Traffic flow prediction method and system based on social value orientation, terminal and medium

Abstract

The invention belongs to the technical field of automatic driving, and discloses a traffic flow prediction method and system based on social value orientation, a terminal and a medium. The dynamic interaction among all vehicle individuals in the traffic flow in the scene is captured by using the game theory, the selfish and the altricity of the driving behavior of the driving vehicle are quantified by using the social value orientation, and the social value orientation is fused into the calculation of traffic flow prediction to predict the driving behavior of the driving vehicle. According to the method, dynamic interaction among all vehicle individuals in a scene is captured by using a game theory, a social value orientation parameter is introduced to quantify selfish and altricity of driving behaviors of human drivers, and the parameter is fused into calculation of traffic flow prediction; therefore, the driving behavior can be predicted stably and effectively. According to the method, the selfish or alliance degree of the vehicle driver is quantified, and how the vehicle driver interacts and cooperates with others can be better predicted.

Description

technical field [0001] The invention belongs to the technical field of automatic driving, and in particular relates to a traffic flow prediction method, system, terminal and medium based on social value orientation. Background technique [0002] At present, urban traffic flow has the characteristics of non-linearity, self-organization, space-time, stochastic time-varying, and cycle similarity, which make short-term traffic flow forecasting technology complex and difficult. In 2015 alone, there were more than 450,000 lane change / merge accidents and 1.4 million right / left turn accidents in the United States. In the analysis of traffic accidents in California, 57% of the accidents were caused by rear-end collisions. The main reason is that the driving behavior of the vehicle in front was beyond the expectation of the driver of the rear vehicle, that is, it is difficult to predict the traffic flow. [0003] The main reason for the above problems is that human drivers will adopt...

AI Technical Summary

Problems solved by technology

[0012] (1) Most of the existing related patents are based on some basic data processing based on the existing sensor data collection, or adopt the method based on rule logic, or adopt the method based on learning. The mapping of future behavior, but the basic object of traffic flow operation is human drivers, who have random personality characteristics, so the prediction method in the form of a fixed rule framework cannot achieve strong robustness in the actual application process, making predictive driving The accuracy of behavioral data is low

[0013] (2) The driving style of human drivers is affected by many factors such as their personal personality and mood, and it is difficult to accurately capture and predict. Therefore, there will be unpredictable driving behaviors in the traffic flow, which will cause traffic safety hazards

[0014] (3) The interaction between vehicles in the traffic flow is the interaction between people. The possible results in this game scene are random, and with the increase of the number of participants, the results of the process can be observed. Sexuality is further reduced

[0015] (4) At present, in the field of autonomous driving, the research process of planning and decision-making has a long periodicity. From information acquisition and processing to the actual action of the vehicle, the entire planning and decision-making process covers a long time series. During this period, the vehicle The state of the surrounding vehicles in the environment will have strong random changes, and the current trajectory prediction scheme based on the vehicle dynamics model is only effective immediately, which in turn affects the effectiveness of the entire planning decision-making process

Human drivers are often affected by the environment and will have strong random driving behavior changes, so the effective time of the basic model prediction results based on vehicle dynamics is limited to 1 to 2s, which is the time required for human drivers to generate new driving behaviors. conversion time

In the face of more complex scenarios, the long-term traffic flow forecast required contradicts the immediate validity of existing model-based forecasts

[0018] (2) In complex scenarios, as the number of vehicle nodes gradually increases, the complexity of the environment increases

The relative positional relationship and social relationship between nodes in the scene have strong random variability, and it is difficult to cover all possible scene features using common classification-based mathematical models

It is also difficult to effectively describe the social relationship between vehicle nodes using mathematical models

[0019] (3) The driving style of the human driver plays a decisive role in the formation of the future behavior trajectory of the driving vehicle, but it is difficult to quantify the driving style of the human driver in the actual prediction process, so this item Factors cannot be directly and efficiently incorporated into existing mathematical models

[0022] (1) In actual application scenarios, traffic flow prediction requires that the obtained data be continuous and effective, so instantaneous effective prediction methods are not applicable

Images