International Journal of Business and Management Technology in Society 2025, 3.
, 38-41 International Journal of Business and Management Technology in Society Available at https://journal.
id/index.
php/ijbmts/ Daily Power Plant Operation Prediction Using Adaptive Filter Based on Wavelet Symlet Milla Kartikasari1.
Nida Ul Hasanah2, and Widya Yunita3
1,2,3
Institut Teknologi Perusahaan Listrik Negara Article History Received : 2025-06-13 Revised : 2025-08-01 Accepted : 2026-01-09 Published : 2026-01-13 Keywords:
Adaptive filter.
NLMS, operation prediction, power plant, symlet.
Corresponding author:
widyayunita@gmail.
Paper type:
Research paper Cite this article:
Kartikasari.
Hasanah.
, &
Yunita.
Daily Power Plant Operation Prediction Using Adaptive Filter Based on Wavelet Symlet.
International Journal of Business and Management Technology in Society.
International Journal of Business and Management Technology in Society, 3.
, 38-41.
Abstract Purpose Ae This study aims to develop an accurate method for predicting the daily operation of power plants to support optimal scheduling of generation and maintenance activities.
Methodology Ae An adaptive filter based on wavelet symlet .
is applied using the Normalized Least Mean Square (NMLS) algorithm.
The model adjusts its coefficients dunamically based on historical operational data to minimize prediction error.
Findings Ae The method was tested on Indonesian power plant operation data and achieved a Mean Square Error (MSE) of 0.
Segment-based evaluation confirmed the modelAos ability to provide consistent prediction accuracy across different time frames.
Originality Ae This research introduces a novel approach by combining wavelet symlet and adaptive filtering in the context of power plant operation prediction, which allows accurate forecasting using limited data.
Research limitations Ae The study focuses on short-term prediction .
p to 3 days ahea.
and does not include external influencing factors such as weather or system demand.
Only the NLMS algorithm was utilized, without comparison to other adaptive Practical implications Ae The proposed method enables operators to generate more accurate and reliable schedules, improving overall system performance and reducing outage risks.
Social implications Ae Enhancing the reliability of power plant operations contributes to a more stable electiricty supply, indirectly supporting public services and economic activities.
Introduction Electric power systems involve generating electricity to generate electrical energy.
Power plant operations are one of the important things to plan carefully so that the electric power system can operate optimally and reduce the risk of disruptions.
One of the important factors in plant operation planning is the prediction of the plant's daily operations Marsudi .
Prediction of daily operation of a plant is the process of estimating the value of plant operation variables such as output power, fuel consumption, and efficiency in the coming day.
ISSN 3025-4256
Copyright @2025 Authors.
This is an open-access article distributed under the terms of the Creative Commons Attribution License .
ttp://creativecommons.
org/licences/by-sa/4.
Daily Power Plant A Prediction of the daily operation of the plant can be done by various methods.
In this study, we propose a method of predicting the daily operation of adaptive filter-based power plants wavelet symlet (Putra, 2008.
Mallat, 1.
This method uses adaptive filtering to adjust to the historical data patterns of plant operations.
Research Methods Adaplet (Adaptif filter based on wavele.
Adaplet (Adaptive filter based on wavele.
is an adaptive filter where the initial coefficient used is wavelet (Douglass & Mathews, 1.
The adaptive filter is shown Figure 1, the adaptive filter works using e.
, the difference from the output y.
and the expected output d.
, for the algorithm that will affect the adaptive filter so that the coefficient changes over time.
Figure 1.
Adaptive Mat The prediction process uses adaptive filtering with an adaptive algorithm used, namely NLMS with the equation used to update the coefficient as follows (Haykin, 2.
LMS
ea ya y.
= ya C .
yyAya.
yeI O .
NLMS ya C.
ea y.
= ya C .
ya C.
ea y.
= ya C .
ya C.
ea y.
= ya C .
ya |.
|| yyA |.
|| yyA yeC |.
|| ya ya.
yeI O .
ya ya.
yeI O .
ya ya.
yeI O .
In Figure 1, the daily operating data signal is fed into an adaptive filter where the initial wavelet coefficient will then produce an output y.
The output y.
is compared to the expected output d.
The difference between the two results in an error value e.
The algorithm is used to calculate the value of the new coefficient based on the error value of e.
so that it affects the adaptive filter and the value of e.
is getting smaller and closer to zero.
The general equation of filtering Finite Impluse Response (FIR) with many taps is as follows:
= OcycAOe1 ycn=0 ycycu .
cu Oe yc.
= ycycu .
cu Oe .
U ycycu .
cA Oe .
cu Oe ycA Oe .
= .
From equation .
it is implemented into the program to predict N the next day with the program Results and Discussion Segmentation Analysis The prediction uses daily operation data of the plant in the form of data on the realization of the plant load for 221 days.
From 221 days, it is divided into 3 segments to obtain daily prediction data for the next 3 days of plant operations in each segment.
From the prediction of each segment, the average error value or MSE is obtained by comparing the original data and prediction data.
Segment 1 : Prediction of Daily Operation of Plant N 3 day from data
DATA ASLI
PREDIKSI
Sym 1 Sym 2 Sym 3 Sym 4 Sym 5 Segment 2 : Prediction of Daily Operations of Plant Day N 3 from data Sym 1 Sym 2 Sym 3 Sym 4 Sym 5 Segment 2 : Prediction of Daily Operations of Plant Day N 3 from data Sym 1 Sym 2 Sym 3 Sym 4 Sym 5 Figure 2.
Prediction results in each segment of the plant's daily operation data From the results of segmentation analysis, prediction data and the average value of errors were obtained as follows.
Tabel 1.
Mean Square Error Wavelet Symlet
Orde
MSE Segment 1
0,063
0,061
0,06
0,075
0,064
0,071
0,071
0,07
0,098
0,093
MSE Segment 2
MSE Segment 3
70 3 141 1 141 2 141 3 221 1 221 2 221 3
0,081 0,063 0,071 0,081 0,063 0,071 0,081
0,079 0,061 0,071 0,079 0,061 0,071 0,079
0,079 0,06
0,07 0,079 0,07 0,079 0,070
0,105 0,075 0,098 0,105 0,075 0,098 0,105
0,109 0,064 0,093 0,109 0,064 0,093 0,109
MSE Average Average 0,072 0,070 0,071 0,093 0,089 0,079 Daily Power Plant A Conclusion and Suggestion Symlet-based wavelet-based adaptive filter is a method that can be used to predict the daily operation of the plant with a mean square error value of 0.
This method can provide accurate predictions using limited data.
Henceforth, the prediction of the daily operation of the plant can be done by increasing the length of the coefficient of the parent wavelet order used or using the RLS (Recursive Least Squar.
filter implementation.
References