In　this　article,　we　propose　a　dual-mode　event-triggered　predictive　control　method　for　nonlinear　systems　with　bounded　disturbances.　The　proposed　method　contains　two　triggering　mechanisms,　namely,　the　hybrid　threshold-based　event-triggered　model　predictive　control　(HETMPC)　mechanism　and　the　event-triggered　linear　quadratic　regulator　mechanism.　The　former　triggering　mechanism　is　designed　based　on　the　error　between　the　real　state　and　the　optimal　state　and　also　the　disturbance　information　acted　on　the　investigated　system.　Compared　with　the　traditional　fixed　triggering　threshold,　the　designed　HETMPC　has　a　fewer　triggering　numbers　and　reduces　the　computational　burden　of　online　real-time　optimization.　This　event-triggered　mechanism　will　be　adopted　before　the　states　go　into　the　terminal　invariant　set.　The　latter　event-triggered　mechanism　is　designed　based　on　the　derivation　of　the　system　state　and　it　will　be　adopted　after　the　states　enter　the　terminal　invariant　set.　The　feasibility　and　the　input-to-state　practical　stability　analysis　of　the　designed　strategy　is　presented.　Some　simulations,　including　the　application　to　a　mass-spring-damper　system,　are　provided　to　show　the　correctness　and　feasibility　of　the　designed　algorithms.