AI-POWERED ACADEMIC INSIGHTS
Evaluation of Course Achievement for Electrical and Control Systems of Electric Multiple Units Based on Improved Back Propagation Neural Network
This research introduces an intelligent evaluation model for course achievement in 'Electrical and Control Systems of Electric Multiple Units' using an improved Back Propagation Neural Network (BPNN) optimized by the Improved Tornado Optimization with Coriolis force (ITOC) algorithm. The study establishes quality standards, collects teaching data, and demonstrates that the ITOC-BPNN model significantly outperforms the standard BPNN in prediction accuracy for course objectives. Key recommendations for teaching improvements are also provided, focusing on project-based learning, simulation competitions, and tiered digital resources.
Executive Impact
The ITOC-BPNN model demonstrates superior performance in course achievement prediction.
0
RMSE Reduction (O1)
0
MAE Reduction (O1)
0
R² Improvement (O1)
Deep Analysis & Enterprise Applications
Select a topic to dive deeper, then explore the specific findings from the research, rebuilt as interactive, enterprise-focused modules.
Introduction
Methodology
Results
Recommendations
Currently, electric multiple units (EMUs) have become the mainstay of China's railway transportation. The electrical and control system serves as the 'neural center' and 'power heart' of EMUs. This course is a distinctive offering in Electrical Engineering and Automation program, focusing on domain-specific concentration and systematic integration, moving beyond traditional isolated subjects. The need for a tailored quality standard system for evaluation is paramount to assess teaching effectiveness and talent cultivation quality.
The proposed method leverages an Improved Tornado Optimization with Coriolis force (ITOC) algorithm to enhance the Back Propagation Neural Network (BPNN). This ITOC-BPNN model aims to improve the predictive accuracy of course achievement evaluation. The optimization involves a combined Gaussian initialization strategy, an adaptive Coriolis force adjustment, and a multi-stage search strategy to balance global exploration and local exploitation, addressing BPNN's limitations in convergence speed and stability.
Comparative analysis shows that the ITOC-BPNN model significantly outperforms the standard BPNN. For objective O1, RMSE and MAE were reduced by 76.56% and 78.41% respectively, with an R² improvement of 33.52%. Similar improvements were observed for O2 and O3, indicating enhanced rationality of learning rate configuration and increased stability and accuracy of the predictive model. The model accurately aligns predicted values with true target values for course attainment.
Based on the assessment, key recommendations include advancing project-based teaching with deep integration of simulation tools like MATLAB/Simulink and Python. Establishing practice platforms for simulation training competitions, developing tiered digital resource libraries for self-directed learning, and enhancing instructors' practical training abilities are also crucial to align teaching with industry advancements.
76.56%
Reduction in RMSE for Objective O1 with ITOC-BPNN
Enterprise Process Flow
Establish Course Quality Standards
→
Collect Teaching Data
→
Propose ITOC Algorithm
→
Enhance BPNN with ITOC
→
Develop Intelligent Evaluation Method
→
Suggest Teaching Improvements
| Metric | Standard BPNN | ITOC-BPNN |
|---|---|---|
| RMSE (O1) | 0.0465 | 0.0109 |
| MAE (O1) | 0.0403 | 0.0087 |
| R² (O1) | 0.7382 | 0.9857 |
| RMSE (O2) | 0.0435 | 0.0294 |
| MAE (O2) | 0.0356 | 0.0256 |
| R² (O2) | 0.8396 | 0.9516 |
Application in Dalian Jiaotong University
The 'Electrical and Control Systems of EMUs' course at Dalian Jiaotong University serves as the primary subject for this research. The intelligent evaluation model was applied to 120 students' data from the 2024-2025 academic year. The results highlighted areas where students excelled, such as optimizing key system components, and identified areas for improvement, particularly in proficiency with electrical engineering software tools. This practical application demonstrates the model's effectiveness in providing actionable insights for curriculum development and teaching reforms, directly supporting the university's goal of high-quality talent cultivation in rail transit.
Estimate Your Potential AI Impact
Quantify the potential efficiency gains and cost savings for your enterprise with advanced AI solutions.
Annual Cost Savings
$0
Hours Reclaimed Annually
0
Your AI Implementation Roadmap
A phased approach to integrate advanced AI into your academic evaluation processes.
Phase 1: Foundation & Data Collection (2-4 Weeks)
Establish clear course quality standards aligned with industry needs. Collect comprehensive teaching data, including theoretical and practical assessment scores, class participation, and student feedback.
Phase 2: Model Development & Training (4-6 Weeks)
Implement the ITOC algorithm to optimize BPNN hyperparameters. Train the ITOC-BPNN model using collected data to predict course achievement objectives accurately.
Phase 3: Validation & Analysis (2-3 Weeks)
Conduct comparative analysis against standard BPNN. Validate model's predictive accuracy and identify specific areas for curriculum improvement based on assessment results.
Phase 4: Implementation of Teaching Reforms (Ongoing)
Integrate project-based learning and simulation tools. Establish practice platforms and develop tiered digital resources. Provide ongoing training for instructors to align teaching with industry advancements.
Ready to Transform Your Course Evaluations?
Unlock the full potential of AI for precise and intelligent course achievement assessment. Discuss your unique needs with our experts.
Ready to Get Started?
Book Your Free Consultation.
Let's Discuss Your AI Strategy!
Lets Discuss Your Needs
Select a Date
Sun
Mon
Tue
Wed
Thu
Fri
Sat