An Optimized Forecasting Approach Based on Grey Theory and Cuckoo Search Algorithm: A Case Study for Electricity Consumption in New South Wales
Table 1
Electricity consumption forecasting results.
Time-point
Actual value (MWh)
GM
IAGM
CSGM
ARIMA (2,2,1)
Forecasting values
Errora (%)
Forecasting values
Error (%)
Forecasting values
Error (%)
Forecasting values
Error (%)
0:00
4330.270
4397.628
1.56
4397.713
1.56
4411.040
1.87
4299.163
0.72
0:30
4209.265
4276.840
1.61
4276.919
1.61
4280.632
1.70
4249.257
0.95
1:00
4114.000
4185.289
1.73
4185.387
1.74
4202.551
2.15
4025.156
2.16
1:30
3982.655
4040.614
1.46
4040.685
1.46
4052.158
1.75
3985.292
0.07
2:00
3853.080
3905.699
1.37
3905.752
1.37
3904.051
1.32
3849.782
0.09
2:30
3668.835
3735.070
1.81
3735.142
1.81
3734.825
1.80
3722.507
1.46
3:00
3550.665
3614.975
1.81
3615.062
1.81
3634.288
2.36
3591.899
1.16
3:30
3443.545
3489.662
1.34
3489.750
1.34
3475.360
0.92
3431.846
0.34
4:00
3350.500
3413.786
1.89
3413.850
1.89
3407.637
1.71
3397.992
1.42
4:30
3313.335
3390.756
2.34
3390.803
2.34
3382.276
2.08
3264.648
1.47
5:00
3373.360
3429.484
1.66
3429.524
1.66
3427.320
1.60
3137.584
5:30
3530.420
3633.135
2.91
3633.220
2.91
3628.695
2.78
3347.417
6:00
3730.750
3832.116
2.72
3832.177
2.72
3836.934
2.85
3478.720
6:30
4115.280
4236.704
2.95
4236.766
2.95
4220.757
2.56
3765.914
7:00
4502.655
4647.155
3.21
4647.238
3.21
4640.427
3.06
4554.320
1.15
7:30
4712.280
4897.057
3.92
4897.194
3.92
4888.935
3.75
4626.589
1.82
8:00
4941.845
5158.762
4.39
5158.978
4.39
5158.713
4.39
4772.266
3.43
8:30
4969.065
4995.615
0.53
4995.667
0.54
4995.996
0.54
5126.765
3.17
9:00
4915.090
4977.279
1.27
4977.374
1.27
5011.611
1.96
5026.143
2.26
9:30
4874.630
4934.493
1.23
4934.557
1.23
4944.742
1.44
4898.405
0.49
10:00
4845.960
4923.899
1.61
4923.984
1.61
4902.575
1.17
4831.506
0.30
10:30
4810.585
4879.438
1.43
4879.522
1.43
4846.219
0.74
4820.355
0.20
11:00
4758.920
4806.381
1.00
4806.484
1.00
4752.485
0.14
4789.164
0.64
11:30
4665.005
4708.921
0.94
4709.015
0.94
4635.415
0.63
4706.608
0.89
12:00
4615.740
4663.379
1.03
4663.437
1.03
4605.949
0.21
4588.729
0.59
12:30
4560.465
4624.130
1.40
4624.185
1.40
4559.143
0.03
4560.645
0.00
13:00
4522.080
4569.492
1.05
4569.526
1.05
4503.554
0.41
4510.220
0.26
13:30
4510.240
4521.704
0.25
4521.722
0.25
4471.516
0.86
4483.658
0.59
14:00
4486.135
4516.147
0.67
4516.169
0.67
4470.371
0.35
4496.701
0.24
14:30
4478.710
4535.532
1.27
4535.573
1.27
4484.597
0.13
4464.722
0.31
15:00
4461.810
4528.393
1.49
4528.434
1.49
4479.814
0.40
4472.497
0.24
15:30
4437.840
4554.719
2.63
4554.768
2.63
4510.697
1.64
4446.137
0.19
16:00
4541.070
4643.533
2.26
4643.599
2.26
4610.473
1.53
4413.194
2.82
16:30
4614.100
4752.256
2.99
4752.335
3.00
4723.989
2.38
4665.412
1.11
17:00
4799.055
4956.608
3.28
4956.701
3.28
4939.399
2.92
4708.471
1.89
17:30
5122.475
5251.827
2.53
5251.924
2.53
5255.480
2.60
4878.819
4.76
18:00
5375.885
5446.141
1.31
5446.216
1.31
5465.649
1.67
5413.701
0.70
18:30
5350.900
5411.714
1.14
5411.789
1.14
5447.303
1.80
5761.367
19:00
5269.615
5348.454
1.50
5348.550
1.50
5376.871
2.04
5685.330
19:30
5119.510
5225.517
2.07
5225.610
2.07
5252.610
2.60
5264.521
2.83
20:00
4982.405
5131.309
2.99
5131.410
2.99
5158.740
3.54
5045.731
1.27
20:30
4905.600
5057.356
3.09
5057.450
3.10
5087.582
3.71
4846.094
1.21
21:00
4809.135
4999.369
3.96
4999.489
3.96
5036.191
4.72
4807.723
0.03
21:30
4715.035
4883.033
3.56
4883.162
3.57
4918.670
4.32
4798.520
1.77
22:00
4558.515
4757.412
4.36
4757.540
4.37
4781.606
4.89
4681.545
2.70
22:30
4602.065
4744.022
3.08
4744.124
3.09
4759.853
3.43
4430.017
3.74
23:00
4489.415
4641.814
3.39
4641.920
3.40
4655.801
3.71
4651.653
3.61
23:30
4397.565
4573.946
4.01
4574.034
4.01
4575.651
4.05
4398.089
0.01
Maximum forecasting error (%)
4.39
4.39
4.89
8.49
Average forecasting error (%)
2.12
2.13
2.07
2.04
The error is defined as follows: error = /actual value * 100%.
bThe forecasting error value is greater than 5%. The specific time-points are 5:00, 5:30, 6:00, 6:30, 18:30, and 19:00.