A Prediction Model of Forest Preliminary Precision Fertilization Based on Improved GRA-PSO-BP Neural Network

Table 5

Comparison of measured results with predicted results (kg/hm²).

Number	Stand age (t/a)	Site index S	Actual amount of nitrogen fertilizer	BP	Percentage of error (%)	PSO-BP	Percentage of error (%)	GRA-PSO-BP	Percentage of error (%)
1	5a	5	17.38	16.31	6.551	17.67	1.66	17.35	0.1752
2	18a	19	84.19	92.22	8.706	86.84	3.142	84.32	0.155
3	21a	22	218.1	200	9.082	217.3	0.3702	218.3	0.0666
4	23a	25	202.3	219	7.628	203.3	0.4784	201	0.6684
5	5a	5	14.94	16.32	8.416	15.08	0.9074	14.95	0.0589
6	18a	19	119.9	130.5	8.12	124.6	3.876	120	0.0650
7	21a	22	216.5	199.3	8.605	214.6	0.8908	216.4	0.0367
8	23a	25	207	222.1	6.79	209.9	1.402	209.1	1.02
1	5a	5	1.609	1.794	10.32	1.649	2.45	1.611	0.1245
2	18a	19	9.811	10.85	9.558	9.659	1.552	9.818	0.0637
3	21a	22	28.55	30.71	7.027	28.91	1.278	28.4	0.705
4	23a	25	39.58	43.12	8.206	39.75	0.4255	39.88	0.7471
5	5a	5	1.404	1.501	6.442	1.381	1.634	1.394	0.7023
6	18a	19	14.91	13.78	8.236	14.84	0.5151	14.92	0.0377
7	21a	22	24.94	27.45	9.125	24.9	0.1747	24.74	0.8051
8	23a	25	39.73	42.93	7.444	39.84	0.2619	39.45	0.6993
1	5a	5	7.41	6.744	9.874	7.397	0.1693	7.388	0.2956
2	18a	19	111	104.1	6.643	110.6	0.3362	110.2	0.664
3	21a	22	155.9	162	5.541	155.8	0.07406	155.5	0.3252
4	23a	25	237.9	253.8	6.257	241.6	1.557	235	1.212
5	5a	5	8.045	7.483	7.514	8.039	0.07832	8.035	0.1311
6	18a	19	87.88	93.75	6.267	91.21	3.791	87.9	0.0249
7	21a	22	172.8	183.4	5.767	172.6	0.08186	173	0.1017
8	23a	25	281.2	261.8	7.395	269.1	4.289	283.2	0.7244