition monitoring method is a value uniquely determined based on the maximum value actually measured under all operating conditions of the machine, and is used to protect an abnormal state. He had a strong personality. Japanese Unexamined Patent Publication No. Sho 56-129524 relates to this.

[Problems to be solved by the invention]

A plurality of types of measurement values (measurement data) indicating an operation state often have a correlation with each other. In such a case, a specific measurement value to be monitored (this measurement value is defined as a monitoring target item value or a monitoring target amount) is replaced by one or more types of other monitoring values affecting the monitoring target amount. It can be expressed as a function of the measured value (this measured value is defined as the influence item value).

However, as in the past, monitoring tolerances that were uniquely determined ignoring changes in other correlated measurement values (effect item values) can detect the initial stage (signs) of abnormalities and predict future predictions. And it has not been considered to set a fine monitoring allowance suitable for the characteristics of each hydraulic machine.

The present invention has been made in view of the above points, and an operation state monitoring device for a hydraulic machine that performs an operation state monitoring in detail in accordance with the secular change (time-dependent change) of the characteristic of the hydraulic machine and enables prediction in the future. To provide.

[Means for solving the problem]

In order to achieve the above object, the present invention provides a measurement value (monitored item value) to be monitored for an operating state of a hydraulic machine.
Is compared with the monitoring allowable value, and in the operating state monitoring device of the hydraulic machine that outputs a signal indicating abnormality when the value of the inspection target item exceeds the monitoring allowable value, the monitoring target item value and the monitoring target item value are affected. Input means for sequentially inputting other types of measured values (influence item values) serving as factors to be given as measurement data; measurement data storage means for storing these input monitor target item values and influence item values; A correlation coefficient indicating a correlation between the monitoring target item value and the influence item value is calculated using the measurement data group stored in the measurement data storage means, and the correlation coefficient is sequentially input to be updated. A correlation coefficient calculating means which is set so as to be recalculated with the passage of time using the measured data group of the monitored item value and the affected item value, and a phase for storing the calculated correlation coefficient. Relationship number A storage unit; and a monitoring allowable value calculating unit configured to establish a function formula for calculating the monitoring allowable value by using the stored correlation coefficient and to calculate the monitoring allowable value in an updatable manner.

[Action]

With the provision of the above means, the operating state can be monitored in detail including the secular change with a monitoring allowable value suitable for the characteristics of each hydraulic machine, and the characteristics of the operating state can be grasped and predicted.

This will be described in more detail. FIG. 4 shows the relationship between the monitoring target item value Q on the vertical axis and the influence item value P affecting the Q on the horizontal axis.

In the figure, M is a point group having values of P and Q measured simultaneously at an arbitrary time and placed on a graph. The measured point group M is approximated to a curve Qv by the least square method or the like. Specifically, assuming that Q _v = q ₀ + q ₁ P + q ₂ P ² + q ₃ P ³ (1), the coefficients q _{0 to} q ₃ are obtained by the least square method or the like.
Is obtained. Here, q _{0 to} q ₃ are the monitored item values Q ·
It is a correlation coefficient indicating a correlation between the influence item values P.

The reason why the measured point group M fluctuates without being on the curve Qv is not only the measurement error at the time of measurement, but also the influence item value P affecting the monitoring target item value Q is not only P. ,
Because several other items affect Q,
By inserting a plurality of items affecting the right side of the equation (1) and making it a multivalent function, the variation is reduced,
A highly accurate curve Qv can be obtained.

A monitored item values Q were measured by any affected item value P in FIG. (1) and E ₁ the difference between the curve Qv determined by the formula, and E _1max the maximum value of E ₁ for each measurement point I do.

E ₂ is the empirical margin in consideration of the measurement error of the entire monitoring device, plus the E _1max and E ₂ are tolerance E. Monitoring tolerance curve Q _a for the monitored item value Q is least squares with becomes plus margin E curve Qv that _{approximates, Q a = E + Qv =} E + q 0 + q 1 P + q 2 P 2 + q 3 P 3 ...... (2) It is shown by. Equation (2) represents the upper limit of the monitoring target quantity (monitoring target item value) Q with respect to the influence item value P during the past normal operation, and is a state monitoring for finding an operating state deviating from the past operating state. The original monitoring can be realized.

