Analysis of college emergency management events based on intelligent industrial Internet of Things (IIoT)

The Internet of Things information technology was proposed many years after the Internet. In fact, it is a derivative of the Internet. It was proposed at the official meeting in 2005. It took nearly 20 years from its inception to its development[16]. The combination technology of the Internet of Things and other related industries has been continuously innovated and discovered here, including the Internet of Things and teaching, the Internet of Things and medicine, the Internet of Things and automotive engineering, the Internet of Things and energy consumption, the Internet of Things and industrial equipment, etc. The Internet of Things has the technology of integrating various devices, and supports intelligent evaluation and decision-making system with artificial empowerment, which can provide guarantee for more intelligent solutions.

The intelligent decision support system of industrial IoT mainly has the following stages: intelligent upgrading of industrial equipment; Internet of Things connection of industrial equipment; Emergency processing data input processing; Decision making for handling emergency events; Deep learning. The general process is shown in Figure 2 below:

Figure 2.

Intelligent Industrial Internet of Things Decision Support System

The Augmented Lagrangian Coordination (ALC) is a novel distributed optimization method for multi-disciplinary design fields, originally proposed by Tosserams of Eindhoven University of Technology, to handle complex system optimization design issues on a large scale. The basic idea of the ALC method is to decompose complex problems into multiple elements with autonomous decision-making power, and then use specific distributed coordination strategies to obtain the global optimal solution. (1)minz1,z2,…,z7 f=f1+f2=z12+z22 s.t. g1=(z3−2+z42)z5−2−1≤0 g2=(z52+z6−2)z7−2−1≤0 h1=(z32+z4−2+z52)z1−2−1=0 h2=(z52+z62+z72)z2−2−1=0$$\begin{array}{*{20}{c}} {\mathop {\min }\limits_{{z_1},{z_2}, \ldots ,{z_7}} }&{f = {f_1} + {f_2} = z_1^2 + z_2^2} \\ {{\text{s.t.}}}&{{g_1} = \left( {z_3^{ - 2} + z_4^2} \right)z_5^{ - 2} - 1 \le 0} \\ {}&{{g_2} = \left( {z_5^2 + z_6^{ - 2}} \right)z_7^{ - 2} - 1 \le 0} \\ {}&{{h_1} = \left( {z_3^2 + z_4^{ - 2} + z_5^2} \right)z_1^{ - 2} - 1 = 0} \\ {}&{{h_2} = \left( {z_5^2 + z_6^2 + z_7^2} \right)z_2^{ - 2} - 1 = 0} \end{array}$$

After introducing auxiliary variables and consistency constraints, the original geometric programming problem can be transformed as follows: (2)minz1,z2,...z7,z5[1],z5[2]f=f1+f2=z12+z22 s.t. g1=(z3−2+z42)(z5[1])−2−1≤0 g2=((z5[2])2+z6−2)z7−2−1≤0 h1=(z32+z4−2+(z5[1])2)z1−2−1=0 h2=((z5[2])2+z62+z72)z2−2−1=0 c=[c1,c2]T=[z5−z5[1],z5−z5[2]]T=[0,0]T$$\begin{array}{rcl} \mathop {\min }\limits_{{z_1},{z_2},...{z_7},z_5^{[1]},z_5^{[2]}} \quad \quad \quad \quad \quad \quad f = {f_1} + {f_2} = z_1^2 + z_2^2 \\s.t.\begin{array}{*{20}{c}} {{g_1} = \left( {z_3^{ - 2} + z_4^2} \right){{\left( {z_5^{[1]}} \right)}^{ - 2}} - 1 \le 0} \\ {{g_2} = \left( {{{\left( {z_5^{[2]}} \right)}^2} + z_6^{ - 2}} \right)z_7^{ - 2} - 1 \le 0} \\ {{h_1} = \left( {z_3^2 + z_4^{ - 2} + {{\left( {z_5^{[1]}} \right)}^2}} \right)z_1^{ - 2} - 1 = 0} \\ {{h_2} = \left( {{{\left( {z_5^{[2]}} \right)}^2} + z_6^2 + z_7^2} \right)z_2^{ - 2} - 1 = 0} \\ {c = {{\left[ {{c_1},{c_2}} \right]}^T} = {{\left[ {{z_5} - z_5^{[1]},{z_5} - z_5^{[2]}} \right]}^T} = {{[0,0]}^T}} \end{array} \\ \end{array}$$

