Model-Based Particle Swarm Optimization Filtering Algorithm for Mecanum Wheel Car Parameter Identification With Measurement Noise

Abstract

The Mecanum wheel car (MWC) is increasingly becoming the mainstream automated guided vehicle (AGV) in factory automation, replacing traditional transport vehicles due to its flexibility and maneuverability. With its widespread applications, there is a corresponding high demand for system inspection and maintenance policies. However, the estimation of kernel parameters without the system disassembly is less investigated. To solve this problem, this article starts from a framework of nonholonomic constraints and uses the Lagrange equations to derive a complete dynamic model of the MWC. Next, a measurement equation using the signal filtering method (FM) is derived. However, the design of the filtering factors is the key issue of the tradeoff between estimation precision and noise suppression. To effectively solve this design problem, particle swarm optimization (PSO) is used to optimize the filtering factor. The proposed method not only avoids interference from noisy acceleration measurements of the MWC but also significantly improves parameter estimation accuracy. The feasibility of the proposed method was validated through both numerical simulations and experiments. The experimental results demonstrate that the parameter estimation method proposed in this article can accurately estimate the internal parameters of the system, enabling precise prediction of the MWC’s motion behavior.