CN115854258A - On-line Inspection Method for Steam Pipe Network Leakage Points Based on Time Series - Google Patents

Abstract

The invention discloses a steam pipe network leakage point online inspection method based on a time sequence, which comprises the following steps: acquiring multi-source sensing data acquired by all steam pressure sensors on a steam pipe network to obtain an original sensor data set; respectively carrying out data correlation detection and time sequence continuity detection on multi-source sensing data of the same steam pressure sensor in an original data set to obtain a data correlation detection abnormal set omega m and a time sequence continuity detection abnormal set omega c; fusing a data correlation detection abnormal set omega m and a time sequence continuity detection abnormal set omega c; and determining leakage point information in the steam pipe network according to the fusion result. The steam pipe network leakage point online inspection method based on the time sequence provided by the invention utilizes a multi-source sensing data abnormity detection strategy based on the time sequence to carry out intelligent inspection on steam pipe network equipment, detects and analyzes the operation state of the steam pipe network equipment at any time, and judges the future development and operation state of a steam pipe network equipment system.

Description

On-line Inspection Method for Steam Pipe Network Leakage Points Based on Time Series

technical field

The invention relates to the technical field of improving steam pipe network equipment, in particular to an online inspection method for steam pipe network leakage points based on time series.

Background technique

At present, cigarette companies do not need steam energy all the time. Once there is a problem with the steam pipe network system, it will have a great impact on the production of the company. With the development of society, the demand for steam energy by cigarette enterprises is increasing, and the scale of production and construction is also increasing. During the operation of steam pipe network equipment, some equipment safety accidents may occur, which will cause a large amount of energy. waste. Therefore, it is very important to strengthen the inspection work and inspection quality of the steam pipe network.

Steam pipe network inspection is an important link in the entire production system, which can ensure the smooth operation of kinetic energy equipment. With the continuous expansion of cigarette production scale, the workload of pipe network inspection also increases accordingly. In the current patrol inspection of steam pipe network equipment, fault judgment mainly relies on the on-site inspection and work experience of operation and maintenance personnel. Difficult to find faults and hidden dangers in time.

Therefore, there is an urgent need for an online inspection method for leaks in steam pipe networks based on time series.

Contents of the invention

The purpose of the present invention is to provide a time-series-based online inspection method for steam pipe network leaks to solve the above-mentioned problems in the prior art, and use the time-series-based multi-source sensing data anomaly detection strategy to realize the inspection of equipment Intelligent patrol inspection can detect and analyze the operation status of steam pipe network equipment at any time, and judge the future development and operation status of steam pipe network equipment system.

The present invention provides a time-series-based online inspection method for steam pipe network leaks, which includes:

Obtain the multi-source sensing data collected by all steam pressure sensors on the steam pipe network, and obtain the original sensor data set. The multi-source sensing data includes various types of steam data;

In the original sensor data set, data correlation detection and time series continuity detection are respectively performed on the multi-source sensing data of the same steam pressure sensor, so as to respectively obtain the data correlation detection abnormal set Ωm and the time series continuity detection abnormal set Ωc;

Fusing the data correlation detection anomaly set Ωm and the time series continuity detection anomaly set Ωc;

Leak point information in the steam pipe network is determined according to the fusion result of the time series continuity detection anomaly set Ωc and the data correlation detection anomaly set Ωm.

In the time-series-based online inspection method for steam pipe network leaks as described above, preferably, the multi-source sensing data collected by each of the steam pressure sensors includes pressure values, flow velocity values and steam volumes.

In the time-series-based online inspection method for leaks in the steam pipeline network as described above, preferably, the multi-source sensing data for the same steam pressure sensor are respectively processed in the original sensor data set. Correlation detection and timing continuity detection to obtain data correlation detection anomaly set Ωm and timing continuity detection anomaly set Ωc respectively, including:

Extracting multi-source sensing data corresponding to each of the steam pressure sensors according to the original sensor data set, to obtain a single sensor steam data set of each of the steam pressure sensors;

Perform data correlation detection on the multi-source sensor data in the single-sensor data set of each of the steam pressure sensors to obtain a data correlation detection abnormal set Ωm, wherein the multi-source sensor data represent different types of steam data;

Time-series continuity detection is performed on the unitary sensing data in the single-sensor data set of each steam pressure sensor to obtain a time-series continuity detection anomaly set Ωc, wherein the unitary sensing data represent the same type of steam data.

