Accurate estimation of model parameters plays a very important role in modeling solar photovoltaic (PV) systems. In the past decade, meta-heuristic algorithms (MHAs) have been used as promising methods for solving this problem. However, due to the non-linearity and multi-modality existed in the problem, many HMAs may present unsatisfactory performance due to their premature or slow convergence. Therefore, how to develop algorithms efficiently balancing the exploration and exploitation, and identify the PV model parameters accurately and reliably is still a big challenge. In this paper, to improve parameter estimation of solar photovoltaic models, we propose a hybrid meta-heuristic algorithm, called biogeography-based heterogeneous cuckoo search (BHCS) algorithm. Specifically, BHCS hybridizes cuckoo search (CS) and biogeography-based optimization (BBO) by employing two search strategies, namely heterogeneous cuckoo search and biogeography-based discovery. The cooperation of the two strategies helps BHCS achieve an efficient balance between exploration and exploitation. Furthermore, the proposed algorithm is applied to solve four parameters estimation problems of different photovoltaic models, including single diode model, double diode model and two PV panel modules. Experimental results demonstrate that BHCS has very competitive performance in terms of accuracy and reliability compared with CS, BBO and several other meta-heuristic algorithms.