CN117579625B - Inspection task pre-distribution method for double prevention mechanism - Google Patents
Inspection task pre-distribution method for double prevention mechanism Download PDFInfo
- Publication number
- CN117579625B CN117579625B CN202410067840.3A CN202410067840A CN117579625B CN 117579625 B CN117579625 B CN 117579625B CN 202410067840 A CN202410067840 A CN 202410067840A CN 117579625 B CN117579625 B CN 117579625B
- Authority
- CN
- China
- Prior art keywords
- task
- data
- calculation
- risk
- inspection
- 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
- 238000007689 inspection Methods 0.000 title claims abstract description 67
- 238000000034 method Methods 0.000 title claims abstract description 29
- 230000007246 mechanism Effects 0.000 title claims abstract description 18
- 230000002265 prevention Effects 0.000 title claims abstract description 17
- 238000009826 distribution Methods 0.000 title claims abstract description 15
- 238000004364 calculation method Methods 0.000 claims abstract description 83
- 230000009977 dual effect Effects 0.000 claims abstract description 13
- 210000002569 neuron Anatomy 0.000 description 7
- 230000006870 function Effects 0.000 description 6
- 230000008569 process Effects 0.000 description 6
- 239000011159 matrix material Substances 0.000 description 5
- 238000003860 storage Methods 0.000 description 5
- 239000000243 solution Substances 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 3
- 230000008878 coupling Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 238000012502 risk assessment Methods 0.000 description 3
- 230000001360 synchronised effect Effects 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 238000007405 data analysis Methods 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 230000007613 environmental effect Effects 0.000 description 2
- 230000000977 initiatory effect Effects 0.000 description 2
- 238000007726 management method Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000000638 solvent extraction Methods 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 230000009466 transformation Effects 0.000 description 2
- 238000012795 verification Methods 0.000 description 2
- 239000002699 waste material Substances 0.000 description 2
- 230000004913 activation Effects 0.000 description 1
- 238000009412 basement excavation Methods 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 239000003245 coal Substances 0.000 description 1
- 230000001010 compromised effect Effects 0.000 description 1
- 239000000470 constituent Substances 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 239000002360 explosive Substances 0.000 description 1
- 238000007667 floating Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000002347 injection Methods 0.000 description 1
- 239000007924 injection Substances 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 230000002035 prolonged effect Effects 0.000 description 1
- 238000013468 resource allocation Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000012954 risk control Methods 0.000 description 1
- 238000013058 risk prediction model Methods 0.000 description 1
- 230000008054 signal transmission Effects 0.000 description 1
- 239000000779 smoke Substances 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
- H04L67/1001—Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers
- H04L67/1004—Server selection for load balancing
-
- 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
- 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
- G06Q50/02—Agriculture; Fishing; Forestry; Mining
-
- G—PHYSICS
- G07—CHECKING-DEVICES
- G07C—TIME OR ATTENDANCE REGISTERS; REGISTERING OR INDICATING THE WORKING OF MACHINES; GENERATING RANDOM NUMBERS; VOTING OR LOTTERY APPARATUS; ARRANGEMENTS, SYSTEMS OR APPARATUS FOR CHECKING NOT PROVIDED FOR ELSEWHERE
- G07C1/00—Registering, indicating or recording the time of events or elapsed time, e.g. time-recorders for work people
- G07C1/20—Checking timed patrols, e.g. of watchman
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/12—Protocols specially adapted for proprietary or special-purpose networking environments, e.g. medical networks, sensor networks, networks in vehicles or remote metering networks
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Tourism & Hospitality (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- General Business, Economics & Management (AREA)
- Entrepreneurship & Innovation (AREA)
- Signal Processing (AREA)
- Computer Networks & Wireless Communication (AREA)
- Marketing (AREA)
- Game Theory and Decision Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Operations Research (AREA)
- Medical Informatics (AREA)
- Educational Administration (AREA)
- Development Economics (AREA)
- Computing Systems (AREA)
- Quality & Reliability (AREA)
- Agronomy & Crop Science (AREA)
- Animal Husbandry (AREA)
- Marine Sciences & Fisheries (AREA)
- Mining & Mineral Resources (AREA)
- Primary Health Care (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to a routing inspection task pre-distribution method for a dual prevention mechanism, which comprises the steps that a remote server configures routing inspection tasks, and the routing inspection tasks acquire calculation results of the routing inspection tasks based on a configured algorithm model. And the edge side server determines input sources configured in the inspection area according to the inspection task and the input of the algorithm model. And the remote server constructs a calculation task containing the risk data according to the historical data acquired by the input source, and the calculation task is executed by the calculation nodes configured at the edge side. The remote server obtains the reliability of the input source data according to the calculation result of the calculation task, determines the input source with the same result of the calculation task and the expected risk attribute of the risk data as the reliable input source, constructs a patrol sequence based on the terminal accessed by the reliable input source, and distributes the patrol sequence based on the load of the terminal. The problem of whether a downhole model is reliable is identified by randomly introducing noise and risk factors.
Description
Technical Field
The invention relates to a data processing method, in particular to a routing inspection task pre-distribution method for a dual prevention mechanism.
Background
Along with the digital transformation of the national energy industry field and the explosive growth of large data in the industry field and the continuous improvement of demands on mass data acquisition, convergence, excavation, analysis and the like, the data scale is rapidly increased, the data attribute dimension is also increased, the data analysis time is prolonged, and the calculation complexity is also rapidly improved. The multi-dimensional characteristic attribute of the data is mined, the design is reasonable and efficient, the expansibility is strong, the self-adaptive big data intelligent analysis and calculation method and the calculation task scheduling management system are adopted, users do not need to pay attention to details of creation, scheduling and execution links of tasks, whether the calculation resource allocation is reasonable, whether the resource dependence is met or not and the like, various intelligent analysis and calculation methods can be flexibly selected, more efforts are put on data analysis and service mining, and powerful guarantee is provided for the industrial data to exert the value of the industrial data.
At present, safety inspection is involved, multiple types of models are deployed near the edge side of a mine, the reliability of the models has a large influence on risk assessment, the models are considered when risk assessment is carried out to determine the reliability of data or models, however, due to the difference of the models, the problem of determining the reliability of the models also relates to the problem of data sources, and no universal solution exists at present.
Disclosure of Invention
In order to solve the problem of reliability of a supervision end on an edge side and an underground configured model, and to identify whether the underground model is reliable or not by randomly introducing noise and risk factors, the invention provides a routing inspection task pre-distribution method for a dual prevention mechanism.
In a first aspect, the present invention provides a method for pre-distributing inspection tasks for a dual prevention mechanism, including:
the remote server configures a patrol task, and the patrol task obtains a calculation result of the patrol task based on a configured algorithm model;
the edge side server determines input sources configured in the inspection area according to the inspection task and the input of the algorithm model;
the remote server constructs a calculation task containing risk data according to the historical data acquired by the input source, and the calculation task is executed by a calculation node configured at the edge side;
the remote server obtains the reliability of the input source data according to the calculation result of the calculation task, determines the input source with the same result of the calculation task and the expected risk attribute of the risk data as the reliable input source, constructs a patrol sequence based on the terminal accessed by the reliable input source, and distributes the patrol sequence based on the load of the terminal.
In some embodiments, the inspection tasks are performed based on multi-source sensors, and one inspection task is matched with at least two algorithmic models that receive inputs from the multi-source sensors and output risk identification results.
In some embodiments, the edge side server determining input sources configured within the inspection area based on the inspection task and the input of the algorithm model comprises: the underground control terminal acquires a sensor network arranged underground according to the type of the inspection task, determines an underground available history sensor in a designated time period according to the designated history time length of the inspection task, and selects an underground sensor which accords with the input of the algorithm model as an input source according to the input of the algorithm model.
In some embodiments, a computing node is selected on the edge side, and a computing task is offloaded at the computing node, wherein source data of the computing task is configured by a remote server and an edge side server, and the computing node communicates with the remote server through SSL when executing the computing task.
In some embodiments, the obtaining the reliability of the input source data according to the calculation result of the calculation task includes: and acquiring the risk type of the calculation task containing the risk data, and determining the reliability of the calculation model according to the difference between the calculation result and the risk type.
In some embodiments, constructing a computing task containing risk data from historical data collected by an input source comprises: and receiving the historical data collected by the input source, constructing virtual source data according to the input of the algorithm model, and using a dataset containing the historical data and the virtual source data as the input of the algorithm model to obtain a risk identification result corresponding to the dataset.
In some embodiments, computing nodes are selected on the edge side, computing tasks are offloaded at the computing nodes, wherein source data of the computing tasks are configured by a remote server, and the computing nodes communicate with the remote server through SSL when executing the computing tasks.
In some embodiments, the determining whether the computing model is reliable based on the difference in the computing result and the risk type comprises: when the risk types are consistent, the remote server builds a calculation task according to the historical data source and the algorithm model, the remote server executes the calculation task to obtain a calculation result of the risk, and the reliability of the calculation model is determined according to whether the calculation result is consistent with a report result of the mine side.
In some embodiments, when the computing results of the edge side computing model and the remote server side are inconsistent, parameters of the model are collected at the edge side and saved to the server, and a temporary model is placed at the edge side to receive input of the terminal, wherein the temporary model and the computing model parameters of the remote server side are consistent.
In some embodiments, the constructing a patrol sequence based on the terminal accessed by the reliable input source, and distributing the patrol sequence based on the load of the terminal includes: and acquiring inspection data acquired by the reliable input source, forwarding the acquired inspection data to the data gateway through a terminal accessed by the reliable input source, constructing a calculation task by the data gateway, unloading the calculation task to a calculation node, and identifying risks based on an algorithm model configured on the calculation node.
Drawings
The above and other objects, features and advantages of the present invention will become more apparent from the following description of embodiments of the present invention with reference to the accompanying drawings, in which:
fig. 1 is a flowchart of a method for pre-distributing patrol tasks for a dual prevention mechanism according to an embodiment of the present application.
Detailed Description
The disclosure will now be discussed with reference to several exemplary embodiments. It should be understood that these embodiments are discussed only to enable those of ordinary skill in the art to better understand and thus practice the present disclosure, and are not meant to imply any limitation on the scope of the present disclosure.
As used herein, the term "comprising" and variants thereof are to be interpreted as meaning "including but not limited to" open-ended terms. The term "based on" is to be interpreted as "based at least in part on". The terms "one embodiment" and "an embodiment" are to be interpreted as "at least one embodiment. The term "another embodiment" is to be interpreted as "at least one other embodiment".
In order to solve the problem of reliability of a supervision end on an edge side and an underground configured model, and to identify whether the underground model is reliable or not by randomly introducing noise and risk factors, the embodiment discloses a routing inspection task pre-distribution method for a dual prevention mechanism, which can include:
the remote server configures a patrol task, and the patrol task obtains a calculation result of the patrol task based on a configured algorithm model;
the edge side server determines input sources configured in the inspection area according to the inspection task and the input of the algorithm model;
the remote server constructs a calculation task containing risk data according to the historical data acquired by the input source, and the calculation task is executed by a calculation node configured at the edge side;
the remote server obtains the reliability of the input source data according to the calculation result of the calculation task, determines the input source with the same result of the calculation task and the expected risk attribute of the risk data as the reliable input source, constructs a patrol sequence based on the terminal accessed by the reliable input source, and distributes the patrol sequence based on the load of the terminal.
In the invention, the remote server can be a public cloud or private cloud server arranged at the cloud end, or is arranged far away from the mine in space, or is configured at the safety monitoring end.
The inspection tasks may be initiated manually by an operator, or periodically by a scheduler, or aperiodically to determine the safety of the mine. The manner of execution of the initiation may be directly initiated or may be an associated initiation.
The remote server is used for configuring the inspection task and executing the inspection task on the edge side. The edge side is a position close to a data generation source, particularly a position close to a roadway sensor side, and the data can be processed more quickly by arranging the calculation node at the position close to the roadway side. The computing power of the computing nodes near the edge side is generally weaker than that of the remote server, when the computing power is implemented by a plurality of computing nodes at the edge side, for example, by nodes arranged underground, the nodes store and utilize some but not all data for risk assessment due to the limitation of the computing power, and the sharing of the data is realized in a local industrial ring network, which is similar to the structure of fog computing. In the invention, the computing nodes close to the roadway side are servers at the edge side, the servers at the edge side contain computing nodes, the computing nodes receive the unloaded computing tasks and execute the computing tasks to obtain computing results, the computing results are synchronized with data according to rules pre-configured by the computing tasks or are sent to a designated data gateway, and the data gateway processes risk data based on the pre-configured rules.
The remote server and the edge side server are connected through, for example, a passive optical network or the internet, and perform synchronization of data. When the data is synchronized, the server at the edge side stores the locally stored data to the remote server in a periodical updating mode, so that the real-time bandwidth occupation is reduced, and the high storage cost introduced by real-time storage is reduced. The remote server may obtain historical sensor information within the roadway based on information of events, sensor types, locations of sensors, and the like. Through setting up reasonable update cycle, can make the data that the edge side kept in a reasonable scope for the data of edge side satisfies the requirement of risk calculation, makes the historical data of criticality be saved in the high in the clouds simultaneously.
When the logic subnetwork is used for connection, the servers at the edge side are connected through the route of the third layer, and the data are shared in the logic subnetwork. For example, based on the instruction of the remote server, the first node disposed on the edge side may obtain data from the second node, in order to save and monitor the resource information of the computing node on the edge side, for example, the saved resource information or the information of available computing resources, or the information of the computing model, a data gateway is further disposed on the edge side, where classifying the node according to the function of the node in the network on the edge side may include:
and (3) a terminal: one or more input sources connected to receive and forward the data;
computing node: for performing a computing task;
and (3) a control node: scheduling for computing nodes and resources within the edge-side network;
and (3) a data gateway: the node on the edge side, which is specially used for constructing the computing task, is also used for acquiring the historical data (particularly the historical environmental data collected by the sensor) or the stored real-time data which are stored on the edge side, and receiving the data forwarded by the terminal for constructing the computing task.
The above-described partitioning should be understood as a partitioning that is determined for a computing task, rather than being fixed, such as a node that is a common terminal when it receives and forwards data only to a data gateway, but that corresponds to the actual computing node for the other computing task.
In some embodiments, the inspection tasks are performed based on multi-source sensors, and one inspection task is matched with at least two algorithmic models that receive inputs from the multi-source sensors and output risk identification results.
The remote server configures a patrol task for the edge side server and is used for determining whether the environmental parameters in the coal mine tunnel correspond to the risk level. The inspection task is carried out by depending on data sources collected by a sensor network of the mine, wherein the data sources can be a gas sensor, a humidity sensor, a temperature sensor, a wind speed sensor, an image sensor and the like.
Common methods of risk level determination typically rely on input from one or more sensors, for example, the determination of gas risk levels can be made by an algorithmic model as follows:
processing the gas concentration data, wherein the gas concentration data is derived from measuring points, and a plurality of measuring points are selected on each roadway;
calculating to obtain section gas concentration distribution data of the section according to the processed gas concentration data;
obtaining a gas concentration distribution matrix according to the section gas concentration distribution data;
combining the gas concentration distribution matrix with the gas concentration distribution matrix data acquired 2min before the gas concentration matrix data to obtain a gas concentration distribution sequence;
and according to the gas concentration distribution sequence, predicting the gas concentration trend, and determining the risk level based on the predicted gas concentration.
The model used for prediction is an LSTM neural network and comprises an input layer, a hidden layer and an output layer, wherein the input layer inputs training data; the hidden layer iteratively learns the short-range and long-range characteristics of the time series data; the output layer outputs the prediction result. LSTM network parameters include learning rate, iteration number, batch size, etc.
In some embodiments, the edge side server determining input sources configured within the inspection area based on the inspection task and the input of the algorithm model comprises: the underground control terminal acquires a sensor network arranged underground according to the type of the inspection task, determines an underground available history sensor in a designated time period according to the designated history time length of the inspection task, and selects an underground sensor which accords with the input of the algorithm model as an input source according to the input of the algorithm model.
In some embodiments, a computing node is selected on the edge side, and a computing task is offloaded at the computing node, wherein source data of the computing task is configured by a remote server and an edge side server, and the computing node communicates with the remote server through SSL when executing the computing task.
In some embodiments, the obtaining the reliability of the input source data according to the calculation result of the calculation task includes: and acquiring the risk type of the calculation task containing the risk data, and determining the reliability of the calculation model according to the difference between the calculation result and the risk type.
In some embodiments, constructing a computing task containing risk data from historical data collected by an input source comprises: and receiving the historical data collected by the input source, constructing virtual source data according to the input of the algorithm model, and using a dataset containing the historical data and the virtual source data as the input of the algorithm model to obtain a risk identification result corresponding to the dataset.
The risk prediction model with multiple input sources is listed below, specifically as follows:
setting a plurality of measuring points in a roadway;
the real-time temperature T, the real-time CO concentration and the real-time smoke concentration are obtained by each measuring point;
constructing a BP network structure with forward propagation characteristics and backward propagation characteristics, wherein neurons in an input layer receive data and then transmit the data to a first layer of a hidden layer; neurons between each layer in the hidden layer adopt a fully connected form, each layer of neurons acquire information from a neuron layer before the neurons, and then the information is transmitted to a neuron layer after the neurons by using an activation function; when data arrives at the output layer, a predicted risk result is obtained.
The above algorithm model is only a common algorithm model, and it is obvious that there is a clear difference between the two network results, and when image recognition or voiceprint recognition is involved, a neural network such as a convolution layer and a multi-head attention mechanism is also used, so that multiple models and multiple models need to be deployed on the edge side to realize the judgment of the security risk. In the roadway at present, because of different sources of the models and inconsistent structures of the models, the identification results of the models for risks may be different, and the risk results may be changed due to the fact that model parameters may be changed or changed, so that challenges are brought to safety risk investigation and management in a double prevention mechanism.
Therefore, before actual risk identification is carried out, whether the local data are qualified or not and whether the identification of the algorithm model to the risk is accurate or not can be determined by testing the data with risk properties, so that ambiguity is eliminated, and resource waste caused by repeated investment of enterprises is avoided.
In some embodiments, computing nodes are selected on the edge side, computing tasks are offloaded at the computing nodes, wherein source data of the computing tasks are configured by a remote server, and the computing nodes communicate with the remote server through SSL when executing the computing tasks.
In some embodiments, constructing a computing task containing risk data from historical data collected by an input source comprises: and receiving the historical data collected by the input source, constructing virtual source data according to the input of the algorithm model, and using a dataset containing the historical data and the virtual source data as the input of the algorithm model to obtain a risk identification result corresponding to the dataset.
In some embodiments, when the computing results of the edge side computing model and the remote server side are inconsistent, parameters of the model are collected at the edge side and saved to the server, and a temporary model is placed at the edge side to receive input of the terminal, wherein the temporary model and the computing model parameters of the remote server side are consistent.
In some embodiments, the constructing a patrol sequence based on the terminal accessed by the reliable input source, and distributing the patrol sequence based on the load of the terminal includes: and acquiring inspection data acquired by the reliable input source, forwarding the acquired inspection data to the data gateway through a terminal accessed by the reliable input source, constructing a calculation task by the data gateway, unloading the calculation task to a calculation node, and identifying risks based on an algorithm model configured on the calculation node.
The risk identification is still associated with the inspection task, as described above, the inspection task is associated with one or more input sources and is associated with an algorithm model, and before the risk identification is performed, the used data source can be determined according to the configured inspection task, and corresponds to the input source in the roadway inspection area, and the data source is determined to be the input of the corresponding algorithm model; where multiple algorithm models are present, one or more of them may be selected and the configuration of the data sources performed.
And then, carrying out configuration of a calculation task containing risk data, wherein the configuration specifically comprises the steps of combining new data according to historical data and risk data acquired by an input source, determining the risk degree of the data after construction, carrying out risk identification through a calculation node configured on the edge side, and determining whether the input source data is reliable or not and whether an algorithm model is reliable or not according to a risk identification result. The historical data collected by the input source is used instead of other roadway data, so that deviation caused by overlarge difference between training data and actual data is avoided; further, the risk data here is preferably data set based on risk control indicators in the roadway, such as gas overruns, or a human image is drawn at key locations in the image so that the model can be identified as risk on the remote server side and it should be identified as risk on the edge side.
The identification process is preferably performed using multiple algorithm models, which can be used to detect when a single algorithm model fails, and by detecting multiple algorithm models to further provide reliable risk identification information when the multiple algorithm models are all valid.
When selecting different algorithm models, the corresponding input sources may not be consistent, and when constructing the data of the input sources, the data should be matched with the models according to the rules which are valid recently, and the model is matched to be consistent with the data specification required by the models, and the model is valid recently to be used for selecting available sensors, and the reason for selecting the latter is that the characteristics of part of the network determine that selecting the earlier sensors may cause data jitter, and the situation that the sensors fail may exist.
Upon selection of the completed input source, then the dimensions of the corresponding constructable risk property data are also determined, and the actual history data, herein referred to as first input data, may be obtained based on the history data of the input source, for example, obtained remotely by a remote server or locally by a data gateway, and the value or frame adjustment is performed based on the first input data to obtain data identified as a risk at the remote server side. Such as a sequenceIs identified as a normal value, but by means of a preset risk transformation matrix, it can be transformed into +.>The sequence has a risk level, e.g. divided into 3 classes of risk, where control of the rectification is required, but since the risk is not actually occurring, this proves that the model and data are valid. Other data injection methods for risk can also be realized by adjusting the numerical value, such as respectively floating the pre-sequence by 5%, so that the data obtained by monitoring the pre-sequence at the remote server side can possibly have gas concentration overrun within 5 minutes, thereby making the data risky. Other situations also include, for example, the adjustment of the constituent elements of the sequence such that the whole risk sequence +.>Become a risk sequence. It should be appreciated that the above is merely exemplary in nature and should be adjusted in practice according to the algorithm model on the remote server side; and the data dimensions of the constructed risk properties are not consistent for different models.
When source data exists at both the remote server and the edge side, selection may be made based on traffic and bandwidth, while when partial data exists only at the remote server side, the data is synchronized by the remote server to the edge side for calculation.
The results of the computing task should be communicated with the remote server based on encryption to avoid the presence of information related to production security being compromised, intercepted or otherwise modified.
After the results of the computing task are transmitted to the remote server, the remote server may build a model to detect, where the detection is an identification of a risk level to determine the reliability of the results, rather than simply determining whether it is safe or at risk.
It will be appreciated that if the remote server or edge side is at risk to the outcome of the risk identification, but is actually safe, unless the false positive rate is too high, for example more than 50%, these will be understood to be acceptable as they can be excluded with the aid of other models.
If the remote server identifies a consistent risk result, the remote server can send an instruction to the edge side, determine that the input source is a reliable input source, acquire inspection data acquired by the reliable input source, forward the acquired inspection data to the data gateway through a terminal accessed by the reliable input source, and the data gateway constructs a calculation task and then uninstalls the calculation task to a calculation node, and identifies the risk based on an algorithm model configured on the calculation node. When the inspection data required by the inspection task can be directly accessed to the data gateway, the data gateway can be used for directly acquiring the inspection data, and the real-time data still needs to be acquired by an input source and is transmitted to the data gateway after being acquired by the terminal so as to construct and unload the calculation task. The execution mode of the task can avoid the waste of resources and improve the response speed.
If the remote server identifies inconsistent risk results, the local model is different from the remote server side to a certain extent, a temporary model can be built at the edge by taking the model of the remote server as a reference, then an instruction is sent to the edge side to store key model parameters of the edge side in the server, then an input source is determined to be a reliable input source, inspection data acquired by the reliable input source are acquired, the acquired inspection data are forwarded to a data gateway through a terminal accessed by the input source, the data gateway is unloaded to a computing node after a computing task is built, and risk identification is carried out based on an algorithm model configured on the computing node. When the inspection data required by the inspection task can be directly accessed to the data gateway, the data gateway can be used for directly acquiring the inspection data, and the real-time data still needs to be acquired by an input source and is transmitted to the data gateway after being acquired by the terminal so as to construct and unload the calculation task. The server for storing the key model parameters can be an edge side server or a remote server, and the server is required to be carried out in an isolated container when being deployed, so that unsafe calculation models are prevented from being executed or unexpected attack is prevented; after completion of the computing task, the corresponding resources should be released. In this process, the edge-side model can still identify risks according to its designed algorithm model, since it still processes information according to the planned task.
In both cases, the nodes performing the computing tasks may or may not be identical in performing the verification and in performing the actual risk identification. In particular, after the computing task with verification property is completed, the actual load of the computing node changes, and the node with low load can be selected according to the actual load so as to unload the risk identification computing task. If the waiting time caused by data transmission is longer, the original computing node can still be appointed to execute the computing task.
In the above, risk types and risk results should have different understandings, wherein risk types generally refer to risk or no risk in the present invention, while risk recognition results generally comprise five classes, 1-5 respectively, corresponding to slightly risk, moderately risk, highly risk and huge risk of conventional understanding.
Those of ordinary skill in the art will appreciate that the modules and algorithm steps described in connection with the embodiments disclosed herein can be implemented as electronic hardware, or combinations of computer software and electronic hardware. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the solution. 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 invention.
It will be clearly understood by those skilled in the art that, for convenience and brevity of description, specific working procedures of the apparatus and device described above may refer to corresponding procedures in the foregoing method embodiments, which are not described herein again.
In the embodiments provided in the present application, it should be understood that the disclosed apparatus and method may be implemented in other manners. For example, the apparatus embodiments described above are merely illustrative, and for example, the division of the modules is merely a logical function division, and there may be additional divisions when actually implemented, for example, multiple modules or components may be combined or integrated into another system, or some features may be omitted or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be an indirect coupling or communication connection via some interfaces, devices or modules, which may be in electrical, mechanical, or other forms.
The modules described as separate components may or may not be physically separate, and components shown as modules may or may not be physical modules, i.e., may be located in one place, or may be distributed over a plurality of network modules. Some or all of the modules can be selected according to actual needs to achieve the purpose of the embodiment of the invention.
In addition, each functional module in the embodiment of the present invention may be integrated in one processing module, or each module may exist alone physically, or two or more modules may be integrated in one module.
The functions, if implemented in the form of software functional modules and sold or used as a stand-alone product, may be stored in a computer-readable storage medium. Based on such understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art or in a part of the technical solution in the form of a software product stored in a storage medium, comprising several instructions for causing a computer device (which may be a personal computer, a server, a network device, etc.) to perform all or part of the steps of the method for energy saving signal transmission/reception of the various embodiments of the present invention. And the aforementioned storage medium includes: a usb disk, a removable hard disk, a ROM, a RAM, a magnetic disk, or an optical disk, etc.
The foregoing description is only of the preferred embodiments of the present application and is presented as a description of the principles of the technology being utilized. It will be appreciated by persons skilled in the art that the scope of the invention referred to in this application is not limited to the specific combinations of features described above, but it is intended to cover other embodiments in which any combination of features described above or equivalents thereof is possible without departing from the spirit of the invention. Such as the above-described features and technical features having similar functions (but not limited to) disclosed in the present application are replaced with each other.
It should be understood that, the sequence numbers of the steps in the summary and the embodiments of the present invention do not necessarily mean the order of execution, and the execution order of the processes should be determined by the functions and the internal logic, and should not be construed as limiting the implementation process of the embodiments of the present invention. The foregoing description of implementations of the present disclosure has been presented for purposes of illustration and description. The foregoing description is not intended to be exhaustive or to limit the disclosure to the precise form disclosed, and modifications and variations are possible in light of the above teachings or may be acquired from practice of the disclosure. The embodiments were chosen and described in order to explain the principles of the present disclosure and its practical application to enable one skilled in the art to utilize the present disclosure in various embodiments and with various modifications as are suited to the particular use contemplated.
Claims (8)
1. The routing inspection task pre-distribution method for the double prevention mechanism is characterized by comprising the following steps of:
the remote server configures a patrol task, and the patrol task obtains a calculation result of the patrol task based on a configured algorithm model;
the edge side server determines input sources configured in the inspection area according to the inspection task and the input of the algorithm model;
the remote server constructs a calculation task containing risk data according to the historical data acquired by the input source, and the calculation task is executed by a calculation node configured at the edge side;
the remote server obtains the reliability of the input source data according to the calculation result of the calculation task, determines the input source with the same result of the calculation task and the expected risk attribute of the risk data as the reliable input source, constructs a patrol sequence based on the terminal accessed by the reliable input source, and distributes the patrol sequence based on the load of the terminal;
the obtaining the reliability of the input source data according to the calculation result of the calculation task comprises the following steps: acquiring a risk type of a calculation task containing risk data, and determining the reliability of an algorithm model according to the difference between a calculation result and the risk type;
the determining the reliability of the algorithm model according to the difference between the calculation result and the risk type comprises the following steps: when the risk types are consistent, the remote server builds a calculation task according to the historical data source and the algorithm model, the remote server executes the calculation task to obtain a calculation result of the risk, and the reliability of the algorithm model is determined according to whether the calculation result is consistent with a reporting result of the mine side.
2. The method for pre-distributing inspection tasks for dual prevention mechanisms of claim 1 wherein said inspection tasks are performed based on multi-source sensors and one inspection task is matched with at least two algorithmic models that receive inputs from the multi-source sensors and output risk identification results.
3. The inspection task pre-distribution method for a dual prevention mechanism according to claim 1, wherein the edge side server determining input sources configured in the inspection area according to the inspection task and the input of the algorithm model comprises: the underground control terminal acquires a sensor network arranged underground according to the type of the inspection task, determines an underground available historical sensor in a designated period according to the designated historical time of the inspection task, and selects an underground sensor conforming to the input of the algorithm model as an input source according to the input of the algorithm model and the underground available historical sensor.
4. The method for pre-distributing inspection tasks for dual prevention mechanism of claim 1 wherein constructing a computing task containing risky data from historical data collected from an input source comprises: and receiving the historical data collected by the input source, constructing virtual source data according to the input of the algorithm model, and using a dataset containing the historical data and the virtual source data as the input of the algorithm model to obtain a risk identification result corresponding to the dataset.
5. The method for pre-distributing patrol tasks for dual prevention mechanism according to claim 4 wherein said edge side configured computing node performing a computing task comprises:
selecting a computing node at the edge side, and unloading the computing task at the computing node, wherein the source data of the computing task is configured by a remote server, and the computing node communicates with the remote server through SSL when executing the computing task.
6. The method for pre-distributing inspection tasks for dual prevention mechanism as recited in claim 4 wherein computing nodes are selected at an edge side and computing tasks are offloaded at the computing nodes, wherein source data of the computing tasks are configured by a remote server and an edge side server, and wherein the computing nodes communicate with the remote server via SSL when executing the computing tasks.
7. The method for pre-distributing inspection tasks for dual prevention mechanism according to claim 1, wherein when the result of the edge side algorithm model and the calculation result of the remote server side calculation task are inconsistent, parameters of the model are collected at the edge side and saved to the server, and a temporary model is placed at the edge side to receive input of the terminal, wherein the temporary model and the algorithm model parameters at the remote server side are consistent.
8. The method for pre-distributing patrol tasks for dual prevention mechanism according to claim 1, wherein said constructing a patrol sequence based on a terminal to which a reliable input source is connected, distributing the patrol sequence based on a load of the terminal comprises: and acquiring inspection data acquired by the reliable input source, forwarding the acquired inspection data to the data gateway through a terminal accessed by the reliable input source, constructing a calculation task by the data gateway, unloading the calculation task to a calculation node, and identifying risks based on an algorithm model configured on the calculation node.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410067840.3A CN117579625B (en) | 2024-01-17 | 2024-01-17 | Inspection task pre-distribution method for double prevention mechanism |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410067840.3A CN117579625B (en) | 2024-01-17 | 2024-01-17 | Inspection task pre-distribution method for double prevention mechanism |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117579625A CN117579625A (en) | 2024-02-20 |
| CN117579625B true CN117579625B (en) | 2024-04-09 |
Family
ID=89884902
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410067840.3A Active CN117579625B (en) | 2024-01-17 | 2024-01-17 | Inspection task pre-distribution method for double prevention mechanism |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117579625B (en) |
Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN213399929U (en) * | 2020-11-26 | 2021-06-08 | 兰州理工大学 | Landslide monitoring system |
| CN113282368A (en) * | 2021-05-25 | 2021-08-20 | 国网湖北省电力有限公司检修公司 | Edge computing resource scheduling method for substation inspection |
| CN114169656A (en) * | 2020-09-11 | 2022-03-11 | 中国石油化工股份有限公司 | Drilling stuck risk early warning method and system based on adjacent well historical data |
| CN114240101A (en) * | 2021-12-02 | 2022-03-25 | 支付宝(杭州)信息技术有限公司 | Risk identification model verification method, device and equipment |
| CN114320469A (en) * | 2021-12-27 | 2022-04-12 | 中国矿业大学 | Cloud-edge intelligence-based underground hazard source detection method |
| CN114373245A (en) * | 2021-12-16 | 2022-04-19 | 南京南自信息技术有限公司 | Intelligent inspection system based on digital power plant |
| WO2022213565A1 (en) * | 2021-04-07 | 2022-10-13 | 全球能源互联网研究院有限公司 | Review method and apparatus for prediction result of artificial intelligence model |
| CN115545198A (en) * | 2022-11-25 | 2022-12-30 | 成都信息工程大学 | Edge intelligent collaborative inference method and system based on deep learning model |
| CN115695541A (en) * | 2022-09-28 | 2023-02-03 | 上海东普信息科技有限公司 | Method, device and equipment for monitoring dot polling based on edge calculation and storage medium |
| CN116109058A (en) * | 2022-11-28 | 2023-05-12 | 南方电网电力科技股份有限公司 | Substation inspection management method and device based on deep reinforcement learning |
| CN116310922A (en) * | 2021-12-17 | 2023-06-23 | 中国石油化工股份有限公司 | Petrochemical plant area monitoring video risk identification method, system, electronic equipment and storage medium |
| CN116486507A (en) * | 2023-04-26 | 2023-07-25 | 新奥（中国）燃气投资有限公司 | Municipal pipe network inspection system |
| CN117294022A (en) * | 2023-11-23 | 2023-12-26 | 国网山东省电力公司济南供电公司 | Substation inspection system and method based on cooperation of multi-source sensors |
| CN117409527A (en) * | 2023-10-26 | 2024-01-16 | 深圳微能聚力物联科技有限公司 | Fire safety detection system based on edge cloud cooperation |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| GB2597692A (en) * | 2020-07-29 | 2022-02-09 | Sony Europe Bv | Systems, devices and methods for operating a vehicle with sensors monitoring parameters |
-
2024
- 2024-01-17 CN CN202410067840.3A patent/CN117579625B/en active Active
Patent Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114169656A (en) * | 2020-09-11 | 2022-03-11 | 中国石油化工股份有限公司 | Drilling stuck risk early warning method and system based on adjacent well historical data |
| CN213399929U (en) * | 2020-11-26 | 2021-06-08 | 兰州理工大学 | Landslide monitoring system |
| WO2022213565A1 (en) * | 2021-04-07 | 2022-10-13 | 全球能源互联网研究院有限公司 | Review method and apparatus for prediction result of artificial intelligence model |
| CN113282368A (en) * | 2021-05-25 | 2021-08-20 | 国网湖北省电力有限公司检修公司 | Edge computing resource scheduling method for substation inspection |
| CN114240101A (en) * | 2021-12-02 | 2022-03-25 | 支付宝(杭州)信息技术有限公司 | Risk identification model verification method, device and equipment |
| CN114373245A (en) * | 2021-12-16 | 2022-04-19 | 南京南自信息技术有限公司 | Intelligent inspection system based on digital power plant |
| CN116310922A (en) * | 2021-12-17 | 2023-06-23 | 中国石油化工股份有限公司 | Petrochemical plant area monitoring video risk identification method, system, electronic equipment and storage medium |
| CN114320469A (en) * | 2021-12-27 | 2022-04-12 | 中国矿业大学 | Cloud-edge intelligence-based underground hazard source detection method |
| CN115695541A (en) * | 2022-09-28 | 2023-02-03 | 上海东普信息科技有限公司 | Method, device and equipment for monitoring dot polling based on edge calculation and storage medium |
| CN115545198A (en) * | 2022-11-25 | 2022-12-30 | 成都信息工程大学 | Edge intelligent collaborative inference method and system based on deep learning model |
| CN116109058A (en) * | 2022-11-28 | 2023-05-12 | 南方电网电力科技股份有限公司 | Substation inspection management method and device based on deep reinforcement learning |
| CN116486507A (en) * | 2023-04-26 | 2023-07-25 | 新奥（中国）燃气投资有限公司 | Municipal pipe network inspection system |
| CN117409527A (en) * | 2023-10-26 | 2024-01-16 | 深圳微能聚力物联科技有限公司 | Fire safety detection system based on edge cloud cooperation |
| CN117294022A (en) * | 2023-11-23 | 2023-12-26 | 国网山东省电力公司济南供电公司 | Substation inspection system and method based on cooperation of multi-source sensors |
Non-Patent Citations (2)
| Title |
|---|
| "煤矿智能双重预防机制与智能安全管控平台研究";李爽 等;《煤炭科学技术》;20230131;第464-473页 * |
| "瑞垟二级水电站双重预防机制建设技术探索";毛小付 等;《水电站机电技术》;20230630;第139-141页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117579625A (en) | 2024-02-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107589695B (en) | Train set fault prediction and health management system | |
| WO2024000852A1 (en) | Data processing method and apparatus, device, and storage medium | |
| US7693589B2 (en) | Anomaly anti-pattern | |
| JP6501675B2 (en) | Configurable on-board information processing | |
| CN116468186B (en) | Flight delay time prediction method, electronic equipment and storage medium | |
| US7756593B2 (en) | Anomaly anti-pattern | |
| WO2018197585A1 (en) | Identifying targets within images | |
| CN116052300B (en) | Digital twinning-based power inspection system and method | |
| CN111815946A (en) | Method and device for determining abnormal road section, storage medium and electronic equipment | |
| WO2023159812A1 (en) | Method and apparatus for detecting ami network intrusion, and medium | |
| CN117725619B (en) | Data sharing method, device, computer equipment, chip and readable storage medium | |
| CN116624226A (en) | Coal mine disaster data acquisition, analysis and visual display system | |
| CN117579625B (en) | Inspection task pre-distribution method for double prevention mechanism | |
| CN117523499B (en) | Forest fire prevention monitoring method and system based on Beidou positioning and sensing | |
| CN117555684A (en) | Method, system, equipment and medium for fusing edge computing and cloud computing | |
| CN117632478A (en) | Method, system and device for fusing distributed edge computing and cloud computing | |
| Sundararajan et al. | A tri-modular framework to minimize smart grid cyber-attack cognitive gap in utility control centers | |
| Balaji et al. | CANLite: Anomaly detection in controller area networks with multitask learning | |
| CN114039837B (en) | Alarm data processing method, device, system, equipment and storage medium | |
| CN115829536A (en) | Gradual faults in power networks | |
| EP2909687B1 (en) | System testing algorithm and apparatus | |
| CN110891039A (en) | LM neural network-based coal mine monitoring system safety early warning method | |
| Guzairov et al. | Simulation modelling of the transmission system of the telemetric information on the status of the on-board aircraft status | |
| CN118710009A (en) | Maritime search and rescue strength scheduling command method | |
| CN117956498A (en) | Unmanned aerial vehicle communication resource and track joint optimization method, device and server |
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 |