RAM analysis on Power systems using Monte Carlo & Matlab

Name: RAM analysis on Power systems using Monte Carlo & Matlab
Rating: 4.0 (8 reviews)

Reliability, Availability & Mantainability analysis on electrical power systems using Monte Carlo simulations and MATLAB

Created byFernando Alberto Tellez Miotta

Last updated 6/2023

English

What you'll learn

Learn reliability, availability and mantainaility fundamental concepts
Learn how to detect the failing modes responsible for having a complete system shutdown
Build individual components failure and reparing time period modeling, from historical data and/or other references using random number generation
Learn to model a whole system's failures and reparing times from its individual building blocs modeling
Learn to estimate realiability, availability and mantainability of a system using Matlab and monte carlo
Create software capable of simulating any kind of system from scratch

Course content

7 sections • 39 lectures • 4h 30m total length

Course scope and objectives9:11

Why Monte Carlo?4:36
Mathematical concepts: Probability density function (PDF).6:14
Explore probability density functions (pdfs) to model component failures for Monte Carlo simulations, including Gaussian (normal) pdf, exponential pdf with memoryless property, Weibull, and uniform pdf.
Mathematical concepts: CPF and ICPF.8:39
Practical examples: Random sample generation through ICPFs10:55
Generate exponential random samples with a mean around 900 using the inverse cumulative probability function in Matlab, building a micro monte carlo approach to estimate revenue losses in ram analysis.
Mathematical concepts: STD, SEM and CI.7:49
Practical examples: Commands for acquiring STD, SEM and CI.3:38
Demonstrate practical use of standard deviation, standard error of the mean, and confidence intervals with a custom function on a 100,000-sample exponential data set, yielding a 95% interval around 897–908.
Reliability, Availability and Mantainability definitions (TTF, TTR ,MTTF, MTTR)8:23
Practical examples: Adjusting a sample to a probability density function7:18
Simulating an individual component's on and off status switches over tmission 14:45
Simulate an individual component's on and off behavior over 8760 hours using Monte Carlo and Matlab, generating random on/off intervals and summing to the total mission time.
Simulating an individual component's on and off status switches over tmission 28:42
Creating our first Monte Carlo RAM analysis code for an individual component12:55
How many simulations are enough?6:59
Exponential probability density function (''memoryless'' function)3:44

What is a reliability block diagram?8:04
Simulating series and parallel RBDs3:31
Use built-in matlab functions to simulate series and parallel configurations, load series_status.mat and parallel_status.mat, and plot the resulting system status for both scenarios.
Minimum cut sets (applied to electrical power systems) for RBD construction.6:24
The adjecency matrix of a graph3:57
Generating the adjacency matrix of a power system using Matlab6:34
Practical case: applying the adjacency matrix function to our big case of study.4:20
Example of elaboration of an RBD for a power system.8:17
Build and test an adjacency matrix from an input file to map generation and load nodes in a power system, then derive RBD failure modes via path and combination analysis.
How do systems fail and how to simulate this failings?4:57

Simulating a series system26:16
Simulating a parallel system13:49
Demonstrate Monte Carlo reliability analysis of a parallel power system by implementing the Weibull and exponential pdfs, modeling repair times, and computing reliability and availability from a reliability block diagram.
Making sure our Monte Carlo RAM Analysis software works3:20
Finding the weak links within the RBD11:23
Final project of section 4 (part I)6:19
Final project of section 4 (part II)10:36

Loss of load probability, Loss of load expectation and how to calculate them.10:30
Applying LOLP and LOLE to build a deterministic MATLAB simulation2:38
Why Monte Carlo? Part I1:48
Why Monte Carlo? Part II1:30
This lecture explains why Monte Carlo methods are needed for power system analysis, citing ram limits and massive combination counts that exceed traditional computation.
Explaining a matlab function to calculate LOLE and LOLP4:57
Analyze a MATLAB function for RAM analysis that calculates LOLE and LOLP for generator on/off configurations using available capacity, load, and loss of load probability.
Creating a Monte Carlo software to calculate LOLP and LOLE15:22
Develop a MATLAB based Monte Carlo tool to compute LOLP and LOLE for power systems, using generator on/off simulations, available capacity and load, and confidence intervals.
LOLE and LOLP as a function of the target load4:59
Explore how LOLE and LOLP vary with target load using a Monte Carlo MATLAB calculator, simulating loads from zero to available capacity and estimating confidence intervals for Lola and Lolb.
Wrapping up LOLE and LOLP3:43

Requirements

Basic programming knowledge using Matlab (or Octave)
Basic engineering knowledge
Basic probability and statistics knowledge

Description

Reliability centered maintenance has become a common practice on maintenance departments over all kinds of fields, from pretoleum fields to power system, passing through naval and aeronautical industry. Reliability centered maintenance allows for a cost effective maintenance policy that is focused on a system's different modes of failure and consecuences.

Most of the times, the failures of a system are fundamentaly random in their behaviour, thus, having a tool capable of simulating this random behaviour thousands or even millions of times in order to get a statistical trend is extemely valuable so that we can plan maintenance policies that tackle the most likely failing modes and the most catastrophic ones.

Monte Carlo methods is an umbrella terms that covers all the studies that rely on many similations of random systems in order to get their most likely behaviour over the span of several tries.

I had the opportunity to work this specific topic on my undergraduate thesis ''RAM analysis of electrical power system on the operational context using sequential Monte Carlo'' back in 2016 and got awarded with honors upon my disertation. I'll thrive myself to pour everything I learned into this course. I'm looking forward for your questions and feedback!

Who this course is for:

Engineers and engineering students
Matlab practitioners
Reliability centered maintenance engineers
Reliability engineers
Maintenance engineers
Maintenance managers

RAM analysis on Power systems using Monte Carlo & Matlab

What you'll learn

Explore related topics

Course content

Introduction1 lecture • 9min

Basic concepts13 lectures • 1hr 35min

How do systems fail?8 lectures • 46min

Building our Monte Carlo RAM analysis software6 lectures • 1hr 12min

LOLP and LOLE8 lectures • 45min

Wrapping up all1 lecture • 1min

Congratulations and course closing2 lectures • 2min

Requirements

Description

Who this course is for: