Design of urban intelligent traffic control system based on multi-agent

Based on the current traffic situation and problems faced by China, this paper explains the important role of traffic control in economic development and urbanization, and puts forward the important purpose and practical significance of developing intelligent transportation systems. Intelligent transportation systems are a wide-ranging research field that integrates various high-tech technologies. The following focuses on the classification of intelligent transportation systems and the design of urban traffic control systems based on multi-agents.

1 Current situation of domestic urban transportation and the problems it faces

The current situation of urban transportation in my country is: the speed of urban transportation infrastructure construction cannot keep up with the rapidly growing transportation demand; conventional public transportation is shrinking; taxis and private cars are increasing rapidly; rail transportation is just starting; and the level of traffic management technology is low. The above problems will lead to increasingly serious traffic congestion, traffic accidents, environmental pollution, and energy problems.

The problems we face are: a wide variety of vehicle types and serious mixed traffic; an alarming number of non-motor vehicles such as bicycles; incompatibility between urban layout and traffic; a single mode of travel with no choice; difficulty in walking and frequent accidents.

2 Purpose and significance of urban traffic control

With the development of social economy, the acceleration of urbanization and the rapid increase of motor vehicles, urban traffic problems are becoming increasingly serious. Urban traffic congestion not only causes frequent traffic accidents and increased vehicle delays, but also further leads to increased energy waste and environmental pollution. The adverse social consequences caused by this are difficult to estimate. At present, urban traffic problems have become a bottleneck for global economic development and one of the global "urban diseases". The purpose of traffic control is to apply advanced technical means and adopt appropriate operating methods under the constraints of certain administrative regulations to ensure that public and private transportation modes have the best traffic conditions. Specifically, the purpose of traffic control is manifested in the following aspects: For more information, please log in to the Electronic Fans Network ( http://www.elecfans.com )

(1) Reduce traffic accidents and increase traffic safety. By implementing traffic control, conflicting traffic flows can be separated in time and space, thereby reducing traffic accidents and increasing traffic safety.

(2) Alleviate traffic congestion and improve traffic efficiency. Reasonable traffic control can effectively guide and dispatch traffic flow, keep urban traffic flow in a stable operating state, thereby avoiding or alleviating traffic congestion, shortening traffic delays for vehicles on the road, and improving the overall efficiency of transportation.

(3) Reduce environmental pollution and energy consumption. Implementing good traffic control can reduce the number of stops of vehicles on the road, keep vehicles running in a better state, and greatly reduce exhaust emissions and energy consumption.

3 Classification of urban traffic control systems

Urban road traffic control systems can be classified from different perspectives. Here, urban road traffic control systems are briefly classified from the perspectives of spatial relationship and control method.

3.1 Classification by spatial relationship

From the perspective of spatial relations, the urban traffic system can be divided into three forms: single intersection control (point control), coordinated control of traffic arteries (line control) and network control of regional intersections (surface control).

(1) Point control of a single intersection

Point control at a single intersection is the most basic control method. The control parameters of point control at an isolated intersection are the signal cycle and the green-to-signal ratio. The control objectives are generally vehicle delay and the traffic capacity of the intersection. Ideally, it is hoped that the total delay time will be minimized and the traffic capacity of the intersection will be maximized. Since point control equipment is simple, investment-saving, and maintenance-friendly, it is still a widely used signal control method. Technically speaking, it is divided into two forms: offline point control and online point control. The former uses timed signal timing technology, which is still the timing basis for other control methods; the latter is traffic response control or vehicle sensing control, which reasonably allocates green light time to each phase according to the actual distribution of each person's traffic flow at the intersection, thereby meeting traffic needs.

(2) Coordinated control of trunk line traffic

Traffic arteries in urban road networks bear a heavy traffic load. Ensuring smooth traffic on arteries often plays a vital role in improving the traffic conditions of a region or even a city. In urban traffic networks, intersections are sometimes very close to each other, and the distance between two adjacent intersections is usually not enough to allow a small group of vehicles to be completely evacuated within a limited time. When single-point signal control is set at each intersection, vehicles often encounter red lights, stop and start from time to time, and traffic is not smooth, causing serious environmental pollution. In order to reduce the number of vehicle stops at each intersection, especially when the traffic on the arterial is relatively smooth, the control scheme between adjacent intersections should adopt a mutually coordinated control strategy. The initial method of coordinating signal timing is based on the concept of green waves. Adjacent intersections execute the same signal control cycle, and the green light opening time of the main line phase is staggered for a certain time. The secondary arterials of the intersection obey the traffic of the main line to a certain extent.

