Probabilistic forecast of electric vehicle charging demand: analysis of different aggregation levels and energy procurement

Energy Informatics

Table 6 Metrics of the models for the TZO zones

TZO zone	Model	nRMSE*	MASE*	R2**	PS low Q***	PS high Q***	IS***
Tennet	LinR	0.883	0.806	134.983	0.793
	Bagging	0.534	0.431	81.516	0.924	1.955	4.259
	GradientB	0.536	0.523	81.820	0.924	3.887	4.527
	Ada	0.467	0.414	71.388	0.942	1.945	4.142
	Random forest	0.545	0.456	83.233	0.921	3.650	6.052
	LSTM	0.493	0.513	75.266	0.936	34.397	69.342
	CNN	0.589	0.519	89.997	0.908	38.683	76.597
	NN	0.587	0.660	89.752	0.908	141.111	89.672
50Hertz	LinR	0.825	0.860	91.940	0.787
	Bagging	0.648	0.662	72.269	0.869	2.782	3.765
	GradientB	0.628	0.686	70.013	0.877	3.481	4.488
	Ada	0.623	0.659	69.436	0.879	2.498	3.808
	Random forest	0.664	0.669	74.091	0.862	5.227	6.575
	LSTM	0.654	0.721	72.922	0.866	97.063	101.574
	CNN	0.788	0.838	87.828	0.806	88.751	92.326
	NN	0.734	0.791	81.815	0.832	115.450	108.417
Amprion	LinR	0.914	0.904	122.239	0.780
	Bagging	0.427	0.473	57.084	0.952	2.115	2.819
	GradientB	0.525	0.599	70.274	0.927	3.555	4.009
	Ada	0.417	0.473	55.783	0.954	2.248	3.034
	Random forest	0.447	0.507	59.807	0.947	3.796	4.939
	LSTM	0.584	0.616	78.181	0.910	46.069	55.681
	CNN	0.718	0.683	96.105	0.864	61.516	60.636
	NN	0.561	0.652	75.076	0.917	117.008	91.710
TransnetBW	LinR	0.845	0.768	173.191	0.768
	Bagging	0.568	0.467	116.333	0.895	3.571	5.436
	GradientB	0.533	0.514	109.334	0.908	4.468	6.936
	Ada	0.579	0.470	118.761	0.891	3.123	5.592
	Random forest	0.555	0.477	113.681	0.900	6.029	9.262
	LSTM	0.574	0.611	117.723	0.893	52.813	111.074
	CNN	0.634	0.600	129.904	0.869	63.730	108.037
	NN	0.643	0.673	131.848	0.866	188.707	125.658