Installation effect evaluation system applied to household energy storage equipment

Abstract

The invention discloses an installation effect evaluation system applied to household energy storage equipment, and relates to the technical field of equipment installation evaluation.The installation effect evaluation system is characterized in that three-dimensional point cloud data, image features, technical parameters and environmental parameters of an installation space are acquired through an installation parameter acquisition module, and a structured case database is constructed; the comprehensive evaluation module is combined with aesthetic evaluation and utility evaluation to generate a comprehensive score so as to evaluate the installation effect; the preferable case library construction module is used for screening high-quality cases, carrying out feature coding, constructing a case similarity map and dynamically maintaining a set of the first ten cases; the collaborative recommendation module receives a new installation demand, and outputs an optimized installation parameter package through k-nearest neighbor search, parameter fusion optimization and virtual installation verification; according to the invention, collaborative optimization of technical performance and humanity requirements is realized, the installation decision-making efficiency is improved, and the installation of the household energy storage equipment is promoted to develop towards the intelligentization and individuation directions.

Description

Technical Field

[0001] The present invention relates to the technical field of equipment installation evaluation, and in particular to an installation effect evaluation system applied to household energy storage equipment. Background Art

[0002] With the popularization of household energy storage equipment, the evaluation of its installation effect has become a key link in improving user experience and equipment performance. Traditional evaluation systems focus more on technical parameters (such as equipment power, heat dissipation efficiency, etc.) and select installation solutions through static indicators. However, such methods have significant defects in multi-dimensional comprehensive evaluation and reuse of historical experience: first, there is a lack of quantitative evaluation of the coordination between the installation solution and the home environment, resulting in the equipment meeting the functionality standards but lacking in aesthetics; second, it relies on manual experience decisions, making it difficult to dynamically optimize installation parameters, resulting in repeated trial and error and waste of resources in similar scenarios. Existing technologies have not yet effectively solved the problem of coordinated optimization of technical parameters and human needs, limiting the personalized adaptation capabilities of household energy storage equipment.

[0003] The current mainstream solutions have the following core problems: First, the aesthetic evaluation system is missing. The traditional method has not established a quantitative model for aesthetic indicators such as decoration style matching and color coordination. It only makes subjective judgments through manual visual inspection, and the consistency of the evaluation results is poor (the error rate is as high as 40% or more). For example, the impact of the reflectivity of the wall material on the visual integration of the equipment appearance is not considered, or the user's personalized aesthetic preferences are ignored. Secondly, the utilization rate of case data is low. Historical installation data is stored in the form of isolated cases, and there is a lack of an intelligent recommendation mechanism based on similarity retrieval, resulting in a reuse rate of less than 15% for high-quality installation solutions. In addition, the static weight allocation model cannot adapt to dynamic environmental changes (such as furniture layout adjustments, equipment aging, etc.), causing the evaluation results to lag behind actual needs.

[0004] The above defects lead to two major difficulties in the installation of household energy storage equipment: the imbalance between technical performance and aesthetic requirements, and user satisfaction is generally lower than industry expectations; the installation decision-making efficiency is low, and the average solution generation takes more than 2 hours. In the existing technology, although some studies have attempted to introduce three-dimensional modeling or user scoring mechanisms, closed-loop optimization of multi-source data fusion analysis and self-evolution case library has not yet been achieved. Therefore, it is urgent to build an intelligent evaluation system that integrates aesthetic evaluation, dynamic case matching and parameter self-optimization to break through the existing technical bottleneck and promote the installation of household energy storage equipment towards intelligence and personalization.

[0005] In order to solve the above defects, a technical solution is now provided. Summary of the invention

[0006] The purpose of the present invention is to solve the problems in the prior art of the installation effect evaluation of household energy storage equipment, such as the lack of multi-dimensional comprehensive evaluation, low utilization rate of high-quality case data, and difficulty in dynamically optimizing installation parameters, and to propose an installation effect evaluation system for household energy storage equipment.

[0007] The purpose of the present invention can be achieved through the following technical solutions: The installation effect evaluation system for household energy storage equipment includes: Installation parameter acquisition module, used to use the equipment to collect 3D point cloud data, image features, technical parameters and environmental parameters of the installation space, and build a structured case database to store the data; A comprehensive evaluation module is used to calculate the aesthetic score and the utility score through aesthetic evaluation and utility evaluation, and then generate a comprehensive score to evaluate the installation effect of the home energy storage device; Optimize the case library construction module, screen high-quality cases, encode their features, build a case similarity map, and dynamically maintain the top ten case sets to provide reference for recommendations; The collaborative recommendation module receives new installation requirements, extracts spatial feature vectors, and outputs the optimized installation parameter package through k-nearest neighbor search, parameter fusion optimization, and virtual installation verification.

