Match 2 columns from different columns

First step you need to do is to deduplicate records in table APE 02 2022.

then you can merge both table based on name to get all combinations.

it depends of usecase. 

please see M-code for deduplication

let
    Source = Table.FromRows(Json.Document(Binary.Decompress(Binary.FromText("tZHBasMwEER/ZfE5B9mtS3NU7KmzibQbJCutCfn/3+jKlEJ6KDQ0CHTZtzPD7OXS9P1265xrNs3ICcNMu5JZkHNz3Txy6lvX1Xln8wgZtUQQEg909gEys6zY9rlz3ZNrjcLHKdk2seSSvAxYgV3v2qpTiUVNhyIvkMmTD/zFmFdl2ldjAvbRCw17nwIyvSWM1fWGq5kOPO1Rgqd3BH9i0UIIrKLpG7W32griQhp1oow0wTQtnXK+0ewrOBUcLRsdNPAZ8wwKZWDcY/4w1L2sSVWOGqxMnxj2jwiw490F/rv5X4pfwbjzJADNEKscetIw/gz5+4Gunw==", BinaryEncoding.Base64), Compression.Deflate)), let _t = ((type nullable text) meta [Serialized.Text = true]) in type table [#"Numero Agents " = _t, #"Noms Agents" = _t]),
    #"Changed Type" = Table.TransformColumnTypes(Source,{{"Numero Agents ", type text}, {"Noms Agents", type text}}),
    #"Removed Duplicates" = Table.Distinct(#"Changed Type", {"Numero Agents "})
in
    #"Removed Duplicates"

see M-code for merge

let
    Source = Table.NestedJoin(#"New Agents codes", {"EAAGTNAME"}, #"APE 02 2022", {"Noms Agents"}, "APE 02 2022", JoinKind.FullOuter),
    #"Expanded APE 02 2022" = Table.ExpandTableColumn(Source, "APE 02 2022", {"Numero Agents ", "Noms Agents"}, {"Numero Agents ", "Noms Agents"})
in
    #"Expanded APE 02 2022"

result looks like