CN117828378B - Digital intelligent green building design evaluation method - Google Patents
Digital intelligent green building design evaluation method Download PDFInfo
- Publication number
- CN117828378B CN117828378B CN202410240760.3A CN202410240760A CN117828378B CN 117828378 B CN117828378 B CN 117828378B CN 202410240760 A CN202410240760 A CN 202410240760A CN 117828378 B CN117828378 B CN 117828378B
- Authority
- CN
- China
- Prior art keywords
- clustering
- data point
- data
- domain
- data points
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000013461 design Methods 0.000 title claims abstract description 56
- 238000011156 evaluation Methods 0.000 title claims abstract description 24
- 238000000034 method Methods 0.000 claims abstract description 46
- 238000012216 screening Methods 0.000 claims abstract description 5
- 238000012163 sequencing technique Methods 0.000 claims description 4
- 238000000638 solvent extraction Methods 0.000 claims description 4
- 238000012545 processing Methods 0.000 abstract description 2
- 238000005265 energy consumption Methods 0.000 description 6
- 239000013598 vector Substances 0.000 description 5
- 238000010276 construction Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000004378 air conditioning Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000004566 building material Substances 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000010438 heat treatment Methods 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention relates to the technical field of data processing, in particular to a digital intelligent green building design evaluation method, which comprises the following steps: obtaining data points corresponding to a building, obtaining local radiuses according to the extremely poor of each item of data of the building and the number of the data points, obtaining the local density of each data point in a clustering domain according to the local radiuses in the process of clustering all the data points, further obtaining the referenceability of each data point, screening discrete data points, obtaining the comprehensive delocalization degree of each discrete data point relative to the nearest clustering domain, reclassifying the discrete data points to obtain outliers and new clustering domains, obtaining clustering results and real outliers through continuous iteration, and carrying out design rating on each building according to the clustering results and the real outliers. The invention eliminates the interference of the true outlier, the clustering result is more accurate, and the building design rating according to the clustering result is more accurate.
Description
Technical Field
The invention relates to the technical field of data processing, in particular to a digital intelligent green building design evaluation method.
Background
With the growing global interest in sustainable development and environmental protection, green construction has become an important trend in the construction industry. For the existing green building, the evaluation is directly carried out according to the green energy-saving data, and the subjectivity is unavoidable, so that a set of more perfect digital intelligent green building design evaluation method is needed.
The conventional intelligent green building design evaluation method carries out comprehensive evaluation rating on various indexes of the intelligent green building according to artificial subjective factors, the determined segmentation threshold is subjective, differences among various building designs are difficult to reflect, or building designs with large differences can be classified into the same evaluation grade, so that inaccuracy on the intelligent green building design evaluation is caused. Therefore, the data of each building needs to be clustered, so that the buildings with more similar data are clustered into the same type of clusters, and each type of clusters is subjected to building design evaluation.
The iterative self-organizing clustering algorithm can be used for clustering the intelligent green building data, but in the iterative self-organizing clustering algorithm, clustering domain division is carried out according to the distances between all data points and the centers of the clustering domains, some outliers are divided into the clustering domains, and under the influence of the discrete data points, the center position of the final clustering domain is caused to deviate, so that the accuracy of the clustering domain is poor, and the accuracy of building design evaluation is influenced.
Disclosure of Invention
In order to solve the problems, the invention provides a digital intelligent green building design evaluation method, which comprises the following steps:
collecting various data of intelligent green buildings, and obtaining data points corresponding to each building; obtaining the local radius of each data point according to the extreme difference of each data point of all buildings and the number of the data points;
in the process of clustering data points corresponding to all buildings by using an iterative self-organizing clustering algorithm, acquiring two new clustering centers of a clustering domain needing splitting for each clustering domain obtained by each iteration;
For a clustering domain needing splitting, obtaining the local density of each data point in the clustering domain according to the local radius of each data point, obtaining the referenceable degree of each data point according to the local density and the distance between each data point and two new clustering centers, and screening discrete data points according to the referenceable degree; dividing the rest data points except the discrete data points in the cluster domain into two new cluster domains according to the two new cluster centers;
Acquiring the comprehensive delocalization degree of each discrete data point relative to the nearest cluster domain, and dividing the discrete data points according to the comprehensive delocalization degree to obtain outliers and a new cluster domain; obtaining a clustering result and a real outlier through continuous iteration;
and carrying out design rating on each building according to the clustering result and the real outlier.
