Abstract:In this work, we consider the Fokker-Planck equation of the Nonlinear Noisy Leaky Integrate-and-Fire (NNLIF) model for neuron networks. Due to the firing events of neurons at the microscopic level, this Fokker-Planck equation contains dynamic boundary conditions involving specific internal points. To efficiently solve this problem and explore the properties of the unknown, we construct a flexible numerical scheme for the Fokker-Planck equation in the framework of spectral methods that can accurately handle the dynamic boundary condition. This numerical scheme is stable with suitable choices of test function spaces, and asymptotic preserving, and it is easily extendable to variant models with multiple time scales. We also present extensive numerical examples to verify the scheme properties, including order of convergence and time efficiency, and explore unique properties of the model, including blow-up phenomena for the NNLIF model and learning and discriminative properties for the NNLIF model with learning rules.
What problem does this paper attempt to address?
### The problems the paper attempts to solve
This paper aims to solve the Fokker - Planck equation in the non - linear noisy leaky integrate - and - fire (NNLIF) neuronal network model. Specifically, this equation describes the evolution of the probability density function \(p(v, t)\) of the membrane potential in the neuronal network, where \(v\) represents the membrane potential and \(t\) represents time. Due to the firing events of neurons at the microscopic level, this Fokker - Planck equation contains dynamic boundary conditions at specific internal points.
### Main problems and challenges
1. **Dynamic boundary conditions**:
- The dynamic boundary conditions of this equation involve specific internal points \(V_R\), and the conditions at these points need special treatment to ensure the accuracy of the numerical solution.
- The form of the dynamic boundary conditions is as follows:
\[
\begin{cases}
p(V_R^-, t)=p(V_R^+, t), \\
\partial_v p(V_R^-, t)=\partial_v p(V_R^+, t)+\frac{N(t)}{a}.
\end{cases}
\]
2. **Efficient numerical methods**:
- Traditional numerical methods have difficulties in dealing with dynamic boundary conditions and may introduce large errors.
- It is necessary to construct a flexible and efficient numerical scheme that can accurately handle dynamic boundary conditions and maintain numerical stability and asymptotic - preservation.
3. **Exploration of model properties**:
- Verify the convergence and time - efficiency of the scheme through numerical experiments.
- Explore the unique properties of the model, such as blow - up phenomena, learning and discrimination properties, etc.
### Solutions
The author proposes a numerical scheme based on the spectral method, which has the following characteristics:
1. **Flexibility of the spectral method**:
- Use semi - globally differentiable and integrable basis functions to accurately capture dynamic boundary conditions and avoid complicating the equation into a form containing δ - functions.
- The choice of basis functions ensures that the approximate solution satisfies the dynamic boundary conditions.
2. **Numerical stability**:
- By choosing an appropriate test function space, the method remains stable and asymptotically - preserving during the time - evolution process.
- Ensure the conservation of total mass, that is:
\[
\int_{V_{\min}}^{V_F} p(v, t)\,dv = 1.
\]
3. **Extension of application**:
- Apply this method to the NNLIF model containing learning rules, test its discrimination properties and further explore the learning behavior of the model.
### Conclusion
This paper successfully solves the problem of dynamic boundary conditions of the Fokker - Planck equation in the NNLIF model by constructing an efficient numerical scheme based on the spectral method, and verifies the effectiveness of the scheme and the properties of the model through extensive numerical experiments.