Hot water supply network energy efficiency system optimization method and system based on artificial intelligence

Abstract

The embodiment of the invention discloses a heat supply water network energy efficiency system optimization method and system based on artificial intelligence, and the method comprises the steps: building a heat supply operation knowledge network comprising a plurality of energy efficiency monitoring knowledge points, building a target knowledge association link between knowledge elements according to the cooperation characteristics of heat supply operation activities under a time sequence, and carrying out the optimization of the energy efficiency system of a heat supply water network. According to the invention, comprehensive coverage and refined description of energy efficiency monitoring knowledge points of the hot water supply network unit in the operation process are realized, and deep optimization evaluation is carried out on the energy efficiency of the hot water supply network unit through the target artificial intelligence model. Therefore, whether the to-be-detected hot water supply network unit needs to be subjected to energy efficiency optimization in the system tracking optimization stage or not can be accurately judged, so that making and implementation of an energy efficiency improvement strategy of the hot water supply network are effectively guided, the overall energy efficiency level of the hot water supply network system is remarkably improved, energy consumption is reduced, and the environmental protection target of energy conservation and emission reduction is achieved.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to an artificial intelligence-based hot water supply network energy efficiency system optimization method and system. Background Art

[0002] With the acceleration of urbanization and the continuous growth of energy demand, the energy efficiency level of the hot water supply network system, as an important part of urban infrastructure, is directly related to energy utilization efficiency and environmental protection. However, traditional hot water supply network system management methods often rely on manual experience and simple data analysis, which makes it difficult to fully and accurately grasp the energy efficiency status of the hot water supply network during operation, resulting in frequent problems such as low energy efficiency and energy waste.

[0003] In order to improve the energy efficiency of the hot water supply network, relevant technologies have begun to try to apply artificial intelligence technology to the management and optimization of the hot water supply network system. However, most of the existing AI-based hot water supply network energy efficiency optimization methods have the following problems: First, there is a lack of comprehensive monitoring and detailed description of heating operation activities, which makes it impossible to accurately identify energy efficiency bottlenecks and potential optimization space; second, the timing correlation and synergy characteristics between heating operation activities are not fully considered, which makes the formulation of optimization strategies lack scientific basis. Summary of the invention

[0004] In view of this, an embodiment of the present invention provides a method and system for optimizing the energy efficiency system of a hot water supply network based on artificial intelligence.

[0005] In a first aspect, an embodiment of the present invention provides an artificial intelligence-based hot water supply network energy efficiency system optimization method, which is applied to an artificial intelligence-based hot water supply network energy efficiency system optimization system, including:

[0006] Obtain X heating operation activities corresponding to the hot water supply network unit to be detected in the system tracking and optimization stage;

[0007] Based on the energy efficiency monitoring knowledge points respectively corresponding to the X heating operation activities and the time series of the X heating operation activities in the system tracking optimization stage, a heating operation knowledge network including Y knowledge elements is generated, where different knowledge elements represent different energy efficiency monitoring knowledge points. Under the time series, for a heating operation activity combination consisting of the mth heating operation activity and the m+1th heating operation activity, when the mth heating operation activity is the first energy efficiency monitoring knowledge point and the m+1th heating operation activity is the second energy efficiency monitoring knowledge point, there is a target knowledge association link from the first knowledge element to the second knowledge element in the heating operation knowledge network, the first knowledge element represents the first energy efficiency monitoring knowledge point, the second knowledge element represents the second energy efficiency monitoring knowledge point, and the knowledge association parameter of the target knowledge association link represents the collaborative feature vector of the heating operation activity combination;

[0008] Based on the heating operation knowledge network and the energy efficiency monitoring feature vectors corresponding to the Y knowledge elements, the target artificial intelligence model is used to perform energy efficiency optimization evaluation on the hot water supply network unit to be tested, and an evaluation result is generated to characterize whether the hot water supply network unit to be tested needs to be energy efficiency optimized during the system tracking and optimization stage.

[0009] For example, in a possible implementation of the first aspect, the step of generating a heating operation knowledge network including Y knowledge elements based on the energy efficiency monitoring knowledge points corresponding to the X heating operation activities and the time series of the X heating operation activities in the system tracking optimization stage includes:

[0010] Classify the energy efficiency monitoring knowledge points corresponding to the X heating operation activities according to their nature, scope of application, and degree of association with the energy efficiency of the hot water supply network, and mark the energy efficiency monitoring knowledge points in each category after classification;

[0011] According to the time sequence of X heating operation activities in the system tracking and optimization stage, the heating operation activities and their corresponding energy efficiency monitoring knowledge points are grouped in chronological order, and for each energy efficiency monitoring knowledge point in each group, a knowledge element corresponding to the energy efficiency monitoring knowledge point is defined, and each knowledge element represents a corresponding energy efficiency monitoring knowledge point;

[0012] Based on the defined knowledge elements, a framework of a heating operation knowledge network is preliminarily constructed, and a preliminary connection relationship between various knowledge elements is determined in the framework, wherein the preliminary connection relationship is determined based on the temporal sequence of heating operation activities;

[0013] For each group consisting of adjacent heating operation activities, analyzing the collaborative characteristics between the groups, specifically including analyzing the collaborative characteristics between the groups under different combinations of heating operation activities, and determining the target knowledge association link from one knowledge element to another knowledge element according to the analysis results of the collaborative characteristics, and then determining the collaborative characteristic vector of each knowledge association link, wherein the collaborative characteristic vector is used to reflect the collaborative characteristics between the combinations of heating operation activities;

[0014] After adding the target knowledge association link and the collaborative feature vector of each knowledge association link to the framework, check whether the framework has any energy efficiency monitoring knowledge points or heating operation activities that are not included. If so, supplement the missing parts with defined knowledge elements, and determine the position and connection relationship of the supplemented defined knowledge elements in the heating operation knowledge network based on the relationship between the supplemented defined knowledge elements and the existing heating operation activities and knowledge elements, and finally generate a complete heating operation knowledge network including Y knowledge elements.

[0015] In a second aspect, an embodiment of the present invention provides an artificial intelligence-based hot water supply network energy efficiency system optimization system, comprising:

[0016] processor;

[0017] A memory, wherein a computer program is stored in the memory, and when the computer program is executed, the method for optimizing the energy efficiency system of a hot water supply network based on artificial intelligence described in the first aspect is implemented.

[0018] As mentioned above, in the embodiment of the present invention, the embodiment of the present application constructs a heating operation knowledge network containing multiple energy efficiency monitoring knowledge points, and establishes target knowledge association links between knowledge elements based on the collaborative characteristics of heating operation activities under the time series. It not only achieves comprehensive coverage and refined description of the energy efficiency monitoring knowledge points of the hot water supply network unit during operation, but also conducts in-depth optimization and evaluation of the energy efficiency of the hot water supply network unit through the target artificial intelligence model. It can accurately determine whether the hot water supply network unit to be tested needs to be energy-efficiently optimized during the system tracking and optimization stage, thereby effectively guiding the formulation and implementation of energy efficiency improvement strategies for the hot water supply network, significantly improving the overall energy efficiency level of the hot water supply network system, reducing energy consumption, and achieving the environmental protection goals of energy conservation and emission reduction. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 A schematic flow chart of the steps of a method for optimizing a hot water supply network energy efficiency system based on artificial intelligence provided by an embodiment of the present invention;

[0020] Figure 2 The embodiment of the present invention provides a method for executing Figure 1A structural schematic block diagram of the hot water supply network energy efficiency system optimization method based on artificial intelligence and the hot water supply network energy efficiency system optimization system based on artificial intelligence.

[0021] Specific steps

[0022] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Through the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0023] See also Figure 1 As shown:

[0024] Step S110, obtaining X heating operation activities corresponding to the hot water supply network unit to be detected in the system tracking optimization stage.

[0025] In this embodiment, the server is used to manage a large regional heating system, which includes multiple hot water network units. In this regional heating system, there is a hot water network unit A to be detected. The system tracking optimization phase is set to a specific time period, for example, from January 1, 2023 to March 31, 2023.

[0026] During this time period, the hot water network unit A will perform a variety of heating operation activities. For example, the first is the activity of starting the heating equipment. On January 1, as the temperature dropped, in order to meet the user's heating needs, the server sent a command to the hot water network unit A to start the heating boiler. This is the first heating operation activity, which involves the startup operation of the equipment and the setting of initial parameters, such as the initial water temperature is set to 50 degrees Celsius and the initial water flow is set to 10 cubic meters per hour.

[0027] Then, on January 5, due to some users' feedback that the indoor temperature was low, the hot water network unit A adjusted the heating parameters. After the server monitored the user's feedback, it controlled the hot water network unit A to increase the water temperature to 55 degrees Celsius and adjust the water flow to 12 cubic meters per hour. This is the second heating operation activity, which involves adjusting key parameters in the heating system to meet the user's comfort needs.

