|
shallowdeep -> RE: What happens if . . . (2/18/2010 2:29:32 AM)
|
quote:
ORIGINAL: velt
Aren't Sd(n) and Ss(n) in the wrong places in the diagram? ... making sure the output would be unbounded was half the fun, of course.
Unstable systems can indeed be amusing. And yes, it does look like I reversed the output labels. Sorry about that. I need to remind myself to take more time with serious analysis problems like this… I wonder if VC's love can survive this.
I would fix it, but I came up with a modified system I like better (see below). Some of the states may not have been well defined in the last one, leading to some issues with what differentiates initial stimulation versus output stimulation and such. The input and output are all defined in terms of the participants' energy now and the network includes adjustable feedback gains to allow for Lady Angelika's "adaptation." I'm not touching the potential nonlinearity of that; there's no sense throwing out perfectly good analysis tools simply because reality isn't doing a good job living up to the model. :)
quote:
ORIGINAL: VaguelyCurious
You don't hit them once and then just stop, after all, do you?
Actually, if you're trying to characterize a system, hitting it with an impulse is often what you do. Really. Fortunately velt set things up in the discrete domain so you can use a finite Kronecker Delta; if it was continuous, you'd need to deliver a blow with infinite intensity in an infinitesimal amount of time to get the Dirac version. Sounds painful.
quote:
Aaaand incidentally, Si is going to increase with every stroke-although that won't affect the end value, it should affect the rate of convergence, right?
Technically yes, but it actually sort of depends on the level of system instability. If the loop gain is much higher than one, the impact of later impacts (sorry) really isn't that significant: e.g. If you have a gain of 10, by the third blow you'd need to be hitting 1000 times harder than the initial stroke just to double the system's rate of divergence.
While this was mostly for amusement, it is actually how I tend to think about things, so I have been thinking. I can't contribute much, but here's an attempt to offset my derailment:
This all reminded me a bit of a podcast I listened to a few months ago that termed play a "Reciprocal Energy Loop." You can listen to it here if you want – the relevant part starts at about 39:10. Basically, it was talking about the need to provide some sort of feedback to keep things interesting and prevent an 'energy suck' from emerging. In the hypothetical situation of the OP, there really isn't necessarily a mismatch in potential interests (i.e. the expressivity and perceptivity functions could still align fairly well) – all that's lacking is some kickstart to get the feedback going. Providing someone can take the initiative to get that started, I think it could work out. One of the sub guys in the interview talked about how he's "manipulative" in the sense that he's "energetically seducing you in a way." I know some posters here (Akasha comes to mind) like to talk about how being a good bottom takes work. If you have an interest in seeing a woman enjoy dominance, I see a need to try to figure out what she enjoys and then how to press those buttons. Even if it doesn't immediately do it for you, there's probably a way to get the right energy cascade started. I don't see that sort of 'manipulation' as being subversive to the dynamic, but rather helping it.
Of course I'm not a mistress and I'm very, very far from an expert… so take all the above with a big boulder of salt.
Back to engineering...
[image]local://upfiles/324704/D63CF8E375F84C13AAD444BBEE63391D.jpg[/image]
Edited to add that the last item in the key should be K_SFB, not K_DFB
|
|
|
|
