PV Power Systems Losses and Failures AI Analyser

chevron-icon
Back
project-presentation-img
Presentation
AGENOR RORIS FILHO
Project Owner

PV Power Systems Losses and Failures AI Analyser

Funding Requested

$30,000 USD

Expert Review
Star Filled Image Star Filled Image Star Filled Image Star Filled Image Star Filled Image 0
Community
Star Filled Image Star Filled Image Star Filled Image Star Filled Image Star Filled Image 3.5 (10)

Overview

A convolutional neural network (CNN) model capable of identifying whether the operating condition of a photovoltaic system (PV system) is normal or under some anomaly. The model will be trained using digital images (I-V curve) produced through mathematical modeling and simulation of the operation of a PV system’s equivalent circuit. Normal operational conditions, mismatch (tolerance in electrical parameters depending on manufacturing methods and materials, degradation due to age or external events or agents, etc.), short circuit, open circuit, and partial shading were evaluated alone or combined. The model could be available via API and be accessed by management platforms or APPs.

Proposal Description

How Our Project Will Contribute To The Growth Of The Decentralized AI Platform

Solar panel APPs or management platforms worldwide can submit data gathered from their photovoltaic systems. Engineering students can do that for free.

Our Team

I trained the AI model during my Master´s and proved its full application. Now, to improve training and expand the model's use, we only have to increase the number of solar panels and arrangements during image generation. I am pursuing a PhD in Intelligent Systems, focusing on AI.

View Team

AI services (New or Existing)

SolarPowerSystemAIAnalyzer

Type

New AI service

Purpose

Analyze the operational condition of a photovoltaic system through a measured IV curve.

AI inputs

Vector of points or digital image of the IV curve

AI outputs

Operating condition (normal mismatch open circuit short circuit shadow or combination)

The core problem we are aiming to solve

The growth of investments in renewable energies, especially solar energy, increases the search for more reliable and profitable generation systems. Monitoring the operation of these systems to analyze power losses and identify anomalies and their possible causes is one of the alternatives,  and automated methods, including those supported by machine learning algorithms, are increasingly relevant in this scenario. 

Our specific solution to this problem

This research resulted in a convolutional neural network (CNN) model capable of identifying whether the operating condition of a photovoltaic system (PV system) is normal or under some anomaly. The model was trained using digital images with graphs of the voltage behavior as a function of the electric current generated  (I-V  curve),  produced through mathematical modeling and simulation of the operation of a PV system’s equivalent circuit. Operation of a PV system’s equivalent circuit. Normal operational conditions,  mismatch  (tolerance in electrical parameters depending on manufacturing methods and materials, degradation due to age or external events or agents, etc.), short circuit, open circuit, and partial shading were evaluated alone or combined. To ensure wide use and offer adequate precision in the identification of the operating condition, the work was developed in the most varied scenarios, with several arrangements of thousands of photovoltaic panels operating under various temperatures and radiations, resulting in the generation of hundreds of thousands of images of the respective characteristic curves I-V. 

Competition and USPs

Solar energy plants are growing rapidly every year, but only a few solar panel apps or management platforms have losses and failure detection capabilities. We can offer them the API.

Engineering students can access for free.

 

Proposal Video

DF Spotlight Day - DFR4 - AGENOR RORIS FILHO - PV Power Systems Losses And Failures AI Analyser

7 June 2024
  • Total Milestones

    3

  • Total Budget

    $30,000 USD

  • Last Updated

    8 Jun 2024

Milestone 1 - API Calls & Hostings

Description

This milestone represents the required reservation of 25% of your total requested budget for API calls or hosting costs. Because it is required we have prefilled it for you and it cannot be removed or adapted.

Deliverables

You can use this amount for payment of API calls on our platform. Use it to call other services or use it as a marketing instrument to have other parties try out your service. Alternatively you can use it to pay for hosting and computing costs.

Budget

$7,500 USD

Milestone 2 - HW acquisition

Description

HW acquisition to generate images and train the model with millions of digital images of PV Systems generated with thousands of panels and dozens of arrangements: - SERVER i7/i9 128 GB 4 Tb SSD - NVIDIA 48 GB GPU Uploading and maintaining millions of images to train the CNN model is hard and should be done locally. Only the final model could by installed on cloud later..

