ĐÁNH GIÁ CÁC MÔ HÌNH DỰ BÁO BỨC XẠ MẶT TRỜI: LIGHTGBM, LSTM VÀ GRU=EVALUATION OF SOLAR RADIATION FORECAST MODELS: LIGHTGBM, LSTM AND GRU

0 Người đánh giá. Xếp hạng trung bình 0

Tác giả: Thành Doanh Lê, Ngọc Trung Nguyễn, Tuấn Anh Nguyễn, Mạnh Hải Phạm, Trọng Tuấn Trần, Minh Pháp Vũ, Thị Anh Thơ Vũ

Ngôn ngữ: vie

Ký hiệu phân loại:

Thông tin xuất bản: Tạp chí Khoa học & Công nghệ - Trường Đại học Công nghiệp Hà Nội, 2024

Mô tả vật lý: tr.3

Bộ sưu tập: Metadata

ID: 482258

Thêm vào giỏ

Accurate solar radiation forecasting is a crucial technical factor in optimizing the performance of solar power systems. This paper evaluates the performance of three advanced machine learning models: Light Gradient Boosting Machine (LightGBM), Long Short-Term Memory (LSTM), and Gated Recurrent Unit (GRU) in solar radiation forecasting. The models were trained and tested on a real-world dataset, including meteorological parameters and measured solar radiation. Evaluation criteria such as Root Mean Square Error (RMSE) and Mean Absolute Error (MAE) were used to compare the models' performance. Results show that each model has its strengths and weaknesses. The LightGBM model demonstrated superior performance during training with faster training times and higher accuracy (0,5s
RMSE = 54.8 W/m2và MAE = 27.6 W/m2) so với LSTM (456,5s
RMSE = 59.2W/m2
MAE = 34.7W/m2) và GRU (397,2s
RMSE= 59.3W/m2
MAE = 34.7W/m2). The three models showed comparable accuracy in forecasting scenarios, but LightGBM had significantly lower prediction times than LSTM and GRU. LSTM and GRU, although more complex and requiring longer training times, also demonstrated good forecasting capabilities with complex time series data characteristics. The paper provides a comprehensive view of these models' performance and recommends selecting appropriate models in solar radiation forecasting applications.Dự báo bức xạ mặt trời chính xác là yếu tố kỹ thuật quan trọng trong việc tối ưu hóa hiệu suất các hệ thống điện mặt trời. Bài báo này đánh giá hiệu suất của ba mô hình học máy tiên tiến là Light Gradient Boosting Machine (LightGBM), Long Short-Term Memory (LSTM) và Gated Recurrent Unit (GRU) trong việc dự báo bức xạmặt trời. Các mô hình được huấn luyện và kiểm tra thử trên một bộ dữ liệu thực tế, bao gồm các thông số khí tượng và bức xạ mặt trời đo được. Các tiêu chí đánh giá như sai số trung bình bình phương gốc (RMSE), và sai số tuyệt đối trung bình (MAE) được sử dụng để so sánh hiệu suất của các mô hình. Kết quả cho thấy mỗi mô hình có những ưu điểm và nhược điểm riêng. Mô hình LightGBM khi huấn luyện thể hiện hiệu suất vượt trội với thời gian huấn luyện nhanh hơn và độ chính xác cao hơn (0,5s
RMSE = 54,8W/m2và MAE = 27,6W/m2) so với LSTM (456,5s
RMSE = 59,2W/m2
MAE = 34,8W/m2) và GRU (397,2s
RMSE = 59,3W/m2
MAE = 34,7W/m2). Trong các kịch bản dự báo, ba mô hình có độ chính xác khá tương đương nhưng LightGBM có thời gian dự báo thấp hơn LSTM và GRU khá nhiều. LSTM và GRU, mặc dù phức tạp hơn và đòi hỏi thời gian huấn luyện lâu hơn, nhưng cũng cho thấy khả năng dự báo tốt với những đặc điểm dữ liệu thời gian phức tạp. Bài báo cung cấp cái nhìn toàn diện về hiệu suất của các mô hình này và đưa ra các khuyến nghị cho việc lựa chọn mô hình phù hợp trong các ứng dụng dự báo bức xạ mặt trời.

1. Đọc trực tuyến

1. Đọc trực tuyến

Tạo bộ sưu tập với mã QR