Abstract: To rationally allocate reactive power resources between grid-following(GFL) and grid-forming(GFM) wind turbines in hybrid wind farms and mitigate reactive power imbalance caused by factors such as spatial distribution and control differences, a reactive power and voltage optimization control strategy considering reactive power balancing is proposed. First, control models are established for both types of units: fixed reactive power control for GFL turbines and linearized droop control for GFM turbines. In the day-ahead stage, a stochastic wind power model based on scenario generation and reduction is developed to handle uncertainty, yielding an optimal plan that balances economy and security. In the intraday stage, using day-ahead schedules as input, a fluctuation-level-based strategy is proposed. Under mild wind variations, optimization minimizes voltage deviation and network losses. During severe fluctuations, the strategy switches to a reactive power balancing mode, where balance zones are partitioned to adjust GFL reactive power references, thereby coordinating reactive power distribution and enhancing voltage stability. Case studies on a 34-turbine hybrid wind farm validate the method's effectiveness.