Unattended heat exchange station operation and maintenance method and device, computer program and storage medium

Abstract

The invention discloses an unattended heat exchange station operation and maintenance method and device, a computer program and a storage medium, and relates to the field of heat exchange station operation and maintenance, in particular to the technical field of unattended heat exchange station operation and maintenance. The method comprises the following steps: S1, acquiring process parameters of the heat exchange station according to the reading of an instrument pointer; s2, establishing a load prediction model according to the process parameters; and S3, controlling output parameters of each heat source device according to the load prediction model. According to the method, the heat exchange station is unattended, the labor cost is remarkably reduced, the operation and maintenance efficiency is improved, the accuracy and reliability of data acquisition are ensured, interference of manual errors is avoided, the energy waste of heat source equipment is greatly reduced, the energy consumption is effectively reduced, and the working efficiency is improved. The emergency response capability of the heat exchange station is greatly improved, stable operation of the heat exchange station is ensured, and the heat exchange station can be widely applied to the heat supply industry and can also be applied to smart city construction or green building and energy conservation and emission reduction projects.

Description

Technical Field

[0001] The present invention relates to the field of operation and maintenance of heat exchange stations, and specifically to a technology for unattended operation and maintenance of heat exchange stations. Background Art

[0002] The operation and maintenance mode of traditional heat exchange stations mainly relies on manual operation and monitoring. The common practice is to conduct manual inspections, visually read the parameters of mechanical instruments, and manually control the heat source equipment in the station. This method faces a series of problems, including low operation efficiency, large data acquisition errors, and high energy consumption of heat source equipment, making it difficult to meet the high demands of the modern heating industry for intelligent and unmanned operation and maintenance. In traditional heat exchange stations, mechanical pointer instruments are used as the main data acquisition tools and are widely used for measuring parameters such as temperature and pressure. However, due to the complexity of the working environment of heat exchange stations and the dynamic changes of instrument data, traditional mechanical pointer instruments have defects such as unstable accuracy and susceptibility to environmental interference. Although high-precision electronic instruments can be replaced, these traditional mechanical pointer instruments are usually in a working state throughout the year, and shutdown maintenance and replacement will cause a wide range of impacts.

[0003] With the progress of modern technology, especially the rapid development of technologies such as machine vision, the Internet of Things, and artificial intelligence, the traditional manual monitoring method of heat exchange stations has gradually been unable to meet the industry's requirements. Manually reading instrument parameters is not only susceptible to human factors, but also has low efficiency for monitoring the equipment of large heat exchange stations, resulting in the inability of the heat source equipment in the station to respond quickly to process parameters, greatly affecting the operation efficiency and economic benefits of the heat exchange station. At the same time, intelligent and unmanned operation and maintenance have become a key direction for the transformation of the current heating industry. By introducing intelligent technologies, remote monitoring and automatic diagnosis of equipment can be achieved, significantly improving operation and maintenance efficiency and reducing labor costs. The intelligent identification and integrated monitoring technology of traditional heat exchange station equipment is not yet perfect, especially in the intelligent identification and data integration of mechanical pointer instruments. Existing technologies have not been able to effectively solve complex problems such as environmental changes and dynamic changes in instrument data. Summary of the Invention

[0004] In order to solve the above problems, the present invention proposes the following solutions:

[0005] An unattended operation and maintenance method for a heat exchange station, the method comprising:

[0006] S1. A step of obtaining process parameters of the heat exchange station according to the readings of instrument pointers;

[0007] S2. A step of establishing a load prediction model according to the process parameters;

[0008] S3. A step of controlling the output parameters of each heat source equipment according to the load prediction model.