[0008] Furthermore, the execution process of the installation parameter collection module is as follows: Use laser scanners to obtain 3D point cloud data of the installation space and reconstruct the BIM model including wall structure and furniture layout; Use a wide-angle camera to capture a panoramic image of the installation area and extract HSV color space histogram features; Record technical parameter set: including battery pack arrangement, ground clearance, and heat dissipation channel inclination; Collect environmental parameters: Use a spectrophotometer to measure the reflectivity of wall materials and the reverberation time of the space; Build a structured case database, where each case is stored as a tuple Case = {coordinates (x, y, z), technical parameters T, aesthetic features A, environmental parameters E}; Finally, semantic data annotation and retrieval are performed.

[0009] Furthermore, the specific operation steps for semantic data annotation and retrieval in the installation parameter collection module are as follows: Semantically annotate the stored case data and use the ontology model to define the semantic relationship between the data; When building a structured case database, efficient retrieval based on semantics is achieved; When users can query related cases through natural language descriptions, the search results are sorted according to the degree of match, and cases with high matching degrees are displayed first; Present the case to the user in a visual way, including: displaying key information of the case in list form, including installation coordinates, technical parameters, aesthetic features and environmental parameter summary.

[0010] Furthermore, the specific process of the comprehensive evaluation module is as follows: Aesthetic evaluation: The ResNet-50 model is used to extract deep features of the installation area image. The input installation area image is subjected to feature extraction through multiple convolutional layers, pooling layers, and fully connected layers. Calculate the cosine similarity with the preset decoration style template, assuming that the image feature vector of the installation area extracted from the ResNet-50 model is , the preset feature vector of a certain decoration style template is , then the cosine similarity calculation formula between them is: ;in, is a vector and The dot product of and They are vectors and The similarity value range is between [-1,1], and the similarity result is used as the AI ​​score, recorded as ; Integrating user subjective ratings, users can view the virtual effects of home energy storage devices in the installation area through the augmented reality interface and rate them from 1 to 5 stars based on their aesthetic feelings, which is recorded as ; Output aesthetic score ; Where 0.6 and 0.4 are weight coefficients; Utility evaluation: Calculate installation compactness: ; Measuring maintenance convenience: simulate the maintenance path length, and use the A* algorithm to calculate the shortest accessible path. When measuring maintenance convenience, the maintenance personnel's starting position is used as the starting point, and the part of the household energy storage device that needs to be repaired is used as the target point. The shortest accessible path length L calculated by the A* algorithm is the simulated maintenance path length; Evaluate thermal management performance: Use fluid dynamics simulation to calculate the uniformity of air velocity distribution in the heat dissipation channel to evaluate thermal management performance. In fluid dynamics simulation, by establishing a physical model of the heat dissipation channel, setting boundary conditions, and solving the fluid mechanics Navier-Stokes equation, the distribution of air velocity in the heat dissipation channel is obtained. Output utility score: First, the installation compactness C, simulated maintenance path length L, and air velocity variation coefficient CV are standardized using the linear normalization method: , , ,in are the minimum and maximum values ​​of installation compactness in all evaluated cases, respectively; are the minimum and maximum values ​​of the simulated maintenance path lengths in all evaluated cases, respectively; are the minimum and maximum values ​​of the coefficient of variation of air velocity in all evaluated cases, respectively; They are the normalized values ​​of installation compactness C, simulated maintenance path length L, and coefficient of variation CV of air velocity, respectively; The utility score calculation formula is: ;in are the dynamic weights of installation compactness, simulated maintenance path length and air velocity variation coefficient, respectively; and ; Generate a composite score: ,in, and is the weight coefficient, and .