[0028] Then, on February 10, the hot water network unit A carried out a restart activity after equipment maintenance. Prior to this, the heating equipment had undergone some routine maintenance, such as cleaning pipes and checking valves. After the maintenance was completed, the server restarted the heating equipment and adjusted the water temperature to 52 degrees Celsius and the water flow rate to 11 cubic meters per hour. This is the third heating operation activity, which is related to the maintenance status of the equipment and the parameter settings after it was put back into operation.

[0029] On March 1, as the weather gradually warmed up, the hot water network unit A implemented an activity to reduce the heating intensity. Based on the weather forecast and the user's historical heat usage data, the server controlled the hot water network unit A to reduce the water temperature to 45 degrees Celsius and the water flow rate to 8 cubic meters per hour. This is the fourth heating operation activity, which mainly adjusts the operating status of the heating system according to changes in the external environment and changes in user demand.

[0030] In this system tracking and optimization phase, the four heating operation activities of the hot water network unit A are the X heating operation activities that the server needs to obtain. The server obtains detailed information of these activities, including the time when the activities occurred, the equipment operations involved, parameter adjustments, etc., through the communication interface with the hot water network unit A and the collection system of various operation data, so as to perform subsequent operations related to energy efficiency evaluation.

[0031] Step S120: Based on the energy efficiency monitoring knowledge points corresponding to the X heating operation activities and the time series sequence of the X heating operation activities in the system tracking and optimization stage, a heating operation knowledge network including Y knowledge elements is generated, where different knowledge elements represent different energy efficiency monitoring knowledge points. Under the time series sequence, for a heating operation activity combination consisting of the mth heating operation activity and the m+1th heating operation activity, when the mth heating operation activity is a first energy efficiency monitoring knowledge point and the m+1th heating operation activity is a second energy efficiency monitoring knowledge point, there is a target knowledge association link from the first knowledge element to the second knowledge element in the heating operation knowledge network, the first knowledge element represents the first energy efficiency monitoring knowledge point, the second knowledge element represents the second energy efficiency monitoring knowledge point, and the knowledge association parameter of the target knowledge association link represents a collaborative feature vector of the heating operation activity combination.

[0032] Continuing with the above hot water network unit A as an example, for the four heating operation activities obtained, the server begins to build a heating operation knowledge network.

[0033] First, the server analyzes the energy efficiency monitoring knowledge points corresponding to each heating operation activity. For the activity of starting the heating equipment on January 1, the energy efficiency monitoring knowledge points may include the initial energy consumption efficiency when the equipment is started, the impact of the startup process on the pressure of the entire hot water network, etc. The server classifies these knowledge points according to their nature, scope of action, and the degree of correlation with the energy efficiency of the hot water network. For example, the initial energy consumption efficiency when the equipment is started belongs to the energy consumption category, and its scope of action is the initial stage of equipment startup, which is directly related to the energy efficiency of the hot water network. The energy efficiency monitoring knowledge points in this category are then marked, such as "Start-Energy Consumption-1".

[0034] For the activity of adjusting heating parameters on January 5, the energy efficiency monitoring knowledge points may include the impact of parameter adjustment on heat energy transmission efficiency, energy waste under different parameter combinations, etc. The same classification and labeling are performed, assuming that the label is "Parameter Adjustment-Heat Energy Transmission-2".

[0035] For the restart activity after equipment maintenance on February 10, the energy efficiency monitoring knowledge points include the energy efficiency recovery of the equipment after maintenance, the difference in energy efficiency between the restart and the previous operating state, etc., marked as "Restart-Energy Efficiency Recovery-3".

[0036] For the activity of reducing heating intensity on March 1, the energy efficiency monitoring knowledge points include the relationship between reducing heating intensity and energy savings, the balance between user comfort and energy consumption, etc., marked as "Reducing Intensity-Energy Savings-4".

[0037] Then, according to the time series of heating operation activities in the system tracking and optimization stage, the heating operation activities and their corresponding energy efficiency monitoring knowledge points are grouped in chronological order. For example, the startup activities on January 1 and the parameter adjustment activities on January 5 are grouped together. For each energy efficiency monitoring knowledge point in each group, the knowledge element corresponding to the energy efficiency monitoring knowledge point is defined. For example, for the energy efficiency monitoring knowledge point "Startup-Energy Consumption-1", a knowledge element K1 is defined, which represents this energy efficiency monitoring knowledge point.

[0038] Next, based on the defined knowledge elements, the framework of the heating operation knowledge network is preliminarily constructed. In this framework, the preliminary connection relationship between the various knowledge elements is determined based on the temporal sequence of the heating operation activities. For example, since the startup activity precedes the parameter adjustment activity, there is a preliminary connection relationship in the framework from the knowledge element K1 representing the energy efficiency monitoring knowledge point of the startup activity to the knowledge element K2 representing the energy efficiency monitoring knowledge point of the parameter adjustment activity.

[0039] For each group consisting of adjacent heating operation activities, analyze the collaborative characteristics between each group. For example, for the group of the startup activity on January 1 and the parameter adjustment activity on January 5, analyze the collaborative characteristics. From the perspective of collaborative frequency, if this situation of adjusting parameters shortly after startup often occurs in similar heating cycles in the past, then this collaborative frequency can be used as part of the collaborative feature vector. In addition, from the perspective of thermal energy value, the initial thermal energy supply after the startup activity and the thermal energy adjustment amount corresponding to the parameter adjustment activity can be reflected in the collaborative feature vector. Assuming that the initial thermal energy supply after the startup activity is 1000 kcal, the parameter adjustment activity adjusts the thermal energy supply to 1200 kcal. These values ​​can be obtained through the server's thermal energy monitoring system for the hot water network.

[0040] According to the analysis results of the collaborative features, after determining the target knowledge association link from one knowledge element to another knowledge element, the collaborative feature vector of each knowledge association link is determined. For example, the collaborative feature vector of the target knowledge association link from K1 to K2 may be [5 (collaborative frequency), 200 (heat energy value change)].

[0041] Finally, the target knowledge association link and the collaborative feature vector of each knowledge association link are added to the framework. Check whether there are any energy efficiency monitoring knowledge points or heating operation activities that are not included in the framework. If so, supplement the missing parts with knowledge elements. For example, if there is a temporary water quality testing activity on February 5 that is not included, the server analyzes the energy efficiency monitoring knowledge points corresponding to this activity, such as the impact of water quality on thermal energy transfer efficiency, marked as "water quality testing-thermal energy transfer-5", defines knowledge element K5, and determines the position and connection relationship of K5 in the heating operation knowledge network based on its relationship with existing heating operation activities and knowledge elements. Finally, a complete heating operation knowledge network including Y knowledge elements is generated.

[0042] Step S130, based on the heating operation knowledge network and the energy efficiency monitoring feature vectors corresponding to the Y knowledge elements, use the target artificial intelligence model to perform energy efficiency optimization evaluation on the hot water supply network unit to be detected, and generate an evaluation result representing whether the hot water supply network unit to be detected needs to be energy efficiency optimized during the system tracking and optimization stage.

[0043] In this embodiment, after the heating operation knowledge network is constructed and the energy efficiency monitoring feature vectors corresponding to Y knowledge elements are determined, the target artificial intelligence model is used to perform energy efficiency optimization evaluation.

[0044] First, determine the association array corresponding to the heating operation knowledge network. Assume that Y=5 (according to the heating operation knowledge network constructed previously, there are 5 knowledge elements K1-K5), and the dimension of the association array is 5*5. For n, a represents different knowledge elements (n, a∈5), in the association array, the array member in the nth column and the ath row represents the knowledge association parameter of the knowledge association link that associates the knowledge element referred to by n to the knowledge element referred to by a. For example, the array member in the 1st column and the 2nd row may represent the knowledge association parameter of the knowledge association link from K1 to K2. This parameter may be the result of some quantization of the previously determined collaborative feature vector, such as [0.5 (quantized value of collaborative frequency), 0.2 (quantized value of thermal energy value change)].

[0045] Based on the energy efficiency monitoring feature vectors corresponding to the association array and Y knowledge elements, the target artificial intelligence model is used to perform energy efficiency optimization evaluation on the hot water supply network unit to be tested.

[0046] Based on the energy efficiency monitoring feature vectors corresponding to the five knowledge elements, the knowledge element feature array is determined. The knowledge element feature array includes five columns of array members, b∈5, where b represents the knowledge element, and the array member in the bth column represents the energy efficiency monitoring feature vector corresponding to the knowledge element referred to by b. For example, the energy efficiency monitoring feature vector corresponding to K1 may be [0.3 (energy consumption efficiency index), 0.1 (pressure impact index)]. This vector is obtained by analyzing and quantifying various relevant data in the startup activity, such as energy consumption data, pressure sensor data, etc.

[0047] For the energy efficiency monitoring feature vector of each knowledge element in the knowledge element feature array, the energy efficiency monitoring feature vector of the knowledge element is fused with all the knowledge association parameters of the corresponding row in the association matrix. Taking K1 as an example, the energy efficiency monitoring feature vector [0.3, 0.1] of K1 is fused with all the knowledge association parameters of the first row of the association matrix. Assume that the knowledge association parameters of the first row of the association matrix include [0.5, 0.2] to K2 and [0.3, 0.1] to K3, etc. The fusion method can be weighted summation and other methods to obtain the fused feature vector. For example, the fused feature vector may be [0.3*0.5+0.3*0.3 (energy consumption efficiency related fusion), 0.1*0.2+0.1*0.1 (pressure impact related fusion)].

