Some remarks on a conjecture in parameter adaptive control

A.S. Morse has raised the following question: Do there exist differentiable functions

$\text{f:R}{}^2\text{→ R}\text{and}\text{g:R}{}^2\text{→ R}$

with the property that for every nonzero real number λ and every (x₀, y₀) ∈ $\text{R}$² the solution (x(t),y(t)) of

$\text{x}\text{?}\text{(t) = x(t) + λf(x(t),y(t))}$

,

$\text{y}\text{?}\text{(t) = g(x(t),y(t))}$

,

$\text{x(0) = x}{}_0\text{, y(0) = y}{}_0$

, is defined for all t ? 0 and satisfies

$\text{lim}\text{t → + ∞}$

and y(t) is bounded on [0,∞)? We prove that the answer is yes, and we give explicit real analytic functions f and g which work. However, we prove that if f and g are restricted to be rational functions, the answer is no.     

Asymptotic expansions for the Landau density and distribution functions

Asymptotic expansions are given for the Landau function

$\text{φ(λ)=}\text{1}\text{2πi}\text{∫}\text{σ+i∞}\text{σ?i∞}\text{e}{}^{\mathrm{\mu ln\mu +\lambda \mu }}\text{dμ}$

and for its integral, as λ → ± ∞. Further, it is shown that φ(λ) is indeed a probability density function, as is always assumed in applications.     

Fractional power approximation and its generation

For an analog simulating system, an approximating system is proposed. Its mathematical form is expressed by an algebraic equation:

$\text{? (x) ≈ α + β χ + γχ}{}^{\mathrm{k}}$

with four parameters given by real numbers. Their values can be determined so as to satisfy a best fit in a Chebyshev sense. Then, the accuracy is of the same order with that obtained by any kind of ordinary power series up to terms o f the third order. It is noticeable that a given function can be accurately approximated by this equation without destroying its uniform continuity.As its electronic system, it is possible to utilize forward current-voltage characteristics o f p-n diodes near their origins. Different functions can be produced by adjusting an amount o f ordinary resistances connected to diodes in series.     

Averaging shifted histograms

A method which consists in shifting different histograms of the same spectrum and then taking their average is presented in order to smooth the data and to increase the localization accuracy and separation of the peaks. The statistical properties of this method are investigated. The average of two histograms with shifted bin limits is studied. It is shown that for histograms with random bin limits, distributed according to

$\text{F}{}_{\mathrm{i}}\text{(x)=}\text{∫}\text{?∞}\text{x}\text{?}{}_{\mathrm{i}}\text{(ξ, μ}{}_{\mathrm{i}}\text{, σ)}\text{d}\text{ξ}$

; where the standard deviation σ is very small compared to the difference of the means (μ_i+1 ? μ_i) for ll i the zero order approximation to the variance of this histogram is given by:

$\text{var}\text{∽}\text{(H)=}\text{∑}\text{i=0}\text{m}\text{(A}{}_{\mathrm{i+1}}\text{?a}{}_{\mathrm{i}}\text{)}{}^2\text{F}{}_{\mathrm{i+1}}\text{(x)(1?F}{}_{\mathrm{i+1}}\text{(x))}$

, where

$\text{a}{}_{\mathrm{i}}\text{=}\text{1}\text{x}{}_{\mathrm{i=1}}\text{?x}{}_{\mathrm{i}}\text{∫}\text{x}{}_{\mathrm{i}}\text{x}{}_{\mathrm{i+1}}\text{g(ξ)}\text{d}\text{ξ}$

and g is an unknown function fitted by the histogram. Formula (1) gives also the relation:

$\text{v}\text{a}\text{?}\text{r}\text{(}\text{(H}{}_1\text{+ H}{}_2\text{)}\text{2}\text{) =}\text{1}\text{4}\text{(}\text{v}\text{a}\text{?}\text{r}\text{(H}{}_1\text{(x)) +}\text{v}\text{a}\text{?}\text{r}\text{(H}{}_2\text{(x))}$

, when H₁ and H₂ have stochastically independent bin limits.When the histogram H is considered as a spline function S of order one it is shown that for the minimization criterion with respect to the coefficient of the spline:

