The invention provides a distributed robust
belief propagation method and equipment with
odometer pre-integration for multi-
robot cooperative positioning, and belongs to the technical field of multi-
robot cooperative positioning. The method comprises the following steps: firstly, providing a
speedometer pre-integration measurement model, and calculating a message from a
speedometer pre-integration factor node based on the
speedometer pre-integration measurement model; secondly, modeling the probability density of external measurement factor nodes into a student t
distribution method, deriving an improved measurement
noise covariance matrix by using variational
inference, and calculating messages from
absolute measurement and relative measurement factor nodes based on the improved measurement
noise covariance matrix; and finally, carrying out iterative updating on belief of all variable nodes in the sliding window by utilizing messages from a speedometer pre-integration factor node, an
absolute measurement factor node and a relative measurement factor node. According to the distributed robust
belief propagation method with
odometer pre-integration provided by the invention, the influence of measurement nonlinearity and measurement outliers can be remarkably reduced, the precision of
pose estimation is improved, and the distributed robust
belief propagation method has practical
engineering application significance.