Unmatched reliability for forecasting price, load, and generation
Hitachi Energy’s Forecaster is a true technical innovation representing a substantial leap forward in the ability to deliver more detailed, relevant, and accurate power generation, consumption, and pricing forecasts. Forecaster arms utilities and energy companies with data-driven insights that help them make better operational and financial decisions for short- and long-term planning to help mitigate risk and seize opportunity.
Based on advanced models, Forecaster can be easily configured to meet specific needs of any country or energy market. Predictions are derived using two methods: neural network technology for short-term forecasts and multiple linear regression techniques for longer time horizons.
Forecaster enables organizations to improve forecast accuracy by tracking forecasts with actual values and integrating new data as soon as it becomes available. The neural network algorithm becomes progressively smarter as it learns load and price variables triggered by weather events, time-of-day cycles, holidays and other historical data.
“A core part of Hitachi Energy’s score for innovation derives from its combination of operational analytics with analytics for finance and risk (bringing together load forecasting and trading, for example). This is relatively rare in the power space and provides tangible benefits.”
- Improved accuracy. Machine learning capabilities are continuously improved to deliver the most accurate forecast data.
- Neural network technology. Recognizes relationships / learned patterns and applies this knowledge to produce accurate daily and hourly/sub-hourly demand, generation and price forecasts.
- Sensitivity analysis and comprehensive statistics. Track changes in data for a better understanding of relevant inputs and more accurate capture of forecast peaks and valleys.
- Highly configurable. High degree of user configurability meets the need for flexibility and change.
- Automation and batch processing. Automation and batch processing capabilities are run across a variety of scheduling parameters and types of forecasting models that enables training and running forecasts with minimal intervention.
- Ease of Integration. Designed to easily integrate with external systems and common file types (e.g., flat file templates, MS Excel, APIs, and web services options) to streamline integration.
- Customizable reporting. Get the data the way you want it. Model training and forecast results can be displayed in grid, pivot and graph formats. Integration with Microsoft Power BI provides further reporting and analysis capabilities on a common platform.
- Exporting data. All data, including the inputs from customized forecast models, can be easily exported in different formats or it can be sent by email.
Submit your enquiry and we will contact you