[0011] Furthermore, the specific operation steps of the preferred case library construction module are as follows: High-quality case screening: Continuously monitor the cases newly entered into the case database in real time; when each new case is entered into the database, it is judged based on the comprehensive score S given by the comprehensive evaluation module; Compare the comprehensive score S with the preset threshold θ; if S>θ, the case is marked as a high-quality case and enters the subsequent processing flow; If S≤θ, it will not enter the high-quality case-related processing, but will only be stored in the case library as an ordinary case for subsequent reference; Feature encoding: Dimensionality reduction of three-dimensional space coordinates. For high-quality cases, the t-SNE algorithm is used to process the three-dimensional space coordinates; the t-SNE algorithm maps high-dimensional data to low-dimensional space while maintaining the relative distance relationship between data points; One-hot encoding of technical parameters: One-hot encoding is used to process the technical parameters T of high-quality cases; Construct case similarity graph: Based on the encoded high-quality case features, calculate the Mahalanobis distance between the feature vectors of family scenarios; Set a distance threshold δ. When the calculated Mahalanobis distance d of two cases is less than δ, the two cases are judged to be similar. At this time, an edge connection is established for the two cases in the case similarity graph to associate them. By continuously calculating and comparing the Mahalanobis distances between all high-quality cases, a graph structure reflecting the case similarity relationship is gradually constructed, which facilitates the rapid search for similar cases. Dynamically maintain the top ten case collection: maintain a collection that only retains the top ten high-quality cases; use the LRU algorithm to manage the collection; Whenever a new high-quality case is added to the collection, when the collection has reached ten cases, the LRU algorithm will check the call time record of each case in the collection; eliminate the old cases that are more than 30 days away from the current time and have the lowest frequency of use within 30 days to make room for the new case; Through a dynamic maintenance mechanism, we ensure that the top ten case collections always contain the most valuable high-quality cases, providing case data support that is more in line with actual needs for subsequent collaborative recommendation functions.

[0012] Furthermore, the specific operation steps of the collaborative recommendation module are as follows: Receive new installation requirements and extract spatial feature vectors: When there is a new home energy storage device installation requirement, the system first extracts the spatial feature vector from the current home’s relevant data ;The data comes from the three-dimensional point cloud data of the installation space obtained by the laser scanner and the panoramic image of the installation area taken by the wide-angle camera; Perform k-nearest neighbor search: Take the extracted spatial feature vector Based on, perform k nearest neighbor search in the case library; calculate The similarity between the spatial feature vectors of all historical cases in the case library. The similarity calculation method includes Euclidean distance or cosine similarity. By calculation and comparison, find the similarity The five most similar historical cases; Parameter fusion optimization: For the five most similar historical cases found, their technical parameters are weighted averaged; the weight is calculated by multiplying the similarity by the comprehensive score. That is, the higher the similarity between each case and the current demand and the higher its own comprehensive score, the greater its weight in the weighted average; Suppose these five historical cases are ; Five historical cases and The similarities are The comprehensive scores are , the technical parameters are ; When analyzing the technical parameters of the battery pack arrangement, assume that the value is series or parallel. When performing weighted average calculation, first represent the series and parallel with numerical values, specifically 0 for series and 1 for parallel, and calculate the weighted average value. The formula is: ; Each technical parameter is calculated by weighted average calculation to finally generate a candidate parameter set. ,in The arrangement of the battery pack, is the height above the ground, It is the inclination angle of the heat dissipation channel, which is the parameter combination with the highest weight after considering the similarity and comprehensive score.

[0013] Furthermore, the collaborative recommendation module also includes: Virtual installation verification: Inject the generated candidate parameter sets into the digital twin system for conflict detection; During the conflict detection process, check whether there is a pipeline collision or the safety distance does not meet the specification requirements; When a problem is detected, parameter adjustment is automatically triggered; the adjustment method is based on preset rules and algorithms to adjust, recalculate and optimize the parameters until the installation requirements are met; Output optimized installation parameter package: After virtual installation verification and parameter adjustment, output optimized installation parameter package, which contains the following information: Three-dimensional positioning coordinates: Provide the three-dimensional positioning coordinates of the home energy storage device in the installation space; Recommended wiring scheme: presented in the form of a color-coded path heat map. Different colors represent different wiring paths. The color depth of the heat map reflects the priority or difficulty of the wiring. Aesthetic optimization suggestions: Based on the aesthetic evaluation of the installation area and the matching analysis with the preset decoration style, specific aesthetic optimization suggestions are given.

