A smart heating system and a smart heating system control method

By introducing neural network models and GPUs to process real-time data in geothermal heating systems, optimizing the operation of pressurized equipment, the problem of existing systems failing to effectively consider the energy loss of secondary heating pipelines and user feedback, achieving more efficient heating management and user-side thermal effects.

CN117190284BActive Publication Date: 2024-12-06XIAN XINCHENG DISTRICT RENEWABLE ENERGY CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202311312263.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2023-10-11
Publication Date
2024-12-06
Estimated Expiration
2043-10-11

AI Technical Summary

Technical Problem

The existing geothermal heating system fails to effectively consider the energy loss and user feedback information in the secondary heating pipeline network during regulation, resulting in inefficiency.

Method used

A smart heating system was designed, using neural network model and GPU for real-time data processing. By collecting basic information and user-side changes of heating pipelines, branch neural networks and comprehensive neural network units were established to form a regulatory scheme to optimize the addition of The operation of the pressure equipment realizes intelligent control of pressure, temperature and flow.

Benefits of technology

The efficiency and stability of the heating system are improved, and adaptive adjustments can be made based on real-time information. The pressurized equipment selects appropriate control plans at different stages, optimizes heating management, reduces energy losses, and improves the thermal effect on the user side.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN117190284B_ABST
    Figure CN117190284B_ABST
Patent Text Reader

Abstract

The present invention discloses a smart heating system and a smart heating system control method, wherein the smart heating system comprises: a heating pipeline for supplying heat to a user end, a pressurizing device capable of pressurizing the heating pipeline, and a control signal capable of controlling the pressurizing device to pressurize the heating pipeline. When the execution direction of the control signal is to control the pressurizing device to pressurize the heating pipeline and there is no human interference control, the control device controls the pressurizing device to enter a restricted control state based on the control signal; when the pressurizing device fails to enter the restricted control state, it automatically enters an automatic control state. The present application takes into account basic parameters of the heating pipelines in the primary heating network and the secondary heating network, namely, water pressure, temperature, and flow; and also includes regulation of the heating system by temperature changes due to weather factors, temperature differences between day and night, and user feedback.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of smart heating technology, and in particular to a smart heating system and a smart heating system control method. Background Art

[0002] Compared with traditional thermal heating, geothermal energy is a clean energy. Although geothermal utilization started late, in emerging areas, geothermal heating and intelligent technology have been fully combined to achieve the purpose of automatic regulation. However, the current regulation is only to set the minimum flow, pressure and temperature in the primary heating network, and it does not take into account the energy loss and user feedback information in the secondary heating network. Summary of the invention

[0003] In order to solve the problems of the prior art, the present invention provides an intelligent heating system and an intelligent heating system control method.

[0004] Its main technical solutions are as follows:

[0005] A smart heating system comprises: a heating pipeline for providing heating to a user end;

[0006] a pressurizing device capable of pressurizing the heating pipeline, and

[0007] Acquire a control signal capable of controlling the pressurizing device to pressurize the heating pipeline, wherein the control signal is formed by a control device making a comprehensive judgment based on the real-time comprehensive information collected by the collecting device and the heating model;

[0008] When the execution direction of the control signal is to control the pressurizing device to pressurize the heating pipeline and there is no human interference control, the control device controls the pressurizing device to enter the restricted control state based on the control signal. At this time, the pressurizing device pressurizes the heating management according to the first adjustment scheme of the control device to reach the set pressure value;

[0009] When the pressurizing device fails to enter the restricted control state, it automatically enters the automatic control state, so that when the pressurizing device pressurizes the heating pipeline, the pressurizing parameters formed by the pressurizing device based on human interference control are adaptively compared with the real-time comprehensive information collected by the collection device according to the set period, so as to select the corresponding second control scheme to complete the pressurization of the pressurizing device at different stages during the execution of the pressurization of the pressurizing device.

[0010] Furthermore, the collection device is used to collect basic information of the heating pipeline and / or changes based on the number of user terminals and / or changes in environmental factors.

[0011] Furthermore, the heating model has:

[0012] A configuration module is established based on the real-time comprehensive information collected by the collection device, and based on the configuration module, the neural network model is configured into a plurality of branch neural network units matching corresponding tasks and at least one group of comprehensive neural network units, and the branch neural network units and the comprehensive neural network units run on a GPU that supports basic control and matrix operations;

[0013] A plurality of storage units are provided on the GPU, and are used to connect the real-time data information collected by the collection device to different storage units according to different collection sources through a serial port or an interface, and each storage unit is configured to correspond one-to-one with the branch neural network unit. The branch neural network unit performs prediction by loading the real-time data information in the corresponding storage unit and obtaining the configuration information of the matrix operation in the GPU to obtain a single prediction result, and the single prediction result is cached accordingly;

[0014] Loading two or more of the single prediction results into the comprehensive neural network unit for comprehensive prediction to obtain a comprehensive prediction result;

[0015] The first control scheme for limiting the control state is formed by combining the basic configuration information of the heating pipeline control setting and calling a single prediction result and a comprehensive prediction result obtained based on the single prediction result.

[0016] Further, the comprehensive neural network unit performs prediction by loading a cached single prediction result and obtaining configuration information of matrix operations in the GPU to obtain a first comprehensive prediction result, and caches the first comprehensive prediction result;

[0017] And / or, the integrated neural network unit performs prediction by loading the cached first integrated prediction result and obtaining configuration information of matrix operations in the GPU to obtain a second integrated prediction result.

[0018] The present invention also provides a smart heating system control method, including the above-mentioned smart heating system comprising the following steps:

[0019] Acquire a control signal capable of controlling the pressurizing device to pressurize the heating pipeline;

[0020] When the execution direction of the control signal is to control the pressurizing device to pressurize the heating pipeline and there is no human interference control, the control device controls the pressurizing device to enter the restricted control state based on the control signal. At this time, the pressurizing device pressurizes the heating management according to the first adjustment scheme of the control device to reach the set pressure value;

[0021] When the pressurizing device fails to enter the restricted control state, it automatically enters the automatic control state, so that when the pressurizing device pressurizes the heating pipeline, the pressurizing parameters formed by the pressurizing device based on human interference control are adaptively compared with the real-time comprehensive information collected by the collection device according to the set period, so as to select the corresponding second control scheme to complete the pressurization of the pressurizing device at different stages during the execution of the pressurization of the pressurizing device.

[0022] Furthermore, the control signal is formed by the control device through comprehensive judgment based on the real-time comprehensive information collected by the collection device through the heating model.

[0023] Furthermore, the heating model has:

[0024] A plurality of branch neural network units and at least one group of integrated neural network units are constructed based on the acquisition source corresponding to the acquisition device and the configuration factors in the basic configuration information,

[0025] Encode the branch neural network unit once according to the acquisition source and the configuration factor in the basic configuration information, and obtain a single prediction result with the branch neural network unit;

[0026] Using a single encoding of the branch neural network unit, any two and / or any two or more branch neural network units are linked to a first integrated neural network unit, and the first integrated neural network unit is encoded twice, and a first integrated prediction result is obtained using the first integrated neural network unit;

[0027] and / or linking any two and / or any two or more first integrated neural network units to a second integrated neural network unit by secondary coding, and obtaining a second first integrated prediction result through the second integrated neural network unit;

[0028] The first control scheme for limiting the control state is formed by combining the basic configuration information of the heating pipeline control setting and calling a single prediction result and a first comprehensive prediction result and / or a second comprehensive prediction result obtained based on the single prediction result.

[0029] Furthermore, when the execution direction of the control signal is to control the pressurizing device to pressurize the heating pipeline and there is human interference control, the pressurizing device fails to enter the restricted control state, and a feedback signal is formed at the automatic control end of the pressurizing device. The pressurizing device inputs the feedback signal into the control device, and the control device enters the self-check state and controls the automatic control end of the pressurizing device to enter the automatic control state;

[0030] When the control device enters the self-check state, the existing human interference control is judged, and the pressurization parameters formed by the human interference control and the corresponding work log formed after the human interference control are obtained, the work log is parsed and the factor information pointed to by the human interference control is comprehensively obtained through the corresponding pressurization parameters, the pointed factor information is recorded, and the pointed factor information is judged, and based on the judgment result, it is selected whether to use it as the basic configuration information of the heating pipeline control setting;

[0031] When judging the pointed factor information, the recorded frequency of occurrence of the pointed factor information is synchronously queried. When the frequency of occurrence of the pointed factor information exceeds the set frequency, the judgment result is used as the basic configuration information of the heating pipeline control setting, and is written into the basic configuration information using the setting program.

[0032] Compared with traditional intelligent heating systems, this application takes into account basic parameters of heating pipelines in primary and secondary heating networks, such as water pressure, temperature, and flow; it also includes the regulation of the heating system due to weather factors, temperature changes caused by day and night temperature differences, and user feedback.

[0033] The present application sets up two control modes. One is that when the execution direction of the control signal is to control the pressurizing device to pressurize the heating pipeline and there is no human interference control, the control device controls the pressurizing device to enter a restricted control state based on the control signal. At this time, the pressurizing device pressurizes the heating management according to the first adjustment scheme of the control device to reach the set pressure value; at this time, it can be used to regulate the pressure when the temperature and flow in the heating pipeline are stable. The second mode is that when the pressurizing device fails to enter the restricted control state, it automatically enters the automatic control state, so that when the pressurizing device pressurizes the heating pipeline, the pressurizing parameters formed by the pressurizing device based on human interference control and the real-time comprehensive information collected by the collection device are adaptively compared according to the set period, so that during the execution process of the pressurizing device performing pressurization, the corresponding second control scheme is selected to complete the pressurization of the pressurizing device at different stages. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0035] Figure 1 It is a system framework principle diagram provided by the present invention;

[0036] Figure 2It is a schematic flow chart of the method provided by the present invention. DETAILED DESCRIPTION

[0037] In order to make the objectives, technical solutions and advantages of the present invention more clear, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings.

[0038] Embodiment 1, a smart heating system, comprising:

[0039] The heating pipeline used to supply heat to the user end,

[0040] a pressurizing device capable of pressurizing the heating pipeline, and

[0041] Acquire a control signal capable of controlling the pressurizing device to pressurize the heating pipeline, wherein the control signal is formed by a control device making a comprehensive judgment based on the real-time comprehensive information collected by the collecting device and the heating model;

[0042] When the execution direction of the control signal is to control the pressurizing device to pressurize the heating pipeline and there is no human interference control, the control device controls the pressurizing device to enter the restricted control state based on the control signal. At this time, the pressurizing device pressurizes the heating management according to the first adjustment scheme of the control device to reach the set pressure value;

[0043] When the pressurizing device fails to enter the restricted control state, it automatically enters the automatic control state, so that when the pressurizing device pressurizes the heating pipeline, the pressurizing parameters formed by the pressurizing device based on human interference control are adaptively compared with the real-time comprehensive information collected by the collection device according to the set period, so as to select the corresponding second control scheme to complete the pressurization of the pressurizing device at different stages during the execution of the pressurization of the pressurizing device.

[0044] In the above, the collection device is used to collect basic information of the heating pipeline and / or based on changes in the number of user terminals and / or changes in environmental factors.

[0045] In the above, the heating model has: a configuration module established according to the real-time comprehensive information collected by the collection device, and based on the configuration module, the neural network model is configured to have multiple branch neural network units matching corresponding tasks and at least one group of comprehensive neural network units, and the branch neural network units and the comprehensive neural network units run on a GPU that supports basic control and matrix operations;

[0046] A plurality of storage units are provided on the GPU, and are used to connect the real-time data information collected by the collection device to different storage units according to different collection sources through a serial port or an interface, and each storage unit is configured to correspond one-to-one with the branch neural network unit. The branch neural network unit performs prediction by loading the real-time data information in the corresponding storage unit and obtaining the configuration information of the matrix operation in the GPU to obtain a single prediction result, and the single prediction result is cached accordingly;

[0047] Loading two or more of the single prediction results into the comprehensive neural network unit for comprehensive prediction to obtain a comprehensive prediction result;

[0048] The first control scheme for limiting the control state is formed by combining the basic configuration information of the heating pipeline control setting and calling a single prediction result and a comprehensive prediction result obtained based on the single prediction result.