[0009] Further, S1 includes: obtaining instrument pictures with dials of all instruments in the heat exchange station, obtaining the parameters of each instrument based on the instrument pictures of each instrument, and finally obtaining the process parameters of the heat exchange station. The process parameters include the supply water temperature, return water temperature, supply water pressure, and return water pressure parameters of the heat exchange station. The method for obtaining the parameters of each instrument based on the instrument pictures of each instrument is as follows:

[0010] S11. Use the YOLOv8 algorithm to detect and segment the instrument area in the instrument picture to obtain the segmented instrument picture;

[0011] S12. Perform grayscale transformation on the segmented instrument picture to obtain a grayscale picture;

[0012] S13. Use the Hough transform algorithm to detect and mark the instrument dial area in the grayscale picture to obtain a dial picture;

[0013] S14. Use the Canny algorithm to perform edge detection on the dial picture to obtain an edge picture;

[0014] S15. Use the Hough transform algorithm to detect and mark the position of the instrument pointer in the edge picture to obtain a pointer picture;

[0015] S16. Use the template matching method to obtain the upper and lower limits of the instrument scale in the dial picture, combine with the pointer picture, calculate the angle between the instrument needle and the lower limit of the instrument scale, and use the angle method to obtain the instrument pointer reading, that is, obtain a process parameter of the heat exchange station collected by this instrument.

[0016] Further, the Canny algorithm includes:

[0017]

[0018] where G is the pixel gradient of each dial picture, G x is the horizontal gradient, G y is the vertical gradient, is the gradient direction.

[0019] Further, the instrument pointer reading obtained in S16 is based on:

[0020]

[0021] where W is the instrument pointer reading, θ is the angle between the instrument needle and the lower limit of the instrument scale, θ max is the angle between the upper and lower limits of the instrument scale, W max is the upper limit of the instrument scale, W min is the lower limit of the instrument scale.

[0022] Further, S2 includes:

[0023] S21. Collect the heating area, historical heat load data, outdoor temperature, and weather change information of the heat exchange station, and combine with the process parameters obtained in S1 to form a data set D;

[0024] S22. Detect the outliers in the data set D using the box plot method, and replace the outliers with the data of the first 10 moments;

[0025] S23. Obtain the delay time of the data set D using the mutual information coefficient method;

[0026] S24. Combine the process parameters corresponding to the delay time and the output variables into a reconstructed data set;

[0027] S25. Use the reconstructed data set D c to train an extreme learning machine neural network and construct a load prediction model M:

[0028]

[0029] where y is the model output, W i is the input weight, β i is the output weight, b i is the input bias, g is the activation function, and X i is the i-th variable of the X column data.

[0030] Further, the S3 includes:

[0031] S31. Obtain the load prediction result within the next 72 hours of the heat exchange station according to the load prediction model;

[0032] S32. Optimize and reallocate the output parameters of each heat source device according to the load prediction result.

[0033] Based on the same inventive concept, the present invention further provides an unattended heat exchange station operation and maintenance device, and the device includes:

[0034] A module for obtaining the process parameters of the heat exchange station according to the reading of the instrument pointer;

[0035] A module for establishing a load prediction model according to the process parameters;

[0036] A module for controlling the output parameters of each heat source device according to the load prediction model.

[0037] Based on the same inventive concept, the present invention further provides a computer storage medium for storing a computer program, and when the computer program is read by a computer, the computer executes the method of the present invention.

[0038] Based on the same inventive concept, the present invention also provides a computer, comprising a processor and a storage medium. When the processor reads the computer program stored in the storage medium, the computer executes the method of the present invention.

[0039] Based on the same inventive concept, the present invention also provides a computer program product. As a computer program, when the computer program is read, the method of the present invention is implemented.

[0040] Beneficial effects

[0041] 1. The unattended heat exchange station operation and maintenance method proposed by the present invention realizes the intelligent recognition and automatic reading of mechanical pointer instruments by combining advanced computer vision technologies such as the YOLOv8 algorithm, Hough transform, and Canny edge detection algorithm. Initially, there were problems with unstable recognition accuracy under different instrument dial forms and lighting conditions. By continuously adjusting the cooperation logic of the algorithms, after many experiments and optimizations, the YOLOv8 is finally used to accurately locate the instrument, and the Hough transform and Canny algorithm are used to accurately determine the pointer position. This innovative solution breaks through the dependence on manual inspection and data reading, enables the heat exchange station to be unattended, significantly reduces labor costs, and improves operation and maintenance efficiency.

[0042] 2. The present invention successfully solves the error problem existing in the manual reading of instrument data in traditional heat exchange stations through image processing and algorithm optimization. During the R & D process, inaccurate recognition of sensors and image processing occurred due to different instrument types and aging problems. To overcome these problems, the team repeatedly tried to adjust the sensitivity of the algorithm and the filtering method, especially in complex environments such as low temperature and strong light, and conducted a large number of debuggings. Finally, the process parameters of the heat exchange station were accurately collected, ensuring the accuracy and reliability of data collection and avoiding the interference of manual errors.

[0043] 3. The present invention combines various parameters such as historical data and external weather information, and uses the extreme learning machine neural network (ELM) to construct a load prediction model for the heat exchange station. In the initial stage of R & D, the accuracy of the model prediction did not meet the actual requirements. Especially when the load fluctuated violently, the prediction deviation of the model was relatively large. The R & D team continuously tried to adjust the network structure, training algorithm, and input data set, and improved the model accuracy by introducing new external factors, such as real-time meteorological data. After multiple rounds of debugging and model iteration, these difficulties were finally overcome, realizing high-precision load prediction, being able to accurately identify and process abnormal data, and greatly reducing the energy waste of heat source equipment.

[0044] 4. The present invention realizes the dynamic adjustment of the output parameters of heat source equipment based on a load prediction model. In practice, the initial model failed to perfectly cope with extreme changes in load demand, resulting in situations where the equipment load was too large or too small in some periods. The team optimized the algorithm many times during this process, added a real-time adjustment mechanism, and precisely adjusted the equipment load through a multi-objective optimization algorithm to adapt to changes in different load demands. This innovative solution maximally improves the operating efficiency of the equipment and effectively reduces energy consumption, achieving the expected energy-saving effect.

[0045] 5. By combining intelligent instrument recognition and data acquisition methods with efficient load prediction and control strategies, the present invention realizes the rapid response of the heat exchange station in case of emergencies or load changes. During the R & D process, in the face of complex emergencies, such as sudden changes in external temperature, abnormal system load, etc., the initially designed emergency control strategy did not fully match the actual needs, resulting in a slow response speed. After many failed attempts and strategy adjustments, finally, by introducing a real-time feedback mechanism and an adaptive control algorithm, the emergency response ability of the heat exchange station has been greatly improved, ensuring the stability of the system, being able to handle emergencies, and ensuring the smooth operation of the heat exchange station.

[0046] 6. The present invention obtains accurate load prediction results through a load prediction model and reasonably allocates the output parameters of heat source equipment according to the load prediction results. In the early model, when facing special weather conditions, there were deviations in the calculation of heat supply, resulting in overheating or insufficient heating phenomena. The R & D team finally achieved accurate load prediction by repeatedly adjusting the load prediction algorithm and increasing the consideration of real-time environmental factors, avoiding overheating or insufficient heating, and optimizing the energy use efficiency. This innovation not only effectively reduces the overall energy consumption of the heat exchange station but also meets the requirements of energy conservation and emission reduction in the modern heating industry.

[0047] An unattended operation and maintenance method for a heat exchange station provided by the present invention can be widely applied to the heating industry and can also be applied to smart city construction or green building and energy conservation and emission reduction projects. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 is a flowchart for obtaining the process parameters of the heat exchange station according to the reading of the instrument pointer in Embodiment 2;

[0049] Figure 2 is a flowchart for establishing a load prediction model according to the process parameters in Embodiment 5;

[0050] Figure 3 is the schematic diagram of training an extreme learning machine neural network with the reconstructed dataset D c in Embodiment 5. Embodiment

[0052] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0053] Embodiment 1

[0054] This embodiment provides an unattended operation and maintenance method for a heat exchange station. The method includes:

[0055] S1. A step of obtaining the process parameters of the heat exchange station according to the readings of the instrument pointers;

[0056] S2. A step of establishing a load prediction model according to the process parameters;

[0057] S3. A step of controlling the output parameters of each heat source device according to the load prediction model.

[0058] This embodiment proposes an unattended operation and maintenance method for a heat exchange station. By automatically obtaining the process parameters of the heat exchange station and using the load prediction model to optimize the output control of the heat source device, the purpose of reducing manual intervention and improving operation and maintenance efficiency is achieved. First, the process parameters such as the supply water temperature, return water temperature, supply water pressure, and return water pressure are automatically obtained, avoiding the errors and inefficiencies caused by traditional manual inspections. Then, based on these process parameters, a load prediction model is constructed to provide accurate inputs for the intelligent control of the heat source device, thereby optimizing the operation of the device and ensuring that the heat exchange station is always in the best operating state. This embodiment reduces labor costs, improves operating efficiency, and optimizes the scheduling of heat source devices in a data-driven manner, avoiding energy waste and system overload problems, and promoting the realization of unattended intelligent operation and maintenance of the heat exchange station.

[0059] Embodiment 2

[0060] This embodiment is a further limitation of Embodiment 1. As Figure 1 shown, further, the S1 includes: obtaining instrument pictures with dials of all instruments in the heat exchange station, obtaining the parameters of each instrument according to the instrument pictures of each instrument, and finally obtaining the process parameters of the heat exchange station. The process parameters include the supply water temperature, return water temperature, supply water pressure, and return water pressure parameters of the heat exchange station. The method for obtaining the parameters of each instrument according to the instrument pictures of each instrument is:

[0061] S11. Using the YOLOv8 algorithm to detect and segment the instrument area in the instrument picture to obtain the segmented instrument picture;

[0062] S12. Performing gray-scale transformation on the segmented instrument picture to obtain a gray-scale picture;

[0063] S13. Detect and mark the instrument dial area in the grayscale image using the Hough transform algorithm to obtain the dial image;

[0064] S14. Perform edge detection on the dial image using the Canny algorithm to obtain the edge image;

[0065] S15. Detect and mark the position of the instrument pointer in the edge image using the Hough transform algorithm to obtain the pointer image;

[0066] S16. Use the template matching method to obtain the upper and lower limits of the instrument scale in the dial image. Combine with the pointer image, calculate the angle between the instrument needle and the lower limit of the instrument scale, and use the angle method to obtain the reading of the instrument pointer, that is, obtain a process parameter of the heat exchange station collected by the instrument.

[0067] This embodiment is further extended on the basis of Embodiment 1. Specifically, machine vision technology is used to intelligently identify the instrument pointer of the heat exchange station. By taking pictures of the instrument image and using the YOLOv8 algorithm to detect and segment the instrument area, a clear instrument image can be obtained. Then, use the Hough transform and Canny algorithms to process the image, accurately identify the edge of the instrument dial, and further calculate the reading of the pointer. This process greatly improves the accuracy of data collection in the operation and maintenance of the heat exchange station and avoids errors in manual operations. Especially in complex environments such as changes in light or environmental interference, this method can still efficiently and accurately extract the instrument readings and has strong adaptability. This embodiment significantly improves the intelligent level of the system, ensures the reliability of data collection, reduces human operation errors, and thus optimizes the accuracy and efficiency of the entire operation and maintenance process.

[0068] Embodiment 3

[0069] This embodiment further limits Embodiment 2. Further, the Canny algorithm includes:

[0070]

[0071] where G is the pixel gradient of each dial image, G x is the horizontal gradient, G y is the vertical gradient, is the gradient direction.

