A building energy consumption optimization method based on BIM

Abstract

The present application relates to the technical field of data processing, and more particularly to a building energy consumption optimization method based on BIM. The content includes: obtaining the original data of the target building and preprocessing, mapping the preprocessed original data to the BIM model, obtaining the dynamic BIM data model, and constructing the five-dimensional space-time feature tensor; the five-dimensional space-time feature tensor is processed by dimension reduction compression transformation, the three-dimensional tensor after compression transformation is used to construct the perturbation response function; based on the perturbation response function, the control inversion processing is carried out to obtain the preliminary control vector, the control path evolution mechanism is introduced to optimize and adjust the preliminary control vector; the optimized and adjusted control vector is converted into a control signal to realize building energy consumption optimization. The traditional building energy consumption optimization method in building energy consumption management has the problems of difficult spatio-temporal fusion of multi-source data, difficult accurate modeling of dynamic coupling relationship and lack of real-time adaptive ability of control vector.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, and in particular to a building energy consumption optimization method based on BIM. BACKGROUND

[0002] With the continuous improvement of building scale and function, building energy consumption has become an important part of the total energy consumption. Especially in large public buildings, complexes and smart parks, building operation involves HVAC systems, lighting systems, fresh air systems and other types of energy equipment. Its running state is influenced by external climate conditions, building structure characteristics, user behavior and other factors. The energy consumption changes show high spatio-temporal dynamics and multi-dimensional coupling characteristics. The existing building energy consumption management method usually relies on static BIM model for structural and equipment modeling analysis, or relies on energy consumption data collected by single type sensor for energy consumption evaluation. These methods often cannot realize the deep fusion of multi-source data, and lack the ability to globally correlate the building operation state under the unified spatio-temporal framework. At the same time, the existing energy consumption prediction and optimization technology is mostly based on experience rules or single-dimensional mathematical model, and fails to fully consider the dynamic interaction between structure information, equipment parameters, personnel behavior and environmental factors, resulting in limited prediction accuracy and difficulty in achieving stable and significant energy saving effect in actual operation.

[0003] In summary, the traditional building energy consumption optimization method still has the technical problems of difficulty in spatio-temporal fusion of multi-source data, difficulty in accurate modeling of dynamic coupling relationship and lack of real-time adaptive ability of control vector in building energy consumption management. SUMMARY

[0004] The present application provides a building energy consumption optimization method based on BIM to solve the technical problems of difficulty in spatio-temporal fusion of multi-source data, difficulty in accurate modeling of dynamic coupling relationship and lack of real-time adaptive ability of control vector in building energy consumption management.

[0005] The building energy consumption optimization method based on BIM of the present application specifically includes the following technical solutions:

[0006] A building energy consumption optimization method based on BIM includes the following steps:

[0007] S1. Obtain the original data of the target building and perform preprocessing to obtain the preprocessed original data; map the preprocessed original data to the BIM model to obtain a dynamic BIM data model, and construct a five-dimensional spatio-temporal feature tensor; perform dimensionality reduction compression transformation processing on the five-dimensional spatio-temporal feature tensor to obtain a compressed and transformed three-dimensional tensor; based on the compressed and transformed three-dimensional tensor, introduce an acceleration term, a nonlinear growth term and a suppression term, combine with the behavior-equipment coupling suppression weight, and construct a perturbation response function;

[0008] S2. Based on the perturbation response function, a control inversion process is performed to obtain a preliminary control vector, and a preliminary control vector trajectory is analyzed to obtain a path energy consumption quantification index; based on the path energy consumption quantification index, a control path evolution mechanism is introduced to optimize and adjust the preliminary control vector to obtain an optimized and adjusted control vector; the optimized and adjusted control vector is converted into a control signal to realize building energy consumption optimization.

[0009] Preferably, S1 specifically comprises:

[0010] Based on the dynamic BIM data model, the geometric structure, material parameters, equipment parameters, behavior mode and multi-dimensional time sequence sensing data of the component are exported, combined in the form of five-tuple, and a five-dimensional space-time feature tensor is generated.

[0011] Preferably, S1 specifically comprises:

[0012] The behavior-time weight is introduced, a logarithmic function is combined, a five-dimensional space-time feature tensor is processed by dimension reduction compression transformation to obtain a compressed and transformed three-dimensional tensor, and a sliding window is introduced to construct a time-based compressed and transformed three-dimensional tensor sequence.

[0013] Preferably, S2 specifically comprises:

[0014] Based on the perturbation response function, a target function is constructed in combination with a perturbation domain mapping function, a control inversion process is performed, a control vector solution required to minimize the target function is back calculated, and a preliminary control vector is obtained; the perturbation domain mapping function is constructed based on a differentiable physical control model.