Preferably, the local radius of each data point is obtained according to the range of each data point and the number of the data points of all buildings, and the specific steps are as follows:
，
Where r represents the local radius of the data point, n represents the number of data points collected, M represents the total number of data items collected, The j-th item of data representing all buildings is extremely bad.
Preferably, the step of obtaining the local density of each data point in the cluster domain according to the local radius of each data point includes the following specific steps:
For each data point, acquiring the distance from the rest data points to the data point, and taking the data point with the distance smaller than the local radius of the data point as the data point in the local radius range of the data point; the number of other data points included in the local radius range of each data point is taken as the local density of each data point.
Preferably, the step of obtaining the referenceable degree of each data point includes the following specific steps:
，
wherein, Representing the referenceability of the ith data point of all data points, i get pass/>In, n represents the number of data points collected,/>Representing the local density of the ith data point,/>Representing the smallest Euclidean distance of the Euclidean distances of the ith data point to all cluster centers.
Preferably, the dividing the rest data points except the discrete data points in the cluster domain into two new cluster domains according to the two new cluster centers comprises the following specific steps:
For each data point except for discrete data points in the cluster domain, acquiring the distance from the data point to two new cluster centers, and dividing the data point into a new cluster domain to which the new cluster center closest to the data point belongs; two new cluster domains are obtained by partitioning all but the discrete data points in the cluster domain.
Preferably, the method for obtaining the nearest cluster domain comprises the following steps:
for each discrete data point, a cluster center closest to the discrete data point is acquired, and a cluster domain to which the cluster center belongs is used as the closest cluster domain of the discrete data point.
Preferably, the step of obtaining the comprehensive delocalization degree of each discrete data point relative to the nearest cluster domain includes the following specific steps:
，
wherein, The degree of delocalization of the jth item of data relative to its nearest cluster domain representing the kth discrete data point,Jth item data representing kth discrete data point,/>Average value of jth item data representing all data points in the nearest cluster domain of kth discrete data point,/>Standard deviation of the jth item of data representing all data points in the nearest cluster domain of the kth discrete data point;
For each discrete data point, multiplying all items of data of the discrete data point by the degree of delocalization of the nearest cluster domain, and obtaining the comprehensive degree of delocalization of the discrete data point relative to the nearest cluster domain.
Preferably, the discrete data points are divided according to the comprehensive delocalization degree to obtain outliers and new cluster domains, and the method comprises the following specific steps:
Incorporating discrete data points with the comprehensive delocalization degree smaller than a preset comprehensive delocalization degree threshold value into the nearest cluster domain of the discrete data points, and updating the cluster domain; and taking the discrete data points with the comprehensive delocalization degree larger than or equal to a preset comprehensive delocalization degree threshold value as outliers.
Preferably, the clustering result and the true outlier are obtained through continuous iteration, and the method comprises the following specific steps:
The outlier obtained by each iteration is still used as a normal data point to participate in the division of the data point again when the next iteration is performed, the iteration is stopped until the iteration stopping condition of the iterative self-organizing clustering algorithm is reached by continuous iteration, all clustering domains obtained when the iteration is stopped are used as clustering results, and the outlier obtained when the iteration is stopped is used as a real outlier.
Preferably, the step of grading the design of each building according to the clustering result and the real outlier comprises the following specific steps:
Acquiring a clustering center of each clustering domain in a clustering result, comparing green energy-saving levels of the clustering centers of different clustering domains in the clustering result, sequencing all the clustering centers according to the green energy-saving levels, respectively setting building design ratings for each clustering center according to the sequencing, and taking the building design ratings of the clustering centers as the building design ratings of the rest data points in the clustering domain; and for each real outlier, acquiring a cluster center with the closest distance from the real outlier, and taking the building design rating corresponding to the closest cluster center as the building design rating corresponding to the real outlier.
The technical scheme of the invention has the beneficial effects that: according to the method, the influence of the outliers on the clustering result is considered, in the clustering process, the referenceable degree of the data points is obtained according to the local density of the data points and the distance between the data points and the clustering center, the discrete data points are screened according to the referenceable degree, in the process of carrying out clustering domain division on the data points, the discrete data points do not participate in division, the discrete data points are re-divided according to the comprehensive delocalization degree of the discrete data points relative to the nearest clustering domain, the real outliers are obtained through continuous iteration, the interference of the real outliers on the clustering result is eliminated, the overall similar building clusters of all data are the same type of clusters, the clustering result is accurate, and the building design rating is accurate according to the clustering result.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of the steps of a method for evaluating a digital intelligent green building design according to the present invention.
