Meta-Stability of Pulsed Load Microgrids

Sandia National Labs

Statement of Work
NAVSEA/ Military microgrids
Using the HSSPFC (Hamiltonian Surface Shaping and Power Flow Control) derived MATLAB/Simulink
tools develop a Reduced Order Model (ROM) to support control designs for pulse load applications for i)
up to (3) key ship modes of a ship power system operation and ii) a stable and unstable modes of
switching operations as a part of a survivability scenario.
Deliverables Tasks:
1. Provide ROM of meta-stable ship system.
2. Analyses and control design (feedforward and feedback) of meta-stable system.
3. Analyses and control design for multi-pulse load systems.
4. Analyses of the effects and potential benefits of non-linear magnetics in meta-stable system.
5. Develop and perform hardware testing on metastable laboratory benchtop system.
6. Develop networked Microgrid model for KIER/LUXCO scenario

Investigator: Wayne Weaver

Distributed and Decentralized Control of Aircraft Energy Systems

U.S. Dept of Defense, Air Force Research Lab -InfoSciTex Corp (AFRL)

Aircraft energy system components, including sources, loads and distribution, have multiple commitments and responsibilities. Often much of the system is comprised of power electronic converters for sources, loads, and energy storage (chemical, mechanical and thermal). For example, a point of load power converter has the commitment to serve the energy needs of the end load. However, if the power system collapses, the needs of the load cannot be met. Therefore, it is also in the interest of the conve1ter to contribute to the global stability of the power system by reducing nonlinear dynamics and incremental negative impedance. One method to mitigate the destabilizing effects of constant power loads is the power buffer concept. A power buffer is a device that mitigates a destabilizing event by presenting controlled impedance to the supply during the transient while local energy is used to maintain constant power to the load until the system can recover. A power buffer may include additional hardware, or may merely be a modification of the controls of an existing active front end power converter. However to date the use of a load as an energy asset in a power buffer has been limited to traditional chemical (capacitor and battery) storage devices in the electrical network. Next generation aircraft may have a broad range of potential assets in the form of loads, including inertial spinning devices and thermal systems, which could be utilized in the overall energy strategy.

Research with AFRL researchers to investigate distributed and decentralized control of aircraft energy systems. This effort will include using models and simulations to formulate decentralized control and study the effects. Specifically,
• Develop and document a mathematical model of the aircraft energy systems including thermal and inertial loads.
• Formulate a decentralized power buffer control including inertial and thermal loads as energy storage assets.
• Develop and document nume1ic simulation models in MATLAB/Simulink and/or
• Modelica. The models will include aircraft system and controls.
• Validate theoretic results through simulation under stressing scenarios.

Investigator: Wayne Weaver

Distributed Agent-Based Management of Agile Microgrids

US Department of Defense, Army Research Laboratory

This project plan (APP) describes the third year of the four year program for distributed agent-based management of agile microgrids. In year 1, the team has evaluated modeling and forecasting techniques for renewable energy sources as well as developed relevant case studies. In year 2 the further developed the models and forecasting techniques as well as begin implementation of simulations and hardware test cases.

The existing simulation models a user-definable, network ofmicrogrids and the Autonomous Agile Microgrid (AAM) control system. The AAM has three main components – (1) a low-level, asset control system (Decentralized Closed Loop Controller agent- DCLC), (2) a mid-level, optimal, grid state-change solver (Decentralized Model Based Control agent – DMBC) and the highest level reasoning layer, (Distributed Grid Management agent- DGM).
The entire system is “driven” by a user-configurable, time-history of prioritized loads and events based on field data.

The focus of the year three plan is to (1) increase the reasoning capability of the DGM, (2) develop an optimal power flow strategy at the DMBC level and (3) design a human-in-the-loop interface that permits real-time interaction with the simulation.

Deliverable 1. The AAM uses a command line approach to execute the simulation and observe the grid’s evolution based on a pre-defined time history scenario of events and loads. While the process for designing rich scenarios has a well-defined workflow, the system currently lacks the ability to respond to real-time inputs from a user. The deliverable is demonstration of a new human-in-the-loop capability for the AAM simulator. It will permit one user to “actuate” the power grid manually, or in an AAM-assist mode where the user can optionally decide to implement the AAM’s recommended actions. A second user will be able to trigger events and load changes in real-time, including policy and scenario changes in the DGM. The intent for year three is to create an environment for more complete testing of the AAM and demonstrate its capabilities. This feature would then be available for future studies to increase the reasoning functions of the DGM using human-in-the-loop training.

Deliverable 2. The DMBC currently computes an optimal solution to transition the power grid from its current state to a new state as requested by the DGM. These requests are based on load, generation and storage forecast agent calculations. The DMBC also triggers a new solution based on high-tempo changes to the bus voltage, independent of the DGM, due to unforecasted changes in loads or generation. All DMBC solutions are based on the assumption of a fully functioning, well-defined set of loads, generation, and storage assets. The DMBC does not compute optimal redirection of power flow based on catastrophic generation or load failures. The deliverable is the development and demonstration of a scheme to optimally redistribute power flow for contingency and catastrophic events including equipment faults and attack damage. This redistribution strategy control may be at the level of the DMBC, DCLC or both. The human-in-the-loop capability, described in Deliverable I, will be used to demonstrate this new feature by instantaneously removing generation assets and loads.

Deliverable 3. The DGM relies on data-driven load and generation forecasts to compute grid state change requests for the DMBC. The forecasts will be improved with the inclusion of additional knowledge of inventory, asset models, and situational information. While load prioritization is accommodated, there is not functionality for addressing situations where there is not enough power to accommodate all of the highest priority loads. Policies and negotiation protocols for the DGM multi-agent system that enable power sharing among microgrids will be explored. Additional policies for fine control of load shedding will be examined and simulated. The ERDC-CERL VFOB project can potentially provide a rich source of data and models to the DGM design that support more elaborate forecasts and reasoning under conflict. In addition, the data can drive methods for scenario classification (prediction of the current and future events of the base, e.g., patrols, heightened alerts, etc.). The deliverable is a report and demonstration of improved forecasting agents and conflict resolution handling, through the power scheduling agent, based on knowledge-based reasoning mechanisms and statistical risk analysis metrics. The report will document the agent models and reasoning strategies along with a description of opportunities and gaps for implementing a fully autonomous, resilient power grid.

Investigators: Gordon Parker, Laura Brown, Wayne Weaver, Steven Goldsmith

Agent Based Control with Application to Microgrids with High Penetration Renewables

Sandia National Laboratory

Abstract
Prior Work is leveraged; MTU has developed and demonstrated through simulation a prototype multiagent system that coordinates the life cycle operations of a microgrid collective composed of independent electric power sources, loads, and storage. MTU has performed simulations of DC micro grids of varying compositions and characteristics. MTU has analyzed simulation results, and developed candidate architectures and protocols for agent-based microgrid controls.

Objective
Execution of this project will further technical innovations associated with multi-agent software controlling microgrid collectives. The microgrid control algorithms for microgrid collectives will be developed and refined using Michigan Tech microgrid models and simulations validated against the MTU test bench. The algorithms will then be applied to SNL hardware models in simulation and finally against the SNL hardware test bed.

Scope
Agent-based control systems will be further developed by MTU in Matlab/Simulink blocks, tested, and refined through simulations. Once control performance objectives have been achieved, the systems will be ported to the MTU situated multi-agent system (MAS) and supporting servo loop controllers on the MTU test bench for evaluation. New Matlab simulations will be tailored and tuned to control the SNL test bed models and verified in simulation. SNL will re-apply the MTU MAS to the physical SNL test bed. SNL will collaborate with MTU on implementation and validation. Collaborative efforts will ensure that SNL attains the technology necessary to achieve the final project objectives for the SNL test bed

Required Research Innovations:
1. Identify control system performance issues between agent informatics and DC nonlinear controls. Since global computations require input from various points, processor speed and network bandwidth may dominate the performance of collaborative protocols that rely on nonlinear control approaches. Research must identify the computational and communication limits for porting nonlinear controls to agent control layers.
2. Investigate scaling properties for controls applied to increasing the number of interconnected DC microgrids. Trading power between microgrids may not be feasible due to geographical distances or communication time latencies. There may also be thresholds identified for collaboration considerations, such as partnering with 10 microgrids or less, due to the global computation requirements. Control scaling results should describe the appropriate considerations at various time scales (seconds, minutes, hours, and days). Additional considerations for scalability may include increasing the number of components within a single microgrid and increasing the variety of components within the microgrid.