[0049] In the above, when the execution direction of the control signal is to control the pressurizing equipment to pressurize the heating pipeline and there is human interference control, the pressurizing equipment fails to enter the restricted control state, and a feedback signal is formed at the automatic control end of the pressurizing equipment. The pressurizing equipment inputs the feedback signal into the control device, the control device enters the self-check state, and controls the automatic control end of the pressurizing equipment to enter the automatic control state.

[0050] When the control device enters the self-check state, the existing human interference control is judged, and the pressurization parameters formed by the human interference control and the corresponding work log formed after the human interference control are obtained. The work log is parsed and the factor information pointed to by the human interference control is comprehensively obtained through the corresponding pressurization parameters. The pointed factor information is recorded, and the pointed factor information is judged, and based on the judgment result, it is selected whether to use it as the basic configuration information for the heating pipeline control setting.

[0051] When judging the pointed factor information, the recorded frequency of occurrence of the pointed factor information is synchronously queried. When the frequency of occurrence of the pointed factor information exceeds the set frequency, the judgment result is used as the basic configuration information of the heating pipeline control setting, and is written into the basic configuration information using the setting program.

[0052] The setting program uses the GPU as an operating carrier.

[0053] The configuration module links the branch neural network units and the integrated neural network unit according to set rules; and the number of the branch neural network units is determined by configuring the neural network model based on tasks, wherein the tasks are formed by the collection source corresponding to the collection device and the configuration factors in the basic configuration information.

[0054] The configuration factors are relatively constant parameters configured on the heating pipeline, including: water pressure, temperature, and flow rate in the heating pipeline.

[0055] In the above, the comprehensive neural network unit performs prediction by loading a cached single prediction result and obtaining configuration information of matrix operations in the GPU to obtain a first comprehensive prediction result, and caches the first comprehensive prediction result;

[0056] And / or, the integrated neural network unit performs prediction by loading the cached first integrated prediction result and obtaining configuration information of matrix operations in the GPU to obtain a second integrated prediction result.

[0057] In the above, the setting rules are:

[0058] Ⅰ: Encode the branch neural network unit once according to the task;

[0059] II: linking any two and / or any two or more branch neural network units to a first integrated neural network unit by first encoding the branch neural network unit, and performing secondary encoding on the first integrated neural network unit;

[0060] And / or III: Link any two and / or any two or more first integrated neural network units to a second integrated neural network unit using secondary coding.

[0061] The present application sets up two control modes. One is that when the execution direction of the control signal is to control the pressurizing device to pressurize the heating pipeline and there is no human interference control, the control device controls the pressurizing device to enter a restricted control state based on the control signal. At this time, the pressurizing device pressurizes the heating management according to the first adjustment scheme of the control device to reach the set pressure value; at this time, it can be used to regulate the pressure when the temperature and flow in the heating pipeline are stable. The second mode is that when the pressurizing device fails to enter the restricted control state, it automatically enters the automatic control state, so that when the pressurizing device pressurizes the heating pipeline, the pressurizing parameters formed by the pressurizing device based on human interference control and the real-time comprehensive information collected by the collection device are adaptively compared according to the set period, so that during the execution process of the pressurizing device performing pressurization, the corresponding second control scheme is selected to complete the pressurization of the pressurizing device at different stages.

[0062] This application collects basic information of the heating pipeline and / or forms temperature adjustment for the user end based on changes in the number of user ends and / or changes in environmental factors, and regulates the pressure when the water temperature and flow rate in the primary heating pipeline and the secondary heating pipeline are stable. The heating model is also used to adjust a single factor, such as weather and temperature difference between day and night, and can be adjusted comprehensively under multiple factors. It should be noted that the general heating flow rate and water temperature are set according to the conventional settings, and only the pressure is different. For example, the pressure in the secondary heating pipeline directly affects the heating cycle state in the entire area. In theory, the higher the water pressure, the faster the circulation, and the more obvious the thermal effect at the user end. However, the water pressure cannot be too large and must be within the set range. Of course, this application can also control the flow rate when the temperature and pressure are stable, and can also adjust the temperature of hot water in the heating pipeline when the pressure and flow are stable.