[0015] Preferably, S2 specifically comprises:

[0016] Based on the perturbation domain mapping function, behavior-driven perturbation and equipment response perturbation are obtained, and the preliminary control vector trajectory is integrated and analyzed in combination with a behavior influence weight coefficient to obtain a path energy consumption quantification index.

[0017] Preferably, S2 specifically comprises:

[0018] In the implementation process of the control path evolution mechanism, based on the perturbation response function, the preliminary control vector is optimized and adjusted in combination with an adjustment rate factor to obtain an optimized and adjusted control vector.

[0019] Preferably, S2 specifically comprises:

[0020] The adjustment rate factor is calculated by normalizing the path energy consumption quantification index in combination with a preset adjustment rate range threshold.

[0021] The beneficial effects of the technical solutions of the present application are:

[0022] 1. By regarding the building energy consumption evolution process as a disturbance flow driven by the coupling of geometric structure, material parameters and equipment parameters, a disturbance response function with clear physical meaning is constructed, realizing the quantification and tracking of the dynamic change trend of energy consumption. The disturbance response function integrates acceleration terms, nonlinear growth terms and suppression terms, and can reflect sudden load changes, cumulative energy effects and the suppression effect of behavior-equipment interaction on energy consumption, thereby more accurately depicting the energy consumption evolution mechanism of buildings under different operating conditions.

[0023] 2. By constructing a disturbance domain mapping function, a differentiable physical correlation model between equipment control parameters and energy consumption disturbance is established, and a preliminary control vector is obtained by inverse optimization, realizing the direct derivation from the dynamic energy consumption target to specific executable control instructions, avoiding the solidification problem of traditional strategies based on experience or static optimization, and enabling the control vector to adapt to the energy consumption characteristic changes under different time periods and operating conditions. At the same time, the path energy consumption quantification index is introduced to integrate and evaluate the energy consumption trend of the preliminary control vector trajectory in the disturbance field, and the control path with low execution cost and high stability can be selected from multiple candidate paths that meet the disturbance suppression target, reducing energy consumption fluctuations and comfort degradation caused by improper control vector selection. BRIEF DESCRIPTION OF DRAWINGS

[0024] Figure 1 A flowchart of a building energy consumption optimization method based on BIM according to the present application. DETAILED DESCRIPTION

[0025] In order to further illustrate the technical means and effects adopted by the present application to achieve the predetermined invention purpose, the technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0026] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present application belongs.

[0027] The specific scheme of the building energy consumption optimization method based on BIM provided by the present application will be specifically described below in conjunction with the drawings.

[0028] Referring to the drawings Figure 1 , which shows a flowchart of a building energy consumption optimization method based on BIM provided by an embodiment of the present application, the method comprises the following steps:

[0029] S1. Obtain the raw data of the target building and preprocess it to obtain the preprocessed raw data; map the preprocessed raw data into the BIM model to obtain the dynamic BIM data model, and construct a five-dimensional spatiotemporal feature tensor; perform dimensionality reduction and compression transformation on the five-dimensional spatiotemporal feature tensor to obtain the compressed three-dimensional tensor; based on the compressed three-dimensional tensor, introduce acceleration terms, nonlinear growth terms, and suppression terms, and combine them with behavior-equipment coupling suppression weights to construct a disturbance response function;

[0030] Raw data is acquired in various functional areas of the target building using various types of IoT sensors (such as environmental sensors, load sensors, and behavioral sensors). The raw data is then preprocessed to obtain preprocessed raw data. The raw data includes temperature and humidity, illuminance, CO2 concentration, personnel activity status, and power consumption. The preprocessing includes data cleaning, time alignment, missing data completion, normalization, and standardization. These preprocessing processes are all techniques well-known to those skilled in the art and will not be elaborated upon here.

[0031] The preprocessed raw data is spatially registered with the existing BIM model. Using existing three-dimensional coordinate mapping algorithms and component coding rules, the preprocessed raw data is fused and mapped onto the specific building components in the BIM model to construct a semantically enhanced dynamic BIM data model. This is a technical method well known to those skilled in the art and will not be elaborated here.

