CN114819560A - Three-in-one place identification method, identification model construction method and related equipment - Google Patents
Three-in-one place identification method, identification model construction method and related equipment Download PDFInfo
- Publication number
- CN114819560A CN114819560A CN202210384496.1A CN202210384496A CN114819560A CN 114819560 A CN114819560 A CN 114819560A CN 202210384496 A CN202210384496 A CN 202210384496A CN 114819560 A CN114819560 A CN 114819560A
- Authority
- CN
- China
- Prior art keywords
- data
- place
- model
- identification model
- signaling
- 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
- 238000000034 method Methods 0.000 title claims abstract description 51
- 238000010276 construction Methods 0.000 title claims abstract description 28
- 239000013598 vector Substances 0.000 claims abstract description 93
- 230000011664 signaling Effects 0.000 claims abstract description 60
- 238000012549 training Methods 0.000 claims description 28
- 238000003860 storage Methods 0.000 claims description 21
- 230000004927 fusion Effects 0.000 claims description 20
- 238000013527 convolutional neural network Methods 0.000 claims description 15
- 238000000605 extraction Methods 0.000 claims description 12
- 238000013528 artificial neural network Methods 0.000 claims description 6
- 238000004590 computer program Methods 0.000 claims description 5
- 238000007405 data analysis Methods 0.000 claims description 4
- 230000000306 recurrent effect Effects 0.000 claims description 4
- 238000005070 sampling Methods 0.000 claims description 4
- 230000004931 aggregating effect Effects 0.000 claims description 2
- 125000004122 cyclic group Chemical group 0.000 claims description 2
- 230000005611 electricity Effects 0.000 description 14
- 239000007789 gas Substances 0.000 description 13
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 12
- 238000010586 diagram Methods 0.000 description 8
- 230000002354 daily effect Effects 0.000 description 6
- 230000006870 function Effects 0.000 description 6
- 238000010801 machine learning Methods 0.000 description 6
- 230000008569 process Effects 0.000 description 6
- 238000004891 communication Methods 0.000 description 5
- 238000004458 analytical method Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 238000012544 monitoring process Methods 0.000 description 4
- 230000002159 abnormal effect Effects 0.000 description 2
- 230000004308 accommodation Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000003203 everyday effect Effects 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- VNWKTOKETHGBQD-UHFFFAOYSA-N methane Chemical compound C VNWKTOKETHGBQD-UHFFFAOYSA-N 0.000 description 2
- 230000002123 temporal effect Effects 0.000 description 2
- 206010000369 Accident Diseases 0.000 description 1
- 230000006978 adaptation Effects 0.000 description 1
- 230000002776 aggregation Effects 0.000 description 1
- 238000004220 aggregation Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- -1 electricity Substances 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000003345 natural gas Substances 0.000 description 1
- 238000003058 natural language processing Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005192 partition Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 235000013311 vegetables Nutrition 0.000 description 1
Images
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
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0635—Risk analysis of enterprise or organisation activities
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
- G06F18/253—Fusion techniques of extracted features
-
- 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
- G06N3/08—Learning methods
-
- 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
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Human Resources & Organizations (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Economics (AREA)
- Data Mining & Analysis (AREA)
- Health & Medical Sciences (AREA)
- Strategic Management (AREA)
- Entrepreneurship & Innovation (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- Marketing (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Biophysics (AREA)
- Educational Administration (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Bioinformatics & Computational Biology (AREA)
- Development Economics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Biomedical Technology (AREA)
- Game Theory and Decision Science (AREA)
- Computational Linguistics (AREA)
- Evolutionary Biology (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- Primary Health Care (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a three-in-one place identification method, an identification model construction method and related equipment, which are used for acquiring life energy use data, signaling data and network consumption data of a three-in-one place to be identified; extracting the characteristic vectors of the various data, inputting the characteristic vectors into a three-in-one place recognition model for recognition to obtain risk scores corresponding to the three-in-one places to be recognized, wherein the three-in-one place recognition model is a model constructed by performing space-time federal and meta learning by using sample data with space-time attributes collected by a data owner; and if the risk score is larger than the preset risk score, determining that potential safety hazards exist in the three-in-one place to be identified. In the scheme, the data difference and the data space-time attribute of three-in-one places in different industries are fully considered, a unified identification model is constructed on the basis of data with the space-time attribute in the three-in-one places in advance, and the three-in-one places with potential safety hazards are accurately and effectively determined.
Description
Technical Field
The invention relates to the technical field of data analysis, in particular to a three-in-one place identification method, an identification model construction method and related equipment.
Background
The three-in-one place is formed by mixedly arranging the personnel accommodation place and places such as production, storage, operation and the like in the same building, and an effective fireproof partition is not arranged between the accommodation place and the places with other functions. Therefore, in order to prevent fire accidents in three-in-one sites, it is important to first identify which sites belong to the category of the three-in-one sites.
The existing identification mode of the three-in-one place mainly analyzes the electricity utilization characteristics of the place through electric power monitoring equipment, and when the abnormal electricity utilization condition of a user is detected, the alarm information is pushed by a background, so that decision assistance is provided for relevant management departments. However, this identification method has the following disadvantages: firstly, the analysis data is single; secondly, the differences of three-in-one places in different industries are easily ignored by adopting predefined rule model analysis, and the three-in-one places with different characteristics cannot be effectively identified.
Therefore, the existing three-in-one place identification method cannot accurately and effectively identify the three-in-one places.
