Beruflich Dokumente
Kultur Dokumente
ISGT 2019
Washington DC
2
Industrial
7
Winter
9
• Assigning Multi-
Timescale Models
to Different Typical
Load Behavior
Patterns Obtained
From Observed
Loads
13
Tracking the typical daily consumption Using inferred load for accurate
pattern of unobserved customers system monitoring (branch
using a Bayesian learning approach current state estimation)
15
Conclusions
• Access to Real Utility Data:
Designing More Accurate and Realistic Load Models
Discovering Strengths and Shortcomings of Available
Algorithms in the Literature
Active Collaboration with Industry
• Using Data-Driven Load Inference for Monitoring
Distribution Systems:
Identifying Temporal Correlations for Load Estimation
Exploiting Latent Trends in Customer Behavior at Different
Time-Scales for Enhancing Inference Accuracy
Facilitating Computational Efficiency for Real-Time Application
17
• Deep Deka
Distribution Grid
Learning Problems
• Structure Learning
• Configuration of switches
Substation
Load Nodes
Learning Problems
• Structure Learning
• Learning Line Impedances
R, X
Substation
Load Nodes
Learning Problems
• Structure Learning
• Learning Line Impedances
• Incomplete observations
Substation
Load Nodes
Missing Node
Applications of Learning Problems
• Structure:
• Failure/Fault Identification
• Connection/phase verification
• Impedance Estimation:
• Non-intrusive control
• Use in DSO optimization
• Optimizing flows
Our approach:
Advantage: Provable results, Missing data extensions
IEEE TCNS’17, PSCC’18
Physics
(Power-Systems) Physics Informed
Interpretability
Physics-Free
PMU data-to-
predictions Machine Learning
approaches (automatic, training & execution efficient,
but lacking Power System interpretability)
Speed
Advantage: Provable results, Missing data extensions
Learning with nodal voltages
𝑐𝑐 a
𝑏𝑏 𝑏𝑏
𝑏𝑏
𝑐𝑐
𝑎𝑎 𝑎𝑎
𝑐𝑐
Topology Reconstruction
Advantages:
• NO additional information needed a
(no impedance/injection statistics)
𝑏𝑏
• Complexity is O(ElogE)
• Works for monotonic flows (gas,water, heating)
• Can be extended to missing nodes
𝑐𝑐
Topology Learning (No missing nodes)
Cross-covariances:
Injection Data assumption
•• Assumption (Now)
(prior work)
: :
Observed are
Injections nodes are uncorrelated/independent
correlated and follow a Graphical Model
15- min detrended injections from households in DiSC data-set (Aalborg Univ.)
Injection Data assumption
• Assumption (Now) :
Injections are correlated and follow a Graphical Model
Graphical Model:
Correlation Conditional Dependence
Learning with end-users
b
Learning with end-users
b
Learning with end-users
• Algorithm:
Compute effective impedances between leaves
b
Learning with end-users
• Algorithm:
Compute effective impedances between leaves
Recursive Grouping Algo (Willsky’11) to learn topology & distances
from known effective impedances
1. Learn siblings
Recursive Grouping Algo
2. Introduce parents
3. Update distance
Recursive Grouping Algo
1. Learn siblings
Recursive Grouping Algo
2. Introduce parents
3. Update distance
Recursive Grouping Algo
After Iterations
Learning with end-users
• Algorithm:
Compute effective impedances between leaves
Recursive Grouping Algo to learn topology & distances from known
effective impedances
• Algorithm:
Compute effective impedances between leaves
Recursive Grouping Algo to learn topology & distances from known
effective impedances
Effect of Correlated Injection
• Algorithm:
Compute effective impedances between leaves
Uncorrelated Injections
• Algorithm:
Compute effective impedances between leaves
Correlated Injections
• Algorithm:
Compute effective impedances between leaves
Correlated Injections
Computing inverse
Graphical Model for injections
Effect of Correlated Injection
• Algorithm:
Compute effective impedances between leaves
Correlated Injections
Computing inverse
Graphical Model for injections
ML estimate (SPICE) for inverse:
Sample Complexity
Uncorrelated : Correlated :
For a grid with constant depth 𝑂𝑂( 𝑉𝑉 2 log 𝑉𝑉 /𝜂𝜂 ) samples
and sub-Gaussian complex power recovers the true topology with
injections, 𝑂𝑂( 𝑉𝑉 log 𝑉𝑉 /𝜂𝜂 ) probability 1 − 𝜂𝜂.
samples recovers the true
topology with probability 1 − 𝜂𝜂.
• Steps:
Bound decay of (covariance of injections with voltages)
Bounds between empirical and true values of covariances
Bound distance using Union bounds
• Learning with End-Users in Distribution Grids (PSCC 2018, TCNS under review)
• Learning with Correlated Injections (Asilomar 2018)
Simulations: IEEE 33 bus graphs (Matpower samples)
500 600 700 800 900 1000 500 600 700 800 900 1000 500 600 700 800 900 1000
Extensions:
• Straightforward:
Three-phase unbalanced systems (under linearized flows)
Change detection: Fault localization, nodal change
Dynamical setting (LTI system) with swing equations
(CDC 2018, PSCC 2018)
• More Work:
Radial Loopy grids
Other Physics-Informed AI approaches in grid
• Neural Network features for ARC: Average Rank of Correct faulted line
fault localization
Reactive power
Neural Networks for
time-series data
OpenIPSL Generator ----- True
(12th order) ----- predicted
Web: https://cnls.lanl.gov/gridscience
Email: ansi-info@lanl.gov for post-doc positions and school.
Support from:
Linearized Power Flow (LinDistFlow)
• Non-linear AC flow:
c b
•
Technical Opportunity
• SCUC is solved repeatedly, multiple times per day:
– Day-ahead market: once per day
– Resource reliability and commitment: twice per day
– Reliability assessment commitment: 24 times per day
• Data only changes slightly from run to run:
– Almost same network topology and set of generators
– Slightly different demands and production costs
– Market operators have large amounts of historical data
• Perfect setting for Machine Learning
5
Proposed Architecture
Current Approach Proposed AI Approach
Goal: Develop methods and tools that learn from past experience and become
progressively more efficient.
6
Previous ML Research
Simplified Formulation
9
• Hyperplanes considered:
– xgt= 0
– xgt = 1
– xgt= xg,t+1
• Classifier: Support Vector Machines
• Training high-quality models:
– Discard hyperplanes with very unbalanced labels
– Measure precision and recall using k-fold cross validation
– Discard models with low recall or precision not significantly
better than dummy predictor
13
• Computational Environment:
– Java, Python, using pandas and scikit-learn
– IBM ILOG CPLEX 12.8.0 as MIP solver
– Intel Xeon E5-2695v4, 36 cores, 128GB DDR4
• Realistic, large-scale instances (adapted from MATPOWER):
– 9 instances in total
– Up to 6,515 buses
– Up to 1,388 generation units
– Up to 9,037 transmission lines
– Scale comparable to PJM and MISO
14
Acknowledgments