[0063] The present application utilizes a neural network model as the basic unit of the heating model. The neural network model can be trained based on historical data to obtain the optimal control value. The present application can also analyze the artificial adjustment of the secondary heating pipeline formed by user feedback information. In this case, the user feedback information can obtain a very specific regional range, and the overall thermal effect (heating meets the standard and comfort situation) feedback information of users in a certain building can be understood. Therefore, artificial interference can be carried out according to the specific situation. When the frequency of artificial interference in a specific area is large, it means that the thermal effect in this area is poor. The logs of the heating operation after previous artificial interference can be analyzed to obtain basic information in different time periods after any adjustment (which can be user feedback information or data monitored by sensors such as temperature, pressure, etc.), so as to obtain a more satisfactory adjustment range. At this time, the adjustment range can be written to the control end of the secondary heating pipeline in this area.

[0064] Example 2

[0065] The present invention also provides a smart heating system control method, including the above-mentioned smart heating system comprising the following steps:

[0066] Acquire a control signal capable of controlling the pressurizing device to pressurize the heating pipeline;

[0067] When the execution direction of the control signal is to control the pressurizing device to pressurize the heating pipeline and there is no human interference control, the control device controls the pressurizing device to enter the restricted control state based on the control signal. At this time, the pressurizing device pressurizes the heating management according to the first adjustment scheme of the control device to reach the set pressure value;

[0068] When the pressurizing device fails to enter the restricted control state, it automatically enters the automatic control state, so that when the pressurizing device pressurizes the heating pipeline, the pressurizing parameters formed by the pressurizing device based on human interference control are adaptively compared with the real-time comprehensive information collected by the collection device according to the set period, so as to select the corresponding second control scheme to complete the pressurization of the pressurizing device at different stages during the execution of the pressurization of the pressurizing device.

[0069] Furthermore, the control signal is formed by the control device through comprehensive judgment based on the real-time comprehensive information collected by the collection device through the heating model.

[0070] Furthermore, the heating model has:

[0071] A plurality of branch neural network units and at least one group of integrated neural network units are constructed based on the acquisition source corresponding to the acquisition device and the configuration factors in the basic configuration information,

[0072] Encode the branch neural network unit once according to the acquisition source and the configuration factor in the basic configuration information, and obtain a single prediction result with the branch neural network unit;

[0073] Using a single encoding of the branch neural network unit, any two and / or any two or more branch neural network units are linked to a first integrated neural network unit, and the first integrated neural network unit is encoded twice, and a first integrated prediction result is obtained using the first integrated neural network unit;

[0074] and / or linking any two and / or any two or more first integrated neural network units to a second integrated neural network unit by secondary coding, and obtaining a second first integrated prediction result through the second integrated neural network unit;

[0075] The first control scheme for limiting the control state is formed by combining the basic configuration information of the heating pipeline control setting and calling a single prediction result and a first comprehensive prediction result and / or a second comprehensive prediction result obtained based on the single prediction result.

[0076] Furthermore, when the execution direction of the control signal is to control the pressurizing device to pressurize the heating pipeline and there is human interference control, the pressurizing device fails to enter the restricted control state, and a feedback signal is formed at the automatic control end of the pressurizing device. The pressurizing device inputs the feedback signal into the control device, and the control device enters the self-check state and controls the automatic control end of the pressurizing device to enter the automatic control state;

[0077] When the control device enters the self-check state, the existing human interference control is judged, and the pressurization parameters formed by the human interference control and the corresponding work log formed after the human interference control are obtained, the work log is parsed and the factor information pointed to by the human interference control is comprehensively obtained through the corresponding pressurization parameters, the pointed factor information is recorded, and the pointed factor information is judged, and based on the judgment result, it is selected whether to use it as the basic configuration information of the heating pipeline control setting;

[0078] When judging the pointed factor information, the recorded frequency of occurrence of the pointed factor information is synchronously queried. When the frequency of occurrence of the pointed factor information exceeds the set frequency, the judgment result is used as the basic configuration information of the heating pipeline control setting, and is written into the basic configuration information using the setting program.

