A green building information data analysis system and method based on artificial intelligence

Abstract

The present invention relates to the field of green buildings, and specifically to a green building information data analysis system and method based on artificial intelligence; the system comprises an information collection module, a resource prediction module, a resource allocation module, and an over-capacity warning module; the information collection module is used to collect weather parameter change data, input resource acquisition node resource acquisition data, and output resource consumption node resource consumption data; the resource prediction module is used to predict the resource acquisition amount of each input resource acquisition node in a future time period, and is also used to predict the resource amount required to be consumed by each output resource consumption node in a future time period; the resource allocation module is used to calculate the resource acquisition loss and allocation contribution between each node, and analyze and calculate the maximum contribution allocation strategy; the over-capacity warning module issues an alarm message when the difference between the actual resource acquisition consumption and the predicted amount exceeds a threshold.

Description

Technical Field

[0001] The present invention relates to the technical field of green buildings, and specifically to a green building information data analysis system and method based on artificial intelligence. Background Art

[0002] As people pay more attention to health and environmental protection, they tend to prefer green buildings that are healthy, environmentally friendly and energy-saving. The transformation of energy structure has also made the popularization of green buildings an irreversible trend, which places high demands on the analysis and distribution of energy data in building energy management systems.

[0003] In today's green buildings, energy regeneration and sustainability mostly adopt the traditional method of providing resources on demand in real time. There is a lack of reasonable analysis of resource collection data and reasonable allocation strategies, and it is impossible to coordinate and manage resources and allocate resources reasonably. Different buildings have different functions, and the resource allocation weights and environmental benefits are also different. Generalized management lacks specificity and cannot reasonably adjust the focus of resource allocation. When resource consumption fluctuates, it is impossible to determine whether waste occurs, and it is impossible to save human resources while warning of node problems to effectively avoid waste. Summary of the invention

[0004] The purpose of the present invention is to provide a green building information data analysis system and method based on artificial intelligence to solve the problems raised in the above background technology.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0006] A green building information data analysis method based on artificial intelligence includes the following analysis steps:

[0007] Step S100: Divide the resources in the building into input resources, output resources, circulating resources, and non-circulating resources; wherein input resources are resources used by the building, output resources are resources consumed by the building, circulating resources are resources collected and utilized in a circular manner, and non-circulating resources are resources that are not collected naturally and are not recycled;

[0008] Step S200: Collect historical data collected by each input resource acquisition node and related weather parameter change data, use linear fitting to obtain the resource acquisition curve of each input resource acquisition node along with weather change, and then predict the input resource acquisition amount in each time period according to the weather forecast data of the geographical area;

[0009] Step S300: Collect consumption data of each output resource consumption node, perform periodic time series prediction, and predict resource consumption requirements of each output resource consumption node of the building in different time periods;

[0010] Step S400: Analyze the unit acquisition loss of each input resource, the unit allocation contribution of the output resource and the cyclic allocation contribution of the cyclic resource, and calculate the total allocation contribution of each unit of resource consumption during resource allocation according to the unit acquisition loss, unit allocation contribution and cyclic allocation contribution of each type of resource;

[0011] Step S500: Based on the predicted input resource acquisition and output resource consumption, comprehensively calculate the resource allocation contribution of each time period and derive the maximum contribution allocation strategy;

[0012] Step S600: Set a floating threshold coefficient for each input resource acquisition node and output resource consumption node. When the actual resource acquisition and consumption is greater than the product of the predicted resource change and the floating threshold coefficient, an alarm message is issued to remind the management personnel to check the nodes to avoid continuous waste.

[0013] According to the above technical solution, in step S200, weather parameters are selected according to resource types, corresponding values ​​of the weather parameters and the resource acquisition amount of the input resource acquisition node in each time period of historical time are collected, and a corresponding relationship curve is drawn using linear fitting;

[0014] Obtain forecast data of weather parameters for each time period in the future, and predict the amount of resources acquired by the input resource acquisition node in each time period based on the corresponding relationship curve;

[0015] Through linear fitting of weather parameters and the amount of sustainable resources collected in nature, the changing trend of building resource acquisition with weather is analyzed, and then the resource collection amount of each input resource acquisition node of the system is accurately predicted according to future weather changes.