$\text{M}{}_1\text{=}\text{min}\text{∫}\text{x}{}_1\text{x}{}_{\mathrm{m+1}}\text{(g(x) ? S}{}_1\text{(x))}{}^2\text{d}\text{x}$

, the following result holds: $\text{M}{}_{\mathrm{a}}\text{?}{}^1{}_2\text{(M}{}_1\text{+ M}{}_2\text{)}$, where $\text{S}{}_{\mathrm{a}}\text{(x) =}{}^1{}_2\text{(S}{}_1\text{(x) + S}{}_2\text{(x))}$. If the number of shifted histograms tends to infinity, then

$\text{S}{}_{\mathrm{\infty }}\text{(x) = [Γ(x + h) + Γ(x ? h) ? 2Γ(x)]/h}{}^2$

, where $\text{Γ(x) =}\text{∫}\text{?∞}\text{x}\text{∫}\text{?∞}\text{η}\text{g(ξ)}\text{d}\text{ξ}\text{d}\text{η}$, and h is a constant bin size. Then

$\text{M}{}_{\mathrm{\infty }}\text{≈}\text{h}{}^4\text{144}\text{∫}\text{x}{}_1\text{x}{}_{\mathrm{m+1}}\text{g″}{}^2\text{(x)}\text{D}\text{x}$

. Extensions to two-dimensional histograms and to higher order (empirical distributions) are presented.     

Computable Analysis of a Boundary-Value Problem for the Korteweg-de Vries Equation

The initial-boundary-value problem for the Korteweg-de Vries (KdV) equation:

$u_{t}+u_x+uu_{x}+u_{xxx}=0, \quad t, x\geq 0,$

$u(x, 0)=\varphi (x),\quad u(0,t)=h(t), \qquad\varphi(0)=h(0),$

defines a nonlinear continous map from the space where the auxiliary data are drawn to the space of solutions. By making use of modern methods for the study of nonlinear dispersive equations, it is shown that the solution map\(H^{3m-1}({\Bbb R}^+)\times H^{m}(0,T)\to C([0,T];H^{3m-1}({\Bbb R}^+))\) is Turing computable for any integer m ≥ 2 and computable real number T > 0. This result provides yet another affirmative answer to the open question raised by Pour-El and Richards [PER]: Is the solution operater of the KdV equation computable?

     

Global stability for separable nonlinear delay differential equations

Consider the following separable nonlinear delay differential equation

, where we assume that, there is a strictly monotone increasing function f(x) on (−∞, +∞) such that

In this paper, to the above separable nonlinear delay differential equation, we establish conditions of global asymptotic stability for the zero solution. In particular, for a special wide class of f(x) which contains a case of f(x) = e^x−1, we give more explicit conditions. Applying these, we offer conditions of global asymptotic stability for solutions of nonautonomous logistic equations with delays.     

A method for the calculation of eigenfunction expansions

In this paper, a method is presented for the calculation of the coefficients of the series expansion of a function f(t), in the base orthonormal set made up by the eigenfunctions of the self-adjoint operator $\text{L: L(x(t)) = (}\text{d}\text{dt}\text{)( p(t)(}\text{dx(t)}\text{dt}\text{))?q(t)x(t)}$. We show that the values of the numbers $\text{t}\text{}$x_k> can be obtained by solving the differential equation L + λ) y(t) = Kf(t), in the interval of definition, for each of the eigenvalues λ of L and by using as initial conditions those which determine one of its associated orthonormal functions. This makes the method specially interesting for its implementation on a hybrid computer: One advantage of the proposed method is that the analysis of f(t) does not require the simultaneous presence of the functions of the base set and the problem signal, thus eliminating both the problems of the synchronized generation of signals and the need for storing it in memory.     

An upper bound on the mean-square error in the sampling expansion for non-band-limited processes

An upper bound is obtained on the mean-square error involved when a real-valued non-band-limited nonstationary random process x(t) is approximated by the sampling expansion

$\text{∑}\text{n=?∞}\text{∞}\text{x(nT)}\text{sin}\text{π(t?nT)/T}\text{π(t?nT)/T}$

for some T > 0. When the process x(t) is band-limited over [$\text{?}\text{1}\text{2T}\text{,}\text{1}\text{2T}$], this error bound reduces to zero.     

