In surrogate-based optimisation methods, computationally burden yet precise fine model is utilised alongside less accurate however quick coarse model to decrease the general computational exertion. Surrogate-based optimisation methods are firstly applied to reconfigurable antenna design problems in this work. Space mapping (SM) with inverse difference method enables productive procedure to decrease computational exertion while enhancing the convergence. Inverse difference mapping is constituted by difference knowledge that is obtained by input and output of the problem space. In addition, inverse coarse model that is obtained by feed forward multi-layer perceptron can generate necessary knowledge to form inverse mapping. This mapping is used to eliminate the direct optimisation-based parameter extraction process. The efficiency of SM with inverse difference technique will be demonstrated by reconfigurable antenna design examples in terms of their convergence and accuracy. Moreover, this technique will be compared to aggressive SM with regard to the convergence efficiency.