[0079] Compared with traditional intelligent heating systems, this application takes into account basic parameters of heating pipelines in primary and secondary heating networks, such as water pressure, temperature, and flow; it also includes the regulation of the heating system due to weather factors, temperature changes caused by day and night temperature differences, and user feedback.

[0080] This application collects basic information of the heating pipeline and / or forms temperature adjustment for the user end based on changes in the number of user ends and / or changes in environmental factors, and regulates the pressure when the water temperature and flow rate in the primary heating pipeline and the secondary heating pipeline are stable. The heating model is also used to adjust a single factor, such as weather and temperature difference between day and night, and can be adjusted comprehensively under multiple factors. It should be noted that the general heating flow rate and water temperature are set according to the conventional settings, and only the pressure is different. For example, the pressure in the secondary heating pipeline directly affects the heating cycle state in the entire area. In theory, the higher the water pressure, the faster the circulation, and the more obvious the thermal effect at the user end. However, the water pressure cannot be too large and must be within the set range. Of course, this application can also control the flow rate when the temperature and pressure are stable, and can also adjust the temperature of hot water in the heating pipeline when the pressure and flow are stable.

[0081] The present application utilizes a neural network model as the basic unit of the heating model. The neural network model can be trained based on historical data to obtain the optimal control value. The present application can also analyze the artificial adjustment of the secondary heating pipeline formed by user feedback information. In this case, the user feedback information can obtain a very specific regional range, and the overall thermal effect (heating meets the standard and comfort situation) feedback information of users in a certain building can be understood. Therefore, artificial interference can be carried out according to the specific situation. When the frequency of artificial interference in a specific area is large, it means that the thermal effect in this area is poor. The logs of the heating operation after previous artificial interference can be analyzed to obtain basic information in different time periods after any adjustment (which can be user feedback information or data monitored by sensors such as temperature, pressure, etc.), so as to obtain a more satisfactory adjustment range. At this time, the adjustment range can be written to the control end of the secondary heating pipeline in this area.

[0082] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A smart heating system, characterized in that: It has: a heating pipeline for supplying heat to the user end, A pressurizing device capable of pressurizing the heating pipeline, and obtaining a control signal capable of controlling the pressurizing device to pressurize the heating pipeline, wherein the control signal is formed by a control device based on the real-time comprehensive information collected by the collection device and a comprehensive judgment made by a heating model; When the execution direction of the control signal is to control the pressurizing device to pressurize the heating pipeline and there is no human interference control, the control device controls the pressurizing device to enter the restricted control state based on the control signal. At this time, the pressurizing device pressurizes the heating management according to the first adjustment scheme of the control device to reach the set pressure value; When the pressurizing device fails to enter the restricted control state, it automatically enters the automatic control state, so that when the pressurizing device pressurizes the heating pipeline, the pressurizing parameters formed by the pressurizing device based on the human interference control are adaptively compared with the real-time comprehensive information collected by the collection device according to the set period, so as to select the corresponding second control scheme to complete the pressurization of the pressurizing device at different stages during the pressurization execution of the pressurizing device; The heating model has: A configuration module is established based on the real-time comprehensive information collected by the collection device, and based on the configuration module, the neural network model is configured into a plurality of branch neural network units matching corresponding tasks and at least one group of comprehensive neural network units, and the branch neural network units and the comprehensive neural network units run on a GPU that supports basic control and matrix operations; A plurality of storage units are provided on the GPU, and are used to connect the real-time data information collected by the collection device to different storage units according to different collection sources through a serial port or an interface, and each storage unit is configured to correspond one-to-one with the branch neural network unit. The branch neural network unit performs prediction by loading the real-time data information in the corresponding storage unit and obtaining the configuration information of the matrix operation in the GPU to obtain a single prediction result, and the single prediction result is cached accordingly; Loading two or more of the single prediction results into the comprehensive neural network unit for comprehensive prediction to obtain a comprehensive prediction result; The first control scheme for limiting the control state is formed by combining the basic configuration information of the heating pipeline control setting and calling a single prediction result and a comprehensive prediction result obtained based on the single prediction result.

