CN110031152B - Thermal heating system and method for detecting temperature deviation accumulation leakage of adjacent days - Google Patents

Abstract

The invention provides a thermal heating system and a leak detection method thereof, comprising a boiler, a heat exchanger and a heating radiator, wherein the boiler, the heat exchanger and the heating radiator are connected through a heating pipe network, and the boiler generates a The steam enters the heat exchanger, exchanges heat with the water in the heat exchanger, and then the water enters the heating radiator for heating, the heating pipe network has a plurality of nodes, and a thermal imager is set at at least one node; When the change of the accumulated sum of temperature offsets at the same time on adjacent days exceeds the threshold, a node leakage alarm will be triggered; the alarm method adopts the accumulated sum of temperature offsets at the same time on adjacent days. The invention provides a new heating pipe network system for intelligently detecting leakage and alarming. The invention automatically detects the temperature deviation of the adjacent days and the time of the leakage by monitoring the changes of the infrared temperature field at the nodes of the heating pipe network in real time through an infrared thermal imager. Shift accumulation and alarms make the detection results more accurate and the error smaller.

Description

A thermal heating system and a method for cumulative leakage detection of adjacent daily temperature offsets

technical field

The invention relates to a heating system, in particular to the technical field of a heating pipe network for intelligent leak detection.

Background technique

The leakage of the central heating pipe network will directly lead to a large loss of high-temperature media in the pipe, thermal pollution of the environment, serious leakage may even lead to geological collapse, resulting in casualties, which has always been the main fault affecting the safe and economic operation of the pipe network. With the rapid development of domestic central heating in recent years, the scale of heating areas and pipe networks has continued to expand. In particular, traditional thermal power plants are actively developing cogeneration (such as low-vacuum transformation, circulating water, etc.) under the guidance of national energy-saving and emission reduction policies. Waste heat utilization, etc.), making the operation safety of the heating network have an increasing impact on the safe operation of the power plant units. Once a large leakage occurs in the heating pipe network, it will directly lead to the unit tripping, resulting in major safety accidents. For the branch nodes of the heating pipe network, due to the connection of branch pipelines, on-site welding and insulation are required, and the processing quality is difficult to reach the technical level of the prefabricated thermal insulation pipelines in the factory. In addition, branch pipeline valves and instruments are installed at the branch nodes. The stress concentration point, which causes the leakage failure probability of the branch nodes of the pipeline network is much greater than that of ordinary pipelines. According to engineering practice statistics, more than 60%-80% of the leakage faults in the heating pipe network occur at branch nodes.

Heating pipe network leak detection, especially the research and application of leakage fault real-time monitoring methods, has always been the focus of domestic and foreign scholars and heating pipe network operators. This method can be divided into two categories: direct method and indirect method. The direct method mainly includes the direct buried early warning line method, the distributed optical fiber temperature measurement method and the infrared imaging detection method. At present, the European direct buried early warning line monitoring system already has a relatively mature design and process method. This method is divided into two types: impedance type and resistance type. Both need to embed the alarm line in the prefabricated insulation layer, and diagnose the fault point and its location by detecting the pulse reflection signal and resistance value respectively, and can detect internal and external leakage. However, this method needs to arrange the detection point within a certain distance (500m is recommended in China), and the field installation process of the detection point and the entire monitoring system have high requirements on the design and process of the pipeline network; the distributed optical fiber temperature measurement method is mainly based on Raman light. Based on the principles of reflection, Brillouin light reflection and fiber grating, a temperature measurement system composed of temperature-measuring optical fiber sensors arranged in series on the outside of the pipeline can sense the temperature change caused by the leakage, so as to find the leakage and accurately locate it. Among them, the British York company's distributed optical fiber temperature sensing system based on Raman light reflection is widely used, but compared with the direct buried early warning line method, the cost is higher and the technology maturity is lower; the infrared imaging detection method adopts thermal infrared imaging technology, which will be The infrared radiation energy distribution image of the measured target is converted into a standard video signal of the temperature field of the measured target. As one of the manual inspection methods of the heating pipe network, this method does not have any influence on the operation of the pipe network, and is mainly used for burying shallow directly buried thermal pipes. At present, there are studies at home and abroad using drone-borne infrared cameras to monitor the leakage of the entire urban pipeline network, but it is impossible to distinguish the temperature rise around the pipeline caused by leakage and pipeline insulation damage, and the high-altitude flying of drones is currently controlled by national security, which is difficult to implement. big. The indirect method mainly includes model method, neural network method and statistical detection method. The model method is to establish a steady-state or transient model of the heating pipe network, and compare and analyze the simulated value of the pipe network with the actual operating data (flow or pressure) to determine whether there is leakage. The accuracy of this method mainly depends on the accuracy of the pipe network model. The neural network method relies on learning the normal and faulty operation data of the pipe network, independently analyzes the operation status of the pipe network and establishes the ability to judge the leakage of the pipe network. The method has strong anti-interference ability, but requires a large amount of leakage data to learn modeling; the statistical detection method is based on statistical theory, analyzes the operating data of leakage conditions, and establishes a functional relationship with normal operating conditions to estimate the leakage amount and leakage location. This method does not need to establish a model, only needs to perform a small amount of pressure and flow probability calculations, and has wide adaptability, but it has strict requirements on the accuracy of the instrument. Due to the wide application of the online monitoring system of the heating pipe network in China and the continuous improvement of the instrument accuracy, a good material foundation has been laid for the application of the statistical detection method. At present, this method has received continuous attention in the field of leakage detection of the water supply pipe network.