Detailed Description
In order to further describe the technical means and effects adopted by the invention to achieve the preset aim, the following is a detailed description of a specific implementation, structure, characteristics and effects of a digital intelligent green building design evaluation method according to the invention with reference to the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" means that the embodiments are not necessarily the same. Furthermore, the particular features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
The following specifically describes a specific scheme of the digital intelligent green building design evaluation method provided by the invention with reference to the accompanying drawings.
Referring to fig. 1, a flowchart of a method for evaluating a digital intelligent green building design according to an embodiment of the invention is shown, the method includes the following steps:
S001, collecting intelligent green building data.
It should be noted that, through gathering intelligent green building related data, can acquire intelligent green building's related energy consumption data to follow-up energy consumption condition according to intelligent green building is scored the building, judges whether intelligent green building exists the direction that needs to improve, thereby carries out energy consumption structural optimization to the building in follow-up use and provides the reference in the intelligent green building design in future, avoids similar problem to appear once more.
In this embodiment, data of an intelligent green building is collected, including power consumption data and building indoor environment parameter data, such as power consumption peaks and troughs, lighting equipment energy consumption, lighting intensity, heating equipment and air conditioning equipment energy consumption, indoor temperature, capacity and utilization condition of renewable energy sources, building materials and structures, and the like.
It should be noted that, gather intelligent green building power consumption data, can directly carry out all kinds of energy consumption data's collection through ammeter and the energy monitoring system that installs on the building, gather building indoor environment parameter data, the indoor temperature sensor of installation of accessible building, humidity transducer, illumination sensor carry out the collection of building indoor environment parameter.
All the collected data of each building are quantized to form a feature vector of each building, and each dimension of the feature vector corresponds to each item of collected data respectively. The number of buildings is denoted as n, the total number of collected data items is denoted as M, and then the feature vector of each building is an M-dimensional vector, which can be expressed as a data point in an M-dimensional space. Since each building corresponds to one data point, the number of data points is also n.
Therefore, intelligent green building data acquisition is realized, and data points corresponding to each building are acquired.
S002, presetting iterative self-organizing clustering algorithm parameters.
It should be noted that, conventionally, for the intelligent green building design evaluation method, the score is obtained by comprehensively calculating according to the size of each item of data of the building, and the evaluation grade is divided according to the score threshold value, and the evaluation method is too subjective. Therefore, in the embodiment, a large amount of building data is clustered, buildings with similar indexes are divided into the same cluster domain in a clustering mode, and the buildings are subjected to evaluation grade division by comparing the energy saving conditions of the buildings in each cluster domain, so that a plurality of buildings with similar whole indexes are divided into the same evaluation grade, and more accurate evaluation grade division is realized.
It should be further noted that, in this embodiment, building data is clustered by using an iterative self-organizing clustering algorithm, and appropriate parameters are preset for the iterative self-organizing clustering algorithm, where the preset parameters include the number of expected clustering centers K, and the minimum number of samples in each clustering domainCluster domain sample distance distribution standard deviation/>Minimum distance between two clustering centers/>Iterative operation times/>. Because all preset parameters can influence the clustering result to a large extent, the parameters should be selected according to the collected building data.
Specifically, the number of expected cluster centers K is preset, the empirical value is k=5, the embodiment is not limited in particular, and the operator can set the number of expected cluster centers according to the data scale or priori knowledge in the specific implementation process. Dividing the number of data points n by 2K to obtain the minimum number of samples in the cluster domain. Obtaining the range of the same item of data of all buildings, dividing the average value of the range of all items by 2K to be used as the cluster domain sample distance distribution standard deviation/>Wherein/>Representing the extreme difference of the j-th data of all buildings, M is the total number of the collected data items. Obtaining the minimum distance/>, of two clustering centers according to the range of each term. And taking 2K as the iterative operation times.
It should be noted that, as the initial condition of iterative self-organizing clustering, the number of initial clustering centers should be presetAnd clustering center positions, wherein the initial clustering center number/>, because the local density of the data points is needed to be judged in the clustering domain splitting processSet to 1. And obtaining the average value of the same item of data of all buildings, taking the average value of the same item as a representative value of the item, and taking data points of feature vectors formed by the representative values of all items in an M-dimensional space as initial clustering centers.