[0032] Based on the dynamic BIM data model, the geometric structure of the components (discretized into a set of node geometric voxel coordinates), material parameters (thermal conductivity, density, specific heat capacity, etc.), equipment parameters (cooling load, fan power, operating frequency, etc.), behavioral patterns (personnel paths, work and rest cycles, door and window operation status, etc.), and multi-dimensional time-series sensor data (temperature, humidity, carbon dioxide concentration, illuminance, etc.) are derived and combined into a five-dimensional spatiotemporal feature tensor in the form of quintuples. The five dimensions of the five-dimensional spatiotemporal feature tensor correspond to: the number of components. Material parameters quantity Equipment parameters and quantity Number of behavioral patterns and time series length ;

[0033] Furthermore, based on BIM-IoT fusion technology, the five-dimensional spatiotemporal feature tensor is subjected to dimensionality reduction and compression transformation to obtain the compressed three-dimensional tensor. The specific calculation formula is as follows:

[0034] ;

[0035] in, represents the three-dimensional tensor after compression transformation, describes the feature strength of the first component, the first material parameter, the first equipment parameter, for subsequent disturbance response function construction process, and retains the nonlinear cross characteristics between different dimensions of data; represents the element value corresponding to the first component, the first material parameter, the first equipment parameter, and the first behavior mode in the five-dimensional spatiotemporal feature tensor at the time ; ; is a behavior-time weight, representing the fusion weight of the first behavior mode at the time , reflecting the relative contribution of the behavior mode to the energy consumption situation at the time, and is determined according to expert experience method, with a reference value range of , and the cumulative sum is 1; is a minimum stability term, used to prevent the logarithmic term from exploding or dividing by zero when the denominator is very small, and can be valued at ; is a combination of logarithmic terms, based on the robustness of the logarithmic scale, to ensure that the resolution of large values is compressed and small values are improved;

[0036] The calculation method of the three-dimensional tensor after compression transformation is based on the three-dimensional tensor after compression transformation, and a sliding window is preset according to specific application requirements. The five-dimensional spatiotemporal feature tensor is processed by dimension reduction compression transformation in the sliding window, to generate the three-dimensional tensor after compression transformation. With each sliding step, the three-dimensional tensor after compression transformation corresponding to the window is obtained, and a time-based sequence of the three-dimensional tensor after compression transformation is further obtained;

[0037] Further, the energy consumption evolution is regarded as a disturbance flow driven by the coupling of geometric structure, material parameter and equipment parameter. Based on physical disturbance flow modeling and nonlinear control function, a disturbance response function that can be inverted at the control link is constructed, and the specific formula is as follows:

[0038] ;

[0039] wherein, is a disturbance response function; is an overall energy consumption disturbance response scalar at the time , that is, a disturbance intensity. The greater the value of the overall energy consumption disturbance response scalar, the more obvious the "intensification trend" of energy consumption at the time; is a disturbance intensity adjustment coefficient, used to control the steepness of the nonlinear curve, and is determined according to expert experience method, with a reference value range of ; is the behavior-device coupling inhibition weight used to control the inhibition term The strength of the inhibition term is determined according to expert experience method, and the reference value range is ; is the characteristic strength of the component-material-device at the moment, that is the three-dimensional tensor after compression transformation at the moment, obtained through dimension reduction compression transformation processing in the sliding time window, indicating the energy consumption influence degree at the moment ; is the second-order change rate (acceleration term) of the characteristic strength with time, used to represent the dynamic change rate of the disturbance, reflecting the instantaneous change trend of the building energy consumption response; is the nonlinear growth term of energy consumption, and the Sigmoid function in the denominator can compress high-value disturbances to a stable interval to avoid response divergence.

[0040] S2. Based on the disturbance response function, control inversion processing is performed to obtain a preliminary control vector, and the preliminary control vector trajectory is analyzed to obtain a path energy consumption quantization index; based on the path energy consumption quantization index, a control path evolution mechanism is introduced to optimize and adjust the preliminary control vector to obtain an optimized and adjusted control vector; the optimized and adjusted control vector is converted into a control signal to realize building energy consumption optimization.

[0041] Based on the disturbance response function, a target function is constructed in combination with a disturbance domain mapping function, control inversion processing is performed, and a control vector solution required to minimize the target function is back calculated to obtain a preliminary control vector; the disturbance domain mapping function is generated based on a differentiable physical control model and is used to quantify the correction effect of the disturbance at the moment; the control vector is represented as , , represents the i-th device control parameter (such as air supply temperature, supply fan speed ratio, lighting brightness ratio, fresh air valve opening ratio, shading device angle ratio, etc.), and represents the total number of device control parameters; the target function formula is as follows:

[0042] ;

