Abstract:Multidimensional systems coupled via complex networks are widespread in nature and thus frequently invoked for a large plethora of interesting applications. From ecology to physics, individual entities in mutual interactions are grouped in families, homogeneous in kind. These latter interact selectively, through a sequence of self-consistently regulated steps, whose deeply rooted architecture is stored in the assigned matrix of connections. The asymptotic equilibrium eventually attained by the system, and its associated stability, can be assessed by employing standard nonlinear dynamics tools. For many practical applications, it is however important to externally drive the system towards a desired equilibrium, which is resilient, hence stable, to external perturbations. To this end we here consider a system made up of $N$ interacting populations which evolve according to general rate equations, bearing attributes of universality. One species is added to the pool of interacting families and used as a dynamical controller to induce novel stable equilibria. Use can be made of the root locus method to shape the needed control, in terms of intrinsic reactivity and adopted protocol of injection. The proposed method is tested on both synthetic and real data, thus enabling to demonstrate its robustness and versatility.
Physics and Society,Statistical Mechanics,Adaptation and Self-Organizing Systems,Biological Physics
What problem does this paper attempt to address?
The problem that this paper attempts to solve is to achieve stable control of an externally specified equilibrium state in multi - dimensional systems on complex networks. Specifically, the paper proposes a method to manipulate the asymptotic dynamics of the system by introducing an additional species (controller) into the system, so as to reach an existing and stable fixed point. This method can be used not only for theoretical analysis but also in practical applications, such as in ecology, genetic regulatory networks, social community dynamics, learning mechanisms in brain tissues, and robot navigation in network groups, effectively driving the system to reach the required stable state.
### Main contributions of the paper
1. **Proposed a general control strategy**: The paper developed a general control strategy for multi - dimensional systems (these systems are composed of a large number of components interacting through complex networks). By introducing an additional species as a controller, the asymptotic dynamics of the system can be manipulated to achieve a specific stable equilibrium state.
2. **Used the root - locus method for control design**: To achieve the above - mentioned goals, the paper adopted the root - locus method to design the parameters of the controller. This method adjusts the coupling strength between the controller and the existing species, enabling the system to reach and maintain the desired stable equilibrium state.
3. **Verified the robustness and generality of the method**: The paper tested the proposed control method using synthetic data and real - world data sets, demonstrating its robustness and generality in different application scenarios. For example, in synthetic gene networks, the activity of specific nodes can be enhanced or suppressed through control; in the gut microbiome, the concentration of the pathogen Clostridium difficile can be reduced through control or a certain species can be made extinct.
### Specific technical details
- **System model**: The paper considered a system composed of \( N \) interacting populations, whose activities \( \mathbf{x}=(x_1, x_2,\ldots, x_N)^T \) follow the coupled nonlinear equations:
\[
\dot{x}_i = f_i(x_i)+\sum_j A_{ij}g_i(x_i, x_j)\quad(i = 1,\ldots, N)
\]
where \( f_i(x_i) \) represents the auto - dynamics of species \( i \), \( A_{ij} \) is the weighted connection matrix, and \( g_i(x_i, x_j) \) describes the interaction between species \( i \) and \( j \).
- **Introduction of the controller**: To achieve control of the system's equilibrium state, an additional species \( u \) was introduced, and its dynamic equations are:
\[
\begin{cases}
\dot{x}_i = f_i(x_i)+\sum_j A_{ij}g_i(x_i, x_j)+\alpha_i h_i(x_i, u)\\
\dot{u}=-(u - u^*) - \rho\sum_j \beta_j(x_j - x_j^*)
\end{cases}
\]
where \( \alpha_i \) and \( \beta_j \) are control parameters, and \( h_i(x_i, u) \) describes the interaction pattern between the controller and the existing species.
- **Stability analysis**: By linearizing the perturbed system, the Jacobian matrix \( J \) was obtained, and the real parts of its eigenvalues were analyzed. If the real parts of all eigenvalues are negative, it indicates that the system is stable at the selected fixed point. The paper used the root - locus method to determine the appropriate control parameters \( \rho \) and \( \beta \) to ensure the stability of the system.
### Application examples
- **Synthetic gene network**: By controlling to enhance or suppress the activities of peripheral nodes or central nodes in a dendritic network.
- **Gut microbiome**: By controlling to reduce the concentration of the pathogen Clostridium difficile or make a certain species extinct.
In conclusion, this paper provides a general and effective method that can achieve stable control of an externally specified equilibrium state in a variety of complex multi - dimensional systems and has broad application prospects.