Energy-saving control method and system based on building group spatial form

Abstract

The invention discloses an energy-saving control method and system based on a building group spatial form, and relates to the technical field of intelligent building energy-saving control, and the method comprises the steps: inputting a building form characteristic parameter matrix into a multi-scale generative adversarial network, respectively extracting the overall layout characteristics, local spacing characteristics and surface detail characteristics of a building group, and calculating the overall layout characteristics, local spacing characteristics and surface detail characteristics of the building group; generating an energy consumption probability distribution thermodynamic diagram corresponding to the building group space distribution; and inputting the energy consumption probability distribution thermodynamic diagram into a discriminator of the multi-scale generative adversarial network, calculating an optimization gradient of each building morphological parameter, and generating a morphological adjustment instruction set in combination with a dynamic constraint function. The method not only improves the accuracy of energy consumption prediction, but also can identify a high-energy-consumption area and provide targeted energy-saving measures. The problems that in the prior art, energy consumption prediction precision is limited, and it is difficult to efficiently convert into operable energy-saving measures are solved.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent building energy-saving control, and in particular to an energy-saving control method and system based on the spatial form of building complexes. Background Art

[0002] With the acceleration of global urbanization, building energy consumption has increasingly become one of the key factors affecting energy consumption and environmental protection. Early research focused on the design optimization of single buildings, such as reducing energy consumption by improving building materials and optimizing building appearance design. In recent years, with the development of computer technology, especially the advancement of artificial intelligence algorithms, the research focus has gradually shifted to the analysis of the overall spatial layout and morphological characteristics of building complexes, aiming to achieve more efficient energy-saving control strategies through intelligent means.

[0003] The existing technology has shortcomings: 1. Traditional methods often rely on static data for energy consumption assessment, which makes it difficult to reflect the dynamic impact of changes in the internal and external environment of the building complex in real time, resulting in limited accuracy in energy consumption prediction. 2. In addition, the current technical solutions are still insufficient for how to efficiently transform complex building form characteristics into actionable energy-saving measures. Summary of the invention

[0004] In view of the above existing problems, the present invention is proposed.

[0005] Therefore, the present invention provides an energy-saving control method based on the spatial morphology of building complexes to solve the problems in the prior art of limited energy consumption prediction accuracy and difficulty in efficiently converting complex building morphological features into operational energy-saving measures.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions: In the first aspect, the present invention provides an energy-saving control method based on the spatial morphology of a building complex, which includes collecting the three-dimensional spatial coordinates, volume, orientation angle and compound curvature parameters of each building in the building complex, and constructing a building morphology feature parameter matrix; inputting the building morphology feature parameter matrix into a multi-scale adversarial generation network, extracting the overall layout characteristics, local spacing characteristics and surface detail characteristics of the building complex respectively, and generating an energy consumption probability distribution heat map corresponding to the spatial distribution of the building complex; inputting the energy consumption probability distribution heat map into the discriminator of the multi-scale adversarial generation network, calculating the optimization gradient of each building morphology parameter, and generating a morphology adjustment instruction set in combination with a dynamic constraint function; distributing the morphology adjustment instruction set to each control terminal in the building complex through a low-latency communication protocol, and dynamically adjusting the operating parameters of the equipment based on a multi-scale adaptive control algorithm.

[0007] As a preferred solution of the energy-saving control method based on the spatial form of building complexes described in the present invention, wherein: the composite curvature parameter refers to the result of fusing the microscopic surface curvature, the mesoscopic spacing curvature and the macroscopic layout curvature through a multi-scale fusion mechanism; The microscopic surface curvature is the ratio of the Gaussian curvature to the average curvature of the triangular mesh of the building surface; The mesoscopic spacing curvature is a statistical dispersion characteristic of the side lengths of the Delaunay triangulation of adjacent buildings; The macro layout curvature is the eigenvalue difference relationship of the building complex coordinate covariance matrix.

[0008] As a preferred solution of the energy-saving control method based on the spatial form of building complexes described in the present invention, wherein: the generator of the multi-scale adversarial generative network is composed of a three-level pyramid hierarchical attention mechanism; The macro layer adopts a decaying spatial attention mechanism; The meso layer uses a windowed local attention mechanism; The micro layer adopts a curvature modulated multi-head attention mechanism.