Thus, the setting of the iterative self-organizing clustering algorithm parameters is realized.
S003, clustering all the corresponding data points of the building, screening discrete data points according to the local density of each data point, and dividing the discrete data points according to the comprehensive delocalization degree of the discrete data points relative to the nearest clustering domain to obtain a clustering result and real outliers.
In the process of clustering building data by the iterative self-organizing clustering algorithm, data points are required to be divided into new clustering domains again in each clustering domain splitting, and in the new clustering domain dividing process, all data points are divided into clustering domains according to the distance from the data points to the split clustering center by the conventional method. However, for building data, it is desirable to divide data points with smaller local density and smaller euclidean distance into the same cluster domain, and a conventional data point dividing method may divide some discrete data points into the cluster domain, and under the influence of the discrete data points, the center position in the final cluster domain is deviated. Therefore, in the embodiment, the local density of the data points is analyzed, the discrete data points are screened, other dividing methods are adopted for the discrete data points, and the influence of the discrete data points on the clustering center is eliminated.
Specifically, the iterative self-organizing clustering algorithm is utilized to cluster the data points corresponding to all the buildings, and in the clustering process, whether the clustering domains need to be split or not is judged for each clustering domain obtained by each iteration. It should be noted that, a specific method for determining whether the clustering domain needs to be split is the same as a method in the iterative self-organizing clustering algorithm, and will not be described in detail herein.
For a clustering domain needing splitting, acquiring standard deviation of each item of data of all data points in the clustering domain, wherein the maximum standard deviation is larger than the standard deviation of the sample distance distribution of the clustering domainData point splitting is carried out on the clustering domain of the clustering domain, the data item with the largest standard deviation is added and subtracted with half of the largest standard deviation respectively, and the data item and the numerical value of other data items of the original clustering center of the clustering domain form two new clustering centers respectively. It should be noted that, the method for obtaining two new cluster centers by splitting data points in the cluster domain is the same as the method in the iterative self-organizing clustering algorithm, and will not be described in detail here.
It should be noted that, after the original cluster center is split into two new cluster centers, the data points in the original cluster domain should be reassigned to the cluster domains corresponding to the two new cluster centers. However, when clustering the data points corresponding to all the buildings, the distance between every two data points in one clustering domain is expected to be smaller, so that the local density of each data point needs to be considered, the data points with smaller local density are excluded from the clustering domain, and the influence of the data points with smaller local density on the clustering center of the clustering domain is avoided.
Specifically, the local radius of each data point is obtained according to the extremely poor data of each building and the number of data points:
，
Where r represents the local radius of the data point, n represents the number of data points collected, M represents the total number of data items collected, i.e., the number of data items each data point contains, The j-th item of data representing all buildings is extremely bad. /(I)The maximum Euclidean distance achievable between all data points is reflected, divided by the number of data points, n, representing the radius of space each data point can occupy.
For each data point, the distance from the remaining data points to the data point is obtained, and the data point with the distance smaller than the local radius of the data point is taken as the data point within the local radius range of the data point. The larger the number of other data points included in the local radius range of any one data point, the larger the local density of the data point is. Conversely, the fewer the number of other data points contained within the local radius of any one data point, the more discrete the data point distribution is illustrated.
The number of other data points included in the local radius range of each data point is taken as the local density of each data point.
It should be noted that, when assigning data points to a cluster domain, some data points with smaller local density and farther from any cluster center should be used as discrete data points, and for discrete data points, they are not assigned to any cluster domain, so that the local density of the data points and the euclidean distance from the cluster center are comprehensively considered, and the referenceof each data point is obtained, so as to measure the possibility that each data point is a discrete data point:
，
wherein, Representing the referenceability of the ith data point of all data points, i get pass/>In, n represents the number of data points collected,/>Representing the local density of the ith data point,/>Representing the smallest Euclidean distance of the Euclidean distances of the ith data point to all cluster centers. The lower the referenceable data point is the more likely it is a discrete data point, which is not included in any cluster domain.
Specifically, a referenceable threshold is presetEmpirical value is/>The embodiment does not limit the referent threshold, and an operator can set the referent threshold according to the actual implementation situation. Data points with a referenceability greater than the referenceability threshold are noted as discrete data points.
