Abstract:<p>A modified adaptive butterfly optimization algorithm is established with the aim of addressing the "early search blindness" and the relatively poor adaptability of the sensory modality. A normal‐distribution‐based model and a Weibull‐distribution‐based adaptive model of sensory modalities are respectively proposed for the global search process and iteration process. Among them, the Weibull‐distribution‐based adaptive model of sensory modalities is mainly manifested as the c value, that is, the adaptive change trend based on the Weibull model. The performance of the modified butterfly optimization algorithm is validated using a 14‐benchmark test function and compared with performances of some latest algorithms. The experimental results indicate that the modified algorithm performs competitively in terms of accuracy and stability. Following the experiment, the modified algorithm is further tested by running a support‐vector‐machine prediction model based on engineering data of a pipe belt conveyor's flat‐pipe/pipe‐flat transition segment. The results of the modified algorithm are then compared with test‐run outcomes of the back‐propagation prediction model and KCV‐SVM model. The results show that the prediction error is well within 10%, demonstrating the method's competence as a reliable reference for future designs of pipe belt conveyors.</p>

What problem does this paper attempt to address?

The paper attempts to address two main issues: 1. **Improving the performance of the Butterfly Optimization Algorithm (BOA)**: - The paper proposes an improved adaptive Butterfly Optimization Algorithm aimed at addressing the issues of "blind search" in the early search stages and poor adaptability to perception patterns in the original BOA. - By introducing perception pattern models based on normal distribution and Weibull distribution, the improved algorithm demonstrates better adaptability and stability in global search and iteration processes. 2. **Enhancing the accuracy of the Support Vector Machine (SVM) prediction model**: - In practical engineering applications, especially in the design of the flat-to-pipe/pipe-to-flat transition section of tubular belt conveyors, a reliable prediction model is needed to determine the length of the transition section. - The paper applies the improved BOA to SVM parameter optimization, generating a prediction model, and experimentally verifies the superiority of this model in terms of prediction accuracy and stability. Specifically, the paper addresses the above issues through the following methods: - **Improved search strategy**: Introducing random numbers based on normal distribution, making the butterflies closer to the optimal solution during the search process, thereby reducing the blindness in the search process. - **Increased adaptability**: Proposing a c-value growth model based on Weibull distribution, allowing the c-value to adaptively change during the iteration process, thus improving the algorithm's adaptability and search efficiency. - **Experimental validation**: The performance of the improved algorithm is validated through 14 benchmark test functions and compared with some existing state-of-the-art algorithms, showing that the improved BOA has significant advantages in terms of convergence speed and accuracy. - **Practical application**: Applying the improved BOA to the SVM prediction model of the tubular belt conveyor, the effectiveness of the model is verified through engineering data, with the prediction error controlled within 10%, demonstrating its application value in practical engineering. In summary, the paper improves the adaptability and search efficiency of the Butterfly Optimization Algorithm and successfully applies it to the parameter optimization of the Support Vector Machine, providing a reliable solution for practical engineering problems.

A modified butterfly optimization algorithm: An adaptive algorithm for global optimization and the support vector machine

Hybrid optimization algorithm for load model parameter identification

Butterfly optimization algorithm: a novel approach for global optimization

A modified butterfly optimization algorithm for mechanical design optimization problems

Optimization of support vector machine through the use of metaheuristic algorithms in forecasting TBM advance rate

A new monarch butterfly optimization with an improved crossover operator

An Enhanced Monarch Butterfly Optimization with Self-adaptive Butterfly Adjusting and Crossover Operators.

An enhanced hybrid seagull optimization algorithm with its application in engineering optimization

A Novel Monarch Butterfly Optimization with Greedy Strategy and Self-Adaptive

Butterfly Intelligent Optimization Algorithm Based on Good Point Set and Adaptive Weight Factor

Improved Butterfly Optimisation Algorithm Based on Guiding Weight and Population Restart

A Hybrid Butterfly Optimization Algorithm for Numerical Optimization Problems

An Adaptive Particle Swarm Algorithm for Unconstrained Global Optimization of Multimodal Functions

A Modified Sine Cosine Algorithm with Teacher Supervision Learning for Global Optimization.

Feature selection using binary monarch butterfly optimization

A Cuckoo Search-Support Vector Machine Model For Predicting Dynamic Measurement Errors Of Sensors

A Modified Sand Cat Swarm Optimization Algorithm Based on Multi-Strategy Fusion and Its Application in Engineering Problems

Support vector machine optimization via an improved elephant herding algorithm for motor energy efficiency rating

Slime Mould Algorithm with Horizontal Crossover and Adaptive Evolutionary Strategy: Performance Design for Engineering Problems

Gorilla optimization algorithm combining sine cosine and cauchy variations and its engineering applications

Hybrid Algorithm for Global Optimization Based on Periodic Selection Scheme in Engineering Computation