Grasshopper　optimization　algorithm　is　one　of　the　latest　swarm　intelligence　algorithms,　however　it　is　easy　to　fall　into　the　local　optimum　and　has　the　defect　of　poor　search　ability.　Therefore,　Gaussian　and　chaos　theory　are　combined　to　improve　the　original　grasshopper　optimization　algorithm.　Firstly,　the　Sinusoidal　chaotic　map　is　used　to　initialize　the　solution　space　of　the　population;　then,　Gaussian　and　Logistic　chaotic　map　are　used　to　improve　the　position　update　formula　of　grasshopper　optimization　algorithm;　finally,　eight　benchmark　numerical　test　functions　are　used　to　verify　and　test　the　search　performance　of　this　algorithm　with　ALO,　PSO,　GOA.　It　can　be　seen　from　the　experimental　data　that　the　improved　grasshopper　optimization　algorithm　has　better　convergence　accuracy　and　stability　than　the　other　three　swarm　intelligent　algorithms.　Besides,　GC-GOA　has　good　performance　in　the　design　of　three-bar　truss　problem.　©　2019　IEEE.