Auto sync: 2026-02-10 21:15:03 (1 files changed)

M  3-research-approach/v2.tex
This commit is contained in:
Dane Sabo 2026-02-10 21:15:03 -05:00
parent 20c499ac35
commit 28091b89a9

View File

@ -347,7 +347,62 @@ according to the control input and control law, while the operational level is a
discrete state evolution which determines the tactical control law to reach
different operational states.
This operational control level is the main reason for the requirement of human
opeartors in nuclear control today. The hybrid nature of this control system
makes it difficult to prove that a controller will perform according to the
strategic requirements, as the infrastructure to build hybrid systems today
dooes not exist. Humans have been used for this layer because the general
intelleigence of humans has be relied upon as a safe way to manage the hybrid
nature of this system. But, these operators are using prescriptive operating
manuals to perform their control with strict procedures on what control to
implement at a given time. These procedures are the key to the operational
control scope.
The method of constructing a HAHACS in this proposal leverages two key points of
the way this control scope is done today: first, the operational scope control
is effectively discrete control. Second, the rules of implementing this control
are described a priori to their implementation in operating procedures. We can
make great use of these facts by formalizing the rules for transitioning between
discrete states. To do this, we will use temporal logic to formalize discrete
switching behavior.
Temporal logic is a rich syntax that allows for the definition of logical
calculations including time related bounds. For this reason, we can make
statements relating discrete control modes to one another. Using temporal logic,
we can effectively describe all of the requirements of a HAHACS. The guard
conditions \(G\) are easily defined by determining boundary conditions between
discrete states and defining their behavior, while continuous mode invariants
can be defined using temporal logic statements as well. These form the basis of
any proofs about a HAHACS, and are the fundamental 'truth' statements about what
the behavior of the system is designed to be.
To build these temporal logic statements, an intermediary tool called FRET is
planned to be used. FRET stands for Formal Requirements Elicitation Tool, and
was designed by NASA to build high assurance timed systems. FRET is an
intermediarly language between temporal logic and natural language that allows
for rigid definitions of temporal behvarior while using a logic-novice friendly
syntax. This benefit is crucial for the feasibility of this methodology for
industry, as minimizing the barrier to formal methods is a critical component of
their scucess. By reducing the expert knowledge required to use these tools,
their adoption with current workforce becomes easier.
A key feature of FRET is the ability to start with logically imprecise
statements and consecutively refine them into a well-posited specification. We
can use this to our advantage by directly dumping in operating procedures and
design requirements into FRET in natural language, and iteratively refining them
into the specifications for a HAHACS. This has two distinct but important
benefits. First, it allows us to draw a direct link from the design
documentation to the digital system implementation. Second, it clearly
demonstrates where the natural language documents are insufficient. These
procedures may still be used by human operators, so any wiggle room for
interpretation is a weakness that must be addressed.
%Talk about how we go from temp logic to reactive synth. Metnion fret can
%export, or naturlly support reactive synth solver ltlsynt, a sota react synth
%solver
% talk about what the benefits of reactive synth are. Proof chain, machine
% checkable, blah blah blah
%%%%%%%%%%%% Building continuous controllers