Allow adaptable risk management for fuzzy logic

So for example, say you have 5 indicator blocks and 3/5 of them need to activate in order for a trade to occur

• Indicator Block 1
• Indicator Block 2
• Indicator Block 3 
• Indicator Block 4
• Indicator Block 5

In this case, whether you have Block 1, Block 2 & Block 3 active, it would have the same risk management as Block 5, Block 2, and Block 3 active. 


My proposal is that SQ would be able to have separate instances of Risk management based on which blocks are active. So the final strategy would have a different TPSL when Block 1, Block 2 & Block 3 active, vs when Block 5, Block 2, and Block 3 active.


This runs into a permutation issue because of the many different possible combinations that can have different risk management settings. To solve this I have a few suggestions



1, Allow the user to set a max number of batches -- Say there are 9 possible permutations but the user only wants to look at 2 different combinations of indicators to have different risk management. In this case, only 2 batches would be made. 
 
Each batch would contain a unique set of blocks. 

• If a batch of Block 1, 2, & 3 is generated, the second batch of Block 3, 2, & 1 cannot be generated because it is a different organization of the same blocks. However, a set of Block 1, 2, and 4 would be considered a "unique set of blocks" in this example. (If the first set is stored in an array, the next set can check its current array to see if it includes all three elements of the previous array)
  

2. Allow a max amount of Indicator blocks to share between batches.

• If a user sets the max amount of 1 for shared blocks, a batch of blocks 1, 2, 3 and batch of blocks 2, 3, 4 can't can't occur because 2 blocks are shared between them
• Also limiting the number of blocks that can be shared will significantly reduce permutations of similar combinations, thus allowing for SQ to identify more unique situations between batches

After these filters are in place, the blocks in each batch will be selected at random to meet the filter criteria. 


3. Eventually a Neural net can be used to organize batches based on available blocks from learned data.