Semi-linear evolution equations via positive semigroups

Wolfgang Arendt,Daniel Daners

2024-07-08

Abstract:We study semi-linear evolutionary problems where the linear part is the generator of a positive $C_0$-semigroup. The non-linear part is assumed to be quasi-increasing. Given an initial value in between a sub- and a super-solution of the stationary problem we find a solution of the semi-linear evolutionary problem. Convergence as $t\to\infty$ is also studied for the solutions. Our results are applied to the logistic equation with diffusion, to a Lotka-Volterra competition model and the Fisher equation from population genetics.

Analysis of PDEs

What problem does this paper attempt to address?

The problem that this paper attempts to solve is about the existence and properties of solutions to semilinear evolution equations. In particular, under the condition that the given initial value lies between a subsolution and a supersolution, find the solutions of these equations and study the convergence of the solutions as time \(t \to \infty\). Specifically, the paper focuses on the semilinear Cauchy problem of the form:

\[
\begin{cases}
u'(t)+A u(t) = F(u(t)), & t > 0, \\
u(0)=u_0,
\end{cases}
\]

where \(A\) is the generator of a positive \(C_0\)-semigroup and \(F\) is a quasi - increasing function. By introducing the concepts of subsolutions and supersolutions and using the properties of positive operators and ordered Banach spaces, the author proves the existence, uniqueness and asymptotic behavior of solutions to such problems.

### Main problems
1. **Existence of solutions**: For the case where the given initial value \(u_0\) lies between a subsolution \(u\) and a supersolution \(u\), prove the existence of the minimal mild solution \(u_{\text{min}}\) and the maximal mild solution \(u_{\text{max}}\).
2. **Properties of solutions**: Study the monotonicity and stability of these mild solutions.
3. **Asymptotic behavior**: Study the convergence of the solutions as time \(t \to \infty\), in particular, prove that the solutions will converge to an equilibrium point.

### Solutions
- **Subsolutions and supersolutions**: Define the concepts of subsolutions and supersolutions and use them to construct iterative sequences.
- **Fixed - point mapping**: Define the initial - value - related fixed - point mapping \(G\) and prove the existence and uniqueness of solutions by studying its properties.
- **Monotone iteration**: Prove the existence and convergence of solutions by the method of monotone iterative sequences.
- **Compactness conditions**: Consider the compactness of the semigroup and the properties of the ordered Banach space to ensure the convergence of solutions.

### Applications
The results of the paper are applied to specific mathematical models, such as the logistic equation with diffusion, the Lotka - Volterra competition model and the Fisher equation, which are of great significance in ecology and population genetics.

### Conclusions
The main conclusion of the paper is that, under certain assumptions (such as the compactness of the semigroup or the norm - continuity of the ordered Banach space), for the case where the given initial value \(u_0\) lies between the subsolution and the supersolution, the semilinear evolution equation has a minimal mild solution and a maximal mild solution, and these solutions will converge to an equilibrium point as time \(t \to \infty\).