The　effects　of　excitation　system　on　the　performance　of　magnetic　flux　leakage　(MFL)-type　non-destructive　testing　were　investigated.　The　investigation　employed　parameterized　finite　element　analysis　of　the　MFL　testing　of　ferromagnetic　samples　to　reveal　the　effects　of　the　structural　design　and　parameters　of　the　excitation　system,　including　the　geometry　and　dimensions　of　the　yoke,　lengths　and　positions　of　coils　and　thickness　of　the　shielding　layer.　Moreover,　a　parameter　optimization　method,　combining　parameterized　finite　element　analysis　with　a　genetic　algorithm,　was　proposed　to　optimize　the　dimensions　of　the　yoke　and　thus　obtain　maximum　magnetic　flux　leakage.　Simulation　and　optimization　results　were　used　to　propose　an　MFL　sensor　with　a　preferred　structural　design　and　preferred　parameters　of　the　excitation　system.　Comparative　experiments　were　conducted　to　validate　the　numerical　simulation　results.　The　experimental　results　agreed　well　with　the　results　of　numerical　simulation,　and　the　superior　performance　of　the　MFL　sensor　with　the　preferred　excitation　system　was　validated.　(C)　2017　Elsevier　B.V.　All　rights　reserved.