Skip to main content

Compute Terminal

Compute Terminal is a powerful tasks engine with a web interface and RESTful API designed specifically for energy community administrators, providing a secure and efficient environment for running advanced scripted tasks, from billing calculations to unsupervised data discovery. This platform deploys AWS EC2 instances on the fly, enabling the execution of complex, resource-intensive models essential for modern energy management.

Key Features:​

  1. Automated Task Execution:

    • Compute Terminal’s task engine interface allows administrators to automate and manage Ansible playbooks effortlessly, streamlining complex energy management tasks with minimal manual intervention.

  2. Comprehensive Energy Management Tools:

    • The platform comes with pre-configured playbooks to perform critical energy simulations and calculations, such as:
      • Optimal Scheduling of Energy Storage: Calculate the best times for charging and discharging based on predictive models and real-time data.
      • Load-Flow Modeling with Artificial Intelligence: Run load-flow simulations using a Recurrent Neural Network (RNN) model to predict and optimize power flow, ensuring stability and efficiency throughout the grid.
      • Voltage Regulation: Implement AI-driven real-time voltage regulation to maintain system balance and minimize disruptions.

  3. Tariff and Financial Simulations:

    • Conduct extensive tariff simulations to analyze different pricing models and calculate the Return on Investment (ROI) for various energy projects. This supports strategic financial planning and decision-making for the community.
    • Issue accurate bills through automated billing processes, ensuring transparency and consistency.

  4. High Performance and Scalability:

    • Powered by AWS EC2 instances, Compute Terminal efficiently handles even the most demanding AI models and simulations. It dynamically scales to meet the computational needs of each task, providing flexibility for both real-time and scheduled processing.

  5. User-Friendly Interface:

    • Compute Terminal’s web interface is intuitive, enabling administrators to schedule, monitor, and manage playbooks with ease. Designed for accessibility, the interface caters to users with diverse technical backgrounds.

Compute Terminal revolutionizes energy management for community administrators, offering an all-in-one platform that leverages artificial intelligence and automation for optimized energy resource management and financial analysis, ensuring a scalable, efficient, and user-friendly solution.

SageMaker​

Administrators can also utilize Amazon’s Unsupervised Built-in SageMaker Algorithms, e.g.:

  • K-Means Algorithm—finds discrete groupings within data, where members of a group are as similar as possible to one another and as different as possible from members of other groups.
  • Principal Component Analysis (PCA) Algorithm—reduces the dimensionality (number of features) within a dataset by projecting data points onto the first few principal components. The objective is to retain as much information or variation as possible. For mathematicians, principal components are eigenvectors of the data’s covariance matrix.
  • Random Cut Forest (RCF) Algorithm—detects anomalous data points within a data set that diverge from otherwise well-structured or patterned data.

Example: RCF Algorithm

Data StreamRCF AlgorithmRandom TreesAnomaly Score CalculatorStream data pointInsert data point into treesRandomly cut feature spaceUpdate forest structureloop[Random Forest]Pass data point informationCalculate anomaly scoreFlag as anomalyNo anomalyalt[Anomaly Detected][Normal Behavior]Data StreamRCF AlgorithmRandom TreesAnomaly Score Calculator

Explanation

  • Data Stream: Represents the continuous stream of incoming data points.
  • RCF Algorithm: Manages the processing of data, including insertion into trees and anomaly score calculation.
  • Random Trees: Multiple trees (forest) are built using random cuts in the feature space to partition the data.
  • Anomaly Score Calculator: Computes the anomaly score based on how “isolated” a data point is within the trees.
  • Outcome: Depending on the anomaly score, the algorithm flags data points as anomalies or normal behavior.

This diagram provides a high-level overview of how RCF processes incoming data and determines anomalies. Adjust the structure based on specific details or requirements if needed.