CSWOA for Enhanced Temperature Compensation in Uninterruptible Power Meter Replacements

Owing to the difficulties in directly measuring and observing contact temperatures within uninterruptible power meter replacement devices, this research puts forward a transient temperature compensation approach relying on the Compressive Sensing–Whale Optimization Algorithm (CSWOA). Initially, a second-order transient heat conduction model is constructed utilizing the Dual Reciprocity Boundary Element Method (DRM-BEM). Subsequently, an enhanced whale optimization algorithm is employed to inversely reconstruct unknown boundary conditions, thus facilitating indirect temperature compensation. Taking advantage of the a priori sparsity of thermal boundaries, the method projects boundary parameters into a sparse coefficient space. This action reduces the optimization dimensionality and speeds up the convergence process. Meanwhile, an adaptive parameter adjustment strategy based on individual sparsity dynamically balances exploration and exploitation, so as to improve stability and accuracy. Electro-thermal coupled simulations demonstrate that CSWOA significantly outperforms Particle Swarm Optimization (PSO) in terms of performance.