For each discrete data point, the discrete data point is not divided into any cluster domain in the process of dividing the data point into the post-split cluster domains. And dividing the data points into the new cluster domains which are closest to the clustering center according to the distance from the data points to the two new clustering centers after splitting, and dividing all the rest data points in the original cluster domains to obtain two new cluster domains. It should be noted that, the partitioning method for the rest of the data points in the original clustering domain is the same as the partitioning method in the iterative self-organizing clustering algorithm, which is a known technology and will not be described in detail herein.
It should be noted that, the discrete data points are obtained only according to the local density and the distance from the data points to the clustering center, so that the obtaining method is ideal, and some discrete data points may be obtained inaccurately. Since a new cluster domain has been generated after the discrete data points are excluded, the present embodiment performs a secondary screening on the discrete data points by correlation between each discrete data point and each cluster domain, and repartitions the discrete data points having a larger correlation with the cluster domain into the cluster domain.
Specifically, for any discrete data point, a cluster center closest to the discrete data point is acquired, and a cluster domain to which the cluster center belongs is used as the closest cluster domain of the discrete data point. Acquiring the degree of delocalization of each item of data of the discrete data points relative to the nearest clustering domain according to each item of data of all data points in the nearest clustering domain and the corresponding item of data of the discrete data points:
，
wherein, The degree of delocalization of the jth item of data relative to its nearest cluster domain representing the kth discrete data point,Jth item data representing kth discrete data point,/>Average value of jth item data representing all data points in the nearest cluster domain of kth discrete data point,/>Standard deviation of the jth term data representing all data points in the nearest cluster domain of the kth discrete data point. /(I)The difference between the jth data of the kth discrete data point and the average value of the jth data of all data points in the nearest cluster domain reflects the integral difference between the kth discrete data point and the jth data in the nearest cluster domain, the integral difference is compared with the standard deviation of the jth data in the nearest cluster domain, and the larger the ratio is, the smaller the correlation between the jth data of the kth discrete data point and the nearest cluster domain is, and at the moment, the discrete data point cannot be included in the nearest cluster domain. Conversely, when the ratio of the overall gap to the standard deviation of the jth data in the nearest cluster domain is smaller, the j-th data of the kth discrete data point has larger correlation with the nearest cluster domain, and if the correlation between the rest of the kth discrete data point and the nearest cluster domain is also larger, the discrete data point should be included in the nearest cluster domain.
It should be noted that, in order to analyze whether the discrete data points can be added into the nearest cluster domain, the overall degree of delocalization of all items of data of the discrete data points relative to the nearest cluster domain should be considered.
Specifically, the comprehensive delocalization degree of each discrete data point relative to the nearest clustering domain is obtained according to the delocalization degree of each piece of data of each discrete data point relative to the nearest clustering domain.
，
Wherein,Representing the combined degree of delocalization of the kth discrete data point relative to its nearest cluster domain, M represents the total number of data items collected, i.e., the number of data items each data point contains,/>The degree of delocalization of the jth item of data relative to its nearest cluster domain, representing the kth discrete data point.
Presetting a comprehensive delocalization degree thresholdEmpirical value is/>The embodiment does not limit the comprehensive delocalization degree threshold, and an implementation person can set the comprehensive delocalization degree threshold according to the actual implementation condition.
Specifically, the discrete data points with the comprehensive delocalization degree smaller than the comprehensive delocalization degree threshold are included in the nearest cluster domain corresponding to the discrete data points, and the repartition of the discrete data points is completed. Discrete data points which are not included in any cluster domain after repartitioning are taken as outliers, namely, the discrete data points with the comprehensive delocalization degree being greater than or equal to a preset comprehensive delocalization degree threshold value are taken as outliers.
It should be noted that, in the iterative self-organizing clustering algorithm, the separation and merging of multiple clustering domains are involved, each clustering domain division involves the clustering domain division of data points, in the process of each data point clustering domain division, besides the data points in the original clustering domain, whether the existing outliers can be divided into the new clustering domain after division should also be considered, that is, the outliers obtained in the previous iteration still serve as normal data points to participate in the data point division by the method in the next iteration, the iteration is continued until the iteration stopping condition of the iterative self-organizing clustering algorithm is reached, all the clustering domains obtained in the iteration stopping process serve as clustering results, and the outliers obtained in the iteration stopping process serve as real outliers.
Thus, classification of building data is achieved.
S004, carrying out design rating on each building according to the clustering result and the real outlier.
