This　work　reports　a　path-tracking　control　method　based　on　tracking　error　constraints　and　integral　guidance　mechanisms　with　anti-interference　for　a　robotic　fish.　A　virtual　control　input　is　presented,　and　the　adaptive　integral　light-of-sight　rule　is　developed　to　eliminate　the　deviation　of　the　sideslip　position　and　improve　the　anti-interference　ability　of　the　robot.　The　raw　and　large　adaptive　controllers　of　a　robotic　fish　are　constructed　to　fit　the　model　uncertainty　and　flow　field　disturbance　brought　by　neural　network　functions.　The　approximation　values　compensate　for　the　system＇s　control　input　to　improve　the　environmental　adaptability　of　the　body.　Moreover,　this　work　employs　event-triggering　mechanisms　to　reduce　the　trigger　frequency　of　the　actuator　and　improve　tracking　speed.　Barrier　Lyapunov　theory　confirms　that　the　proposed　strategy　has　uniform　ultimate　boundedness.　Simulation　and　experimental　results　indicate　that　the　scheme　improves　the　error　convergence　rate　and　steady-state　characteristics　of　the　robotic　fish.