Lucas R. C. Farias,Aluizio F. R. Araújo
Abstract:This paper introduces the inverse modeling constrained multi-objective evolutionary algorithm based on decomposition (IM-C-MOEA/D) for addressing constrained real-world optimization problems. Our research builds upon the advancements made in evolutionary computing-based inverse modeling, and it strategically bridges the gaps in applying inverse models based on decomposition to problem domains with constraints. The proposed approach is experimentally evaluated on diverse real-world problems (RWMOP1-35), showing superior performance to state-of-the-art constrained multi-objective evolutionary algorithms (CMOEAs). The experimental results highlight the robustness of the algorithm and its applicability in real-world constrained optimization scenarios.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is to propose an Inverse Modeling Constrained Multi - Objective Evolutionary Algorithm based on Decomposition (IM - C - MOEA/D) for Constrained Multi - Objective Optimization Problems (CMOPs). Such problems are very common in practical applications, such as robot gripper optimization, water resource management, gearbox design and process synthesis. CMOPs are more challenging than unconstrained multi - objective optimization problems because they contain multiple objectives and constraints simultaneously.
Specifically, the objectives of the paper include:
1. **Introduce a new algorithm framework**: Combine inverse modeling and decomposition methods to develop a new evolutionary algorithm IM - C - MOEA/D to more effectively handle multi - objective optimization problems with constraints.
2. **Improve optimization performance**: Verify the performance of the new algorithm in solving various practical problems through experiments, and prove that it is superior to several existing advanced Constrained Multi - Objective Evolutionary Algorithms (CMOEAs).
3. **Deal with complex constraints**: Pay special attention to how to effectively guide the optimization process in the presence of multiple inequality and equality constraints, ensure the feasibility of solutions and improve the ability to converge to the Pareto Front.
### Formula Representation
The formulas involved in the paper are as follows:
1. **Definition of Constrained Multi - Objective Optimization Problem (CMOP)**:
\[
\min f(x)=(f_1(x), f_2(x), \dots, f_m(x))^T
\]
\[
\text{subject to} \begin{cases}
g_j(x)\leq0, & j = 1, \dots, l \\
h_j(x)=0, & j = l + 1, \dots, k \\
x=(x_1, x_2, \dots, x_d)^T\in\Omega
\end{cases}
\]
where \(x\) is a \(d\)-dimensional real - valued decision vector in the decision space, \(f(x)\) is an \(m\)-dimensional objective vector in the objective space, \(\Omega\subset\mathbb{R}^d\) defines the feasible region of the decision space, \(g_j(x)\leq0\) is the \(j\)-th inequality constraint, \(h_j(x) = 0\) is the \((j - l)\)-th equality constraint, and \(l\) and \(k - l\) represent the number of inequality and equality constraints respectively.
2. **Tchebycheff Decomposition**:
\[
\min g_{\text{TCH}}(x|\lambda, z)=\max_{1\leq j\leq m}(\lambda_j|f_j(x)-z_j|)
\]
\[
\text{subject to }x\in\Omega
\]
where \(m\) is the number of objectives, \(z\) is the ideal reference point, that is, \(z_j=\min\{f_j(x)|x\in\Omega\}\) for each \(j = 1, \dots, m\). The \(m\)-dimensional weight vector is defined as \(\lambda=(\lambda_1, \dots, \lambda_m)^T\), and satisfies \(\sum_{j = 1}^{m}\lambda_j=1\) and \(\lambda_j\geq0\).
3. **Global Replacement Scheme**:
\[
i=\arg\min_{1\leq