[0048] The principal component analysis and linear discriminant analysis are used to extract the main features of the fused feature vector and generate a set of feature vectors after dimensionality reduction. For example, after principal component analysis, it is found that some dimensions in the fused feature vector contribute less to the energy efficiency evaluation. These dimensions can be removed to obtain the feature vector after dimensionality reduction. Assume that the feature vector corresponding to K1 after dimensionality reduction becomes [0.2 (main energy consumption efficiency feature)].

[0049] Arrange the reduced eigenvectors in a set order to construct a feature input matrix. Each row of this feature input matrix represents the main energy efficiency feature of a knowledge element, and each column represents the value of different knowledge elements on the same feature dimension. For example, the feature input matrix may be:

[0050] |0.2(K1 main feature)|0.3(K2 main feature)|0.1(K3 main feature)|0.25(K4 main feature)|0.15(K5 main feature)|

[0051] Input the feature input matrix into the target artificial intelligence model. The target artificial intelligence model is previously trained. It optimizes and evaluates the energy efficiency of the hot water network unit to be tested based on the input feature matrix and generates corresponding evaluation results. For example, if the model output is 1, it means that energy efficiency optimization is required; if the output is 0, it means that energy efficiency optimization is not required.

[0052] Based on the evaluation results, the energy efficiency status of the hot water network unit to be tested can be further analyzed. If the evaluation results indicate that energy efficiency optimization is required, the server will determine the specific heating operation activities or links that need to be optimized. For example, if the optimization is required due to low energy consumption efficiency in the startup activity, the server will conduct a detailed analysis of the equipment and parameters related to the startup activity, determine areas for improvement, such as adjusting the fuel supply strategy at startup, and then perform energy efficiency optimization adjustments on the hot water network unit to be tested.

[0053] Based on the above steps, the embodiment of the present application constructs a heating operation knowledge network containing multiple energy efficiency monitoring knowledge points, and establishes target knowledge association links between knowledge elements based on the collaborative characteristics of heating operation activities under the time series. It not only achieves comprehensive coverage and refined description of the energy efficiency monitoring knowledge points of the hot water supply network unit during operation, but also conducts in-depth optimization and evaluation of the energy efficiency of the hot water supply network unit through the target artificial intelligence model. It can accurately determine whether the hot water supply network unit to be tested needs energy efficiency optimization during the system tracking and optimization stage, thereby effectively guiding the formulation and implementation of energy efficiency improvement strategies for the hot water supply network, significantly improving the overall energy efficiency level of the hot water supply network system, reducing energy consumption, and achieving the environmental protection goals of energy conservation and emission reduction.

[0054] For example, in a possible implementation, step S120 includes:

[0055] Step S121, classify the energy efficiency monitoring knowledge points corresponding to the X heating operation activities according to their properties, scope of application, and degree of association with the energy efficiency of the hot water supply network, and mark the energy efficiency monitoring knowledge points in each category after classification.

[0056] Step S122: according to the time series of the X heating operation activities in the system tracking and optimization stage, the heating operation activities and their corresponding energy efficiency monitoring knowledge points are grouped in chronological order, and for the energy efficiency monitoring knowledge points in each group, knowledge elements corresponding to the energy efficiency monitoring knowledge points are defined, and each knowledge element represents a corresponding energy efficiency monitoring knowledge point.

[0057] Step S123, preliminarily constructing a framework of a heating operation knowledge network based on the defined knowledge elements, and determining a preliminary connection relationship between various knowledge elements in the framework, wherein the preliminary connection relationship is determined based on the temporal sequence of heating operation activities.

[0058] Step S124, for each group composed of adjacent heating operation activities, analyze the collaborative characteristics between each group, specifically including analyzing the collaborative characteristics between each group under different combinations of heating operation activities, and according to the analysis results of the collaborative characteristics, determine the target knowledge association link from one knowledge element to another knowledge element, and then determine the collaborative characteristic vector of each knowledge association link, wherein the collaborative characteristic vector is used to reflect the collaborative characteristics between the combinations of heating operation activities.

[0059] Step S125, after adding the target knowledge association link and the collaborative feature vector of each knowledge association link to the framework, check whether there are energy efficiency monitoring knowledge points or heating operation activities that are not included in the framework. If so, supplement the missing parts with defined knowledge elements, and determine the position and connection relationship of the supplemented defined knowledge elements in the heating operation knowledge network based on the relationship between the supplemented defined knowledge elements and the existing heating operation activities and knowledge elements, and finally generate a complete heating operation knowledge network including Y knowledge elements.

[0060] In this embodiment, taking the previously mentioned hot water network unit A as an example, first, for the energy efficiency monitoring knowledge points corresponding to each heating operation activity of the hot water network unit A in the system tracking and optimization stage, the server classifies and marks them according to their nature, scope of action, and the degree of association with the energy efficiency of the hot water network. For example, for the heating operation activity of starting the heating equipment on January 1, the relevant energy efficiency monitoring knowledge points include the initial energy consumption efficiency when the equipment is started. The nature of this knowledge point is energy-related, and the scope of action is the moment when the equipment is started. It is directly related to the energy efficiency of the hot water network and can be marked as "Start-up-Energy Consumption-Initial Efficiency-1"; there is also the impact of the startup process on the pressure of the entire hot water network. The nature is the equipment operation impact type, the scope of action is the entire hot water network, and it is indirectly related to energy efficiency, and it is marked as "Start-up-Pressure Impact-1". For the activities of adjusting heating parameters on January 5, the impact of parameter adjustment on heat energy transmission efficiency is of energy transmission nature, and its scope of action is the heat energy transmission link in the heating network, which is directly related to energy efficiency, and is marked as "parameter adjustment-heat energy transmission-efficiency impact-2". The energy waste under different parameter combinations is of energy management nature, and its scope of action is the relevant range of heating operation parameters, which is directly related to energy efficiency, and is marked as "parameter adjustment-energy waste-2", etc.

[0061] Next, according to the time sequence of the heating operation activities in the system tracking and optimization stage, the server groups the heating operation activities and their corresponding energy efficiency monitoring knowledge points in chronological order. For example, the two activities of starting the heating equipment on January 1 and adjusting the heating parameters on January 5 and their corresponding energy efficiency monitoring knowledge points are grouped together. For the energy efficiency monitoring knowledge points in this group, define the corresponding knowledge elements. For the energy efficiency monitoring knowledge point marked as "Start-up-Energy Consumption-Initial Efficiency-1", define the knowledge element K1, which represents the energy efficiency monitoring knowledge point "Start-up-Energy Consumption-Initial Efficiency-1"; define the knowledge element K2 for "Start-Pressure Impact-1"; define the knowledge element K3 for "Parameter Adjustment-Heat Energy Transfer-Efficiency Impact-2"; and define the knowledge element K4 for "Parameter Adjustment-Energy Waste-2".

[0062] Then, based on these defined knowledge elements, the server preliminarily constructs the framework of the heating operation knowledge network and determines the preliminary connection relationship between each knowledge element based on the temporal sequence of heating operation activities. Since the activity of starting the heating equipment precedes the activity of adjusting the heating parameters, in the framework, the knowledge elements K1 and K2 representing the energy efficiency monitoring knowledge points related to starting the heating equipment and the knowledge elements K3 and K4 representing the energy efficiency monitoring knowledge points related to adjusting the heating parameters have a preliminary connection relationship. For example, preliminary connection relationships such as K1 connected to K3, K1 connected to K4, K2 connected to K3, and K2 connected to K4 reflect the correlation effect of the sequence of heating operation activities on the energy efficiency monitoring knowledge points.

[0063] Afterwards, for each group consisting of adjacent heating operation activities, the server deeply analyzes the collaborative features between each group. For the group of starting the heating equipment on January 1 and adjusting the heating parameters on January 5, its collaborative features are analyzed from different combinations of heating operation activities. For example, from the historical data of heating operation, in the past multiple similar heating cycles, the proportion of parameter adjustments within a period of time after each equipment start-up can be used as the collaborative frequency in the collaborative feature, assuming that this proportion is 30%. From the perspective of thermal energy value, the initial thermal energy supply value at the start of the activity is 1000 kcal, and the thermal energy supply after adjusting the parameters becomes 1200 kcal. The difference of 200 kcal is the collaborative feature related to the thermal energy value. Based on the analysis results of these collaborative features, the target knowledge association link from one knowledge element to another is determined. For example, there is a target knowledge association link from K1 to K3, and there is also a target knowledge association link from K2 to K4, and so on. Then determine the collaborative feature vector of each knowledge association link. For the knowledge association link from K1 to K3, the collaborative feature vector may be [0.3 (quantized value of collaborative frequency), 200 (quantized value of thermal energy value change)].

