Grey Model Optimized by Particle Swarm Optimization for Data Analysis and Application of Multi-Sensors

Area of Science:

• Engineering
• Data Science

Background:

• Pumping station operations require accurate temperature monitoring.
• Complex unit structures and coupled factors complicate temperature prediction.
• Limited data hinders effective operational analysis.

Purpose of the Study:

• To enhance the prediction accuracy of multivariable grey models for pumping station temperature.
• To investigate the impact of parameter optimization on model performance.
• To provide a reliable prediction method for complex systems with scarce data.

Main Methods:

• Utilized a multivariable grey system prediction model.
• Employed particle swarm optimization (PSO) to optimize the model's 'q' parameters.
• Processed multi-sensor temperature data from a pumping station unit.
• Compared results against unoptimized grey models and BP neural networks.

Main Results:

• Optimizing 'q' parameters with PSO significantly improved prediction accuracy.
• The optimized multivariable grey model demonstrated smaller relative errors.
• The method proved effective even with insufficient training data.

Conclusions:

• Particle swarm optimization enhances the predictive capability of multivariable grey models.
• The optimized model offers a superior approach for temperature trend analysis in pumping stations.
• This approach is valuable for complex industrial systems with data limitations.