adding minET parameters to Topo ADVAE algorithm

This MR adds some more L1Menu configurable parameters to the newly introduced AnomalyDetection algorithm on L1Topo.

The purpose of these parameters is to have ET cut capabilities inside this algorithm similar to other decision algorithms in L1Topo. For the AD algorithm this may become particularly relevant as the underlying NN is trained on previously taken data and thus is likely to have learned artificial features such as existing ET cuts applied before L1Topo (e.g. jFEX currently applies a 15 GeV cut on jJets for both, sending them to L1Topo which also reflects in the contents of the RoI readout data that is effectively what the NN sees during training - as the NN have never seen a 12 GeV jJet during its training the occurrence of such an object would seem highly "anomalous").

If those thresholds were to be lowered for the benefit of other signatures/algorithms the AD algo may likely create a much higher trigger rate unless we ensure that at least we can independently keep consistent minET thresholds for the AD algorithm (increasing thresholds upstream of L1Topo is not addressed here but may have less severe implications and also does not seem realistic to be desirable for the remainder of Run 3).

Related jira ticket: https://its.cern.ch/jira/browse/ATR-30618