Disclosure of Invention
In view of this, embodiments of the present invention provide a three-in-one location identification method, an identification model construction method, and related devices, so as to solve the problem that the existing three-in-one location cannot be accurately and effectively identified.
In order to achieve the above purpose, the embodiments of the present invention provide the following technical solutions:
the first aspect of the embodiment of the invention discloses a three-in-one place identification method, which comprises the following steps:
acquiring life energy use data, signaling data and network consumption data of a three-in-one place to be identified;
extracting the characteristic vectors of the life energy use data, the signaling data and the network consumption data, and inputting the characteristic vectors into a three-in-one place identification model, wherein the three-in-one place identification model is a model constructed by performing space-time federal learning by using sample data with space-time attributes collected by a data owner;
identifying the characteristic vectors by using the three-in-one place identification model to obtain risk scores corresponding to the three-in-one places to be identified;
and if the risk score is larger than the preset risk score, determining that potential safety hazards exist in the three-in-one place to be identified.
Optionally, extract the feature vector of life energy usage data, signaling data and network consumption data, and will the feature vector input trinity place identification model includes:
converting the life energy use data into life characteristic vectors through a transformer model;
converting the signaling data into a signaling feature vector by using a Convolutional Neural Network (CNN) and a cyclic neural network (RNN);
acquiring address information in the network consumption data, analyzing the consumption data under the address information, and constructing a corresponding consumption characteristic vector based on the consumption data;
and features are fused with the life feature vector, the signaling feature vector and the consumption feature vector, and obtained fusion features are input into the three-in-one place recognition model.
Optionally, the method further includes:
and sending alarm prompt information for prompting the existence of the potential safety hazard in the three-in-one place to be identified.
The second aspect of the embodiment of the invention discloses a method for constructing a three-in-one site recognition model, which comprises the following steps:
extracting characteristic vectors corresponding to sample data respectively collected by a data owner of the combined training three-in-one site recognition model, wherein the sample data comprises life energy use data, signaling data and network consumption data;
performing feature fusion and storage on the feature vectors based on federal learning to obtain a data set corresponding to the three-in-one place;
obtaining gradient updating parameters fed back by each data owner, wherein the gradient updating parameters are obtained by performing space-time federal learning training on each data owner based on the data set;
based on gradient update parameter updates initial multitask identification model, obtains trinity place identification model, initial multitask identification model is based on each initial gradient parameter construction that the data owner uploaded, the input of trinity place identification model is characteristic data, and the output is the risk score.
Optionally, the feature vector that the data owner who extracts the three-in-one site recognition model of joint training corresponds to the sample data that each collected includes:
acquiring sample data collected by a data owner of a current combined training three-in-one site recognition model in a sampling period;
if the sample data is life energy use data, converting the life energy use data into life energy feature vectors through a transformer model;
if the sample data is signaling data, converting the signaling data into signaling characteristic vectors by using a Convolutional Neural Network (CNN) and a Recurrent Neural Network (RNN);
if the sample data is network consumption data, acquiring address information in the network consumption data, analyzing consumption data under the address information, and constructing a corresponding consumption characteristic vector based on the consumption data.
Optionally, update the initial multitask identification model based on the gradient update parameter until obtaining a three-in-one place identification model, including:
aggregating the gradient updating parameters to obtain a new gradient;
and updating the initial multi-task recognition model based on the new gradient, feeding the updated multi-task recognition model back to each data owner, and updating the local model by each data owner.
The third aspect of the embodiment of the invention discloses a three-in-one place recognition device, which comprises:
the three-in-one site recognition model is constructed according to the construction method of the three-in-one site recognition model disclosed by the second aspect of the embodiment of the invention;
the system comprises a characteristic acquisition module, a three-in-one place identification module and a three-in-one place identification module, wherein the characteristic acquisition module is used for acquiring the life energy use data, the signaling data and the network consumption data of a three-in-one place to be identified, extracting the characteristic vectors of the life energy use data, the signaling data and the network consumption data, and inputting the characteristic vectors into the three-in-one place identification model;
the three-in-one place identification model is used for identifying the input characteristic vector to obtain a risk score corresponding to the three-in-one place to be identified;
and the determining module is used for determining that potential safety hazards exist in the three-in-one places to be identified if the risk score is larger than the preset risk score.
The fourth aspect of the embodiment of the invention discloses a device for constructing a three-in-one site recognition model, which comprises:
the system comprises an extraction module, a data acquisition module, a data analysis module and a data analysis module, wherein the extraction module is used for extracting characteristic vectors corresponding to sample data respectively collected by a data owner of a joint training three-in-one site recognition model, and the sample data comprises life energy use data, signaling data and network consumption data;
the fusion module is used for performing feature fusion and storage on the feature vectors based on federal learning to obtain a data set corresponding to the three-in-one places;
the training module is used for obtaining gradient updating parameters fed back by each data owner, and the gradient updating parameters are obtained by performing space-time federal learning training on each data owner based on the data set;
the updating module is used for updating the initial multi-task identification model based on the gradient updating parameters to obtain a three-in-one site identification model, the initial multi-task identification model is based on each initial gradient parameter uploaded by the data owner is constructed, the input of the three-in-one site identification model is characteristic data, and the output is a risk score.
The fifth aspect of the embodiment of the invention discloses an electronic device, which comprises a processor and a memory;
the memory for storing a computer program;
the processor is configured to implement the triple-play site identification method disclosed in the first aspect of the embodiment of the present invention or implement the construction method of the triple-play site identification model disclosed in the second aspect of the embodiment of the present invention when the computer program stored in the memory is called and executed.
