Internet of things
The Internet of things is the “Internet-connected with everything,” a vast network formed by the combination of various information sensing devices and the Internet, which can realize the interconnection of people, machines, and things anytime and anywhere18,19. The realization of the Internet of things mainly depends on specific communication protocols, which can unify various information formats. At this time, there is a need for various data collectors and sensor equipment to support the development and application of the Internet of Things. The Internet of Things architecture mainly includes the perception layer, network transmission layer, information processing layer, and application layer. Figure 1 shows the architecture.
Internet of things architecture.
In Fig. 1, the perception layer mainly collects relevant data for the Internet of things system, including various data acquisition technologies and sensing technologies, and can realize the dynamic connection and information collection. The network layer is mainly a variety of communication systems, which is the critical part of the Internet of things, providing a more efficient communication protocol to realize the Internet of things. The application layer is mainly the interface part of the Internet of things and users. It can obtain the corresponding data according to different industries, process, classify, screen, process the data according to the industry’s needs, and finally present the results to the users. The implementation of this part mainly depends on various databases and professional software20,21.
Data mining technology
Data mining technology mainly relies on four fundamental technologies: database, artificial intelligence, mathematical statistics, and visualization. Data mining technology’s algorithm input is the database, algorithm output is knowledge or pattern extraction and discovery, and algorithm processing is the specific design of search methods22. The description or illustration of algorithm design is mainly divided into three parts: input, output, and processing. The data mining algorithm mainly involves three aspects: mining objects, mining tasks, and mining methods.
The process of data mining is also called KDD (Knowledge Discovery in Database). Data mining refers to the extraordinary process of automatically extracting useful information hidden in data sets. Data mining technology in the mining methods is divided into four categories: statistical methods, machine learning methods, neural network methods, and database methods23. Statistical methods can be subdivided into regression analysis and discriminant analysis. Data mining technology comprises many mining objects such as relational databases, spatial databases, and text databases. Neural network methods can be subdivided into a forward neural network and a self-organizing neural network. Machine learning in data mining technology mainly means the genetic algorithm. Data mining technology is mainly based on multi-dimensional data analysis. Data mining is a long process and a specific step, which is mainly produced in KDD. Its main feature is that it can extract, transform, analyze, and model big data and extract critical data.
In order to meet the needs of information fusion, it is necessary to standardize it. Various signal indices are standardized by constructing transformation functions. Interval standardization functions can be expressed as:
$${a_i}={u_{{d_i}}}({x_i})\left\{ \begin{gathered} 1 – \frac{{\hbox{min} (m_{1}^{i} – {x_i},{x_i} – m_{2}^{i})}}{{\hbox{max} (m_{1}^{i} – {x_i},{M_i} – m_{2}^{i})}}\mathop {}\nolimits^{{}} {x_i} \notin [m_{1}^{i},m_{2}^{i}] \hfill \\ 1\mathop {}\nolimits^{{}} \mathop {}\nolimits^{{}} \mathop {}\nolimits^{{}} \mathop {}\nolimits^{{}} \mathop {}\nolimits^{{}} \mathop {}\nolimits^{{}} \mathop {}\nolimits^{{}} \mathop {}\nolimits^{{}} \mathop {}\nolimits^{{}} \mathop {}\nolimits^{{}} \mathop {}\nolimits^{{}} {x_i} \in [m_{1}^{i},m_{2}^{i}] \hfill \\ \end{gathered} \right.$$
(1)
In (1), \({d_i}=[m_{i}^{{}},{M_i}]\) indicates the threshold of \(i\), \({x_i}\) is the actual measured value of the target, and \(\operatorname{m}\) is a fixed value.
System requirements analysis
In the system’s design, it is necessary to consider the system requirements, which is also the key to determining the system’s success or failure. Insufficient analysis of system requirements may lead to impaired function and system design capability, and excessive analysis may lead to waste. Therefore, considering the feasibility of an expressway intelligent transportation system based on the Internet of things, the discussion is carried out from user demand and technology.
User demand analysis
The system will involve many subjects, such as relevant administrative departments, expressway operators, equipment suppliers, and related derivative service providers, so it is necessary to conduct a more comprehensive analysis when considering the needs of users. First of all, the administrative department demands to tell the traffic management department that it needs to know the real-time traffic flow and understand the traffic flow of each traffic entrance and exit and the overall traffic situation of critical areas through visualization technology. Also, traffic flow modeling and grooming model are needed to analyze and predict the data flow to complete the natural dredging work24. When there is abnormal traffic flow, it is also necessary to build an abnormal traffic accident model to monitor the abnormal flow. When some traffic sections have bottlenecks due to the influence of climate, facilities, and other factors, or when there are violations and accidents, they can be recorded, reported, and stored.
It is necessary for expressway operators to effectively manage the relevant facilities and equipment on the expressway to ensure the timely acquisition of traffic flow, the regular operation of traffic facilities and equipment, the push of traffic information, and the collection of various information of toll station information.
The demand for expressway service demanders represents the future development trend of intelligent transportation. The automated driving technology, expressway traffic flow, and road environment information acquisition provide the basis for coordinating travel plans, driving routes, and route arrangements. Simultaneously, the weather conditions, the distribution of service areas, and the traffic information in front of the owners need to be informed the first time so that all parties can make a timely response.
Expressway derivative service providers, namely navigation, ETC, financial services, and vehicle rental enterprises, provide more convenient expressway services for car owners.
Technical requirement analysis
The realization of intelligent transportation of expressway needs to consider the characteristics of openness, compatibility, dynamic, global, and intelligent to realize the intelligent system. The first is openness, which needs to open the national information system to meet customers’ needs. Compatibility mainly refers to that in the intelligent transportation system, data acquisition, processing, mining, interaction, and other technologies need a lot of equipment support, so the system needs absolute compatibility. Dynamic is that the vehicles in high-speed traffic are dynamic, so nodes and systems must collect relevant information and signals in real-time to further identify the traffic operation status25,26.
Global mainly considers the economic problem of deploying sensor nodes on the expressway. Different application fields and application scenarios are completed on the information processing platform to ensure the equipment operation, information recording, storage, global system, and local system analysis.
Intelligence mainly refers to collaborative processing and pattern recognition. Many heterogeneous nodes are deployed to complete the information collection so that the intelligent transportation system can automatically judge, control, and manage the traffic information to realize the overall intelligent development.
