Building power resource intelligent allocation method based on BIM (Building Information Modeling)

Abstract

The invention relates to the technical field of data processing, in particular to a BIM-based building power resource intelligent allocation method. The method comprises the following steps: acquiring building modeling data and real-time power consumption data of each building area, and integrating the building modeling data and the real-time power consumption data to obtain the integrated building modeling data and real-time power consumption data; the historical building modeling data and the historical power consumption data are integrated, and a power consumption demand prediction model is constructed; processing the integrated building modeling data and the real-time power consumption data through a power consumption demand prediction model to obtain a power consumption demand prediction result of each building area; generating an optimal distribution strategy through a multi-objective optimal distribution algorithm based on the power consumption demand prediction result of each building area; and dynamically adjusting the power supply quantity of each building area based on an optimal distribution strategy, and carrying out intelligent distribution of power resources. The technical problems of incorrect data integration, inaccurate electricity demand prediction and low distribution efficiency in intelligent distribution of building electric power resources are solved.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a method for intelligent allocation of building power resources based on BIM. Background Art

[0002] Building power resource management is an important part of modern building energy consumption optimization. Its main goal is to reduce energy consumption, improve energy efficiency and ensure power supply safety by accurately predicting and reasonably allocating power resources inside the building. However, the current technical solutions have significant deficiencies in data integration, power demand forecasting and power distribution optimization. With the rise of smart buildings and green building concepts, the demand for intelligent and refined building power resource management is increasing, which requires not only the effective integration of multi-source data, but also the realization of high-precision power demand forecasting and dynamic and efficient power distribution.

[0003] However, the above technologies have the following technical problems: incorrect data integration, inaccurate electricity demand prediction and inefficient allocation in the intelligent allocation of building power resources. Summary of the invention

[0004] The present invention provides a method for intelligent allocation of building power resources based on BIM, so as to solve the technical problems of incorrect data integration, inaccurate prediction of power demand and inefficient allocation in intelligent allocation of building power resources.

[0005] The present invention provides a BIM-based intelligent allocation method for building power resources, which specifically includes the following technical solutions:

[0006] A method for intelligent allocation of building power resources based on BIM, comprising the following steps:

[0007] S1. Obtain building modeling data and real-time power consumption data of each building area; integrate the building modeling data and real-time power consumption data of each building area to obtain integrated building modeling data and real-time power consumption data;

[0008] S2. Obtain and integrate historical building modeling data and historical electricity consumption data to obtain integrated historical building modeling data and historical electricity consumption data; construct an electricity demand forecasting model based on the integrated historical building modeling data and historical electricity consumption data; process the integrated building modeling data and real-time electricity consumption data through the electricity demand forecasting model to obtain electricity demand forecasting results for each building area;

[0009] S3. Based on the electricity demand forecast results of each building area, an optimized allocation strategy is generated through a multi-objective optimization allocation algorithm; based on the optimized allocation strategy, the power supply of each building area is dynamically adjusted to perform intelligent allocation of power resources.

[0010] Preferably, the S1 specifically includes:

[0011] The specific implementation steps of the integration process are:

[0012] The first step is to preprocess the building modeling data and the real-time power consumption data of each building area to obtain the preprocessed building modeling data and the real-time power consumption data;

[0013] In the second step, the preprocessed building modeling data and real-time electricity consumption data are matched using an association algorithm based on spatial regions and equipment numbers to obtain a matching matrix;

[0014] The third step is to perform time series alignment on the matching matrix to obtain time-aligned integrated data; the time-aligned integrated data is structured to obtain integrated building modeling data and real-time electricity consumption data.

[0015] Preferably, the S1 specifically includes:

[0016] In the process of implementing the association algorithm based on spatial area and equipment number, a multidimensional matching algorithm based on spatial weight allocation is introduced to calculate the matching degree between the preprocessed real-time power consumption data and the preprocessed building modeling data to obtain the matching weight; based on the matching weight, a matching weight matrix is ​​constructed; the specific calculation formula of the matching weight is:

[0017]

[0018] Among them, w ij is the matching weight, which indicates the matching degree between the i-th record of the preprocessed real-time power consumption data and the j-th object of the preprocessed building modeling data; λ1 is the distance decay parameter; α, β, γ are the feature weight coefficients, which are used to control the influence of spatial distance, power similarity and regional area similarity on the matching weight respectively; (x j ,y j ) represents the spatial coordinates of the building area of ​​the jth object of the preprocessed building modeling data; represents the real-time spatial position of the ith record of the preprocessed real-time power consumption data; P j represents the equipment power of the jth object of the preprocessed building modeling data; A represents the real-time power of the ith record of the preprocessed real-time power consumption data; j represents the area of ​​the jth object of the preprocessed building modeling data; represents the area of ​​the i-th record of the preprocessed real-time electricity consumption data; It is an impact factor based on spatial distance; is the power similarity factor; is the area similarity factor.

[0019] Preferably, the S1 specifically includes:

[0020] A binary variable reflecting the object matching between the preprocessed real-time electricity consumption data and the preprocessed building modeling data is introduced, and a global optimization objective function is constructed in combination with constraint conditions to maximize the sum of all matching weights. After the optimization solution, a matching matrix is ​​constructed based on the binary variable. The matching matrix is ​​post-processed, and when the matching weight is less than or equal to the preset matching weight threshold, the specific calculation formula of the matching weight is readjusted.

[0021] Preferably, the S2 specifically includes:

[0022] Based on the integrated historical building modeling data and historical electricity consumption data, a multi-layer nested time series characteristic association model is constructed as the electricity demand prediction model through multi-level characteristic association and nonlinear nested structure.

[0023] Preferably, the S2 specifically includes:

[0024] In the process of constructing the electricity demand prediction model, feature extraction is performed on the integrated historical building modeling data and historical electricity consumption data to obtain feature data; the feature data includes a building feature matrix and a time series feature matrix; the time series feature matrix is ​​feature decomposed to obtain different frequency characteristics; the building feature matrix is ​​subjected to principal component analysis, and the main components of building characteristics are extracted by dimensionality reduction; in the mapping unit in the hidden layer, the frequency characteristics are cross-correlated with the main components of building characteristics through mapping processing to obtain a nonlinear mapping matrix.

[0025] Preferably, the S2 specifically includes:

[0026] In the dynamic optimization unit in the hidden layer, based on the nonlinear mapping matrix, the time delay interaction processing is introduced, the time delay interaction results are calculated, and the time-space interaction characteristic matrix is ​​obtained.

[0027] Preferably, the S2 specifically includes:

[0028] A multi-level nested structure is introduced to nest and combine the time-space interaction characteristic matrices of different frequency levels to obtain the output of the hidden layer; the output of the hidden layer is input into the prediction layer, and the hyperbolic tangent function is used to predict the electricity demand in the future time period.

[0029] Preferably, the S3 specifically includes:

[0030] Based on the electricity demand forecast results of each building area and combined with the preprocessed building modeling data, a multi-objective optimization function is constructed; the specific definition of the multi-objective optimization function is:

[0031]

[0032] in, is a multi-objective optimization function; It is The power resource allocation power of each building area at time t; It is The predicted value of electricity demand for each building area at time t; is the number of zones defined in the building; is the weight coefficient of the supply-demand balance target; It is Building area to electrical equipment Transmission loss at time t; It is Electrical equipment in building areas The power resource allocation power at time t; It is Building area to electrical equipment The physical distance between is the distance attenuation coefficient, which is used to control the impact of long-distance transmission energy consumption in optimization; ∈ is the decimal bias to prevent the denominator from being zero; is the number of all electrical equipment in the building; is the weight coefficient of the loss minimization objective; It is Building area and The coupling phase angle of each building area at time t; It is Building area and The fixed phase offset of each building area at time t; is the logarithm of the regions participating in the coupling in the building; η is the weight coefficient of the regional coupling target; T is the total duration of the optimization time window;

[0033] The multi-objective optimization function is solved to obtain the optimal allocated power; based on the optimal allocated power, the optimal allocation strategy is generated.

[0034] Preferably, the S3 specifically includes:

[0035] Collect actual electricity consumption data, compare and analyze the actual electricity consumption data with the electricity demand forecast results, and obtain comparative analysis results; based on the comparative analysis results, optimize the electricity demand forecast model and improve the power distribution strategy.