[0064] Finally, the server adds the target knowledge association link and the collaborative feature vector of each knowledge association link to the framework. Then check whether there are any energy efficiency monitoring knowledge points or heating operation activities that are not included in this framework. Assume that before the restart activity after equipment maintenance on February 10, there is a temporary water quality detection activity that is not included in the previous analysis. The server analyzes the energy efficiency monitoring knowledge points corresponding to this water quality detection activity, such as the impact of water quality on thermal energy transfer efficiency, marked as "water quality detection-thermal energy transfer-efficiency impact-5", and defines the knowledge element K5. According to the relationship between K5 and the existing heating operation activities and knowledge elements, its position and connection relationship in the heating operation knowledge network are determined. Since the water quality detection activity is before the restart activity after equipment maintenance, and the water quality will affect the thermal energy transfer efficiency and then affect the subsequent startup and operation, K5 has a certain connection relationship with K1, K2, and subsequent and restart-related knowledge elements. After such supplementary operations, a complete heating operation knowledge network including Y knowledge elements is finally generated. In this process, the server comprehensively and meticulously considers all aspects of heating operation activities and the relationships between them, thereby constructing a knowledge network that accurately reflects the energy efficiency-related situations of the hot water supply network.

[0065] In a possible implementation, step S130 includes:

[0066] Step S131, determine the association array corresponding to the heating operation knowledge network, the dimension of the association array is Y*Y, n, a represent different knowledge elements, n, a∈Y, in the association array, the array member in the nth column and the ath row represents the knowledge association parameters of the knowledge association link that associates the knowledge element referred to by n to the knowledge element referred to by a.

[0067] Step S132, based on the association array and the energy efficiency monitoring feature vectors corresponding to the Y knowledge elements, the target artificial intelligence model is used to perform energy efficiency optimization evaluation on the hot water supply network unit to be detected.

[0068] In a possible implementation, step S132 includes:

[0069] Step S1321, based on the energy efficiency monitoring feature vectors corresponding to the Y knowledge elements respectively, determine the knowledge element feature array, the knowledge element feature array includes Y columns of array members, b∈Y, b represents the knowledge element, and the bth column array member represents the energy efficiency monitoring feature vector corresponding to the knowledge element referred to by b.

[0070] Step S1322, for the energy efficiency monitoring feature vector of each knowledge element in the knowledge element feature array, the energy efficiency monitoring feature vector of the knowledge element is fused with all the knowledge association parameters of the corresponding row in the association matrix to generate a fused feature vector, and each feature vector in the fused feature vector set fuses the association information between the corresponding knowledge element and other knowledge elements.

[0071] Step S1323, principal component analysis and linear discriminant analysis are used to extract the main features of the fused feature vectors, and a reduced-dimensional feature vector set is generated, wherein each feature vector in the reduced-dimensional feature vector set contains the main energy efficiency features of the corresponding knowledge element.

[0072] Step S1321, arrange the reduced eigenvectors in a set order to construct a feature input matrix, wherein each row of the feature input matrix represents the main energy efficiency feature of a knowledge element, and each column represents the values ​​of different knowledge elements on the same feature dimension. The feature input matrix is ​​used as input data for the target artificial intelligence model.

[0073] Step S1321, input the feature input matrix into the target artificial intelligence model, so that the target artificial intelligence model optimizes and evaluates the energy efficiency of the hot water supply network unit to be detected according to the feature input matrix, and generates a corresponding evaluation result, which indicates whether the hot water supply network unit to be detected needs to be optimized in the system tracking optimization stage. And

[0074] The method further includes: analyzing the energy efficiency status of the hot water supply network unit to be tested according to the evaluation result; if the evaluation result indicates that energy efficiency optimization is required, after determining the specific heating operation activities or links that need to be optimized, energy efficiency optimization adjustment is performed on the hot water supply network unit to be tested.

[0075] In this embodiment, taking the previously mentioned hot water network unit A as an example, the association array corresponding to the heating operation knowledge network is first determined. Since the previously constructed heating operation knowledge network contains Y knowledge elements, assuming that Y=5 (corresponding to the knowledge elements K1-K5 in the previously constructed network), the dimension of the association array is 5*5. For n and a, which represent different knowledge elements (n, a∈5), for example, n=1 represents the knowledge element K1, and a=2 represents the knowledge element K2. In this association array, the array member in the first column and the second row represents the knowledge association parameter of the knowledge association link associated with K1 to K2. These knowledge association parameters are quantified based on information such as the collaborative feature vector between each knowledge element determined when the heating operation knowledge network was previously constructed. For example, the knowledge association link from K1 to K2, its collaborative feature vector may contain relevant characteristics of the heating operation activity such as collaborative frequency, thermal energy value change, etc., which have been quantified. Assume that this quantized knowledge association parameter is [0.3 (collaborative frequency quantification), 0.2 (thermal energy value change quantification)], which is used as the value of the array member in the first column and the second row. For the entire association array, the server will determine the knowledge association parameters of the knowledge association link corresponding to each position one by one. These parameters comprehensively reflect the characteristics such as the association relationship and association strength between various knowledge elements.

[0076] Next, based on the energy efficiency monitoring feature vectors corresponding to the association array and the five knowledge elements, the target artificial intelligence model is used to perform energy efficiency optimization evaluation on the hot water network unit to be tested. First, based on the energy efficiency monitoring feature vectors corresponding to the five knowledge elements, the knowledge element feature array is determined. This knowledge element feature array contains five columns of array members, b∈5, where b represents the knowledge element. For each knowledge element, the array member in the bth column represents the energy efficiency monitoring feature vector corresponding to the knowledge element referred to by b. Taking K1 as an example, it is assumed that the energy efficiency monitoring feature vector corresponding to K1 is obtained by analyzing the data related to the startup of the heating equipment, such as the energy consumption efficiency of the equipment when it is started, the impact on the pressure of the hot water network, and other data quantified results. If the energy consumption efficiency at startup is 0.3 (quantified by the proportion of energy consumption to total energy, etc.), and the impact on pressure is 0.1 (quantified by the ratio of pressure change to the standard pressure range), then the energy efficiency monitoring feature vector corresponding to K1 is [0.3, 0.1]. For other knowledge elements K2-K5, they will also be quantified according to the data of their corresponding energy efficiency monitoring knowledge points related to heating operation activities to obtain corresponding energy efficiency monitoring feature vectors. For example, K2 may correspond to [0.2, 0.15], K3 may correspond to [0.4, 0.05], etc., so as to construct a complete knowledge element feature array.

[0077] For the energy efficiency monitoring feature vector of each knowledge element in the knowledge element feature array, the server fuses the energy efficiency monitoring feature vector of the knowledge element with all the knowledge association parameters of the corresponding row in the association matrix to generate a fused feature vector. Taking K1 as an example, the energy efficiency monitoring feature vector of K1 is [0.3, 0.1], and the first row of the association matrix corresponds to the knowledge association parameters of K1 to other knowledge elements (K2-K5). Assume that the knowledge association parameters from K1 to K2 are [0.3, 0.2], the knowledge association parameters from K1 to K3 are [0.2, 0.1], the knowledge association parameters from K1 to K4 are [0.1, 0.05], and the knowledge association parameters from K1 to K5 are [0.05, 0.03]. The fusion method can be weighted summation, etc. The fusion calculation for the first element may be 0.3*0.3+0.3*0.2+0.3*0.1+0.3*0.05 (here it is assumed that the weights of the elements in the energy efficiency monitoring feature vector are the same, and the weights can be set according to different importance in practice). The fusion calculation for the second element is similar to obtain the fused feature vector. This operation is performed on each knowledge element in the knowledge element feature array, so that each feature vector in the fused feature vector set integrates the association information between the corresponding knowledge element and other knowledge elements.

[0078] Afterwards, the server uses principal component analysis and linear discriminant analysis to extract the main features of the fused feature vector and generate a set of feature vectors after dimensionality reduction. For the fused feature vector, for example, the fused feature vector corresponding to K1, it may contain information of multiple dimensions, but some of these dimensions contribute less to the final energy efficiency evaluation. Through principal component analysis, the server can determine which dimensions have a larger variance, that is, they have a greater impact on the overall data, and these dimensions are often the main features. Linear discriminant analysis further determines more distinguishing features based on the principle of maximizing the difference between different categories (here, it may be different state categories of energy efficiency, etc.). Assume that after analysis, it is found that some dimensions in the fused feature vector contribute very little to the energy efficiency evaluation. After the dimensionality reduction operation, the dimensionality reduction feature vector corresponding to K1 may become [0.25] (only the main energy efficiency features are retained). The same dimensionality reduction operation is performed on the fused feature vectors corresponding to other knowledge elements to obtain a set of dimensionality reduction feature vectors. Each feature vector in this set contains the main energy efficiency features of the corresponding knowledge element.

[0079] Then, the server arranges the reduced feature vectors in a set order to construct a feature input matrix. Each row of this feature input matrix represents the main energy efficiency feature of a knowledge element, and each column represents the value of different knowledge elements on the same feature dimension. For example, assuming that the feature vectors corresponding to K1-K5 after dimension reduction are [0.25], [0.3], [0.2], [0.15], [0.22] respectively, then the constructed feature input matrix is:

[0080] |0.25|0.3|0.2|0.15|0.22|

[0081] This feature input matrix comprehensively reflects the main energy efficiency characteristics of each knowledge element after processing, and is presented in a format suitable for model input as input data for the target artificial intelligence model.

