How to exclude rows with unwanted values

[#1]

Hello,

My table (from 2 sources:   TAB_1    <- KEY ->   TAB_2   >>>  ideally 1...1) may contain more row from TAB_2. In most case   column named "Druh dokladu" has only one value (1...1), but some rows have more values "Druh dokladu", but I know if there is value "D3" and "AA" so row with "D3" is irrelevant and I need only "AA" row. If there would be only one row, value of "Druh dokladu" may have any value.

Any suggestion?

nazov (TAB_1)	Obdobie plán. úhrady (TAB_1)	tyzden_plan_uhrada (TAB_1)	pred_splatnost (TAB_1)	popis (TAB_1)	c_faktura (KEY)	Druh dokladu (TAB_2)	Plán platby (bez DPH) (TAB_1)	Plán platby (bez DPH) (TAB_2)	Header 10
MEL	2016.06	24	13.6.16	Vyčistenie objektu	F2016007	AA	15412	15412,16
MEL	2016.06	24	13.6.16	Vyčistenie objektu	F2016007	D3	15412	15412,16
MEL	Total						15412	30824,32
Total							15412	30824,32