Multi-intelligent vehicle formation method based on distributed finite-time state observer

Abstract

The invention relates to a multi-intelligent vehicle formation method based on a distributed finite-time state observer, comprising the steps of: S1, describe the kinematic characteristics of the intelligent vehicle according to the kinematics model of the intelligent vehicle; S2, construct a master-slave architecture of the multi-intelligent vehicle , and establish a communication network structure between multi-intelligent vehicles based on the topology structure of the directed spanning tree; S3. Design a distributed finite-time state observer that does not depend on the global coordinate information to obtain the observed value; S4. Input the observed value into multiple Intelligent vehicle distributed formation controller to realize automatic formation of multi-intelligent vehicles. Compared with the prior art, the present invention obtains the relative position between vehicles through the laser radar and the gyroscope, so that the error is controlled within the range of millimeters, ensuring that the observed value is closer to the actual value, and at the same time, a directed spanning tree is used to construct a communication network structure. The convergence of the observer is guaranteed, and the accuracy and safety of the intelligent vehicle formation can be effectively improved.

Description

technical field [0001] The invention relates to the field of coordinated control of multiple intelligent vehicles, in particular to a multi-intelligent vehicle formation method based on a distributed finite-time state observer. Background technique [0002] With the continuous deepening of research on intelligent driving vehicles, the limitations of the performance of a single intelligent vehicle (such as load capacity, work efficiency, detection field of view, etc.) Overall system performance, compared with a single agent, a multi-agent system has irreplaceable advantages: [0003] For complex tasks and environments, the multi-agent system can decompose complex tasks into multiple simple sub-tasks, and then use multiple agents to work simultaneously in different areas, thereby improving work efficiency; [0004] Multiple agents are distributed in an environment with a large area, each perceives its surrounding environment, and shares the information perceived by each agent...

AI Technical Summary

Problems solved by technology

[0011] In the existing distributed finite-time state observers, the slave often directly observes the global position information of the master, but since the global position information provided by the master itself is generally obtained by GPS, its accuracy can only reach the meter level. The positioning error of the upper-slave itself and the influence of environmental factors (such as tunnels and high-rise buildings) on the positioning accuracy will lead to poor observation accuracy of the existing distributed finite-time state observers. When applied to the distributed formation control of multi-intelligent vehicles, it may cause vehicle collisions and fail to achieve the expected formation goal

Images