[0014] Compared with the prior art, the present invention has the following beneficial effects: (1) The present invention integrates functions such as installation parameter collection, comprehensive evaluation, construction of a preferred case library and collaborative recommendation, and innovates the way of evaluating the installation effect of household energy storage equipment. The installation parameter collection module accurately obtains multi-dimensional data of the installation space and builds a structured case database. The comprehensive evaluation module integrates aesthetic and utility evaluation, uses the ResNet-50 model to extract image features, calculates the similarity with the preset style template as the AI ​​score, and combines the user's subjective score to derive the aesthetic score. The utility is evaluated from three aspects: installation compactness, maintenance convenience and thermal management efficiency, and finally a comprehensive score is generated. This series of operations effectively solves the defects of traditional evaluation systems that only focus on technical parameters and lack multi-dimensional comprehensive evaluation, realizes the coordinated optimization of technical performance and humanistic needs, and improves the overall evaluation level of the installation effect of household energy storage equipment. (2) In the present invention, the preferred case library construction module continuously screens high-quality cases, constructs a case similarity map through feature coding, and dynamically maintains the top ten case sets to provide support for the collaborative recommendation module; after receiving new installation requirements, the collaborative recommendation module extracts spatial feature vectors, uses k-nearest neighbor search to locate similar cases in the case library, generates candidate parameter sets through parameter fusion optimization, and finally outputs the optimized installation parameter package through virtual installation verification; this process makes full use of historical high-quality cases to achieve intelligent recommendation and optimization of installation parameters, avoids repeated trial and error and waste of resources, significantly improves installation decision-making efficiency, and promotes the installation of home energy storage equipment to develop in the direction of intelligence and personalization; (3) The present invention focuses on the combination of aesthetic evaluation and actual installation effect. Through semantic data annotation and retrieval, users can query related cases through natural language descriptions and present them in a visual way, which is convenient for users to quickly screen and view, thereby improving user participation and satisfaction. In the process of optimizing installation parameters, the matching degree between the aesthetic characteristics of the installation area and the preset decoration style is fully considered, and specific aesthetic optimization suggestions are given to effectively solve the problem that the equipment functionality meets the standards but the aesthetics is insufficient in the traditional method. The installation effect of the home energy storage equipment can better integrate into the home environment while meeting the technical performance requirements, thereby improving users' satisfaction with the installation effect and promoting the development of the home energy storage equipment installation industry in the direction of taking into account both technology and aesthetics. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to facilitate understanding by those skilled in the art, the present invention is further described below in conjunction with the accompanying drawings; Figure 1 This is the overall system block diagram of the present invention. DETAILED DESCRIPTION

[0016] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0017] It should be understood that the terms "include" and "comprising" used in the specification and claims of the present disclosure indicate the presence of described features, integers, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or collections thereof.

[0018] It should also be understood that the terms used in this disclosure are only for the purpose of describing specific embodiments and are not intended to limit the disclosure. As used in this disclosure and claims, the singular forms of "a", "an", and "the" are intended to include the plural forms unless the context clearly indicates otherwise. It should also be further understood that the term "and / or" used in this disclosure and claims refers to any combination of one or more of the associated listed items and all possible combinations, including these combinations.

[0019] like Figure 1 As shown, the installation effect evaluation system applied to household energy storage equipment includes an installation parameter acquisition module, a comprehensive evaluation module, a preferred case library construction module and a collaborative recommendation module; The installation parameter acquisition module uses the equipment to collect the three-dimensional point cloud data, image features, technical parameters and environmental parameters of the installation space, and builds a structured case database to store the data; The laser scanner is used to obtain the three-dimensional point cloud data of the installation space and reconstruct the BIM model including the wall structure and furniture layout; a wide-angle camera is used to capture a panoramic image of the installation area and extract the HSV color space histogram features; a set of technical parameters is recorded: including the arrangement of the battery pack (series / parallel), the height from the ground (accurate to the millimeter level), and the inclination of the heat dissipation channel; environmental parameters are collected: the reflectivity of the wall material is measured (using a spectrophotometer) and the spatial reverberation time (to evaluate the noise propagation of the equipment); Build a structured case database, where each case is stored as a tuple Case = {coordinates (x, y, z), technical parameters T, aesthetic features A, environmental parameters E}; Semantic data annotation and retrieval: Semantically annotate the stored case data and use the ontology model to define the semantic relationship between the data; in this way, when building a structured case database, efficient retrieval based on semantics is achieved. When users can query related cases through natural language descriptions, the search results are sorted according to the degree of matching, and cases with high matching degrees are displayed first. The cases are presented to users in a visual way, such as displaying key case information in a list format, including installation coordinates, technical parameters, aesthetic features, and environmental parameter summaries, so that users can quickly filter and view.

