Abstract
A dynamic diagnosis system of mine safety based on multi-data fusion is designed by combining the advantages of cloud computing and the Internet of things. The system’s framework is built with a three-tier model, and the construction of the cloud computing service platform provides essential support for mass data storage, processing, and security diagnosis reasoning. Simultaneously, using the relational database SQL Server 2017 and the object-oriented language C# to complete the design of an expert knowledge base and reasoning mechanism, and establish the diagnostic scoring system in the way of weighted sum. Therefore, based on the logical matching and reasoning between the collected data and the safety rules, the dynamic diagnosis of mine safety is realized.
Introduction
The coal mine production system develops dynamically in space and time [1]. Coal mine accidents such as a gas explosion, water disaster, fire, and rockburst are dynamic, random, and fuzzy [2]. Different accidents are related to time and space. To prevent and control the occurrence of coal mine accidents from the source, it is necessary to prevent and control coal mine safety risks. However, coal mine safety risk prevention and control generally have the following problems: risk prevention and control is lack of systematization [3, 4] It is difficult to grasp the risk factors and change characteristics entirely; lack of dynamic monitoring of risk quantitative change process, the timeliness of risk prevention and control is poor [5]; management at all levels cannot timely grasp the safety conditions within the scope of risk prevention and control responsibility, and only respond to the abnormal events [6, 7], and risk prevention and control is in the passive response mode. Thus, this paper designs a coal mine safety risk prevention and early warning system, which can improve the systematicness, timeliness, and initiative of coal mine safety risk prevention and control [8–10].
To effectively ensure the accuracy of coal mine safety monitoring, its monitoring structure will involve many fields, but in fact, the environment in the coal mine production process is terrible, which will quickly lead to various problems in safety monitoring. At the same time, the traditional wired monitoring system relies on specific test points for particular application laying. This monitoring mode has low mobility and networking, which significantly affects the quality of coal mine production to a certain extent. With the development of science and technology, wireless sensor technology and unique bus technology are gradually introduced into coal mine monitoring. The coal mine monitoring system is developing towards the direction of intelligence and precision. The wireless sensor network has the advantages of strong networking ability and no blind spot. Through wireless sensor, the comprehensive monitoring of coal mine data can be effectively carried out. Wireless tracking has been widely used in coal mine safety monitoring. Besides, the traditional monitoring methods only deal with the data thoroughly and can’t process and analyze the data in-depth, quickly causing the omission of data information. Moreover, the intelligent means are not adopted in the security early warning decision-making, mainly relying on manual work, which significantly reduces the work efficiency. In the coal mine safety monitoring, this paper combines mostly the tt-can bus transmission technology with the data acquisition technology of wireless sensor network, uses the improved Apriori coal mine safety early warning software module analysis of the upper computer numerical control to get the relationship between various factors in coal mine safety, and then improve the efficiency of coal mine safety monitoring [11].
The system is composed of three layers. The first layer is mainly data acquisition node and data preprocessing circuit. The system collects data from various factors related to coal mine safety through special sensors. After the data acquisition is completed, the data is stored in the original database through data preprocessing, to facilitate the later use of data. In the second layer, a reliable TTCAN compatible controller based on DSP and mco2515 is adopted, effectively reducing the error rate and having high real-time performance. The third layer is the data host computer analysis system, which analyzes the factors associated with coal mine safety factors, studies the correlation between these factors and early warning to ensure the accuracy of coal mine safety early warning. In the real-time analysis of the coal mine upper computer data, the data collected after the formation of the system needs to be uploaded. The data collected by the sensor can effectively realize the coverage of no blind area, covering all factors of coal mine safety comprehensively, which dramatically improves the accuracy of coal mine safety prediction. Besides, the system can also provide a relatively complete database, which can analyze the correlation between coal mine safety factors and effectively help the staff carry out real-time monitoring of coal mine safety, which significantly guarantees the accurate tracking of coal mine safety [12–15].