A sixth aspect of the embodiments of the present invention discloses a computer storage medium, in which computer-executable instructions are stored, and when the computer-executable instructions are loaded and executed by a processor, the method for identifying a three-in-one location identification model as disclosed in the first aspect of the embodiments of the present invention is implemented, or the method for constructing a three-in-one location identification model as disclosed in the second aspect of the embodiments of the present invention is implemented.
The three-in-one place identification method, the identification model construction method and the related equipment provided by the embodiment of the invention have the following advantages or beneficial effects: acquiring life energy use data, signaling data and network consumption data of a three-in-one place to be identified; extracting the characteristic vectors of the various data, inputting the characteristic vectors into a three-in-one place recognition model for recognition to obtain risk scores corresponding to the three-in-one places to be recognized, wherein the three-in-one place recognition model is a model constructed by performing space-time federal and meta learning by using sample data with space-time attributes collected by a data owner; and if the risk score is larger than the preset risk score, determining that potential safety hazards exist in the three-in-one place to be identified. In the scheme, the data differences of three-in-one places in different industries and the characteristics of data space-time attributes of the three-in-one places are fully considered, and under the technical framework of space-time federation and meta-learning, a unified identification model is constructed in advance according to the life energy use data, the signaling data and the network consumption data of the three-in-one places in different industries to perform feature identification, so that the three-in-one places with potential safety hazards are accurately and effectively determined.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the provided drawings without creative efforts.
Fig. 1 is a schematic structural diagram of a construction system of a three-in-one site recognition model disclosed in an embodiment of the invention;
FIG. 2 is a schematic diagram of feature extraction performed on data collected by each data owner according to an embodiment of the present invention;
FIG. 3 is a spatiotemporal federated meta-learning training framework disclosed in an embodiment of the present invention;
FIG. 4 is a schematic flow chart of a method for constructing a three-in-one site recognition model according to an embodiment of the present invention;
FIG. 5 is a schematic flow chart of a method for identifying a three-in-one location according to an embodiment of the present invention;
fig. 6 is a schematic structural diagram of an electronic device according to an embodiment of the disclosure.
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 obtained by a person skilled in the art without making any creative effort based on the embodiments in the present invention, belong to the protection scope of the present invention.
In this application, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.
The terms "first," "second," "third," "fourth," and the like in the description and in the claims of the present application and in the above-described drawings are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It will be appreciated that the data so used may be interchanged under appropriate circumstances such that the embodiments described herein may be practiced otherwise than as specifically illustrated or described herein.
The following are terms of art used in the examples of the present invention:
federal learning, also known as federal machine learning, joint learning, league learning. Federal machine learning is a machine learning framework which can effectively help a plurality of organizations to perform data use and machine learning modeling under the condition of meeting the requirements of user privacy protection, data safety and government regulations.
Meta learning: the method is to summarize the essential rules of a series of similar tasks by learning the tasks. When a brand-new task is faced, fine adjustment can be performed according to the learned rule, and the quick adaptation can be realized.
According to the background technology, the existing identification mode of the three-in-one places mainly analyzes the electricity utilization characteristics of the places through the power monitoring equipment, the analysis data is single, and differences among the three-in-one places in different industries are not considered in the analysis model, so that the three-in-one places with different characteristics cannot be effectively and accurately identified.
Therefore, the embodiment of the invention discloses a three-in-one place identification method, an identification model construction method and related equipment, which fully consider the data differences of three-in-one places in different industries and the characteristics of the data space-time attributes of the three-in-one places, and construct a unified identification model for feature identification according to the life energy use data, signaling data and network consumption data of the three-in-one places in different industries under the technical framework of space-time federation and meta-learning, thereby realizing the accurate and effective determination of the three-in-one places with potential safety hazards. The details are explained in the following examples.
As shown in fig. 1, a schematic structural diagram of a system for constructing a three-in-one site recognition model disclosed in an embodiment of the present invention mainly includes: a server 11 and a data owner.
The data owner is deployed in the local of the data side needing to collect data, and includes but is not limited to a terminal 12 arranged in a government department, a terminal 13 arranged in a communication operator and a terminal 14 arranged in an internet company.
The government departments include, but are not limited to, the electric power department, the water conservancy department, and the natural gas department, and may also relate to other civilian departments.
The communication operator includes Unicom, Mobile and Telecommunications.
Because data among the data owners are kept secret, a data processing framework is built through federal machine learning, the privacy of each data owner is protected, meanwhile, data use and machine learning modeling are carried out, and a three-in-one place recognition model is obtained.
The terminal 12 is used for collecting and recording data of water consumption, electricity consumption, gas consumption and the like of a user every day. Whether the user has abnormal water consumption, electricity consumption, gas consumption and other conditions in a certain period is obtained by analyzing the data of water consumption, electricity consumption, gas consumption and the like of the user; or the type of a certain area can be determined by analyzing the data of water use, electricity use, gas use and the like in the area.
For example, there may be significant differences in water, electricity, and gas usage data per unit area for business and residential areas with a single function; the three-in-one place and the single-function place have different data such as water consumption, electricity consumption, gas consumption and the like.
The terminal 13 is used for recording communication data of the user. The communication data is signaling data having temporal and spatial attributes. The population activity in a certain area and the characteristics of the area can be obtained by analyzing the signaling data in the certain area in a certain period of time.
Similarly, the communication data of the business area and the residential area with single function and the three-in-one place can be different because of different places and different people flow.
The terminal 14 is used for recording network consumption data generated when a user performs corresponding operations based on the internet. The network consumption data includes, but is not limited to, consumption data generated based on the internet. For example, the online shopping data generated when the user performs online shopping, such as express delivery data generated when the merchant delivers goods. The characteristics of the area where a certain address is located can be obtained by analyzing the network consumption data under the address.
Similarly, the consumption data of single-function business and residential areas, as well as triple play areas, will vary.
In the embodiment of the present invention, N three-in-one places with determined industry types are used as sample places, and the server 11 respectively obtains data corresponding to the N three-in-one places by using the terminal 12, the terminal 13, and the terminal 14. And a three-in-one site recognition model is constructed under the technical framework of space-time federation and meta learning. The value of N is a positive integer greater than 1.
Firstly, for each three-in-one place, data from different data owners are subjected to feature extraction to obtain corresponding feature vectors.
Fig. 2 is a schematic diagram illustrating feature extraction of data collected by each data owner.
Data such as electricity and gas (gas and gas) for water use and the like for each hour each day collected by the terminal 11 is input as input data to a transform (natural language processing) model, and the data is converted into a corresponding feature vector X1 by the transform model. The feature vector not only contains basic statistics of water use, electricity use and gas use in each hour every day, but also contains correlation trend features of water use, electricity use and gas use in one day.
Signaling data acquired by the terminal 12 is converted into feature vectors in each period within 24 hours by using a Convolutional Neural Network (CNN), the feature vectors are used as input vectors of a Recurrent Neural Network (RNN), and a relationship between a period and a period is extracted through the RNN, so that a feature vector X2 representing the population activity condition of the place within 24 hours a day is finally obtained.
Based on the consumption data collected by the terminal 13, the consumption situation at the address of the place is analyzed. Specifically, firstly, analyzing and converting data such as daily product consumption period, frequency and usage amount in a normal period to construct a feature vector for describing the purchasing condition of daily products at the address; then, the purchase data of other non-daily supplies are analyzed to obtain the characteristics of the non-daily supplies at the site address. The daily article characteristics and the non-daily article characteristics are combined to form a consumption characteristic vector X3 for describing consumption information of the place.
And then, fusing the feature vectors extracted from the data provided by the different data owners based on the federal learning technology so as to obtain the cross-domain fusion features required by the monitoring of the three-in-one site. Storing the cross-domain fusion characteristics of the three-in-one places of the N industries to obtain a data set corresponding to the three-in-one places of each industry type. The data set comprises a support data set and a query data set, wherein the support data set is a labeled data set; the query dataset is an unlabeled dataset. In a specific task, query data is classified with a model that supports dataset learning.
In the embodiment of the present invention, the ways of feature fusion include, but are not limited to: and the modes of feature splicing, feature addition, feature dot product and the like.
And then, performing meta-model training on the feature vectors which are distributed on different data owners in a different way in a space-time federal metadata mode based on the data sets corresponding to the three-in-one places of the N different industries.
FIG. 3 illustrates a spatiotemporal federated meta-learning training framework disclosed in an embodiment of the present invention.
The three-in-one places with different N industries are divided into different tasks, for example, the three-in-one places engaged in catering operation, the three-in-one places engaged in vegetable storage and the like are all used as a classification task.
For each classification task, the three data owners corresponding to the data sets based on the three data owners (client, equivalent to terminal 12, terminal 13, and terminal 14), for example, the industry 1 feature data set of fig. 3, are government, Unicom, and Jingdong. And three data owners corresponding to the industry N characteristic data set are government departments, signaling data parties and Internet companies.
Local gradient updating is carried out by utilizing the support data set, and local calculation is carried out by utilizing the updated local gradient to obtain new parameters; and calculating a loss function and a new gradient by using the query data set, transmitting the new gradient to the server 11, and performing gradient aggregation and gradient update on the jointly trained three-in-one site recognition model.
The method is specifically shown in formula (1):
wherein α is a learning rate;is a gradient; theta is a parameter;for the loss function, there is a loss for each task; e is average meaning, and loss of a plurality of tasks is expected; t refers to different tasks; Γ refers to the set of tasks.
Under the condition that each classification task is completed, a trained three-in-one site recognition model is obtained, and the three-in-one site recognition model is used for calculating the three-in-one site risk score of the unknown site.
Based on the building system of the three-in-one site recognition model disclosed by the embodiment of the invention, the embodiment of the invention also discloses a building method of the three-in-one site recognition model, which aims at building corresponding three-in-one site recognition models for three-in-one sites of different industries, the method is suitable for the server 11 disclosed in the figure 1, and as shown in figure 4, the method mainly comprises the following steps:
s401: and extracting the characteristic vectors corresponding to the sample data respectively collected by the data possessor of the combined training three-in-one place recognition model.
In S401, the sample data includes life energy usage data, signaling data, and network consumption data.
In particular, the life energy usage data includes, but is not limited to, water usage data, electricity usage data, and gas usage data.
Network consumption data includes, but is not limited to, consumption data.
In the process of implementing S401 specifically, refer to a schematic diagram of feature extraction performed on data collected by each data owner as shown in fig. 2.
Firstly, acquiring sample data collected by a data owner of a current combined training three-in-one site recognition model in a sampling period.
Second, the type of sample data is confirmed.
And if the sample data is the life energy use data, converting the life energy use data into a life characteristic vector through a transformer model.