[0016] According to the above technical solution, in step S300, the consumption data of each output resource consumption node in the building in the historical period is collected, and the data is normalized according to the formula:

[0017]

[0018] Among them, ω n Normalized data for the consumption of the nth output resource consumption node, is the amount of resources consumed in the i-th time period, S min is the minimum resource consumption of the nth output resource consumption node in the historical data, S max The maximum value of resource consumption of the nth output resource consumption node in the historical data;

[0019] The consumption data after processing is used to train the LSTM time series prediction model, and then the output resource consumption required by the nth output resource consumption node in the future time period is predicted;

[0020] Through time series prediction, we analyze the consumption data of each output resource consumption node in the building over time, analyze the time period relationship between the data and the building function, and accurately predict the resource consumption of the building in different time periods.

[0021] According to the above technical solution, in step S400, the unit acquisition loss of input resources refers to the loss of the input resource acquisition node acquiring one unit of input resources, the unit allocation contribution of output resources refers to the resource saving benefit when the output resource consumption node consumes one unit of the allocated output resources, and the cycle allocation contribution of the cycle resources refers to the resource cycle benefit obtained by the output resource node when consuming one unit of resources.

[0022] Calculate the resource allocation contribution between nodes according to the formula:

[0023] W m,n =P m,n +C m,n -E m ;

[0024] Among them, W m,n It refers to the allocation contribution obtained from obtaining one unit of resource usage from the mth input resource acquisition node to the nth output resource consumption node, P m,n The allocation contribution brought by the consumption of one unit of resources by the mth input resource acquisition node for the nth output resource consumption node, C m,n The cyclic allocation contribution obtained by consuming one unit of the resources obtained by the mth input resource acquisition node after the nth output resource consumption node recycles, E m Get the unit loss of the node for the mth input resource;

[0025] By refining the acquisition, loss and allocation contribution data of each unit resource acquisition, allocation, consumption and circulation in the building, the total allocation contribution of resource allocation in the building is calculated; among them, the allocation contribution of consumption and circulation is introduced when allocating resources between nodes, and the allocation contribution of each type of resource between nodes is adjusted to adjust the emphasis of the allocation strategy on different allocation paths and resource recycling.

[0026] When the quality of the resources obtained by the input resource acquisition node a cannot meet the needs of the output resource consumption node b, the calculation will be skipped. a,b Set to negative infinity; when the output resource consumption node a consumes the output resource consumption node b resources and cannot be recycled, set C a,b is 0;

[0027] Divide the predicted input resource acquisition and output resource consumption in the future time period by unit quantity, and calculate the allocation contribution of each unit resource; treat the acquisition and stock resources and the consumed resource quantities equally. When the input resource acquisition is greater than the output resource consumption, store the excess resources with large unit acquisition loss in the input resources as the stock resources allocated in the next time period, and do not allocate them in the current time period; when the input resource acquisition is less than the output resource consumption, add the stock resources to the acquisition resources for calculation. If it still cannot meet the resource consumption demand, use non-circulating resources to supplement them;

[0028] By dividing the resources obtained by different input resource acquisition nodes according to demand, the supply relationship between the acquired resources and the consumption nodes is refined to avoid general management that may result in substandard resource quality and the inability of the resources allocated to the output resource consumption nodes to meet demand during allocation. The acquired resources and consumed resources are allocated in quantity based on the resource supply and demand relationship. When too many resources are acquired, the surplus is stored to maximize the utilization of resources. When the acquired resources and stock resources cannot meet demand, external resources are introduced to balance supply and demand.

[0029] According to the above technical solution, in step S500, the predicted input resources, replenished stock resources and non-circulating resources are taken as upstream resources, the consumption predicted by the output resource consumption node is taken as downstream resources, the upstream resources and downstream resources are divided by units, and a resource consumption contribution matrix is ​​calculated and generated, wherein the rows of the matrix represent the upstream resources of each unit, the columns represent the downstream resources of each unit, and the elements of the xth row and yth column in the matrix represent the total contribution of the upstream resources of the xth unit to the yth downstream resource;

[0030] The upstream resources and downstream resources are combined into a bipartite graph. The elements in the resource consumption contribution matrix are used as the weights of the edges when matching the bipartite graph. The KM algorithm is used to analyze the weighted best perfect matching strategy of the bipartite graph to find the allocation strategy with the maximum total contribution of all upstream resources during allocation.