Among the above two types of methods, the direct buried early warning line method in the first type of direct method has relatively mature technology and high detection efficiency, but has high process requirements and high cost, which is difficult to popularize and apply in China in the short term. Even if the new pipe network can be considered, it is more difficult to apply and implement the heating pipe network that has been built and operated at present due to the high cost; although the distributed optical fiber temperature measurement method has been accumulated to a certain extent in research and engineering applications, and the The detection efficiency of the method is high, but compared with the direct buried early warning line method, its cost is higher and the technology maturity is lower; the infrared imaging detection method has been widely used in the field of manual detection due to its simple and fast characteristics. However, even if domestic conditions permit, the current research and development of the UAV-borne infrared camera detection method can only achieve the purpose of regular inspection and inspection, and this method also needs to solve how to distinguish and confirm the supply under the complex background and environmental interference. Heat pipe leak point; the second type of indirect method, the first is the model method. Whether it is a steady-state or transient model method, it is necessary to further improve the model accuracy and study how to quickly and effectively establish a specific heating pipe network model; the main problem faced by the neural network method is not only how to obtain effective operating data, but also how to Research the optimization algorithm to ensure the fast and effective convergence of the neural network; the statistical detection method has simple operation and wide adaptability, and the online monitoring system is widely used in the domestic heating pipe network, which provides a solid application basis for it. However, it is still necessary to rely on the further improvement of the instrument accuracy in the instrument industry, as well as the research and development of suitable methods, so that it can be applied in the field of large and complex heating pipe networks.

This project is based on the currently relatively mature infrared thermal imaging technology (infrared thermal imaging technology converts invisible infrared energy emitted by objects into thermal images visible to the human eye through optics and detectors), and combines visible light image processing methods, thermal infrared image processing methods The method is organically integrated with pattern recognition technology, and a real-time leakage detection system and method for heating pipe network nodes based on infrared thermal imaging technology are proposed, and the corresponding software and hardware systems are researched and developed. The overall efficiency of the leakage detection of the heating pipe network ensures the safe operation of the heating pipe network and the power plant units.

The inventor has developed and tested the leakage detection technology in the previous research and applied for the patent, and adopts the alarm method of the temperature difference on the day, but this alarm method has large errors and false alarms. Improved and developed a new alarm method, which can improve the accuracy of the alarm and reduce errors.

SUMMARY OF THE INVENTION

Aiming at the deficiencies in the prior art, the present invention provides a heating system and method for intelligently detecting leakage, which can detect the leakage of pipe network nodes in real time, so as to solve the technical problem of real-time leakage detection at heating pipe network nodes.

