CN112434361A - Intelligent ventilation system suitable for office buildings and data processing method - Google Patents
Intelligent ventilation system suitable for office buildings and data processing method Download PDFInfo
- Publication number
- CN112434361A CN112434361A CN202011280986.4A CN202011280986A CN112434361A CN 112434361 A CN112434361 A CN 112434361A CN 202011280986 A CN202011280986 A CN 202011280986A CN 112434361 A CN112434361 A CN 112434361A
- Authority
- CN
- China
- Prior art keywords
- dimensional
- low
- sub
- error
- air
- 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.)
- Pending
Links
- 238000009423 ventilation Methods 0.000 title claims abstract description 27
- 238000003672 processing method Methods 0.000 title claims abstract description 23
- 238000006243 chemical reaction Methods 0.000 claims abstract description 37
- 238000004088 simulation Methods 0.000 claims abstract description 30
- 230000009467 reduction Effects 0.000 claims abstract description 23
- 238000004378 air conditioning Methods 0.000 claims abstract description 22
- 238000004364 calculation method Methods 0.000 claims abstract description 21
- 230000003044 adaptive effect Effects 0.000 claims abstract description 15
- 238000000034 method Methods 0.000 claims description 51
- 230000008569 process Effects 0.000 claims description 17
- 230000009466 transformation Effects 0.000 claims description 15
- 238000013473 artificial intelligence Methods 0.000 claims description 9
- 238000011156 evaluation Methods 0.000 claims description 9
- 230000006798 recombination Effects 0.000 claims description 8
- 238000005215 recombination Methods 0.000 claims description 8
- 238000000638 solvent extraction Methods 0.000 claims description 8
- 230000011218 segmentation Effects 0.000 claims description 7
- 238000004422 calculation algorithm Methods 0.000 claims description 6
- 230000008859 change Effects 0.000 claims description 5
- 238000012545 processing Methods 0.000 claims description 5
- 230000010354 integration Effects 0.000 claims description 4
- 238000003491 array Methods 0.000 claims description 3
- 238000013519 translation Methods 0.000 claims description 3
- 108020001568 subdomains Proteins 0.000 claims description 2
- 238000004891 communication Methods 0.000 claims 1
- 238000011160 research Methods 0.000 description 5
- 238000013528 artificial neural network Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 201000010099 disease Diseases 0.000 description 2
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 2
- 239000006185 dispersion Substances 0.000 description 2
- 239000003344 environmental pollutant Substances 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 231100000719 pollutant Toxicity 0.000 description 2
- 238000012216 screening Methods 0.000 description 2
- 238000000844 transformation Methods 0.000 description 2
- 238000012935 Averaging Methods 0.000 description 1
- 208000024172 Cardiovascular disease Diseases 0.000 description 1
- 231100000597 Sick building syndrome Toxicity 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 239000012530 fluid Substances 0.000 description 1
- 230000006870 function Effects 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
- 230000008521 reorganization Effects 0.000 description 1
- 208000023504 respiratory system disease Diseases 0.000 description 1
- 230000033764 rhythmic process Effects 0.000 description 1
- 208000008842 sick building syndrome Diseases 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/10—Geometric CAD
- G06F30/13—Architectural design, e.g. computer-aided architectural design [CAAD] related to design of buildings, bridges, landscapes, production plants or roads
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/23—Design optimisation, verification or simulation using finite element methods [FEM] or finite difference methods [FDM]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/28—Design optimisation, verification or simulation using fluid dynamics, e.g. using Navier-Stokes equations or computational fluid dynamics [CFD]
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Geometry (AREA)
- General Physics & Mathematics (AREA)
- Computer Hardware Design (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Analysis (AREA)
- Pure & Applied Mathematics (AREA)
- Mathematical Optimization (AREA)
- Structural Engineering (AREA)
- Computational Mathematics (AREA)
- Civil Engineering (AREA)
- Architecture (AREA)
- Algebra (AREA)
- Computing Systems (AREA)
- Fluid Mechanics (AREA)
- Mathematical Physics (AREA)
- Air Conditioning Control Device (AREA)
Abstract
The invention discloses an intelligent ventilation system suitable for office buildings and a data processing method, wherein the ventilation system comprises an air supply unit arranged in the upper area of a wall on one side of a building, the air supply unit comprises a strip-seam type air supply outlet and an air supply pipeline arranged on one side of the air supply outlet, an air conditioning unit arranged in the outer area of the top layer of the building, and a return air unit arranged in the upper area of the wall on the other side of the building, the return air unit comprises a strip-seam type return air inlet and a return air pipeline arranged on one side of the return air inlet, and the air conditioning unit is respectively connected with the air supply unit, the return air unit and a fresh air unit to; the data processing method comprises the following steps: an input and output part of the calculation domain is used for reading an original CFD simulation result and outputting the original CFD simulation result in a dimensionality reduction mode; a physical field volume average conversion and error calculation part; and a low-dimensional calculation domain division and adaptive updating part. The data processing method is based on open source numerical simulation software OpenFOAM, and has strong expandability and transportability.