[0072] This embodiment further optimizes the Canny algorithm in Embodiment 2, clarifies the calculation methods of pixel gradient and gradient direction, thereby improving the accuracy of edge detection. Specifically, the Canny algorithm can more accurately extract the edge information in the image by calculating the pixel gradient and direction of each dial image. This optimization can ensure the accuracy of edge detection under complex environmental conditions, especially in the case of low light and large image noise, and still can stably identify the edges of the instrument dial and the pointer. This optimization scheme further improves the quality and stability of image processing, provides a more accurate basis for subsequent extraction of instrument pointer readings, and ensures that the data collection of the entire heat exchange station intelligent operation and maintenance system is more reliable and efficient.

[0073] Embodiment 4

[0074] This embodiment is a further limitation of Embodiment 2. Further, the instrument pointer reading obtained in S16 is based on:

[0075]

[0076] where W is the instrument pointer reading, θ is the angle between the instrument pointer and the lower limit of the instrument scale, θ max is the angle between the upper limit and the lower limit of the instrument scale, W max is the upper limit of the instrument scale, and W min is the lower limit of the instrument scale.

[0077] This embodiment calculates the angular relationship between the instrument pointer and the scale by the angle method, thereby accurately extracting the reading of the instrument pointer. In this embodiment, the reading of the instrument pointer is calculated by calculating the angle between the pointer and the lower limit of the scale, combining the upper and lower limits of the instrument, and using a formula. This method can effectively eliminate the measurement inaccuracy caused by instrument display errors or environmental interference, and ensure the high precision of instrument data. Through the accurate angle method, the accuracy of the data during the operation of the heat exchange station is improved, and the accuracy of the load prediction model and the regulation of the heat source equipment is enhanced.

[0078] Embodiment 5

[0079] This embodiment is a further limitation of Embodiment 1. As Figure 2 shown, further, the S2 includes:

[0080] S21. Collect the heating area, historical heat load data, outdoor temperature, and weather change information of the heat exchange station, and combine with the process parameters obtained in S1 to form a data set D;

[0081] S22. Detect the outliers in the data set D by the box plot method, and replace the outliers with the data of the first 10 moments;

[0082] S23. Obtain the delay time of the dataset D using the mutual information coefficient method;

[0083] S24. Combine the process parameters and output variables corresponding to the delay time into a reconstructed dataset;

[0084] S25. Use the reconstructed dataset D c to train an extreme learning machine neural network to construct a load prediction model M:

[0085]

[0086] where y is the model output, W i is the input weight, β i is the output weight, b i is the input bias, g is the activation function, and X i is the i-th variable of the X column data.

[0087] This embodiment further expands the construction process of the load prediction model. By using a multi-dimensional dataset for training, the accuracy of load prediction is improved. Specifically, considering factors such as the heating area of the heat exchange station, historical heat load data, outdoor temperature, and weather change information, a comprehensive dataset is established, and outlier detection is performed using the box plot method to ensure the quality of the dataset. At the same time, an extreme learning machine neural network is used to train this dataset to construct an efficient load prediction model. Through accurate load prediction, it can help the heat exchange station better grasp the future change trend of heat load, provide a scientific basis for the scheduling of heat source equipment, avoid energy waste and over-operation of the system, and improve the energy use efficiency of the system.

[0088] Embodiment Six

[0089] This embodiment further limits Embodiment One. Further, the S3 includes:

[0090] S31. Obtain the load prediction results of the heat exchange station within the next 72 hours according to the load prediction model;

[0091] S32. Optimize and re-allocate the output parameters of each heat source equipment according to the load prediction results.

[0092] In this embodiment, the heat source equipment is dynamically optimized and scheduled by combining the load prediction results, ensuring that the heat exchange station intelligently adjusts the output parameters of the heat source equipment according to the load demand within the next 72 hours. This not only improves the response speed of the heat source equipment but also enables timely adjustments when the load demand changes, ensuring the stable operation of the system. Through intelligent scheduling based on load prediction, the heat exchange station can reduce energy waste and avoid problems of unstable heat supply caused by over - operation or under - operation of the heat source equipment. Through intelligent equipment scheduling and optimization, the energy utilization efficiency of the system is further improved, energy consumption is reduced, and at the same time, it is ensured that the heat exchange station can more flexibly adapt to external environmental changes and achieve more efficient and accurate unattended operation.

[0093] Embodiment Seven

[0094] This embodiment presents an example for explaining Embodiments 1 to 6 above. Specifically, the example is as follows:

[0095] This embodiment aims to provide an intelligent operation and maintenance technology for unattended heat exchange stations. Relying on machine vision, intelligent control, and Internet of Things technologies, it realizes intelligent monitoring, equipment diagnosis, and energy - saving operation of the heating system, providing efficient and reliable intelligent operation and maintenance technology for the heating industry.

[0096] To achieve the above - mentioned purpose, this embodiment provides the following technical solutions: An intelligent operation and maintenance technology for unattended heat exchange stations, including an intelligent perception module, an intelligent operation and maintenance module, and an intelligent control module:

[0097] Step 1: The intelligent perception module relies on machine vision technology to realize the recognition of pointer - type instrument readings, including:

[0098] Step 1 - 1: Detect the instrument position through the target detection algorithm and extract the image area;

[0099] Step 1 - 2: Grayscale and denoise the image;

[0100] Step 1 - 3: Use the Hough circle transform to mark the circular area of the dial, and combine edge detection and Hough line transform to determine the pointer position;

[0101] Step 1 - 4: Obtain the upper and lower limits of the instrument based on the template matching method and calculate the reading according to the pointer angle.

[0102] Step 2: The intelligent operation and maintenance module realizes multi - heat - source collaborative control based on load prediction, including:

[0103] Step 2 - 1: Obtain the real - time parameters of the heat exchange station based on the intelligent perception module and various sensors;

[0104] Step 2 - 2: Based on historical data, real - time parameters, and neural network algorithms, establish a load prediction model for the heat exchange station;

[0105] Step 2-3: Based on the load prediction model, an optimization algorithm is used to optimize the control strategy to achieve the coordinated operation of multi-heat source equipment such as air source heat pumps, ground source heat pumps, and water source heat pumps.

[0106] Step 3: The intelligent control module realizes the automatic start and stop and remote control of the system-related equipment based on the real-time demand and load prediction;

[0107] Step 3-1: According to the real-time load demand and environmental parameters, the automatic start and stop of the heat exchange station system are realized;

[0108] Step 3-2: Through the cloud platform, remote monitoring and operation control of the unmanned heat exchange station are realized.

[0109] The invention will be further described below in conjunction with the attached drawings and specific implementation examples.

[0110] Step 1: Identify the readings of pointer-type instruments based on machine vision technology;

[0111] Step 1-1: Use a camera to take pictures of on-site instruments, and detect and segment the instrument area based on the VOLOv8 object detection algorithm;

[0112] Step 1-2: Perform gray-scale transformation on the segmented pictures, and use the Hough transform to detect the dial area in the image and mark it in the original image;

[0113] Step 1-3: Use the Canny algorithm for edge detection, and its calculation formula is expressed as follows:

[0114]

[0115] where G is the gradient of each pixel, G x is the horizontal gradient, G y is the vertical gradient, is the gradient direction.

[0116] Step 1-4: Use the Hough transform to detect the pointer position of the detected graph and mark it;

[0117] Step 1-5: Use the template matching method to obtain the upper and lower limits of the scale of the instrument dial, calculate the angle θ between the pointer and the lower limit, and calculate the pointer reading W using the angle method. Its calculation formula is expressed as follows:

[0118]

[0119] where θmax is the angle between the upper and lower limits of the instrument, W max is the upper limit value of the instrument, W min is the lower limit value of the instrument.

[0120] Step 2: Establish a heat network load prediction model M based on a deep neural network to achieve the coordinated operation of multi-heat source devices such as air source heat pumps, ground source heat pumps, and water source heat pumps;

[0121] The construction process of the initial prediction model M is as follows:

[0122] Step 2-1: Collect the process parameter dataset D related to the load, including heating area, historical heat load data, prediction time period, outdoor temperature, weather change, supply water temperature, return water temperature, supply water pressure, return water pressure, and circulating water mass flow rate;

[0123] Step 2-2: Use the box plot method to detect the outliers in the dataset D. If the outliers are directly removed, it will lead to discontinuous data in time. Replacing the outliers with the data of the first 10 moments can ensure data continuity.

[0124] Step 2-3: Use the mutual information coefficient method (Mutual Information, MI) to calculate the delay time of each variable in Step 2-1. The calculation formula is as follows:

[0125]

[0126] where p(x) is the marginal probability density function of X, p(y) is the marginal probability density function of Y, and p(x,y) is the joint probability density function of X and Y. X and Y are any two different columns of data in the dataset D

[0127] Step 2-4: Combine each group of process parameters corresponding to the delay time and the output variable Y t to form a reconstructed dataset Dc, where the first column represents the output data, and each remaining column is the data reconstructed according to the optimal delay moment;

[0128]

[0129] In the formula is the value of the mth group of variables at the corresponding delay moment km;

[0130] Step 2-5: As Figure 3 shown, use the reconstructed data to train the extreme learning machine neural network to obtain (W i , b i , β i ), and construct the load prediction model M. The prediction model M is expressed as follows:

[0131]

[0132] In the formula, W i = [w1, w2,... w L Tis the input weight, g(x) is the activation function, b i = [b1, b2, … b L T is the input bias, β i = [β1, β2, … β L T is the output weight, and y is the model output. X i is the i-th variable of the X column data.

[0133] Step 3: Based on the real-time demand and the optimized calculation results, realize the automatic start / stop and remote control of the system-related devices;

[0134] Step 3-1: The sensors in the heat exchange station will collect key parameters such as indoor temperature, outdoor temperature, supply and return water temperature, flow rate, and pressure in real time, and obtain the load prediction results;

[0135] Step 3-2: According to the load prediction results, optimize the calculation and re-allocate the output of each heat source. The intelligent control module will make a decision on whether to start or stop the operation of the heat exchange station. When the load demand is higher than the preset threshold, the module will trigger the start command of the heat exchange station. On the contrary, when the load demand continues to be lower than a certain threshold and it is predicted that it will not increase significantly in the future for a period of time, the module will issue a stop command;

[0136] Step 3-3: The sensors and control units in the heat exchange station will upload the real-time data to the data center of the cloud platform, and the operator can view the operation status and key parameters of the heat exchange station in real time through the cloud platform interface;

[0137] Step 3-4: The devices can be remotely started or stopped through the cloud platform.

[0138] The technical solutions provided by the present invention are further described in detail through the above several specific embodiments to highlight the advantages and beneficial effects of the technical solutions provided by the present invention. However, the above several specific embodiments are not used as a limitation to the present invention. Any reasonable modifications and improvements, combinations of embodiments, and equivalent replacements based on the spirit and principles of the present invention should be included within the protection scope of the present invention.

[0139] ​​Those skilled in the art can understand that the above is only the preferred embodiment of the present invention. The features described in each embodiment and / or claim of the present disclosure can be combined or combined in various ways, even if such combinations or combinations are not explicitly recorded in the present disclosure. It is not used to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments, or equivalently replace some of the technical features. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

[0140] Although the preferred embodiments of the present invention have been described, those skilled in the art can make additional changes and modifications once they know the basic creative concept. Therefore, the appended claims are intended to be construed as including the preferred embodiments and all changes and modifications falling within the scope of the present invention. Obviously, those skilled in the art can make various changes and variations to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention is also intended to include these modifications and variations.

Claims

1. An unattended heat exchange station operation and maintenance method, characterized in that: The method comprises: S1. A step of obtaining process parameters of a heat exchange station according to the reading of an instrument pointer; S2, a step of establishing a load forecasting model according to the process parameters; S3. A step of controlling output parameters of each heat source device according to the load prediction model.

2. The method according to claim 1, characterized in that S1 includes: obtaining instrument pictures containing dials of all instruments in the heat exchange station, obtaining parameters of each instrument according to the instrument picture of the instrument, and finally obtaining process parameters of the heat exchange station, wherein the process parameters include the supply water temperature, return water temperature, supply water pressure and return water pressure parameters of the heat exchange station. The method for obtaining the parameters of each instrument according to the instrument picture of the instrument is: S11, using the YOLOv8 algorithm to detect and segment the instrument area in the instrument image to obtain a segmented instrument image; S12, performing grayscale transformation on the segmented instrument image to obtain a grayscale image; S13, using the Hough transform algorithm to detect and mark the instrument dial area in the grayscale image to obtain a dial image; S14, using the Canny algorithm to perform edge detection on the dial image to obtain an edge image; S15, using the Hough transform algorithm to detect and mark the position of the instrument pointer in the edge image to obtain a pointer image; S16. Use the template matching method to obtain the upper and lower limits of the instrument scale in the dial image, combine the pointer image, calculate the angle between the instrument needle and the lower limit of the instrument scale, and use the angle method to obtain the instrument pointer reading, that is, obtain a heat exchange station process parameter collected by the instrument.

3. The method according to claim 2, characterized in that The Canny algorithm includes: Among them, G is the pixel gradient of each dial image, G x is the horizontal gradient, G y is the vertical gradient, is the gradient direction.

4. The method according to claim 2, characterized in that: The instrument pointer readings obtained in S16 are based on: accomplish, Where W is the meter pointer reading, θ is the angle between the meter needle and the lower limit of the meter scale, and θ max is the angle between the upper and lower limits of the instrument scale, W max is the upper limit of the instrument scale, W min The lower limit of the instrument scale.

5. The method according to claim 1, characterized in that: The S2 includes: S21, collect the heating area, historical heat load data, outdoor temperature, and weather change information of the heat exchange station, and combine them with the process parameters obtained in S1 to form a data set D; S22, using a box plot method to detect outliers in the data set D, and replacing the outliers with the data at the previous 10 moments; S23, obtaining the delay time of the data set D using the mutual information coefficient method; S24, combining the process parameters and output variables corresponding to the delay time into a reconstructed data set; S25, using the reconstructed dataset D c Train the extreme learning machine neural network and build the load forecasting model M: Among them, y is the model output, W i is the input weight, β i is the output weight, b i is the input bias, g is the activation function, X i is the i-th variable in the X column data.

6. The method according to claim 1, characterized in that: The S3 includes: S31. Obtaining load forecast results of the heat exchange station within the next 72 hours according to the load forecast model; S32. Optimize and reallocate output parameters of each heat source device based on load forecast results.

7. An unattended heat exchange station operation and maintenance device, characterized in that: The device comprises: A module for obtaining process parameters of the heat exchange station according to the readings of the instrument pointer; A module for establishing a load forecasting model according to the process parameters; A module for controlling the output parameters of each heat source device according to the load prediction model.

8. A computer storage medium for storing a computer program, characterized in that: When the computer program is read by a computer, the computer executes the method according to any one of claims 1 to 6.

9. A computer, comprising a processor and a storage medium, characterized in that: When the processor reads the computer program stored in the storage medium, the computer executes the method according to any one of claims 1 to 6.

10. A computer program product, being a computer program, characterized in that When the computer program is read, the method according to any one of claims 1 to 6 is implemented.

Images