In the time-series-based online inspection method for leaks in the steam pipe network as described above, preferably, the multi-source sensing data corresponding to each of the steam pressure sensors is extracted according to the original sensor data set to obtain The single-sensor steam data set of each of the steam pressure sensors specifically includes:

Extracting multi-source sensing data corresponding to a single steam pressure sensor from the original sensor data set as a single sensor abnormality detection ID set, the number of the single abnormality detection ID set is consistent with the number of the steam pressure sensor;

The single sensor anomaly detection ID set is used as the single sensor steam data set.

In the time-series-based online inspection method for steam pipe network leaks as described above, preferably, the fusion of the data correlation detection abnormality set Ωm and the time-series continuity detection abnormality set Ωc is carried out, specifically include:

Using the time series continuity detection anomaly set Ωc to locate the steam pipe network corresponding to the correlation sensing data in the data correlation detection anomaly set Ωm, and according to the time series continuity detection anomaly set Ωc, multiple Source sensor abnormal data, eliminating the steam data that does not have abnormality in the data correlation detection abnormal set Ωm, and storing the steam data that does not have abnormality in the normal sensor data set Ωdel;

Using the time series continuity detection anomaly set Ωc to re-screen the correct data set Ωr that meets the correlation relationship in the data correlation detection anomaly set Ωm, and filter out the sensor data that conforms to the correct correlation relationship and has abnormal data Stored in the new abnormal data set Ωadd;

Using the normal sensor data set Ωdel and the newly added abnormal data set Ωadd, the results of the time series continuity detection abnormal set Ωc and the data correlation detection abnormal set Ωm are fused.

In the time-series-based online inspection method for steam pipe network leakage points as described above, preferably, the correlation in the anomaly set Ωc detected by using the time-series continuity to the data correlation detection anomaly set Ωm The steam pipe network corresponding to the sensing data is positioned, including:

Draw a time-varying curve of the multi-source sensing data in the time-series continuity detection anomaly set Ωc according to the time-series continuity detection anomaly set Ωc, to obtain a multi-source sensing data change curve;

According to the change of the slope of the curve in the multi-source sensing data change curve, the steam pipe network corresponding to the correlation sensing data in the data correlation detection abnormal set Ωm is located.

In the time-series-based online inspection method for steam pipe network leaks as described above, preferably, using the normal sensor data set Ωdel and the newly added abnormal data set Ωadd, the time series continuity The results of detecting anomaly set Ωc and the data correlation detection anomaly set Ωm are fused, specifically including:

In the union set of timing continuity detection anomaly set Ωc and data correlation detection anomaly set Ωm, the normal sensor data set Ωdel is removed, and the new abnormal data set Ωadd is added to obtain the fusion result.

In the time-series-based online inspection method for steam pipe network leaks as described above, preferably, the determination is made according to the fusion results of the time-series continuity detection abnormality set Ωc and the data correlation detection abnormality set Ωm Leak point information in the steam pipe network, specifically including:

According to the fusion result of the sequence continuity detection anomaly set Ωc and the data correlation detection anomaly set Ωm, the location of the leak point in the steam pipe network and the abnormal sensor data of the location of the leak point are determined.

In the time-series-based online inspection method for leaks in steam pipe networks as described above, preferably, the time-series-based online inspection for leaks in steam pipe networks further includes:

According to the fusion results, the pressure value in the steam pipe network is screened in the 3D intelligent system, and the location of the leak point in the steam pipe network is marked and warned in the 3D intelligent system.