And if the sample data is signaling data, converting the signaling data into a signaling characteristic vector by using a CNN (common network node) and an RNN (radio network node).
If the sample data is network consumption data, acquiring address information in the network consumption data, analyzing consumption data under the address information, and constructing a corresponding consumption characteristic vector based on the consumption data.
S402: and performing feature fusion and storage on the feature vectors based on federal learning to obtain a data set corresponding to the three-in-one place.
In the specific implementation of S402, the following steps may be adopted: and fusing the feature vectors by feature fusion modes such as feature splicing, feature addition, feature dot product and the like, and storing to obtain a data set corresponding to the three-in-one place.
The data set includes a support data set and a query data set.
S403: and obtaining the gradient updating parameters fed back by each data owner.
In S403, the gradient update parameters are obtained by performing space-time federal learning training on the data set by each of the data owners.
Specifically, each data owner performs local gradient update by using a support data set, and performs local calculation by using the updated local gradient to obtain a new parameter; a loss function and a new gradient are calculated using the query dataset.
S404: and updating the initial multi-task identification model based on the gradient updating parameters to obtain a three-in-one site identification model.
In S404, the initial multi-tasking identification model is constructed based on the initial gradient parameters uploaded by each of the data owners.
The input of the three-in-one place recognition model is characteristic data, and the output is a risk score.
In the process of implementing S404, first, the gradient update parameters are aggregated to obtain a new gradient.
And then updating the initial multi-task recognition model based on the new gradient, feeding the updated multi-task recognition model back to each data owner, and updating the local model by each data owner.
Based on the construction method of the three-in-one place identification model disclosed by the embodiment of the invention, the data differences of the three-in-one places in different industries and the characteristics of the data space-time attributes of the three-in-one places are fully considered, and a unified identification model is constructed according to the living energy use data, the signaling data and the network consumption data of the three-in-one places in different industries under the technical framework of space-time federation and meta-learning, so that the three-in-one places with potential safety hazards are accurately and effectively determined.
Based on the construction method of the three-in-one site recognition model disclosed by the embodiment of the invention, the embodiment of the invention also correspondingly discloses a construction device of the three-in-one site recognition model, and the device mainly comprises: an extraction module 21, a fusion module 22, a training module 23 and an update module 24.
The extraction module 21 is configured to extract feature vectors corresponding to sample data collected by data owners of the joint training three-in-one site recognition model, where the sample data includes life energy usage data, signaling data, and network consumption data.
And the fusion module 22 is used for carrying out feature fusion and storage on the feature vectors based on federal learning to obtain a data set corresponding to the three-in-one places.
And the training module 23 is configured to obtain gradient update parameters fed back by each data owner, where the gradient update parameters are obtained by performing space-time federal learning training on each data owner based on the data set.
Update module 24 for based on the gradient update parameter is updated initial multitask identification model, obtains trinity place identification model, initial multitask identification model is based on each the initial gradient parameter that the data owner uploaded constructs, the input of trinity place identification model is characteristic data, and the output is the risk score.
In a specific embodiment, the extraction module 21 is specifically configured to obtain sample data collected by data owners of the current joint training three-in-one site recognition model within a sampling period; if the sample data is life energy use data, converting the life energy use data into life characteristic vectors through a transformer model; if the sample data is signaling data, converting the signaling data into signaling characteristic vectors by using a Convolutional Neural Network (CNN) and a Recurrent Neural Network (RNN); if the sample data is network consumption data, acquiring address information in the network consumption data, analyzing consumption data under the address information, and constructing a corresponding consumption characteristic vector based on the consumption data.
In a specific embodiment, the updating module 24 is specifically configured to aggregate the gradient updating parameters to obtain a new gradient; and updating the initial multi-task identification model based on the new gradient to obtain a three-in-one place identification model, feeding the updated multi-task identification model back to each data owner, and updating the local model by each data owner.
Based on the three-in-one site identification model construction device disclosed by the embodiment of the invention, the data differences of three-in-one sites in different industries and the characteristics of data space-time attributes of the three-in-one sites are fully considered, and a unified identification model is constructed according to the life energy use data, the signaling data and the network consumption data of the three-in-one sites in different industries under the technical framework of space-time federation and meta-learning, so that the three-in-one sites with potential safety hazards are accurately and effectively determined.
Based on the construction method of the three-in-one site identification model disclosed by the embodiment of the invention and the three-in-one site identification model constructed by the construction model, the embodiment of the invention also discloses a three-in-one site identification method, which is suitable for the three-in-one site identification model disclosed by the embodiment, and as shown in fig. 5, the method mainly comprises the following steps:
s501: and acquiring the life energy use data, the signaling data and the network consumption data of the three-in-one place to be identified.
In S501, the life energy usage data includes, but is not limited to, water usage data, electricity usage data, and gas usage data.
Network consumption data includes, but is not limited to, consumption data.
S502: and extracting the feature vectors of the life energy use data, the signaling data and the network consumption data, and inputting the feature vectors into the three-in-one place identification model.
In the process of implementing S502 specifically, reference may be made to a schematic diagram of feature extraction performed on data collected by each data owner, shown in fig. 2.
Firstly, the life energy utilization data is converted into a life characteristic vector through a transformer model.
The signaling data is then converted into a signaling feature vector using the CNN and RNN.
Then, address information in the network consumption data is obtained, consumption data under the address information is analyzed, and a corresponding consumption feature vector is constructed based on the consumption data.
And finally, fusing the characteristics of the life characteristic vector, the signaling characteristic vector and the consumption characteristic vector, and inputting the obtained fusion characteristics into the three-in-one place recognition model.
S503: utilize trinity place identification model is right the eigenvector is discerned, obtains the risk score that the trinity place of waiting to discern corresponds.
In S503, the three-in-one site recognition model is obtained by training different data owners having temporal and spatial attributes in combination according to the construction method and apparatus disclosed in the embodiments of the present invention under the technical framework of spatio-temporal federation and meta learning.
The input of this trinity place determination module is characteristic data, and the output is the risk score. The feature data is a feature vector or a fusion feature.
S504: and judging whether the risk score is greater than or equal to a preset risk score, if so, executing S505, otherwise, returning to S501 to execute next identification.
S505: and if the risk score is larger than the preset risk score, determining that potential safety hazards exist in the three-in-one place to be identified.
In an embodiment, after confirming that there is the potential safety hazard in the trinity place of waiting to discern, can also generate warning prompt message, demonstrate on current system, perhaps send to monitoring personnel's terminal equipment on for the suggestion has the potential safety hazard the trinity place of waiting to discern.
The three-in-one place identification method disclosed by the embodiment of the invention is characterized in that a three-in-one place identification model constructed according to life energy use data, signaling data and network consumption data of three-in-one places in different industries is utilized to identify places to be identified under the technical framework of space-time federation and meta learning, if the risk score output by the three-in-one place identification model is greater than or equal to a preset risk score, whether the current three-in-one place to be identified has risks can be accurately, timely and effectively found, the relevant government departments can be helped to effectively eliminate potential safety hazards, and the property and life safety of residents can be protected.
Based on the three-in-one place identification method disclosed by the embodiment of the invention, the embodiment of the invention also discloses a three-in-one place identification device, which comprises the following steps: a feature acquisition module 31, a triple play location determination module 32, and a determination module 33.
The three-in-one place determination module 32 is constructed based on the construction method of the three-in-one place identification model disclosed in the embodiment of the invention.
The feature acquisition module 31 is used for acquiring the life energy use data, the signaling data and the network consumption data of the three-in-one place to be identified, extracting the feature vectors of the life energy use data, the signaling data and the network consumption data, and inputting the feature vectors into the three-in-one place identification model.
The three-in-one place recognition model 32 is used for recognizing the input characteristic vector to obtain the risk score corresponding to the three-in-one place to be recognized.
And the determining module is used for determining that potential safety hazards exist in the three-in-one places to be identified if the risk score is larger than the preset risk score.
In an embodiment, the feature obtaining module 31 is specifically configured to obtain life energy usage data, signaling data, and network consumption data of a to-be-identified three-in-one place, and convert the life energy usage data into a life feature vector through a transformer model; converting the signaling data into a signaling feature vector by using a CNN and an RNN; acquiring address information in the network consumption data, analyzing the consumption data under the address information, and constructing a corresponding consumption characteristic vector based on the consumption data; and features are fused with the life feature vector, the signaling feature vector and the consumption feature vector, and obtained fusion features are input into the three-in-one place recognition model.
In one embodiment, the triple play location identification apparatus further comprises: an alarm module 34.
The alarm module 34 is used for generating and sending alarm prompt information, and the alarm prompt information is used for prompting the three-in-one places to be identified, which have potential safety hazards.
The three-in-one place recognition device disclosed by the embodiment of the invention is used for recognizing places to be recognized by utilizing a three-in-one place recognition model constructed according to life energy use data, signaling data and network consumption data of three-in-one places in different industries under the technical framework of space-time federation and meta learning, and if the risk score output by the three-in-one place recognition model is greater than or equal to a preset risk score, whether risks exist in the current three-in-one places to be recognized can be accurately, timely and effectively found, the device helps relevant government departments to effectively eliminate potential safety hazards, and the property and life safety of residents are protected.
Based on the construction device of the three-in-one place recognition model and the three-in-one place recognition device disclosed by the embodiment of the disclosure, the modules can be realized by a hardware device consisting of a processor and a memory. Specifically, the modules are stored in a memory as program units, and a processor executes the program units stored in the memory to realize thread control.
The processor comprises a kernel, and the kernel calls a corresponding program unit from the memory. The kernel can be set to be one or more than one, and the construction of a three-in-one place recognition model or the recognition of the three-in-one place can be realized by adjusting the kernel parameters.
The embodiment of the disclosure provides a computer storage medium, wherein the storage medium comprises a construction program of a three-in-one place identification model, and the program is executed by a processor to realize the construction method of the three-in-one place identification model disclosed by the embodiment of the disclosure.
The embodiment of the disclosure provides a computer storage medium, wherein the storage medium comprises an identification program of three-in-one places, and the program is executed by a processor to realize the three-in-one place identification method disclosed by the embodiment of the invention.
The embodiment of the present disclosure, which is disclosed by the foregoing embodiments of the present invention, provides a processor for running a program, wherein the program runs to execute the method for constructing a three-in-one place recognition model or the method for recognizing a three-in-one place disclosed in the foregoing embodiments.
The disclosed embodiment of the present invention provides an electronic device, and as shown in fig. 6, is a schematic structural diagram of an electronic device provided in the disclosed embodiment of the present invention.
The electronic device 60 in the disclosed embodiments of the invention may be a server, a PC, a PAD, a mobile phone, etc.
The electronic device 60 comprises at least one processor 601, and at least one memory 602 coupled to the processor, and a bus 603.
The processor 601 and the memory 602 communicate with each other via a bus 603.
A processor 601 for executing the programs stored in the memory.
A storage 602, configured to store a program, where the program is at least configured to extract feature vectors corresponding to sample data collected by data owners of the joint training three-in-one location identification model, where the sample data includes life energy usage data, signaling data, and network consumption data; performing feature fusion and storage on the feature vectors based on federal learning to obtain a data set corresponding to the three-in-one place; obtaining gradient updating parameters fed back by each data owner, wherein the gradient updating parameters are obtained by performing space-time federal learning training on each data owner based on the data set; based on gradient update parameter updates initial multitask identification model, obtains trinity place identification model, initial multitask identification model is based on each initial gradient parameter construction that the data owner uploaded, the input of trinity place identification model is characteristic data, and the output is the risk score.
Or the program is at least used for acquiring the life energy use data, the signaling data and the network consumption data of the three-in-one place to be identified; extracting feature vectors of the life energy use data, the signaling data and the network consumption data, and inputting the feature vectors into the three-in-one place identification model; identifying the characteristic vectors by using the three-in-one place identification model to obtain risk scores corresponding to the three-in-one places to be identified; and if the risk score is larger than the preset risk score, determining that potential safety hazards exist in the three-in-one place to be identified.
In a typical configuration, a device includes one or more processors (CPUs), memory, and a bus. The device may also include input/output interfaces, network interfaces, and the like.
The memory may include volatile memory in a computer readable medium, Random Access Memory (RAM) and/or nonvolatile memory such as Read Only Memory (ROM) or flash memory (flash RAM), and the memory includes at least one memory chip. The memory is an example of a computer-readable medium.
Computer-readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), Read Only Memory (ROM), Electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), Digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information that can be accessed by a computing device. As defined herein, a computer readable medium does not include a transitory computer readable medium such as a modulated data signal and a carrier wave.
The embodiment of the invention also provides a computer program product which is suitable for executing the construction method of the three-in-one site identification model or the identification method of the three-in-one site disclosed by the embodiment of the invention when being executed on the electronic equipment.
The embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from the other embodiments. In particular, the system or system embodiments are substantially similar to the method embodiments and therefore are described in a relatively simple manner, and reference may be made to some of the descriptions of the method embodiments for related points. The above-described system and system embodiments are only illustrative, wherein the units described as separate parts may or may not be physically separate, and the parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.
Those of skill would further appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, computer software, or combinations of both, and that the various illustrative components and steps have been described above generally in terms of their functionality in order to clearly illustrate this interchangeability of hardware and software. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the implementation. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present disclosure.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present disclosure. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the disclosure. Thus, the present disclosure is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (10)
1. A three-in-one place identification method is characterized by comprising the following steps:
acquiring life energy use data, signaling data and network consumption data of a three-in-one place to be identified;
extracting the characteristic vectors of the life energy use data, the signaling data and the network consumption data, and inputting the characteristic vectors into a three-in-one place identification model, wherein the three-in-one place identification model is a model constructed by performing space-time federal learning by using sample data with space-time attributes collected by a data owner;
identifying the characteristic vectors by using the three-in-one place identification model to obtain risk scores corresponding to the three-in-one places to be identified;
and if the risk score is larger than the preset risk score, determining that potential safety hazards exist in the three-in-one place to be identified.
2. The method of claim 1, wherein extracting feature vectors of the renewable energy usage data, signaling data, and network consumption data and inputting the feature vectors into the triple play site identification model comprises:
converting the living energy use data into a living characteristic vector through a transformer model;
converting the signaling data into a signaling feature vector by using a Convolutional Neural Network (CNN) and a cyclic neural network (RNN);
acquiring address information in the network consumption data, analyzing the consumption data under the address information, and constructing a corresponding consumption characteristic vector based on the consumption data;
and features are fused with the life feature vector, the signaling feature vector and the consumption feature vector, and obtained fusion features are input into the three-in-one place recognition model.
3. The method according to claim 1 or 2, characterized in that the method further comprises:
and sending alarm prompt information for prompting the existence of the potential safety hazard in the three-in-one place to be identified.
4. A construction method of a three-in-one site recognition model is characterized by comprising the following steps:
extracting characteristic vectors corresponding to sample data respectively collected by a data owner of the combined training three-in-one site recognition model, wherein the sample data comprises life energy use data, signaling data and network consumption data;
performing feature fusion and storage on the feature vectors based on federal learning to obtain a data set corresponding to the three-in-one place;
obtaining gradient updating parameters fed back by each data owner, wherein the gradient updating parameters are obtained by performing space-time federal learning training on each data owner based on the data set;
based on gradient update parameter updates initial multitask identification model, obtains trinity place identification model, initial multitask identification model is based on each initial gradient parameter construction that the data owner uploaded, the input of trinity place identification model is characteristic data, and the output is the risk score.
5. The method of claim 4, wherein the extracting the feature vectors corresponding to the sample data collected by each of the data owners of the joint training three-in-one site recognition model comprises:
acquiring sample data collected by a data owner of a current combined training three-in-one site recognition model in a sampling period;
if the sample data is the life energy use data, converting the life energy use data into a life energy feature vector through a transformer model;
if the sample data is signaling data, converting the signaling data into signaling characteristic vectors by using a Convolutional Neural Network (CNN) and a Recurrent Neural Network (RNN);
if the sample data is network consumption data, acquiring address information in the network consumption data, analyzing consumption data under the address information, and constructing a corresponding consumption characteristic vector based on the consumption data.
6. The method of claim 4 or 5, wherein the updating the initial multi-tasking identification model based on the gradient update parameters until a three-in-one site identification model is obtained comprises:
aggregating the gradient updating parameters to obtain a new gradient;
and updating the initial multi-task recognition model based on the new gradient, feeding the updated multi-task recognition model back to each data owner, and updating the local model by each data owner.
7. A three-in-one venue identification device, the device comprising:
a three-in-one site recognition model constructed according to the construction method of the three-in-one site recognition model of claim 4;
the system comprises a characteristic acquisition module, a three-in-one place identification module and a three-in-one place identification module, wherein the characteristic acquisition module is used for acquiring the life energy use data, the signaling data and the network consumption data of a three-in-one place to be identified, extracting the characteristic vectors of the life energy use data, the signaling data and the network consumption data, and inputting the characteristic vectors into the three-in-one place identification model;
the three-in-one place identification model is used for identifying the input characteristic vector to obtain a risk score corresponding to the three-in-one place to be identified;
and the determining module is used for determining that potential safety hazards exist in the three-in-one places to be identified if the risk score is larger than the preset risk score.
8. The utility model provides a device for constructing trinity place discernment model which characterized in that, the device includes:
the system comprises an extraction module, a data acquisition module and a data analysis module, wherein the extraction module is used for extracting characteristic vectors corresponding to sample data respectively collected by a data owner of a combined training three-in-one site recognition model, and the sample data comprises life energy use data, signaling data and network consumption data;
the fusion module is used for performing feature fusion and storage on the feature vectors based on federal learning to obtain a data set corresponding to the three-in-one places;
the training module is used for obtaining gradient updating parameters fed back by each data owner, and the gradient updating parameters are obtained by performing space-time federal learning training on each data owner based on the data set;
the updating module is used for updating the initial multi-task identification model based on the gradient updating parameters to obtain a three-in-one site identification model, the initial multi-task identification model is based on each initial gradient parameter uploaded by the data owner is constructed, the input of the three-in-one site identification model is characteristic data, and the output is a risk score.
9. An electronic device comprising a processor and a memory;
the memory for storing a computer program;
the processor, when calling and executing the computer program stored in the memory, implements the triple play site identification method of any one of claims 1 to 3, or implements the construction method of the triple play site identification model of any one of claims 4 to 6.
10. A computer storage medium having stored thereon computer-executable instructions that, when loaded and executed by a processor, implement the triple play location identification method of any one of claims 1 to 3 or the method of constructing the triple play location identification model of any one of claims 4 to 6.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210384496.1A CN114819560A (en) | 2022-04-13 | 2022-04-13 | Three-in-one place identification method, identification model construction method and related equipment |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210384496.1A CN114819560A (en) | 2022-04-13 | 2022-04-13 | Three-in-one place identification method, identification model construction method and related equipment |
Publications (1)
Publication Number | Publication Date |
---|---|
CN114819560A true CN114819560A (en) | 2022-07-29 |
Family
ID=82535014
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202210384496.1A Pending CN114819560A (en) | 2022-04-13 | 2022-04-13 | Three-in-one place identification method, identification model construction method and related equipment |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN114819560A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115795354A (en) * | 2023-02-06 | 2023-03-14 | 北京志翔科技股份有限公司 | Three-in-one place identification method and identification device |
-
2022
- 2022-04-13 CN CN202210384496.1A patent/CN114819560A/en active Pending
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115795354A (en) * | 2023-02-06 | 2023-03-14 | 北京志翔科技股份有限公司 | Three-in-one place identification method and identification device |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN112104899B (en) | Information recommendation method and device in live broadcast, electronic equipment and storage medium | |
CN103606093A (en) | Intelligent chain VIP customer service system based on human characteristics | |
CN105512790A (en) | Integrated operation and maintenance management system | |
KR20170099078A (en) | Management Estimating Platform System Based on Big Data Using Manufacturing Facilities-Related Information from MES | |
CN110727761B (en) | Object information acquisition method and device and electronic equipment | |
WO2020019488A1 (en) | Staff demission warning method and related devices | |
CN103559619A (en) | Response method and system for garment size information | |
CN108230016B (en) | Agricultural product market price transmission analysis method and analysis device | |
CN104850955B (en) | The user information intelligent management and system of Network Environment | |
CN106572186B (en) | Production supervision method and system | |
CN113157947A (en) | Knowledge graph construction method, tool, device and server | |
CN110111167A (en) | A kind of method and apparatus of determining recommended | |
KR20200039852A (en) | Method for analysis of business management system providing machine learning algorithm for predictive modeling | |
CN107169081A (en) | All-purpose card big data analysis system | |
CN108388563A (en) | Information output method and device | |
CN103970891A (en) | Method for inquiring user interest information based on context | |
JP2018500669A (en) | Customer value evaluation method and system | |
CN115423429A (en) | Multimode integrated distribution network operation system based on image and sound information | |
CN106230985A (en) | A kind of based on the big data processing method of Internet of Things, system and service processing end | |
CN109447674A (en) | Electronic device, insurance agent target service area determine method and storage medium | |
CN110213342A (en) | Message pushing method and device of purchasing system, storage medium and terminal | |
CN114819560A (en) | Three-in-one place identification method, identification model construction method and related equipment | |
CN107977678A (en) | Method and apparatus for output information | |
CN104199889A (en) | RTLogic big data processing system and method based on CEP technology | |
CN105005852A (en) | Image analysis based intelligent monitoring system for dormitory environment |
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 |