2. The intelligent heating system according to claim 1, characterized in that: The collection device is used to collect basic information of the heating pipeline and / or changes based on the number of user terminals and / or changes in environmental factors.

3. The intelligent heating system according to claim 1, characterized in that: The comprehensive neural network unit performs prediction by loading a cached single prediction result and obtaining configuration information of matrix operations in the GPU to obtain a first comprehensive prediction result, and caches the first comprehensive prediction result; And / or, the integrated neural network unit performs prediction by loading the cached first integrated prediction result and obtaining configuration information of matrix operations in the GPU to obtain a second integrated prediction result.

4. A smart heating system control method, comprising any one of the smart heating systems of claims 1-3, characterized in that: The steps include: Acquire a control signal capable of controlling the pressurizing device to pressurize the heating pipeline; When the execution direction of the control signal is to control the pressurizing device to pressurize the heating pipeline and there is no human interference control, the control device controls the pressurizing device to enter the restricted control state based on the control signal. At this time, the pressurizing device pressurizes the heating management according to the first adjustment scheme of the control device to reach the set pressure value; When the pressurizing device fails to enter the restricted control state, it automatically enters the automatic control state, so that when the pressurizing device pressurizes the heating pipeline, the pressurizing parameters formed by the pressurizing device based on the human interference control are adaptively compared with the real-time comprehensive information collected by the collection device according to the set period, so as to select the corresponding second control scheme to complete the pressurization of the pressurizing device at different stages during the pressurization execution of the pressurizing device; The heating model has: A plurality of branch neural network units and at least one group of integrated neural network units are constructed based on the acquisition source corresponding to the acquisition device and the configuration factors in the basic configuration information, Encode the branch neural network unit once according to the acquisition source and the configuration factor in the basic configuration information, and obtain a single prediction result with the branch neural network unit; Using a single encoding of the branch neural network unit, any two and / or any two or more branch neural network units are linked to a first integrated neural network unit, and the first integrated neural network unit is encoded twice, and a first integrated prediction result is obtained using the first integrated neural network unit; and / or linking any two and / or any two or more first integrated neural network units to a second integrated neural network unit by secondary coding, and obtaining a second first integrated prediction result through the second integrated neural network unit; The first control scheme for limiting the control state is formed by combining the basic configuration information of the heating pipeline control setting and calling a single prediction result and a first comprehensive prediction result and / or a second comprehensive prediction result obtained based on the single prediction result.

5. The intelligent heating system control method according to claim 4, characterized in that: The control signal is formed by the control device through comprehensive judgment based on the real-time comprehensive information collected by the collection device and the heating model.

6. The intelligent heating system control method according to claim 4, characterized in that: When the execution direction of the control signal is to control the pressurizing device to pressurize the heating pipeline and there is human interference control, the pressurizing device fails to enter the limited control state, and a feedback signal is formed at the automatic control end of the pressurizing device. The pressurizing device inputs the feedback signal into the control device, and the control device enters the self-check state and controls the automatic control end of the pressurizing device to enter the automatic control state; When the control device enters the self-check state, the existing human interference control is judged, and the pressurization parameters formed by the human interference control and the corresponding work log formed after the human interference control are obtained, the work log is parsed and the factor information pointed to by the human interference control is comprehensively obtained through the corresponding pressurization parameters, the pointed factor information is recorded, and the pointed factor information is judged, and based on the judgment result, it is selected whether to use it as the basic configuration information of the heating pipeline control setting; When judging the pointed factor information, the recorded frequency of occurrence of the pointed factor information is synchronously queried. When the frequency of occurrence of the pointed factor information exceeds the set frequency, the judgment result is used as the basic configuration information of the heating pipeline control setting, and is written into the basic configuration information using the setting program.