It should be noted that after the iterative self-organizing clustering algorithm is used to classify the data points corresponding to all the buildings, all the data of the data points in the same cluster domain are similar in level, and the building data of each cluster domain can be evaluated easily to realize the building design grade evaluation.
Specifically, the clustering centers of each clustering domain in the clustering result are obtained, the green energy-saving levels of the clustering centers of different clustering domains in the clustering result are compared, all the clustering centers are ranked according to the green energy-saving levels, the building design rating is set for each clustering center according to the ranking, and the building design rating of the clustering center is used as the building design rating of each data point in the clustering domain. It should be noted that, the green energy-saving level of the clustering center may be assessed by a professional according to professional knowledge, and the method for assessing the green energy-saving level in this embodiment is not limited.
And for each real outlier, acquiring a cluster center, which is closest to the real outlier, of the real outlier, and taking the building design rating corresponding to the closest cluster center as the building design rating corresponding to the real outlier.
After the new building design is generated, the building design can be rated directly according to the building design rating of the cluster center with the closest data point corresponding to the building when the new building design is evaluated.
It should be noted that, the clustering center obtained by the traditional method is affected by outliers, and has offset, and the building design rating of the clustering center is used as the building design rating of other data points in the clustering domain.
Through the steps, green building design evaluation is completed.
According to the embodiment of the invention, the influence of the outlier on the clustering result is considered, in the clustering process, the referenceable degree of the data points is obtained according to the local density of the data points and the distance from the data to the clustering center, the discrete data points are screened according to the referenceable degree, in the process of carrying out clustering domain division on the data points, the discrete data points do not participate in division, the discrete data points are re-divided according to the comprehensive delocalization degree of the discrete data points relative to the nearest clustering domain, the real outlier is obtained through continuous iteration, the interference of the real outlier on the clustering result is eliminated, so that the overall more similar building clusters of all data are in the same type of cluster, the clustering result is more accurate, and the building design rating is more accurate according to the clustering result.
The above description is only of the preferred embodiments of the present invention and is not intended to limit the invention, but any modifications, equivalent substitutions, improvements, etc. within the principles of the present invention should be included in the scope of the present invention.
Claims (6)
1. The digital intelligent green building design evaluation method is characterized by comprising the following steps of:
collecting various data of intelligent green buildings, including power consumption data and building indoor environment parameter data, and obtaining data points corresponding to each building; obtaining the local radius of each data point according to the extreme difference of each data point of all buildings and the number of the data points;
in the process of clustering data points corresponding to all buildings by using an iterative self-organizing clustering algorithm, acquiring two new clustering centers of a clustering domain needing splitting for each clustering domain obtained by each iteration;
For a clustering domain needing splitting, obtaining the local density of each data point in the clustering domain according to the local radius of each data point, obtaining the referenceable degree of each data point according to the local density and the distance between each data point and two new clustering centers, and screening discrete data points according to the referenceable degree; dividing the rest data points except the discrete data points in the cluster domain into two new cluster domains according to the two new cluster centers;
Acquiring the comprehensive delocalization degree of each discrete data point relative to the nearest cluster domain, and dividing the discrete data points according to the comprehensive delocalization degree to obtain outliers and a new cluster domain; obtaining a clustering result and a real outlier through continuous iteration;
carrying out design rating on each building according to the clustering result and the real outlier;
the design rating of each building is carried out according to the clustering result and the real outlier, and the method comprises the following specific steps:
Acquiring a clustering center of each clustering domain in a clustering result, comparing green energy-saving levels of the clustering centers of different clustering domains in the clustering result, sequencing all the clustering centers according to the green energy-saving levels, respectively setting building design ratings for each clustering center according to the sequencing, and taking the building design ratings of the clustering centers as the building design ratings of the rest data points in the clustering domain; for each real outlier, acquiring a cluster center with the closest distance from the real outlier, and taking the building design rating corresponding to the closest cluster center as the building design rating corresponding to the real outlier;
The method for obtaining the local radius of each data point according to the extreme difference of each data point of all buildings and the number of the data points comprises the following specific steps:
，
Where r represents the local radius of the data point, n represents the number of data points collected, M represents the total number of data items collected, The very poor of the j-th item of data representing all buildings;
the method for acquiring the referenceable degree of each data point comprises the following specific steps:
，
wherein, Representing the referenceability of the ith data point of all data points, i get pass/>In, n represents the number of data points collected,/>Representing the local density of the ith data point,/>Representing the minimum Euclidean distance of the Euclidean distances of the ith data point to all cluster centers;
The method for obtaining the comprehensive delocalization degree of each discrete data point relative to the nearest cluster domain comprises the following specific steps:
，
wherein, The degree of delocalization of the jth item of data relative to its nearest cluster domain, representing the kth discrete data point,/>Jth item data representing kth discrete data point,/>Average value of jth item data representing all data points in the nearest cluster domain of kth discrete data point,/>Standard deviation of the jth item of data representing all data points in the nearest cluster domain of the kth discrete data point;
For each discrete data point, multiplying all items of data of the discrete data point by the degree of delocalization of the nearest cluster domain, and obtaining the comprehensive degree of delocalization of the discrete data point relative to the nearest cluster domain.