Although the local constraints of the decomposed elements are independent of each other, the introduction of consistency constraints still results in coupling relationships between the decomposed elements. (3)ϕ(c)=vTc+∥w∘c∥22=ϕ1(c1)+ϕ2(c2)=v1(z5−z5[1])+∥w1(z5−z5[1])∥22+v2(z5−z5[2])+∥w2(z5−z5[2])∥22ϕ1(c1)=v1(z5−z5[1])+∥w1(z5−z5[1])∥22ϕ2(c2)=v2(z5−z5[2])+∥w2(z5−z5[2])∥22$$\begin{array}{*{20}{c}} {\phi (c)}&{ = {v^T}c + \parallel w \circ c\parallel _2^2 = {\phi _1}\left( {{c_1}} \right) + {\phi _2}\left( {{c_2}} \right)} \\ {}&{ = {v_1}\left( {{z_5} - z_5^{[1]}} \right) + \parallel {w_1}\left( {{z_5} - z_5^{[1]}} \right)\parallel _2^2 + {v_2}\left( {{z_5} - z_5^{[2]}} \right) + \parallel {w_2}\left( {{z_5} - z_5^{[2]}} \right)\parallel _2^2} \\ {}&{{\phi _1}\left( {{c_1}} \right) = {v_1}\left( {{z_5} - z_5^{[1]}} \right) + \parallel {w_1}\left( {{z_5} - z_5^{[1]}} \right)\parallel _2^2} \\ {}&{{\phi _2}\left( {{c_2}} \right) = {v_2}\left( {{z_5} - z_5^{[2]}} \right) + \parallel {w_2}\left( {{z_5} - z_5^{[2]}} \right)\parallel _2^2} \end{array}$$

After relaxing the consistency constraint, the original geometric programming problem can be expressed as: (4)minz1,z2,…,z7,z5[1],[]52]z12+z22+ϕ1(c1)+ϕ2(c2)s.t.g1=(z3−2+z42)(z5[1])−2−1≤0g2=((z5[2])2+z6−2)z7−2−1≤0h1=(z32+z4−2+(z5[1])2)z1−2−1=0h2=((z5[2])2+z62+z72)z2−2−1=0$$\begin{array}{*{20}{l}} {\mathop {\min }\limits_{{z_1},{z_2}, \ldots ,{z_7},z_5^{[1]},[]_5^{2]}} z_1^2 + z_2^2 + {\phi _1}\left( {{c_1}} \right) + {\phi _2}\left( {{c_2}} \right)} \\ {{\text{s.t.}}\quad {g_1} = \left( {z_3^{ - 2} + z_4^2} \right){{\left( {z_5^{[1]}} \right)}^{ - 2}} - 1 \le 0} \\ {{g_2} = \left( {{{\left( {z_5^{[2]}} \right)}^2} + z_6^{ - 2}} \right)z_7^{ - 2} - 1 \le 0} \\ {{h_1} = \left( {z_3^2 + z_4^{ - 2} + {{\left( {z_5^{[1]}} \right)}^2}} \right)z_1^{ - 2} - 1 = 0} \\ {{h_2} = \left( {{{\left( {z_5^{[2]}} \right)}^2} + z_6^2 + z_7^2} \right)z_2^{ - 2} - 1 = 0} \end{array}$$ (5)minϕ(c) =vTc+∥w∘c∥22=ϕ1(c1)+ϕ2(c2) =v1(z5−z5[1])+∥w1(z5−z5[1])∥22+v2(z5−z5[2])+∥w2(z5−z5[2])∥22$$\begin{array}{*{20}{c}} {\min \phi (c)}&{ = {v^T}c + \parallel w \circ c\parallel _2^2 = {\phi _1}\left( {{c_1}} \right) + {\phi _2}\left( {{c_2}} \right)} \\ {}&{ = {v_1}\left( {{z_5} - z_5^{[1]}} \right) + \parallel {w_1}\left( {{z_5} - z_5^{[1]}} \right)\parallel _2^2 + {v_2}\left( {{z_5} - z_5^{[2]}} \right) + \parallel {w_2}\left( {{z_5} - z_5^{[2]}} \right)\parallel _2^2} \end{array}$$

Formulation of decomposed element P₁: (6)minz1,z3,z4,z511z12+v1(z5−z5[1])+∥w1(z5−z5[1])∥22 s.t.g1=(z3−2+z42)(z5[1])−2−1≤0 h1=(z32+z4−2+(z5[1])2)z1−2−1=0$$\begin{array}{*{20}{l}} {\mathop {\min }\limits_{{z_1},{z_3},{z_4},z_5^{11}} z_1^2 + {v_1}\left( {{z_5} - z_5^{[1]}} \right) + \parallel {w_1}\left( {{z_5} - z_5^{[1]}} \right)\parallel _2^2} \\ {{\text{s}}{\text{.t}}{\text{.}}\quad {g_1} = \left( {z_3^{ - 2} + z_4^2} \right){{\left( {z_5^{[1]}} \right)}^{ - 2}} - 1 \le 0} \\ {{h_1} = \left( {z_3^2 + z_4^{ - 2} + {{\left( {z_5^{[1]}} \right)}^2}} \right)z_1^{ - 2} - 1 = 0} \end{array}$$

Formulation of decomposed element P₂: (7)minz2,z6,z7,z5[3]z22+v2(z5−z5[2])+∥w2(z5−z5[2])∥22 s.t.g2=((z5[2])2+z6−2)z7−2−1≤0 h2=((z5[2])2+z62+z72)z2−2−1=0$$\begin{array}{*{20}{l}} {\mathop {\min }\limits_{{z_2},{z_6},{z_7},{z_5}^{[3]}} z_2^2 + {v_2}\left( {{z_5} - z_5^{[2]}} \right) + \parallel {w_2}\left( {{z_5} - z_5^{[2]}} \right)\parallel _2^2} \\ {{\text{s}}{\text{.t}}{\text{.}}\quad {g_2} = \left( {{{\left( {z_5^{[2]}} \right)}^2} + z_6^{ - 2}} \right)z_7^{ - 2} - 1 \le 0} \\ {} \\ {\quad {h_2} = \left( {{{\left( {z_5^{[2]}} \right)}^2} + z_6^2 + z_7^2} \right)z_2^{ - 2} - 1 = 0} \end{array}$$

The deep learning mechanism of industrial Internet of Things is mainly to analyze the handling measures and results of college emergency events, and then use the corresponding data value extraction technology to improve the deficiencies. This research uses the deep learning mechanism based on convolutional neural system computing to combine the intelligent industrial Internet of Things learning mechanism to improve its expression[17].

Only the industrial Internet of Things in-depth learning system first uses WK-IRT data processing technology to screen the data transmitted by different industrial equipment for emergency handling events, then uses FENR convolution network technology to process the data set and complete the conversion of two-dimensional data, then uses convolution calculation method to obtain the effective value after data processing, and then uses CCD neural network data processing descending connection to connect too TXT connection, Obtain emergency treatment data that can reflect the real situation. Its learning mechanism is shown in Figure 3 below:

Figure 3.

Deep learning process of intelligent industrial Internet of Things

Data processing steps for in-depth learning of intelligent industrial Internet of Things:

Analyze the IP address of the data to find out whether the data propagation address is true and whether the data propagation meets the conditions at that time. Then use the no pot technology to analyze the two-dimensional data of data conversion, which includes the input result surface, the convolution result processing surface, the data output processing surface and the final decision result output surface. The number of several different surfaces needs to be set according to the actual situation of deep learning, which may be majority or minority. The main calculation formulas are shown in Formula 8 to Formula 10: (8)K(z,x,c,s)=∑i=1nzn0.5+logx(x−c)lncn$$K(z,x,c,s)=\frac{\sum \limits_{i=1}^n\,{{z}_{n}}0.5+\log x(x-c)\ln c}{n}$$ (9)p(zn=a|xc,c〉)=exc∑i=1nxnce$$p\left( {{z}_{n}}=a|xc,c \rangle \right)=\frac{{{e}^{xc}}}{\sum \limits_{i=1}^n\,{{x}_{n}}{{c}^{e}}}$$ (10)M=(m−xc)∇z(z,x,c)$$M=(m-xc)\nabla z(z,x,c)$$

Where, K is the entropy change value of data that may be lost; Z Data elimination of abnormal values; X is the true value of data discrimination; C is the output condition threshold; M is the optimal solution of data fitting parameters; M is the result theory of data parameter fitting.

The final in-depth learning mechanism model of intelligent industrial Internet of Things is shown in Figure 4 below.

Figure 4.

Deep learning model of intelligent industrial Internet of Things

This deep learning model is based on the recognition of the results of convolutional neural analysis data processing. Although it can filter the data transmitted by the Industrial Internet of Things, the encoding process is easy to lose its authenticity. If the number of neural network training is too small, or the setting of the input threshold is unreasonable, the final output results will also be affected. This process needs to be optimized and improved.

Different models have different computational depths. Data based on the state of intelligent industrial Internet presents an unstable and uncertain form, which brings considerable difficulty to data processing. Different models have different processing methods, and the results are different.

The above two summaries introduce the evaluation system setting and deep learning mechanism of industrial IoT only. Next, we need to set up the decision-making management model of intelligent IoT according to these two aspects[18].

In colleges and universities, the handling of industrial equipment emergencies mainly includes short circuit and spontaneous combustion of industrial equipment, forced stay of industrial equipment in automatic operation, and uncontrolled industrial equipment in laboratories, which must be controlled.

The decision management model of only industrial Internet of Things proposed in this study mainly includes the following three types: XT-A decision management model; GT-U decision management model; IIT decision-making management model[19].

The biggest advantage of XT-A decision-making management model is that it can predict the occurrence probability of various emergency events before the occurrence of emergency events with high accuracy. It can provide more manual enabling conditions when emergency events occur, which will be conducive to the decision-making of emergency event management on the surface. However, this decision-making management model cannot make choices in the shortest time, and may not reflect its advantages in the process of more refined processing.

The biggest difference between GT-U management decision-making model and XT-A decision-making management model is that it cannot accurately predict the probability before an emergency event occurs. Although this decision-making management model can also predict the occurrence of an emergency event, the prediction result is biased from the actual situation because the data transmission and storage efficiency of the industrial Internet of Things is not high. However, this decision-making management model can give a more reasonable solution to a considerable extent, and ensure the safety of teachers and students to the greatest extent.

IIT decision-making management model integrates the unified advantages of XT-A and GT-U management decision-making models, and is a very balanced intelligent industrial Internet of Things decision-making management model. In the process of responding to emergency events, we can not only consider the accuracy of prediction results, but also put forward corresponding solutions when emergency events occur. However, its prediction accuracy is not as high as XT-A decision-making model, and the processing plan is not as reasonable as GT-U decision-making management. The only thing is equilibrium.

These models restrict and achieve each other. Emergency events can be classified into different levels in different industrial IoT emergency events in a university, and appropriate decision-making management models can be used according to different levels, as shown in Figure 5 below:

Figure 5.

Intelligent Industrial Internet of Things Decision Model

Heinrich’s Law tells us that any emergency accident occurs after the accident hidden danger given by Sabah is not handled in time and there are nearly 30 small accidents. This rule tells us to put the safety valve in the first place at any time[20]. The probability of safety accidents in colleges and universities is small, but once they occur, they will affect a wide range of areas, such as stampede, fire, explosion, especially in the places where science and engineering experiments are gathered. Once an emergency event occurs, it is likely to cause a large number of casualties.

And the intelligent industrial Internet of Things technology has solved the two major conflicts in college emergency handling. The intelligent industrial Internet of Things makes it possible to reflect the real situation by connecting more industrial equipment in colleges and universities, and completes the handling of emergency events through the artificial intelligent processing model. The intelligent Internet of Things emergency response of colleges and universities with artificial empowerment mainly includes student dormitories, industrial laboratories, and college fire protection.

The application of industrial Internet of Things technology can guarantee the personal safety of teachers and students in colleges and universities, and reduce the emergency management accidents due to lack of experience. Before the emergency management accident, the Industrial Internet of Things can comprehensively monitor colleges and universities through data analysis and type comparison, and can also complete the prediction of emergency events through data analysis. Industrial IoT technology can reduce the probability of safety accidents caused by human negligence; When an accident occurs, the powerful intelligent analysis technology of the Industrial Internet of Things can make scientific analysis and evaluation of campus security accidents, automatically generate feasible suggestions for accident disposal, and provide reference for further resource allocation and work development decisions..

The college emergency management model of intelligent Internet of Things with artificial empowerment is mainly realized through data collection, data preprocessing, RFID technology and FTA IOT technology[21]. The intelligent industrial Internet of Things technology takes colleges and universities as a whole, and different industrial equipment as a sub target. These target equipment can collect data from funny different regions when an emergency occurs, and then complete the regional definition through the analysis of the data, notify the corresponding competent departments to respond, and also complete the accident probability prediction before the emergency occurs through resource interaction and intercommunication, The living space of teachers and students in colleges and universities is quite high.

The basic terms of RFID technology are people-oriented. When an emergency occurs, we will judge which decision can help more teachers and students through data and make a choice. The use of RFID technology also needs to pay attention to its system relevance. It is necessary to select an appropriate operation model among managers to solve various problems in emergency management leadership, management authority and management level. Its security also needs to be guaranteed. The characteristics of security functions are reflected in the daily lives of college teachers and students. In order to avoid false judgments caused by information leakage or abnormal data errors, it needs certain security. Table 1 and Table 2 show the user personnel and basic information.

FTA-IOT analysis method needs to improve the data processing of intelligent industrial Internet of Things through complete logical analysis. The main analysis objectives include the number of teachers and students affected, accident prediction and handling, accident decision-making, etc. The number of teachers and students affected mainly includes how many teachers and students will be affected by the emergency accident. The way to reduce this injury may involve trade-offs, which requires manual empowerment to set the intelligent industrial Internet of Things technology in advance. On the other hand, it is necessary to make decisions on the handling of accidents. For example, when fire emergency accidents occur. The main comparison is shown in Figure 6 below:

Figure 6.

Intelligent Industrial Internet of Things Model with Artificial Enabling

The emergency management model of intelligent industrial Internet of Things colleges and universities with artificial empowerment is introduced above, which also applies to remote sensing and monitoring systems of industrial equipment, but its application degree is low, and it is not fully used. This section will improve the emergency management model according to the remote sensing and intelligent monitoring system of industrial equipment in colleges and universities[22].

The number of industrial equipment for emergency management in colleges and universities is large in science and engineering laboratories. Many colleges and universities have a large number of chemical agents in their laboratories. An accident will break out in a flash, which will also have unpredictable consequences. Therefore, most chemical laboratory equipment has remote sensing and monitoring systems of varying degrees, and other systems without remote sensing and monitoring should be installed in place. The remote sensing and monitoring system of industrial equipment is used to track and trace all equipment and reagents, and the RFID system mentioned above is used to label different equipment and reagents.

The intelligent monitoring system should not only include the monitoring equipment in colleges and universities, but also include the intelligent APP system that can be used by all teachers and students. This kind of monitoring equipment can be used by industrial equipment of schools, dormitories, teachers and libraries, and can be connected with them through smart phones, so that we can know where it comes from, where it is now, who is using it, and it is not allowed to use it or when the probability of accidents during use is high, so that we can control the occurrence of emergency accidents to a large extent.

The dynamic receiving data of the fire protection system in colleges and universities needs the support of remote sensing and monitoring systems of equipment. The remote sensing and monitoring system of fire protection equipment collects relevant data, transmits it to the intelligent industrial Internet, forms data exclusion nodes, uses convolution algorithm for deep learning, and gives appropriate emergency management decisions, which can avoid such accidents[23].

The use of remote sensing and intelligent monitoring system in the intelligent industrial Internet of Things can provide more accurate data, and can also remotely control the operation of equipment through decision-making after data analysis, which largely controls the source of emergency events..

Machine learning mechanism is a refined branch of the deep learning mechanism of intelligent industrial Internet of Things. Machine learning mechanism is constantly self updating and self-learning in daily life by setting intelligent mechanisms for different industrial equipment. Machine learning mechanism is a learning phenomenon[24]. If an industrial equipment cannot provide good quality conditions, it will not be able to complete the whole stage of machine learning. The most important part of machine learning is to judge whether it is part of emergency event management through daily behavior observation, which mainly includes: emergency response event analysis and learning, emergency treatment scheme learning, and emergency equipment data updating[25].

The emergency response event analysis of intelligent industrial Internet of Things is to help industrial equipment determine whether there is a need for emergency treatment by observing the activities of teachers and students and their impacts in daily life. For example, in the emergency event analysis of industrial equipment in the laboratory, the dangerous actions and dangers that occur in the process of students’ experiments should be input into the program in a timely manner. If these actions are found in the experiment, focus on whether they will cause harm, so as to learn.

Machine learning of emergency response plan is to judge whether an emergency response measure needs to be included in the scope of review procedure by comparing all emergency response measures encountered in daily life and their final results. The next time a similar event occurs, the emergency machine learning event handling measures can be improved through the previous event analysis.