To achieve the above object, the present invention adopts the following technical solutions:

A thermal heating system, comprising a boiler, a heat exchanger and a heating radiator, the boiler, the heat exchanger and the heating radiator are connected through a heating pipe network, and the steam generated by the boiler enters the heat exchanger and is connected with the heat exchange The water in the heat exchanger conducts heat exchange, and then the water enters the heating radiator for heating. The heating pipe network has multiple nodes, and is characterized in that a thermal imager is installed at at least one node; the thermal imager is installed on the manhole cover. Detect the data of the position of the manhole cover; calculate the change of the cumulative sum of temperature offsets at the same time on adjacent days, and trigger a node leakage alarm when the temperature exceeds the threshold; The cumulative sum of the temperature offset values at the same time of day, when the cumulative sum value exceeds the set threshold, the adjacent time of day temperature offset cumulative sum alarm will be triggered.

Preferably, the thermal imager is arranged on the column.

A real-time detection method for node leakage of a heating system, comprising the following steps:

Data collection and monitoring steps: use a thermal imager to monitor and collect infrared video monitoring data and visible light video monitoring data at the well cover of the heating pipe network;

Data transmission steps: communicate with the data acquisition and monitoring subsystem, and transmit the infrared video monitoring data and visible light video monitoring data of the monitoring point to the server through optical fibers;

Manhole cover integrity detection step: judge the integrity of the manhole cover according to the visible light video monitoring data transmitted to the server;

Leak confirmation step: For the image frames that meet the integrity detection of the manhole cover, calculate the change of the cumulative sum of temperature offsets at the same time of the adjacent days, and trigger the node leakage alarm when the threshold value is exceeded.

Preferably, the integrity detection of the manhole cover includes the following steps:

Define the standard image frame of the manhole cover in the visible light video monitoring data under various working conditions of each monitoring point, which is called the reference image frame R;

1) Calculate the grayscale mean μ _r and the grayscale standard deviation δ _r of each reference image frame according to the following formulas respectively;

where M, N are the image resolutions, and I _ij represents the grayscale value at the corresponding coordinate

2) take a frame in the visible light monitoring video, calculate the grayscale mean μ _t and the grayscale standard deviation δ _t of the current image frame T;

3) Calculate the grayscale mean difference Δμ and the grayscale standard deviation difference Δδ between the current image frame T and the corresponding reference image frame R;

4) When the values of Δμ and Δδ are greater than the set threshold, the current frame is regarded as a suspected frame, and the processing of step 5) is continued; when the values of Δμ and Δδ are less than the set threshold, the current frame is a normal manhole cover frame, and step 2 is continued. ) processing;

如果S_i的值大于设定阈值时，则认为当前帧没有通过井盖完整性检测，丢弃该帧，返回步骤2)继续下一帧的处理；5) For the suspected frame, continue to calculate the sum S _i of the absolute value of the difference in the number of grayscale pixels of each level of the current image frame T and the corresponding reference image frame R,

If the value of S _i is greater than the set threshold, it is considered that the current frame has not passed the integrity detection of the manhole cover, the frame is discarded, and the process is returned to step 2) to continue the processing of the next frame;

6) If the image frames within the specified continuous time fail to pass the integrity detection of the manhole cover, an abnormal integrity alarm will be triggered, and the management personnel will be notified for manual processing.

Preferably, for the image frames that satisfy the integrity detection of the manhole cover, calculate the change of the accumulated sum of temperature offsets at the same time of the adjacent days, and trigger the node leakage alarm when the threshold value is exceeded.

As an option, the accumulation and specific calculation steps of the temperature offset at the same time of the adjacent days are as follows:

与方差

将数据序列标准化为y_i＝(x_i-μ₀)/σ₀。1) According to the monitoring data series x _i of manhole cover temperature at the same time of day (2h group, 12 groups in total), where i=1, 2, Λn, calculate the mean value

with variance

Normalize the data series to y _i =( _xi -μ ₀ )/σ ₀ .

其中，

2) Select the CUSUM cumulative sum parameter k=1.376 according to experience, and the value of h is preferably set to different values according to the three-level alarm level. Then calculate the upper offset cumulative sum

in,

是否大于设定的报警阈值h，若某个

则认为在该时刻温度偏移累积和超过阈值，报警。警报发生后，除非人工干预，否则一级报警时间15min、二级报警时间30min，三级报警一直保持报警状态。3) Judgment

Whether it is greater than the set alarm threshold h, if a certain

Then it is considered that the accumulated sum of temperature offset exceeds the threshold at this moment, and an alarm is issued. After the alarm occurs, unless there is manual intervention, the first-level alarm time is 15 minutes, the second-level alarm time is 30 minutes, and the third-level alarm remains in the alarm state.

4) After the third-level alarm is manually intervened, the temperature data offset is accumulated and cleared, and the next round of calculation and detection is restarted.

的值超过设定阈值h时，触发温度偏移累积和报警。作为优选，根据阈值h的大小，设定一级报警，二级报警以及三级报警。when

When the value exceeds the set threshold h, the temperature offset accumulation and alarm are triggered. Preferably, according to the size of the threshold h, a first-level alarm, a second-level alarm and a third-level alarm are set.

The present invention has the following advantages:

1) A new heating pipe network system for intelligently detecting leakage is provided. The present invention monitors the changes of the infrared temperature field at the manhole covers of the nodes of the heating pipe network in real time through an infrared thermal imager. The jump of the field temperature and the change of the cumulative sum of the temperature offset, the alarm method adopts the accumulation and alarm of the temperature offset at the same time of the adjacent day to determine the leakage accident of the node, and alarm to notify the management personnel. Compared with the leak detection method of the previous application, the present application changes the alarm mode, adopts the accumulation and alarm of the temperature offset at the same time of the adjacent days, so that the result is more accurate, the error is smaller, and it is more suitable for the detection of small leaks.

2) The present invention proposes a new idea of monitoring the occurrence of leakage by detecting the temperature change at the node. By detecting the position of the manhole cover, and firstly by detecting the damage of the manhole cover, the structure is simple and the cost is low.

3) In order to ensure the reliability and accuracy of the provided method, the present invention uses the visible light data monitored at the node to process the abnormal condition (damage or occlusion) of the monitoring node manhole cover to avoid false alarms.

3) The method organically integrates the visible light image processing method, the thermal infrared image processing method and the pattern recognition technology, which can improve the leakage detection efficiency of the heating pipe network node and ensure the safe operation of the heating pipe network and the power plant units.

Description of drawings:

Fig. 1 shows the principle block diagram of the real-time detection system for leakage of heating pipe network nodes based on infrared thermal imaging technology;

Figure 2 shows a schematic diagram of the engineering implementation of the real-time detection system for leakage of heating pipe network nodes based on infrared thermal imaging technology;

Fig. 3 shows the implementation flow chart of the real-time detection method for leakage of heating pipe network nodes based on infrared thermal imaging technology of the present invention;

Fig. 4 shows the flow chart of the manhole cover integrity inspection algorithm in the real-time detection method of node leakage of heating pipe network based on infrared thermal imaging technology of the present invention;

5 shows the flow chart of the temperature offset accumulation and alarm algorithm at the same time of day adjacent to the real-time detection method for leakage of heating pipe network nodes based on infrared thermal imaging technology of the present invention;

Fig. 6 shows the general algorithm flow chart of the real-time detection method for leakage of heating pipe network nodes based on infrared thermal imaging technology of the present invention;

7 shows a flowchart of an alarm algorithm used in conjunction with multiple alarm modes in the real-time detection method for leakage of heating pipe network nodes based on infrared thermal imaging technology of the present invention;

Fig. 8 shows the general algorithm flow chart used in conjunction with multiple alarm modes of the method for real-time detection of leakage of heating pipe network nodes based on infrared thermal imaging technology of the present invention;

Detailed ways

The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are all simplified schematic diagrams, and only illustrate the basic structure of the present invention in a schematic way, so they only show the structure related to the present invention,

A thermal heating system, comprising a boiler, a heat exchanger and a heating radiator, the boiler, the heat exchanger and the heating radiator are connected through a heating pipe network, and the steam generated by the boiler enters the heat exchanger and is connected with the heat exchange The water in the radiator exchanges heat, and then the water enters the heating radiator for heating. The heating pipe network has a plurality of nodes, and a thermal imager is arranged at at least one node.

Preferably, as shown in FIG. 2 , the thermal imager is arranged at the manhole cover to detect the data of the position of the manhole cover. The thermal imager is set on the column.

The invention provides a new heating pipe network system for intelligently detecting leakage. The invention monitors the changes of the infrared temperature field at the nodes of the heating pipe network in real time through an infrared thermal imager. Changes, determine the node leakage accident, and alarm to notify the management personnel.

The detection method will be described in detail below.