On a fourth-order unconditionally stable scheme for damped second-order systems

We consider the damped second-order system $\text{M}\text{u}\text{?}\text{+ C/.u + Ku = F(t)}$ (M, C, K symmetric, positive definite n × n matrices) by conversion to an equivalent first-order system

$\text{I}\text{O}\text{O}\text{M}\text{d}\text{d}\text{t}\text{O}\text{?K}\text{I}\text{?C}\text{u}\text{u}\text{?}\text{+}\text{o}\text{F(t)}$

We demonstrate that an algorithm proposed by Fairweather for the implementation of the (2, 2) Padé approximation of the exponential matrix for approximating the solution of homogeneous first-order systems extends advantageously to this case, yielding an unconditionally stable fourth-order scheme with the feature that the approximating equations decouple. As a result we are required only to solve one symmetric complex n × n system of linear algebraic equations at each time step, with a fixed matrix which may be decomposed into triangular factors at the outset. We also consider iterative schemes involving only real, positive definite, symmetric n × n matrices. Numerical results are included.     

Cubature formulae for nearly singular and highly oscillating integrals

The paper deals with the approximation of integrals of the type

$$\begin{aligned} I(f;{\mathbf {t}})=\int _{{\mathrm {D}}} f({\mathbf {x}}) {\mathbf {K}}({\mathbf {x}},{\mathbf {t}}) {\mathbf {w}}({\mathbf {x}}) d{\mathbf {x}},\quad \quad {\mathbf {x}}=(x_1,x_2),\quad {\mathbf {t}}\in \mathrm {T}\subseteq \mathbb {R}^p, \ p\in \{1,2\} \end{aligned}$$

where \({\mathrm {D}}=[-\,1,1]^2\), f is a function defined on \({\mathrm {D}}\) with possible algebraic singularities on \(\partial {\mathrm {D}}\), \({\mathbf {w}}\) is the product of two Jacobi weight functions, and the kernel \({\mathbf {K}}\) can be of different kinds. We propose two cubature rules determining conditions under which the rules are stable and convergent. Along the paper we diffusely treat the numerical approximation for kernels which can be nearly singular and/or highly oscillating, by using a bivariate dilation technique. Some numerical examples which confirm the theoretical estimates are also proposed.

     

Sulle formule di quadratura di Tschebyscheff

For the Tschebyscheff quadrature formula: $$\int\limits_{ - 1}^1 {\left( {1 - x^2 } \right)^{\lambda - 1/2} f(x) dx} = K_n \sum\limits_{k = 1}^n {f(x_{n,k} )} + R_n (f), \lambda > 0$$ it is shown that the degre,N, of exactness is bounded by: $$N \leqslant C(\lambda )n^{1/(2\lambda + 1)} $$ whereC(λ) is a convenient function of λ. For λ=1 the complete solution of Tschebyscheff's problem is given.     

Truncation error for band-limited nonstationary processes

Let x(t) be a real-valued random process band-limited to the interval [$\text{?1}\text{2T}\text{,}\text{1}\text{2T}$] for some T > 0. In this note we find an upper bound on the mean square of the truncation error involved when x(t) is approximated in the interval $\text{|t| ?}\text{T}\text{2}$ by the finite selection

$\text{∑}\text{n=?N}{}_1\text{N}{}_2\text{x(nt)}\text{sin}\text{π(t?nT)}\text{T}\text{π(t?nT)}\text{T}$

of terms from its sampling expansion representation.     

Lower bounds for polynomials with algebraic coefficients

We give a method, based on algebraic geometry, to show lower bounds for the complexity of polynomials with algebraic coefficients. Typical examples are polynomials with coefficients which are roots of unity, such as

$\text{Σ}\text{j=1}\text{d}\text{e}{}^{\mathrm{<mtext>2\pi i</mtext><mtext>i</mtext>}}\text{X}{}^{\mathrm{i}}$

and

$\text{Σ}\text{j=i}\text{d}\text{e}{}^{\mathrm{<mtext>2\pi i</mtext><mtext>pi</mtext>}}\text{X}{}^{\mathrm{j}}$

where p_j is the jth prime number.We apply the method also to systems of linear equations.     