[0020] The comprehensive evaluation module calculates the aesthetic score and utility score through aesthetic evaluation and utility evaluation, and then generates a comprehensive score to evaluate the installation effect of home energy storage equipment; Aesthetic evaluation: The ResNet-50 model is used to extract deep features of the installation area image. The input installation area image is subjected to feature extraction through multiple convolutional layers, pooling layers, and fully connected layers. Calculate the cosine similarity with the preset decoration style template (modern / classical / industrial style), assuming that the installation area image feature vector extracted from the ResNet-50 model is , the preset feature vector of a certain decoration style template is , then the cosine similarity calculation formula between them is: ;in, is a vector and The dot product of and They are vectors and The similarity value range is between [-1,1]. The closer the value is to 1, the more similar the installation area image is to the preset decoration style template; the closer the value is to -1, the greater the difference between the two. This similarity result is used as the AI ​​score, recorded as ; Integrating user subjective ratings, users can intuitively view the virtual effects of home energy storage devices in the installation area through the augmented reality (AR) interface, and rate them from 1 to 5 stars based on their own aesthetic feelings, which is recorded as ; Output aesthetic score ; 0.6 and 0.4 are weight coefficients used to adjust the relative importance of AI ratings and user ratings in aesthetic ratings; Utility evaluation: Calculate installation compactness: ; The higher the installation compactness, the more efficient the equipment is in space utilization; Measuring maintenance convenience: simulate the maintenance path length, and use the A* algorithm to calculate the shortest accessible path. When measuring maintenance convenience, the maintenance personnel's starting position is used as the starting point, and the key parts of the home energy storage device that need to be repaired are used as the target point. The shortest accessible path length L calculated by the A* algorithm is the simulated maintenance path length. The shorter the simulated maintenance path length, the smaller the operating space and moving distance required when repairing the equipment, and the higher the maintenance convenience; Evaluate thermal management performance: Use computational fluid dynamics (CFD) simulation to calculate the uniformity of air velocity distribution in the heat dissipation channel to evaluate thermal management performance. In CFD simulation, by establishing a physical model of the heat dissipation channel, setting boundary conditions (such as inlet air velocity, temperature, etc.), and solving fluid mechanics equations (such as the Navier-Stokes equation), the distribution of air velocity in the heat dissipation channel is obtained; Output utility score: First, the installation compactness C, simulated maintenance path length L, and air velocity variation coefficient CV are standardized using the linear normalization method: , , ,in are the minimum and maximum values ​​of installation compactness in all evaluated cases, respectively; are the minimum and maximum values ​​of the simulated maintenance path lengths in all evaluated cases, respectively; are the minimum and maximum values ​​of the coefficient of variation of air velocity in all evaluated cases, respectively; , , are respectively the normalized values ​​of installation compactness C, simulated maintenance path length L, and air velocity coefficient of variation CV; the utility score calculation formula is: ;in They are the dynamic weights of installation compactness, simulated maintenance path length, and air velocity variation coefficient; ; Generate a composite score: ,in, and is the weight coefficient, and , initial values ​​α=0.4,β=0.6.

[0021] The optimal case library construction module selects high-quality cases, encodes their features, constructs a case similarity map, and dynamically maintains the top ten case sets to provide reference for recommendations; High-quality case screening: Continuously monitor the cases newly entered into the case library in real time. When each new case is entered into the library, it will be judged based on the comprehensive score S given by the comprehensive evaluation module; the comprehensive score S will be compared with the preset threshold θ (the default value of θ is 80); if (S>θ), ​​the case will be marked as a high-quality case and enter the subsequent processing flow; if (S≤θ), it will not enter the high-quality case-related processing and will only be stored in the case library as an ordinary case for subsequent general reference.

[0022] Feature encoding: Dimensionality reduction of three-dimensional space coordinates: For cases marked as high-quality, the t-SNE (t-distributed stochastic neighbor embedding) algorithm is used to process their three-dimensional space coordinates. The t-SNE algorithm can map high-dimensional data to low-dimensional space (here, the three-dimensional space coordinates are reduced to two-dimensional embedding vectors), while trying to maintain the relative distance relationship between data points for subsequent analysis and visualization; One-hot encoding of technical parameters: One-hot encoding is used for the technical parameters T of high-quality cases. Taking "parallel connection_30 degree tilt" as an example, assuming that among all possible technical parameter combinations, "parallel connection_30 degree tilt" is one of the specific combinations, it is encoded as [0,1,0,0,...,1]. In this encoding vector, only the position corresponding to "parallel connection_30 degree tilt" is 1, and the other positions representing different technical parameter combinations are all 0. Through this encoding, the non-numerical data such as technical parameters are converted into a numerical vector form that is easy for computer processing and analysis, which is convenient for subsequent calculation of the similarity between cases.

[0023] Construct case similarity graph: Based on the encoded high-quality case features, calculate the Mahalanobis distance between the feature vectors of the family scenes; set the distance threshold δ (δ=0.35); when the calculated Mahalanobis distance d of two cases is less than δ, the two cases are judged to be similar cases. At this time, establish an edge connection for the two cases in the case similarity graph to associate them. By continuously calculating and comparing the Mahalanobis distances between all high-quality cases, a graph structure reflecting the case similarity relationship is gradually constructed to facilitate the rapid search for similar cases; Dynamically maintain the top ten case collection: maintain a collection that only retains the top ten high-quality cases. The LRU (least recently used) algorithm is used to manage the collection; whenever a new high-quality case is added to the collection, if the collection is full (reaches ten cases), the LRU algorithm will check the call time record of each case in the collection. Eliminate old cases that are more than 30 days away from the current time and have the lowest frequency of use (i.e., least recently used) in the past 30 days to make room for new cases. Through this dynamic maintenance mechanism, it is ensured that the top ten case collection always contains the most valuable high-quality cases in the near future, providing case data support that is more in line with actual needs for subsequent collaborative recommendation and other functions.

[0024] The collaborative recommendation module receives new installation requirements, extracts spatial feature vectors, and outputs optimized installation parameter packages through k-nearest neighbor search, parameter fusion optimization, and virtual installation verification. Receive new installation requirements and extract spatial feature vectors: When there is a new home energy storage device installation requirement, the system first extracts the spatial feature vector from the current home’s relevant data ; These data come from the three-dimensional point cloud data of the installation space obtained by the laser scanner, the panoramic image of the installation area taken by the wide-angle camera, etc.; For example, the BIM model containing the wall structure and furniture layout is reconstructed by the laser scanner, and the space size, shape, obstacle location and other information are extracted from it; the image taken by the wide-angle camera is used to analyze the color distribution and lighting conditions of the space, and these multi-source information are integrated and feature extracted to finally obtain a vector that can represent the characteristics of the current home installation space. ; Perform k-nearest neighbor search: Take the extracted spatial feature vector Based on, perform a k-nearest neighbor search (k=5) in the case library; calculate The similarity between the spatial feature vectors of all historical cases in the case library is calculated by Euclidean distance and cosine similarity. The five most similar historical cases; historical cases are accumulated from previous installation practices. Their installation solutions and effects have been evaluated and verified, and they have important reference value for current new installation needs; Parameter fusion optimization: For the five most similar historical cases found, the system performs a weighted average of their technical parameters. The weight is calculated by multiplying the similarity by the comprehensive score. That is, the higher the similarity between each case and the current demand and the higher its own comprehensive score, the greater its weight in the weighted average. Suppose these five historical cases are ; They are related to The similarities are The comprehensive scores are , the technical parameters are ; Taking the battery pack arrangement as an example, assuming that its value is series or parallel, when performing weighted average calculation, first represent the series and parallel with numerical values ​​(such as series is 0 and parallel is 1), and calculate the weighted average value The formula is: ; Calculate each technical parameter by weighted average calculation, and finally generate a candidate parameter set ,in The arrangement of the battery pack, is the height above the ground, It is the inclination angle of the heat dissipation channel. The parameter combination with the highest weight after comprehensive consideration of similarity and comprehensive score is considered to be the most suitable parameter combination for the current new installation needs.

[0025] Virtual installation verification: Inject the generated candidate parameter set into the digital twin system for conflict detection; the digital twin system is a virtual model that is highly similar to the actual physical system, simulating various situations in the actual installation process of home energy storage equipment; during the conflict detection process, check whether there are problems such as pipeline collisions or safety spacing that does not meet regulatory requirements; for example, simulate the installation location and wiring of the equipment to see whether the pipeline will collide with walls, furniture or other equipment, and whether the safety distance between the equipment and surrounding objects meets relevant standards and specifications. If a problem is detected, parameter adjustment is automatically triggered; the adjustment method is based on preset rules and algorithms to adjust, recalculate and optimize the parameters until the installation requirements are met; Output optimized installation parameter package: After virtual installation verification and parameter adjustment, the system outputs an optimized installation parameter package. This parameter package contains the following information: Three-dimensional positioning coordinates: Provide accurate three-dimensional positioning coordinates of household energy storage devices in the installation space, with an accuracy of up to ±2mm. This ensures that the equipment can be accurately placed during the installation process, avoiding installation problems and subsequent inconvenience caused by position deviation; Recommended wiring scheme: Presented in the form of a color-coded path heat map, different colors represent different wiring paths, and the color depth of the heat map can reflect the priority or difficulty of wiring; This visual method makes it easy for installers to clearly understand the recommended wiring method and improve wiring efficiency and accuracy; Aesthetic optimization suggestions: Based on the aesthetic evaluation of the installation area and the matching analysis with the preset decoration style, specific aesthetic optimization suggestions are given. For example, "using a matte shell to improve the style matching by 12%" helps users improve the aesthetics of the installation effect while meeting the functional requirements of the device, so that it can better integrate into the home environment.

[0026] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to only specific implementation methods. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can understand and use the present invention well. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. An installation effect evaluation system for household energy storage equipment, characterized in that: include: Installation parameter acquisition module, used to use the equipment to collect 3D point cloud data, image features, technical parameters and environmental parameters of the installation space, and build a structured case database to store the data; A comprehensive evaluation module is used to calculate the aesthetic score and the utility score through aesthetic evaluation and utility evaluation, and then generate a comprehensive score to evaluate the installation effect of the home energy storage device; Optimize the case library construction module, screen high-quality cases, encode their features, build a case similarity map, and dynamically maintain the top ten case sets to provide reference for recommendations; The collaborative recommendation module receives new installation requirements, extracts spatial feature vectors, and outputs the optimized installation parameter package through k-nearest neighbor search, parameter fusion optimization, and virtual installation verification.

2. The installation effect evaluation system for household energy storage equipment according to claim 1, characterized in that: The execution process of the installation parameter acquisition module is as follows: Use laser scanners to obtain 3D point cloud data of the installation space and reconstruct the BIM model including wall structure and furniture layout; Use a wide-angle camera to capture a panoramic image of the installation area and extract HSV color space histogram features; Record technical parameter set: including battery pack arrangement, ground clearance, and heat dissipation channel inclination; Collect environmental parameters: Use a spectrophotometer to measure the reflectivity of wall materials and the reverberation time of the space; Construct a structured case database, where each case is stored as a tuple Case = {coordinates (x, y, z), technical parameters T, aesthetic features A, environmental parameters E}; Finally, semantic data annotation and retrieval are performed.

3. The installation effect evaluation system for household energy storage equipment according to claim 1, characterized in that: The specific steps for semantic data annotation and retrieval in the installation parameter acquisition module are as follows: Semantically annotate the stored case data and use the ontology model to define the semantic relationship between the data; When building a structured case database, efficient retrieval based on semantics is achieved; When users can query related cases through natural language descriptions, the search results are sorted according to the degree of match, and cases with high matching degrees are displayed first; Present the case to the user in a visual way, including: displaying key information of the case in list form, including installation coordinates, technical parameters, aesthetic features and environmental parameter summary.

4. The installation effect evaluation system for household energy storage equipment according to claim 1, characterized in that: The specific process of the comprehensive evaluation module is as follows: Aesthetic evaluation: The ResNet-50 model is used to extract deep features of the installation area image. The input installation area image is subjected to feature extraction through multiple convolutional layers, pooling layers, and fully connected layers. Calculate the cosine similarity with the preset decoration style template, assuming that the image feature vector of the installation area extracted from the ResNet-50 model is , the preset feature vector of a certain decoration style template is , then the cosine similarity calculation formula between them is: ;in, is a vector and The dot product of and They are vectors and The similarity value range is between [-1,1], and the similarity result is used as the AI ​​score, recorded as ; Integrating user subjective ratings, users can view the virtual effects of home energy storage devices in the installation area through the augmented reality interface and rate them from 1 to 5 stars based on their aesthetic feelings, which is recorded as ; Output aesthetic score ; Where 0.6 and 0.4 are weight coefficients; Utility evaluation: Calculate installation compactness: ; Measuring maintenance convenience: simulate the maintenance path length, and use the A* algorithm to calculate the shortest accessible path. When measuring maintenance convenience, the maintenance personnel's starting position is used as the starting point, and the part of the household energy storage device that needs to be repaired is used as the target point. The shortest accessible path length L calculated by the A* algorithm is the simulated maintenance path length; Evaluate thermal management performance: Use fluid dynamics simulation to calculate the uniformity of air velocity distribution in the heat dissipation channel to evaluate thermal management performance. In fluid dynamics simulation, by establishing a physical model of the heat dissipation channel, setting boundary conditions, and solving the fluid mechanics Navier-Stokes equation, the distribution of air velocity in the heat dissipation channel is obtained. Output utility score: First, standardize the installation compactness C, simulated maintenance path length L, and air velocity variation coefficient CV using the linear normalization method: , , ,in are the minimum and maximum values ​​of installation compactness in all evaluated cases, respectively; are the minimum and maximum values ​​of the simulated maintenance path lengths in all evaluated cases, respectively; are the minimum and maximum values ​​of the coefficient of variation of air velocity in all evaluated cases, respectively; They are the normalized values ​​of installation compactness C, simulated maintenance path length L, and coefficient of variation CV of air velocity, respectively; The utility score calculation formula is: ;in are the dynamic weights of installation compactness, simulated maintenance path length and air velocity variation coefficient, respectively; and ; Generate a composite score: ,in, and is the weight coefficient, and .

