Abstract: According to design requirements,in order to solve the problem of operating concrete wet spraying machinery to substitute for manual tools under an abominable tunnel construction environment,certain manipulator machine had been designed.Three sections telescopic boom are key component of the manipulator machine and it had a conservative design with surplus strength and stiffness.Parameter sensitivity analysis are usually regarded as basis for completing lightweight design,including reducing structure weight and manufacture cost,making good use of materials’ mechanical properties,etc.Considering strength and stiffness as constraint conditions,structure weight as design objective,researched the effects of 6 variables on design indexes.First of all,utilizing software MATLAB established the BP neural network model with 6 design variables,3 constraint conditions and 1 design objective.After that,integrating the multidisciplinary optimization design software ISIGHT and software MATLAB achieved the Monte Carlo numerical simulation for parameter sensitivity analysis.By this analysis the differences parameters sensitivity are clarified,which will provide great reference for the design of similar structures.

Key words: three section telescopic boom, neural network model, descriptive sampling method, parameter sensitivity analysis