Abstract:The Marine Predator Algorithm (MPA) has unique advantages as an important branch of population-based algorithms. However, it emerges more disadvantages gradually, such as traps to local optima, insufficient diversity, and premature convergence, when dealing with complex problems in practical industrial engineering design applications. In response to these limitations, this paper proposes a novel Improved Marine Predator Algorithm (IMPA). By introducing an adaptive weight adjustment strategy and a dynamic social learning mechanism, this study significantly improves the encounter frequency and efficiency between predators and preys in marine ecosystems. The performance of the IMPA was evaluated through benchmark functions, CEC2021 suite problems, and engineering design problems, including welded beam design, tension/compression spring design, pressure vessel design, and three-bar design. The results indicate that the IMPA has achieved significant success in the optimization process over other methods, exhibiting excellent performance in both solving optimal parameter solutions and optimizing objective function values. The IMPA performs well in terms of accuracy and robustness, which also proves its efficiency in successfully solving complex industrial engineering design problems.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is the limitations of the existing Marine Predator Algorithm (MPA) when dealing with complex industrial engineering design optimization problems. Specifically, MPA is prone to getting trapped in local optimal solutions, lacking in diversity, and converging prematurely. To address these issues, the author proposes an Improved Marine Predator Algorithm (IMPA). By introducing an adaptive weight adjustment strategy and a dynamic social learning mechanism, the frequency and efficiency of encounters between predators and prey are significantly improved.
### Main contributions of the paper
1. **Adaptive weight parameter adjustment**: Adjust the adaptive weight parameter according to the proportion of fitness values, which enhances the global exploration ability of the algorithm. When the optimization conditions are stable, it shifts to the exploitation of local solutions, achieving a balance between global and local exploration.
2. **Dynamic social learning mechanism**: By balancing the potential neighborhood information of elite predators with the information of other individuals, the algorithm's search space optimization ability and the effectiveness of the search process are improved, reducing the risk of getting trapped in local optimal solutions.
3. **Performance evaluation**: IMPA is evaluated through benchmark functions, CEC2021 suite problems, and engineering design problems (such as welded beam design, tension - compression spring design, pressure vessel design, and three - bar design). The results show that IMPA has achieved remarkable success in the optimization process and performs excellently, especially in solving the optimal parameter solutions and optimizing the objective function values.
### Formula summary
- **Generation of initial solutions**:
\[
X_0 = X_{\text{min}}+ \text{rand}\times(X_{\text{max}} - X_{\text{min}})
\]
- **Update of elite matrix**:
\[
\text{Elite}=\begin{bmatrix}
X_{I_{1,1}}, X_{I_{1,2}}, \ldots, X_{I_{1,d}}\\
X_{I_{2,1}}, X_{I_{2,2}}, \ldots, X_{I_{2,d}}\\
\vdots\\
X_{I_{n,1}}, X_{I_{n,2}}, \ldots, X_{I_{n,d}}
\end{bmatrix}_{n\times d}
\]
- **Update of prey matrix**:
\[
\text{Prey}=\begin{bmatrix}
X_{1,1}, X_{1,2}, \ldots, X_{1,d}\\
X_{2,1}, X_{2,2}, \ldots, X_{2,d}\\
\vdots\\
X_{n,1}, X_{n,2}, \ldots, X_{n,d}
\end{bmatrix}_{n\times d}
\]
- **Step - size and position updates in different stages**:
- **High - velocity - ratio stage (HVR)**:
\[
\text{stepsize}_i=\text{RB}\otimes(\text{Elite}_i - \text{RB}\otimes\text{Prey}_i),\quad i = 1,\ldots,n
\]
\[
\text{Prey}_i=\text{Prey}_i+P\cdot R\otimes\text{stepsize}_i
\]
- **Uniform - velocity - ratio stage (UVR)**:
\[
\text{stepsize}_i=\text{RL}\otimes(\text{Elite}_i - \text{RL}\otimes\text{Prey}_i),\quad i = 1,