Optimization of Process Control Parameters for Fully Mechanized Mining Face Based on ANN and GA

Area of Science:

• Mining Engineering
• Artificial Intelligence
• Optimization Theory

Background:

• Traditional process control in mechanized mining relies on experience, leading to inefficiencies and long adjustment cycles.
• Lack of scientific basis in traditional methods causes significant production losses.
• Improving mine production efficiency is crucial for economic viability.

Purpose of the Study:

• To develop a scientific and efficient method for optimizing process control parameters in fully mechanized mining faces.
• To overcome the limitations of experience-based parameter determination.
• To enhance overall mine production efficiency.

Main Methods:

• A hybrid optimization strategy combining artificial neural networks (ANN) and genetic algorithms (GA).
• Utilizing a cross-entropy cost function to optimize the ANN, improving learning speed and accuracy.
• Establishing a mixed-strategy optimization model with specific control parameters (shearer speed, support speed, conveyor speeds, pump pressures) as variables and coal output per minute as the objective.

Main Results:

• The proposed method demonstrates high accuracy in optimizing process control parameters.
• The optimization process time is significantly reduced compared to traditional methods.
• Effective improvement in the production efficiency of the mining face.

Conclusions:

• The hybrid ANN-GA method provides a scientifically sound and efficient approach to process control optimization in mechanized mining.
• This method effectively addresses the limitations of traditional, experience-based techniques.
• The optimized parameters lead to tangible improvements in mining productivity.