Description
Technical Field
The invention relates to the field of office building ventilation systems, in particular to an intelligent ventilation system and a data processing method suitable for office buildings.
Background
People are influenced by current working and living habits, the time in a building room accounts for about 90 percent, especially for young people, the time in an office building is far longer than that in a residential building due to the influence of the rhythm of life. Research shows that building-related diseases, sick building syndrome and other diseases are easily caused in uncomfortable buildings for a long time, and when the indoor air quality of the buildings is poor, the incidence rate of respiratory diseases, cardiovascular diseases and the like of indoor personnel is increased. Therefore, there is an increasing demand for indoor air quality and thermal comfort in office buildings.
At present, researches on indoor thermal comfort and indoor air quality of office buildings mainly depend on experiments and numerical simulation methods, wherein the numerical simulation methods are widely applied to effective researches on environmental parameters such as wind speed, temperature and pollutant concentration of the office buildings. However, due to the high resolution data and time consuming calculations of CFD (computational fluid dynamics) simulations, the current numerical simulation techniques still face the problems of "inefficiency" and "high cost" in building the initial database of fast predictive models. Therefore, the combination of CFD and Artificial Neural Networks (ANN) offers an incentive to address this problem. Cao and Do propose a fast prediction model using linear low-dimensional ventilation model (LLVM) and Artificial Neural Network (ANN) prediction, i.e., a case study that incorporates CFD into artificial intelligence for fast prediction. LLVM is able to create new concentration fields from known fields by dimension reduction methods and process high resolution CFD data to "low dimension" levels, which can effectively reduce computation time and data volume.
However, there is a limitation in applying low dimensional data in consideration of the dilemma of the prediction accuracy and speed of the ventilation system. Inconsistencies between the low dimensional data and the actual monitored data may result in incorrect estimates of real time conditions, which may affect the immediate decision of ventilation control. On the one hand, such inconsistencies may be caused by simulation errors and monitoring errors, which can be solved by applying more accurate, more advanced simulation and sensor techniques. On the other hand, even with accurate simulation results, conversion errors between CFD data and LLVM data cannot be ignored, which is a major concern in current work. This conversion error will decrease as the number of low dimensional regions increases (low dimensional region refers to the integration of data in a sparse grid obtained from high resolution CFD simulation). For office buildings, using several low dimensional regions to represent high resolution data with acceptable conversion errors can greatly solve the problems of low efficiency and high cost. Therefore, in view of the limited number of accurate transformations of the size transformation, it is very necessary and important to improve the low-dimensional model (LVM).
Disclosure of Invention
The invention aims to provide an intelligent ventilation system suitable for office buildings and a data processing method, wherein the data processing method is based on open source numerical simulation software OpenFOAM and has strong expandability and transportability; the low-dimensional process is completely integrated and is directly butted with a CFD simulation result; the non-uniform method can effectively improve the low-dimensional conversion precision under the limited low-dimensional subdomain quantity, and the screening process of the non-uniform method occupies less memory; a low-dimensional error index considering 'volume contribution rate' is provided, and the conversion condition of the low-dimensional process can be more accurately reflected; the indoor air quality and the thermal comfort of the office building are improved, an intelligent data processing method is provided, and the work efficiency is greatly saved for the follow-up research of the office building.
The purpose of the invention can be realized by the following technical scheme:
the utility model provides an intelligent ventilation system suitable for official working building, ventilation system are located the building, set up the official working region in the building, ventilation system includes air conditioning unit, the last intercommunication of air conditioning unit is equipped with air supply unit, return air unit and new trend unit, air supply unit and return air unit are located air conditioning unit's both sides respectively, air supply unit includes the supply-air duct, the one end and the air conditioning unit intercommunication of supply-air duct, the other end extends to and is equipped with strip seam type supply-air outlet in the building, the inside fan subassembly that is equipped with of air conditioning unit, the return air unit includes the return air pipeline, the one end and the air conditioning unit intercommunication of return air pipeline, the other end extends to and is equipped with strip seam type return-air inlet in the building, air conditioning unit and air supply unit, return.
A data processing method of an intelligent ventilation system suitable for office buildings, the data processing method comprising:
s1: an input and output part of the calculation domain is used for reading an original CFD simulation result and outputting the original CFD simulation result in a dimensionality reduction mode;
s2: a physical field volume average conversion and error calculation part;
s3: and a low-dimensional calculation domain division and adaptive updating part.
Further, the S1 includes:
input and output of the computational domain: reading and outputting original CFD simulation results, software user settings and dimension reduction by using numerical simulation open source software OpenFOAM and a dimension reduction tool;
computing a domain input: directly reading a CFD result of the OpenFOAM simulation example, wherein the CFD result comprises grid and physical field data and user setting data required by software operation;
computing domain output: the computationally obtained low-dimensional translation results, including the low-dimensional grid and physical field data, are exported to subdirectories of the OpenFOAM algorithm.
Further, the S2 includes:
s21: volume average conversion of physical field:
n volumes of total volume omega are added to form omegaiIs divided into N volumes of omega by' cut points PcA grid of sub-regions of;
when the position of the cutting point P is determined, the division of the subdomain is determined, and the number of subdomain grids is expressed as N-Nx×Ny×NzWhere the subscripts x, y, z denote the number of subfields per direction, and the number of cut points P per direction is Nj+1(j ═ x, y, z) comprising a start point and an end point;
after the sub-domain division is finished, grid information corresponding to the (x, y, z) coordinates of the sub-domain grid position is recorded into the array artificial intelligence;
according to the array artificial intelligence, the physical field values C corresponding to the grids corresponding to the (x, y, z) coordinates are also sequentially divided into corresponding arrays, and then the volume average value of all data in each sub-domain is calculated based on the grid data in each array artificial intelligence and is expressed as<C>c；
The volume average data result is used for expressing all grid data results of the corresponding sub-domains, and the discrete processing of high-precision grid data is realized;
s22: and (3) error calculation:
the performance of low-dimensional conversion is evaluated by adopting a conversion error epsilon, the volume influence of the grid must be considered by using a finite volume method, otherwise, the error calculation is inaccurate, and particularly, when the volume change of a single grid of the CFD grid is large, the method is very common in practical simulation;
the transformation error ε is defined as follows:
the grid volume ratio is introduced into the equation so that it represents more accurately the actual contribution of the individual grids to the low-dimensional conversion, the volume mean values are distributed uniformly over the area, the volume consists of the respective grid volumes which are met, and the difference between the low-dimensional field and the CFD field is quantified using an oa.
Further, c is 1-N, i is 1-N, and N < < N.
Further, the transformation error ε is used to represent the low-dimensional dispersion performance of a specific region by changing the volume limit Ω in the above equation to ΩcA regional error index is obtained that can represent the performance of the conversion error epsilon in the sub-region:
calculating a transformation error variance sigma according to the regional error index, wherein the transformation error variance sigma is used as a reference for analyzing error distribution in the whole volume when different low-dimensional magnitudes and dividing methods are analyzed;
the low-dimensional discrete error evaluation is used for evaluating the number range of the regions to be divided, the dimension reduction processing is also an important parameter, and an automatic and self-adaptive low-dimensional method is realized through a self-adaptive updating module according to the error index.
Further, the S3 includes:
s31: a non-uniform partitioning module:
the dividing method is to determine the tangent point P setting of each direction, and the tangent point P setting is N in each directionjAn array of +1(j ═ x, y, z), specifying boundaries of low-dimensional regions in one dimension, the boundaries being uniformly distributed by default; increasing the number of sub-fields, reducing the deviation between the calculated low-dimensional field and the original CFD field, and controlling the error within a certain range; a non-uniform division method is adopted as an important component of a self-adaptive low-dimensional method to improve the precision of low-dimensional conversion;
determining the tangent point P by combining an accurate sublayer recombination algorithm, realizing the non-uniform dimensionality reduction of a computational domain,
s311: generation of sublayers: dividing the whole calculation domain omega into N in each direction (X, Y, Z) respectivelySubIndividual sub-layers, wherein the sub-layers in one direction are formed by equal thickness criteria lSubGenerating, then calculating therein, a volume-averaged scalar value C for each sub-layerSub；
S312: recombination and selection of sublayers: the sub-layers are directionally recombined according to a vector passing through the average scalar field value CSubSelecting the combination with the minimum recombination error according to the sub-layer volume, and converting the sub-layer into a one-way segmentation mode after obtaining the optimal combination;
s313: the integration of the unidirectional segmentation modes integrates the unidirectional segmentation modes of each direction to construct a non-uniform low-dimensional field over the entire computational domain;
s32, self-adaptive updating:
the self-adaptive updating module determines whether to increase the number N of low-dimensional subdomains after low-dimensional conversion in different modesjJ is x, y, z; the adaptive CFD data dimension reduction tool provides two dimension reduction modes, called direct and automatic/adaptive respectivelyImplemented through common inheritance in C + +, which means that the "automatic" mode is a high-level "direct" mode; with the difference that the "direct" mode does not increase the number of low-dimensional subfields Nj(ii) a It will accomplish the low dimensional conversion directly for a given number of low dimensional subdomains and partitioning mode, while for the automatic mode, additional evaluation and update procedures will be performed;
the evaluation result in automatic mode is given according to the conversion error epsilon and another index 'Iter' recording the number of update iterations; the user has specified a conversion error limit value oalimAnd iteration limit Iterlim(ii) a Furthermore, if oa is givenlimGreater than or equal to lim and Iter < IterlimThen the number of low-dimensional subfields N will be updatedj；
NjThe update procedure for (j ═ x, y, z) can be expressed as follows:
Nj-newis the updated number of low-dimensional subdomains; n is a radical ofj-newIs the initial value of the number of low-dimensional subdomains given by the user in the first iteration; α is an "update ratio" specified for the user to control the stride at the time of update;
is a dimensionless length based on a given direction scale of the computational domain,
round (a) is a short representation, meaning rounded to the nearest integer to a;
by the update process, the number of low-dimensional subdomains NjWill expand proportionally to the relative size of the computational domain; although N isjMust be an integer, but the addition process will also depend on the update ratio α and the number of iterations; when the relatively small update ratio α does not change in the update step, the number of iterations is increased and the next update process is performed.
The invention has the beneficial effects that:
1. the data processing method is based on open source numerical simulation software OpenFOAM, and has strong expandability and transportability;
2. the low-dimensional process proposed by the invention is completely integrated and is directly butted with a CFD simulation result;
3. the non-uniform method provided by the invention can effectively improve the low-dimensional conversion precision under the limited low-dimensional subdomain quantity, and the screening process of the non-uniform method occupies less memory;
4. the invention provides a low-dimensional error index considering 'volume contribution rate', which can more accurately reflect the transformation condition of the low-dimensional process;
5. the invention improves the indoor air quality and thermal comfort of the office building, provides an intelligent data processing method, and greatly saves the working efficiency for the follow-up research of the office building.
Drawings
The invention will be further described with reference to the accompanying drawings.
FIG. 1 is a schematic view of a ventilation system according to the present invention;
FIG. 2 is a flow chart of a data processing method of the present invention;
FIG. 3 is a schematic diagram of the volume-averaged conversion of the physical field according to the present invention;
FIG. 4 is a low-dimensional computational domain partitioning diagram of the present invention;
FIG. 5 is a flow diagram of an adaptive update module of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
An intelligent ventilation system suitable for office buildings is arranged on a building, as shown in figure 1, an office area 8 is arranged in the building, the ventilation system comprises an air conditioning unit 1, an air supply unit, an air return unit and a fresh air unit 7 are communicated with the air conditioning unit 1, the air supply unit and the air return unit are respectively arranged on two sides of the air conditioning unit 1, the air supply unit comprises an air supply pipeline 3, one end of the air supply pipeline 3 is communicated with the air conditioning unit 1, the other end of the air supply pipeline 3 extends to a strip seam type air supply opening 5 arranged in the building, a fan assembly 2 is arranged in the air conditioning unit 1, the air return unit comprises an air return pipeline 4, one end of the air return pipeline 4 is communicated with the air conditioning unit 1, the other end of the air return pipeline extends to a strip seam type, the return air unit and the fresh air unit 7 form a circulation loop to realize air circulation of the office area 8.
A data processing method of an intelligent ventilation system suitable for office buildings, which relates to a building with a geometric size of 8m (length) x 5m (width) x 4m (height), as shown in fig. 1, the data processing method comprises the following steps:
s1: an input and output part of the calculation domain is used for reading an original CFD simulation result and outputting the original CFD simulation result in a dimensionality reduction mode;
input and output of the computational domain: and reading the original CFD simulation result (including grid and data) and outputting the original CFD simulation result in a dimensionality reduction mode by using numerical simulation open source software OpenFOAM.
Computing a domain input: the CFD results of the OpenFOAM simulation example are read directly, including grid and physical field data and user-set data required for software operation.
Computing domain output: the computationally obtained low-dimensional translation results, including the low-dimensional grid and physical field data, are exported to subdirectories of the OpenFOAM algorithm.
S2: physical field volume average conversion and error calculation part
S21 volume average conversion of physical field:
n volumes of total volume omega are added to form omegaiIs divided into N volumes of omega by' cut points PcA grid of sub-regions (the sub-region sizes need not be identical) as shown in fig. 3, where c is 1-N, i is 1-N, N<<n。
The number of sub-region grids can be expressed as N-Nx×Ny×NzWhere the subscripts x, y, z denote the number of subfields per direction, and the number of cut points P per direction is Nj+1(j ═ x, y, z) contains a start point and an end point.
After the sub-field division is completed, the grid information (grid volume, physical field value C of the corresponding position) corresponding to the sub-field grid position (x, y, z) coordinates is recorded into the array artificial intelligence.
According to the array artificial intelligence, the physical field values C (such as pollutant content) corresponding to the grids corresponding to the (x, y, z)) coordinates can also be sequentially divided into corresponding arrays, and the volume average value of all data in each sub-domain is calculated based on the grid data in each array artificial intelligence (the volume average value is: (the volume average value is calculated)<C>c)。
The volume averaged data result is used to represent the entire grid data result for the corresponding sub-field, enabling discrete (low-dimensional) processing of high-precision grid data.
S22 error calculation:
to evaluate the performance of low-dimensional transformations, we present a transformation error ε, which can be used to represent the low-dimensional dispersion performance of a particular region (for example).
The grid volume ratio is introduced into the equation so that it more accurately represents the actual contribution of a single grid to the low-dimensional transformation (i.e., the volume averaging process). The volume mean values are considered to be uniformly distributed in the region and the volumes consist of satisfied corresponding grid volumes, so that we can use the oa to quantify the difference between the low-dimensional field and the CFD field. The transformation error variance σ can be used as a reference for the distribution of errors throughout the volume when analyzing different low dimensional magnitudes and partitioning methods.
The low-dimensional discrete error evaluation is very important in the practical application of dimension reduction, and provides a method for evaluating the number range of the regions to be divided. Performing dimension reduction is also an important parameter. According to the error index, an automatic and adaptive low-dimensional method is realized through an adaptive updating module.
S3: and a low-dimensional calculation domain division and adaptive updating part.
S31 non-uniform partitioning module:
the division method is to determine the cut of each directionThe point P is set. As shown in fig. 4, it is a size N in each directionjAn array of +1(j ═ x, y, z), which specifies the boundaries of the low-dimensional regions in one dimension. These boundaries are by default uniformly distributed. However, in some cases it may lose much precision during the low-dimensional transformation. To control the error to a certain extent, we have to increase the number of subfields to reduce the deviation between the calculated low-dimensional field and the original CFD field. Therefore, in order to improve the precision of the low-dimensional conversion, a non-uniform partitioning method is proposed as an important component of the adaptive low-dimensional method.
To achieve non-uniform dimensionality reduction of the computational domain, the following "exact sub-layer reorganization" algorithm is proposed to determine the tangent point P.
S311: generation of sublayers: dividing the whole calculation domain omega into N in each direction (X, Y, Z) respectivelySubIndividual sub-layers, wherein the sub-layers in one direction are formed by equal thickness criteria lSubGenerating, then calculating therein, a volume-averaged scalar value C for each sub-layerSub。
S312: recombination and selection of sublayers: the sub-layers are directionally recombined according to a vector passing through the average scalar field value CSubAnd sublayer volumes to select the combination with the smallest recombination error. And after the optimal combination is obtained, converting the sub-layer into a one-way segmentation mode.
S313: the integration of the unidirectional split modes integrates the unidirectional split modes for each direction to construct a non-uniform low-dimensional field over the entire computational domain.
S32: self-adaptive updating:
the function of the adaptive updating module is to determine whether to increase the number N of low-dimensional subdomains after low-dimensional conversion in different modesjAnd j is x, y, z. As shown in FIG. 5, the adaptive CFD data dimension reduction tool provides two dimension reduction modes, referred to as "direct" and "auto/adaptive," respectively, which are implemented through common inheritance in C + +, which means that the "auto" mode is an advanced "direct" mode. With the difference that the "direct" mode does not increase the number of low-dimensional subfields Nj. It will be directly at a given number of low-dimensional subfields and division pattern (uniform or non-uniform)Level) while for the automatic mode, additional evaluation and update processes will be performed.
The evaluation result in the automatic mode is given based on the conversion error e and another index 'Iter' of the number of record update iterations. The user has specified a conversion error limit value oalimAnd iteration limit Iterlim. Furthermore, if oa is givenlimGreater than or equal to lim and Iter < IterlimThen the number of low-dimensional subfields N will be updatedj。
NjThe update procedure for (j ═ x, y, z) can be expressed as follows:
Nj-newis the updated number of low-dimensional subdomains; n is a radical ofj-newIs the initial value of the number of low-dimensional subdomains given by the user in the first iteration; α is an "update ratio" specified for the user to control the stride at the time of update;
is a dimensionless length based on a given direction scale of the computational domain,
round (a) is a short representation that is rounded to the nearest integer to a.
By the update process, the number of low-dimensional subdomains NjWill expand in proportion to the relative size of the computational domain. Although N isjMust be an integer, but the addition process will also depend on the update ratio alpha and the number of iterations. In some cases, due to the relatively small update ratio α, it does not change in one update step, at which point the number of iterations will increase and the next update process will be performed.
In the description herein, references to the description of "one embodiment," "an example," "a specific example" or the like are intended to mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, the schematic representations of the terms used above do not necessarily refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.
The foregoing shows and describes the general principles, essential features, and advantages of the invention. It will be understood by those skilled in the art that the present invention is not limited to the embodiments described above, which are described in the specification and illustrated only to illustrate the principle of the present invention, but that various changes and modifications may be made therein without departing from the spirit and scope of the present invention, which fall within the scope of the invention as claimed.
Claims (7)
1. The utility model provides an intelligent ventilation system suitable for official working building, ventilation system are located building upper portion, set up office area (8) in the building, a serial communication port, ventilation system includes air conditioning unit (1), the intercommunication is equipped with air supply unit, return air unit and new trend unit (7) on air conditioning unit (1), air supply unit and return air unit are located the both sides of air conditioning unit (1) respectively, air supply unit includes supply-air duct (3), the one end and the air conditioning unit (1) intercommunication of supply-air duct (3), the other end extends to and is equipped with strip seam type supply-air outlet (5) in the building, air conditioning unit (1) inside is equipped with fan subassembly (2), return air unit includes return air duct (4), the one end and the air conditioning unit (1) intercommunication of return air duct (4), the other end extends to and is equipped with strip seam type return-air outlet (6) in the building, air conditioning, The air return unit and the fresh air unit (7) form a circulation loop to realize air circulation of an office area (8).
2. The data processing method of the intelligent ventilation system for office buildings according to claim 1, wherein the data processing method comprises the following steps:
s1: an input and output part of the calculation domain is used for reading an original CFD simulation result and outputting the original CFD simulation result in a dimensionality reduction mode;
s2: a physical field volume average conversion and error calculation part;
s3: and a low-dimensional calculation domain division and adaptive updating part.
3. The data processing method of the intelligent ventilation system for office buildings according to claim 2, wherein the S1 includes:
input and output of the computational domain: reading and outputting original CFD simulation results, software user settings and dimension reduction by using numerical simulation open source software OpenFOAM and a dimension reduction tool;
computing a domain input: directly reading a CFD result of the OpenFOAM simulation example, wherein the CFD result comprises grid and physical field data and user setting data required by software operation;
computing domain output: the computationally obtained low-dimensional translation results, including the low-dimensional grid and physical field data, are exported to subdirectories of the OpenFOAM algorithm.
4. The data processing method of the intelligent ventilation system for office buildings according to claim 2, wherein the S2 includes:
s21: volume average conversion of physical field:
n volumes of total volume omega are added to form omegaiIs divided into N volumes of omega by' cut points PcA grid of sub-regions of;
when the position of the cutting point P is determined, the division of the subdomain is determined, and the number of subdomain grids is expressed as N-Nx×Ny×NzWhere the subscripts x, y, z denote the number of subfields per direction, and the number of cut points P per direction is Nj+1(j ═ x, y, z) comprising a start point and an end point;
after the sub-domain division is finished, grid information corresponding to the (x, y, z) coordinates of the sub-domain grid position is recorded into the array artificial intelligence;
according to the array artificial intelligence, the physical field values C corresponding to the grids corresponding to the (x, y, z) coordinates are also sequentially divided into corresponding arrays, and then all the numbers in each sub-domain are calculated based on the grid data in each array artificial intelligenceAccording to the volume average value, expressed as<C>c；
The volume average data result is used for expressing all grid data results of the corresponding sub-domains, and the discrete processing of high-precision grid data is realized;
s22: and (3) error calculation:
the performance of low-dimensional conversion is evaluated by adopting a conversion error epsilon, the influence of the grid volume must be considered by using a finite volume method, otherwise, the error calculation is inaccurate, and particularly, when the volume change of a single grid of a CFD grid is large, the method is very common in practical simulation;
the transformation error ε is defined as follows:
the grid volume ratio is introduced into the equation so that it represents more accurately the actual contribution of the individual grids to the low-dimensional conversion, the volume mean values are distributed uniformly over the area, the volume consists of the respective grid volumes which are met, and the difference between the low-dimensional field and the CFD field is quantified using an oa.
5. The data processing method of the intelligent ventilation system for office buildings according to claim 4, wherein c is 1-N, i is 1-N, and N < < N.
6. The data processing method of the intelligent ventilation system for office buildings according to claim 4, wherein the conversion error epsilon is used for representing the low-dimensional discrete performance of a specific area by changing the volume limit omega in the above formula into omegacA regional error index is obtained that can represent the performance of the conversion error epsilon in the sub-region:
calculating a transformation error variance sigma according to the regional error index, wherein the transformation error variance sigma is used as a reference for analyzing error distribution in the whole volume when different low-dimensional magnitudes and dividing methods are analyzed;
the low-dimensional discrete error evaluation is used for evaluating the number range of the regions to be divided, the dimension reduction processing is also an important parameter, and an automatic and self-adaptive low-dimensional method is realized through a self-adaptive updating module according to the error index.
7. The data processing method of the intelligent ventilation system for office buildings according to claim 1, wherein the S3 includes:
s31: a non-uniform partitioning module:
the dividing method is to determine the tangent point P setting of each direction, and the tangent point P setting is N in each directionjAn array of +1(j ═ x, y, z), specifying boundaries of low-dimensional regions in one dimension, the boundaries being uniformly distributed by default; increasing the number of sub-fields, reducing the deviation between the calculated low-dimensional field and the original CFD field, and controlling the error within a certain range; a non-uniform division method is adopted as an important component of a self-adaptive low-dimensional method to improve the precision of low-dimensional conversion;
determining the tangent point P by combining an accurate sublayer recombination algorithm, realizing the non-uniform dimensionality reduction of a computational domain,
s311: generation of sublayers: dividing the whole calculation domain omega into N in each direction (X, Y, Z) respectivelySubIndividual sub-layers, wherein the sub-layers in one direction are formed by equal thickness criteria lSubGenerating, then calculating therein, a volume-averaged scalar value C for each sub-layerSub；
S312: recombination and selection of sublayers: the sub-layers are directionally recombined according to a vector passing through the average scalar field value CSubSelecting the combination with the minimum recombination error according to the sub-layer volume, and converting the sub-layer into a one-way segmentation mode after obtaining the optimal combination;
s313: the integration of the unidirectional segmentation modes integrates the unidirectional segmentation modes of each direction to construct a non-uniform low-dimensional field over the entire computational domain;
s32, self-adaptive updating:
the self-adaptive updating module determines whether to increase the number N of low-dimensional subdomains after low-dimensional conversion in different modesjJ is x, y, z; the adaptive CFD data dimension reduction tool provides two dimension reduction modes, called "direct" and "automatic/adaptive", respectively, implemented by common inheritance in C + +, which means that the "automatic" mode is an advanced "direct" mode; with the difference that the "direct" mode does not increase the number of low-dimensional subfields Nj(ii) a It will accomplish the low dimensional conversion directly for a given number of low dimensional subdomains and partitioning mode, while for the automatic mode, additional evaluation and update procedures will be performed;
the evaluation result in automatic mode is given according to the conversion error epsilon and another index 'Iter' recording the number of update iterations; the user has specified a conversion error limit value oalimAnd iteration limit Iterlim(ii) a Furthermore, if oa is givenlimGreater than or equal to lim and Iter < IterlimThen the number of low-dimensional subfields N will be updatedj；
NjThe update procedure for (j ═ x, y, z) can be expressed as follows:
Nj-newis the updated number of low-dimensional subdomains; n is a radical ofj-newIs the initial value of the number of low-dimensional subdomains given by the user in the first iteration; α is an "update ratio" specified for the user to control the stride at the time of update;
is a dimensionless length based on a given direction scale of the computational domain,
round (a) is a short representation, meaning rounded to the nearest integer to a;
by the update process, the number of low-dimensional subdomains NjWill expand proportionally to the relative size of the computational domain; although N isjMust be an integer, but the addition process will also depend on the update ratio α and the number of iterations; when the relatively small update ratio α does not change in the update step, the number of iterations is increased and the next update process is performed.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011280986.4A CN112434361A (en) | 2020-11-16 | 2020-11-16 | Intelligent ventilation system suitable for office buildings and data processing method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011280986.4A CN112434361A (en) | 2020-11-16 | 2020-11-16 | Intelligent ventilation system suitable for office buildings and data processing method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112434361A true CN112434361A (en) | 2021-03-02 |
Family
ID=74700187
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011280986.4A Pending CN112434361A (en) | 2020-11-16 | 2020-11-16 | Intelligent ventilation system suitable for office buildings and data processing method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112434361A (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107357999A (en) * | 2017-07-19 | 2017-11-17 | 云南电网有限责任公司电力科学研究院 | The method for numerical simulation and system of a kind of wind field |
| CN108563867A (en) * | 2018-04-13 | 2018-09-21 | 华中科技大学 | A method of WRF and CFD coupled simulation wind fields are realized based on OpenFOAM |
| CN108763794A (en) * | 2018-06-01 | 2018-11-06 | 北京航空航天大学 | A kind of heat analysis method being used for the thermotropic failure of circuit board product and reliability assessment |
| CN110532687A (en) * | 2019-08-29 | 2019-12-03 | 哈尔滨工业大学 | One kind cold ground gymnasium building nature ventilation parameterization design method |
| CN111928407A (en) * | 2020-08-10 | 2020-11-13 | 东南大学 | Intelligent flow guiding and ventilating device and control method |
-
2020
- 2020-11-16 CN CN202011280986.4A patent/CN112434361A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107357999A (en) * | 2017-07-19 | 2017-11-17 | 云南电网有限责任公司电力科学研究院 | The method for numerical simulation and system of a kind of wind field |
| CN108563867A (en) * | 2018-04-13 | 2018-09-21 | 华中科技大学 | A method of WRF and CFD coupled simulation wind fields are realized based on OpenFOAM |
| CN108763794A (en) * | 2018-06-01 | 2018-11-06 | 北京航空航天大学 | A kind of heat analysis method being used for the thermotropic failure of circuit board product and reliability assessment |
| CN110532687A (en) * | 2019-08-29 | 2019-12-03 | 哈尔滨工业大学 | One kind cold ground gymnasium building nature ventilation parameterization design method |
| CN111928407A (en) * | 2020-08-10 | 2020-11-13 | 东南大学 | Intelligent flow guiding and ventilating device and control method |
Non-Patent Citations (1)
| Title |
|---|
| JIE REN等: "Development of self-adaptive low-dimension ventilation models using OpenFOAM: Towards the application of AI based on CFD data", WEB OF SCIENCE, 16 January 2020 (2020-01-16), pages 1 - 12 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Zhang et al. | Optimal design of building environment with hybrid genetic algorithm, artificial neural network, multivariate regression analysis and fuzzy logic controller | |
| CN103049612B (en) | Building indoor environment optimization method based on model order reduction technology | |
| Zhou et al. | Comparison of different deep neural network architectures for isothermal indoor airflow prediction | |
| CN114662201B (en) | Optimizing method for intelligent regulation and control of natural ventilation | |
| CN111881505A (en) | Multi-objective optimization transformation decision method for existing building based on GA-RBF algorithm | |
| CN114200986B (en) | Greenhouse environment optimization design method considering crop production benefits and energy saving | |
| Li et al. | Multi-objective optimization of HVAC system using NSPSO and Kriging algorithms—A case study | |
| CN115907191A (en) | Adaptive building photovoltaic skin model prediction control method | |
| CN112432230A (en) | Intelligent ventilation monitoring system and control method based on pollution source position information | |
| CN115310350A (en) | Chicken coop air age prediction method based on POD-ANN order reduction model | |
| CN112152840A (en) | Sensor deployment method and system based on BIM and analog simulation | |
| Zhang et al. | Research on the combined prediction model of residential building energy consumption based on random forest and BP neural network | |
| CN116147154A (en) | Method and device for adjusting temperature of air conditioner in machine room and electronic equipment | |
| CN113158593B (en) | Intelligent design method of hospital ward ventilation system based on artificial intelligence | |
| CN112434361A (en) | Intelligent ventilation system suitable for office buildings and data processing method | |
| JP7459935B2 (en) | Estimation method, simulation method, estimation device, and estimation program | |
| Wang et al. | Research on intelligent regulation of air conditioning energy saving based on human thermal comfort | |
| Menzel et al. | Application of free form deformation techniques in evolutionary design optimisation | |
| CN116744634A (en) | Data center temperature prediction control method | |
| CN113505929B (en) | Topological optimal structure prediction method based on embedded physical constraint deep learning technology | |
| CN115540252A (en) | Heating ventilation air conditioner energy-saving control system and method based on machine learning | |
| CN110196987A (en) | The dimensionally-optimised method of air channel structure based on agent model | |
| WO2024073937A1 (en) | Method for assessing air conditioner control algorithm, and apparatus, device and storage medium | |
| Fragkia et al. | Occupant-oriented adaptive building envelopes: A hybrid design framework for human, material, environment synergies | |
| CN111967090A (en) | Dynamic improvement method for optimizing design space |
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 |