5. The installation effect evaluation system for household energy storage equipment according to claim 1, characterized in that: The specific operation steps of the preferred case library construction module are as follows: High-quality case screening: Continuously monitor the cases newly entered into the case database in real time; when each new case is entered into the database, it is judged based on the comprehensive score S given by the comprehensive evaluation module; Compare the comprehensive score S with the preset threshold θ; if S>θ, the case is marked as a high-quality case and enters the subsequent processing flow; If S≤θ, it will not enter the high-quality case-related processing, but will only be stored in the case library as an ordinary case for subsequent reference; Feature encoding: Dimensionality reduction of three-dimensional space coordinates. For high-quality cases, the t-SNE algorithm is used to process the three-dimensional space coordinates; the t-SNE algorithm maps high-dimensional data to low-dimensional space while maintaining the relative distance relationship between data points; One-hot encoding of technical parameters: One-hot encoding is used to process the technical parameters T of high-quality cases; Construct case similarity graph: Based on the encoded high-quality case features, calculate the Mahalanobis distance between the feature vectors of family scenarios; Set a distance threshold δ. When the calculated Mahalanobis distance d of two cases is less than δ, the two cases are judged to be similar. At this time, an edge connection is established for the two cases in the case similarity graph to associate them. By continuously calculating and comparing the Mahalanobis distances between all high-quality cases, a graph structure reflecting the case similarity relationship is gradually constructed, which facilitates the rapid search for similar cases. Dynamically maintain the top ten case collection: maintain a collection that only retains the top ten high-quality cases; use the LRU algorithm to manage the collection; Whenever a new high-quality case is added to the collection, when the collection has reached ten cases, the LRU algorithm will check the call time record of each case in the collection; eliminate the old cases that are more than 30 days away from the current time and have the lowest frequency of use within 30 days to make room for the new case; Through a dynamic maintenance mechanism, we ensure that the top ten case collections always contain the most valuable high-quality cases, providing case data support that is more in line with actual needs for subsequent collaborative recommendation functions.

6. The installation effect evaluation system for household energy storage equipment according to claim 1, characterized in that: The specific operation steps of the collaborative recommendation module are as follows: Receive new installation requirements and extract spatial feature vectors: When there is a new home energy storage device installation requirement, the system first extracts the spatial feature vector from the current home’s relevant data ;The data comes from the three-dimensional point cloud data of the installation space obtained by the laser scanner and the panoramic image of the installation area taken by the wide-angle camera; Perform k-nearest neighbor search: Take the extracted spatial feature vector Based on, perform k nearest neighbor search in the case library; calculate The similarity between the spatial feature vectors of all historical cases in the case library. The similarity calculation method includes Euclidean distance or cosine similarity. By calculation and comparison, find the similarity The five most similar historical cases; Parameter fusion optimization: For the five most similar historical cases found, their technical parameters are weighted averaged; the weight is calculated by multiplying the similarity by the comprehensive score. That is, the higher the similarity between each case and the current demand and the higher its own comprehensive score, the greater its weight in the weighted average; Suppose these five historical cases are ; Five historical cases and The similarities are The comprehensive scores are , the technical parameters are ; When analyzing the technical parameters of the battery pack arrangement, assume that the value is series or parallel. When performing weighted average calculation, first represent the series and parallel with numerical values, specifically 0 for series and 1 for parallel, and calculate the weighted average value. The formula is: ; Each technical parameter is calculated by weighted average calculation to finally generate a candidate parameter set. ,in The arrangement of the battery pack, is the height above the ground, It is the inclination angle of the heat dissipation channel, which is the parameter combination with the highest weight after considering the similarity and comprehensive score.

7. The installation effect evaluation system for household energy storage equipment according to claim 6, characterized in that: The collaborative recommendation module also includes: Virtual installation verification: Inject the generated candidate parameter sets into the digital twin system for conflict detection; During the conflict detection process, check whether there is a pipeline collision or the safety distance does not meet the specification requirements; When a problem is detected, parameter adjustment is automatically triggered; the adjustment method is based on preset rules and algorithms to adjust, recalculate and optimize the parameters until the installation requirements are met; Output optimized installation parameter package: After virtual installation verification and parameter adjustment, output optimized installation parameter package, which contains the following information: Three-dimensional positioning coordinates: Provide the three-dimensional positioning coordinates of the home energy storage device in the installation space; Recommended wiring scheme: presented in the form of a color-coded path heat map. Different colors represent different wiring paths. The color depth of the heat map reflects the priority or difficulty of the wiring. Aesthetic optimization suggestions: Based on the aesthetic evaluation of the installation area and the matching analysis with the preset decoration style, specific aesthetic optimization suggestions are given.

Images