For the parameter coordination and optimization of PSS in a multi-machine power system, this paper proposes a new method based on the Phillips-Heffron model to search the optimal install locations of PSS, which employs the sensitivity of the system eigenvalue to the PSS gain as the indicator of its install location, i.e. the larger this sensitivity is, the more the system damping can be improved effectively when the PSS is installed at the corresponding generator. Moreover, based on the block matrix theory and the Newton-Raphson iterative algorithm, the proposed method optimizes the parameters of multiple PSS, where the unknown parameters of each PSS is incorporated into the system state equations and solved by the Newton-Raphson iterative algorithm. Such a method takes the interaction among multiple PSS into account, so that the parameters of multiple PSS can be optimally coordinated. The proposed method is applied to the Sichuan Power Grid under different operating modes, and engineering practices shows that the proposed method is effective and robust.