[0009] As a preferred solution of the energy-saving control method based on the spatial form of building complexes described in the present invention, the energy consumption probability distribution heat map is input into the discriminator of the multi-scale adversarial generation network to calculate the optimization gradient of each building form parameter. The specific steps are as follows: Normalize the energy consumption probability distribution heat map and perform building location mask overlay processing; Perform multi-scale convolution feature extraction on the processed energy consumption probability distribution heat map and generate a confidence matrix; Based on the confidence matrix, the optimization gradient of each building morphology parameter is identified.

[0010] As a preferred solution of the energy-saving control method based on the spatial form of building complexes described in the present invention, wherein: the form adjustment instruction set is generated by combining the dynamic constraint function, and the specific steps are as follows: The optimization gradient of the building morphology parameters is processed through the application of dynamic constraint functions and gradient correction; Based on the processed building morphology parameters, a morphology adjustment instruction set is generated through the Adam optimizer.

[0011] As a preferred solution of the energy-saving control method based on the spatial form of the building complex described in the present invention, the multi-scale adaptive control algorithm includes the overall orientation adjustment of the building complex at the macro level, the ventilation coordination between adjacent buildings at the meso level, and the angle adjustment of the facade sunshade at the micro level.

[0012] As a preferred solution of the energy-saving control method based on the spatial form of a building complex described in the present invention, the low-latency communication protocol adopts a zero message queue architecture to achieve end-to-end transmission, and dynamically selects the primary and backup transmission paths through the discriminator's confidence evaluation of the energy consumption probability distribution heat map.

[0013] In the second aspect, the present invention provides an energy-saving control system based on the spatial morphology of a building complex, including a data acquisition module for collecting the three-dimensional spatial coordinates, volume, orientation angle and compound curvature parameters of each building in the building complex, and constructing a building morphology feature parameter matrix; an energy consumption analysis module for inputting the building morphology feature parameter matrix into a multi-scale adversarial generation network, extracting the overall layout characteristics, local spacing characteristics and surface detail characteristics of the building complex respectively, and generating an energy consumption probability distribution heat map corresponding to the spatial distribution of the building complex; an optimization and adjustment module for inputting the energy consumption probability distribution heat map into the discriminator of the multi-scale adversarial generation network, calculating the optimization gradient of each building morphology parameter, and generating a morphology adjustment instruction set in combination with a dynamic constraint function; a control execution module for distributing the morphology adjustment instruction set to each control terminal in the building complex through a low-latency communication protocol, and dynamically adjusting the operating parameters of the equipment based on a multi-scale adaptive control algorithm.

[0014] In a third aspect, the present invention provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program is executed by the processor, any step of the energy-saving control method based on the spatial form of a building complex as described in the first aspect of the present invention is implemented.

[0015] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of the energy-saving control method based on the spatial form of a building complex as described in the first aspect of the present invention.

[0016] The beneficial effects of the present invention are as follows: by accurately collecting the three-dimensional spatial coordinates, volume, orientation angle and compound curvature parameters of the building complex, a detailed architectural morphological characteristic parameter matrix is ​​constructed, and the overall layout, local spacing and surface detail features of the building complex are extracted using a multi-scale adversarial generative network to generate an energy consumption probability distribution heat map. This not only improves the accuracy of energy consumption prediction, but also identifies high energy consumption areas and provides targeted energy-saving measures. Through intelligent data processing and analysis, the goal of significantly improving the energy efficiency management of the building complex is achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0018] Figure 1 is a flow chart of the energy-saving control method based on the spatial form of a building complex in Example 1; Figure 2 This is a flow chart of the calculation of the compound curvature parameters in Example 1; Figure 3 This is a flowchart of the multi-scale adversarial generation network in Example 1; Figure 4 This is a flowchart of the morphology adjustment instruction generation in Example 1. DETAILED DESCRIPTION

[0019] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0020] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0021] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0022] Example 1, reference Figure 1~Figure 4 , this embodiment provides an energy-saving control method based on the spatial form of a building complex, comprising the following steps: S1: Collect the three-dimensional spatial coordinates, volume, orientation angle and compound curvature parameters of each building in the building complex to construct the architectural morphological characteristic parameter matrix; S1.1: Collect the three-dimensional coordinates, volume, orientation angle and compound curvature parameters of each building in the building complex; Use drones to fly in a cross-circling path, obtain multi-view images of the building complex through RTK positioning, and deploy ground control points; It should be noted that a five-lens oblique photography module was installed on a drone to set a cross-circular flight path; the differential signal of the Qianxun location network was received through the built-in RTK module of the drone to achieve real-time dynamic positioning, synchronously record the POS data of each image, sort the image acquisition order by timestamp, generate the original sequence index of the aerial photography path, and generate the spatial topological sequence index through spatial proximity reordering; a ground control point was arranged every 200 square meters within the building complex, and a 30×30 cm cross-shaped coding mark was used for the ground control point. The three-dimensional coordinates were measured using a Leica TS16 total station, and each ground control point was covered by at least 3 images from different angles; after the flight was completed, the image and POS data were exported and imported into the ContextCapture software together with the ground control point coordinate table as the input source for subsequent dense point cloud generation.

[0023] Use ContextCapture to generate a dense point cloud, and align the dense point cloud with the ground control points using the ICP algorithm (iterative closest point); It should be noted that the multi-view images taken by drones are imported into the ContextCapture software, and the image matching parameters including image overlap rate, feature point matching accuracy, and point cloud resolution are set, and the initial dense point cloud is generated by the multi-view stereo matching algorithm; the three-dimensional coordinates of the ground control points are measured by the total station, and the corresponding ground control point positions in the dense point cloud are manually annotated in ContextCapture; based on the dense point cloud and the ground control points, the ICP algorithm (iterative closest point) is called to calculate the rigid body transformation matrix, and the Euclidean distance error between the dense point cloud and the ground control points is minimized, and iterative optimization is performed until convergence; the registered dense point cloud is output.

[0024] The registered dense point cloud is used to generate a triangular mesh model of the building through the Poisson surface reconstruction algorithm, and the geometric parameters of each building including three-dimensional coordinates, volume, orientation angle and compound curvature parameters are extracted; It should be noted that the specific process of generating a building triangular mesh model from the registered dense point cloud through the Poisson surface reconstruction algorithm is as follows: based on the registered dense point cloud data, the Poisson surface reconstruction algorithm is used to calculate the point cloud normal vector and construct an implicit surface function, and the building triangular mesh model is generated through octree space partitioning and isosurface extraction; In the triangular mesh model, the three-dimensional coordinates are obtained by calculating the coordinates of the geometric center point of each building triangular mesh model. The building triangular mesh model is decomposed into non-overlapping tetrahedral units through the Delaunay tetrahedron partitioning algorithm and the volume is obtained by accumulating the volumes of all tetrahedrons. The covariance matrix of the triangular mesh projection point set of the building bottom surface is calculated by the principal component analysis method, and the eigenvector direction corresponding to the maximum eigenvalue is extracted, and the orientation angle is calculated in combination with the angle to the true north direction. The specific process of the composite curvature parameters fusing the micro surface curvature, meso spacing curvature and macro layout curvature through a multi-scale fusion mechanism is as follows: based on the Gaussian curvature and mean curvature of the building surface triangular mesh, the ratio of Gaussian curvature to mean curvature is processed by the normalized inverse tangent function to generate the micro surface curvature; based on the statistical characteristics of the Delaunay triangulation side lengths between adjacent buildings, the ratio of the square of the standard deviation of the triangulation side lengths to the mean is used to characterize the discreteness of the spacing distribution, and the meso spacing curvature is generated; based on the eigenvalue difference of the covariance matrix of the main axis direction of the building complex, the ratio of the absolute difference of the first two eigenvalues ​​of the covariance matrix to the sum is used to characterize the directional strength of the layout, and the macro layout curvature is generated; the micro surface curvature, meso spacing curvature and macro layout curvature are weighted summed to generate the composite curvature parameters.

[0025] S1.2: Construct a building morphology parameter matrix based on the geometric parameters of each building, including three-dimensional coordinates, volume, orientation angle, and compound curvature parameters.

[0026] S2: Input the architectural morphological feature parameter matrix into the multi-scale adversarial generative network to extract the overall layout characteristics, local spacing characteristics and surface detail characteristics of the building complex, and generate an energy consumption probability distribution heat map corresponding to the spatial distribution of the building complex; S2.1: Normalize and embed the architectural form parameter matrix into high dimensions; Normalize the three-dimensional coordinates, volume, orientation angle and compound curvature parameters in the architectural form parameter matrix; It should be noted that the three-dimensional coordinate parameters in the architectural morphological characteristic parameter matrix are converted into relative coordinates with the geometric center of the building complex as the origin, and normalized to the interval [-1, 1]; the volume parameter is divided by the maximum volume value in the building complex and constrained to the interval [0, 1]; the orientation angle parameter (radian system) is divided by π and constrained to the interval [-1, 1].

[0027] The normalized architectural morphology parameter matrix is ​​mapped into a high-dimensional embedding matrix through a learnable weight matrix.

[0028] Furthermore, the normalized 6-dimensional parameters are mapped to 512-dimensional vectors through the learnable weight matrix in the multi-scale adversarial generative network to generate a high-dimensional embedding matrix; It should be noted that the normalized architectural morphology parameter matrix, which contains 6 columns of parameters, including 3D coordinates, volume, orientation angle, and compound curvature, is input into the multi-scale adversarial generative network. The learnable weight matrix in the multi-scale adversarial generative network is used to perform linear transformation, and the 6-dimensional parameters of each building are mapped into a 512-dimensional vector to generate an initial high-dimensional embedded architectural morphology parameter matrix. Based on the spatial topological sequence index of the drone aerial photography path, sine and cosine position codes are generated and superimposed element by element by row-by-row addition to a high-dimensional embedding matrix; It should be noted that based on the spatial topological sequence of the drone aerial photography path (i.e., the order of image acquisition), a unique position index value is assigned to each building. According to the Transformer position encoding rule, the even dimensions of the 512-dimensional embedding vector are filled with sine function values, and the odd dimensions are filled with cosine function values. In the specific calculation, the position index is divided by the power of 10,000 and then trigonometric functions are performed; the generated sine and cosine position codes are superimposed element by element on the high-dimensional embedding matrix, so that the high-dimensional embedding matrix contains both building morphological characteristics and spatial order information. The high-dimensional embedding matrix is ​​used as the input of the three-level pyramid Transformer generator for subsequent multi-scale feature extraction and heat map generation.

[0029] S2.2: Through the three-level pyramid Transformer generator, the overall layout features, local spacing features and surface detail features of the building complex are extracted respectively; The overall layout feature extraction of the macro layer is to combine the logarithmic attenuation term of the building spacing in the Softmax function. For every 50-meter increase in the building spacing, the association weight of the distant building is reduced. Finally, the overall layout features of the macro layer are output to characterize the global layout correlation of the building complex.

[0030] The local spacing feature extraction at the meso-level is to divide the buildings into 7×7 grid windows according to normalized plane coordinates through window division and orientation constraints; the self-attention weights between buildings are calculated in each window, and an exponential decay penalty is imposed on building pairs with orientation angle differences exceeding 30° to reduce their interaction weights; the meso-level feature matrix is ​​output to characterize the local building spacing and shadow occlusion effects.

[0031] It should be noted that window partitioning is to project the three-dimensional spatial coordinates of each building in the building complex onto a plane, ignore the height dimension, generate a two-dimensional plane coordinate set, normalize the two-dimensional plane coordinate set, and divide it into a uniform grid; calculate the absolute value of the orientation angle difference of each pair of buildings in the grid, and combine the exponential decay penalty term to perform orientation constraints; The specific steps for calculating the self-attention weights between buildings are as follows: based on the high-dimensional embedding matrix, a query matrix, a key matrix and a value matrix are generated through linear transformation; for each building in the grid window, the dot product of the transposed matrix of the query matrix and the key matrix is ​​calculated, and the result is divided by the scaling factor to obtain the initial attention score matrix; for each pair of buildings in the window, if the orientation angle difference exceeds 30°, the initial attention score is multiplied by an exponential decay weight to reduce the interaction strength of the building pair; the attenuated attention score matrix is ​​normalized by the Softmax function to generate a normalized attention weight matrix; the normalized attention weight matrix is ​​multiplied by the value matrix to generate the feature representation of the building in the window; all grid windows are traversed, and the feature representations output by each window are spliced ​​in the order of the original building positions to form a complete mesoscopic feature matrix to characterize the local building spacing and shadow occlusion effects.

[0032] The surface detail feature extraction at the micro level is achieved by inputting the compound curvature parameters into a two-layer fully connected network through curvature modulation attention calculation to generate a modulation coefficient in the interval [0, 1]. In three independent attention heads, the modulation coefficient is used as a scaling factor of the key-value pair to enhance the feature contribution of high-curvature buildings. The micro feature matrix is ​​output to characterize the local impact of surface curvature on energy consumption.