[0043] wherein, is the preliminary control vector, that is, the optimal control vector, representing the combination of device control parameters that minimizes the energy consumption disturbance of the target building in the entire time interval ; is the optimization time interval length, which is determined according to specific application scenarios and will not be described here; is the disturbance domain mapping function between the control vector and the disturbance, representing​​ At all times in the control vector Under the influence of the building equipment, the predicted energy consumption disturbance value generated by the building equipment is generated by the differentiable physical control model. The differentiable physical control model includes equipment power-state function (e.g., the functional relationship between air conditioning power and supply air temperature and fan speed), heat transfer and ventilation model (heat transfer coefficient, volume air exchange rate, etc. calculated by BIM model), personnel behavior response model (obtained by coupling occupancy rate and equipment response), etc. The construction of the differentiable physical control model is a technical means well known to those skilled in the art, and will not be described in detail here. This is the gradient regularization weight coefficient, used to balance fitting accuracy and control smoothness to prevent drastic changes in equipment control parameters between adjacent time steps. It is set using expert experience based on the on-site equipment response capability and comfort requirements. A larger gradient regularization weight coefficient results in smoother changes in equipment control parameters, but may reduce convergence speed. A reference value range is [insert range here]. ; It is the disturbance domain mapping function to the control vector The gradient, representing the gradient at... The sensitivity of the device control parameters to energy consumption disturbances is obtained through numerical differentiation, a technique well-known to those skilled in the art, and will not be elaborated here.

[0044] After obtaining the initial control vector Subsequently, to avoid instability caused by the optimal control vector in the disturbance space, a formula for calculating the path energy consumption quantification index is introduced. An integral analysis is performed on the evolution trajectory of the control path in the disturbance field to obtain the path energy consumption quantification index. The specific formula is as follows:

[0045] ;

[0046] in, It is along the initial control vector trajectory The corresponding path energy consumption quantification index is used to measure the overall energy consumption trend of the initial control vector trajectory; It is the path integral domain, which is the set of trajectories formed by the evolution of the initial control vector in the disturbance phase space (with the disturbance intensity and control vector as coordinate axes). It is obtained based on existing path reconstruction and mapping techniques, which are well known to those skilled in the art and will not be elaborated here. Is Moment, behavior-driven disturbance Direction and device response disturbance The angle between the directions in the perturbation phase space reflects the phase matching degree of their interaction, with a reference range of values. ; is a behavior influence weight coefficient, used to adjust the proportion of behavior disturbance in the path energy consumption quantization index, determined according to expert experience method, and the reference value range is ; is a disturbance component caused by user behavior (such as window opening, occupant number change, light manual adjustment, etc.) at time , i.e. behavior-driven disturbance, obtained by disturbance domain mapping function ; is a disturbance component caused by device execution (air conditioner, lighting, fresh air system, etc.) at time , i.e. device response disturbance, obtained by disturbance domain mapping function ; is used to measure the disturbance change speed when the behavior and the device response direction are consistent; is used to measure the proportion of behavior-driven disturbance relative to device disturbance; is a very small positive number, used to prevent the denominator from being 0; the calculation method of the behavior-driven disturbance and the device response disturbance is a technical means familiar to those skilled in the art, which will not be described here;

[0047] Further, in order to ensure that the preliminary control vector has time dynamics and strain capacity, a control path evolution mechanism is introduced, so that the preliminary control vector can be optimized and adjusted according to the disturbance trend to obtain the optimized and adjusted control vector, and the specific adjustment formula is as follows:

[0048] ;

[0049] wherein, is the optimized and adjusted control vector; is an adjustment rate factor, calculated based on the path energy consumption quantization index, and the specific calculation process is: the path energy consumption quantization index is normalized to obtain the normalized path energy consumption quantization index , combined with the adjustment rate range threshold , to obtain , and respectively represent the minimum value and the maximum value of the adjustment rate, which are obtained according to expert experience preset; represents the disturbance response function corresponding to the th device control parameter at time , based on the gradient information of the disturbance response function and the physical mapping relationship of each device control parameter, the disturbance response function is projected to each device control parameter to obtain ; is the th device control parameter in the preliminary control vector;For the logarithmic compression term of disturbance intensity, the influence of large disturbance is mapped to sub-linear growth, avoiding too large update step; For the phase-based nonlinear shaping function, it is used to compress the disturbance intensity to [−1, 1] and introduce periodic suppression to prevent coupled resonance of multiple device control parameters when updating in the same direction; is the global pressure amplitude multiplied by a smooth saturation function, which is used to ensure that the aggregated amount in the parentheses falls in , multiplied by determines the final adjustment;

[0050] Finally, the optimized control vector is converted into a control signal using the existing PID technology to achieve building energy optimization. The PID technology is a mature and general means, and will not be described here.

[0051] In summary, a building energy optimization method based on BIM is completed.

[0052] The order of the embodiments of the application is only for description, 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, multi-task processing and parallel processing are also possible or may be advantageous.

