基于BSO算法优化的悬架系统神经网络控制研究OA

Vehicle passive suspensions have limited vibration reduction performance.Existing inerter-based suspension control strategies have insufficient adaptive ability.To solve these problems,this paper integrates an inerter into a variable-stiffness variable-damping suspension structure.It combines a 5-degree-of-freedom(5-DOF)human-seat model to establish a human-seat-suspension coupled dynamic model.This paper proposes an adaptive fuzzy neural PID control strategy optimized by beetle swarm optimization(BSO).A five-layer adaptive fuzzy neural network(AFNN)realizes full-process mapping of fuzzy inference.A hybrid RLS-BP learning algorithm completes online adaptive update of the network parameters.This paper takes the integral of time-weighted absolute error(ITAE)as the fitness function.It uses the BSO algorithm to achieve global optimization of the controller's core parameters.Simulation results verify the effectiveness of the proposed strategy.Under the Class-C random road condition,the RMS values of vertical acceleration of human head and pelvis are reduced by 55.76%and 27.38%respectively compared with the passive suspension,and the resonance peak in the 4～8 Hz frequency band sensitive to human body is effectively suppressed.Under the bump impact condition,the peak value of human head acceleration is controlled within 3 m/s2,and the vibration attenuation time is shortened to less than 2 s.It greatly improves the anti-vibration performance of the suspension under impact conditions,fully taps the vibration reduction potential of the inerter-based suspension,and significantly improves vehicle ride comfort.