Feed Error Prediction and Compensation of CNC Machine Tools Based on Whale Particle Swarm Backpropagation Neural Network

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The paper titled "Feed Error Prediction and Compensation of CNC Machine Tools Based on Whale Particle Swarm Backpropagation Neural Network" presents a novel approach to enhancing the accuracy of CNC (Computer Numerical Control) machine tools through advanced modeling and compensation techniques. This research aims to address the limitations of current feed error modeling methods under complex machining conditions, proposing a solution that significantly improves the predictive accuracy and effectiveness of actual feed error compensation.

Interesting Aspects

The study introduces an innovative combination of the Whale Particle Swarm Optimization (WPSO) algorithm with a Backpropagation Neural Network (BPNN) to optimize the network's weights and thresholds. This hybrid approach, which incorporates elements of nature-inspired optimization and machine learning, demonstrates a creative strategy to tackle the prediction and compensation of feed errors in CNC machine tools.

The proposed model achieves a high prediction accuracy. It significantly reduces the maximum backward and forward errors highlighting the effectiveness of the model in real-time compensation experiments.

Practical validation under different operating conditions showcases the model's robustness and potential for enhancing manufacturing processes.

Suggested improvements:

The computational complexity associated with the WPSO algorithm and BPNN might still pose challenges for real-time applications in some scenarios. Some discussion about this is needed.

The model's performance, while tested on specific machine tool configurations and conditions, may vary when applied to different types of CNC machines or under vastly different operating conditions. Some discussion about generalization of this approach is needed.

The dynamic and ever-evolving nature of machining environments, including wear and tear of machine components, changes in material properties, and varying ambient conditions, may affect the model's predictive accuracy over time. Ongoing adaptation and recalibration of the model could be necessary to maintain its effectiveness. What is the robustness level, some discussion insights about this needed.

Comments on the Quality of English Language

Consistency in Hyphenation: Technical terms and compound modifiers sometimes require hyphens for clarity (e.g., "feed error prediction" might be clearer when hyphenated as "feed-error prediction" if used as an adjective).

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

This study addresses the limitations of current machine tool feed error modelling methods, particularly in meeting the demands of high-precision machining under complex conditions. Introducing the Whale Particle Swarm Optimization (WPSO) algorithm to optimise the Backpropagation Neural Network (BPNN) enhances predictive accuracy and compensatory effectiveness. The integration of screw elongation and feed position as inputs further refine the feed error prediction model. Comparative analysis against other models, followed by real-time compensation experiments, demonstrates the superiority of the proposed model. The proposed model exhibits substantial improvement with an impressive accuracy of 93.12% and minimal errors ranging from -3.80 μm to 4.57 μm, averaging at -0.30 μm. Under diverse operational scenarios, significant reductions in both maximum and average errors signify the model's robustness and efficacy. Overall, this research underscores the potential of the WPSO-optimized BPNN model in advancing precision machining capabilities and optimising real-time error compensation strategies.

I have a couple of questions, the answers to which I think should be included in the text so that the paper is understandable for readers who do not deal with artificial intelligence.

1. How does the Whale Particle Swarm Optimization (WPSO) algorithm enhance the Backpropagation Neural Network (BPNN) in the context of machine tool feed error modelling?

2. What were the key findings regarding the accuracy and error range of the proposed prediction model?

3. Can you explain the significance of the observed reductions in maximum and average errors under different operating conditions?

4. What implications do the research results have for improving precision machining capabilities?

5. How might the WPSO-optimized BPNN model be further refined or applied in future research or industrial settings?

6. What challenges or limitations were encountered during the development and evaluation of the prediction model?

7. How does this study contribute to advancing understanding and techniques for addressing feed error in high-precision machining processes?

Author Response

Please see the attachment.

Author Response File: Author Response.pdf