[0031] A green building information data analysis system based on artificial intelligence using a green building information data analysis method based on artificial intelligence in the above technical solution, the system comprising an information collection module, a resource prediction module, a resource allocation module, and an over-capacity warning module;

[0032] The information collection module is used to collect weather parameter change data, input resource acquisition node resource acquisition data, and output resource consumption node resource consumption data; the resource prediction module is used to predict the resource acquisition amount of each input resource acquisition node in the future time period, and is also used to predict the resource amount required to be consumed by each output resource consumption node in the future time period; the resource allocation module is used to calculate the resource acquisition loss and allocation contribution between each node, and analyze and calculate the maximum contribution allocation strategy; the over-consumption warning module issues an alarm message when the difference between the actual resource acquisition consumption and the predicted amount exceeds a threshold;

[0033] An over-alarm module is introduced into the system, which issues an alarm message when the resource collection or consumption of a node exceeds the expected threshold, promptly reminds you to check the relevant nodes, implements the system self-check function, and avoids continuous node failures or resource waste during use that affects the sustainable efficiency of resources.

[0034] The output end of the information collection module is connected to the input end of the resource prediction module and the excess warning module; the output end of the resource prediction module is connected to the input end of the resource allocation module and the excess warning module.

[0035] According to the above technical solution, the information collection module includes a weather parameter collection unit, an input resource collection unit, and an output resource collection unit;

[0036] The weather parameter collection unit is used to collect weather parameter change data in each time period around the building and predicted weather parameter values ​​in future time periods; the input resource collection unit is used to collect input resources used in the building resource circulation system and collect historical data on the collection amount of each input resource acquisition node; the output resource collection unit is used to collect historical data on the resource consumption amount of each output resource consumption node.

[0037] According to the above technical solution, the resource prediction module includes an input resource prediction unit and an output resource prediction unit;

[0038] The input resource prediction unit is used to perform linear fitting based on the weather parameters obtained by the information collection module and the historical data of the collection amount of each input resource acquisition node, analyze the relationship between the collection amount of the input resource acquisition node and the weather parameters, and then predict the collection amount of each input resource acquisition node in the future time period based on the predicted weather parameters; the output resource prediction unit uses the LSTM model to perform time series analysis and prediction based on the historical data of the consumption of each output resource consumption node obtained by the information collection module, and then predicts the resource amount required for each output resource consumption node in the future time period.

[0039] According to the above technical solution, the resource allocation module includes a contribution analysis unit, an allocation strategy analysis unit, a resource storage unit, and a non-circular resource introduction unit;

[0040] The contribution analysis unit is used to calculate the total contribution of each unit resource allocation; the allocation strategy analysis unit analyzes the maximum contribution allocation strategy when all upstream resources are allocated according to the total contribution of each unit resource allocation; the resource storage unit is used to store excess input resources for resource replenishment in the next time period; the non-circulating resource introduction unit is used to introduce non-circulating resources as a supplement when the input resources and stock resources cannot meet the output resource demand;

[0041] The contribution analysis unit input end is connected to the resource prediction module output end; the contribution analysis unit output end is connected to the allocation strategy analysis unit input end; the allocation strategy analysis unit output end is connected to the resource storage unit and the non-circular resource introduction unit input end.

[0042] Compared with the prior art, the beneficial effects achieved by the present invention are:

[0043] The scheme predicts resource supply and demand data for future time periods based on historical weather data, input resources, and output resource collection and consumption data, conducts resource data analysis in combination with actual conditions, refines and quantifies resources for reasonable allocation, uses the KM algorithm to analyze the best perfect matching strategy for resource supply and demand, maximizes the allocation contribution of resource utilization, improves the sustainable utilization rate of resources, and avoids resource waste caused by existing real-time management; introduces the allocation contribution of consumption and recycling of resource allocation between nodes, and adjusts the emphasis on resource allocation to different consumption nodes and the emphasis on recycling by adjusting the resource allocation contribution between nodes; the system also introduces an over-alarm module to realize the alarm prompt function when resource data is abnormal, and issues an alarm when the actual collected data of the node deviates too much from the predicted data, effectively avoiding continuous node failures and unintentional waste in the resource sustainability system. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0045] Figure 1 It is a structural diagram of a green building information data analysis system based on artificial intelligence;