Figure 1 shows the principle block diagram of a real-time detection system for node leakage in a heating pipe network based on infrared thermal imaging technology.

As shown in Figure 1, the real-time detection system for node leakage of heating pipe network based on infrared thermal imaging technology of the present invention includes:

It includes a data acquisition and monitoring subsystem, which is used to collect and transmit in real time the infrared video monitoring data and visible light video monitoring data at the nodes of the heating pipe network (preferably manhole covers);

The data transmission subsystem is used to communicate with the data acquisition and monitoring subsystem, and transmit the infrared video monitoring data and visible light video monitoring data of the monitoring point to the server;

The manhole cover integrity detection subsystem uses the monitored visible light data to determine whether there is damage or occlusion at the monitoring point (preferably the manhole cover). The integrity detection data frame is discarded directly. If the image frames within the specified continuous time fail to pass the integrity detection of the manhole cover, the abnormality alarm of the integrity of the manhole cover will be triggered, and the management personnel will be notified to deal with it manually.

The infrared data processing and alarm subsystem uses the monitored temperature field data of infrared imaging to calculate the change of the accumulated sum of temperature offsets at the same time in adjacent days. When the threshold value is exceeded, the node leakage alarm is triggered.

Figure 2 shows a schematic diagram of the engineering implementation of a real-time leak detection system for heating pipe network nodes based on infrared thermal imaging technology.

The statistical data of engineering practice shows that: in the case of leakage in the heating pipe network, most of the leakage occurs at the nodes of the heating pipe network. As shown in Figure 2, an infrared thermal image monitor is placed near the primary pipe network node (manhole cover) of the urban central heating, and the change information of the infrared temperature field at the monitoring point is transmitted to the server in real time through the optical fiber. The server automatically monitors the occurrence of leaks in real time and notifies managers.

Preferably, the present invention also provides a real-time detection method for node leakage of a heating pipe network based on infrared thermal imaging technology. Fig. 3 shows the implementation flow chart of the method for real-time detection of leakage of heating pipe network nodes based on infrared thermal imaging technology of the present invention, as shown in Fig. 3, which specifically includes the following steps:

1) Extract a frame of visible light image from the data transmitted from the monitoring point to the server, and perform the integrity detection of the manhole cover according to the frame image. Infrared temperature field imaging is easily affected by surrounding objects or the environment. Manhole cover integrity inspection can eliminate abnormalities such as missing or occluded manhole covers, and ensure the accuracy of the infrared temperature field data transmitted from the monitoring point back to the server. The specific method of manhole cover integrity inspection will be described in detail later.

2) For the data frame that does not pass the detection, directly discard it and take the next frame of visible light data;

3) For the data frame that passes the detection, extract the infrared temperature field data corresponding to the frame, save the temperature field data to the database, and calculate the cumulative sum of temperature offsets at the same time on adjacent days, and determine whether there is leakage through threshold judgment. occur. If there is, a leak alarm will be triggered, and the relevant management personnel will be notified to deal with it, otherwise, the calculation will continue at the next moment of the adjacent day. The specific methods for accumulating temperature offsets at the same time of day and for detecting and alarming thresholds will be described in detail in the following content.

The method for detecting the integrity of the manhole cover will be described in detail in this embodiment below.

Infrared imaging data is easily affected by the external environment. Manhole cover integrity detection can eliminate abnormal conditions such as manhole cover damage and occlusion, and ensure that the infrared temperature field distribution at the monitoring point (preferably the manhole cover) can be accurately obtained in the follow-up. Manhole cover integrity detection uses the visible light data transmitted from the monitoring point to the server, and is divided into two steps: suspected frame search and suspected frame confirmation. The steps for finding suspected frames are as follows:

1) Define the standard image frame of the manhole cover in the visible light video monitoring data under various working conditions of each monitoring point, which we call the reference image frame R;

2) Calculate the grayscale mean μ _r and the grayscale standard deviation δ _r of each reference image frame according to the following formulas respectively;

where M, N are the image resolutions, and I _ij represents the grayscale value at the corresponding coordinate

3) take a frame in the visible light monitoring video, calculate the grayscale mean μ _t and the grayscale standard deviation δ _t of the current image frame T;

4) Calculate the difference Δμ of the mean gray value and the difference Δδ of the standard deviation of gray between the current image frame T and the corresponding reference image frame R;

5) When the values of Δμ and Δδ are greater than the set threshold, the current frame is regarded as a suspected frame, and the confirmation of subsequent suspected frames is continued; when the values of Δμ and Δδ are less than the set threshold, the current frame is a normal manhole cover frame, and proceed to step 3 processing.

