Introduction

Join us for a transformative webinar on Predictive Maintenance with MATLAB and Simulink, exploring a captivating case study of a packaging machine. Discover how this innovative approach revolutionizes maintenance strategies, enhances efficiency, and leads to substantial cost savings. The use of AI techniques on time-series data is growing in popularity across electric utilities sector for asset management, demand response, outage management, customer services, energy storage, renewable resources, and many other areas in the power generation and delivery system. 

In this webinar, we will present a case study on “Identifying Risk in Underground Utility Cable Systems Using Machine Learning and Deep Learning”. Predictive maintenance begins with understanding how cable system failures occur. Analyzing and interpreting results from partial discharge (PD) measurements taken in the field can be a complex task for humans. Machine learning algorithms and deep learning algorithms are used to automatically identify and categorize markers of defects contained in the PD measurements. These algorithms are used to categorize different defect types by the risk of going to failure soon. Differentiating cables with “high to low-risk defects” along with those that are "defect-free" enables predictive maintenance. Examples of identified defects will be presented.

You will also learn how to apply AI using MATLAB® for asset condition monitoring and find out about tools and fundamental approaches for developing advanced predictive models on time series data. Using a real-world faulty dataset, we will show two approaches of building deep learning models using convolution neural networks and recurrent neural networks, and finally deploying the models on an edge device or the cloud.

Highlights:

1. Acquiring, creating and annotating faulty datasets
   
2. Applying time-frequency transformations, extracting  established signal features, and automating deep feature extraction using Invariant Scattering Convolutional Networks
3. Building and comparing Deep Learning models with CNNs and LSTMs.
4. Generating stand-alone CUDA code to deploy models on edge devices.
5. Deploying and integration on the cloud