[0036] The beneficial effects of the technical solution of the present invention are:

[0037] 1. The multi-dimensional matching algorithm based on spatial weight allocation is used to significantly improve the accuracy of data matching and solve the problem of inaccurate matching caused by repeated equipment numbers and fuzzy spatial areas.

[0038] 2. Use a multi-layer nested time series characteristic correlation model to explore the deep correlation between the power consumption characteristics of the building area and the external time series characteristics. Through the nested correlation processing of frequency decomposition and the main components of building characteristics, the multi-dimensional modeling of power demand is realized, and the multi-frequency characteristics are fully captured, making the power demand forecast results more robust and adaptable. By using time delay interaction processing and multi-level nested structure, the ability to capture short-term fluctuations, cyclical changes and long-term trends is improved, and the power demand forecast results are refined in the time series, providing accurate data support for dynamic power supply decisions.

[0039] 3. By constructing a multi-objective optimization function that includes supply and demand balance, loss minimization and regional coupling, and using the particle swarm optimization algorithm to generate the optimal allocation strategy for power resources, while ensuring supply and demand balance, the transmission loss is reduced and the power utilization efficiency is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 This is a flow chart of a method for intelligent allocation of building power resources based on BIM described in the present invention. DETAILED DESCRIPTION

[0041] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the technical scheme in the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiment of the present invention. Obviously, the described embodiment is only a part of the embodiment of the present invention, not all 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.

[0042] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0043] The following is a detailed description of a BIM-based intelligent allocation method for building power resources provided by the present invention in conjunction with the accompanying drawings.

[0044] See attached Figure 1 , which shows a flow chart of a method for intelligent allocation of building power resources based on BIM provided by an embodiment of the present invention, the method comprising the following steps:

[0045] S1. Obtain building modeling data and real-time power consumption data of each building area; integrate the building modeling data and real-time power consumption data of each building area to obtain integrated building modeling data and real-time power consumption data;

[0046] Use BIM technology to obtain building modeling data, including building structure information (such as space area number, floor distribution, area), equipment layout (equipment number, equipment type, location, power parameters), initial power design load, power resource allocation power, etc. At the same time, use IoT sensors to collect real-time power consumption data in various areas of the building, including power consumption parameters (such as voltage, current, power factor, active power, reactive power), timestamp information (collection time of each data point), etc.

[0047] Furthermore, the building modeling data and the real-time power consumption data are integrated to obtain the integrated building modeling data and the real-time power consumption data. The specific implementation steps are as follows:

[0048] The first step is to normalize the building modeling data and real-time power consumption data (i.e., preprocessing) to obtain the preprocessed building modeling data and real-time power consumption data; for example, for building modeling data, perform data integrity check, and use the mean filling method to fill in simple attributes (such as area), and use the nearest neighbor interpolation method to fill in spatial distribution related attributes (such as power parameters of equipment layout); for real-time power consumption data, remove abnormal data (such as noise points with too large or too small power values), and use linear interpolation or time series prediction to fill in if there is a collection interruption in the IoT sensor;

[0049] In the second step, the preprocessed building modeling data and real-time power consumption data are matched using an association algorithm based on spatial regions and equipment numbers to obtain a matching matrix; the preprocessed real-time power consumption data is bound to the corresponding spatial region according to the spatial region number in the preprocessed building modeling data, and the preprocessed real-time power consumption data is mapped to a specific device according to the equipment number; the SQL JOIN operation or the merge function based on Python Pandas is used to associate the preprocessed building modeling data and real-time power consumption data according to the spatial region number and equipment number to achieve data matching processing.

[0050] In order to improve the matching accuracy of pre-processed building modeling data and real-time power consumption data and avoid ambiguity in the correspondence between spatial areas and equipment numbers, such as multiple devices in the same spatial area or repeated equipment numbers leading to inaccurate matching, a multi-dimensional matching algorithm based on spatial weight distribution is introduced. By utilizing the spatial weights of area and equipment distribution and combining the multi-dimensional attributes of building modeling data, a matching matrix is ​​constructed to improve matching accuracy.

[0051] The specific process of the multidimensional matching algorithm based on spatial weight allocation is as follows:

[0052] The preprocessed building modeling data is denoted as B = {b1, b2, ..., b N}, use b j represents any object of the preprocessed building modeling data, j ranges from 1 to N; N represents the total number of preprocessed building modeling data; each object b j Contains the spatial area number id j , Equipment No.d j 、Equipment power P j 、Area A j , the spatial coordinates of the building area (x j ,y j ), power resource allocation power etc.; the real-time power consumption data after preprocessing is recorded as R = {r1, r2, ..., r M}, use r i represents any record in the preprocessed real-time power consumption data, i ranges from 1 to M; M represents the total number of records in the preprocessed real-time power consumption data; each record r i The sampling time t i , real-time power Real-time spatial location Area Furthermore, the pre-processed building modeling data and real-time electricity consumption data are uniquely associated through a matching matrix, thereby achieving accurate data integration;

[0053] In order to accurately quantify the matching degree between each record in the preprocessed real-time power consumption data and each object in the preprocessed building modeling data, a matching weight is constructed; based on the matching weight, a matching weight matrix is ​​constructed; the matching weight is composed of spatial distance, power similarity and area similarity, and the specific calculation formula is:

[0054]

[0055] Among them, w ij is the matching weight, which represents the i-th record r of the preprocessed real-time electricity consumption data i and the jth object b of the preprocessed building modeling data j The matching degree between them reflects the similarity between them in terms of spatial position, device power and regional area; λ1 is the distance attenuation parameter, which is used to control the attenuation speed of spatial distance on matching weight. The longer the spatial distance, the smaller the contribution to matching weight; α, β, γ are feature weight coefficients, which are used to control the influence of spatial distance, power similarity and regional area similarity on matching weight respectively. The feature weight coefficients satisfy the normalization condition α+β+γ=1; It is an influencing factor based on spatial distance, which aims to reduce the contribution of distant data to the matching weight. represents the Euclidean space distance; (x j ,y j ) represents the jth object b of the preprocessed building modeling data j The spatial coordinates of the building area; Represents the i-th record r of the preprocessed real-time electricity consumption data i The real-time spatial location. is the power similarity factor, maximizing the contribution of similarity power to the matching weight; It is the regional area similarity factor, which reflects the degree of spatial overlap between two regions. The more similar the regional areas are, the higher the matching weight is.

[0056] Furthermore, in order to ensure the optimization of the overall matching weight matrix, a global optimization objective function Z is constructed, the purpose of which is to maximize the sum of all matching weights. The specific formula is:

[0057]

[0058] Among them, c ij is a binary variable, representing the i-th record r of the preprocessed real-time electricity consumption data record i Is it consistent with the jth object b of the preprocessed building modeling data? j Match, if it matches, then c ij =1, otherwise c ij = 0. This formula controls the uniqueness of the match through a uniqueness constraint, that is, each record of preprocessed real-time electricity consumption data can only be matched with one object of preprocessed building modeling data:

[0059]

[0060] Similarly, each object of preprocessed building modeling data can only be matched to one record of preprocessed real-time electricity consumption data:

[0061]

[0062] Through the constraints, the binary variable c ij ∈{0, 1}, which ensures the discreteness of the matching results and avoids the ambiguity of partial matching;

[0063] In the optimization solution process, a linear programming method (such as the Hungarian algorithm) or an integer linear programming (ILP) method is used to solve the global optimization objective function Z;

[0064] Based on the binary variables, construct the matching matrix C = [c ij], and post-process the matching results; if it is found that some records have a matching weight less than or equal to the matching weight threshold preset according to the expert experience method due to missing or abnormal data, the specific calculation formula of the matching weight can be readjusted. For example, by adjusting the distance attenuation parameter λ1 or recalibrating the equipment power P of the pre-processed building modeling data j and real-time power of preprocessed real-time power consumption data Iteratively optimize items with low matching weights;

[0065] The third step is to perform time series alignment processing based on the matching matrix using the existing timestamp mapping and interpolation technology to ensure that the preprocessed real-time electricity consumption data and the building modeling data are aligned in the time dimension to obtain the time-aligned integrated data; the time-aligned integrated data is structured, that is, the time-aligned integrated data is organized into a tabular form, where each row corresponds to the regional data of a time point (fields include spatial area number, equipment number, equipment power parameters); real-time electricity consumption information (power, voltage, current); timestamp, and finally the integrated building modeling data and real-time electricity consumption data are obtained.

[0066] S2. Obtain and integrate historical building modeling data and historical electricity consumption data to obtain integrated historical building modeling data and historical electricity consumption data; construct an electricity demand forecasting model based on the integrated historical building modeling data and historical electricity consumption data; process the integrated building modeling data and real-time electricity consumption data through the electricity demand forecasting model to obtain electricity demand forecasting results for each building area;

[0067] The historical building modeling data and historical electricity consumption data are obtained from the existing database, and the integration method of step S1 is used to integrate and process the historical building modeling data and historical electricity consumption data after integration; the integrated historical building modeling data and historical electricity consumption data are used as training data for building an electricity demand forecasting model, and the model is built and trained to finally obtain an electricity demand forecasting model. The specific process is as follows:

[0068] A multi-layer nested time series characteristic association model is constructed as a power demand forecasting model. The multi-layer nested time series characteristic association model mines the deep relationship between the power consumption characteristics of the building area and the external time series influence through multi-level characteristic association and nonlinear nested structure. The construction process is as follows:

[0069] First, in the feature extraction layer, the integrated historical building modeling data and historical electricity consumption data are extracted using existing feature engineering technology to obtain feature data; the feature data includes a building feature matrix and a time series feature matrix; the building feature matrix includes spatial distribution features (such as area, floor height, window ratio, etc.) and equipment load features (such as power, electricity consumption duration), expressed as a two-dimensional matrix Where m is the number of regions in the historical building modeling data; n is the number of building characteristics, which represents the total number of spatial distribution characteristics and equipment load characteristics, including all feature quantities; the time series feature matrix, including historical power consumption data (such as power, current) and environmental time series data (such as temperature, humidity), is represented as a three-dimensional tensor Where q is the length of the time series; r is the number of regions detected in the historical electricity consumption data; s is the number of time series characteristics (such as voltage, current, temperature and humidity, etc.);

[0070] In order to make the feature data available for subsequent calculations, the time series feature matrix is ​​decomposed in the feature optimization layer; the time series feature matrix T is decomposed into different frequency characteristics F using Fourier transform. k , represents the Fourier component of the time series feature matrix T at the frequency component k; further, F k It is divided into low-frequency (long-term trend), medium-frequency (periodic change) and high-frequency (short-term fluctuation) features to capture the multi-level characteristics of electricity consumption in the building area. Conduct principal component analysis (PCA) to reduce the dimension and extract the principal components of building characteristics

[0071] Furthermore, in the hidden layer, the frequency characteristics of the time series feature matrix are deeply associated with the main components of the architectural characteristics of the architectural feature matrix. First, in the mapping unit under the hidden layer, the frequency characteristics F k Principal components of building characteristics Perform cross-correlation and the mapping processing formula is as follows:

[0072]

[0073] in, It is the nonlinear feature matrix of the output of the mapping unit under the hidden layer, which represents the frequency characteristic F k Principal components of building characteristics The deep connection between them after being processed by the nonlinear activation function (ReLU); is the linear mapping weight matrix of the frequency characteristic, which is used to transform the frequency characteristic F k Mapped into the feature space corresponding to the principal components of building characteristics, dynamically adjusted through model training; is the linear mapping weight matrix of the principal components of building characteristics, which is used to transform the principal components of building characteristics Mapped into the feature space corresponding to the frequency characteristics, and dynamically adjusted through model training; It is a bias term, which is used to adjust the linear mapping results of the frequency characteristics and the main components of the building characteristics, and increase the flexibility of the electricity demand forecasting model;

[0074] Furthermore, in order to capture the dynamic dependencies of time series, time delay interaction processing is introduced in the dynamic optimization unit under the hidden layer. The nonlinear feature matrix based on the output of the hidden layer mapping unit is Calculate the time delay interaction results and obtain the time-space interaction characteristic matrix:

[0075]

