Photovoltaic Generation Forecast: Model Training and Adversarial Attack Aspects
Forecasting photovoltaic (PV) power generation, as in many other time series scenarios, is a challenging task. Most current solutions for time series forecasting are grounded on Machine Learning (ML) algorithms, which usually outperform statistical-based
- PDF / 784,863 Bytes
- 16 Pages / 439.37 x 666.142 pts Page_size
- 78 Downloads / 201 Views
Computer Science Department, Londrina State University, Londrina, PR, Brazil {evertonsantana,brunozarpelao,barbon}@uel.br 2 Electrical Engineering Department, Londrina State University, Londrina, PR, Brazil [email protected]
Abstract. Forecasting photovoltaic (PV) power generation, as in many other time series scenarios, is a challenging task. Most current solutions for time series forecasting are grounded on Machine Learning (ML) algorithms, which usually outperform statistical-based methods. However, solutions based on ML and, more recently, Deep Learning (DL) have been found vulnerable to adversarial attacks throughout their execution. With this in mind, in this work we explore four time series analysis techniques, namely Naive, a baseline technique for time series, Autoregressive Integrated Moving Average (ARIMA), from the statistical field, and Long Short-term Memory (LSTM) and Temporal Convolutional Network (TCN), from the DL family. These techniques are used to forecast the power generation of a PV power plant 15 minutes and 24 hours ahead, having as input only power generation historical data. Two main aspects were analyzed: i) how training size influenced the performance of the forecasting models and ii) how univariate time series data could be modified by an adversarial attack to decrease models’ performance through cross-technique transferability. For i), the mentioned methods were used and evaluated with monthly updates. For ii), Fast Gradient Sign Method (FGSM), along with a logistic regression substitute model and past data, were used to perform attacks against DL models at test time. LSTM and TCN decreased the error as the training sample size increased and outperformed Naive and ARIMA models. Adversarial samples were able to reduce the performance of LSTM and TCN, particularly for short-term forecasts. Keywords: Time series forecast · Solar photovoltaic generation Deep learning · Adversarial attack · Smart grid
·
The authors would like to thank the financial support of Coordination for the Improvement of Higher Education Personnel (CAPES) - Finance Code 001 -, the National Council for Scientific and Technological Development (CNPq) of Brazil - Grant of Project 420562/2018-4, and Funda¸ca ˜o Arauc´ aria. c Springer Nature Switzerland AG 2020 R. Cerri and R. C. Prati (Eds.): BRACIS 2020, LNAI 12320, pp. 634–649, 2020. https://doi.org/10.1007/978-3-030-61380-8_43
PV Generation Forecast: Model Training and Adversarial Attack Aspects
1
635
Introduction
Intelligent generation and distribution of electrical energy are beneficial for systems operators, plant managers and consumers [2]. A key aspect in this process is the accurate forecast of produced energy, which is fundamental to enable the integration of several plants to the grid, save costs, make power grids more reliable amid the variation in the demand, avoid power outage, and prevent plant managers from penalties. It is also advantageous for the sake of the environment [10], particularly when renewable sources are employed. For photovoltaic (PV) gene
Data Loading...