[0033] S2.3: Generate a heat map of energy consumption probability distribution corresponding to the spatial distribution of the building complex; Through hierarchical weight calculation and feature weighted fusion, dynamic gated multi-scale feature fusion is performed to generate a multi-scale fusion feature matrix; Furthermore, the macro feature matrix, the meso feature matrix, and the micro feature matrix are globally averaged and pooled to generate scalar weight values. The weights are normalized by the Softmax function to ensure that the sum of the weights of the three is 1. Multiply the macro feature matrix by the macro weight, the meso feature matrix by the meso weight, and the micro feature matrix by the micro weight, and add them element by element to obtain the fused feature matrix; spatially align the fused feature matrix through a 3×3 convolution kernel to eliminate the resolution differences of features at different scales.

[0034] Based on the deconvolution network structure and spatial coordinate mapping, deconvolution upsampling and spatial mapping are performed to generate a single-channel heat map matrix; Furthermore, based on the deconvolution network structure, a single-channel heat map matrix is ​​generated; The first deconvolution layer: input multi-scale fusion feature matrix (dimension n×512), use 3×3 convolution kernel (step size 2), and increase the output resolution to n×256×256.

[0035] The second deconvolution layer: repeat 3×3 convolution (step 2), and the resolution is gradually increased to 64×64 and 128×128.

[0036] Channel adjustment: The last layer of deconvolution uses a 1×1 convolution kernel to reduce the number of channels from 512 to 1 and generate a single-channel heat map matrix.

[0037] Generate heat map pixels through spatial coordinate mapping; The normalized coordinates of the building are aligned with the pixel grid of the heat map, and the discrete building features in the multi-scale fusion feature matrix are mapped to the continuous space through bilinear interpolation. In the deconvolution process, the non-building areas of the heat map (such as green spaces and roads) are constrained to low energy consumption values ​​according to the building location mask (generated by three-dimensional coordinates).

[0038] Combined with the Sigmoid activation function and physical constraint correction, the heat map probability is normalized to generate the energy consumption probability distribution heat map; Furthermore, the single-channel heat map matrix is ​​input into the Sigmoid function, and the original value is constrained to the interval [0, 1] to generate a probability distribution heat map. According to the heat transfer equation (Fourier's law), the heat map gradient is forced to match the material thermal conductivity in the building surface area, and the L2 regularization constraint is applied to the pixels in the non-building area (outside the mask) to avoid noise interference.

[0039] S3: Input the energy consumption probability distribution heat map into the discriminator of the multi-scale adversarial generative network, calculate the optimization gradient of each building morphological parameter, and generate a morphological adjustment instruction set in combination with the dynamic constraint function; S3.1: preprocessing the energy consumption probability distribution heat map; The pixel values ​​of the energy consumption probability distribution heat map are linearly mapped from [0, 1] to the interval [-1, 1] to adapt to the input range of the discriminator of the multi-scale generative adversarial network. The heat map resolution is adjusted from 128×128 to 256×256 using the bicubic interpolation method to improve the discriminator's sensitivity to detail features.

[0040] A binary mask is generated based on the collected three-dimensional coordinates of the building, and the building projection area is marked as 1, and the non-building area (road, green space) is marked as -1; the binary mask is multiplied pixel by pixel with the normalized energy consumption probability distribution heat map to suppress noise interference in the non-building area.

[0041] S3.2: Perform multi-scale convolution feature extraction on the processed energy consumption probability distribution heat map, generate a confidence matrix, and calculate the objective function; The processed energy consumption probability distribution heat map is input into the discriminator of the multi-scale adversarial generative network, and passes through 5 layers of convolution operations in sequence. The resolution of the feature map output by each layer is reduced by half. A channel attention module is inserted after the third layer of convolution to dynamically adjust the weight according to the importance of the feature map channel. The last layer of convolution outputs an 8×8 feature map, which is mapped to a 128×128 confidence matrix through a fully connected layer. Each pixel value ∈ [-1, 1] indicates the degree of match between the heat map at that position and the actual energy consumption distribution. The mean of the confidence matrix is ​​calculated as the global optimization target.

[0042] S3.3: Based on the objective function, calculate the optimization gradient of each building form parameter; Furthermore, the gradient of the building morphology parameter matrix is ​​calculated through the PyTorch automatic differentiation framework with the confidence mean as the optimization target. The gradient of the building morphology parameters of each building is calculated independently. The building morphology parameters are sorted according to the absolute value of the gradient, and the highly sensitive parameters are optimized first. Sign constraints are imposed on the gradients of the compound curvature parameters. The compound curvature parameters whose gradient values ​​exceed the range of [-0.5, 0.5] are truncated to avoid optimization oscillation.