[0082] Next, the server inputs the feature input matrix into the target artificial intelligence model. This target artificial intelligence model was previously trained with a large amount of data and can optimize and evaluate the energy efficiency of hot water network unit A based on the input feature matrix. The model analyzes and processes the input features based on its complex algorithm structure and trained parameters. For example, the model may be based on a neural network algorithm, which calculates and transmits multiple layers of neurons and finally outputs an evaluation result. If the model output is 1, it means that the hot water network unit A needs to be optimized for energy efficiency during the system tracking optimization stage; if the output is 0, it means that no energy efficiency optimization is required.

[0083] Finally, if the evaluation results indicate that energy efficiency optimization is needed, the server will further analyze the energy efficiency status of the hot water network unit A based on the evaluation results. The server will dig deep into the reasons behind the evaluation results and look for clues from the previously constructed heating operation knowledge network, knowledge element feature array and other related data. For example, if the need for optimization is due to the low energy efficiency corresponding to the knowledge elements related to starting the heating equipment, the server will determine the specific heating operation activities or links that need to be optimized, which may be the energy supply strategy at startup, the initial parameter settings of the equipment, etc. After determination, the server will take corresponding measures to optimize the energy efficiency of the hot water network unit A to be tested. For example, adjust the fuel supply speed at startup, optimize the initial water temperature or water flow of the equipment and other parameters to improve the energy efficiency of the hot water network unit A. Throughout the process, the heating operation knowledge network and related feature vectors and other information are fully utilized to accurately evaluate and adjust the energy efficiency optimization of the hot water network unit.

[0084] In a possible implementation, the step of determining the collaborative feature vector includes:

[0085] Based on the X heating operation activities and the time series, the coordination frequency of the heating operation activity combination is analyzed as the coordination feature vector.

[0086] Based on the X heating operation activities and the time series, the heat energy value corresponding to the heating operation activity of the second energy efficiency monitoring knowledge point performed by the heating operation activity combination is analyzed as the collaborative feature vector.

[0087] In this embodiment, first of all, from the perspective of collaborative frequency, the server will conduct a detailed analysis of the combination of heating operation activities. For example, in the system tracking and optimization stage, there is a combination of the activity of starting the heating equipment on January 1 and the activity of adjusting the heating parameters on January 5. The server will query the historical heating data to see how many times the parameter adjustment occurred shortly after the equipment was started in multiple similar heating cycles in the past. Assume that the server has saved the data of the past 10 heating cycles. In these 10 cycles, there were 4 times when the parameters were adjusted within a similar time after the equipment was started (for example, within 5 days after startup). Then the collaborative frequency of this combination of heating operation activities is 4 / 10=0.4, and this 0.4 can be used as the collaborative frequency part of the collaborative feature vector of this combination of heating operation activities.

[0088] When determining the synergistic characteristic vector from the perspective of thermal energy value, the same applies to the combination of heating operation activities. Take the combination of starting the heating equipment on January 1 and adjusting the heating parameters on January 5 as an example. When starting on January 1, according to the thermal energy output monitoring data of the heating equipment, the initial thermal energy supply value corresponding to the activity of starting the heating equipment is 1000 kcal. When adjusting the heating parameters on January 5, the heating operation activity of adjusting the parameters will cause the thermal energy supply to change. Assuming that the thermal energy supply becomes 1200 kcal after adjustment, then under this combination of heating operation activities, the thermal energy value change corresponding to the heating operation activity of adjusting the heating parameters, the second energy efficiency monitoring knowledge point, is 1200-1000=200 kcal. This 200 kcal is used as the synergistic characteristic vector of this combination of heating operation activities in terms of thermal energy value.

[0089] Another example is the combination of the restart activity after equipment maintenance on February 10 and the activity of reducing the heating intensity on March 1. The query of historical data shows that in the past 12 similar heating cycles, the restart after equipment maintenance to reduce the heating intensity occurred 3 times, and the synergy frequency is 3 / 12 = 0.25. In terms of thermal energy value, the thermal energy supply was 800 kcal when it was restarted on February 10, and the thermal energy supply became 600 kcal after reducing the heating intensity on March 1. The thermal energy value changed to 600-800 = -200 kcal, which is the synergy feature vector of this combination in terms of thermal energy value. By analyzing each combination of heating operation activities from the two aspects of synergy frequency and thermal energy value, the server can determine the synergy feature vectors corresponding to each combination of heating operation activities. These synergy feature vectors can accurately reflect the synergy characteristics between the combinations of heating operation activities, and provide an important basis for subsequent operations such as building a heating operation knowledge network.

[0090] In a possible implementation, the method further includes:

[0091] Step A110, matching the heating energy efficiency characteristics corresponding to the heating operation activity with the energy efficiency monitoring feature vector corresponding to the set energy efficiency monitoring knowledge point, the heating energy efficiency characteristics corresponding to the heating operation activity are used to express the heating operation activity, and the energy efficiency monitoring feature vector corresponding to the set energy efficiency monitoring knowledge point is used to express the set energy efficiency monitoring knowledge point.

[0092] Step A120: determining the energy efficiency monitoring knowledge point corresponding to the heating operation activity based on the matched set energy efficiency monitoring knowledge point.

[0093] The step of determining the energy efficiency monitoring knowledge points includes:

[0094] Based on the prior heating operation activities performed in the target heating application environment, Z energy efficiency influencing factors are determined, and the energy efficiency influencing factors include at least one influencing factor subcategory. The heating operation activities are performed by the hot water supply network unit to be detected in the target heating application environment.

[0095] The influencing factor subcategories corresponding to the Z energy efficiency influencing factors are integrated to generate the set energy efficiency monitoring knowledge points.

[0096] In this embodiment, taking the management of hot water network unit A as an example, the heating energy efficiency characteristics corresponding to the heating operation activities are matched with the energy efficiency monitoring characteristic vectors corresponding to the set energy efficiency monitoring knowledge points. For each heating operation activity in the hot water network unit A, such as the activity of starting the heating equipment on January 1, there is a corresponding heating energy efficiency characteristic. This heating energy efficiency characteristic is obtained by quantifying various data in the startup activity, including energy consumption at startup, the impact on the pressure of the entire hot water network, and the initial heat energy supply. For example, the proportion of energy consumption at startup to the total energy supply is 0.3, the impact on pressure is expressed as 0.1 through the ratio of the pressure change value to the standard pressure range, and the initial heat energy supply is 1000 kcal, etc. These data are combined to express the heating energy efficiency characteristics of this heating operation activity.

[0097] At the same time, the server has energy efficiency monitoring feature vectors corresponding to the set energy efficiency monitoring knowledge points. These set energy efficiency monitoring knowledge points are predetermined, and their corresponding energy efficiency monitoring feature vectors are also quantified. For example, there is a set energy efficiency monitoring knowledge point about the ideal state of normal startup of heating equipment. Its energy efficiency monitoring feature vector may include the ideal energy consumption ratio of 0.25, the ideal ratio of pressure impact of 0.08, the ideal initial heat supply range, etc.

[0098] The server matches the heating energy efficiency characteristics of the heating equipment with the energy efficiency monitoring characteristic vector of the set energy efficiency monitoring knowledge point. By comparing the values ​​of each element, such as energy consumption ratio, pressure impact ratio, etc., the best matching set energy efficiency monitoring knowledge point is found. If the energy consumption ratio is close, the pressure impact ratio is close, and the initial heat supply is within the set ideal range, the match is considered successful.

[0099] The energy efficiency monitoring knowledge points corresponding to the heating operation activities are determined based on the matched set energy efficiency monitoring knowledge points. Once the set energy efficiency monitoring knowledge point about the ideal state of normal startup of heating equipment is matched, it can be determined that the energy efficiency monitoring knowledge point corresponding to the activity of starting the heating equipment on January 1 is the energy efficiency monitoring knowledge point under normal startup, which contains relevant information such as energy consumption, pressure impact, and heat supply under ideal conditions.

[0100] For the determination steps of setting energy efficiency monitoring knowledge points. In the target heating application environment (such as the regional heating environment where the hot water network unit A is located), Z energy efficiency influencing factors are determined based on the prior heating operation activities previously performed. For example, for the prior heating operation activities in the previous heating cycle, the determined energy efficiency influencing factors may include heat energy consumption factors, equipment operation time factors, etc.

[0101] Taking the heat energy consumption factor as an example, it contains at least one influencing factor sub-category. This influencing factor sub-category can be the heat energy consumption in different time periods, such as the heat energy consumption during the peak heating period, the heat energy consumption during the valley heating period, etc. The influencing factor sub-category of the equipment operation time factor can be the operation time of different equipment (such as heating boilers, circulating pumps, etc.).

[0102] Then, the influencing factor subcategories corresponding to the Z energy efficiency influencing factors are integrated. The heat energy consumption in the peak heating period and the valley heating period in the heat energy consumption factor, as well as the different equipment operation time in the equipment operation time factor, are comprehensively processed. For example, the upper limit of heat energy consumption in the peak heating period, the lower limit of heat energy consumption in the valley heating period, and the reasonable operation time range of different equipment are integrated together to generate a set energy efficiency monitoring knowledge point. This set energy efficiency monitoring knowledge point contains various energy efficiency-related key information summarized from the prior heating operation activities, so as to facilitate the subsequent matching of the heating energy efficiency characteristics corresponding to the heating operation activities, thereby determining the energy efficiency monitoring knowledge point corresponding to the heating operation activities. In this way, the heating operation activities can be accurately associated with the energy efficiency monitoring knowledge points, providing an important basis for the energy efficiency evaluation and management of the entire hot water network unit.

[0103] In a possible implementation, the energy efficiency influencing factor includes a heat energy consumption factor, the influencing factor subclass corresponding to the heating operation activity under the heat energy consumption factor represents the heat energy consumption corresponding to the heating operation activity, the influencing factor subclass corresponding to the set energy efficiency monitoring knowledge point under the heat energy consumption factor represents the heat energy consumption interval corresponding to the set energy efficiency monitoring knowledge point, and the heat energy consumption corresponding to the heating operation activity is within the heat energy consumption interval corresponding to the energy efficiency monitoring knowledge point corresponding to the heating operation activity.

[0104] In a possible implementation, if the target heating application environment is a regional heating application environment, the energy efficiency influencing factor includes an ambient temperature factor, the influencing factor subclass corresponding to the heating operation activity under the ambient temperature factor represents the external ambient temperature information when the heating operation activity is executed, and the influencing factor subclass corresponding to the set energy efficiency monitoring knowledge point under the ambient temperature factor represents the external ambient temperature information when the heating operation activity belonging to the set energy efficiency monitoring knowledge point is executed.

[0105] In this embodiment, first of all, regarding the heat energy consumption factor. When the server processes the heating operation activity of the hot water network unit A, it will consider the heat energy consumption factor and its corresponding influencing factor subclass. For each heating operation activity, such as the activity of starting the heating equipment on January 1, its corresponding influencing factor subclass under the heat energy consumption factor is the heat energy consumption corresponding to the activity. The server obtains the data of this heat energy consumption in a variety of ways, such as installing a heat energy meter on the energy supply pipeline of the heating equipment. This meter can monitor and record the heat energy value consumed when starting the heating equipment in real time. Assume that this value is 1000 kcal.

[0106] As for the sub-category of influencing factors corresponding to the set energy efficiency monitoring knowledge point under the heat energy consumption factor, it represents the heat energy consumption range corresponding to the set energy efficiency monitoring knowledge point. These set energy efficiency monitoring knowledge points are pre-set based on historical data, industry standards or ideal operation models. For example, for the set energy efficiency monitoring knowledge point of normal startup of heating equipment, based on past normal operation data statistics and industry experience standards, the corresponding heat energy consumption range may be set to 800-1200 kcal.

[0107] The server will determine whether the heat energy consumption corresponding to the heating operation activity is within the heat energy consumption range corresponding to the energy efficiency monitoring knowledge point corresponding to the heating operation activity. In this example, the heat energy consumption of 1000 kcal corresponding to the heating equipment startup activity on January 1 is within the range of 800-1200 kcal. This judgment helps to determine whether the heating operation activity meets the expected energy efficiency standards. If the heat energy consumption corresponding to a heating operation activity exceeds this range, for example, when the heating equipment is started at another time point, the heat energy consumption reaches 1500 kcal, then this means that there may be energy efficiency problems in this startup activity, which may be caused by equipment failure, unreasonable parameter settings, etc., and further investigation is required.

[0108] Next, for the target heating application environment, the ambient temperature factor is used in the regional heating application environment. During the operation of the hot water network unit A, the ambient temperature factor is an important factor affecting energy efficiency. For each heating operation activity, the corresponding influencing factor subclass under the ambient temperature factor characterizes the external ambient temperature information when the heating operation activity is executed. For example, when adjusting the heating parameters on January 5, the server obtained the external ambient temperature of -5°C at that time through multiple temperature sensors installed in the area. These temperature sensors are distributed in different locations and can accurately reflect the average external ambient temperature of the entire heating area.

[0109] For the influence factor subclass corresponding to the set energy efficiency monitoring knowledge point under the ambient temperature factor, it represents the external ambient temperature information when the heating operation activity belonging to the set energy efficiency monitoring knowledge point is executed. For example, for the set energy efficiency monitoring knowledge point of reasonably adjusting heating parameters in cold weather, the corresponding external ambient temperature information may be in the range of -10℃ to 0℃ based on historical data and experience.

[0110] The server will make a comprehensive judgment based on the external ambient temperature information when the heating operation activity is executed and the external ambient temperature information corresponding to the set energy efficiency monitoring knowledge point. In this example, the external ambient temperature of -5°C when the heating parameters were adjusted on January 5th was within the range of -10°C to 0°C. This helps to evaluate the rationality of heating operation activities under specific ambient temperatures. If the external ambient temperature is 5°C during an activity to adjust the heating parameters, and this temperature is not within the set ambient temperature range for reasonable adjustment of the heating parameters, then this adjustment activity may need to be reviewed, as reasonable parameter adjustments may not be made based on the actual ambient temperature, resulting in possible energy waste or poor heating effects.

[0111] In the entire regional heating application environment, the server will continuously monitor and analyze the energy efficiency influencing factors such as the heat energy consumption factor and ambient temperature factor of each heating operation activity. For the heat energy consumption factor, by constantly comparing the heat energy consumption of the heating operation activity with the heat energy consumption range of the set energy efficiency monitoring knowledge point, possible energy efficiency anomalies can be discovered in a timely manner. For the ambient temperature factor, according to the matching of the external ambient temperature during the execution of the heating operation activity and the external ambient temperature information of the set energy efficiency monitoring knowledge point, the rationality of the heating operation activity under different ambient temperatures is ensured.

[0112] For example, in the restart activity after equipment maintenance on February 10, the server obtained the corresponding heat energy consumption of 800 kcal. By comparing it with the heat energy consumption range (assuming 700-900 kcal) of the set energy efficiency monitoring knowledge point corresponding to the restart after normal equipment maintenance, it was found to be within this range, indicating that the heat energy consumption met expectations. At the same time, the external ambient temperature at that time was -3°C, which matched the restart after equipment maintenance within a reasonable ambient temperature range (assuming -5°C to 2°C), indicating that the restart activity met the set energy efficiency monitoring requirements in terms of heat energy consumption and ambient temperature.

[0113] However, in the activity of reducing the heating intensity on March 1, if the heat consumption is 300 kcal, and the corresponding heat consumption range of the set energy efficiency monitoring knowledge point is 400-500 kcal, the heat consumption is lower than this range, which may indicate that there is an excessive reduction in the operation of reducing the heating intensity, which may affect the user's heating experience, and further check whether the parameter settings are reasonable. And if the external ambient temperature at that time is 10℃, and the reasonable external ambient temperature range for reducing the heating intensity is 5℃ to 15℃, although within this range, combined with the heat consumption, the server needs to comprehensively evaluate whether the best energy efficiency balance is achieved when reducing the heating intensity, and whether the parameters can be further optimized to improve energy efficiency while meeting user needs.

[0114] Through in-depth analysis and judgment of energy efficiency influencing factors such as heat energy consumption factor and ambient temperature factor, the heating operation activities of hot water network unit A can be better managed to ensure that it achieves the best energy efficiency while meeting the heating needs of users.

[0115] In a possible implementation, the step of training the target artificial intelligence model includes:

[0116] Step B110, obtaining sample learning data carrying energy efficiency optimization annotation data, wherein the energy efficiency optimization annotation data represents whether the template hot water supply network unit needs to perform energy efficiency optimization during the template system tracking and optimization phase, and the sample learning data represents the energy efficiency monitoring feature vector of the template heating operation knowledge network corresponding to the template hot water supply network unit during the template system tracking and optimization phase and the knowledge elements included in the template heating operation knowledge network.

[0117] Step B120, based on the sample learning data, using the initial artificial intelligence model to determine evaluation data characterizing whether the template hot water supply network unit needs to be energy-efficiently optimized during the template system tracking optimization phase.

[0118] Step B130, based on the loss function value between the evaluation data and the energy efficiency optimization labeled data, perform parameter learning on the initial artificial intelligence model to generate the target artificial intelligence model.

[0119] In this embodiment, it is assumed that there are multiple template hot water network units, and one of the template hot water network units B is taken as an example. During the corresponding template system tracking optimization phase (assuming it is from October 1, 2022 to December 31, 2022), the server will collect relevant sample learning data. This sample learning data includes the template heating operation knowledge network and the energy efficiency monitoring feature vectors of the knowledge elements contained in the network.

[0120] The template heating operation knowledge network is constructed by analyzing the heating operation activities of the template hot water network unit B during the template system tracking and optimization stage. For example, the heating operation activity of the template hot water network unit B starting the heating equipment on October 1, the corresponding energy efficiency monitoring knowledge points include the energy consumption efficiency when the equipment is started, the impact on the pressure of the hot water network, etc. According to the time sequence of the heating operation activities, such as the activity of adjusting the heating parameters on October 5, the corresponding energy efficiency monitoring knowledge points include the impact of parameter adjustment on the efficiency of thermal energy transmission, the energy waste under different parameter combinations, etc. These heating operation activities and their corresponding energy efficiency monitoring knowledge points constitute the template heating operation knowledge network. In addition, each knowledge element has a corresponding energy efficiency monitoring feature vector. For example, the energy consumption efficiency knowledge element when the equipment is started in the activity of starting the heating equipment, its energy efficiency monitoring feature vector may be quantified through energy consumption data, equipment power data, etc., and is assumed to be [0.3 (energy consumption ratio), 0.1 (deviation coefficient from standard energy consumption)].

[0121] At the same time, the energy efficiency optimization annotation data in the sample learning data characterizes whether the template hot water network unit B needs to be optimized for energy efficiency during the template system tracking optimization stage. This annotation data is obtained through manual evaluation or comparison based on historical best operation data. For example, if during this stage, based on the actual heating effect, energy consumption and comparison with the same period in history, it is found that the energy consumption is too high or the heating effect is not good, and it is determined that energy efficiency optimization is needed, then the energy efficiency optimization annotation data is marked as 1; if the heating operation is good and the energy consumption is within a reasonable range, it is marked as 0.

[0122] The initial artificial intelligence model is an initial artificial intelligence model based on a neural network structure, with an input layer, a hidden layer, and an output layer. The server inputs the association array corresponding to the template heating operation knowledge network (determined by the relationship between knowledge elements) and the sample learning data such as the energy efficiency monitoring feature vector of the knowledge element into the input layer of the initial artificial intelligence model. The neurons inside the initial artificial intelligence model process these data according to a pre-set algorithm. For example, in the hidden layer, the neurons perform weighted summation, activation function calculation and other operations on the input data, gradually extract features from the data and integrate them. Finally, an evaluation data is output at the output layer. This evaluation data is a judgment result of the initial artificial intelligence model on whether the template hot water network unit B needs to be optimized for energy efficiency. It may be a value between 0 and 1. A value close to 1 indicates that energy efficiency optimization tends to be required, and a value close to 0 indicates that energy efficiency optimization tends not to be required.

[0123] The loss function is a function used to measure the difference between the evaluation data and the energy efficiency optimization labeled data, such as the mean square error (MSE) function. Therefore, the MSE value between the evaluation data and the energy efficiency optimization labeled data can be calculated. If this value is large, it means that the prediction result of the initial artificial intelligence model is far from the actual labeled result, and the performance of the initial artificial intelligence model needs to be improved. Then, the server uses the back propagation algorithm to adjust the parameters in the initial artificial intelligence model according to the loss function value. For example, by adjusting the parameters such as the weights of the connections between neurons in the neural network, when the initial artificial intelligence model inputs the same sample learning data again, the difference between the output evaluation data and the energy efficiency optimization labeled data is reduced. After many such iterative processes, the parameters of the initial artificial intelligence model are continuously adjusted until the loss function value reaches an acceptable range. At this time, the initial artificial intelligence model becomes the target artificial intelligence model. This target artificial intelligence model can accurately judge whether the hot water network unit needs to be optimized for energy efficiency based on the input heating operation knowledge network and the energy efficiency monitoring feature vector of the knowledge element.

[0124] In a possible implementation, the method further includes:

[0125] If the evaluation result indicates that the hot water supply network unit to be detected needs to be optimized for energy efficiency during the system tracking optimization phase, a target heating operation activity corresponding to the energy efficiency optimization needs to be determined.

[0126] An energy efficiency warning is provided for the target hot water supply network unit associated with the target heating operation activity.

[0127] In this embodiment, taking the previously mentioned hot water network unit A as an example, after completing the energy efficiency evaluation of the hot water network unit A in the system tracking optimization stage, if the evaluation result shows that the hot water network unit A needs to be optimized for energy efficiency in this stage, the previously constructed heating operation knowledge network, the energy efficiency monitoring knowledge points corresponding to each heating operation activity, and the related feature vectors and other data are analyzed to determine the target heating operation activity corresponding to the energy efficiency optimization. For example, in the previous evaluation process, the server found that the energy efficiency corresponding to the knowledge elements related to starting the heating equipment in the heating operation knowledge network was low. In the activity of starting the heating equipment on January 1, by analyzing the energy consumption efficiency at startup, the impact on the pressure of the hot water network and other energy efficiency monitoring knowledge points, it was found that the energy consumption efficiency was 0.3, while the energy consumption efficiency of the normal startup according to the set standard or the same type of hot water network unit was 0.25. At the same time, the impact on the pressure of the hot water network at startup was 0.1. This pressure impact value is relatively high, indicating that it has a certain impact on the operating stability of the entire heating system, thereby determining that the heating operation activity of starting the heating equipment is a target heating operation activity that needs to be optimized for energy efficiency.

[0128] For example, in the activity of reducing heating intensity on March 1, it was found that the heat energy supply value after reducing the heating intensity was 300 kcal. According to comprehensive factors such as the external ambient temperature at that time (10°C) and the user's historical heating demand, the reasonable heat energy supply value should be between 400-500 kcal. This shows that there is an excessive reduction in the activity of reducing heating intensity, which affects the user's heating experience and may lead to poor overall energy efficiency. Therefore, the activity of reducing heating intensity has also been identified as a target heating operation activity.

[0129] After determining the target heating operation activity, the server will issue an energy efficiency warning to the target hot water network unit (here is hot water network unit A) associated with the target heating operation activity. The server will send warning information to the relevant management system or operator in a variety of ways. One way is to pop up a warning window on the interface of the monitoring system, clearly indicating that there is an energy efficiency problem with the hot water network unit A, and that specific heating operation activities (such as starting heating equipment or reducing heating intensity) need to be optimized. The warning window will list the relevant energy efficiency data in detail, such as the energy consumption efficiency and pressure impact value when starting the heating equipment, or the comparison between the actual heat supply value and the reasonable heat supply value when reducing the heating intensity, so that the operator can quickly understand the problem.

[0130] At the same time, the server can also send energy efficiency warning information to relevant personnel responsible for the maintenance and management of hot water network unit A through email or text messages. The content of the email or text message will include the identification of the hot water network unit A, the name of the target heating operation activity that needs to be optimized for energy efficiency, relevant energy efficiency data, and possible suggestions (such as for the activity of starting the heating equipment, it can be suggested to check whether the initial parameter settings of the equipment are reasonable, or for the activity of reducing the heating intensity, it is suggested to re-evaluate the heating strategy based on the ambient temperature and user needs, etc.). Through such energy efficiency warnings, relevant personnel can promptly learn about the energy efficiency problems of the hot water network unit A, so as to take corresponding measures to optimize the target heating operation activities and improve the energy efficiency of the hot water network unit A.

[0131] Based on the above description, in another embodiment, the present invention also provides an artificial intelligence-based hot water network energy efficiency system optimization system, see Figure 2 , Figure 2The structural diagram of the hot water supply network energy efficiency system optimization system 100 based on artificial intelligence provided in an embodiment of the present invention, the hot water supply network energy efficiency system optimization system 100 based on artificial intelligence may have relatively large differences due to different configurations or performances, and may include one or more central processing units (CPU) 112 (for example, one or more processors) and a memory 111. Among them, the memory 111 may be a short-term storage or a persistent storage. The program stored in the memory 111 may include one or more modules, and each module may include a series of instruction operations in the hot water supply network energy efficiency system optimization system 100 based on artificial intelligence. Furthermore, the central processing unit 112 can be configured to communicate with the memory 111, and execute a series of instruction operations in the memory 111 on the hot water supply network energy efficiency system optimization system 100 based on artificial intelligence.

[0132] The artificial intelligence-based hot water supply network energy efficiency system optimization system 100 may also include one or more power supplies, one or more communication units 113, one or more transmission output interfaces, and / or one or more operating systems, such as Windows ServerTM, Mac OS XTM, UnixTM, LinuxTM, FreeBSDTM, etc.

[0133] The steps performed by the hot water network energy efficiency system optimization system based on artificial intelligence in the above embodiment can be combined with Figure 2 The artificial intelligence-based hot water network energy efficiency system optimization system structure is shown.

[0134] In addition, an embodiment of the present invention further provides a storage medium, wherein the storage medium is used to store a computer program, and the computer program is used to execute the method provided by the above embodiment.

[0135] An embodiment of the present invention further provides a computer program product including instructions, which, when executed on a computer, enables the computer to execute the method provided in the above embodiment.

[0136] A person of ordinary skill in the art can understand that: all or part of the steps of implementing the above-mentioned method embodiment can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above-mentioned method embodiment; and the aforementioned storage medium can be at least one of the following media: read-only memory (English: Readonly Memory, abbreviated: ROM), RAM, magnetic disk or optical disk, etc. Various media that can store program codes.