The present invention provides a time-series-based online inspection method for leaks in steam pipe networks, which acquires multi-source sensing data of each steam pressure sensor, obtains distributed sensing data, and utilizes the idea of edge computing for big data processing to achieve It is possible to process the corresponding data on computing resources close to the data source, while reducing the pressure on network transmission bandwidth, it improves the overall efficiency of data processing, and at the same time performs corresponding data fusion on the data detection results of DCD and TCD operation, so as to effectively avoid the defects of the above-mentioned DCD and TCD detection, realize the online detection of leaks in the steam pipe network, and effectively improve the work efficiency of maintenance personnel; it is possible to more quickly judge the approximate position information of impurities and other influences on the pressure curve and analyze them in 3D Marking and early warning in the system can enable staff to view and analyze in real time, communicate effectively with on-site staff, and deal with faults or avoid accidents in a timely and accurate manner.

Description of drawings

In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention will be further described below in conjunction with accompanying drawing, wherein:

Fig. 1 is a flow chart of an embodiment of an online inspection method for steam pipe network leaks based on time series provided by the present invention.

Detailed ways

Various exemplary embodiments of the present disclosure will now be described in detail with reference to the accompanying drawings. The description of the exemplary embodiments is illustrative only, and in no way restricts the disclosure, its application or uses. The present disclosure can be implemented in many different forms and is not limited to the embodiments described here. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art. It should be noted that relative arrangements of parts and steps, compositions of materials, numerical expressions and numerical values set forth in these embodiments should be interpreted as illustrative only and not as limiting, unless specifically stated otherwise.

"First", "second" and similar words used in the present disclosure do not indicate any order, quantity or importance, but are only used to distinguish different parts. Words like "comprising" or "comprising" mean that the elements preceding the word cover the elements listed after the word, and do not exclude the possibility of also covering other elements. "Up", "Down" and so on are only used to indicate the relative positional relationship. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

In the present disclosure, when it is described that a specific component is located between a first component and a second component, there may or may not be an intervening component between the specific component and the first component or the second component. When it is described that a specific component is connected to other components, the specific component may be directly connected to the other component without an intermediate component, or may not be directly connected to the other component but has an intermediate component.

All terms (including technical terms or scientific terms) used in the present disclosure have the same meaning as understood by one of ordinary skill in the art to which the present disclosure belongs, unless otherwise specifically defined. It should also be understood that terms defined in, for example, general-purpose dictionaries should be interpreted as having meanings consistent with their meanings in the context of the relevant technology, and should not be interpreted in idealized or extremely formalized meanings, unless explicitly stated herein Defined like this.

Techniques, methods and devices known to those of ordinary skill in the relevant art may not be discussed in detail, but where appropriate, techniques, methods and devices should be considered part of the description.

As shown in Figure 1, the time-series-based online inspection method for steam pipe network leaks provided by this embodiment specifically includes the following steps in the actual implementation process:

Step S1. Obtain multi-source sensing data collected by all steam pressure sensors on the steam pipe network to obtain a set of original sensor data. The multi-source sensing data includes various types of steam data.

The initialization operation of the corresponding data variables is completed through step S1, that is, the collected parameters are put into a data set, which is convenient for subsequent processing.

Wherein, the multi-source sensing data collected by each steam pressure sensor includes pressure value, flow rate value and steam volume. It should be noted that the present invention does not specifically limit the type of steam data, as long as the parameter values that can reflect the detection of leaks in the pipeline network belong to multi-source sensing data.

By obtaining the multi-source sensing data of each steam pressure sensor, the distributed sensing data is obtained, and the idea of big data processing of edge computing is used to process the corresponding data on the computing resources close to the data source as much as possible. , which can reduce the pressure on network transmission bandwidth and improve the overall efficiency of data processing.

Step S2, performing data correlation detection (data correlation detection, DCD) and temporal continuity detection (temporal conyinuity detection, TCD) respectively on the multi-source sensing data of the same vapor pressure sensor in the original sensor data set, In order to obtain data correlation detection anomaly set Ωm and timing continuity detection anomaly set Ωc respectively.