Deliverables

Millions of images genereted to all operating condition using thousands of panels and dozen of arrangements. Trained model.

Budget

$15,000 USD

Milestone 3 - API Development

Description

Develop the API to get an IV Curve (image or point (xy)) and run the model to show the operating condition (normal or under anomalies).

Deliverables

API ready to run the trained model

Budget

$7,500 USD

Join the Discussion (2)

Sort by

2 Comments
  • 0

    Thank you very much. Let's do it!!!!

  • 0
    commentator-avatar
    CLEMENT
    Jun 1, 2024 | 3:30 PM

    The availability of this AI tool can optimizing energy production and minimize energy losses. I hope the team can really deliver on this.

Reviews & Rating

Sort by

10 ratings
  • 0
    user-icon
    Ana Paula Pereira
    May 29, 2024 | 1:25 PM

    Overall

    5

    • Feasibility 5
    • Viability 5
    • Desirabilty 5
    • Usefulness 5
    Great initiative

    Agenor is an outstanding, qualified Brazilian researcher with a focus on AI applications and solutions. This proposal tackles a real-world problem and aligns Deep Funding with a trend in decentralized infrastructure, called DePIN. In other words, this is a project that not only promises to enhance our understanding of solar system failures through AI analytics but also paves the way for innovative uses of decentralized technologies in critical real-world applications. 

    user-icon
    AGENOR RORIS FILHO
    Jun 9, 2024 | 10:28 PM
    Project Owner

    Thank you very much for the review.

  • 0
    user-icon
    Nicolad2008
    Jun 10, 2024 | 4:28 AM

    Overall

    4

    • Feasibility 3
    • Viability 4
    • Desirabilty 4
    • Usefulness 3
    API integration into management platforms

    This project is capable of distinguishing normal operating conditions and abnormalities through digital images (I-V curves) generated from mathematical modeling and equivalent circuit simulation of the PV system. This not only helps improve operating performance but also contributes to timely maintenance and repairs, minimizes energy loss and enhances system reliability. However, the project also faces challenges. some challenges. Scaling the model to include a variety of solar panel types and configurations can require large amounts of training data and powerful computational resources. Additionally, API integration into management platforms or mobile applications needs to be done carefully to ensure compatibility and ease of access.

    user-icon
    AGENOR RORIS FILHO
    Jun 10, 2024 | 1:28 PM
    Project Owner

    Thank you very much for the review.

  • 0
    user-icon
    CLEMENT
    Jun 1, 2024 | 3:28 PM

    Overall

    4

    • Feasibility 4
    • Viability 4
    • Desirabilty 4
    • Usefulness 4
    Significantly impact the renewable energy sector

    I see this to be of great use for the renewal energy sector. This because it offers features which are critical for maximizing energy production and minimizing downtime. For me, this technology also has the potential to enhance the efficiency and reliability of solar power systems, ultimately contributing to the widespread adoption of renewable energy sources.

    As a contribution to the SNET AI Marketplace, this analyser provides a valuable tool for stakeholders in the solar energy industry. Management platforms, as well as individual users through apps, can access the AI analyser via API to monitor the performance of PV systems in real-time.

    Kudos to the team !

    user-icon
    AGENOR RORIS FILHO
    Jun 9, 2024 | 10:28 PM
    Project Owner

    Thank you very much for the review.

     

  • 0
    user-icon
    Gombilla
    Jun 10, 2024 | 11:12 AM

    Overall

    4

    • Feasibility 4
    • Viability 4
    • Desirabilty 4
    • Usefulness 4
    Enables proactive maintenance of PV systems

    I am concerned with potential issues with false positives (incorrectly identifying normal conditions as anomalies) or false negatives (failing to detect actual anomalies) which could affect the reliability of the system. However, I commend the availability via API which will allow integration with management platforms or apps, facilitating seamless access and integration into existing infrastructure, allowing for timely intervention to prevent or mitigate losses and failures. Great job !

    user-icon
    AGENOR RORIS FILHO
    Jun 10, 2024 | 1:28 PM
    Project Owner

    Thank you very much for the review.