[0043] S3.4: Generate morphology adjustment instruction set in combination with dynamic constraint function; Dynamic constraint function application and gradient correction are performed through minimum spacing constraint correction, curvature safety range constraint and volume cost constraint correction; Furthermore, all building pairs are traversed. If the distance between buildings is less than the planned fire protection distance, a reverse correction is applied to the coordinate gradient. For every 1 meter reduction in building distance, the corresponding coordinate gradient is reduced by the correction factor; if the composite curvature parameter exceeds the material safety range, the curvature gradient is set to zero and further adjustment is prohibited; according to the volume adjustment cost model, the cost penalty item is deducted from the optimization gradient of the building form parameters. The volume gradient is linearly reduced according to the cost weight (for example, 0.01); It should be noted that the volume adjustment cost model collects the economic parameters of building energy-saving transformation, including unit volume adjustment cost (yuan / cubic meter) and material strength limit unit price (yuan / MPa), and stores them in the dynamic constraint database. Based on the economic parameters in the dynamic constraint database, the volume adjustment cost model is constructed. When initializing the Adam optimizer, the learning rate, momentum decay parameter, and second-order moment estimation decay parameter are set, and the first-order moment estimation vector and the second-order moment estimation vector are initialized as zero matrices. The gradients of the architectural morphology parameter matrix are sorted according to the absolute value of the gradient, and the priority order is volume parameter > compound curvature parameter > three-dimensional coordinate parameter. The update operation is performed on each parameter in turn: the current parameter gradient value is extracted from the optimized gradient of the architectural morphology parameter, multiplied by the learning rate, and the superposition momentum term is obtained based on the exponential decay weighted average of the first-order moment estimation vector and the momentum decay parameter. The second-order moment estimation correction term is obtained based on the weighted average of the second-order moment estimation vector and the second-order moment estimation decay parameter, and finally the parameter update amount is generated.

[0044] Input the adjusted morphological parameter matrix into the three-level pyramid Transformer generator to generate a new energy consumption probability distribution heat map and calculate the confidence improvement; call the collision detection interface of the BIM model to check whether the adjusted building spacing and curvature violate safety regulations. If there is a conflict, roll back to the most recent valid parameter version; It should be noted that the adjusted morphological parameter matrix is ​​input into the generator of the pre-trained multi-scale adversarial generative network to generate a new energy consumption probability distribution heat map; the new heat map is input into the discriminator, the new confidence matrix is ​​output and the mean is calculated, and the improvement is obtained by comparing with the original confidence mean; if the improvement is less than 0.1, the conflict detection process is triggered. Call the collision detection interface of the BIM model, traverse the three-dimensional coordinates of all building pairs, calculate the plane Euclidean distance, and if there is a building spacing less than the fire spacing specification value (for example, 6 meters), mark it as a spacing conflict; check whether the composite curvature parameter exceeds the material safety range, and if it exceeds, mark it as a curvature conflict. If there is any conflict, retrieve the most recent conflict-free morphological parameter matrix from the historical version library and replace the current parameter matrix; re-input the parameter matrix after rollback into the generator to generate a heat map and calculate the confidence. If the improvement is ≥ 0.1 and there is no conflict, save it as the latest valid version.

[0045] Generate an instruction list according to the "building ID, parameter type, adjustment amount" triple, sort it in ascending order by building ID, and generate a morphology adjustment instruction set.

[0046] S4: Distribute the morphology adjustment instruction set to each control terminal in the building complex through a low-latency communication protocol, and dynamically adjust the operating parameters of the equipment based on a multi-scale adaptive control algorithm.

[0047] Furthermore, the morphology adjustment instruction set based on the JSON format is transmitted to the edge gateway in the building complex through the MQTT protocol; It should be noted that when the morphological adjustment instruction set in JSON format (including building ID, parameter type, adjustment amount, and timestamp fields) is transmitted to the edge gateway in the building complex through the MQTT protocol, the morphological adjustment instruction set in JSON format is first UTF-8 encoded to generate a binary data stream; a message is published to the parameterized topic path through the MQTT client, in which the topic path parameter of the MQTT protocol is replaced with the building complex; the edge gateway subscribes to the topic path parameter of the MQTT protocol, parses the JSON field after receiving the message, and verifies the legitimacy of the building ID (for example, checking whether the ID is in the pre-registration list). If it is legal, the morphological adjustment instruction set is distributed to the corresponding control terminal according to the building ID; if the reception confirmation (ACK) message is not received, it is retransmitted after 50ms according to the QoS=1 rule, and retransmitted up to 3 times; after the transmission is completed, the edge gateway generates a transmission log (including message ID, timestamp, and building ID list) and stores it in the local database.