Among them, Data Dependency Detection (DCD) is used to detect the correlation between data, and Time Series Continuity Detection (TCD) is used to detect the continuity of data in time series. In one embodiment of the time-series-based online inspection method for leaks in the steam pipe network of the present invention, the step S2 may specifically include:

Step S21 , extracting multi-source sensing data corresponding to each of the steam pressure sensors according to the original sensor data set to obtain a single-sensor steam data set of each of the steam pressure sensors.

In an embodiment of the time-series-based online inspection method for steam pipe network leaks of the present invention, the step S21 may specifically include:

Step S211, extract the multi-source sensing data corresponding to a single steam pressure sensor from the original sensor data set as a single sensor anomaly detection ID set, and the number of the single anomaly detection ID set is consistent with the number of the steam pressure sensors .

Step S212, using the single sensor anomaly detection ID set as the single sensor steam data set.

Step S22, performing data correlation detection on the multi-source sensor data in the single-sensor data set of each of the steam pressure sensors to obtain a data correlation detection abnormal set Ωm, wherein the multi-source sensor data represent different types of steam data .

Step S23, performing time-series continuity detection on the unary sensing data in the single-sensor data set of each steam pressure sensor, and obtaining an anomaly set Ωc of time-series continuity detection, wherein the unary sensing data represent the same type of steam data .

The specific detection methods of Data Dependency Detection (DCD) and Timing Continuity Detection (TCD) can refer to the prior art, and will not be repeated here.

Step S3, merging the data correlation detection anomaly set Ωm and the time series continuity detection anomaly set Ωc.

In one embodiment of the time-series-based online inspection method for leaks in the steam pipe network of the present invention, the step S3 may specifically include:

Step S31, using the time-series continuity detection anomaly set Ωc to locate the steam pipe network corresponding to the correlation sensing data in the data correlation detection anomaly set Ωm, and according to the time-series continuity detection anomaly set Ωc The multi-source sensor abnormal data in , eliminate the steam data that does not have abnormality in the data correlation detection abnormal set Ωm, and save the steam data that does not have abnormality in the normal sensor data set Ωdel.

Wherein, in one embodiment of the present invention, the multi-source sensing data corresponding to each of the steam pressure sensors is extracted according to the original sensor data set to obtain a single-sensor steam data set of each of the steam pressure sensors, Specifically include:

Step S311 , draw a time-varying curve of the multi-source sensing data in the time-series continuity detection anomaly set Ωc according to the time-series continuity detection anomaly set Ωc, to obtain a multi-source sensing data change curve.

Step S312 , according to the change of the slope of the curve in the multi-source sensing data change curve, locate the steam pipe network corresponding to the correlation sensing data in the data correlation detection anomaly set Ωm.

Exemplarily, the leakage point of the steam pipe network can be accurately located according to the sudden change point in the pressure value curve and the flow rate value curve. If a sudden change occurs, it indicates that there is a leak point in the pipe network near this point.

Step S32, using the time series continuity detection anomaly set Ωc to re-screen the correct data set Ωr that conforms to the correlation relationship in the data correlation detection anomaly set Ωm, and selects the correct data set Ωr that conforms to the correct correlation relationship and has data anomalies The sensing data is stored in the newly added anomaly dataset Ωadd.

Among them, the criterion for re-screening the correct data set Ωr that conforms to the correlation relationship in the data correlation detection abnormal set Ωm can be the change of the correlation relationship between the pressure data measured by the pressure sensor and the change of the steam volume and flow rate.

Step S33, using the normal sensor data set Ωdel and the newly added abnormal data set Ωadd to fuse the results of the sequence continuity detection abnormal set Ωc and the data correlation detection abnormal set Ωm.

Exemplarily, the fusion process can be to remove the normal sensor data set Ωdel and add the new abnormal data set Ωadd to obtain the final fusion result in the union of the sequence continuity detection anomaly set Ωc and the data correlation detection anomaly set Ωm, The effective fusion of abnormal data results can be achieved through the fusion operation. By integrating the data sets, not only can the abnormal data not found by TCD detection and DCD detection be supplemented, but also the corresponding data without abnormalities can be eliminated, and the two detection methods can learn from each other to ensure the accuracy of the final result.

The present invention proposes an Anomaly Detection of Multi-source Sensing Date based on Time Series (ADMSD_TS) strategy, which can perform corresponding data fusion operations on the data detection results of DCD and TCD, thereby effectively avoiding Therefore, the detection result of ADMSD_TS algorithm is obviously better than the abnormal data detection method of single TCD and single DCD.

Step S4 , according to the fusion result of the time series continuity detection anomaly set Ωc and the data correlation detection anomaly set Ωm , determine the leakage point information in the steam pipeline network.

Specifically, according to the fusion result of the time series continuity detection anomaly set Ωc and the data correlation detection anomaly set Ωm, the location of the leak point in the steam pipe network and the abnormal sensor data of the location of the leak point are determined. In one embodiment of the present invention, the fused abnormal data set is stored in the result array.

Further, in one embodiment of the present invention, the time-series-based on-line inspection of leaks in the steam pipeline network further includes:

Step S5. Screen the pressure value in the steam pipe network in the 3D AI system according to the fusion result, and mark and warn the location of the leak in the steam pipe network in the 3D AI system.

The time-series-based online inspection method for steam pipe network leaks provided by the embodiment of the present invention obtains multi-source sensing data of each steam pressure sensor, obtains distributed sensing data, and utilizes the idea of big data processing of edge computing, As far as possible, the corresponding data is processed on the computing resources close to the data source, which reduces the pressure on the network transmission bandwidth and improves the overall efficiency of data processing. Fusion operation, so as to effectively avoid the defects of the above-mentioned DCD and TCD detection, realize the online detection of steam pipe network leakage points, and effectively improve the work efficiency of maintenance personnel; it can more quickly judge the approximate position information of impurities and other influences on the pressure curve Marking and early warning in the 3D system can enable staff to view and analyze in real time, communicate effectively with on-site staff, and deal with faults or avoid accidents in a timely and accurate manner.

So far, the embodiments of the present disclosure have been described in detail. Certain details known in the art have not been described in order to avoid obscuring the concept of the present disclosure. Based on the above description, those skilled in the art can fully understand how to implement the technical solutions disclosed herein.

Although some specific embodiments of the present disclosure have been described in detail through examples, those skilled in the art should understand that the above examples are for illustration only, rather than limiting the scope of the present disclosure. Those skilled in the art should understand that the above embodiments can be modified or some technical features can be equivalently replaced without departing from the scope and spirit of the present disclosure. The scope of the present disclosure is defined by the appended claims.

Claims (9)

1. A time-series-based online inspection method for steam pipe network leaks, characterized in that it comprises: Obtain the multi-source sensing data collected by all steam pressure sensors on the steam pipe network, and obtain the original sensor data set. The multi-source sensing data includes various types of steam data; In the original sensor data set, data correlation detection and time series continuity detection are respectively performed on the multi-source sensing data of the same steam pressure sensor, so as to respectively obtain the data correlation detection abnormal set Ωm and the time series continuity detection abnormal set Ωc; Fusing the data correlation detection anomaly set Ωm and the time series continuity detection anomaly set Ωc; Leak point information in the steam pipe network is determined according to the fusion result of the time series continuity detection anomaly set Ωc and the data correlation detection anomaly set Ωm. 2. The time-series-based online inspection method for steam pipe network leaks according to claim 1, wherein the multi-source sensing data collected by each of the steam pressure sensors includes pressure values, flow rate values and steam quantity. 3. The time-series-based online inspection method for steam pipe network leaks according to claim 1, wherein the multi-source sensing data for the same steam pressure sensor in the original sensor data set Carry out data correlation detection and time series continuity detection respectively to obtain data correlation detection anomaly set Ωm and time series continuity detection anomaly set Ωc, specifically including: Extracting multi-source sensing data corresponding to each of the steam pressure sensors according to the original sensor data set, to obtain a single sensor steam data set of each of the steam pressure sensors; Perform data correlation detection on the multi-source sensor data in the single-sensor data set of each of the steam pressure sensors to obtain a data correlation detection abnormal set Ωm, wherein the multi-source sensor data represent different types of steam data; Time-series continuity detection is performed on the unitary sensing data in the single-sensor data set of each steam pressure sensor to obtain a time-series continuity detection anomaly set Ωc, wherein the unitary sensing data represent the same type of steam data. 4. The time-series-based online inspection method for steam pipe network leaks according to claim 3, wherein the multi-source sensor corresponding to each of the steam pressure sensors is extracted according to the original sensor data set. Data, obtain the single-sensor steam data collection of each described steam pressure sensor, specifically include: Extracting multi-source sensing data corresponding to a single steam pressure sensor from the original sensor data set as a single sensor abnormality detection ID set, the number of the single abnormality detection ID set is consistent with the number of the steam pressure sensor; The single sensor anomaly detection ID set is used as the single sensor steam data set. 5. The time-series-based online inspection method for steam pipe network leaks according to claim 1, characterized in that, the data correlation detection abnormal set Ωm and the time-series continuity detection abnormal set Ωc are performed integration, including: Using the time series continuity detection anomaly set Ωc to locate the steam pipe network corresponding to the correlation sensing data in the data correlation detection anomaly set Ωm, and according to the time series continuity detection anomaly set Ωc, multiple Source sensor abnormal data, eliminating the steam data that does not have abnormality in the data correlation detection abnormal set Ωm, and storing the steam data that does not have abnormality in the normal sensor data set Ωdel; Using the time series continuity detection anomaly set Ωc to re-screen the correct data set Ωr that meets the correlation relationship in the data correlation detection anomaly set Ωm, and filter out the sensor data that conforms to the correct correlation relationship and has abnormal data Stored in the new abnormal data set Ωadd; Using the normal sensor data set Ωdel and the newly added abnormal data set Ωadd, the results of the time series continuity detection abnormal set Ωc and the data correlation detection abnormal set Ωm are fused. 6. The time-series-based online inspection method for steam pipe network leaks according to claim 5, wherein the time-series continuity detection anomaly set Ωc is used to detect the data correlation anomaly set Ωm Position the steam pipeline network corresponding to the relevant sensor data, including: Draw a time-varying curve of the multi-source sensing data in the time-series continuity detection anomaly set Ωc according to the time-series continuity detection anomaly set Ωc, to obtain a multi-source sensing data change curve; According to the change of the slope of the curve in the multi-source sensing data change curve, the steam pipe network corresponding to the correlation sensing data in the data correlation detection abnormal set Ωm is located. 7. The time-series-based online inspection method for leaks in steam pipe networks according to claim 5, characterized in that, using the normal sensor data set Ωdel and the newly added abnormal data set Ωadd, the The results of the sequence continuity detection anomaly set Ωc and the data correlation detection anomaly set Ωm are fused, specifically including: In the union set of timing continuity detection anomaly set Ωc and data correlation detection anomaly set Ωm, the normal sensor data set Ωdel is removed, and the new abnormal data set Ωadd is added to obtain the fusion result. 8. The time-series-based online inspection method for steam pipe network leaks according to claim 1, characterized in that, according to the time-series continuity detection abnormality set Ωc and the data correlation detection abnormality set Ωm The fusion results determine the leak point information in the steam pipe network, including: According to the fusion result of the sequence continuity detection anomaly set Ωc and the data correlation detection anomaly set Ωm, the location of the leak point in the steam pipe network and the abnormal sensor data of the location of the leak point are determined. 9. The time-series-based online inspection method for steam pipe network leaks according to claim 1, wherein the time-series-based online inspection of steam pipe network leaks further comprises: According to the fusion results, the pressure value in the steam pipe network is screened in the 3D intelligent system, and the location of the leak point in the steam pipe network is marked and warned in the 3D intelligent system.

Images