Overall framework of the system
Coal mine safety dynamic diagnosis system is based on cloud computing, the Internet of things, data mining technology as the support, based on various safety production regulations and specifications formulated by the industry, group, and mine, combined with comprehensive automation, in the. Line monitoring (water, fire, gas, roof, etc.) and dynamic data and historical data obtained daily by coal mine safety production, based on a dynamic three-dimensional geological model and virtual mine platform, all kinds of coal mine information safety production are analyzed. Through demonstration, analysis, and reasoning, mining the patterns and knowledge in historical data, diagnosing and summarizing the current safety status, predicting the future safety situation, and realizing the dynamic diagnosis and auxiliary decision-making of coal mine safety production. Benefit. Based on the necessary data, real-time monitoring data, and transaction data in the comprehensive database of coal mine safety, and according to the knowledge base of dynamic diagnosis experts of coal mine safety, the status quo of coal mine safety production is analyzed [16–20].
And the trend, forecast the future, and make scientific and reasonable response countermeasures and treatment measures for coal mine emergency phenomenon. The main functions of the coal mine safety dynamic diagnosis system include coal mine safety situation evaluation and scoring, coal mine safety problem reasoning and interpretation, diagnosis task configuration, and management. Based on a three-dimensional virtual mine visualization platform, online monitoring data, business management data, and evaluation scoring and reasoning analysis results are displayed.
The dynamic diagnosis system of mine safety is based on the online tracking and detection of multiple hazard sources such as underground workers, equipment, production environment, etc. It collects the data parameters of potential hazard factors in real-time through the arrangement of sensor equipment with different functions, builds the expert knowledge base, combines the contents of the stored standard procedures, carries out reasoning diagnosis, and evaluates the hidden danger or risk of mine accidents According to the root cause; an effective solution is given. The Internet of things technology can collect and intelligently process massive security information by deploying multiple types of sensors, which can be used as nodes to form a sensing network. Cloud computing has high storage capacity and fast distributed parallel computing ability, which can provide sufficient support for the storage and operation of data collection, expert knowledge, diagnosis, and reasoning principles. Based on this, the mine diagnosis system’s overall framework constructed by combining the two technical advantages is shown in Fig. 1. The diagnosis system consists of an application service layer, network layer, and perception layer [21–23].
Framework of mine safety dynamic diagnosis system.
This layer is based on the call and analysis of data parameters in the perception layer to process complex services such as security state evaluation calculation, logic diagnosis reasoning, and encapsulates the business logic processing function into a series of components including data access, access, query, knowledge call, security state evaluation, rule reasoning, etc., which provides effective services for different application functions, and can It uses intelligent mobile terminal, computer, LED electronic screen and other multi-mode monitoring and diagnosis terminals to carry out dynamic remote security assessment and reasoning diagnosis; at the same time, the layer stores, organizes and manages information such as collected data, expert knowledge, diagnosis and reasoning rules, and provides data services based on Cloud Computing Technology.
Network layer
The network layer is the service support to transmit the data parameters obtained by each sensor node to the application service. The underground tunnels of the mine are complex, and the equipment is diverse, which requires high network communication technology. RFID Wireless communication technology is limited to receiving short distance signals, and the communication rate is not high, which is generally used for receiving the signals of main fixed equipment and short distance equipment in the coal mine; WiFi wireless communication technology has a relatively high data transmission rate, which can support long-distance communication in underground tunnels, and is very suitable for the effective acquisition of mobile equipment and long-distance equipment signals. Therefore, the advantages of the two can be integrated to realize the “comprehensive perception” of equipment signals.
Perception layer
This layer covers the temperature and humidity, currant, smoke, and other sensor nodes associated with the production system, safety system, power supply system, and production scheduling system. It can collect online monitoring data of gas concentration, wind speed direction, temperature and humidity, negative pressure, smoke, door switch, surrounding rock stress, tunnel deformation, etc. in real-time, and integrate these data parameters to form a data source for safety diagnosis According to, and link to WiFi or RFID wireless network in the network layer to realize data transmission.
