<p>Wuhan Advanced Light Source (WALS), a fourth-generation synchrotron radiation light source operating at 1.5 GeV, is currently under design and will use a full-energy linear accelerator (LINAC) as the electron beam injector. The injection beamline adopts a three-stage scheme: First, the beam from the LINAC, which is 6&#xa0;m below the storage ring, is horizontally deflected below the storage ring; second, it gradually climbs from underground to the same altitude as the storage ring; and third, the beam is delivered horizontally into the injection straight section inside the storage ring. Twiss parameter matching between the LINAC and storage ring was also completed. During the construction of the beamline, magnet manufacturing errors, installation errors, and beam injection errors from the LINAC will cause beam deviations from the predetermined ideal orbits and even particle losses. Therefore, electron beam correction is required during beam commissioning. In contrast with the single-plane beam correction used in general transfer lines, the horizontal and vertical directions of the beam are coupled in the WALS injection transfer line, which greatly increases the complexity and difficulty of beam correction. Machine learning technology has been extensively developed in recent years, and the powerful invertible neural network algorithm is expected to solve the beam commissioning challenge of the beam injection transfer line at the WALS. Therefore, an invertible neural network (INN) model has been designed and trained to simulate the beam transport and beam correction of the WALS injection beamline. By optimizing the number and positions of beam profile monitors, the accuracy of both bidirectional prediction and beam correction can be greatly improved. This method has important practical significance for the commissioning and operation of similar complex beam transport systems.</p>