[0053] Each embodiment in the specification is described in a progressive manner, and the same or similar parts between each embodiment can be referred to each other. Each embodiment focuses on the difference from other embodiments.

[0054] The above embodiments are only used to illustrate the technical solutions of the present application, but not limit them. Although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that they can modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for some technical features. The modification or replacement does not make the essence of the corresponding technical solution deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application.

Claims

1. A BIM-based building energy consumption optimization method, characterized in that, Includes the following steps: S1. Obtain the raw data of the target building and preprocess it to obtain the preprocessed raw data; The preprocessed raw data is mapped onto the BIM model to obtain a dynamic BIM data model; Based on the dynamic BIM data model, the geometric structure, material parameters, equipment parameters, behavioral patterns, and multi-dimensional temporal sensing data of components are exported and combined in the form of quintuples to construct a five-dimensional spatiotemporal feature tensor. Behavior-time weights are introduced, and combined with a logarithmic function, the five-dimensional spatiotemporal feature tensor is subjected to dimensionality reduction and compression transformation to obtain a compressed three-dimensional tensor. A sliding window is then introduced to construct a time-varying sequence of the compressed three-dimensional tensor. The specific formula is as follows: ； in, Represents the three-dimensional tensor after compression transformation, describing the first... The first component, the first The material parameter, the first The characteristic intensity of each device parameter; It is the length of the time series; It is the number of behavioral patterns; Indicates in Time of the first The first component, the first The material parameter, the first The device parameters, the first The element values ​​corresponding to each behavioral pattern in the five-dimensional spatiotemporal feature tensor; It is behavior-time weight, representing the time weight. Time of the first The fusion weight of various behavioral patterns; It is a minimal stable term, used to prevent the logarithmic term from exponentially increasing or being divided by zero when the denominator is very small; It is a logarithmic combination term, which, based on the robustness of the logarithmic scale, ensures that large values ​​are compressed and small values ​​are improved in terms of resolution; Based on the compressed 3D tensor, an acceleration term, a nonlinear growth term, and a suppression term are introduced, and a perturbation response function is constructed by combining behavior-device coupling suppression weights. The specific formula is as follows: ； in, Is The overall energy consumption disturbance response scalar at any given time, i.e., the disturbance intensity; It is the disturbance intensity adjustment coefficient; These are behavior-device coupling suppression weights; yes The characteristic strength of the component-material-equipment at a given moment, i.e. The three-dimensional tensor after compression transformation at any time; It is the acceleration term; This is a non-linear growth term for energy consumption; It is a suppressive term; S2. Based on the disturbance response function and the disturbance domain mapping function, construct the objective function, perform control inversion processing, and derive the control vector solution required to minimize the objective function, thus obtaining the preliminary control vector. Decompose the disturbance based on the disturbance domain mapping function to obtain the behavior-driven disturbance and the equipment response disturbance. Combine this with the behavior influence weighting coefficient, perform integral analysis on the preliminary control vector trajectory to obtain the path energy consumption quantification index. The specific formula is as follows: ； in, It is along the initial control vector trajectory Quantify the energy consumption indicators of the corresponding path; It is the path integral domain; Is Moment, behavior-driven disturbance Direction and device response disturbance The angle between the directions in the perturbation phase space; It is the weighting coefficient of behavioral influence; At any moment The disturbance component caused by user behavior, i.e., behavior-driven disturbance; Is The disturbance component caused by the device's execution at any given time, i.e., the device response disturbance; It is a very small positive number, used to prevent the denominator from being 0; Based on the path energy consumption quantification index, a control path evolution mechanism is introduced. Based on the disturbance response function and combined with the adjustment rate factor, the initial control vector is optimized and adjusted to obtain the optimized control vector. The specific adjustment formula is as follows: ； in, It is the optimized and adjusted control vector; It is the rate adjustment factor; Indicates in Time of the first The disturbance response function corresponding to each device control parameter is obtained by projecting the disturbance response function onto each device control parameter based on the gradient information of the disturbance response function and the physical mapping relationship between the two devices control parameters. It is the initial control vector The Middle Each device control parameter; This indicates the total number of equipment control parameters; The optimized control vector is converted into a control signal to achieve building energy consumption optimization.

2. The BIM-based building energy consumption optimization method according to claim 1, characterized in that, S2 specifically includes: The perturbation domain mapping function is constructed based on a differentiable physical control model.

3. The BIM-based building energy consumption optimization method according to claim 1, characterized in that, S2 specifically includes: The adjustment rate factor is calculated by normalizing the path energy consumption quantification index and combining it with a preset adjustment rate range threshold.

Images