Building cloud computing service platform of the system
In the design of mine safety dynamic diagnosis, online data monitoring, and evaluation reasoning are the keys to the system’s distributed sensing function. To realize the real-time tracking monitoring, early warning analysis, and reasoning prediction of mine safety state, it needs the support of a cloud computing platform, which should include data collection, expert knowledge base, reasoning mechanism, and remote control Monitoring and diagnosis equipment, data fusion management, diagnosis evaluation, reasoning, and other functions [24].
The coal mine safety diagnosis system needs to manage and deal with the massive data of coal mine safety production and online monitoring in time. The system needs to be configured and deployed on the hardware platform with a strong performance to ensure system performance, access efficiency, and subsequent scalability and scalability. At present, the popular cloud computing technology is a computing method based on the Internet, which provides shared hardware-level software resources to computers and related equipment according to their specific needs. Based on the network’s help, the corresponding computing resource pool is constructed, and unified management is implemented to provide services to users according to their actual needs. Cloud computing provides scalable and cheap distributed computing capability through the network. In the cloud computing environment, data is stored in a distributed way to achieve high availability and economy [25].
Cloud computing
The coal mine safety diagnosis system needs to manage and deal with the massive coal mine safety production data and online monitoring in time. The system needs to be configured and deployed on the hardware platform with a strong performance to ensure system performance, access efficiency, and subsequent scalability and scalability. At present, the popular cloud computing technology is a computing method based on the Internet, which provides shared hardware-level software resources to computers and related equipment according to their specific needs. Based on the network’s help, the corresponding computing resource pool is constructed, and unified management is implemented to provide services to users according to their actual needs. Cloud computing delivers scalable and cheap distributed computing capability through the network. In the cloud computing environment, data is stored in a distributed way to achieve high availability and economy. According to the content, it can be divided into infrastructure cloud (IAAs), platform cloud (PAAS), and software cloud (SaaS). Infrastructure cloud is mainly for underlying hardware integration, while platform cloud and software cloud are business integration based on enterprise requirements. The rapid development of cloud computing is based on the development of virtualization technology and data-intensive computing. Through distributed computing, storage, data management, and Internet technology, it provides users with cloud computing and cloud storage functions, with high reliability and high scalability [5–7]. The emergence and rapid development of cloud computing, on the one hand, is the result of the development of virtualization technology, data-intensive computing, and other technologies, on the other hand, it is also the embodiment of the inevitable trend that the development of the Internet needs to constantly enrich its application.
Internet of things
Internet of things is the third-largest information industry in the world after the computer and the Internet. The second wave, an important part of the new generation of information technology, is a network-based on computer Internet to realize remote monitoring and management by integrating global positioning technology, radiofrequency technology, and infrared technology, and carry out information exchange and communication, to realize the intelligent identification, positioning, tracking, monitoring and management of goods. It’s about perception. In order to realize the comprehensive interconnection between people, people, and things, things, and things, it is an extension of ubiquitous sensor networks. The Internet of things has three distinct characteristics: the deployment of a large number of types of sensors, the importance of technology. The foundation and core is still the Internet, which has the ability of intelligent processing. The essence of the Internet of things is a micro intelligent sensor with sensing, computing, and communication capabilities and a sensor network formed by its nodes. It is the specific realization of sensor network technology in the process of social production and life. It uses intelligent technologies such as cloud computing and pattern recognition to expand its application field. The mine Internet of things is to realize the visualization, digitization, and intellectualization of the whole mine and related phenomena, digitize the comprehensive information of mine production and construction process and safety management and realize the dynamic, collaborative control of mine safety production process [26].
Virtualization of infrastructure
My safety production is associated with various equipment and facilities. To improve the efficiency of monitoring and diagnosis and simplify the system design, Hyper-V virtualization software system can be used to integrate multiple back-end servers into one logical server, which can be divided into the web server, application server. Database service Virtual machine resources provide intelligent service support for data monitoring, real-time diagnosis, expert knowledge matching reasoning, data access, remote monitoring equipment maintenance, etc.
The coal mine safety dynamic diagnosis system is logically divided into the front-end performance layer, middle application logic layer, and background data source layer. The data source layer stores, organizes, and manages the data, knowledge, metadata, algorithm model, and other information, and provides consistent and efficient data access services through the data access layer. The middle application logic layer calls the data access layer’s data services to carry out complex business logic processing, such as coal mine safety condition evaluation calculation and safety rules logic reasoning. It encapsulates these business logic processing functions into a series of components. It includes three-dimensional virtual environment integration components, data access & query components, unified data access components, unified knowledge access components, security condition assessment components, rule reasoning components, and so on [27].
Service of software application
The system architecture of mine safety dynamic diagnosis system based on Internet of things technology includes the application service layer, network layer, and perception layer, which can store, organize and manage safety data, diagnosis rules, reasoning mechanism, and other information, and based on the access and call of data, carry out rating calculation, safety rule logic reasoning and other complex business logic processing for mine safety status To realize these functions, we can use the service of software application to encapsulate the services provided by each layer in the form of Web services into a series of components such as 3D virtual environment integration component, data access extraction component, data unified access component, security status evaluation component, rule reasoning component, etc., to provide functional component services for different application requirements of the system.
Key technologies of system design
Establishment of expert knowledge base
Object-oriented programming (OOP) can define an expert knowledge base’s attributes to realize the induction, classification, storage, and representation of safety rules, regulations, and standards. Therefore, the expert knowledge base of mine safety diagnosis can be constructed by this method, and the function of the knowledge base can be divided into modules. The expert knowledge base is mainly composed of three modules: fact base, rule base, and standard base. SQL Server 2017 is used to store the knowledge through a series of two-dimensional tables. Among them, the fact base stores the factual knowledge such as mine safety production status and inference conclusion in the form of a fact table; The rule base uses a rule table, atomic forward table, premise table, and conclusion table to store, which includes the diagnostic rules of smoke, gas, water inrush, water inrush, low pressure of mine pressure impact, personnel, equipment, and environmental hazards, as the basis for the assessment and diagnosis of mine safety status; The standard library contains the relevant safety rules, regulations, and standards, including the three violation behavior standards, accident potential standards, equipment operation regulations, public operation regulations, etc. At present, there are production and frame methods to express safety knowledge. Its storage structure is generally as follows: if x=>thy, or X - >y (confidence degree), if x is established, y conclusion can be obtained. X is the rule antecedent, which can be the combination of some propositions. Y is the rule antecedent, which is a specific conclusion. To express incomplete knowledge, confidence should be introduced.
Diagnostic evaluation method
The safety diagnosis evaluation method can be classified according to the quantification degree of evaluation structure and the reasoning process of evaluation. In this paper, the quantitative evaluation method is selected, and the diagnosis scoring system is designed by way of weighted sum, as shown in Fig. 2.
Safety diagnosis scoring system.
The dynamic diagnosis of mine safety involves the scoring content of operators, equipment, and the environment. The total score of diagnosis can be set as 100 points, and 35, 30, and 35 can be assigned, respectively. The total score of 100 points is used for detailed classification calculation for the three parts, among which the diagnosis of operators is divided into three violation behaviors and downhole overtime scoring modules, with 50 assigned respectively. The score of the two modules and the product of their respective weights together constitute the operator’s diagnosis score; the equipment diagnosis focuses on the fault score; the environment diagnosis includes two modules, accident hidden danger, and detection alarm, each with a value of 50 points. The score of three violations can be divided into three grades: slight, general, and profound. Each grade is assigned with 4 points, 2 points, and 1 point. The deduction standard is determined by the product of unit score and frequency of behaviors. Taking the scoring method of accident hidden danger in the environmental diagnosis model as an example, the design method is as follows, which can be divided into four grades: A, B, C, and d according to the severity, and the deduction weight standard of each kind of hidden danger is 10, 5, 3 and 1. Combined with the situation of hidden danger protection or not, the unprotected hidden danger is counted into minutes, and the total deduction method is as follows:
In formula (1), Ma, Mb, Mc, and Md are unprotected class A, B, C, and D hidden dangers, Ka, Kb, Kc, and Kd are class A, B, C, and D classification weights, Na, Nb, Nc, and Nd are all class A, B, C and D hidden dangers.
The reasoning mechanism uses object-oriented, rule-based, and rule-based clips development tools to embed them in the system. The net platform uses C# and clips’ mixed programming and encapsulates them as a method through clips net component. When calling clips in. Net platform, you need to initialize environment object environment = new mommosoft. Expert system. Environment(); Environment. Reset(), then load the environment, factaddressvalue, float value, instanceaddressvalue, and other classes of system operation in the mommosoft.expertsystem.dll library, to use the environment.assertstring method to assign the collected fact parameters, use the multifieldvalue class to retrieve the inference rule content of the expert knowledge base, and use the multifieldvalue class to retrieve the inference rule content of the expert knowledge base To manipulate objects. Inclusion module of the security reasoning process based on clips reasoning mechanism
Function modules such as pattern matching, conflict resolution, rule activation, and rule execution. Specifically, firstly, scan all rule patterns in the knowledge base, put the patterns matching the collected data into the agenda, and the activated rules stored will be pressed into the agenda according to the priority sequence. This agenda mechanism controls the rules’ execution sequel scheduled. The diagnosis conclusion can be drawn by adjusting the tscheduleed rules. The reasoning mechanism includes forward and reverses reasoning. The former takes the data parameters collected by online monitoring as the real-time basis of safety diagnosis, takes the rules and standards in the expert knowledge base as the diagnosis basis, and obtains the mine safety state through the fusion, matching, and reasoning of the two. The latter takes the conclusion as the logical starting point of safety diagnosis to find the rules that can reach the objective conclusion.
The coal mine safety dynamic diagnosis system is based on cloud computing, the Internet of things, data mining technology as the support, based on various safety production regulations and specifications formulated by the industry, group, and mine, combined with comprehensive automation, online monitoring (water, fire, gas, roof, etc.) and dynamic data and historical data obtained daily in coal mine safety production, Based on the dynamic three-dimensional geological model and virtual mine platform, various information of coal mine safety production is displayed, analyzed and reasoned. The patterns and knowledge contained in historical data are mined, the current safety status is diagnosed and summarized, the future safety situation is predicted, and the dynamic diagnosis and auxiliary decision-making of coal mine safety production are realized. Based on the basic data, real-time monitoring data, and transaction data in the comprehensive database of coal mine safety, the evaluation, reasoning, and deduction are carried out according to the knowledge base of dynamic diagnosis experts of coal mine safety. The status and trend of coal mine safety production are analyzed, and the future is predicted. Scientific and reasonable response countermeasures and treatment measures are made according to the emergency of a coal mine. The main functions include evaluating and scoring coal mine safety situation, reasoning, and interpretation of coal mine safety problems, configuration and management of diagnosis tasks, displaying online monitoring data, business management data, and evaluation scoring and reasoning analysis results based on 3D virtual mine visualization platform.
Conclusions
Using the massive data resources to deeply mine the hidden safety hazards and make accurate predictions and analysis is the basic premise for effective evaluation of mine production safety. The dynamic system of safety diagnosis designed in this paper meets this demand. Based on the Internet of things, multi-element sensors are deployed to collect safety monitoring data, and data storage, analysis, and diagnosis reasoning are carried out based on the cloud computing platform. Thus, the information on safety production factors such as equipment, personnel, and environment are related, and the diagnosis of mine production safety can be realized by using forward and reverse reasoning policy decisions.