In machine learning, emergency event processing data update is the most important means of machine learning mechanism. Data update is based on effective measures for emergency event processing and accurate identification of emergency events. The updating of data is very cautious. Generally speaking, the learning process of industrial equipment is the industrialization product of manual empowerment, and the setting of its programs is generally the ability of manual empowerment. However, in some cases, some uncontrollable factors may occur, such as judging whether the handling of an emergency event is effective or how many levels an emergency event is divided into, which may lead to situations where artificial intelligence cannot judge. This requires judgment through machine learning. For example, if similar situations occur in life, managers’ handling methods and effects are recorded. If the results are good, they can participate in learning, If the results are not displayed, it is necessary to predict such emergencies to avoid such emergencies..

he above chapters have obtained the analysis model of intelligent industrial Internet of Things in college emergency management practice through the analysis of only industrial Internet of Things. Next, we use this model to predict the probability of an emergency event in a college. The results are shown in Figure 7 below:

Figure 7.

Probability prediction of college emergency time

From the data shown in Figure 7, we can see that the prediction results of the college emergency event prediction model obtained through remote sensing and intelligent monitoring are consistent with the actual results, but the prediction results in some areas are idealistic and cannot represent the actual probability of occurrence. However, after updating the number of machine learning models that can only be used for industrial IoT through manual empowerment, more and more data show that the results are representative and can be well integrated with the actual results. All the data in the figure are prediction results, which cannot reflect the real situation. Some prediction values are relatively rational within the regional scope, and may also have strong process response integration, which cannot map to the probability of emergency events in real life, but they have certain reference value. The prediction of emergency events can be used as a reference for daily life, and the evaluation of its results can also be achieved using only industrial Internet of Things technology. However, it is better for experienced administrators of colleges and universities to regularly check and handle the occurrence of emergency events. After all, machine learning needs a certain amount of data to improve its fundamentals and needs to wait for maturity.

The above results are better by comparing the predicted emergency event values of different models with the actual values. Now we analyze the impact of the six measures of hard extrusion after the emergency event, as shown in Figure 8:

Figure 8.

Comparison of emergency response measures

From Figure 8, we can see that in most cases, the emergency response measures taken by using intelligent industrial IoT technology can achieve good results, but in some cases, they can not play their advantages. This may be because some situations have certain particularity, and the probability of occurrence in daily life is very small.

If they are not captured by intelligent industrial IoT algorithms, there will be no in-depth learning and machine learning, It is impossible to propose effective solutions when such emergencies occur, which is also a problem that needs to be solved in the follow-up study.

Funny emergency response measures can use contact databases, not necessarily regional databases, but international databases or actual cases. Input relevant data in cases, and get better results through in-depth learning.

Whether in colleges and universities or in other regions that need industrial Internet, human beings need to endow the industrial Internet of Things with certain intelligence to ensure that only the industrial Internet of Things will not be rigid when dealing with emergencies, can be sustainable and will not be abandoned by the times.

The most important sustainable development strategy is: intelligent industrial Internet of Things through artificial empowerment, then use deep learning strategies to constantly learn emergency event prediction, evaluation and decision-making, and then use machine learning mechanisms to constantly update the database, replace bad prediction methods and emergency event handling measures that cannot meet the reality, so as to obtain a sustainable intelligent industrial Internet of Things. Different prediction results use different emergency plans.

Under the combined use of deep learning and industrial Internet only, machine learning accounts for a large proportion. In order to get more realistic prediction results and emergency response plans, we improve the efficiency and accuracy of intelligent industrial Internet by studying its sustainable development.

Its sustainable development process includes artificial empowerment, deep learning and machine learning, and its impact on sustainable development is shown in Figure 9 and Figure 10 below:

Figure 9.

Analysis of factors affecting sustainable development

Figure 10.

Application of sustainable development results of intelligent industrial Internet of Things

From Figure 9, we know that in most cases, manual empowerment is the most critical step in the handling of emergency events in colleges and universities of the Industrial Internet of Things. Without all the steps behind this step, it would be empty talk. However, this step does not have the greatest impact, and the biggest impact is machine learning. This can also reflect from the side that the data update of machine learning is a necessary means for the sustainable development of the intelligent industrial Internet of Things in emergency events. The data update represents that the system can complete the transformation of the times and improve the system perception.

The results in Figure 10 show that the sustainable development of the intelligent industrial Internet of Things in college emergency handling accounts for a large proportion. The sustainable development can be reflected not only in the update of system data, but also in the service life of industrial equipment. The number of funny industrial equipment is increasing, and most of them are expensive. If we can increase the service life of industrial equipment, it will undoubtedly reduce some of the funding pressure of colleges and universities. The sustainable development process of industrial Internet of Things will be emphasized in the follow-up research, which is closely related to the national policy and the importance of ensuring the safety of college teachers and students. If we can play a good leading role in this process, the prediction and handling of emergency events in colleges and universities will become easier.