3rd ICAI 2024

International Conference on Automotive Industry 2024

Mladá Boleslav, Czech Republic

Figure 8: SOURCE_1 average load prediction with confidence interval

Source: Own elaboration

Figure 9: SOURCE_2 average load prediction with confidence interval

Source: Own elaboration

4. Conclusion The presented solution is in implementation at AUTOPART battery plant located in Poland. Integrating the proposed solution with planning processes can reduce the variance of power demand over time and reduce costs related to electricity distribution fees. Currently, based on the research carried out, the potential for reducing the difference between the maximum value of power used and the average value in individual months by nearly 47% has been defined. However, this value requires linear power use and is the maximum value. It indicates a huge optimization potential in the field of ordering active power and creates space for further research in the field of production planning schemes considering energy efficiency. The possibility of using the models described in this publication without the need for high-performance computing equipment provides a space for implementation in the scheduling of operations in companies with varying energy consumption in their manufacturing processes. However, it should be noted that it may be necessary to use blended models to achieve the highest quality solution, as discussed in this paper.

91