Lateral　control　is　the　key　technology　of　autonomous　driving.　The　traditional　model-based　lateral　control　algorithm　relies　on　the　mathematical　model　of　the　controlled　vehicle,　and　the　inaccuracy　of　the　model　results　in　poor　algorithm　performance.　In　order　to　further　improve　the　tracking　accuracy　of　the　target　path,　a　fractional-order　PID　control　algorithm　based　on　data-driven　control　is　proposed　for　control,　and　the　particle　swarm　optimization　(PSO)　algorithm　is　used　to　optimize　its　parameters.　In　order　to　prove　the　performance　of　the　algorithm,　based　on　the　MATLAB/CarSim　joint　simulation　platform,　it　was　compared　with　the　PID　control　and　linear　quadratic　regulator　(LQR)　under　the　same　working　conditions.　The　experimental　results　show　that　the　vehicle　based　on　the　FOPID　control　can　effectively　overcome　the　increase　in　tracking　error　caused　by　the　change　of　path　curvature,　while　the　vehicle　can　maintain　good　comfort,　stability　and　safety.　©　2021　IEEE.