Incidentally Q _c of the figure shows the monitoring tolerance monitoring method according to the conventional constant monitoring value (fixed value), for example, monitoring 'monitored item value Q in' in FIG △ mark display effect field values P Tolerance
Symptoms less than qc, thus resulting in overlooked as normal, but the curve Q _a according to the present invention, greater than _a 'monitoring the amount Q of' P on the curve Q _a, thus that could not happen in the past An abnormality can be found as an error, and the occurrence of an abnormality can be found at an early stage, and the state change can be clearly understood.

Furthermore, the correlation function q ₀ to q ₃ are sequentially input as the measurement data group (monitored item values, affected item value) because we are updated, even functional expression monitoring tolerance Q _a use of this correlation function update Therefore, the monitoring allowable value can be set according to the secular change (temporal change) of the hydraulic machine to be monitored, and the future trend can be predicted. That is,
As the monitoring period becomes longer, it becomes possible to set a monitoring tolerance value most suitable for the characteristics of each machine.

Further, the measurement data group is increasing as monitoring is continued, and the storage capacity for storing the measurement data becomes enormous. In order to cope with such a problem, the measurement data stored in the monitoring amount storage device at the time when the correlation coefficient of the approximate expression is obtained from the measurement data by the above equation (1) may be deleted.

〔Example〕

Hereinafter, the present invention will be described based on the illustrated embodiments.
1 to 3 show one embodiment of the present invention.

As shown in FIG. 1, a hydraulic machine, for example, a device for monitoring the operating state of a turbine main bearing temperature inputs a plurality of types of measured values (monitored item values, influence item values) as signals (input device 1). ), A measurement data storage means (monitoring amount storage means 2) for storing the input measurement data, and compares the monitored item value and the permissible value under the current operating conditions, and when the monitored item value exceeds the permissible value, A means (comparing device 3) for outputting a signal indicating an abnormality is provided.

In the operating state monitoring device thus machined, in the present embodiment, the monitoring target item value / influence item value is obtained from the measurement data (monitoring target item value, influence item value) group stored by the monitoring amount storage device 2. correlation coefficient showing the correlation between the correlation coefficient calculating means for calculating a (correlation coefficient calculation unit 4) [the above (1) coefficient q ₀ to q ₃ in formula] the computed for monitoring A correlation coefficient storage means (monitoring coefficient storage device 5) for storing a correlation coefficient, and a function formula for calculating a monitoring allowable value using the stored correlation coefficient is established to perform monitoring at a predetermined time. A monitoring allowable value calculating means (monitoring allowable value calculating device 6) for calculating the allowable value is formed. Among them, the correlation coefficient calculation device 4 is recalculated as time elapses using a new measurement data group of the monitoring target item value and the influence item value which are sequentially input in order to update the correlation coefficient. It has been set.

The monitoring allowable value calculating device 6 calculates the monitoring allowable value by using the above-mentioned correlation coefficient by, for example, the function formula of the above-described equation (2), and this calculation is stored at a predetermined time and before. From the measured data, the data in the range of the operating state where there was no abnormality was taken out, and based on the data, the allowable monitoring value after that point was set (calculated), and the operating state of the machine was monitored. By doing so, the operating state can be monitored with the monitoring allowable value suitable for the characteristics of the turbine main bearing, and the characteristic of the operating state that changes with time can be grasped.
Prediction becomes possible, and detailed monitoring of the operating state of the hydraulic machine can be realized.

For example, as shown in FIG. 2, a bearing cooling water temperature B, a bearing ambient temperature C, and a generator output D are selected as factors influencing the state of the turbine wheel bearing temperature A, and the turbine wheel bearing temperature A is selected. determining the allowable value (curve) a _a. First, as described above, a curve corresponding to the above-described curve Qv of the water turbine bearing temperature A
The av, with the bearing cooling water temperature B, the bearing ambient temperature C, and generator output D, for example, _{Av = a 0 + a 1 B} + a 2 B 2 + a 3 B 3 + a 4 C + a 5 C 2 + a 6 D + a 7 D 2 + A ₈ D ³ ………………………………………………………………………………………………………………………………………………………… (3) A _{0 to} a _{8 by} least square method
Is obtained (correlation coefficient). The curve A _a is an allowable amount of monitoring based on the curve Av, determined by the following equation by adding the aforementioned margin E.

