I'm not sure - I thought that limma had general functionality for Agilent arrays, though (both single and two-colour arrays). I wish that I had more time to check myself; otherwise, you can obtain the normalised data with the following code:

library(Biobase)
library(GEOquery)

# load series and platform data from GEO
gset <- getGEO("GSE36864", GSEMatrix =TRUE, getGPL=FALSE)
if (length(gset) > 1) idx <- grep("GPL8687", attr(gset, "names")) else idx <- 1
gset <- gset[[idx]]

With that, you can do DEA and any other downstream stuff.