[0048] The edge gateway routes the command to the corresponding control terminal based on the building ID; It should be noted that after receiving the morphology adjustment instruction set in JSON format, the edge gateway parses the building ID field and queries the predefined building ID-IP address mapping table; verifies whether the building ID exists in the registration list, and extracts the target IP address if it does; publishes the instruction message to the exclusive topic of the target control terminal through the MQTT protocol; the control terminal subscribes to the topic corresponding to its building ID, receives and parses the instruction content; the edge gateway records the transmission log, including the building ID, target IP address, and timestamp, and stores it in the local SQLite database.

[0049] After receiving the command, the control terminal analyzes the parameter type and adjustment amount, and dynamically adjusts the equipment operating parameters through a multi-scale adaptive control algorithm; Furthermore, the deployment percentage of the building facade shading components is adjusted proportionally according to the volume adjustment amount at a macro time scale (hour level); At the mesoscopic time scale (minute level), the opening and air volume of the ventilation system are adjusted by linear interpolation according to the compound curvature parameter adjustment amount; At a microscopic time scale (seconds), the temperature setting value of the air conditioning equipment is dynamically corrected through the PID control algorithm based on the direction angle adjustment combined with the real-time light intensity and temperature sensor data; All device parameter adjustment results are fed back to the edge gateway via the Modbus-TCP protocol, generating real-time operation logs and synchronizing them to the building complex management platform.

[0050] The present embodiment also provides an energy-saving control system based on the spatial morphology of a building complex, including: a data acquisition module, used to collect the three-dimensional spatial coordinates, volume, orientation angle and compound curvature parameters of each building in the building complex, and construct a building morphology feature parameter matrix; an energy consumption analysis module, used to input the building morphology feature parameter matrix into a multi-scale adversarial generation network, respectively extract the overall layout characteristics, local spacing characteristics and surface detail characteristics of the building complex, and generate an energy consumption probability distribution heat map corresponding to the spatial distribution of the building complex; an optimization and adjustment module, used to input the energy consumption probability distribution heat map into the discriminator of the multi-scale adversarial generation network, calculate the optimization gradient of each building morphology parameter, and generate a morphology adjustment instruction set in combination with a dynamic constraint function; a control execution module, used to distribute the morphology adjustment instruction set to each control terminal in the building complex through a low-latency communication protocol, and dynamically adjust the operating parameters of the equipment based on a multi-scale adaptive control algorithm.

[0051] This embodiment also provides a computer device, which is suitable for the energy-saving control method based on the spatial form of a building complex, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute computer-executable instructions to implement the energy-saving control method based on the spatial form of a building complex proposed in the above embodiment.

[0052] The computer device may be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. The display screen of the computer device may be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device may be a touch layer covered on the display screen, or a key, trackball or touchpad provided on the housing of the computer device, or an external keyboard, touchpad or mouse, etc.

[0053] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, the energy-saving control method based on the spatial form of a building complex as proposed in the above embodiment is implemented; the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (Static Random Access Memory, referred to as SRAM), electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, referred to as EEPROM), erasable programmable read-only memory (Erasable Programmable Read Only Memory, referred to as EPROM), programmable read-only memory (Programmable Red-Only Memory, referred to as PROM), read-only memory (Read-Only Memory, referred to as ROM), magnetic storage, flash memory, magnetic disk or optical disk.

[0054] In summary, the present invention achieves the goal of significantly improving the energy efficiency management of building complexes by: accurately collecting the three-dimensional spatial coordinates, volume, orientation angle and compound curvature parameters of the building complex, constructing a detailed architectural morphological feature parameter matrix, and using a multi-scale adversarial generative network to extract the overall layout, local spacing and surface detail features of the building complex, and generating an energy consumption probability distribution heat map. This not only improves the accuracy of energy consumption prediction, but also identifies high-energy consumption areas and provides targeted energy-saving measures. Through intelligent data processing and analysis, the goal of significantly improving the energy efficiency management of building complexes is achieved, solving the problem of limited energy consumption prediction accuracy and difficulty in efficiently converting into operational energy-saving measures in the prior art.