[0046] Figure 2 It is a flow chart of green building information data analysis method based on artificial intelligence;

[0047] Figure 3 It is a weighted bipartite graph of a KM algorithm in a green building information data analysis method based on artificial intelligence of the present invention. DETAILED DESCRIPTION

[0048] 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 only part of the embodiments of the present invention, not all of the embodiments. Based on 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.

[0049] See also Figure 1-Figure 3 , the present invention provides a technical solution:

[0050] A green building information data analysis method based on artificial intelligence includes the following analysis steps:

[0051] Step S100: Divide the resources in the building into input resources, output resources, circulating resources, and non-circulating resources; wherein input resources are resources used by the building, output resources are resources consumed by the building, circulating resources are resources collected and utilized in a circular manner, and non-circulating resources are resources that are not collected naturally and are not recycled;

[0052] In practical application, taking the water cycle in buildings as an example, the collected rainwater, drainage water, etc. are classified as input resources, the irrigation water, domestic water, etc. are consumed as output resources, and the recycled water is regarded as a circulating resource;

[0053] Step S200: Collect historical data collected by each input resource acquisition node and related weather parameter change data, use linear fitting to obtain the resource acquisition curve of each input resource acquisition node along with weather change, and then predict the input resource acquisition amount in each time period according to the weather forecast data of the geographical area;

[0054] Step S300: Collect consumption data of each output resource consumption node, perform periodic time series prediction, and predict resource consumption requirements of each output resource consumption node of the building in different time periods;

[0055] Step S400: Analyze the unit acquisition loss of each input resource, the unit consumption contribution of the output resource and the circulation contribution of the circulation resource, and calculate the total allocation contribution of each unit of resource consumption during resource allocation according to the unit acquisition loss, unit consumption contribution and circulation contribution of each type of resource;

[0056] Step S500: Based on the predicted input resource acquisition and output resource consumption, the total contribution of resource allocation in each time period is comprehensively calculated and the maximum contribution allocation strategy is derived;

[0057] Step S600: Set a floating threshold coefficient for each input resource acquisition node and output resource consumption node. When the actual resource acquisition and consumption is greater than the product of the predicted resource change and the floating threshold coefficient, an alarm message is issued to remind the management personnel to check the nodes to avoid continuous waste.

[0058] In step S200, weather parameters are selected according to resource types, corresponding values ​​of the weather parameters and resource acquisition amounts of input resource acquisition nodes in each time period of historical time are collected, and a corresponding relationship curve is drawn using linear fitting;

[0059] Obtain forecast data of weather parameters for each time period in the future, and predict the amount of resources acquired by the input resource acquisition node in each time period based on the corresponding relationship curve;

[0060] Through linear fitting of weather parameters and the amount of sustainable resources collected in nature, the changing trend of building resource acquisition with weather is analyzed, and then the resource collection amount of each input resource acquisition node of the system is accurately predicted according to future weather changes.

[0061] In actual application, relevant weather parameters can be collected according to the allocated resources. For example, water resources can collect precipitation data around the building, and light resources can collect light intensity around the building. The collected parameter data are linearly fitted with the resource amount collected by each input resource acquisition node, and the relationship between parameter changes and resource collection amount is analyzed. Then, the resource amount collected by each input resource acquisition node is predicted based on the weather parameters predicted in the future time period.

[0062] In step S300, the consumption data of each output resource consumption node in the building in the historical period is collected, and the data is normalized according to the formula:

[0063]

[0064] Among them, ω n Normalized data for the consumption of the nth output resource consumption node, is the amount of resources consumed in the i-th time period, S min is the minimum resource consumption of the nth output resource consumption node in the historical data, S max The maximum value of resource consumption of the nth output resource consumption node in the historical data;

[0065] The consumption data after processing is used to train the LSTM time series prediction model, and then the output resource consumption required by the nth output resource consumption node in the future time period is predicted;

[0066] Through time series prediction, we analyze the consumption data of each output resource consumption node in the building over time, analyze the time period relationship between the data and the building function, and accurately predict the resource consumption of the building in different time periods.