  • 0
    user-icon
    Max1524
    Jun 9, 2024 | 9:38 AM

    Overall

    3

    • Feasibility 3
    • Viability 3
    • Desirabilty 4
    • Usefulness 3
    Waiting for good results for energy sources

    I look forward to the result of improved energy production, while maximizing energy loss. If these two things are realized together, the product will be a great success. This AI analysis tool is gradually asserting its role in regional energy in particular and the world in general - it is still early for us to confirm the usefulness of the Convolutional Neural Network (CNN) Model. But we can absolutely hope for it when it comes to energy issues.

    user-icon
    AGENOR RORIS FILHO
    Jun 9, 2024 | 10:30 PM
    Project Owner

    Thank you very much for the review. The research and model creation results were very good, and I hope I can improve them soon.

  • 0
    user-icon
    Devbasrahtop
    May 20, 2024 | 9:49 AM

    Overall

    3

    • Feasibility 3
    • Viability 3
    • Desirabilty 4
    • Usefulness 4
    Lack of detailed info on technical methodology

    Feasibility

    The project aims to develop a CNN model to identify anomalies in photovoltaic (PV) systems using I-V curve images. The concept is technically feasible, given the proponent's background in AI and their experience with the model during their Master's studies. However, the proposal lacks details on the technical methodology for generating and simulating the I-V curves, as well as the specifics of how the CNN model will be trained and validated. Furthermore, the requirement for significant hardware acquisition and the associated costs might present logistical challenges that could impact the project's feasibility.

    Viability

    The project's capacity to scale and maintain the model and API will determine its profitability. The plan does not adequately address the practical concerns of implementing and sustaining the service, particularly the API, despite the proponent's strong academic background. Furthermore, no clear plan for upgrades and maintenance is mentioned, which is important for a service that is supposed to be linked to systems for managing solar panels. Although the budgetary allotment appears appropriate, it may be insufficient given the extent of hardware procurement and ongoing expenses.

    Desirability

    The project is very attractive since it fills a big gap in the quickly expanding solar energy industry. The success of the sector depends on increased efficiency, decreased downtime, and overall superior performance, all of which can be achieved with automated anomaly detection in PV systems. Giving engineering students unrestricted access is a great idea that can encourage learning and creativity. A more thorough market analysis and a more defined strategy for involving stakeholders and potential users would, however, improve the concept.

    Usefulness

    The proposed solution has substantial potential for utility. By providing a reliable and automated method for detecting anomalies in PV systems, the project can significantly improve the management and maintenance of solar energy installations. The API's integration into existing solar panel apps and management platforms could streamline operations and provide valuable insights for operators. But then, its usefulness will ultimately depend on the model's accuracy and reliability in real-world conditions, which require thorough validation.

    user-icon
    AGENOR RORIS FILHO
    May 22, 2024 | 2:41 PM
    Project Owner

    Thank you very much for the review. If the project is selected, I hope we can make adjustments taking advantage of the community's contributions and make it more effective.

  • 0
    user-icon
    TrucTrixie
    Jun 9, 2024 | 2:25 PM

    Overall

    3

    • Feasibility 4
    • Viability 3
    • Desirabilty 3
    • Usefulness 3
    Milestone #3 should be presented more

    The first 2 milestones are presented quite thoroughly, but milestone 3 about API Development should be presented more thoroughly. I find it a bit brief and difficult to understand for those who do not have technical majors because it requires certain knowledge to understand. If the author can do this, it's good for Viability.

    user-icon
    AGENOR RORIS FILHO
    Jun 9, 2024 | 10:31 PM
    Project Owner

    I know, sorry, my fault! I've missed the timeout. Thank you for the review.

  • 0
    user-icon
    Joseph Gastoni
    May 22, 2024 | 9:12 AM

    Overall

    3

    • Feasibility 3
    • Viability 3
    • Desirabilty 2
    • Usefulness 3
    a CNN model for identifying anomalies

    This proposal outlines a CNN model for identifying anomalies in photovoltaic systems using I-V curve images. Here's a breakdown of its strengths and weaknesses:

    Feasibility:

    • High: The core concept (using CNNs for anomaly detection) is feasible with existing tools and libraries.
      • Strengths: Leverages established techniques in computer vision and machine learning.
      • Weaknesses: Training a robust model might require a large dataset of real-world I-V curve images, which could be challenging to obtain.

