Self-powered flexible devices with skin-like multiple sensing ability have attracted great attentions due to their broad applications in the Internet of Things (IoT). Various methods have been proposed to enhance mechano-optic or electric performance of the flexible devices; however, it remains challenging to realize the display and accurate recognition of motion trajectories for intelligent control. Here, we present a fully self-powered mechanoluminescent-triboelectric bimodal sensor based on micro-nanostructured mechanoluminescent elastomer, which can patterned-display the force trajectories. The deformable liquid metals used as stretchable electrode make the stress transfer stable through overall device to achieve outstanding mechanoluminescence (with a gray value of 107 under a stimulus force as low as 0.3 N and more than 2000 cycles reproducibility). Moreover, a microstructured surface is constructed which endows the resulted composite with significantly improved triboelectric performances (voltage increases from 8 to 24 V). Based on the excellent bimodal sensing performances and durability of the obtained composite, a highly reliable intelligent control system by machine learning has been developed for controlling trolley, providing an approach for advanced visual interaction devices and smart wearable electronics in the future IoT era.