The steps to confirm the suspected frame are as follows:

如果S_i的值大于设定阈值时，则认为当前帧没有通过井盖完整性检测，丢弃该帧对应的红外数据帧，返回疑似帧查找步骤3)；1) For the suspected frame, continue to calculate the sum of the absolute values of the difference between the current image frame T and the corresponding reference image frame R in each level of grayscale pixel counts

If the value of S _i is greater than the set threshold, it is considered that the current frame has not passed the integrity detection of the manhole cover, the infrared data frame corresponding to the frame is discarded, and the suspected frame search step 3) is returned;

2) If the image frames in the continuous time fail to pass the integrity detection of the manhole cover, an abnormal integrity alarm will be triggered, and the management personnel will be notified to deal with the abnormality at the manhole cover manually.

The infrared data processing and alarm methods will be described in detail in this embodiment below.

The leakage of heating pipe network can be divided into two types: burst pipe and leakage. Pipe burst is a sudden event, with large leakage, which will cause the temperature of the manhole cover to rise rapidly; while leakage is a gradual event, with small leakage and slow temperature change. Therefore, as a preference, the infrared data processing and alarm method steps are as follows:

For the image frames that meet the integrity detection of the manhole cover, calculate the change of the cumulative sum of temperature difference or temperature offset, and trigger the node leakage alarm when it exceeds the threshold.

Specifically, the following three alarm modes are included:

1) Current temperature difference alarm

According to the set time interval (5min), calculate the difference between the current temperature field matrix T _i and the previous frame temperature field matrix T _i-1 : ΔT=T _i -T _i-1 , when the value of ΔT exceeds the set threshold, Trigger the temperature difference alarm. According to the size of the ΔT value, the first-level alarm, the second-level alarm and the third-level alarm are set to realize the rapid detection of the pipe burst in the pipe network.

2) 24-hour temperature offset accumulation and alarm

For the image frames that meet the integrity detection of the manhole cover, calculate the change of the cumulative sum of the 24-hour continuous monitoring temperature offset, and when the threshold is exceeded, the node leakage alarm is triggered.

Specifically, it includes the following three-level alarm modes:

与方差

将数据序列标准化为y_i＝(x_i-μ₀)/σ₀。1) Calculate the mean value of the manhole cover temperature monitoring data sequence x _i at a set time interval of 24 hours, where i=1, 2, Λn

with variance

Normalize the data series to y _i =( _xi -μ ₀ )/σ ₀ .

其中，

2) According to experience, select the CUSUM cumulative sum parameter k=1.425, and the value of h is set to different values according to the three-level alarm. Then calculate the upper offset cumulative sum

in,

是否大于设定的三级报警阈值h，若某个

则认为在该时刻温度偏移累积和超过阈值，报警。警报发生后，除非人工干预，否则一级报警时间15min、二级报警时间30min，三级报警一直保持报警状态。3) Judgment

Whether it is greater than the set three-level alarm threshold h, if a certain

Then it is considered that the accumulated sum of temperature offset exceeds the threshold at this moment, and an alarm is issued. After the alarm occurs, unless there is manual intervention, the first-level alarm time is 15 minutes, the second-level alarm time is 30 minutes, and the third-level alarm remains in the alarm state.

4) After the third-level alarm is manually intervened, the temperature data offset is accumulated and cleared, and the calculation and detection are restarted.

5) Start from 0:00 every day and end in 24 hours, complete the detection task of the day, accumulate and automatically reset, and re-enter the detection calculation for the next day at the same time.

的值超过设定阈值h时，触发温度偏移累积和报警。作为优选，根据阈值h的大小，设定一级报警，二级报警以及三级报警。when

When the value exceeds the set threshold h, the temperature offset accumulation and alarm are triggered. Preferably, according to the size of the threshold h, a first-level alarm, a second-level alarm and a third-level alarm are set.