    Viability:

    • Moderate: Success depends on the accuracy and efficiency of the model, as well as the business model for API access.
      • Strengths: The proposal addresses a need for automated anomaly detection in solar energy systems.
      • Weaknesses: The proposal lacks details on the model's performance (accuracy, generalizability) and the pricing strategy for the API.

    Desirability:

    • Moderate: For solar panel app and management platform developers, this could be desirable.
      • Strengths: The proposal offers a potentially valuable tool for improving solar system monitoring and maintenance.
      • Weaknesses: The proposal needs to address concerns about the model's reliability and the potential cost of API access.

    Usefulness:

    • Moderate-High: The project has the potential to improve the efficiency and profitability of solar energy systems, but its impact depends on the model's effectiveness and adoption rate.
      • Strengths: The proposal offers a way to automate anomaly detection and potentially reduce maintenance costs.
      • Weaknesses: The proposal lacks details on how the model will be integrated with existing monitoring systems and how it will handle real-world variations in I-V curves.

    Overall, this project has a sound technical approach, but focus on:

    • Data Acquisition: Developing a strategy to obtain a large and diverse dataset of real-world I-V curve images for training the model.
    • Model Performance: Evaluating the model's accuracy and generalizability on real-world data and showcasing its performance metrics.
    • API Design and Pricing: Defining a clear API design and pricing strategy that is attractive to solar panel app and management platform developers.
    • Integration Considerations: Addressing how the model will be integrated with existing monitoring systems and how it will account for variations in I-V curves.

    By addressing these considerations, this project can increase its chances of success and become a valuable tool for the solar energy industry.

    Here are some strengths of this project:

    • Addresses a relevant problem in the solar energy industry - automated anomaly detection in photovoltaic systems.
    • Leverages advancements in deep learning (CNNs) for image-based anomaly detection.
    • Offers a potential solution that could be integrated with existing solar panel apps and management platforms.

    user-icon
    AGENOR RORIS FILHO
    May 22, 2024 | 2:41 PM
    Project Owner

    Thank you very much for the review. If the project is selected, I hope we can make adjustments taking advantage of the community's contributions and make it more effective.

  • 0
    user-icon
    Tu Nguyen
    May 23, 2024 | 1:42 AM

    Overall

    3

    • Feasibility 3
    • Viability 4
    • Desirabilty 3
    • Usefulness 4
    PV Power Systems Losses And Failures AI Analyser

    This proposal addresses a problem in the solar energy industry, which is monitoring anomalies in photovoltaic systems. This proposal created a convolutional neural network model capable of determining whether the operating conditions of a photovoltaic system are normal or have some abnormalities. These are useful solutions in practice. Hope they will implement it successfully.
    Currently, the proposal only has 1 member. This creates the risk that they will fall behind the project schedule. They should also determine the start and end times of milestones. Additionally, they should also define a more detailed budget for each milestone.

    user-icon
    AGENOR RORIS FILHO
    May 24, 2024 | 9:14 AM
    Project Owner

    Thank you very much for the review. If the project is selected, I hope we can make adjustments taking advantage of the community's contributions and make it more effective.

  • 0
    user-icon
    BlackCoffee
    Jun 10, 2024 | 1:23 AM

    Overall

    3

    • Feasibility 3
    • Viability 3
    • Desirabilty 3
    • Usefulness 3
    Is the author alone making suggestions?

    I see that there is only one member, Roris Filho, who is completely revealing his identity through the introduction and linkedin social network link. I guess there will be others working with Roris to complete this interesting proposal. That's why I hope the author will try to publicize the complete identities of the co-participants so the community can know.

    user-icon
    AGENOR RORIS FILHO
    Jun 10, 2024 | 1:30 PM
    Project Owner

    Thank you very much for the review. The proposal was my Master's degree research and its why I am alone. More details about the team you can see in the proposal video. 

Summary

Overall Community

3.5

from 10 reviews
  • 5
    1
  • 4
    3
  • 3
    6
  • 2
    0
  • 1
    0

Feasibility

3.5

from 10 reviews

Viability

3.6

from 10 reviews

Desirabilty

3.6

from 10 reviews

Usefulness

3.6

from 10 reviews

Get Involved

Contribute your talents by joining your dream team and project. Visit the job board at Freelance DAO for opportunites today!

View Job Board