2 ways that I can see immediately:
If you added VLANs to monitoring during your initial discovery for NPM, you can go to 'Manage Nodes' and simply search for the VLAN # as an interface and you will get a result set of all instances you are currently monitoring.
You could also do this as a custom report using sql as so:
SELECT n.Caption, i.FullName FROM Nodes n JOIN Interfaces i ON n.NodeID = i.NodeID WHERE i.FullName LIKE '%VLAN%'
Just change the VLAN to whatever you're searching for (VLAN #)
The second way would use NCM. I would probably recommend this because you're limited in NPM to what you've actively decided to monitor. Using NCM, you are looking at actual device configurations.
In NCM, you would make a Policy Report to search all configurations for the string (you can use RegEx as well).
Essentially, you make a rule that says "look for this"; you assign that rule to a policy and then select the devices you want to search through; then apply that policy to a report.
It's explained fairly well in the NCM admin guide under Managing Policy Reports.
In NPM, VLANs are polled on nodes and are linked to specific interfaces on that node, so when you discover your network it will discover also all VLANs. Then on Interface filtering page during the Discovery import wizard, you can import interfaces based on VLAN ID or Port Mode (access/trunk). So when you filter out only access ports and import them, you will have only VLANs with access ports in NPM.