Journal of Northeastern University(Natural Science) ›› 2023, Vol. 44 ›› Issue (12): 1726-1733.DOI: 10.12068/j.issn.1005-3026.2023.12.008

• Mechanical Engineering • Previous Articles     Next Articles

Genetic Algorithm for Solving Nonlinear Equilibrium Points of Automobile Steering and Its Improvement

LI Jie1, JIA Chang-wang1, QIAO Bin2, LIU Jia-yong2   

  1. -
  • Published:2024-01-30
  • Contact: -
  • About author:-
  • Supported by:
    -

Abstract: Aiming at the problem of solving nonlinear equilibrium points of automobile steering, the solving effect of the genetic algorithm was studied and an improved method was proposed. A two-degree-of-freedom model of automobile steering was established, and the reason why the nonlinear equilibrium points of automobile steering can only be solved by numerical iteration was explained, a fitness function suitable for the intelligent optimization method was constructed, and a method to determine the feasible solution range was proposed. Under the driving conditions of a vehicle speed of 70km/h, road adhesion coefficient of 0.5 and a front wheel angle of 0~0.2rad, three equilibrium points were solved by using genetic algorithm. By comparing the fitness value surfaces of large and small steering angles, the reason why the genetic algorithm can not meet the solving accuracy of small steering angle equilibrium point was explained, and the solution flow of the fusing genetic algorithm and BFGS quasi-Newton method was presented. The results show that the proposed fusion solution process can solve the small steering angle equilibrium points that meet the solving accuracy requirements and have higher solving efficiency, which makes up for the shortage of the genetic algorithm solving alone.

Key words: steering nonlinearity; equilibrium point; genetic algorithm; BFGS quasi-Newton method; fusion solution; two-degree-of-freedom model of automotive steering

CLC Number: