Real-time temperature field reconstruction using a few measurement points and RPIM-AGQ6 interpolation

Abstract

In thermal management systems, temperature field monitoring is vital for the structural safety of electronic equipment. However, there are practical challenges when it comes to installing temperature sensors on some monitoring points directly. To solve this problem, a novel temperature field reconstruction method is proposed. This strategy offers the advantage of using only a limited number of discrete values to predict temperature fields across an entire plane. Initially, the approach constructs a composite interpolation function, named RPIM-AGQ6, by coupling the mesh-free radial point interpolation method (RPIM) with the quadrilateral area coordinate method (AGQ6) The temperature field is discretized and represented as the product of the nodal temperatures and the RPIM-AGQ6. Next, ill-posed equations are established and solved using Bayesian Multivariate Linear Regression (BMLR) to obtain the temperature in the key points. Subsequently, a novel optimization strategy is proposed to achieve the cooperative optimization of the number of sensors and reconstruction accuracy. Finally, the simulated calculation and experimental tests are used to demonstrate the effectiveness and accuracy of the proposed approach. The results show that the proposed methodology can accurately reconstruct the plane temperature field of plate structure with one and two hot sources and the relative percentage error is within 2.6 %. Therefore, it can be utilized as an effective tool to real-time monitor structural surface temperature.