Investigators: Gordon Parker, Wayne Weaver, Steven Goldsmith

Microgrid Modeling and Optimization for High Penetration Renewables Integration

Sandia National Laboratory

Abstract
Future microgrids are envisioned having a large renewable energy penetration. While this feature is attractive it also produces design and control challenges that are currently unsolved. To help solve this dilemma, development of analysis methods for design and control of microgrids with high renewable penetration is the general focus of this activity. The specific foci are (1) reduced order microgrid modeling and (2) optimization strategies to facilitate improved design and control. This will be investigated over a multi-year process that will include simplified microgrid modeling and control, single microgrid modeling and control, collective microgrid modeling and control, and microgrid (single and collective) testing and validation.

Microgrid Reduced Order Modeling (ROM)
Model development is one of the first steps in the microgrid control design process and incurs trade-offs between fidelity and computational expense. Models used for model-based control implementation must be real-time while having sufficient accuracy so that feedforward information can be maximized to achieve specified requirements. The expected outcomes of this study are (1) quantification of model uncertainty as a function of the assumptions with particular interest given to reduced order models (2) determination of appropriate time scales for reduced order modeling and (3) a MATLAB / Simulink reduced order model library of microgrid components. Contrasting different microgrid reduced order modeling approaches and simulation results that demonstrate the reduced order microgrid simulation.

Microgrid Optimization
Demonstrating microgrids with robust and high renewable penetration requires system-level extremization. This includes both its physical and control system designs. The expected outcomes of this study are (1) energy-optimal design methods suitable for microgrid design and control and (2) integration of these strategies with the microgrid reduced order model environment described above. How energy-optimal design can be exploited for microgrid design and control.

Investigators: Gordon Parker, Wayne Weaver

CPS: Breakthrough: Toward Revolutionary Algorithms for Cyber-Physical Systems Architecture Optimization

National Science Foundation

Design optimization of cyber-physical systems (CPS) includes optimizing the system architecture (topology) in addition to the system variables. Optimizing the system architecture renders the dimension of the design space variable (the number of design variables to be optimized is a variable.) This class of Variable-Size Design Space (VSDS) optimization problems arises in many CPS applications including (1) microgrid design, (2) automated construction, (2) optimal grouping, and (3) space mission design optimization.

Evolutionary Algorithms (EAs) present a paradigm for statistical inference that implements a simplified computational model of the mechanisms embedded in natural evolution, with potential to solve this problem. However, existing EAs cannot optimize among solutions of different architectures because of the inherent strategy for coding the variables in EAs. Existing EAs resembles natural evolution in which a given architecture can evolve by improving the state of its variables but cannot be revolutionized. Inspired by the concept of hidden genes in biology, this project investigates revolutionary optimization algorithms that can optimize among different solution architectures and autonomously develop new architectures that might not be known a priori, yet are more fit solution architectures. Efficacy of the new algorithms for CPS is evaluated in the context of space mission design optimization.

Intellectual Merit:
There is an increasing demand in the scientific community for autonomous design optimization tools that can revolutionize systems designs and capabilities. Most existing optimization algorithms can only search for optimal solutions in a fixed-size design space; and hence they cannot be used for solution architecture optimization. Few existing algorithms can search for optimal solutions in VSDS problems; however these are problem-specific algorithms and cannot be used as a general framework for VSDS optimization. This project investigates the novel concept of hidden genes in coding the variables in evolutionary algorithms so that the resulting algorithms can be used for optimizing VSDS problems. The key innovation in these new algorithms is the new coding strategies. In addition, in this project, the standard operations in EAs will be replaced by new operations that are defined to enable revolutionizing a current population of solution architectures using the new coding strategy. The Pl’s recent research results, in the context of space mission design optimization, demonstrate that the hidden genes optimization algorithms can search for optimal solutions among different solution architectures, revolutionize an initial population of solutions, and construct new solution architectures that are more fit than the initial population solutions.

Investigator: Ossama Abdelkhalik

Distributed Agent-Based Management of Agile Microgrids Research

Overview

A remote microgrid is a class of stand-alone power grids that services diverse loads, employs distributed generation with renewable resources, and requires on-line control and optimization to maintain stability and power flow. The grid control system is both agile and autonomous, accommodating rapid changes in generation and load resources with minimal training or intervention on the part of human operators.

Active Research Projects

Applications

  • Control based on a hybrid approach that marries novel model-predictive control strategies with multi-agent systems.
  • Utilizes artificial intelligence and machine learning techniques.
  • By imbuing software agents with component models and knowledge about grid operations the collective can cooperatively plan and execute coordinated operations that essentially re-organize grid structure in real-time while maintaining uninterrupted service.

Distributed Agent Based Management Layout

High Order Nonlinear Droop

High Order Nonlinear Droop

Distributed Flowchart