Study for process reliability modeling and controlling method of space product

The inner collection between process reliability and stable process of space product is analyzed, and the model of controlling-detecting process parameters is built by multiple linear regression technology. Then the extremum and mean value of multivariable controlling parameters' expected value, which meet the requirement of process detecting parameters' stability, are used to predict each mean response and new response, and also prediction ellipsoid and joint confidence intervals of detecting parameters are obtained. Furthermore, three optimization problems based on genetic algorithm, which focused on controlling parameters' fluctuation interval, are studied. Finally, taking the coating process of space product ferrite phase shifter as an example, the method is analyzed and verified. Therefore, the significant guarantee is provided for improving space products' quality.