CN109685246B - Environment data prediction method and device, storage medium and server - Google Patents
Environment data prediction method and device, storage medium and server Download PDFInfo
- Publication number
- CN109685246B CN109685246B CN201811348276.3A CN201811348276A CN109685246B CN 109685246 B CN109685246 B CN 109685246B CN 201811348276 A CN201811348276 A CN 201811348276A CN 109685246 B CN109685246 B CN 109685246B
- Authority
- CN
- China
- Prior art keywords
- data
- target site
- environmental data
- site
- target
- 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
- 238000000034 method Methods 0.000 title claims abstract description 66
- 230000007613 environmental effect Effects 0.000 claims abstract description 224
- 239000003344 environmental pollutant Substances 0.000 claims abstract description 93
- 231100000719 pollutant Toxicity 0.000 claims abstract description 63
- 238000013527 convolutional neural network Methods 0.000 claims abstract description 56
- 230000008878 coupling Effects 0.000 claims abstract description 51
- 238000010168 coupling process Methods 0.000 claims abstract description 51
- 238000005859 coupling reaction Methods 0.000 claims abstract description 51
- 239000000356 contaminant Substances 0.000 claims description 32
- 230000007246 mechanism Effects 0.000 claims description 31
- 230000015654 memory Effects 0.000 claims description 30
- 238000004364 calculation method Methods 0.000 claims description 14
- 238000013075 data extraction Methods 0.000 claims description 13
- 230000001537 neural effect Effects 0.000 claims description 12
- 238000004590 computer program Methods 0.000 claims description 4
- 230000008859 change Effects 0.000 abstract description 28
- 238000012544 monitoring process Methods 0.000 abstract description 20
- 238000001514 detection method Methods 0.000 abstract description 5
- 238000007405 data analysis Methods 0.000 abstract description 2
- 230000004044 response Effects 0.000 abstract description 2
- 230000008569 process Effects 0.000 description 28
- 230000006870 function Effects 0.000 description 16
- 238000000605 extraction Methods 0.000 description 14
- RAHZWNYVWXNFOC-UHFFFAOYSA-N Sulphur dioxide Chemical compound O=S=O RAHZWNYVWXNFOC-UHFFFAOYSA-N 0.000 description 8
- 230000007774 longterm Effects 0.000 description 6
- 230000008901 benefit Effects 0.000 description 5
- 230000002093 peripheral effect Effects 0.000 description 5
- 238000010586 diagram Methods 0.000 description 3
- 229910052739 hydrogen Inorganic materials 0.000 description 3
- 239000011159 matrix material Substances 0.000 description 3
- 210000005036 nerve Anatomy 0.000 description 3
- MWUXSHHQAYIFBG-UHFFFAOYSA-N nitrogen oxide Inorganic materials O=[N] MWUXSHHQAYIFBG-UHFFFAOYSA-N 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- 229910052698 phosphorus Inorganic materials 0.000 description 3
- 230000009466 transformation Effects 0.000 description 3
- 230000006978 adaptation Effects 0.000 description 2
- 238000010219 correlation analysis Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000003912 environmental pollution Methods 0.000 description 2
- 230000005291 magnetic effect Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 239000013598 vector Substances 0.000 description 2
- UGFAIRIUMAVXCW-UHFFFAOYSA-N Carbon monoxide Chemical compound [O+]#[C-] UGFAIRIUMAVXCW-UHFFFAOYSA-N 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 229910002091 carbon monoxide Inorganic materials 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000010339 dilation Effects 0.000 description 1
- 229930195733 hydrocarbon Natural products 0.000 description 1
- 150000002430 hydrocarbons Chemical class 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 230000035772 mutation Effects 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- Economics (AREA)
- General Physics & Mathematics (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Entrepreneurship & Innovation (AREA)
- Biomedical Technology (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Game Theory and Decision Science (AREA)
- Development Economics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Operations Research (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The application relates to the technical fields of environment detection, data analysis and predictive estimation, and provides an environment data prediction method, which comprises the following steps: acquiring a target site and environment data of each site within a preset range of the target site, and extracting key data affecting the target site based on a convolutional neural network and the environment data; acquiring a first pre-estimated model, a target site and historical environment data of each site within a preset range of the target site, and determining the first pre-estimated environment data based on the first pre-estimated model and the historical environment data; and coupling the key data and the first estimated environmental data to obtain second estimated environmental data of the target site. According to the application, the traditional environmental pollutant prediction model is combined on the basis of the convolutional neural network, so that the prediction precision is improved, the prediction time period is shortened, the environmental pollutant change trend is more accurate, the precision of the predicted environmental pollutant is improved, and the rapid response to the pollutant change of the monitoring station is facilitated.
Description
Technical Field
The invention relates to the technical fields of environment detection, data analysis and predictive estimation, in particular to an environment data prediction method and device, a storage medium and a server.
Background
With the development of science and technology, the national and social individuals pay more and more attention to environmental protection, so that more materials and people are invested for detecting environmental pollutants. In the related technology, most of commonly adopted environmental pollution prediction models only predict the data of a single monitoring site to obtain the variation trend of the monitored single site pollutants; or simply calculating the influence of the nearby sites on the pollutant change trend of the monitoring site, and the above calculation method cannot well reflect the influence of the space information on the target monitoring site. In addition, the traditional numerical model can comprehensively consider the influence of the space information on the target monitoring point to give more accurate change trend prediction, but the traditional numerical model is difficult to give short-term change prediction, is commonly used for the change of the pollutant value in a longer period, and consumes huge calculation power when the period is shortened, such as for an hour-level period, and is difficult to realize.
Disclosure of Invention
In order to solve the technical problems, particularly the problems that only a single monitoring station is predicted by the existing environmental pollution prediction model, the influence of peripheral stations on target station pollutants is simple to calculate, and the prediction period of the traditional prediction model is long, the following technical scheme is specifically provided:
the environment data prediction method provided by the embodiment of the invention comprises the following steps:
acquiring a target site and environment data of each site within a preset range of the target site, and extracting key data affecting the target site based on a convolutional neural network and the environment data;
Acquiring a first pre-estimated model, a target site and historical environment data of each site within a preset range of the target site, and determining the first pre-estimated environment data based on the first pre-estimated model and the historical environment data;
and coupling the key data and the first estimated environmental data to obtain second estimated environmental data of the target site.
Optionally, the coupling the key data and the first estimated environmental data to obtain second estimated environmental data of the target site includes:
Coupling the key data and the first estimated environmental data through an attention mechanism to obtain coupling data;
And inputting the coupling data into a convolutional neural network for operation to obtain the second estimated environmental data.
Optionally, the acquiring the target site and the environmental data of each site within the preset range of the target site, extracting the key data affecting the target site based on the convolutional neural network and the environmental data, includes:
Acquiring a target site and environmental data of each site within a preset range of the target site, and determining an influence value of the environmental data of each site on the environmental data of the target site based on a convolutional neural network;
and extracting key data influencing the environmental data of the target site based on the influence value.
Optionally, the acquiring the target site and the environmental data of each site within the preset range of the target site, determining, based on a convolutional neural network, an influence value of the environmental data of each site on the environmental data of the target site, includes:
Acquiring the environmental data of the target site and each site within a preset range of the target site, wherein the environmental data comprises meteorological data and pollutant data of at least one pollutant;
Determining a first impact value of the environmental data of the target site on the target contaminant based on the meteorological data of the target site, the contaminant data of the target contaminant, and an attention mechanism;
And determining a second influence value of each station on the target pollutant of the target station based on the meteorological data of each station, the pollutant data of the target pollutant, the correlation between the target station and each station and the attention mechanism in a preset range.
Optionally, the extracting key data affecting the environmental data of the target site based on the impact value includes:
And inputting the first influence value and the second influence value into a convolutional neural network to obtain the key data influencing the environmental data of the target site.
Optionally, the environmental data is acquired through a convolutional neural network.
Optionally, the acquiring the target site and the environmental data of each site within the preset range of the target site includes:
and acquiring the target site and the environmental data of each site in the preset range of the target site according to the preset environmental data acquisition time period.
The device for estimating the environmental data provided by the embodiment of the application comprises:
the key data extraction module is used for acquiring the target site and the environment data of each site within the preset range of the target site, and extracting the key data affecting the target site based on the convolutional neural network and the environment data;
the first estimated environmental data determining module is used for acquiring a first estimated model, a target site and historical environmental data of each site within a preset range of the target site, and determining the first estimated environmental data based on the first estimated model and the historical environmental data;
And the coupling module is used for coupling the key data and the first estimated environmental data to obtain second estimated environmental data of the target site.
Optionally, the coupling module includes:
the coupling unit is used for coupling the key data and the first estimated environmental data through an attention mechanism to obtain coupling data;
and the convolutional neural calculation unit is used for inputting the coupling data into a convolutional neural network to perform operation so as to obtain the second estimated environmental data.
The embodiment of the invention also provides a computer readable storage medium, wherein the computer readable storage medium is stored with a computer program, and the program realizes the environmental data prediction method according to any technical scheme when being executed by a processor.
The embodiment of the invention also provides a server, which comprises:
One or more processors;
A memory;
One or more applications, wherein the one or more applications are stored in the memory and configured to be executed by the one or more processors, the one or more applications configured to perform the steps of the environmental data prediction method according to any of the claims.
The embodiment of the invention also provides a computer readable storage medium, wherein the computer readable storage medium is stored with a computer program, and the program realizes the environmental data prediction method according to any technical scheme when being executed by a processor.
The embodiment of the invention also provides a server, which comprises:
One or more processors;
A memory;
One or more applications, wherein the one or more applications are stored in the memory and configured to be executed by the one or more processors, the one or more applications configured to perform the environmental data prediction method according to any of the claims.
Compared with the prior art, the invention has the following beneficial effects:
1. The environment data prediction method provided by the embodiment of the application comprises the following steps: acquiring a target site and environment data of each site within a preset range of the target site, and extracting key data affecting the target site based on a convolutional neural network and the environment data; acquiring a first pre-estimated model, a target site and historical environment data of each site within a preset range of the target site, and determining the first pre-estimated environment data based on the first pre-estimated model and the historical environment data; and coupling the key data and the first estimated environmental data to obtain second estimated environmental data of the target site. According to the application, the attention mechanism and the traditional environmental pollutant pre-estimation model (such as a numerical model, an RNN model, a Sequence2Sequence and the like) are combined on the basis of the convolutional neural network, so that the pre-estimation time of the environmental pollutant change trend is shortened, the prediction precision is improved, the prediction time period is shortened, and meanwhile, the environmental pollutant change trend is obtained more accurately, and the precision of the pre-estimated environmental pollutant is improved, so that the environmental pollutant at the monitoring station can be responded more quickly.
2. The method for estimating the environmental data provided by the embodiment of the application comprises the steps of obtaining the environmental data of each site in a preset range of a target site and determining the influence value of the environmental data of each site on the environmental data of the target site based on a convolutional neural network, wherein the method comprises the following steps: acquiring the environmental data of the target site and each site within a preset range of the target site, wherein the environmental data comprises meteorological data and pollutant data of at least one pollutant; determining a first impact value of the environmental data of the target site on the target contaminant based on the meteorological data of the target site, the contaminant data of the target contaminant, and an attention mechanism; and determining a second influence value of each station on the target pollutant of the target station based on the meteorological data of each station, the pollutant data of the target pollutant, the correlation between the target station and each station and the attention mechanism in a preset range. In the application, not only the influence of the peripheral sites of the target site on the environment of the target site is considered, but also the influence of the environmental data of the target site on related pollutants is considered, the acquired factors influencing the pollutants of the target site are more comprehensive, the acquired prediction result is more accurate, in the process, the influence of each peripheral site is independently extracted by adopting a convolution nerve module, and the correlation analysis among the sites is combined, in the calculation process, more factors influencing the target site are added, the influence on the target site is comprehensively calculated, and the prediction is performed by using spatial information more accurately, so that the calculated environmental pollutants of the target site are more accurate.
Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention.
Drawings
The foregoing and/or additional aspects and advantages of the invention will become apparent and readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings, in which:
FIG. 1 is a flowchart of an exemplary embodiment of an environmental data prediction method according to the present invention;
FIG. 2 is a schematic diagram of a TCN network structure of a cavity convolution mode in the environmental data prediction method of the present invention;
FIG. 3 is a flowchart illustrating another embodiment of an environmental data prediction method according to the present invention;
FIG. 4 is a schematic diagram illustrating an exemplary embodiment of an environmental data prediction apparatus according to the present invention;
Fig. 5 is a schematic structural diagram of an embodiment of a server according to the present invention.
Detailed Description
Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to like or similar elements or elements having like or similar functions throughout. The embodiments described below by referring to the drawings are illustrative only and are not to be construed as limiting the invention.
As used herein, the singular forms "a", "an", "the" and "the" are intended to include the plural forms as well, unless expressly stated otherwise, as understood by those skilled in the art. It will be further understood that the terms "comprises" and/or "comprising," when used in this specification, specify the presence of stated features, integers, steps, operations, but do not preclude the presence or addition of one or more other features, integers, steps, operations.
It will be understood by those skilled in the art that all terms (including 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 unless defined otherwise. It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the prior art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.
It will be appreciated by those skilled in the art that references to "application," "application program," "application software," and similar concepts herein are intended to be equivalent concepts well known to those skilled in the art, and refer to computer software, organically constructed from a series of computer instructions and related data resources, suitable for electronic execution. Unless specifically specified, such naming is not limited by the type, level of programming language, nor by the operating system or platform on which it operates. Of course, such concepts are not limited by any form of terminal.
The environmental data prediction method provided by the embodiment of the application is mainly applied to an environmental pollutant prediction model in an environmental pollutant monitoring system, wherein the model comprises a convolutional neural module, an attention mechanism module and a traditional environmental pollutant prediction module, and the environmental pollutant prediction module is used as the middle input of the whole environmental pollutant prediction model so as to realize superposition of various prediction results and further improve the precision of the whole prediction model. The embodiment of the application mainly realizes the environmental data prediction method through the convolutional neural network and the attention mechanism in the convolutional neural module and the attention mechanism module.
The method for estimating environmental data provided by the embodiment of the application, as shown in fig. 1, comprises the following steps: s100, S200, S300.
S100: acquiring a target site and environment data of each site within a preset range of the target site, and extracting key data affecting the target site based on a convolutional neural network and the environment data;
S200: acquiring a first pre-estimated model, a target site and historical environment data of each site within a preset range of the target site, and determining the first pre-estimated environment data based on the first pre-estimated model and the historical environment data;
S300: and coupling the key data and the first estimated environmental data to obtain second estimated environmental data of the target site.
In the embodiment of the application, the condition of the environmental pollutants of the target site can be predicted more accurately. Acquiring the target site and the environmental data of each site within the preset range of the target site, wherein the environmental data comprises: contaminant data, weather data. The contaminant data includes: PM2.5, PM10, O 3、NO2、SO2, CO, etc., the meteorological data includes: weather data such as temperature, barometric pressure, relative humidity, wind speed, wind direction, weather conditions, and the like. The preset range is set by monitoring personnel. And inputting the environmental data into a convolutional neural network, and extracting key data affecting the target site based on the environmental data of the target site. And performing feature extraction on each site around the target site and the target site in time sequence by using a convolutional neural network, namely extracting environment data, and particularly, acquiring a wider time receptive field by using a cavity convolution mode so as to acquire comprehensive time dimension information. And then, obtaining the influence of the environmental data of the surrounding sites on the different environmental information of the target site through the softmax layer. For example, suppose there are A, B, C, D sites, a being the target site. The method comprises the steps of carrying out feature extraction (features such as pollutant content, temperature, air pressure, wind flow, weather conditions and the like) on a B, C, D pollutant value and space features through a convolution module, obtaining influence of different features of a target site on an A site (for example, the influence of temperature change of a B site on the A site is larger, and the influence of wind speed of a C site on the A site is larger), simultaneously calculating environmental data of the A site through the convolution module, obtaining the influence of the environmental data of the A site on the target pollutant, and extracting key data based on the influence of the B, C, D site on the A site and the environmental data of the A site. The specific process is described in detail later, and will not be described here.
In order to more accurately estimate the change condition of the pollutant value of the target site. The method comprises the steps of obtaining a first pre-estimated model, wherein the model can comprehensively consider the influence of space information on target monitoring points, give a relatively accurate long-term pollutant change trend of a target site, and the first pre-estimated model can comprise one or more of a numerical model, an RNN model and a Sequence2Sequence model. And inputting all the historical environmental data of all the sites into the first estimation model to obtain first estimation environmental data. When the Sequence2Sequence model is adopted, the operation process of the first estimated environmental data can be as follows:
Assume that the first predicted environmental data is
The main task at this stage is to predict the future Y matrix. Using X i as a second pre-estimated model input:
Where X is the total statistics (historical environmental data) of past sites. Since X has different properties, its spatial and temporal properties are different, such as PM2.5 is closely related to the information of NO and SO 2, which is a time dependence; at the same time, there is also an association between different sites, which is a spatial correlation. Thus, spatial attention mechanisms are used to model this correlation. Inputting X into an LSTM module, outputting two vectors of h and s, and calculating attention parameters according to the following calculation formula:
Where v l、bl、Wl and U l are parameters to be learned. I.e. attention coefficients. The final inputs are as follows:
The global attention parameters can also be calculated. After the global attention parameters are calculated, the final characteristics are as follows:
Wherein, Is a global attention coefficient. After obtaining the local attention features and the global attention features, space attention is used to operate to obtain space attenttion features:
Wherein W d, And v d、bd is a learnable parameter. C is the output feature vector of global attention. After c is obtained, the local attention features are spliced with the global attention and input into a self-coding-decoding model together to obtain a final prediction result. The coding-decoding model is a stack of a series of LSTM, and after information is input into the coding model, the decoding model outputs a final first estimated environmental predicted value P.
It should be noted that, the key data extraction process and the first estimated environmental data determination process are not sequential, that is, the two processes may be performed simultaneously, or the first estimated environmental data determination process may be performed by advanced key data extraction, or the two processes may be exchanged. And (3) predicting a more accurate transformation trend of pollutants in the environment of the target site, and coupling and superposing long-term data (first predicted environment data) and short-term data (key data) with long period so as to obtain second predicted environment data of the target site. Specifically, coupling is performed through an attention mechanism, and the coupled data is input into a convolutional neural network to obtain second estimated environmental data. In the operation process, a prediction result of a traditional environmental pollutant prediction model (such as a numerical model, an RNN model and a Sequence2Sequence model) is used as an intermediate input of a convolutional neural network, local mutation characteristics extracted by the convolutional network are overlapped, the prediction accuracy of the traditional model on a change trend and the extraction advantage of the convolutional network on the local characteristics are combined, and the prediction accuracy of the environmental pollutant is further improved. Further, after the foregoing basis, in order to more intuitively determine the possible change condition of the environmental data of the target monitoring site, so as to prepare corresponding response measures in advance based on the change condition, after the second estimated environmental data is acquired, the following procedure may be further performed: obtaining second estimated environmental data of each time period in a preset time period, generating an estimated curve of the target monitoring station, outputting the estimated curve of the target monitoring station, and displaying the curve on a display interface so as to be convenient for knowing the trend of environmental pollutants of the target monitoring station through the curve.
Optionally, the coupling the key data and the first estimated environmental data to obtain second estimated environmental data of the target site includes:
Coupling the key data and the first estimated environmental data through an attention mechanism to obtain coupling data;
And inputting the coupling data into a convolutional neural network for operation to obtain the second estimated environmental data.
And carrying the key information and the first estimated environmental data into an attention mechanism for coupling, and determining second estimated environmental data of the target site through a convolutional neural network. Specifically, the prediction result of the first prediction model and the key data (namely the target site and the environmental data of each surrounding site) extracted by the convolutional neural network are taken as new features extracted by the convolutional neural network, and the two are combined together to obtain the second prediction environmental data of the target site. The first estimated environmental data predicted by the first estimated model is used as an intermediate input for determining the second estimated environmental data. By combining the prediction accuracy of the first prediction model on the change trend and the extraction advantage of the convolution network on the local characteristics, the prediction accuracy of the prediction model is further improved, so that the prediction environment information of the target site is more reliable.
In a specific implementation process, the key data (i.e., data H) extracted from the context is combined, the key data is output from the hidden layer of the convolutional neural network, the predicted result P (the first predicted data) of the target pollutant (such as PM 2.5) in the future is coupled in the form of attention by using the Sequence2Sequence model (which may be an RNN model or a numerical model), where the coupling calculation is performed on H and P, the value obtained after H and P are sequentially calculated through Matmul, scale, mask (opt.), softmask, and the value and H are input Matmul to perform calculation, so as to obtain coupling data after coupling of H and P. And sending the coupling data into a convolutional neural module for feature extraction, and connecting a dense layer to serve as prediction output to obtain second estimated environmental data.
Optionally, in one embodiment, as shown in fig. 3, the acquiring the target site and the environmental data of each site within the preset range of the target site, extracting the key data affecting the target site based on the convolutional neural network and the environmental data includes:
S110, acquiring a target site and environment data of each site within a preset range of the target site, and determining an influence value of the environment data of each site on the environment data of the target site based on a convolutional neural network;
and S120, extracting key data influencing the environmental data of the target site based on the influence value.
In combination with the foregoing overview, in order to accurately estimate the environmental data of the target site, the calculation process of the present application uses a convolutional neural network to perform the calculation. Therefore, in this process, after the environmental data of the target site and each site around the target site are obtained, the influence value of the environmental data of each site on the environmental data of the target site is calculated by the convolution nerve, so that the key data affecting the environmental data of the target site is conveniently extracted according to the influence value, as the influence of the temperature change of the site B on the site a is greater, and the influence of the wind speed of the site C on the site a is greater, when the key data is extracted by the convolution nerve, the temperature change of the site B and the wind speed change of the site C need to be considered seriously.
In the operation process, since each site obtained through the convolutional neural network affects the environmental data of the target site, in order to realize superposition of multiple types of environmental information and improve the accuracy of environmental monitoring, the application also needs to determine the influence value of the environmental data of each site on the environmental data of the target site based on the convolutional neural network to extract key data.
In one embodiment, the application introduces the influence of surrounding sites into a convolutional neural module in attention form to determine the change trend of the target site; i.e., extracting critical data from the environmental data based on the Attention Model, such as one or more of air contaminant values (e.g., PM2.5, carbon monoxide, sulfur dioxide, nitrogen oxides, and hydrocarbons) within the aforementioned site detection range, weather information (e.g., temperature, humidity, wind speed, air pressure, etc.), and/or contaminant values in the water stream, etc. attention model aims to obtain different influence weights of related information of different pollutants, extract key information in the pollutants and improve the accuracy of prediction. This part of the neural network comprises two parts: a Time Convolutional Network (TCN) as feature extraction of contaminant information, followed by a softmax layer as output of attention matrix of contaminant features; the original contaminant information is multiplied by the element-wise of the attention matrix. And acquiring important information of the pollutants input by the station through the operation of the steps. In_X represents the input of the current site contaminant.
Attention(In_X)＝softmax(E(In_X))⊙In_X
Wherein, the Attention (In_X) indicates the output of the current site contaminant information after Attention passes through; e, indicating the operation of implicit characteristic extraction of the input pollutant information by the time convolution network; softmax represents the range in which the feature dimension of the input is normalized to [0,1 ]; as indicated by the element-wise multiplication.
The TCN network structure adopting the cavity convolution mode is shown in fig. 2, and has the main advantages that after dilation times, the convolution receptive field is multiplied, so that the model can learn more historical information, and the learning capacity of the model is improved.
The feature information of the obtained influence forces (local attention and global attention) is combined (see later examples for details), and is sent to a CNN (convolutional neural) model for feature extraction, so as to obtain hidden layer output H (output in the figure) which represents influence factors abstracted from historical time and space information (such as the temperature of the B site and the wind speed of the C site), wherein the extraction process is not limited to the temperature of the B site and the wind speed of the C site, and other environmental data of each site are combined to extract key data, that is, the influence of the temperature of the B site on the a site is larger, and after the temperature of the B site is extracted, other environmental data such as the wind speed of the B site may be extracted.
Optionally, the acquiring the target site and the environmental data of each site within the preset range of the target site, determining, based on a convolutional neural network, an influence value of the environmental data of each site on the environmental data of the target site, includes:
Acquiring the environmental data of the target site and each site within a preset range of the target site, wherein the environmental data comprises meteorological data and pollutant data of at least one pollutant;
Determining a first impact value of the environmental data of the target site on the target contaminant based on the meteorological data of the target site, the contaminant data of the target contaminant, and an attention mechanism;
And determining a second influence value of each station on the target pollutant of the target station based on the meteorological data of each station, the pollutant data of the target pollutant, the correlation between the target station and each station and the attention mechanism in a preset range.
When the environmental data of the sites are estimated, the influence of the environmental data of each site on the environmental data of each site needs to be considered. Therefore, in the process of estimating the specific pollutant value, each pollutant is used as the target pollutant, and the influence of all pollutant data and meteorological data of the target site on the target pollutant data is detected, for example:
The pollutant information PM2.5 concentration of the target site plus six meteorological information (7 characteristic values in total) are taken as the input of the convolutional neural network, and are sent to a convolutional neural network module for characteristic extraction. The magnitude of the impact of these 7 features on the future trend of the target contaminant (PM 2.5) is then obtained by the attention mechanism module. The influence of 7 characteristic values of the target site on the target pollutant is obtained through the process, and the influence is called local attention, which is the first influence value in the application.
Further, in determining the influence of the environmental data of the peripheral station on the environmental data of the target station, the correlation between the target station and the peripheral station needs to be considered. The correlation between sites is a generalized correlation, and is related to not only distance, but also collected environmental information trend. The closer the distance of the site/the closer the acquired environmental information, the higher the correlation of the site is considered.
The correlation is divided into two parts, one is the distance between the surrounding sites and the target site, which is inversely proportional to the correlation between sites, the farther the distance, the less the impact. However, the distance is only used as a correlation judgment basis, and if the geographic positions of the two stations are quite similar but separated by a mountain, the correlation is quite low. Therefore, a linear correlation coefficient DTW (dynamic time plan) calculated from environmental factors between stations is used as a second basis for correlation determination at the same time. The formula is as follows:
k in the denominator is mainly used to compensate for the regular paths of different lengths.
The previous example is assembled, assuming that 34 sites exist around the target site, the PM2.5 concentration of the remaining 34 sites plus the weather information of the corresponding sites are input into the convolutional neural network, and the convolutional neural network is sent to the convolutional neural module for feature extraction. Then, the contribution degree (i.e. influence attention) of different characteristics of 34 sites to the target pollutant is obtained by the attention mechanism module in combination with the correlation degree (including distance factor and linear correlation degree) between the 34 sites and the target site. The impact of the different characteristics of the remaining 34 stations on the target station PM2.5 is obtained through this process, which is referred to in this section as global attention, which is the second impact value in the present application. global attention can reflect the influence of different sites, and can also be specific to the main influence characteristics of the sites, for example, the influence of the temperature change of the site B on the target site is larger, and the influence of the wind speed of the site C on the target site is larger. The influence of each surrounding site is extracted by adopting the convolutional neural module, and in combination with correlation analysis among the sites, more factors influencing the target site are added in the calculation process, and the influence of comprehensive calculation on the target site is predicted by using the spatial information more accurately, so that the calculated environmental pollutants of the target site are more accurate.
Optionally, the extracting key data affecting the environmental data of the target site based on the impact value includes:
And inputting the first influence value and the second influence value into a convolutional neural network to obtain the key data influencing the environmental data of the target site.
The first influence value and the second influence value are combined and fed into a convolutional neural model for feature extraction, and hidden layer output H (as an output result in FIG. 2) is obtained, wherein the hidden layer output H represents influence factors abstracted from historical time and space information.
Optionally, the acquiring the target site and the environmental data of each site within the preset range of the target site includes:
and acquiring the target site and the environmental data of each site in the preset range of the target site according to the preset environmental data acquisition time period.
Further, in order to make the detection result more time-efficient and accurate, the user may set a period of time for acquiring the environmental information of each site around the target site and acquiring the environmental data, for example, the period may be one hour, or may be one day or several monitoring periods in one day, such as: three time periods of 00:00 to 5:00, 15:00 to 17:00, and 20:00 to 11:00 in one day are set as the detection time periods. And estimating the environmental data of the target site based on the environmental data acquired in the preset time period.
Preferably, different time periods can be set according to different areas or different seasons, so that the estimated environmental information is more consistent with the local situation, and further pollutants with larger influence on the environment can be accurately predicted.
In one implementation manner, as shown in fig. 4, an environmental data prediction apparatus provided in an embodiment of the present application includes: the system comprises a key data extraction module 100, a first estimated environmental data determination module 200 and a coupling module 300.
The key data extraction module 100 is configured to obtain the target site and environmental data of each site within a preset range of the target site, and extract key data affecting the target site based on the convolutional neural network and the environmental data;
The first estimated environmental data determining module 200 is configured to obtain a first estimated model, a target site, and historical environmental data of each site within a preset range of the target site, and determine first estimated environmental data based on the first estimated model and the historical environmental data;
And the coupling module 300 is configured to couple the key data and the first estimated environmental data, and obtain second estimated environmental data of the target site.
Further, as shown in fig. 4, the apparatus for estimating environmental data provided in the embodiment of the present invention further includes:
A coupling unit 310, configured to couple the key data and the first estimated environmental data through an attention mechanism, so as to obtain coupling data; the convolutional neural calculation unit 320 is configured to input the coupling data into a convolutional neural network for operation, so as to obtain the second estimated environmental data. An influence value determining unit 110, configured to obtain a target site and environmental data of each site within a preset range of the target site, and determine an influence value of the environmental data of each site on the environmental data of the target site based on a convolutional neural network; and a key data extraction unit 120, configured to extract key data affecting the environmental data of the target site based on the impact value. A first environmental data obtaining unit 111, configured to obtain environmental data of a target site and each site within a preset range of the target site, where the environmental data includes meteorological data and pollutant data of at least one pollutant; a first influence value determining unit 112 for determining a first influence value of the environmental data of the target site on the target contaminant based on the meteorological data of the target site, the contaminant data of the target contaminant, and an attention mechanism; a second influence value determining unit 113, configured to determine a second influence value of each site on the target pollutant of the target site based on the meteorological data of each site, the pollutant data of the target pollutant, the correlation between the target site and each site, and the attention mechanism within a preset range. A key data obtaining unit 121, configured to input the first influence value and the second influence value into a convolutional neural network, and obtain the key data that affects the environmental data of the target site. The second environmental data obtaining unit 130 is configured to obtain the environmental data of the target site and each site within the preset range of the target site according to a preset environmental data obtaining time period.
The embodiment of the environmental data estimation method can be implemented by the environmental data estimation device provided by the embodiment of the invention, and specific function implementation is shown in the description of the embodiment of the method and is not repeated here.
The embodiment of the invention provides a computer readable storage medium, and a computer program is stored on the computer readable storage medium, and when the program is executed by a processor, the method for estimating environmental data according to any technical scheme is realized. The computer readable storage medium includes, but is not limited to, any type of disk including floppy disks, hard disks, optical disks, CD-ROMs, and magneto-optical disks, ROMs (Read-Only memories), RAMs (Random AcceSS Memory, random access memories), EPROMs (EraSable Programmable Read-Only memories), EEPROMs (ELECTRICALLY ERASABLE PROGRAMMABLE READ-Only memories), flash memories, magnetic cards, or optical cards. That is, a storage device includes any medium that stores or transmits information in a form readable by a device (e.g., computer, cell phone), and may be read-only memory, magnetic or optical disk, etc.
According to the embodiment of the application, the estimated time of the environmental pollutant change trend is shortened by combining the attention mechanism and the traditional environmental pollutant estimation model (such as a digital model, an RNN model, a Sequence2Sequence and the like) on the basis of the convolutional neural network, and the environmental pollutant change trend is obtained more accurately, so that the accuracy of the estimated environmental pollutant is improved; the environment data prediction method provided by the embodiment of the application comprises the following steps: acquiring a target site and environment data of each site within a preset range of the target site, and extracting key data affecting the target site based on a convolutional neural network and the environment data; acquiring a first pre-estimated model, a target site and historical environment data of each site within a preset range of the target site, and determining the first pre-estimated environment data based on the first pre-estimated model and the historical environment data; and coupling the key data and the first estimated environmental data to obtain second estimated environmental data of the target site. In the embodiment of the application, the condition of the environmental pollutants of the target site can be predicted more accurately. Acquiring the target site and the environmental data of each site within the preset range of the target site, wherein the environmental data comprises: contaminant data, weather data. The contaminant data includes: PM2.5, PM10, O 3、NO2、SO2, CO, etc., the meteorological data includes: weather data such as temperature, barometric pressure, relative humidity, wind speed, wind direction, weather conditions, and the like. The preset range is set by monitoring personnel. And inputting the environmental data into a convolutional neural network, and extracting key data affecting the target site based on the environmental data of the target site. And performing feature extraction on each site around the target site and the target site in time sequence by using a convolutional neural network, namely extracting environment data, and particularly, acquiring a wider time receptive field by using a cavity convolution mode so as to acquire comprehensive time dimension information. And then, obtaining the influence of the environmental data of the surrounding sites on the different environmental information of the target site through the softmax layer. In order to more accurately estimate the change condition of the pollutant value of the target site. The method comprises the steps of obtaining a first pre-estimated model, wherein the model can comprehensively consider the influence of space information on target monitoring points, give a relatively accurate long-term pollutant change trend of a target site, and the first pre-estimated model can comprise one or more of a numerical model, an RNN model and a Sequence2Sequence model. And inputting all the historical environmental data of all the sites into the first estimation model to obtain first estimation environmental data. It should be noted that, the key data extraction process and the first estimated environmental data determination process are not sequential, that is, the two processes may be performed simultaneously, or the first estimated environmental data determination process may be performed by advanced key data extraction, or the two processes may be exchanged. And (3) predicting a more accurate transformation trend of pollutants in the environment of the target site, and coupling and superposing long-term data (first predicted environment data) and short-term data (key data) with long period so as to obtain second predicted environment data of the target site. Specifically, coupling is performed through an attention mechanism, and the coupled data is input into a convolutional neural network to obtain second estimated environmental data.
In addition, in another embodiment, the present invention further provides a server, as shown in fig. 5, where the server processor 503, the memory 505, the input unit 507, the display unit 509, and other devices. Those skilled in the art will appreciate that the structural elements shown in fig. 5 do not constitute a limitation on all servers, and may include more or fewer components than shown, or may combine certain components. The memory 505 may be used to store an application 501 and various functional modules, and the processor 503 runs the application 501 stored in the memory 505 to perform various functional applications and data processing of the device. The memory 505 may be an internal memory or an external memory, or include both internal and external memories. The internal memory may include read-only memory (ROM), programmable ROM (PROM), electrically Programmable ROM (EPROM), electrically Erasable Programmable ROM (EEPROM), flash memory, or random access memory. The external memory may include a hard disk, floppy disk, ZIP disk, U-disk, tape, etc. The disclosed memory includes, but is not limited to, these types of memory. The memory 505 of the present disclosure is by way of example only and not by way of limitation.
The input unit 507 is used for receiving input of a signal, and site information input by a user. The input unit 507 may include a touch panel and other input devices. The touch panel can collect touch operations on or near the client (such as operations of the client on or near the touch panel using any suitable object or accessory such as a finger, a stylus, etc.), and drive the corresponding connection device according to a preset program; other input devices may include, but are not limited to, one or more of a physical keyboard, function keys (e.g., play control keys, switch keys, etc.), a trackball, mouse, joystick, etc. The display unit 509 may be used to display information input by a client or information provided to the client and various menus of the computer device. The display unit 509 may take the form of a liquid crystal display, an organic light emitting diode, or the like. The processor 503 is the control center of the computer device, connecting the various parts of the overall computer using various interfaces and lines, performing various functions and processing data by running or executing software programs and/or modules stored in the memory 503, and invoking data stored in the memory. The one or more processors 503 shown in fig. 5 are capable of executing, implementing, the functions of the critical data extraction module 100, the functions of the first estimated environmental data determination module 200, the functions of the coupling module 300, the functions of the coupling unit 310, the functions of the convolutional neural calculation unit 320, the functions of the influence value determination unit 110, the functions of the critical data extraction unit 120, the functions of the first environmental data acquisition unit 111, the functions of the first influence value determination unit 112, the functions of the second influence value determination unit 113, the functions of the critical data acquisition unit 121, the functions of the second environmental data acquisition unit 130 shown in fig. 4.
In one embodiment, the server includes one or more processors 503 and one or more memories 505, one or more applications 501, wherein the one or more applications 501 are stored in the memory 505 and configured to be executed by the one or more processors 503, and the one or more applications 301 are configured to perform the environmental data prediction method described in the above embodiments.
According to the embodiment of the application, the embodiment of the environmental data estimation method can be realized, and the estimated time of the environmental pollutant change trend is shortened by combining the attention mechanism and the traditional environmental pollutant estimation model (such as a digital model, an RNN model, a Sequence2Sequence and the like) on the basis of the convolutional neural network, so that the environmental pollutant change trend is obtained more accurately, and the accuracy of the estimated environmental pollutant is further improved; the environment data prediction method provided by the embodiment of the application comprises the following steps: acquiring a target site and environment data of each site within a preset range of the target site, and extracting key data affecting the target site based on a convolutional neural network and the environment data; acquiring a first pre-estimated model, a target site and historical environment data of each site within a preset range of the target site, and determining the first pre-estimated environment data based on the first pre-estimated model and the historical environment data; and coupling the key data and the first estimated environmental data to obtain second estimated environmental data of the target site. In the embodiment of the application, the condition of the environmental pollutants of the target site can be predicted more accurately. Acquiring the target site and the environmental data of each site within the preset range of the target site, wherein the environmental data comprises: contaminant data, weather data. The contaminant data includes: PM2.5, PM10, O 3、NO2、SO2, CO, etc., the meteorological data includes: weather data such as temperature, barometric pressure, relative humidity, wind speed, wind direction, weather conditions, and the like. The preset range is set by monitoring personnel. And inputting the environmental data into a convolutional neural network, and extracting key data affecting the target site based on the environmental data of the target site. And performing feature extraction on each site around the target site and the target site in time sequence by using a convolutional neural network, namely extracting environment data, and particularly, acquiring a wider time receptive field by using a cavity convolution mode so as to acquire comprehensive time dimension information. And then, obtaining the influence of the environmental data of the surrounding sites on the different environmental information of the target site through the softmax layer. In order to more accurately estimate the change condition of the pollutant value of the target site. The method comprises the steps of obtaining a first pre-estimated model, wherein the model can comprehensively consider the influence of space information on target monitoring points, give a relatively accurate long-term pollutant change trend of a target site, and the first pre-estimated model can comprise one or more of a numerical model, an RNN model and a Sequence2Sequence model. And inputting all the historical environmental data of all the sites into the first estimation model to obtain first estimation environmental data. It should be noted that, the key data extraction process and the first estimated environmental data determination process are not sequential, that is, the two processes may be performed simultaneously, or the first estimated environmental data determination process may be performed by advanced key data extraction, or the two processes may be exchanged. And (3) predicting a more accurate transformation trend of pollutants in the environment of the target site, and coupling and superposing long-term data (first predicted environment data) and short-term data (key data) with long period so as to obtain second predicted environment data of the target site. Specifically, coupling is performed through an attention mechanism, and the coupled data is input into a convolutional neural network to obtain second estimated environmental data.
The server provided by the embodiment of the present invention can implement the embodiment of the method for estimating environmental data provided above, and specific functional implementation is described in the method embodiment and is not repeated here.
The foregoing is only a partial embodiment of the present invention, and it should be noted that it will be apparent to those skilled in the art that modifications and adaptations can be made without departing from the principles of the present invention, and such modifications and adaptations are intended to be comprehended within the scope of the present invention.
Claims (7)
1. An environmental data prediction method, comprising:
Acquiring the target site and the environmental data of each site within the preset range of the target site, and extracting key data affecting the target site based on the convolutional neural network and the environmental data, wherein the method comprises the following steps: acquiring the environmental data of the target site and each site within a preset range of the target site, wherein the environmental data comprises meteorological data and pollutant data of at least one pollutant; determining a first impact value of the environmental data of the target site on the target contaminant based on the meteorological data of the target site, the contaminant data of the target contaminant, and an attention mechanism; determining a second influence value of each station on the target pollutant of the target station based on the meteorological data of each station, the pollutant data of the target pollutant, the correlation between the target station and each station and an attention mechanism in a preset range; inputting the first influence value and the second influence value into a convolutional neural network to obtain the key data influencing the environmental data of the target site;
Acquiring a first pre-estimated model, a target site and historical environment data of each site within a preset range of the target site, and determining the first pre-estimated environment data based on the first pre-estimated model and the historical environment data;
and coupling the key data and the first estimated environmental data to obtain second estimated environmental data of the target site.
2. The method of claim 1, wherein the coupling the key data and the first predicted environmental data to obtain second predicted environmental data for the target site comprises:
Coupling the key data and the first estimated environmental data through an attention mechanism to obtain coupling data;
And inputting the coupling data into a convolutional neural network for operation to obtain the second estimated environmental data.
3. The environmental data prediction method according to claim 1 or 2, wherein the obtaining the environmental data of the target site and each site within the preset range of the target site includes:
and acquiring the target site and the environmental data of each site in the preset range of the target site according to the preset environmental data acquisition time period.
4. An environmental data prediction apparatus, comprising:
The key data extraction module is used for acquiring the target site and the environment data of each site within the preset range of the target site, and extracting the key data affecting the target site based on the convolutional neural network and the environment data, and comprises the following steps: acquiring the environmental data of the target site and each site within a preset range of the target site, wherein the environmental data comprises meteorological data and pollutant data of at least one pollutant; determining a first impact value of the environmental data of the target site on the target contaminant based on the meteorological data of the target site, the contaminant data of the target contaminant, and an attention mechanism; determining a second influence value of each station on the target pollutant of the target station based on the meteorological data of each station, the pollutant data of the target pollutant, the correlation between the target station and each station and an attention mechanism in a preset range; inputting the first influence value and the second influence value into a convolutional neural network to obtain the key data influencing the environmental data of the target site;
the first estimated environmental data determining module is used for acquiring a first estimated model, a target site and historical environmental data of each site within a preset range of the target site, and determining the first estimated environmental data based on the first estimated model and the historical environmental data;
And the coupling module is used for coupling the key data and the first estimated environmental data to obtain second estimated environmental data of the target site.
5. The environmental data prediction apparatus of claim 4, wherein the coupling module comprises:
the coupling unit is used for coupling the key data and the first estimated environmental data through an attention mechanism to obtain coupling data;
and the convolutional neural calculation unit is used for inputting the coupling data into a convolutional neural network to perform operation so as to obtain the second estimated environmental data.
6. A computer readable storage medium, wherein a computer program is stored on the computer readable storage medium, which when executed by a processor, implements the environmental data prediction method of any one of claims 1 to 3.
7. A server, comprising:
One or more processors;
A memory;
One or more applications, wherein the one or more applications are stored in the memory and configured to be executed by the one or more processors, the one or more applications configured to perform the steps of the environmental data prediction method of any one of claims 1 to 3.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811348276.3A CN109685246B (en) | 2018-11-13 | 2018-11-13 | Environment data prediction method and device, storage medium and server |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811348276.3A CN109685246B (en) | 2018-11-13 | 2018-11-13 | Environment data prediction method and device, storage medium and server |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109685246A CN109685246A (en) | 2019-04-26 |
| CN109685246B true CN109685246B (en) | 2024-04-23 |
Family
ID=66184650
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811348276.3A Active CN109685246B (en) | 2018-11-13 | 2018-11-13 | Environment data prediction method and device, storage medium and server |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109685246B (en) |
Families Citing this family (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110210681B (en) * | 2019-06-11 | 2023-06-27 | 西安电子科技大学 | Prediction method of PM2.5 value of monitoring station based on distance |
| CN110347192B (en) * | 2019-06-17 | 2020-11-03 | 武汉理工大学 | Glass furnace temperature intelligent prediction control method based on attention mechanism and self-encoder |
| CN112986492B (en) * | 2019-12-17 | 2023-02-03 | 华为云计算技术有限公司 | Method and device for establishing gas concentration prediction model |
| CN113125635B (en) * | 2019-12-31 | 2023-04-28 | 北京蛙鸣华清环保科技有限公司 | Atmospheric pollution early warning method and device and readable storage medium |
| CN111443163A (en) * | 2020-03-10 | 2020-07-24 | 中国科学院深圳先进技术研究院 | Interpolation method and device for ozone missing data and interpolation equipment |
| CN112578089B (en) * | 2020-12-24 | 2023-04-07 | 河北工业大学 | Air pollutant concentration prediction method based on improved TCN |
| CN112801423B (en) * | 2021-03-29 | 2021-07-20 | 北京英视睿达科技有限公司 | Method and device for identifying abnormity of air quality monitoring data and storage medium |
| CN116664582B (en) * | 2023-08-02 | 2023-10-27 | 四川公路桥梁建设集团有限公司 | Road surface detection method and device based on neural vision network |
| CN117251700B (en) * | 2023-11-18 | 2024-01-26 | 拓普思传感器(太仓)有限公司 | Artificial intelligence-based environmental monitoring sensor data analysis method and system |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2002049722A (en) * | 2000-08-04 | 2002-02-15 | Sumitomo Metal Mining Co Ltd | Environment-monitoring apparatus and environment data providing method |
| CN106651036A (en) * | 2016-12-26 | 2017-05-10 | 东莞理工学院 | Air quality forecasting system |
| CN107943928A (en) * | 2017-11-21 | 2018-04-20 | 清华大学 | A kind of ozone concentration Forecasting Methodology and system based on space-time data statistical learning |
| WO2018153322A1 (en) * | 2017-02-23 | 2018-08-30 | 北京市商汤科技开发有限公司 | Key point detection method, neural network training method, apparatus and electronic device |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7233933B2 (en) * | 2001-06-28 | 2007-06-19 | Microsoft Corporation | Methods and architecture for cross-device activity monitoring, reasoning, and visualization for providing status and forecasts of a users' presence and availability |
| US20120054024A1 (en) * | 2010-08-31 | 2012-03-01 | Paul Anthony Polizzotto | Apparatuses, methods and systems for matching environmental advertisements with environmental projects |
| US20110145074A1 (en) * | 2010-08-31 | 2011-06-16 | Paul Anthony Polizzotto | Apparatuses, methods and systems for an environmental advertising, financing and management platform |
| US10438125B2 (en) * | 2015-11-12 | 2019-10-08 | International Business Machines Corporation | Very short-term air pollution forecasting |
| US10340983B2 (en) * | 2016-12-09 | 2019-07-02 | At&T Intellectual Property I, L.P. | Method and apparatus for surveying remote sites via guided wave communications |
-
2018
- 2018-11-13 CN CN201811348276.3A patent/CN109685246B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2002049722A (en) * | 2000-08-04 | 2002-02-15 | Sumitomo Metal Mining Co Ltd | Environment-monitoring apparatus and environment data providing method |
| CN106651036A (en) * | 2016-12-26 | 2017-05-10 | 东莞理工学院 | Air quality forecasting system |
| WO2018153322A1 (en) * | 2017-02-23 | 2018-08-30 | 北京市商汤科技开发有限公司 | Key point detection method, neural network training method, apparatus and electronic device |
| CN107943928A (en) * | 2017-11-21 | 2018-04-20 | 清华大学 | A kind of ozone concentration Forecasting Methodology and system based on space-time data statistical learning |
Non-Patent Citations (1)
| Title |
|---|
| 大气超级站网建设及在江苏区域的集成应用实践;汤莉莉;刀;秦玮;余家燕;张祥志;王晨波;杜嵩山;张;陈诚;秦艳红;丁爱军;;中国环境监测(第05期);第20-23页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109685246A (en) | 2019-04-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109685246B (en) | Environment data prediction method and device, storage medium and server | |
| Fu et al. | Correlation analysis and augmentation of samples for a bidirectional gate recurrent unit network for the remaining useful life prediction of bearings | |
| Peng et al. | Machine learning and deep learning modeling and simulation for predicting PM2. 5 concentrations | |
| CN113570867B (en) | Urban traffic state prediction method, device, equipment and readable storage medium | |
| AU2020102518A4 (en) | A method of gas sensor calibration based on linear optimization | |
| Su et al. | An improved random forest model for the prediction of dam displacement | |
| CN113009077B (en) | Gas detection method, gas detection device, electronic equipment and storage medium | |
| CN112508265A (en) | Time and activity multi-task prediction method and system for business process management | |
| CN112668238B (en) | Rainfall processing method, rainfall processing device, rainfall processing equipment and storage medium | |
| CN110795522B (en) | Method and device for predicting track position of mobile user | |
| CN115654381A (en) | Water supply pipeline leakage detection method based on graph neural network | |
| CN114217025B (en) | Analysis method for evaluating influence of meteorological data on air quality concentration prediction | |
| Zu et al. | A simple gated recurrent network for detection of power quality disturbances | |
| Yang et al. | A novel hybrid prediction model for PM2. 5 concentration based on decomposition ensemble and error correction | |
| CN108694472B (en) | Prediction error extreme value analysis method, device, computer equipment and storage medium | |
| CN114519420A (en) | Air quality prediction method and device, computer equipment and storage medium | |
| Li et al. | A new PM2. 5 concentration forecasting system based on AdaBoost‐ensemble system with deep learning approach | |
| CN112990530B (en) | Regional population quantity prediction method, regional population quantity prediction device, electronic equipment and storage medium | |
| Lei et al. | Prediction of PM2. 5 concentration considering temporal and spatial features: A case study of Fushun, Liaoning Province | |
| CN117189309A (en) | Noise reduction method and device for intelligent computation center diesel generator set | |
| CN117093477A (en) | Software quality assessment method and device, computer equipment and storage medium | |
| CN116522594A (en) | Time self-adaptive transient stability prediction method and device based on convolutional neural network | |
| CN115391746A (en) | Interpolation method, device, electronic device and medium for meteorological element data | |
| CN111027680B (en) | Monitoring quantity uncertainty prediction method and system based on variational self-encoder | |
| Gao et al. | An attention-based spatio-temporal model for methane concentration forecasting in coal mine |
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 |