[0067] In step S400, the unit acquisition loss of input resources refers to the loss of the input resource acquisition node acquiring one unit of input resources, the unit allocation contribution of output resources refers to the resource saving benefit when the output resource consumption node consumes one unit of allocated output resources, and the cycle allocation contribution of cycle resources refers to the resource cycle benefit obtained by the output resource node recycling when consuming one unit of resources;

[0068] Calculate the resource allocation contribution between nodes according to the formula:

[0069] W m,n =P m,n +C m,n -E m ;

[0070] Among them, W m,n It refers to the allocation contribution obtained from obtaining one unit of resource usage from the mth input resource acquisition node to the nth output resource consumption node, P m,n The allocation contribution brought by the consumption of one unit of resources by the mth input resource acquisition node for the nth output resource consumption node, C m,n The cyclic allocation contribution obtained by consuming one unit of the resources obtained by the mth input resource acquisition node after the nth output resource consumption node recycles, E m Get the unit loss of the node for the mth input resource;

[0071] By refining the loss and allocation contribution data of resource acquisition, allocation, consumption, and circulation of each unit in the building, the total allocation contribution of resource allocation in the building is calculated; among them, the allocation contribution of consumption and circulation when allocating resources between nodes is introduced, and the allocation contribution of each type of resource between each node is adjusted to adjust the emphasis of the allocation strategy on different allocation paths and resource recycling;

[0072] During specific implementation, the acquisition loss, distribution contribution and circulating distribution contribution of various types of resources can be adjusted, and the distribution emphasis of various types of resources can be increased during distribution. For example, if the circulating distribution contribution of rainwater allocated to toilets is increased, the emphasis of rainwater resources allocated for toilet use can be increased. Reducing the distribution contribution of tap water allocated to irrigation water can reduce the emphasis of tap water used for irrigation during distribution.

[0073] When the quality of the resources obtained by the input resource acquisition node a cannot meet the needs of the output resource consumption node b, the calculation will be skipped. a,b Set to negative infinity; when the output resource consumption node a consumes the output resource consumption node b resources and cannot be recycled, set C a,b is 0;

[0074] Divide the predicted input resource acquisition and output resource consumption in the future time period by unit quantity, and calculate the allocation contribution of each unit resource; treat the acquisition and stock resources and the consumed resource quantities equally. When the input resource acquisition is greater than the output resource consumption, store the excess resources with large unit acquisition loss in the input resources as the stock resources allocated in the next time period, and do not allocate them in the current time period; when the input resource acquisition is less than the output resource consumption, add the stock resources to the acquisition resources for calculation. If it still cannot meet the resource consumption demand, use non-circulating resources to supplement them;

[0075] By dividing the resources obtained by different input resource acquisition nodes according to demand, the supply relationship between the acquired resources and the consumption nodes is refined to avoid general management that may result in substandard resource quality and the inability of the resources allocated to the output resource consumption nodes to meet demand during allocation. The acquired resources and consumed resources are allocated in quantity based on the resource supply and demand relationship. When too many resources are acquired, the surplus is stored to maximize the utilization of resources. When the acquired resources and stock resources cannot meet demand, external resources are introduced to balance supply and demand.

[0076] In step S500, the predicted input resources, replenished stock resources and non-circulating resources are taken as upstream resources, the consumption predicted by the output resource consumption node is taken as downstream resources, the upstream resources and downstream resources are divided by units, and a resource consumption contribution matrix is ​​calculated and generated, wherein the matrix rows represent the upstream resources of each unit, the columns represent the downstream resources of each unit, and the elements in the xth row and yth column of the matrix represent the total contribution of the upstream resources of the xth unit to the yth downstream resource;

[0077] Combine upstream resources with downstream resources to form a bipartite graph. Use the elements in the resource consumption contribution matrix as the weights of the edges when matching the bipartite graph. Use the KM algorithm to analyze the best weighted perfect matching strategy of the bipartite graph and find the allocation strategy with the maximum total contribution of all upstream resources during allocation.

[0078] like Figure 3FIG. 1 is a weighted bipartite graph of the KM algorithm in a green building information data analysis method based on artificial intelligence; X1, X2, X3, …, Xn on the left side of the figure are the unit resources after the upstream resources are divided into units, and Y1, Y2, Y3, …, Yn on the right side are the unit resources after the downstream resources are divided into units; the connection line between the upstream resources and the downstream resources indicates whether different upstream resources and downstream resources can be allocated, wherein the solid line indicates that the upstream resources on the left side of the connection line can be allocated to the downstream resources on the right side, and the dotted line indicates that the upstream resources on the left side of the connection line cannot be allocated to the downstream resources on the right side. For example, rainwater in reality can be used for irrigation and toilets, but rainwater cannot be directly allocated to drinking water nodes; each connection line has a separate number, which indicates the allocation contribution of the upstream resource on the left side of the connection line to allocate a unit amount to the downstream resource on the right side. When the connection line is a dotted line, the number is -∞.

[0079] A green building information data analysis system based on artificial intelligence, the system comprising an information collection module, a resource prediction module, a resource allocation module, and an over-capacity warning module;

[0080] The information collection module is used to collect weather parameter change data, input resource acquisition node resource acquisition data, and output resource consumption node resource consumption data; the resource prediction module is used to predict the resource acquisition amount of each input resource acquisition node in the future time period, and is also used to predict the resource amount required to be consumed by each output resource consumption node in the future time period; the resource allocation module is used to calculate the resource acquisition loss and allocation contribution between each node, and analyze and calculate the maximum contribution allocation strategy; the over-consumption warning module issues an alarm message when the difference between the actual resource acquisition consumption and the predicted amount exceeds a threshold;

[0081] An over-alarm module is introduced into the system, which issues an alarm message when the resource collection or consumption of a node exceeds the expected threshold, promptly reminds you to check the relevant nodes, implements the system self-check function, and avoids continuous node failures or resource waste during use that affects the sustainable efficiency of resources.

[0082] The output end of the information collection module is connected to the input end of the resource prediction module and the excess warning module; the output end of the resource prediction module is connected to the input end of the resource allocation module and the excess warning module.

[0083] The information collection module includes a weather parameter collection unit, an input resource collection unit, and an output resource collection unit;

[0084] The weather parameter collection unit is used to collect weather parameter change data in each time period around the building and predicted weather parameter values ​​in future time periods; the input resource collection unit is used to collect input resources used in the building resource circulation system and collect historical data on the collection amount of each input resource acquisition node; the output resource collection unit is used to collect historical data on the resource consumption amount of each output resource consumption node.

[0085] The resource prediction module includes an input resource prediction unit and an output resource prediction unit;

[0086] The input resource prediction unit is used to perform linear fitting based on the weather parameters obtained by the information collection module and the historical data of the collection amount of each input resource acquisition node, analyze the relationship between the collection amount of the input resource acquisition node and the weather parameters, and then predict the collection amount of each input resource acquisition node in the future time period based on the predicted weather parameters; the output resource prediction unit uses the LSTM model to perform time series analysis and prediction based on the historical data of the consumption of each output resource consumption node obtained by the information collection module, and then predicts the resource amount required for each output resource consumption node in the future time period.

[0087] The resource allocation module includes a contribution analysis unit, an allocation strategy analysis unit, a resource storage unit, and a non-circular resource introduction unit;

[0088] The contribution analysis unit is used to calculate the total contribution of each unit resource allocation; the allocation strategy analysis unit analyzes the maximum contribution allocation strategy when all upstream resources are allocated according to the total contribution of each unit resource allocation; the resource storage unit is used to store excess input resources for resource replenishment in the next time period; the non-circulating resource introduction unit is used to introduce non-circulating resources as a supplement when the input resources and stock resources cannot meet the output resource demand;

[0089] The contribution analysis unit input end is connected to the resource prediction module output end; the contribution analysis unit output end is connected to the allocation strategy analysis unit input end; the allocation strategy analysis unit output end is connected to the resource storage unit and the non-circular resource introduction unit input end.

[0090] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0091] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein by equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A green building information data analysis method based on artificial intelligence, characterized in that: The analysis steps include: Step S100: Divide the resources in the building into input resources, output resources, circulating resources, and non-circulating resources; wherein input resources are resources used by the building, output resources are resources consumed by the building, circulating resources are resources collected and utilized in a circular manner, and non-circulating resources are resources that are not collected naturally and are not recycled; Step S200: Collect historical data collected by each input resource acquisition node and related weather parameter change data, use linear fitting to obtain the resource acquisition curve of each input resource acquisition node along with weather change, and then predict the input resource acquisition amount in each time period according to the weather forecast data of the geographical area; Step S300: Collect consumption data of each output resource consumption node, perform periodic time series prediction, and predict resource consumption requirements of each output resource consumption node of the building in different time periods; Step S400: Analyze the unit acquisition loss of each input resource, the unit allocation contribution of the output resource and the cyclic allocation contribution of the cyclic resource, and calculate the total allocation contribution of each unit of resource consumption during resource allocation according to the unit acquisition loss, unit allocation contribution and cyclic allocation contribution of each type of resource; Step S500: Based on the predicted input resource acquisition and output resource consumption, comprehensively calculate the resource allocation contribution of each time period and derive the maximum contribution allocation strategy; Step S600: Set a floating threshold coefficient for each input resource acquisition node and output resource consumption node. When the actual resource acquisition and consumption is greater than the product of the predicted resource change and the floating threshold coefficient, an alarm message is issued to remind the management personnel to check the nodes to avoid continuous waste.

2. The green building information data analysis method based on artificial intelligence according to claim 1 is characterized by: In step S200, weather parameters are selected according to resource types, corresponding values ​​of the weather parameters and resource acquisition amounts of input resource acquisition nodes in each time period of historical time are collected, and a corresponding relationship curve is drawn using linear fitting; Obtain forecast data of weather parameters for each time period in the future, and predict the amount of resources acquired by the input resource acquisition node in each time period based on the corresponding relationship curve.

3. The green building information data analysis method based on artificial intelligence according to claim 1 is characterized by: In step S300, the consumption data of each output resource consumption node in the building in the historical period is collected, and the data is normalized according to the formula: Among them, ω n Normalized data for the consumption of the nth output resource consumption node, is the amount of resources consumed in the i-th time period, S min is the minimum resource consumption of the nth output resource consumption node in the historical data, S max The maximum value of resource consumption of the nth output resource consumption node in the historical data; The consumption data after processing is used to train the LSTM time series prediction model, and then the output resource consumption required by the n-th output resource consumption node in the future time period is predicted.

4. The green building information data analysis method based on artificial intelligence according to claim 1 is characterized by: In step S400, the unit acquisition loss of input resources refers to the loss of the input resource acquisition node acquiring one unit of input resources, the unit allocation contribution of output resources refers to the resource saving benefit when the output resource consumption node consumes one unit of allocated output resources, and the cycle allocation contribution of cycle resources refers to the resource cycle benefit obtained by the output resource node recycling when consuming one unit of resources; Calculate the resource allocation contribution between nodes according to the formula: W m,n =P m,n +C m,n -E m ; Among them, W m,n It refers to the allocation contribution obtained from obtaining one unit of resource usage from the mth input resource acquisition node to the nth output resource consumption node, P m,n The allocation contribution brought by the consumption of one unit of resources by the mth input resource acquisition node for the nth output resource consumption node, C m,n The cyclic allocation contribution obtained by consuming one unit of the resources obtained by the mth input resource acquisition node after the nth output resource consumption node recycles, E m Get the unit loss of the node for the mth input resource; When the quality of the resources obtained by the input resource acquisition node a cannot meet the needs of the output resource consumption node b, the calculation will be skipped. a,b Set to negative infinity; when the output resource consumption node a consumes the output resource consumption node b resources and cannot be recycled, set C a,b is 0; The predicted input resource acquisition and output resource consumption in the future time period are divided into units, and the allocation contribution of each unit resource is calculated; the acquisition and stock resources are treated equally with the consumed resource amounts. When the input resource acquisition is greater than the output resource consumption, the excess resources with large unit acquisition loss in the input resources are stored as stock resources for allocation in the next time period, and no allocation is made in the current time period; when the input resource acquisition is less than the output resource consumption, the stock resources are added to the acquisition resources for calculation. If the resource consumption demand cannot be met, non-circulating resources are used for supplementation.

5. The green building information data analysis method based on artificial intelligence according to claim 1 is characterized by: In step S500, the predicted input resources, replenished stock resources and non-circulating resources are taken as upstream resources, the consumption predicted by the output resource consumption node is taken as downstream resources, the upstream resources and downstream resources are divided by units, and a resource consumption contribution matrix is ​​calculated and generated, wherein the matrix rows represent the upstream resources of each unit, the columns represent the downstream resources of each unit, and the elements in the xth row and yth column of the matrix represent the total contribution of the upstream resources of the xth unit to the yth downstream resource; The upstream resources and downstream resources are combined into a bipartite graph. The elements in the resource consumption contribution matrix are used as the weights of the edges when matching the bipartite graph. The KM algorithm is used to analyze the weighted best perfect matching strategy of the bipartite graph to find the allocation strategy with the maximum total contribution of all upstream resources during allocation.

6. A green building information data analysis system based on artificial intelligence using the green building information data analysis method based on artificial intelligence according to any one of claims 1 to 5, characterized in that: The system includes an information collection module, a resource prediction module, a resource allocation module, and an over-capacity warning module; The information collection module is used to collect weather parameter change data, input resource acquisition node resource acquisition data, and output resource consumption node resource consumption data; the resource prediction module is used to predict the resource acquisition amount of each input resource acquisition node in the future time period, and is also used to predict the resource amount required to be consumed by each output resource consumption node in the future time period; the resource allocation module is used to calculate the resource acquisition loss and allocation contribution between each node, and analyze and calculate the maximum contribution allocation strategy; the over-consumption warning module issues an alarm message when the difference between the actual resource acquisition consumption and the predicted amount exceeds a threshold; The output end of the information collection module is connected to the input end of the resource prediction module and the excess warning module; the output end of the resource prediction module is connected to the input end of the resource allocation module and the excess warning module.

7. The green building information data analysis system based on artificial intelligence according to claim 6 is characterized by: The information collection module includes a weather parameter collection unit, an input resource collection unit, and an output resource collection unit; The weather parameter collection unit is used to collect weather parameter change data of each time period around the building and predicted weather parameter values ​​of future time periods; The input resource collection unit is used to collect the input resources used in the building resource circulation system and collect the historical data of the collection amount of each input resource acquisition node; the output resource collection unit is used to collect the historical data of the resource consumption amount of each output resource consumption node.

8. The green building information data analysis system based on artificial intelligence according to claim 6 is characterized by: The resource prediction module includes an input resource prediction unit and an output resource prediction unit; The input resource prediction unit is used to perform linear fitting based on the weather parameters obtained by the information collection module and the historical data of the collection amount of each input resource acquisition node, analyze the relationship between the collection amount of the input resource acquisition node and the weather parameters, and then predict the collection amount of each input resource acquisition node in the future time period based on the predicted weather parameters; the output resource prediction unit uses the LSTM model to perform time series analysis and prediction based on the historical data of the consumption of each output resource consumption node obtained by the information collection module, and then predicts the resource amount required for each output resource consumption node in the future time period.

9. The green building information data analysis system based on artificial intelligence according to claim 6 is characterized by: The resource allocation module includes a contribution analysis unit, an allocation strategy analysis unit, a resource storage unit, and a non-circular resource introduction unit; The contribution analysis unit is used to calculate the total contribution of each unit resource allocation; the allocation strategy analysis unit analyzes the allocation strategy with the maximum contribution when all upstream resources are allocated according to the total contribution of each unit resource allocation; The resource storage unit is used to store excess input resources for resource supplementation in the next time period; the non-circulating resource introduction unit is used to introduce non-circulating resources as supplements when the input resources and stock resources cannot meet the output resource demand; The contribution analysis unit input end is connected to the resource prediction module output end; the contribution analysis unit output end is connected to the allocation strategy analysis unit input end; the allocation strategy analysis unit output end is connected to the resource storage unit and the non-circular resource introduction unit input end.

Images