[0076] in, is the time t at which the frequency component k and the spatial characteristic The time delay interaction result under is the time series-space interaction characteristic matrix, which represents the cumulative impact of the frequency components and spatial characteristic interactions in the past L time steps on the current moment; t is the time step index, which represents the current time point; is the time delay weight, which indicates the influence of time step l on the current time step t; is the time step tl, in the frequency component k and spatial characteristics The output of the hidden layer mapping unit under; L is the window size of the time delay;

[0077] Furthermore, based on the time-space interaction characteristic matrix, a multi-level nested structure is introduced to analyze the time delay interaction results under different frequency components and spatial characteristics. Perform nested combination to get the output of the hidden layer. The formula is as follows:

[0078]

[0079] Among them, O multi is the output of the hidden layer, which is a multi-frequency, multi-level nested characteristic correlation output matrix, representing the comprehensive characteristics after processing of different frequency components and spatial characteristics, and is used to describe the overall electricity demand characteristics of the building area; J is the total number of frequency levels; It is the frequency level number, indicating the level to which the frequency components after Fourier decomposition are divided (such as low frequency, medium frequency or high frequency). Each frequency level Corresponds to a frequency range, not a single specific frequency value; It is The weight of the frequency level, indicating the The contribution of each frequency level in the comprehensive characteristics. The greater the weight, the greater the impact of this frequency level on the overall power demand characteristics. GatedUnit is a gating unit used to The time-space interaction characteristic matrix of frequency levels Perform nonlinear transformation and feature screening, whose function is to dynamically control the passage or suppression of information through a gating mechanism, thereby capturing the key impact of frequency characteristics on the target; It is The time-space interaction characteristic matrix of the frequency level represents the Time delay accumulation characteristics at each frequency level;

[0080] Finally, the output of the hidden layer O multi Input to the prediction layer, use the hyperbolic tangent function Tanh to predict the electricity demand in the future time period, and get the predicted value of electricity demand. The specific formula is:

[0081]

[0082] Among them, W o and b o are weights and biases; It is the predicted value of electricity demand.

[0083] When training the electricity demand forecasting model, the existing composite loss function is used to optimize the electricity demand forecasting model for model training to obtain a fully trained electricity demand forecasting model.

[0084] Furthermore, the integrated building modeling data and real-time electricity consumption data are processed using the electricity demand prediction model to obtain electricity demand prediction results for each building area. The electricity demand prediction results include electricity demand prediction values ​​for each building area, time series trends of electricity demand, electricity characteristic components, and abnormal electricity consumption behavior warnings.

[0085] S3. Based on the electricity demand forecast results of each building area, an optimized allocation strategy is generated through a multi-objective optimization allocation algorithm; based on the optimized allocation strategy, the power supply of each building area is dynamically adjusted to perform intelligent allocation of power resources.

[0086] According to the electricity demand forecast results of each building area and the pre-processed building modeling data, the multi-objective optimization allocation algorithm is used to generate the optimization allocation strategy. The multi-objective optimization allocation algorithm constructs a multi-objective optimization function and uses the particle swarm optimization algorithm to solve it, and finally obtains the optimization allocation strategy. The specific implementation process is as follows:

[0087] A multi-objective optimization function is constructed based on the electricity demand forecast results of each building area and the pre-processed building modeling data; it aims to achieve the goals of supply and demand balance, loss minimization and regional coupling. The specific definition of the multi-objective optimization function is:

[0088]

[0089] in, is a multi-objective optimization function; It is The power resource allocation power of each building area at time t; It is The predicted value of electricity demand for each building area at time t is taken from the electricity demand forecast result; is the number of zones defined in the building; is the weight coefficient of the supply-demand balance target; It is Building area to electrical equipment Transmission loss at time t; It is Electrical equipment in building areas The power resource allocation power at time t; It is Building area to electrical equipment The physical distance between is the distance attenuation coefficient, which is used to control the impact of long-distance transmission energy consumption in optimization; ∈ is the decimal bias to prevent the denominator from being zero; is the number of all electrical equipment in the building; is the weight coefficient of the loss minimization objective; It is Building area and The coupling phase angle of the building areas at time t reflects the synchronization or phase difference of electricity demand between building areas; It is Building area and The fixed phase offset of each building area at time t; is the logarithm of the regions participating in the coupling in the building; η is the weight coefficient of the regional coupling target; T is the total duration of the optimization time window, which is determined according to the expert experience method; is the supply-demand balance objective, and the difference between demand and allocation is characterized by the squared error form; It is the loss target. Considering the distance attenuation relationship of the transmission power, an exponential attenuation function is used to describe the energy consumption characteristics between the building area and the power equipment. It is a regional coupling target that reflects the dynamic interaction between building areas through a sine square function.

[0090] Furthermore, the existing particle swarm optimization algorithm is used to solve the multi-objective optimization function to obtain the optimal allocated power; based on the optimal allocated power, an optimal allocation strategy is generated, and the optimal allocation strategy includes the allocated power value and time period of each building area.

[0091] Finally, the power supply of each building area is dynamically adjusted according to the optimal allocation strategy to realize the intelligent allocation of power resources.

[0092] Collect actual electricity consumption result data, use existing comparative analysis methods (such as error distribution analysis and time correlation analysis) to compare and analyze the actual electricity consumption result data with the electricity demand forecast results, and obtain comparative analysis results; based on the comparative analysis results, optimize the electricity demand forecast model through the existing reinforcement learning algorithm, further improve the power distribution strategy, and improve the distribution accuracy and adaptability.

[0093] In summary, a BIM-based intelligent allocation method for building power resources was completed.

[0094] The order of the embodiments of the invention is for description only and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0095] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

[0096] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that the technical solutions described in the above embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention, and should be included in the protection scope of the present invention.

Claims

1. A method for intelligent allocation of building power resources based on BIM, characterized in that: The following steps are involved: S1. Obtain building modeling data and real-time electricity consumption data of each building area; Integrate and process the building modeling data and the real-time power consumption data of each building area to obtain integrated building modeling data and real-time power consumption data; S2. Obtain and integrate historical building modeling data and historical electricity consumption data to obtain integrated historical building modeling data and historical electricity consumption data; construct an electricity demand forecasting model based on the integrated historical building modeling data and historical electricity consumption data; process the integrated building modeling data and real-time electricity consumption data through the electricity demand forecasting model to obtain electricity demand forecasting results for each building area; S3. Based on the electricity demand forecast results of each building area, an optimized allocation strategy is generated through a multi-objective optimization allocation algorithm; Based on the optimized allocation strategy, the power supply of each building area is dynamically adjusted to carry out intelligent allocation of power resources.

2. The method for intelligent allocation of building power resources based on BIM according to claim 1, characterized in that: The S1 specifically includes: The specific implementation steps of the integration process are: The first step is to preprocess the building modeling data and the real-time power consumption data of each building area to obtain the preprocessed building modeling data and the real-time power consumption data; In the second step, the preprocessed building modeling data and real-time electricity consumption data are matched using an association algorithm based on spatial regions and equipment numbers to obtain a matching matrix; The third step is to perform time series alignment on the matching matrix to obtain time-aligned integrated data; the time-aligned integrated data is structured to obtain integrated building modeling data and real-time electricity consumption data.

3. The method for intelligent allocation of building power resources based on BIM according to claim 2 is characterized in that: The S1 specifically includes: In the process of implementing the association algorithm based on spatial area and equipment number, a multidimensional matching algorithm based on spatial weight allocation is introduced to calculate the matching degree between the preprocessed real-time power consumption data and the preprocessed building modeling data to obtain the matching weight; based on the matching weight, a matching weight matrix is ​​constructed; the specific calculation formula of the matching weight is: Among them, w ij is the matching weight, which indicates the matching degree between the i-th record of the preprocessed real-time power consumption data and the j-th object of the preprocessed building modeling data; λ1 is the distance decay parameter; α, β, γ are the feature weight coefficients, which are used to control the influence of spatial distance, power similarity and regional area similarity on the matching weight respectively; (x j ,y j ) represents the spatial coordinates of the building area of ​​the jth object of the preprocessed building modeling data; represents the real-time spatial position of the ith record of the preprocessed real-time power consumption data; P j represents the equipment power of the jth object of the preprocessed building modeling data; A represents the real-time power of the ith record of the preprocessed real-time power consumption data; j represents the area of ​​the jth object of the preprocessed building modeling data; represents the area of ​​the i-th record of the preprocessed real-time electricity consumption data; It is an impact factor based on spatial distance; is the power similarity factor; is the area similarity factor.

4. The method for intelligent allocation of building power resources based on BIM according to claim 3 is characterized in that: The S1 specifically includes: A binary variable reflecting the object matching between the preprocessed real-time electricity consumption data and the preprocessed building modeling data is introduced, and a global optimization objective function is constructed in combination with constraint conditions to maximize the sum of all matching weights. After the optimization solution, a matching matrix is ​​constructed based on the binary variable. The matching matrix is ​​post-processed, and when the matching weight is less than or equal to the preset matching weight threshold, the specific calculation formula of the matching weight is readjusted.

5. The method for intelligent allocation of building power resources based on BIM according to claim 1, characterized in that: The S2 specifically includes: Based on the integrated historical building modeling data and historical electricity consumption data, a multi-layer nested time series characteristic association model is constructed as the electricity demand prediction model through multi-level characteristic association and nonlinear nested structure.

6. The method for intelligent allocation of building power resources based on BIM according to claim 5 is characterized in that: The S2 specifically includes: In the process of constructing the electricity demand prediction model, feature extraction is performed on the integrated historical building modeling data and historical electricity consumption data to obtain feature data; the feature data includes a building feature matrix and a time series feature matrix; the time series feature matrix is ​​feature decomposed to obtain different frequency characteristics; the building feature matrix is ​​subjected to principal component analysis, and the main components of building characteristics are extracted by dimensionality reduction; in the mapping unit in the hidden layer, the frequency characteristics are cross-correlated with the main components of building characteristics through mapping processing to obtain a nonlinear mapping matrix.

7. The method for intelligent allocation of building power resources based on BIM according to claim 6 is characterized in that: The S2 specifically includes: In the dynamic optimization unit in the hidden layer, based on the nonlinear mapping matrix, the time delay interaction processing is introduced, the time delay interaction results are calculated, and the time-space interaction characteristic matrix is ​​obtained.

8. The method for intelligent allocation of building power resources based on BIM according to claim 7 is characterized in that: The S2 specifically includes: A multi-level nested structure is introduced to nest and combine the time-space interaction characteristic matrices of different frequency levels to obtain the output of the hidden layer; the output of the hidden layer is input into the prediction layer, and the hyperbolic tangent function is used to predict the electricity demand in the future time period.

9. The method for intelligent allocation of building power resources based on BIM according to claim 1, characterized in that: The S3 specifically includes: Based on the electricity demand forecast results of each building area and combined with the preprocessed building modeling data, a multi-objective optimization function is constructed; the specific definition of the multi-objective optimization function is: in, is a multi-objective optimization function; It is The power resource allocation power of each building area at time t; It is The predicted value of electricity demand for each building area at time t; is the number of zones defined in the building; is the weight coefficient of the supply-demand balance target; It is Building area to electrical equipment Transmission loss at time t; It is Electrical equipment in building areas The power resource allocation power at time t; It is Building area to electrical equipment The physical distance between is the distance attenuation coefficient, which is used to control the impact of long-distance transmission energy consumption in optimization; ∈ is the decimal bias to prevent the denominator from being zero; is the number of all electrical equipment in the building; is the weight coefficient of the loss minimization objective; It is Building area and The coupling phase angle of each building area at time t; It is Building area and The fixed phase offset of each building area at time t; is the logarithm of the regions participating in the coupling in the building; η is the weight coefficient of the regional coupling target; T is the total duration of the optimization time window; The multi-objective optimization function is solved to obtain the optimal allocated power; based on the optimal allocated power, the optimal allocation strategy is generated.

10. The method for intelligent allocation of building power resources based on BIM according to claim 1, characterized in that: The S3 specifically includes: Collect actual electricity consumption data, compare and analyze the actual electricity consumption data with the electricity demand forecast results, and obtain comparative analysis results; based on the comparative analysis results, optimize the electricity demand forecast model and improve the power distribution strategy.

Images