2. The method for evaluating the digital intelligent green building design according to claim 1, wherein the step of obtaining the local density of each data point in the cluster according to the local radius of each data point comprises the following specific steps:
For each data point, acquiring the distance from the rest data points to the data point, and taking the data point with the distance smaller than the local radius of the data point as the data point in the local radius range of the data point; the number of other data points included in the local radius range of each data point is taken as the local density of each data point.
3. The method for evaluating the digital intelligent green building design according to claim 1, wherein the dividing the rest of the data points except the discrete data points in the cluster domain into two new cluster domains according to the two new cluster centers comprises the following specific steps:
For each data point except for discrete data points in the cluster domain, acquiring the distance from the data point to two new cluster centers, and dividing the data point into a new cluster domain to which the new cluster center closest to the data point belongs; two new cluster domains are obtained by partitioning all but the discrete data points in the cluster domain.
4. The method for evaluating the digital intelligent green building design according to claim 1, wherein the method for acquiring the nearest cluster domain is as follows:
for each discrete data point, a cluster center closest to the discrete data point is acquired, and a cluster domain to which the cluster center belongs is used as the closest cluster domain of the discrete data point.
5. The method for evaluating the digital intelligent green building design according to claim 1, wherein the step of dividing the discrete data points according to the comprehensive delocalization degree to obtain the outliers and the new clustering domain comprises the following specific steps:
Incorporating discrete data points with the comprehensive delocalization degree smaller than a preset comprehensive delocalization degree threshold value into the nearest cluster domain of the discrete data points, and updating the cluster domain; and taking the discrete data points with the comprehensive delocalization degree larger than or equal to a preset comprehensive delocalization degree threshold value as outliers.
6. The method for evaluating the digital intelligent green building design according to claim 1, wherein the clustering result and the true outlier are obtained through continuous iteration, and the method comprises the following specific steps:
The outlier obtained by each iteration is still used as a normal data point to participate in the division of the data point again when the next iteration is performed, the iteration is stopped until the iteration stopping condition of the iterative self-organizing clustering algorithm is reached by continuous iteration, all clustering domains obtained when the iteration is stopped are used as clustering results, and the outlier obtained when the iteration is stopped is used as a real outlier.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410240760.3A CN117828378B (en) | 2024-03-04 | 2024-03-04 | Digital intelligent green building design evaluation method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410240760.3A CN117828378B (en) | 2024-03-04 | 2024-03-04 | Digital intelligent green building design evaluation method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117828378A CN117828378A (en) | 2024-04-05 |
| CN117828378B true CN117828378B (en) | 2024-05-17 |
Family
ID=90523194
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410240760.3A Active CN117828378B (en) | 2024-03-04 | 2024-03-04 | Digital intelligent green building design evaluation method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117828378B (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2016031546A (en) * | 2014-07-25 | 2016-03-07 | 清水建設株式会社 | Green space evaluation system and green space evaluation method |
| CN109102028A (en) * | 2018-08-20 | 2018-12-28 | 南京邮电大学 | Based on improved fast density peak value cluster and LOF outlier detection algorithm |
| CN109858832A (en) * | 2019-02-25 | 2019-06-07 | 合肥工业大学 | A kind of more attribute green Index grading Coordination Evaluation methods |
| CN110533111A (en) * | 2019-09-03 | 2019-12-03 | 西南交通大学 | A kind of adaptive K mean cluster method based on local density Yu ball Hash |
| CN114818936A (en) * | 2022-04-29 | 2022-07-29 | 山东大学 | Retired battery rapid comprehensive sorting method and system based on K-means clustering algorithm |