When a convoy is traveling on a trunk line with many intersections, coordinated control ensures that the vehicles always arrive at the beginning of the green light when passing through the trunk intersection, so they can pass through the intersection without stopping, forming a green wave belt of traffic flow. Green wave control can effectively improve vehicle speed and road capacity, ensure smooth roads, and reduce vehicle delay time and energy consumption during driving. The control parameters of trunk line traffic coordinated control are cycle length, green signal ratio and phase difference. The control objectives are generally the average delay and number of stops of vehicles. The design flow chart of the trunk line signal coordinated control method is shown in Figure 1.

The object of regional traffic signal control is the traffic signals of all intersections in a city or a certain area. With the development of computer technology, optimization methods, automatic control and vehicle detection technology, people are studying how to combine and coordinate the traffic signals of all intersections in a city area or a local community, so as to minimize the total loss of vehicles in the area when passing through certain intersections. Under this control mode, the traffic signal transmits the traffic volume data to the host computer in real time through the communication network. The host computer continuously adjusts the timing plan being executed according to the real-time changes in the traffic volume of the road network at a certain time step. The host computer controls multiple intersections in a city area at the same time, realizes the unified coordination management between the intersections in the area, and improves the operation efficiency of the road network. Through this control method, it is easy to realize the unified scheduling and optimization management of the traffic road network. The design of regional signal coordination control timing optimization is shown in Figure 2.

3.2 Classification by control method

According to the control method, urban road traffic control can be divided into several types: timing control, sensor control, adaptive control and intelligent control.

(1) Timing control

The timing control method is based on historical traffic flow data, finds out the different traffic flow change patterns of each day/week and time period, and uses artificial methods or computer simulation to prepare the timing schemes for different days/weeks and different time periods in advance. It belongs to open-loop control and it is not easy to adjust the control scheme in real time according to the traffic conditions. Since the timing control has low requirements for traffic signals and does not require real-time traffic volume detection, it is still a widely used control strategy in the current urban road traffic system.

(2) Induction control

The principle of induction control is to adjust the length and time sequence of the corresponding green light time according to the traffic flow data measured by the vehicle detector to adapt to the random changes in traffic flow. This method has greater flexibility than timing control.

(3) Adaptive control

Adaptive control is based on the detected road traffic information and the future traffic demand predicted by the prediction model, and selects the corresponding optimization plan from the system signal timing plan library, or generates the corresponding optimization control plan in real time to achieve automatic traffic control.

(4) Intelligent control

Strictly speaking, intelligent control is not only the control of traffic signals, but also the control of the entire transportation system, that is, the intelligent transportation system. The intelligent transportation system is the highest level of traffic control. It integrates a series of high-tech technologies such as advanced information technology, data communication technology, detection and sensing technology, automatic control theory, operations research, artificial intelligence, computers and their networks into various subsystems of transportation, thereby establishing a real-time, accurate and efficient comprehensive transportation management system that plays a role in a wide range and in all directions. The intelligent transportation system organically integrates various aspects of the transportation system such as people, vehicles, roads and environment, so as to integrate and intelligentize the operation functions of vehicles and roads. The intelligent transportation system is the only way to solve traffic problems. Safe, efficient, environmentally friendly, low-consumption, fast and comfortable green transportation is the development direction of intelligent transportation.

