跟 网 -构 网 混 联 系 统 虚 拟 惯 量 与 阻 尼 参 数 协 同 优 化 方 法

The high proportion of renewable energy connected through power electronic equipment significantly weakens the oscillation damping capability and overall inertia level of the power system. Although grid-forming converters can actively provide system inertia and damping support, their security and stability characteristics are influenced by nonlinear coupling of multi-machine parameters, increasing the difficulty in configuring device inertia and damping. To address this, this paper proposes a multi-converter virtual inertia and damping collaborative optimization method integrating the matrix perturbation theory and alternating optimization mechanism. First, a multi-constraint nonlinear and nonsmooth optimization model is constructed, encompassing small-signal synchronous stability, frequency security, and device physical limits. Subsequently, an alternating update mechanism is employed to decompose it into independent optimization problems for virtual inertia and damping. Then, based on the matrix perturbation theory, the sensitivity expressions of eigenvalues to parameters are derived, transforming the nonsmooth independent optimization problem into a iteratively solvable sequence of linear subproblems. Additionally, a trust-region adaptive step size strategy based on eigenvalue second-order sensitivity is proposed to enhance convergence efficiency and stability. Case studies demonstrate that the proposed method significantly improves small-signal stability under various robustness challenges, including changes in grid-forming converter connection locations, quantities, and power flow fluctuations, with both optimization effectiveness and efficiency surpassing traditional methods.