3) Accumulation and alarm of temperature difference at the same time on adjacent days

For the image frames that meet the integrity detection of the manhole cover, calculate the change of the cumulative sum of temperature offsets at the same time of the adjacent days, and trigger the node leakage alarm when it exceeds the threshold. As a preference, the following three-level alarm modes are specifically included:

与方差

将数据序列标准化为y_i＝(x_i-μ₀)/σ₀。1) According to the monitoring data series x _i of manhole cover temperature at the same time of day (2h group, 12 groups in total), where i=1, 2, Λn, calculate the mean value

with variance

Normalize the data series to y _i =( _xi -μ ₀ )/σ ₀ .

其中，

2) According to experience, select the CUSUM cumulative sum parameter k=1.376, and the value of h is set to different values according to the three-level alarm. Then calculate the upper offset cumulative sum

in,

是否大于设定的三级报警阈值h，若某个

则认为在该时刻温度偏移累积和超过阈值，报警。警报发生后，除非人工干预，否则一级报警时间15min、二级报警时间30min，三级报警一直保持报警状态。3) Judgment

Whether it is greater than the set three-level alarm threshold h, if a certain

Then it is considered that the accumulated sum of temperature offset exceeds the threshold at this moment, and an alarm is issued. After the alarm occurs, unless there is manual intervention, the first-level alarm time is 15 minutes, the second-level alarm time is 30 minutes, and the third-level alarm remains in the alarm state.

4) After the third-level alarm is manually intervened, the temperature data offset is accumulated and cleared, and the next round of calculation and detection is restarted.

的值超过设定阈值h时，触发温度偏移累积和报警。作为优选，根据阈值h的大小，设定一级报警，二级报警以及三级报警。when

When the value exceeds the set threshold h, the temperature offset accumulation and alarm are triggered. Preferably, according to the size of the threshold h, a first-level alarm, a second-level alarm and a third-level alarm are set.

As a preference, the above three alarm modes can be used in combination or independently.

The present invention adopts a new alarm mode. Compared with the current temperature difference alarm in the background technology, the 24-hour temperature offset accumulation and alarm and the adjacent day and time temperature offset accumulation and alarm mode can further improve the accuracy of the alarm and reduce errors. .

Applications

The thermal imager is placed on a column with a height of 3.5 meters, powered by a civilian AC power supply, and connected to the server through an optical fiber. The vertical distance between the thermal imager and the manhole cover is 3 meters, the horizontal distance is 1.5 meters, and the monitoring angle is about 30° obliquely downward. The monitoring video resolution is 384*288, and the frame rate is 12 frames per second.

Other parameter settings of the thermal imager are shown in the table below:

parameter item range temperature range -20℃---150℃ Emissivity 0.81 Reflected temperature 5℃ atmospheric temperature 10℃ Relative humidity 0.33 Transmittance 0.80

The thresholds used for manhole cover integrity inspection are shown in the following table:

parameter item threshold Gray mean difference Δμ 30 Difference Δδ of gray standard deviation 15 The sum S_i of the absolute value of the difference in the number of pixels of each gray level 5500

The 24-hour temperature offset accumulation and alarm thresholds are shown in the following table:

Alarm level Threshold h First class alarm 5 Secondary alarm 10 three-level alarm 15

The temperature offset accumulation and alarm thresholds at the same time of adjacent days are shown in the following table:

Alarm level Threshold h First class alarm 4 Secondary alarm 8 three-level alarm 12

Although the present invention has been disclosed above with preferred embodiments, the present invention is not limited thereto. Any person skilled in the art can make various changes and modifications without departing from the spirit and scope of the present invention. Therefore, the protection scope of the present invention should be based on the scope defined by the claims.

Claims (4)

1. A real-time detection method for node leakage of a heating system, comprising a heating system comprising a boiler, a heat exchanger and a heating radiator, wherein the boiler, the heat exchanger and the heating radiator pass through a heating pipe The steam generated by the boiler enters the heat exchanger and exchanges heat with the water in the heat exchanger, and then the water enters the heating radiator for heating. The heating pipe network has multiple nodes, which is characterized in that , set up a thermal imager at at least one node; the thermal imager is set at the manhole cover to detect the data of the manhole cover position; when the cumulative sum of temperature offset changes at the same time in adjacent days exceeds the threshold, the node leakage alarm will be triggered; the alarm method is as follows: Accumulation and alarm of temperature offsets at the same time on adjacent days, that is, calculating the cumulative sum of temperature offsets at the same time on adjacent days, and triggering the accumulation and alarm of temperature offsets when the value of the accumulated sum exceeds the set threshold; It is characterized in that, comprises the following steps: 1) Data collection and monitoring steps: use a thermal imager to monitor and collect infrared video monitoring data and visible light video monitoring data at the well cover of the heating pipe network; 2) Data transmission step: communicate with the data acquisition and monitoring subsystem, and transmit the infrared video monitoring data and visible light video monitoring data of the monitoring point to the server through optical fibers; 3) Manhole cover integrity detection step: judge the integrity of the manhole cover according to the visible light video monitoring data transmitted to the server; 4) Leak confirmation step: For the image frames that meet the integrity detection of the manhole cover, calculate the cumulative sum of temperature offsets at the same time on adjacent days, and when the cumulative sum exceeds the set threshold, trigger the cumulative sum of temperature offsets to alarm; In the leak confirmation step, the alarm method adopts the temperature offset accumulation and alarm at the same time on adjacent days, that is, the daily manhole cover temperature monitoring data starts from zero, and a calculation time is set to be selected every 2 hours, a total of 12 times, and then the adjacent days are set to be calculated. The data at the same time is composed of 12 groups of stable time data series, and the cumulative sum C _i of temperature excursions of the data series at the same time on adjacent days is calculated. When the value of C _i exceeds the set threshold h, the cumulative sum of temperature excursions is triggered. Specific steps are as follows: 1) According to the monitoring data sequence x _i of the manhole cover temperature at the same time of the adjacent days, where i=1, 2, Λn, calculate the mean value

with variance

Normalize the data series to y _i =( _xi -μ ₀ )/σ ₀ ; 2) Select the CUSUM cumulative sum parameter k=1.376 according to experience, and then calculate the upper offset cumulative sum

in,

3) Judgment

Whether it is greater than the set alarm threshold h, if a certain

Then it is considered that the cumulative sum of the temperature offset exceeds the threshold at this moment, and an alarm is issued; 4) After the alarm is manually intervened, the temperature data offset is accumulated and cleared, and the next round of calculation and detection is restarted. 2. The real-time detection method for node leakage of a heating system as claimed in claim 1, wherein the manhole cover integrity detection comprises the steps: Define the standard image frame of the manhole cover in the visible light video monitoring data under various working conditions of each monitoring point, which is called the reference image frame R; 1) Calculate the grayscale mean μ _r and the grayscale standard deviation δ _r of each reference image frame according to the following formulas respectively;

where M, N are the image resolutions, and I _ij represents the grayscale value at the corresponding coordinate 2) take a frame in the visible light monitoring video, calculate the grayscale mean μ _t and the grayscale standard deviation δ _t of the current image frame T; 3) Calculate the difference Δμ of the mean gray value and the difference Δδ of the standard deviation of gray between the current image frame T and the corresponding reference image frame R; 4) When the values of Δμ and Δδ are greater than the set threshold, the current frame is regarded as a suspected frame, and the process of step 5) is continued; when the values of Δμ and Δδ are less than the set threshold, the current frame is a normal manhole cover frame, and the process continues with step 2 ); 5) For the suspected frame, continue to calculate the sum S _i of the absolute value of the difference in the number of grayscale pixels of each level of the current image frame T and the corresponding reference image frame R,

If the value of S _i is greater than the set threshold, it is considered that the current frame has not passed the integrity detection of the manhole cover, the frame is discarded, and the process is returned to step 2) to continue the processing of the next frame; 6) If the image frames within the specified continuous time fail to pass the integrity detection of the manhole cover, an abnormal integrity alarm will be triggered, and the management personnel will be notified for manual processing. 3. The real-time detection method for node leakage of a heating system according to claim 2, wherein, for the image frame satisfying the integrity detection of the manhole cover, calculate the change of the cumulative sum of temperature offsets at the same time of the adjacent days, and exceed a threshold value. , the node leak alarm is triggered. 4 . The real-time detection method for node leakage of a heating system as claimed in claim 3 , wherein the first-level alarm, the second-level alarm and the third-level alarm are set according to the size of the threshold h value. 5 .

Images