Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

Model updated

...

22nd July 2024

Summary

The retraining process resulted in two new models, one for PD and one for 7day price forecast. After experimentation we retrained the price forecasting models using existing features and hyper parameter settings. The increase in accuracy (using Absolute Mean Error, AME) was approximately 4%5%. (see validation results table below comparing the previous models with the new models).

...

Both PD and 7day price forecast models were trained on data between 01-0407-2021 and 08-0407-2024.

The market suspension period and a period after the Callide explosion were removed from the training data.

...

The model was validated by predicting the price between 09-0407-2024 and 3021-0407-2024.

New features (input data) exceptions

...

6th May

Predispatch model

p7day model

Model_id

20240503_090016_CB_2

20240503_085256_CB_7D_2

Previous model Latest release date

5th February 2024

5th February 2024

Latest release date (after retraining)

6th May 2024

(after retraining)

26th July 2024

Previous model release date

26th July 2024

26th July 2024

Machine Learning Algorithm

Ensemble predictor: gradient boosting

Ensemble predictor: gradient boosting

Training Period

1st A2021 July 2021 to 8th April July 2024

1st April July 2021 to 8th April July 2024

Hyper-parameter type

A

A

Hyper-parameter notes

No weighting across time periods

No weighting across time periods

Target

Dispatchprice (30min resolution)

Dispatchprice (30min resolution)

Input (and training) variables

PREDISPATCHPRICE

rrp

DISPATCHPRICE

last 3 day average actual RRP for same time

PREDISPATCHREGIONSUM (intervention = 0)

hours_out

time_of_day

day_type (weekday_or_weekend)

dispatchableload

netinterchange

(availablegeneration - uigf)

(dispatchablegeneration - scheduled_clearedmw)

ss_solar_clearedmw

ss_solar_uigf

ss_wind_clearedmw

ss_wind_uigf

totaldemand

totalintermittentgeneration

STPASA_REGIONSOLUTION

hours_out

time_of_day

day_type (weekday_or_weekend)

aggregatepasaavailability

aggregatescheduledload

netinterchangeunderscarcity

aggregatecapacityavailable

ss_solar_cleared

ss_wind_uigf

ss_wind_cleared

ss_solar_uigf

demand10

demand50

totalintermittentgeneration

demand_and_nonschedgen

...

Metric/

Region

Actual Mean Price

Mean Prediction New Model

Mean Prediction Previous Model

Absolute Mean Error New Model

Absolute Mean Error Previous Model

Average Error New Model

Average Error Previous Model

NSW1

98124.6556

105146.3864

106148.4464

2741.5351

2844.5601

622.7308

724.7908

QLD1

88111.5298

99135.0552

100138.3238

2649.5909

2752.4939

1023.54

1126.8141

SA1

89178.8734

86172.4826

83177.9712

3192.390531

93.1388

-36.3907

-51.9021

TAS1

81196.3272

82151.9568

86151.90

2962.1424

3263.9975

1-45.6304

5-44.5882

VIC1

92159.6629

84144.5144

82146.9875

2841.6498

2942.1436

-814.1485

-912.6855

7 Day Results

Metric/

Region

Actual Mean Price

Mean Prediction New Model

Mean Prediction Previous Model

Absolute Mean Error New Model

Absolute Mean Error Previous Model

Average Error New Model

Average Error Previous Previous Model

NSW1

94124.6633

88149.8382

82157.2016

2925.9249

3132.7184

-558.8299-12

60.4547

QLD1

83110.5400

80124.4043

73143.4161

2914.4243

3033.3561

-366.1491-10

70.1335

SA1

86187.5049

110423.7172

91410.7566

58236.9523

51223.4817

24304.2011

5317.2539

TAS1

80198.5677

71111.6508

70112.3795

31-87.7069

35-85.3082-8

93.9136

-1095.1967

VIC1

89161.894178

158.8990

68161.3917

37-2.2051

39-0.7524-11

61.0076

-2164.5003

Info

Variables that imply binding interconnectors (such as limits and marginal values) are currently not included as features.

...