Research on virtual measurement of key parameters in reactors can achieve diversified measurements, which is beneficial for improving the reliability and accuracy of measurements. Due to the weak correlation between reactor parameters, meeting the sparsity requirements, and the strong interpretability of sparse reconstruction methods, research on virtual measurement of key reactor parameters based on sparse reconstruction is conducted. Taking the steam generator water level as an example, data is generated based on simulation software. Through data correlation analysis, measurement model construction and solving, the sparse vector of the measurement parameter model is obtained, the virtual measurement of the steam generator water level is realized. The accuracy of the sparse reconstruction method is proved to be high through validation with simulation data and real data, verifying the feasibility of the method.