| CN116167668A (en) * | 2023-04-26 | 2023-05-26 | 山东金至尊装饰工程有限公司 | BIM-based green energy-saving building construction quality evaluation method and system |
| CN117435939A (en) * | 2023-12-14 | 2024-01-23 | 广东力宏微电子有限公司 | IGBT health state evaluation method based on big data |
-
2024
- 2024-03-04 CN CN202410240760.3A patent/CN117828378B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2016031546A (en) * | 2014-07-25 | 2016-03-07 | 清水建設株式会社 | Green space evaluation system and green space evaluation method |
| CN109102028A (en) * | 2018-08-20 | 2018-12-28 | 南京邮电大学 | Based on improved fast density peak value cluster and LOF outlier detection algorithm |
| CN109858832A (en) * | 2019-02-25 | 2019-06-07 | 合肥工业大学 | A kind of more attribute green Index grading Coordination Evaluation methods |
| CN110533111A (en) * | 2019-09-03 | 2019-12-03 | 西南交通大学 | A kind of adaptive K mean cluster method based on local density Yu ball Hash |
| CN114818936A (en) * | 2022-04-29 | 2022-07-29 | 山东大学 | Retired battery rapid comprehensive sorting method and system based on K-means clustering algorithm |
| CN116167668A (en) * | 2023-04-26 | 2023-05-26 | 山东金至尊装饰工程有限公司 | BIM-based green energy-saving building construction quality evaluation method and system |
| CN117435939A (en) * | 2023-12-14 | 2024-01-23 | 广东力宏微电子有限公司 | IGBT health state evaluation method based on big data |
Non-Patent Citations (3)
| Title |
|---|
| Improved ISODATA Clustering Method with Parameter Estimation based on Genetic Algorithm;Kohei Arai等;International Journal of Advanced Computer Science and Applications;20221231;第13卷(第5期);全文 * |
| TM影像分类算法比较与评价;杨洁;魏华锋;刘士文;;黑龙江工程学院学报;20150225(01);全文 * |
| 离群数据挖掘方法研究;蔡江辉, 张华煜;电脑开发与应用;20051230(12);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117828378A (en) | 2024-04-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110991786B (en) | 10kV static load model parameter identification method based on similar daily load curve | |
| CN107730031B (en) | Ultra-short-term peak load prediction method and system | |
| CN110458340B (en) | Building air conditioner cold load autoregressive prediction method based on mode classification | |
| CN110070282B (en) | Low-voltage transformer area line loss influence factor analysis method based on comprehensive relevance | |
| CN111105104A (en) | Short-term power load prediction method based on similar day and RBF neural network | |
| CN109886465B (en) | Power distribution network load prediction method based on intelligent electric meter user cluster analysis | |
| CN110348488B (en) | Modal identification method based on local density peak clustering | |
| CN112926795A (en) | SBO (statistical analysis) -based CNN (continuous casting) optimization-based high-rise residential building group heat load prediction method and system | |
| CN111291822A (en) | Equipment running state judgment method based on fuzzy clustering optimal k value selection algorithm | |
| CN113139880A (en) | Wind turbine generator actual power curve fitting method, device, equipment and storage medium | |
| CN114611842B (en) | Whole-county roof distributed photovoltaic power prediction method | |
| CN115018200A (en) | Power load prediction method and system based on deep learning and considering multiple influence factors | |
| CN115081933A (en) | Low-voltage user topology construction method and system based on improved spectral clustering | |
| CN117828378B (en) | Digital intelligent green building design evaluation method | |
| CN111027841A (en) | Low-voltage transformer area line loss calculation method based on gradient lifting decision tree | |
| CN116826745B (en) | Layered and partitioned short-term load prediction method and system in power system background | |
| CN111553434A (en) | Power system load classification method and system | |
| CN109064353B (en) | Large building user behavior analysis method based on improved cluster fusion | |
| CN117043794A (en) | Building energy consumption prediction method and system based on multiple linear regression and cluster analysis | |
| CN115374991A (en) | Improved statistics upscale-based photovoltaic cluster output prediction method | |
| CN110991743B (en) | Wind power short-term combination prediction method based on cluster analysis and neural network optimization | |
| CN110781538B (en) | Windowing simulation algorithm based on field monitoring | |
| CN112215297A (en) | Production and manufacturing data hierarchical clustering method based on factor analysis | |
| CN113487080B (en) | Wind speed dynamic scene generation method, system and terminal based on wind speed classification | |
| CN112633622B (en) | Smart power grid operation index screening method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |