Abstract
In order to improve the effect of smart city construction, this paper combines smart buildings and ethical computing to conduct research on smart city edge computing. The new smart city architecture based on the flexible deployment of edge computing and data slicing capabilities provides support for the transformation of smart city construction from hardware embedded technology, access means, and software data processing. Moreover, this paper uses information technology to collect, process, analyze, use the information to achieve intelligence, and integrate resources and information of cities and people to build a smart city functional architecture. Moreover, this paper combines simulation technology for experimental research. Through experimental analysis, it can be seen that the smart city edge computing method based on smart buildings and ethical computing proposed in this paper has good results.
1. Introduction
The goal of the smart city is to link, integrate, and improve the internal systems and various services offered by the city, as well as to integrate these efforts with the city’s building plans. It increases the pace at which different resources are allocated in the city, optimizes urban services and urban administration, advances the city to a higher level, and raises the standard of living and productivity for city residents [1]. The newest addition to the digital city is the smart city. The Internet of Things, cloud computing, and wireless mobility are the key representations of the new information technology generation that gives rise to the digital city. Based on digital cities, smart cities have the potential to connect things to other things, perceive each other, and use high-level information in ways that benefit both people and people. People, the Internet, and Things are widely and profoundly interconnected, coexisting, and interacting thanks to the network’s ubiquity, the vast processing of information and data (cloud storage, cloud computing), and the integration of the Internet of Things and the Internet [2].
This allows all aspects of the world to achieve a more thorough perception, wider interconnection, and deeper intelligence. Different angles can be taken to examine the growth of smart cities. Technology-wise, smart cities should build their infrastructure using cutting-edge information technologies like the Internet, cloud computing, and big data in order to achieve urban connectedness and coverage. From the standpoint of social development, smart cities use social networks, mass media, and other modern tools and applications to facilitate communication between residents and the city, realize innovative development, and foster the sustainable and creative growth of the economic, cultural, technological, and spiritual societies. A smart city encompasses technology, industry, application, service, government, humanities, living, etc. as its content. The application of wisdom, which addresses the four major facets of food, clothing, housing, and transportation, is most closely related to people’s lives and primarily reflects changes in agriculture, food, medical care, transportation, schools, communities, agriculture, enterprises, and governments [3].
The creation of smart cities has emerged as a new model and choice for contemporary city construction thanks to the advancement and support of cutting-edge technologies like big data, the Internet of Things, and artificial intelligence. Governments at all levels, businesses, and individuals have adopted the idea of smart cities and tried to integrate new Internet-based technology in order to improve living conditions and preserve the environment in the face of increasingly serious urban problems. The need for a new model of contemporary urban development has prompted the emergence of smart cities. One of the key trends and features of urban development is the improvement of urban governance and innovation. It is also a cutting-edge theory and area of research for furthering the optimization of urban organizational structure and the urban informatization process. In the modern day, improving urban intelligence is essential for urban growth since it improves people’s quality of life and positions oneself to take advantage of fresh prospects for urban development.
This article combines smart buildings and ethical computing to study the edge computing of smart cities, which provides a theoretical reference for the construction and development of smart cities.
2. Related Work
Kocakaya [4] believes that smart urban management is a new urban management model supported by a new generation of information technology and a knowledge society innovation 2.0 environment. Through the support of a new generation of information technology, it can achieve comprehensive and thorough perception, broadband ubiquitous interconnection, intelligent integration applications, and promote people-oriented sustainable innovation characterized by user innovation, open innovation, mass innovation, and collaborative innovation. Smart urban management is an evolution of digital urban management to deal with increasingly complex urban governance issues [5]. To realize the intelligent transformation from digital urban management, Ustinovičius et al. [6] propose to establish a complete urban information database, based on 5G, Internet of Things, City Information Model (CIM), and other technologies, combined with the already relatively complete urban management in digital urban management Component information, urban management industry data, related industry applications, and public appeal data form a city information database that can comprehensively portray city portraits. Fadeyi [7] summarized a set of methodologies that can process urban information data. A complete city information database must have the characteristics of big data, that is, the amount of data is huge, the types are various, the value density is low, and the data speed is fast. Mandičák et al. [8] pointed out: should pay attention to the overall data thinking, tolerate the confounding of data, pay more attention to correlation rather than causality, and so on. The ideal big data methodology should be able to complete the analysis and processing of massive data in accordance with established requirements and also provide reference data support for unspecified requirements. Ustinovichius et al. [9] construct an information framework that can be iteratively upgraded. Data only provides evidence to solve the problem, but cannot really solve the actual problem. The treatment of urban issues should also be under an information framework. This requires a standardized deconstruction of the problems that the city urgently needs to solve and the resources of all parties involved in urban governance and provides optimal solutions through algorithms. Then what we need to do is to continuously optimize iteratively. Wei and Chen [10] created a set of management systems to reduce internal management consumption so that different subjects are no longer single vertical command management, but horizontal multidimensional cooperation, coordination, and information exchange.
Smart cities are considered to be a complex ecosystem with the potential to improve the livability, workability, and sustainability of cities through human, process, and data networks [11]. Clarifying the meaning of “smart” in the urban context, determining the main dimensions and elements of smart cities, and comparing different measurement standards of smart cities are the basic conditions for the development of smart cities [12]. Smart city construction currently necessitates technological advancement and demand stimulation. On the one hand, tremendous technological advancement has resulted in a thriving smart city product. On the other hand, cities must urgently address efficiency and sustainability challenges in order for smart cities to become a fruitful ground for economic development [13–15]. The sustainability of resources is an important component in fostering the growth of cities and regions. Smart cities are characterized as cities that are sustainable and efficient in their use of resources to provide a high quality of life [16, 17]. Smart resource cities have emerged as Europe’s most recent urban development plan. Several smart energy city projects have been established using advancements in information and communication technology. Research has found that the key components of smart energy city projects are strategically sustainable and consistent [18–20]. Chen et al. [21] study how to promote and realize the sustainability of urban resources through smart city initiatives. The study found that although sustainability is not always the main goal of local implementation of smart city projects, smart city construction projects have increased the ambition to achieve resource sustainability goals. Furthermore, smart cities are built on the advancement of new technologies. However, modern technology is rarely used to drive smart city sustainability initiatives [22]. Simultaneously, smart cities have enormous promise in terms of urban mobility, development, and resource (and related greenhouse gas) conservation [23].
The importance of urban sustainability analysis has been extended to the construction of smart cities [24].
The idea of smart, sustainable cities has gained popularity recently. Smart cities have gained widespread attention as a viable solution to urban sustainability issues [25]. In light of this, the idea of a smart sustainable city is put out using studies on urban sustainability and smart cities. This idea focuses on how programs for smart, sustainable cities merge technology and the natural world. It identifies three components of such cities: The creation of spatially advanced, environmentally friendly communities; the digitalization of city infrastructure; and the joint experimentation of digital and low-carbon technology. Smart, sustainable cities are now the new standard for constructing and growing cities. According to studies, the growth of the green economy will have a significant impact on how European smart cities develop in the future [26]. Future smart cities will need to take into account topics like how smart cities support urban sustainability and how sustainability in smart cities supports the growth of smart cities sustainably. Existing research demonstrates that smart cities are directly linked to innovation, technology, and the economy, but that they do nothing to advance sustainability.
3. Intelligent Building and Ethical Decision-Making Feature Calculation
The steps in this paper to calculate the decision features are as follows:(1)This article first classifies the data according to different scene characteristics, and scenes with the same characteristics are classified into the same type of calculation data.(2)Secondly, this paper calculates the value of decision risk and moral strength under different choices according to the above calculation formula on the classified scene data.(3)Finally, this paper uses two decision-making features as coordinates to obtain the distribution map of decision-making features for each type of scene. The following first mainly introduces the specific meaning and calculation formula of the two decision-making characteristics of decision-making risk and moral strength.
The main features used in this paper are decision risk and moral strength . Among them, decision risk can also be understood as construction risk in the road construction area, which is represented by decision risk here. First, the calculation of decision-making risk is explained, which is mainly divided into three aspects to consider People: the above research shows that there are more protected element attributes such as female, young, few people, healthy, low social value, and human beings in the scene selection, which will increase the risk of decision-making. It can also be understood that the presence of certain types of characteristic elements in the scene will increase the risk of construction decision-making in the building area, and the decision should be made cautiously. Construction equipment: if there are people in the automatic control of construction equipment, it will increase the risk of decision-making; Building area: The existence of obstacles in the building area and intelligent building construction will also increase the risk of building construction decision-making for construction equipment. The presence of obstacles will make the decision more complicated. Intelligent building construction will increase the possibility of movement of elements in the scene, and complex behavior will increase the risk of decision-making.
According to the above description, the performance indicators in the decision risk degree are linearly combined, as shown in the following formula:In the formula, : represents the degree of decision-making risk brought by the construction of a smart city. : represents the decision-making risk caused by obstacles in the construction area. Specifically, it characterizes the decision-making conflicts between the construction personnel and the protection of pedestrians in the protection of the construction equipment in the data set, so the construction equipment will increase the decision risk. : represents the decision-making risks caused by different genders in the scene, and women are more likely to bring decision-making risks. : represents the decision-making risks brought by the healthy elements in the scene, and high health is more likely to bring decision-making risks. : represents the decision-making risk brought by the social value of the elements in the scene, and high social value will increase the decision-making risk. : represents the decision-making risk brought by the age of the elements in the scene, and a low age will increase the decision-making risk. : represents the decision-making risk brought by the ethnic category of the elements in the scene, and humans will increase the decision-making risk compared to animals. : represents the decision-making risk brought by the number of elements in the scene, and the larger the number of people, the higher the decision-making risk.
Among them, is the coefficient group of each characteristic index.
The meanings of the above indicators are set according to the aforementioned research and are in line with the moral decision-making judgments of most decision-makers. The specific quantitative values are expressed in the number of attributes. The calculation formula of each indicator is expressed as follows:
In the formula, each indicator is quantified by the number of elements in the data set. Among them, PedPed and barrier are 1 means that the construction equipment vehicle carries passengers and obstacles increase the risk of decision-making, otherwise, there is no passenger. When the crosignal is 1, it means that there is intelligent building construction in the building area, which increases the decision-making risk.
According to the research, the coefficient of each decision index is defined as the selection frequency difference ∆P of each index.
Through calculation, μ = (0.1, 0.15, 0.18, 0.2, 0.35, 0.49, 0.58, 0.65) can be obtained. When it is substituted into formula (1), we can get the following formula:
In the formula, is the decision-making risk brought by the construction of intelligent buildings, and there is no correlation coefficient.
The calculation of moral strength is explained below. Moral strength is mainly from six aspects result size , social theory , effect possibility , instantaneity , intimacy , and affect concentration .Among them, : represents the total harm of the ethical decision-making results to the victim, and represents the total number of elements of the victim in this paper. : represents the impact of ethical decision-making results on society or the degree of social comment on ethical decision-making results. This article uses legal effects to characterize social commentary, and smart building construction will increase the legal effects brought about. : represents the possible harm caused by the result of ethical decision-making. In this paper, the injury possibility of the most protected element is used to express it, and the injury value of the young, adult, and elderly among the elements is estimated, respectively. : represents the time interval between making a decision and causing injury, where changing lanes will increase the duration. : represents the degree of relationship between decision-makers and potential victims. In this paper, the masses are more intimate with construction personnel and automatic control construction equipment. : represents the degree of concentration of injury caused by the decision-making result. The proportion of people in the total number of people in this paper indicates a high degree of concentration of injury.
The above description has corresponding changes according to the element form of the data set. The injury of the decision result is serious injury or death, that is, the degree of injury is high. The specific quantitative value is expressed by the number of attributes, and the specific calculation formula for the six indicators of moral strength is as follows:
In formula (7), . According to the failure research of construction equipment, people of different ages have a different probability of being seriously injured. The specific P can be calculated according to the following formula. If it is assumed that the automatic control of the brake failure is constructed in the urban road construction area.
We can get . When it is substituted into formula (7), we can get the following formula:
The parameters in the above formula are mainly derived from the moral machine test data set. The final parameters are two decision feature values and .
The following uses the obtained feature formulas to process and calculate the data set, and, respectively, use the decision feature values to coordinate plots for analysis. After classifying and processing the result of feature value calculation, the decision feature distribution of each type of scene is obtained as shown in Figures 1(a)–1(f).
(a)
(b)
(c)
(d)
(e)
(f)
The distribution of decision feature values in each feature scene is shown in Figures 1(a)–1(f). The red point set in the figure represents the distribution of decision-making features under the default selection, and the blue point set represents the distribution of decision-making features under the nondefault selection. It can be seen from the figure that there are central effects and hierarchical effects in the distribution of different decision-making and decision-making features. That is, the area of specific decision-making feature values tends to have the same decision-making choice, and each layer has a clear boundary.
Therefore, this paper draws the following conclusions on the regular characteristics of the decision feature distribution map:(1)Decision-making dilemmas with different characteristics have different distributions of decision-making characteristics, that is, different decision-making laws(2)In the same feature decision scenario, different feature value areas have different decisions; that is, the decision rules can be explained(3)In the same feature decision scenario, the distribution of decision feature values under the same decision is discontinuous and has the characteristics of discrete stratification
The analysis of the data processing results shows that the decision results under the six dilemma scenarios are distributed in different decision features. The following characteristics can be summarized from the analysis results and the decision feature map:(1)In the same type of ethical decision-making dilemma scenario, the decision-making results in the specific environment characterized by different decision-making characteristics show a regular and hierarchical distribution(2)The distribution of each type of hierarchical decision feature in the decision feature map has a relatively regular shape as a whole and presents a central effect(3)The decision features of each layer in the decision feature map show a nonlinear distribution
Generally, the following basic concepts are included in the pattern recognition classification problem:
3.1. Features
In the classification problem, the feature is the description of the event or problem to be classified. The characteristics of the event are the common attributes of the event, and the model classifies the event by identifying the characteristics of the event. In ethical decision-making problems, the characteristic attributes of the problem can be represented by the characteristic vector group, where is the characteristic attribute of the model to the current environmental conditions in the corresponding type of environment.
3.2. Class Label
Class is a limited set of categories of classification problems, which represents the result of event classification, which can be represented by . In ethical decision-making problems, the set of classes can be .
3.3. Training Data Set
The training data set of the classification problem is composed of events with known class labels and the characteristic attributes of each event. Therefore, the training data set is a collection of event samples of known categories, which are generally determined by experimental data or empirical data.
3.4. Test Data Set
The main function of the test data set in the classification problem is to verify the feasibility of the classification model. Through the composition of randomly selected samples, the test data set is also a collection of event samples of known categories.
3.5. Classification Decision Model
The classification model is obtained through the learning or training of the training data set through the pattern recognition method, and then the optimized model is verified through the test data set. Generally, the form of classification model includes classification rules, mathematical expressions, or decision trees. The classification model is also the implicit classification law in the data.
The Bayesian classification method is mainly based on the Bayes theorem, which is the Bayes criterion proposed by Bayes. The theorem mainly describes the relationship between the prior probability, posterior probability, and conditional probability of an event through the Bayesian formula. When it is applied to classification problems, the theorem characterizes the relationship between the prior probability, posterior probability, and conditional probability of the category. In practice, two basic premises should be met before applying Bayes’ theorem to solve problems such as decision-making classification.(1)The probability distribution of each category is known; that is, the prior probability of each category and the probability distribution of the event itself(2)The number of classified categories is limited
We assume that we are studying an m-class classification problem, the sample category space is , the sample feature space is , and the number of samples in each category space is . If a sample belongs to the class , then represents the prior probability of class . It is the subjective determination of the probability of belonging to a class according to the number of samples, reflecting the knowledge of the possibility of this class. The prior probability of category can be estimated from sample data, such as . represents the conditional probability of the sample , that is, the probability distribution of samples in a category , which can generally be expressed by a probability density function. The posterior probability represents the probability that a given sample is determined to be a class . Combining the observed information, the prior probability is modified to obtain a more reasonable probability. It also reflects a kind of observational learning ability, which can be calculated by the Bayesian formula.
Thus, according to Bayes’ theorem, the following relationship can be obtained:
Among them, represents the probability density function of the sample feature.
In specific classification problems, the Bayesian learning classification method needs to adopt appropriate decision criteria according to different objectives. Common decision criteria include the minimum error rate criterion and the minimum risk criterion.
3.5.1. Minimum Error Rate Criterion
Misclassification is inevitable in the actual classification. If a sample originally belongs to a class but is classified into a class , there will be a phenomenon of misclassification. Therefore, in order to make the classification optimal, it is necessary to minimize the error rate of the classification, that is, to meet the minimum error rate criterion. It can be expressed as
In the formula, is the probability of dividing sample x into class 1.
In fact, the error rate can be minimized by maximizing the posterior probability, and the minimum error rate can also be expressed as follows:
Among them, is expressed as the probability of dividing sample x by mistake.
For sample point x, if it is classified into class 1, the error rate is .
Thus, the formula for minimizing the error rate can be obtained
Because the posterior probability can be calculated by Bayes’ theorem, the above formula can be transformed to
Therefore, minimizing the error rate can be seen as maximizing the posterior probability, and can also be seen as maximizing the prior probability and conditional probability.
3.5.2. Minimum Risk Criterion
In classification, not only the error rate of classification should be considered but also the risks brought by classification should be considered. Different classifications or decisions will bring different risks. When a penalty is added to the minimum error rate it indicates that the impact of classification errors is also a risk. When applying the minimum risk criterion, a loss matrix or loss function is required to represent the penalty coefficient for misclassification.
For example, for an m-class problem, there are m classifications in the decision space Y to classify the samples into different classes. We assume that the loss of a certain -type sample into a class by the j-th classification is . Then, the risk of misclassification of sample x is expressed as follows:
For a given sample x to minimize the risk, it can be expressed as follows:
Among them, is the loss function, and the commonly used loss functions have 0–1 loss functions, that is, the correct classification loss is 0, and the misclassification loss is 1, as expressed by the following formula:
Generally, when the prior probability and conditional probability of the sample are known, both the minimum error rate criterion and the minimum risk criterion can be used to perform Bayesian optimal classification.
4. Smart City Edge Computing Based on Smart Buildings and Ethical Computing
Based on the policy, technology, and market support of the development of the government and the global information application market for the construction of smart cities, as well as the data explosion brought about by 5G technology, the existing technical architecture has become more and more difficult for the processing and interaction of massive real-time data. The new smart city architecture based on the flexible deployment of edge computing and data slicing capabilities provides strong support for the transformation of smart city construction in terms of hardware-embedded technology, access means, and software data processing. The new smart city system architecture based on edge computing is divided into 6 levels, as shown in Figure 2.
From the perspective of common application scenarios in smart cities, edge computing business types have diversified characteristics, such as industrial operations, smart driving, business orchestration, business data push, online marketing, factory supervision, and video collection. According to implementation methods and business scenarios, the edge computing of smart cities mainly includes three types telecom operators' edge cloud computing, enterprise and IoT edge cloud computing, and industrial edge cloud computing. The three types of edge cloud computing are not mutually exclusive. The specific situation is shown in Figure 3.
In industrial Internet application scenarios, edge computing mainly processes critical business data and improves real-time response speed by installing edge computing chips or small industrial computers on terminal equipment or original data recording nodes, as shown in Figure 4.
With the cooperation of edge computing, cloud computing, and 5G communication, we can extend the existing intelligence from the terminal side of the driving tool to the road environment, as shown in Figure 5.
The schematic diagram of the process of edge computing in the field of smart medical applications and the schematic diagram of the process of edge computing in the field of smart agricultural applications are shown in Figures 6 and 7, respectively.
The smart city management platform designed on the aforementioned basis is shown in Figure 8.
The simulation research of a smart city is carried out through the above system. Figure 9 shows the smart city model proposed in this paper.
(a)
(b)
(c)
(d)
On the basis of the above research, the system of this paper is verified by a simulation test of the smart city effect, and multiple sets of data are counted, and the results are shown in Figure 10 and Table 1.
From the above research, it can be seen that the smart city edge computing method based on smart building-ethical computing proposed in this paper has good results. All the effectiveness verification of the smart city effect are >80.25, and half of them are >90.12, which demonstrates that the Applications of uncertainty models as support in smart buildings and ethical computing in edge computing of smart cities that were proposed in this work have a positive effect on the information extraction process.
5. Conclusion
As a new model of urban development, smart cities have swept the world as soon as they were proposed, and have become an important way for the government to respond to reforms and innovations in all aspects of society. China has advanced the construction of smart cities to a strategic height and strives to develop them as an important way to build a new type of urbanization. Building a smart city has many advantages to urban development. Through the means of information technology, information is collected, processed, analyzed, and used to achieve intelligence, integrate resources and information between cities and people, to promote urban development, speed up information exchange, simplify work procedures, and improve work efficiency. Moreover, the development goal of smart cities is to use advanced information technology to promote the adjustment and upgrading of industrial structure, accelerate economic development, optimize the economic structure, realize new social resource allocation, realize industrial innovation and upgrade, and increase economic development. This paper combines smart buildings-ethical computing to conduct research on smart city edge computing and provides a theoretical reference for the construction and development of smart cities. The research shows that the smart city edge computing method based on smart buildings and ethical computing proposed in this paper has good results.
Based on edge computing, smart city buildings are expected to be intelligent and humanized, and the data generated by buildings can be well processed and utilized. Based on mobile edge computing and assisted by autonomous driving technology with deep learning and big data analysis, driverless services can achieve real-time positioning of vehicles that are not visible in the line of sight, collaborative hazard prediction of urban areas, and 3D map generation of autonomous driving. Edge computing-based multienergy networks can improve the overall efficiency and effectiveness of energy systems in areas of different sizes, such as parks, islands, and towns, including large buildings. Multienergy networks based on edge computing technology can integrate smart grid, heat supply and gas network, and network flow, and realize unified energy management in smart cities. As a new technology, the development of edge computing will also go through the process from scratch, from the beginning to maturity, and will also be affected by social needs and traditional rules. Although edge computing will bring great opportunities to the construction of smart cities from both technical and application levels, its development process will also face challenges from a technical, application, and even legal and ethical levels.
Data Availability
The data used to support the findings of this study are included within the article.
Conflicts of Interest
The authors declare that they have no conflicts of interest.