In view of the problems existing in the measuring process of thermal conductivity of vacuum insulated panel(VIP), such as long test time and high cost, an embedded heat flow meter method for the thermal conductivity measurement of VIP was proposed. First, the feasibility of the measuring principle was verified by ANSYS, and the measurement systems for the thermal conductivity of VIP were established to obtain the output signal frequency change of VIP with different internal pressures. Then, the quality and service life of VIP was evaluated based on the relationship between output signal frequency change of oscillating circuit caused by temperature change and thermal conductivity. Finally, rigid regression was utilized to improve the extreme learning machine model (RRELM) and the generalization ability of extreme learning machine (ELM). Experiments show that the embedded heat flow meter method could realize the rapid thermal conductivity measurement of VIP. Compared with the traditional ELM, the proposed RRELM model could effectively amend the relationship between the thermal conductivity and output signal frequency change with higher precision.