[0137] It should be noted that each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device and system embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiments. The device and system embodiments described above are merely schematic, in which the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0138] The above is only a specific implementation step of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A method for optimizing the energy efficiency system of a hot water supply network based on artificial intelligence, characterized in that: include: Obtain X heating operation activities corresponding to the hot water supply network unit to be detected in the system tracking and optimization stage; Based on the energy efficiency monitoring knowledge points respectively corresponding to the X heating operation activities and the time series of the X heating operation activities in the system tracking optimization stage, a heating operation knowledge network including Y knowledge elements is generated, where different knowledge elements represent different energy efficiency monitoring knowledge points. Under the time series, for a heating operation activity combination consisting of the mth heating operation activity and the m+1th heating operation activity, when the mth heating operation activity is the first energy efficiency monitoring knowledge point and the m+1th heating operation activity is the second energy efficiency monitoring knowledge point, there is a target knowledge association link from the first knowledge element to the second knowledge element in the heating operation knowledge network, the first knowledge element represents the first energy efficiency monitoring knowledge point, the second knowledge element represents the second energy efficiency monitoring knowledge point, and the knowledge association parameter of the target knowledge association link represents the collaborative feature vector of the heating operation activity combination; Based on the heating operation knowledge network and the energy efficiency monitoring feature vectors corresponding to the Y knowledge elements, the target artificial intelligence model is used to perform energy efficiency optimization evaluation on the hot water supply network unit to be tested, and an evaluation result is generated to characterize whether the hot water supply network unit to be tested needs to be energy efficiency optimized during the system tracking and optimization stage.

2. The method for optimizing the energy efficiency system of a hot water supply network based on artificial intelligence according to claim 1, characterized in that: The energy efficiency optimization evaluation of the hot water supply network unit to be detected is performed using a target artificial intelligence model based on the heating operation knowledge network and the energy efficiency monitoring feature vectors corresponding to the Y knowledge elements, including: Determine an association array corresponding to the heating operation knowledge network, the dimension of the association array is Y*Y, n and a represent different knowledge elements, n and a∈Y, in the association array, the array member in the nth column and the ath row represents the knowledge association parameter of the knowledge association link that associates the knowledge element referred to by n to the knowledge element referred to by a; Based on the association array and the energy efficiency monitoring feature vectors corresponding to the Y knowledge elements, the target artificial intelligence model is used to perform energy efficiency optimization evaluation on the hot water supply network unit to be detected.

3. The method for optimizing the energy efficiency system of a hot water supply network based on artificial intelligence according to claim 2, characterized in that: The energy efficiency monitoring feature vectors corresponding to the association array and the Y knowledge elements are used to optimize the energy efficiency of the hot water supply network unit to be detected by using the target artificial intelligence model, including: Based on the energy efficiency monitoring feature vectors corresponding to the Y knowledge elements, respectively, a knowledge element feature array is determined, wherein the knowledge element feature array includes Y columns of array members, b∈Y, b represents a knowledge element, and the bth column of array members represents the energy efficiency monitoring feature vector corresponding to the knowledge element referred to by b; For the energy efficiency monitoring feature vector of each knowledge element in the knowledge element feature array, the energy efficiency monitoring feature vector of the knowledge element is fused with all the knowledge association parameters of the corresponding row in the association matrix to generate a fused feature vector, wherein each feature vector in the fused feature vector set fuses the association information of the corresponding knowledge element with other knowledge elements; Using principal component analysis and linear discriminant analysis to extract the main features of the fused feature vectors, and generating a set of feature vectors after dimensionality reduction, each feature vector in the set of feature vectors after dimensionality reduction contains the main energy efficiency features of the corresponding knowledge element; Arrange the reduced-dimensional feature vectors in a set order to construct a feature input matrix, where each row of the feature input matrix represents the main energy efficiency feature of a knowledge element, and each column represents the value of different knowledge elements on the same feature dimension. The feature input matrix is ​​used as input data for the target artificial intelligence model; Inputting the feature input matrix into the target artificial intelligence model, so that the target artificial intelligence model optimizes and evaluates the energy efficiency of the hot water supply network unit to be detected according to the feature input matrix, and generates a corresponding evaluation result, which indicates whether the hot water supply network unit to be detected needs to be optimized for energy efficiency during the system tracking optimization stage; and The method further comprises: According to the evaluation results, the energy efficiency status of the hot water supply network unit to be tested is analyzed. If the evaluation results indicate that energy efficiency optimization is needed, the energy efficiency optimization adjustment of the hot water supply network unit to be tested is carried out after determining the specific heating operation activities or links that need to be optimized.

4. The method for optimizing the energy efficiency system of a hot water supply network based on artificial intelligence according to claim 1, characterized in that: The step of determining the collaborative feature vector comprises: Based on the X heating operation activities and the time series, analyzing the coordination frequency of the heating operation activity combination as the coordination feature vector; Based on the X heating operation activities and the time series, the heat energy value corresponding to the heating operation activity of the second energy efficiency monitoring knowledge point performed by the heating operation activity combination is analyzed as the collaborative feature vector.

5. The method for optimizing the energy efficiency system of a hot water supply network based on artificial intelligence according to claim 1, characterized in that: The method further comprises: Matching the heating energy efficiency feature corresponding to the heating operation activity with the energy efficiency monitoring feature vector corresponding to the set energy efficiency monitoring knowledge point, wherein the heating energy efficiency feature corresponding to the heating operation activity is used to express the heating operation activity, and the energy efficiency monitoring feature vector corresponding to the set energy efficiency monitoring knowledge point is used to express the set energy efficiency monitoring knowledge point; Determining the energy efficiency monitoring knowledge point corresponding to the heating operation activity based on the matched set energy efficiency monitoring knowledge point; The step of determining the energy efficiency monitoring knowledge points includes: Determine Z energy efficiency influencing factors based on a priori heating operation activities performed in the target heating application environment, wherein the energy efficiency influencing factors include at least one influencing factor subclass, and the heating operation activities are performed by the hot water supply network unit to be detected in the target heating application environment; The influencing factor subcategories corresponding to the Z energy efficiency influencing factors are integrated to generate the set energy efficiency monitoring knowledge points.

6. The method for optimizing the energy efficiency system of a hot water supply network based on artificial intelligence according to claim 5, characterized in that: The energy efficiency influencing factor includes a heat energy consumption factor, and the influencing factor subclass corresponding to the heating operation activity under the heat energy consumption factor represents the heat energy consumption corresponding to the heating operation activity. The influencing factor subclass corresponding to the set energy efficiency monitoring knowledge point under the heat energy consumption factor represents the heat energy consumption interval corresponding to the set energy efficiency monitoring knowledge point, and the heat energy consumption corresponding to the heating operation activity is within the heat energy consumption interval corresponding to the energy efficiency monitoring knowledge point corresponding to the heating operation activity.

7. The method for optimizing the energy efficiency system of a hot water supply network based on artificial intelligence according to claim 5, characterized in that: If the target heating application environment is a regional heating application environment, the energy efficiency influencing factor includes an ambient temperature factor, the influencing factor subclass corresponding to the heating operation activity under the ambient temperature factor represents the external ambient temperature information when the heating operation activity is executed, and the influencing factor subclass corresponding to the set energy efficiency monitoring knowledge point under the ambient temperature factor represents the external ambient temperature information when the heating operation activity belonging to the set energy efficiency monitoring knowledge point is executed.

8. The method for optimizing the energy efficiency system of a hot water supply network based on artificial intelligence according to claim 1, characterized in that: The training steps of the target artificial intelligence model include: Acquire sample learning data carrying energy efficiency optimization annotation data, wherein the energy efficiency optimization annotation data represents whether the template hot water supply network unit needs to be energy-efficiently optimized during the template system tracking optimization phase, and the sample learning data represents the energy efficiency monitoring feature vector of the template heating operation knowledge network corresponding to the template hot water supply network unit during the template system tracking optimization phase and the knowledge elements included in the template heating operation knowledge network; Based on the sample learning data, using the initial artificial intelligence model to determine evaluation data characterizing whether the template hot water supply network unit needs to be energy-efficiently optimized during the template system tracking optimization phase; Based on the loss function value between the evaluation data and the energy efficiency optimization labeled data, parameter learning is performed on the initial artificial intelligence model to generate the target artificial intelligence model.

9. The method for optimizing the energy efficiency system of a hot water supply network based on artificial intelligence according to any one of claims 1 to 8, characterized in that: The method further comprises: If the evaluation result indicates that the hot water supply network unit to be detected needs to be optimized for energy efficiency during the system tracking and optimization phase, then determining a target heating operation activity corresponding to the energy efficiency optimization that needs to be performed; An energy efficiency warning is provided for the target hot water supply network unit associated with the target heating operation activity.

10. An artificial intelligence-based hot water network energy efficiency system optimization system, characterized in that: include: processor; A memory, wherein a computer program is stored in the memory, and when the computer program is executed, the method for optimizing the energy efficiency system of a hot water supply network based on artificial intelligence is implemented.

Images