New stability theorems concerning one-step numerical methods for ordinary differential equations

A measure of the stability properties of numerical integration methods for ordinary differential equations is provided by their stability region, which is that region in the complex (Δtλ) plane for which a given method is stable when applied to the differential equation

$\text{d}\text{y}\text{d}\text{t=λy}$

with a time-step Δt.Free parameters which exist in numerical integration algorithms may be used to maximize, in some sense, the size of the stability region, rather than increasing the order of accuracy, as is usually done. We derive new results which set theoretical limits to this maximization process for one step, explicit methods.Specifically, if K is the number of function evaluation invoked, then:(i) we prove (Theorem 1) that if ? is the radius of the largest circle, tangent to the imaginary axis at the origin of the complex plane that is contained in the stability region S, then ? cannot exceed K.(ii) we also prove (Theorem 2) that the imaginary stability boundary S_I (or maximum stable value of ∣Δtλ∣ with λ imaginary) cannot exceed (K ? 1).While Theorem 1 is to our knowledge new, a limited form of theorem 2 (K odd only) had been established in v.d. Houwen (1977). That the maximum imaginary boundary S_I = (K ? 1) is attainable had been shown (constructively) for K odd. We show that this maximum is also reached for K = 2 and K = 4, and correct in the process an erroneous result in the above reference.     

A general bayes rule and its application to nonlinear estimation

This paper consists of two main results, a general Bayes rule, and a general Bucy representation theorem. The general Bayes rule is a natural generalization of the elementary Bayes rule:

$\text{P(A}\text{B)}\text{P(A)}\text{=}\text{P(B}\text{A)}\text{P(B)}$

. The general Bucy representation theorem plays a central role in nonlinear estimation theory as does the Bucy theorem in nonlinear filtering. A simple and direct proof of the general Bucy representation theorem is obtained by the application of the general Bayes rule.     

Nonlinear superposition principles: A new numerical method for solving matrix Riccati equations

Previously derived superposition formulae, expressing the general solution of the matrix Riccati equation $\text{W}\text{?}\text{W(t) = A + WB + CW + WDW (A(t), B(t), C(t), D(t), W(t) ?}\text{R}{}^{\mathrm{n\times n}}\text{, n ≥ 2, t ?}\text{R}\text{)}$ in terms of 5 particular solutions, are applied to obtain numerical solutions of this coupled system of nonlinear equations.     

Crosscorrelation between linearly and nonlinearly distorted versions of a given signal

A given deterministic signal x(.) is distorted by passing it through a linear time-invariant filter and also by subjecting it to the action of an instantaneous nonlinearity. The resulting time crosscorrelation of the two distorted versions of the original signal is expressed by the function

$\text{R}{}_2\text{(s)?∫}{}_{\mathrm{?\infty }}{}^{\mathrm{\infty }}\text{?∫}{}_{\mathrm{?\infty }}{}^{\mathrm{\infty }}\text{g[x(t)]k(t?t′)x(t?s)dt dt′}$

, where x(.) is the given signal, k(.) is the nonnegative definite impulse response of the linear filter, and g(.) is the output-input characteristic of the zero-memory nonlinear device. The problem considered is that of determining conditions on the pair (x,g) such that R₂(s) ? R₂(0) for all s and any choice of nonnegative definite filter function k; the principal result is the formulation of a necessary and sufficient condition for R₂ to have a global maximum at the origin. In particular, the peak value occurs at the origin if and only if $\text{G}{}_{\mathrm{x}}{}^1\text{(ω)X(ω)}$ is real and nonnegative for all ω ? 0, where G_x(.) and X(.) are the Fourier transforms of g[x(.)] and x(.), respectively. An equivalent condition is that the correlation function

$\text{R}{}_2\text{(s)?∫}{}_{\mathrm{?\infty }}{}^{\mathrm{\infty }}\text{g[x(t)]x(t?s)dt}$

, previously studied by Richardson, be nonnegative definite.Several examples are given, and it is shown that, unlike the case for R₁(.), monotonicity of g(.) is not a sufficient condition for R₂(.) to have a global maximum at s = 0 independently of the choice of filter characteristic k. Certain extensions of these results are given for the case when x(.) is a Gaussian random input.