Abstract:
Power electronics systems are extensively used in many engineering systems,
such as hand-held devices, medical equipment, electric vehicles, aerospace
artifacts, power grids, nuclear power plants and military equipment. Power
electronics are characterized by the highly-nonlinear behavior due to the in-
volvement of semiconductor switches for power conversion. Mathematical
concepts of basic circuit theory, operational calculus, differential theory, in-
tegration theory and control theory are involved in modeling their behavior.
Traditionally, paper-and-pencil proof methods and computer-based simula-
tions are used to perform the analysis and design of power electronics sys-
tems. However, these traditional analysis techniques suffer from several lim-
itations, such as human-error, discretization and truncation errors, and thus
cannot guarantee an accurate analysis and verification of power electronics
circuits. Whereas, accurate analysis and design of power electronic circuits is
mandatory for the safety and mission-critical engineering applications, such
as medical equipment, transportation and smart grids. To overcome the
shortcomings of the traditional analysis techniques, we propose a higher-
order-logic theorem proving based framework to formally analyze and verify
power electronic systems accurately and exhaustively. Higher-order logic is
highly expressive and allows modeling complex system behaviors, whereas,
the soundness of theorem provers ensure accurate analysis and verification
of the systems.
The proposed framework, mainly, enables time and complex-domain anal-
i
ysis and design of power electronics systems within the HOL Light theorem
prover. The formalization in the time-domain paradigm allows to perform
formal periodic steady-state analysis and design of the ideal and non-ideal
equivalent circuits of power electronics circuits. In this regard, we propose:
(i) formal modeling of ideal and non-ideal behavior of the circuit elements
(ii) formal modeling of basic circuit theory notions, such as switching func-
tion technique (iii) formally verified results for the integration and differen-
tiation of piecewise functions (iv) formal modeling of differential equations
(v) formally verified results for the solutions of the differential equations
(vi) formal modeling of steady-state characteristics and design parameters.
To illustrate the usefulness of the proposed formalizations, they are used
to formally analyze the time-domain based periodic steady-state analysis of
ideal and non-ideal DC-DC Buck converters.
Formalization in the complex-domain is aimed at formally analyzing sta-
bility of the control systems in power electronics systems to control the flow
of energy from input to output. Our main contributions in complex-domain
analysis are: (i) formal modeling of stability criterion (ii) formally verified
results of factorization of characteristic polynomials upto the fourth-order
(iii) formally verified results for the stability of the characteristic polyno-
mials. The proposed formalization allows to formally specify and verify the
stability analysis of control systems based on the characteristic polynomials
which are obtained from the transfer function of the system in complex-
domain. To illustrate the practical effectiveness, we formally verify the sta-
bility of controllers in smart grids for efficient energy processing from wind
turbines.