Abstract: In order to optimize the sensor system for a large-scale group-pile foundation and properly select sensor types, a multi-sensor optimization model was established, and the genetic algorithm was used to solve this optimal problem. In this model, the cost of the sensor system is set as the objective function, and the reliability of the voting system is taken as the constraint conditions. The reduction coefficient of environmental factor was put forward to consider the influence of sensor installation environment on the system reliability, its determination way was given, and it was used to revise the sensor reliability under different environments. In addition, the sensor system of Sutong bridge foundation was taken as an example to prove the optimization model. The research result shows that with the help of the established model, the cost can be saved nearly 10% under the precondition of the system reliability.