_{A a = E + a 0 +} a 1 B + a 2 B 2 + a 3 B 3 + a 4 C + a 5 C 2 + a 6 D + a 7 D 2 + a 8 D 3 ...... (4) i.e. hydraulic turbine bearing temperature A inputted sequentially, the bearing cooling water The measured data of the temperature B, the bearing ambient temperature C, and the generator output D are continuously stored in the monitored amount storage device 2. Then, when an instruction to update the monitoring allowable value of Av is issued, the correlation coefficient calculating device 4 uses the measurement data of the normal operation state from the measurement data to obtain (4)
Request E and a ₀ ~a ₈ expression. The calculation of the approximation coefficient by the correlation coefficient calculation device 4 does not need to be performed at all times.
It may be executed at an interval such as once / day or once / two days, so it is executed only when an update command is issued.

The monitoring allowable value calculating device 6 receives the cooling water temperature B, the bearing ambient temperature C, and the generator output D, and E and a _{0 to} a received from the correlation coefficient calculating device 4 via the monitoring coefficient storage device 5. ₈
Substituted Doo in the equation (4), it calculates the current cooling water temperature B, the bearing ambient temperature C, and monitoring tolerance A _a bearing temperature on the generator output D. Comparator 3 compares the bearing temperature A which is measured and monitored tolerance A _a, and outputs a signal indicating an abnormality when the bearing temperature A had exceeded the monitoring tolerance A _a.

Further, as shown in FIG. 3 in which the vertical axis indicates the turbine bearing temperature A, and the horizontal axis indicates the cooling water temperature B, the bearing ambient temperature C, and the generator output D, the value for one year is shown. in measurement data sought (3) of the curve Av, by illustrated as _{Av 1, Av 2, Av 3} ... every year, it is possible to quantitatively confirm the degree of aging of the machine.

In FIG. 1, 8 is a device for detecting a turbine wheel bearing temperature A, 9 is a device for detecting a bearing cooling water temperature B, 10 is a device for detecting an ambient temperature C of a bearing portion, 11
Is a device for detecting the load D of the generator.

As described above, according to the present embodiment, a plurality of other types of items (influence items) that affect a change in one monitored amount (monitored item) are set for each operation condition, and the measurement is performed during normal operation.・
Each measurement data stored (monitored item value, impact item value)
Is used to calculate the correlation (correlation coefficient) between the monitored item value and the affected item value using an approximation formula, add a margin to the correlation, and monitor as an allowable monitoring value. The operating state can be monitored with the allowable value.

The monitoring allowable value is calculated at regular intervals, and the monitoring allowable value is updated, so that the longer the monitoring period is, the more the monitoring allowable value most suitable for the characteristics of each machine is obtained. Thus, a monitoring device that can be called a learning type is obtained.

In the above embodiment, when the coefficients a _{1 to} a ₈ are obtained by the correlation coefficient calculating device 4, the measurement data for each period may be weighted. That is, for example, one point is assigned to the data in the monitoring start period, and ten points to the most recent data, and the data in the period during the period is assigned a weight proportional to the time series. Thus the coefficient a ₁ ~a ₈ which data obtained on the basis of which weighted becomes a value reflecting good recent operating conditions, as monitored by the monitoring tolerance reflecting good recent operating conditions Become.

That is, the characteristics of the operating state of the machine change over time. Therefore, if all of the long-term measurement data is used in the calculation for obtaining the approximate expression with the same weight, the average value will be an average over the entire period, and there is a risk that a monitoring allowable value that is far from the current operating state obtained over time may be set. However, by performing the above, a monitoring allowable value that reflects the recent operating state well is obtained.

Further, when the coefficients a _{1 to} a ₈ are obtained by the correlation coefficient calculation device 4,
Determined coefficient a ₁ ~a ₈ obtained by the measurement data of the last period, the average of the coefficients a ₁ ~a ₈ that has been used up to the previous time, determined coefficients
a a ₁ ~a ₈ was to be used to monitor after the current time. Even in this case, it is possible to obtain a monitoring allowance that well reflects recent operating conditions.

〔The invention's effect〕

INDUSTRIAL APPLICABILITY As described above, the present invention can represent an operating state monitoring device for a hydraulic machine that enables the characteristics of the operating state of the hydraulic machine to change with time and a prediction thereof in the future, thereby enabling fine monitoring.

[Brief description of the drawings]

FIG. 1 is an explanatory view of an embodiment of a hydraulic machine operating state monitoring apparatus according to the present invention, and FIG. 2 is a hydraulic turbine bearing temperature A of the same embodiment.
FIG. 3 is a characteristic diagram showing the relationship among the turbine cooling water temperature B, the bearing ambient temperature C, and the generator output D. FIG. FIG. 4 is a characteristic diagram showing a relationship between a water temperature B, a bearing ambient temperature C, and a generator output D. FIG. 4 is a characteristic diagram showing a relationship between a monitored item value and an affected item value. 1 ... input device (measurement data input means) 2 ... monitoring amount storage device (measurement data storage means) 3 ... comparison device (abnormal signal output means) 4 ... correlation coefficient calculation device (phase Relationship number calculation means), 5... Monitoring coefficient storage device (correlation coefficient storage means), 6... Monitoring allowable value calculation device (monitoring allowable value calculation means).

Claims (6)

(57) [Claims] 1. A monitored value of an operation state of a hydraulic machine (hereinafter, this measured value is referred to as a monitored item value) is compared with a monitoring allowable value, and the monitored item value exceeds the monitoring allowable value. An operating state monitoring device for a hydraulic machine that outputs a signal indicating an abnormality at the time, wherein the monitored item value and another type of measured value that is a factor affecting the monitored item value (hereinafter referred to as an affected item value) ) Are sequentially input as measurement data, measurement data storage means for storing these input monitored item values and influence item values, and measurement data groups stored in the measurement data storage means. Calculating the correlation coefficient indicating the correlation between the monitored item value and the affected item value, and measuring the monitored item value and the affected item value which are sequentially input in order to update the correlation coefficient. De Correlation coefficient calculating means set so as to be recalculated with time using a data group; correlation coefficient storage means for storing the calculated correlation; and storing the stored correlation coefficient. An operating condition monitoring device for a hydraulic machine, comprising: a monitoring allowance value calculating unit that establishes a function formula for calculating the monitoring allowance value by using the monitoring allowance value and calculates the monitoring allowance value in an updatable manner. 2. The method according to claim 1, wherein the effect item value is one or a plurality of types, and the monitoring allowable value calculating means sets a margin for an approximate value of the monitored item value calculated from the effect item value and the correlation coefficient. The operating state monitoring device for a hydraulic machine according to claim 1, wherein the monitoring allowable value is calculated by adding. 3. The operating state monitoring device for a hydraulic machine according to claim 1, wherein the monitoring allowable value calculating means calculates and updates the monitoring allowable value at regular intervals. 4. The measurement data stored in the measurement data storage means is set so as to be deleted when the correlation coefficient is calculated. Operating condition monitoring equipment for hydraulic machinery. 5. The monitoring allowable value is calculated by giving a larger weight to the measurement data of the old period than to the measurement data of the new period and to the measurement data of the new period to the measurement data of the old period. The operating state monitoring device for a hydraulic machine according to any one of claims 1 to 4, wherein the operating state monitoring device is a monitoring device. 6. The monitoring allowable value calculating means calculates a monitoring allowable value at the end of a predetermined period in the past by using the measurement data within the period, and calculates the monitoring allowable value and the current monitoring allowable value. The operating condition monitoring device for a hydraulic machine according to any one of claims 1 to 5, wherein an average value is taken as the monitoring allowable value for the next period.