4 Design of Urban Intelligent Traffic Control System

Intelligence is the ability to apply knowledge to a certain environment, or the ability to think abstractly to measure target criteria. Another definition is the ability to effectively obtain, process and use information for a specific purpose in a certain environment to successfully achieve the goal. Intelligent transportation system is a comprehensive system that uses the theory and methods of artificial intelligence to solve traffic problems. The achievements of artificial intelligence in recent years have provided a solid theoretical foundation for the research of intelligent transportation systems, and these achievements can be used to solve problems that traditional methods cannot solve. This is because: on the one hand, the transportation system is a system with complex structure, many influencing factors and strong randomness. It is very difficult to solve traffic problems using mathematical methods. The established models are often too complex and difficult to solve. At the same time, it is also difficult to summarize the diversity of traffic flow systems with one or several models. On the other hand, the transportation system is a dynamic time-varying system, and the real-time requirements for traffic management and control are very high. Therefore, based on the actual situation, the traffic management and control methods based on mathematical descriptions are difficult to meet the requirements of online real-time control and have poor operability. The artificial intelligence method, drawing on the human problem-solving method, solves complex problems through knowledge expression, reasoning and learning, and transforms the previous description of the transportation system with pure mathematics into a description with knowledge or a combination of knowledge and mathematical models. By gradually adapting to the environment through learning capabilities, we can continuously improve management and control effects.

Multi-agent system is a cutting-edge subject in artificial intelligence today and an important branch of distributed artificial intelligence research. Its goal is to build large complex systems into small, mutually communicating and coordinated, easy-to-manage subsystems, and solve complex system control problems through the autonomy and mutual coordination capabilities of subsystems. Due to the complexity of the scale of road traffic and the real-time nature of the dynamic characteristics of traffic flow in urban area transportation networks, the application of multi-agent systems to urban transportation network control has become a research topic of great concern. Based on this, this paper designs the urban intelligent transportation control structure diagram, as shown in Figure 3.

In Figure 3, the left side is the urban traffic flow system based on multiple agents, and the right side is the signal control system. In the traffic model, the road section agent has the ability to update the traffic flow of a single road section in real time, and can provide traffic flow data for the connected intersections to optimize signal timing; it controls the traffic flow according to the traffic flow information of the superior area, and exchanges data with other sections through the intersection; in the corresponding signal control model, according to the information provided by the section agent, it performs signal timing and coordinates the dynamic balance of traffic flow between sections.

In addition to the same functions as the road section agent, such as obtaining traffic flow information and providing data support for signal control, and being independent and able to operate independently, the regional agent can also coordinate signals at traffic intersections according to the traffic flow operation in the area, so as to optimize the traffic objectives of the corresponding road network operation, independently determine the corresponding control strategy, and send the signal timing results to each signal in the area in a timely manner, and at the same time transmit the traffic demand and control effect to the upper decision-making layer, that is, the traffic management center. The regional agent plays a connecting role between the road section agent and the central management agent, transmitting traffic flow information with a single road and the entire traffic network; and exchanging information with other regional agents at the same level to provide services for traffic control. The central management agent is composed of the urban traffic control decision-making system. Its function is to make an estimate of the operation status of the entire traffic network based on factors such as road network structure, traffic detection and traffic congestion. Its purpose is to seek the optimal performance indicators of the entire system and use its superior reasoning and planning capabilities to make the system operate in the optimal state. In the signal control system, the intersection-level signal control gives the intersection signal timing through control decisions based on the traffic flow information of the intelligent bodies of the connected road sections. The intersection-level signal control transmits the road flow information and signal timing to the upper regional control, and the upper regional control feeds back the control command. If a certain intersection in a certain area is congested in a certain direction, the regional control will adjust the signal timing of other intersections on the same road section by giving a control command to relieve the congestion as soon as possible and reduce the overall delay time of the area. What is transmitted between regional controls is the traffic flow information in the area. If a congested intersection appears in a certain area, the signal timing in the area and adjacent areas is adjusted to guide the traffic flow to disperse to relieve the congestion, and the vehicle diversion is guided by roadside information signs or traffic radio signals. What is transmitted between regional control and the traffic control center is the regional traffic flow information. Adjust the dynamic balance of road network traffic flow and provide information to the central traffic control center to achieve the centralized and decentralized control method of urban traffic. For more information, please visit Electronic Fans Network ( http://www.elecfans.com )

5 Conclusion

Intelligent transportation system is a wide-ranging research field that integrates various high-tech technologies. Due to space limitations, this article only discusses some measures to improve urban road safety from the perspective of urban intelligent traffic control, and does not conduct research from the perspective of vehicles and drivers and passengers. Improving the overall level of intelligent transportation systems requires the coordinated development of various industries to jointly promote the improvement of urban transportation levels.