[0055] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. An energy-saving control method based on the spatial form of a building complex, characterized by: include, Collect the three-dimensional spatial coordinates, volume, orientation angle and compound curvature parameters of each building in the building complex to construct the architectural morphological characteristic parameter matrix; The architectural morphological feature parameter matrix is ​​input into a multi-scale adversarial generative network to extract the overall layout characteristics, local spacing characteristics and surface detail characteristics of the building complex, and generate a heat map of energy consumption probability distribution corresponding to the spatial distribution of the building complex. The energy consumption probability distribution heat map is input into the discriminator of the multi-scale adversarial generative network to calculate the optimization gradient of each building morphological parameter, and the morphological adjustment instruction set is generated in combination with the dynamic constraint function; The morphology adjustment instruction set is distributed to each control terminal in the building complex through a low-latency communication protocol, and the operating parameters of the equipment are dynamically adjusted based on a multi-scale adaptive control algorithm.

2. The energy-saving control method based on the spatial form of building complexes according to claim 1, characterized in that: The composite curvature parameter refers to the result of fusing the microscopic surface curvature, the mesoscopic spacing curvature and the macroscopic layout curvature through a multi-scale fusion mechanism; The microscopic surface curvature is the ratio of the Gaussian curvature to the average curvature of the triangular mesh of the building surface; The mesoscopic spacing curvature is a statistical dispersion characteristic of the side lengths of the Delaunay triangulation of adjacent buildings; The macro layout curvature is the eigenvalue difference relationship of the building complex coordinate covariance matrix.

3. The energy-saving control method based on the spatial form of building complexes according to claim 1, characterized in that: The generator of the multi-scale adversarial generative network is composed of a three-level pyramid hierarchical attention mechanism; The macro layer adopts a decaying spatial attention mechanism; The meso layer uses a windowed local attention mechanism; The micro layer adopts a curvature modulated multi-head attention mechanism.

4. The energy-saving control method based on the spatial form of building complexes according to claim 3 is characterized in that: The energy consumption probability distribution heat map is input into the discriminator of the multi-scale adversarial generation network to calculate the optimization gradient of each building form parameter. The specific steps are as follows: Normalize the energy consumption probability distribution heat map and perform building location mask overlay processing; Perform multi-scale convolution feature extraction on the processed energy consumption probability distribution heat map and generate a confidence matrix; Based on the confidence matrix, the optimization gradient of each building morphology parameter is identified.

5. The energy-saving control method based on the spatial form of building complexes according to claim 1, characterized in that: The specific steps of generating the morphology adjustment instruction set by combining the dynamic constraint function are as follows: The optimization gradient of the building morphology parameters is processed through the application of dynamic constraint functions and gradient correction; Based on the processed building morphology parameters, a morphology adjustment instruction set is generated through the Adam optimizer.

6. The energy-saving control method based on the spatial form of building complexes according to claim 1, characterized in that: The multi-scale adaptive control algorithm includes the overall orientation adjustment of the building complex at the macro level, the ventilation coordination between adjacent buildings at the meso level, and the angle adjustment of the facade sunshades at the micro level.

7. The energy-saving control method based on the spatial form of building complexes according to claim 6, characterized in that: The low-latency communication protocol adopts a zero message queue architecture to achieve end-to-end transmission, and dynamically selects the primary and backup transmission paths through the confidence evaluation of the energy consumption probability distribution heat map by the discriminator.

8. An energy-saving control system based on the spatial form of a building complex, based on the energy-saving control method based on the spatial form of a building complex according to any one of claims 1 to 7, characterized in that: include, The data acquisition module is used to collect the three-dimensional spatial coordinates, volume, orientation angle and compound curvature parameters of each building in the building complex, and construct a matrix of architectural morphological characteristic parameters; Energy consumption analysis module, which is used to input the architectural morphological characteristic parameter matrix into the multi-scale adversarial generative network, respectively extract the overall layout characteristics, local spacing characteristics and surface detail characteristics of the building complex, and generate the energy consumption probability distribution heat map corresponding to the spatial distribution of the building complex; The optimization and adjustment module is used to input the energy consumption probability distribution heat map into the discriminator of the multi-scale adversarial generation network, calculate the optimization gradient of each building form parameter, and generate a form adjustment instruction set in combination with the dynamic constraint function; The control execution module is used to distribute the morphology adjustment instruction set to each control terminal in the building complex through a low-latency communication protocol, and dynamically adjust the operating parameters of the equipment based on the multi-scale adaptive control algorithm.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the energy-saving control method based on the spatial form of a building complex as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the energy-saving control method based on the spatial form of a building complex as described in any one of claims 1 to 7 are implemented.

Images