An Autonomous Co-localization Method for Networked Multi-Agents Actively Changing Topology

Abstract

The invention provides an active variable topology autonomous cooperative positioning method for a networked multi-agent, and belongs to the technical field of signal processing. The method aims at anetworked multi-agent system composed of m standard agents and n common agents, and comprises the steps as follows: before positioning, each common agent acquires a position estimation initial value computing basis and a speed estimation initial value computing basis via m standard agent envelops; and each common agent separately initializes the respective estimated position value and estimated speed value and performs positioning in two stages, and each common agent actively converts a topological structure in a computational level every moment, and computes and coordinates self-positioning and speed estimation according to estimation information of positioning and speed of the neighbor agent selected by variable topology and sensing information about a distance between the adjacent agents, to complete autonomous cooperative positioning. The method provided by the invention has the characteristics of being distributed, self-organizing and extensible, can furthest prevent positioning estimation from being trapped into local optimum, ensures that a positioning result is accurate, and is extremely high in robustness.

Description

technical field [0001] The invention belongs to the technical field of signal processing, in particular to an autonomous collaborative positioning method for networked multi-agents to actively change topology. Background technique [0002] An agent usually refers to an individual with basic capabilities such as computing power, communication ability, response ability, and perception ability. Common agents include satellites, drones, robots, and smart sensors. Using multiple relatively simple agents to coordinate and network to form a networked multi-agent system can complete tasks that were originally difficult for a single agent that was complex and large. Common networked multi-agents such as short-distance satellite clusters, wireless sensor networks, and synthetic aperture cameras have great application potential in various fields of national economy and people's livelihood. [0003] The key basis for the coordinated application of networked multi-agents lies in the pos...

AI Technical Summary

Problems solved by technology

[0005] Chinese patent 200910000855.3 discloses a wireless positioning method for sensor networks. Each wireless sensor transmits the signal obtained by itself to the main server through a tree network. The main server centrally processes all the signals, and then calculates the location of each sensor. location, and then distributed back to each agent through a tree network, so as to realize the positioning of multiple agents. The calculation cost, communication cost and cost cost of the server are huge, which is severely restricted by the network topology, and once the main server fails, its positioning function will completely fail, so it is not suitable for large-scale multi-agent networks

However, on the one hand, this method is only suitable for relatively static scenes between multiple agents. On the other hand, this method is extremely sensitive to the initial value of data fusion, so this method can only be used in the high-precision positioning stage, and must rely on Provide fusion initial value for "relatively low precision" and "non-fusion" positioning methods

Images