Solar　energy　applications　and　research　are　becoming　increasingly　popular,　and　photovoltaics　(PVs)　are　among　the　most　significant　solar　energy　applications.　To　simulate　and　optimise　PV　system　performance,　the　optimal　parameters　of　the　solar　cell　models　should　be　estimated　exactly.　In　this　study,　improved　simplified　swarm　optimisation　(iSSO),　a　recently　introduced　soft　computing　method　based　on　simplified　swarm　optimisation,　is　proposed　to　minimise　the　least　square　error　between　the　extracted　and　the　measured　data　for　the　solar　cell　models　parameter　estimation　of　the　single-　and　double-diode　model　problems.　Based　on　the　new　all-variable　difference　update　mechanism　and　survival　of　the　fittest　policy,　the　proposed　algorithm　is　able　to　find　an　improved　approximation　for　estimating　the　parameters　of　single-　and　double-diode　solar　cell　models.　As　evidence　of　the　utility　of　the　proposed　iSSO,　the　authors　present　extensive　computational　results　for　two　benchmark　problems.　The　comparison　of　the　computational　results　supports　the　proposed　iSSO　algorithm　outperforms　the　previously　developed　algorithms　for　all　of　the　experiments　in　the　literature.