State Recognition of Surface Discharges by Visible Images and Machine Learning

Yuan, Zhe; Ye, Qizheng; Wang, Yuwei; Guo, Zeqing

doi:10.1109/tim.2020.3031543

Cited by 15 publications

(10 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Because of the low fault current produced, it is below the sensitivity of current circuit breakers, so they are only able to trip when the problem is in an advanced stage, and a certain level of damage has been produced. Arc tracking activity releases energy and generates different measurable effects [13], including radio waves, sound, visible and ultraviolet light, heat or chemicals such as ozone, NOx and nitric acid, among others [14]. Arc tracking activity tends to pyrolyze the insulation materials surrounding the core of the wire, thus producing a partially conductive path that promotes discontinuous arcing activity of low magnitude that overheats the insulation [15], thus producing more damage.…”

Section: Introductionmentioning

confidence: 99%

Performance Evaluation of Solar-Blind Gas-Filled Sensors to Detect Electrical Discharges for Low-Pressure Aircraft Applications

Ruiz

Moreno-Eguilaz

Boizieau

et al. 2022

Sensors

View full text Add to dashboard Cite

Unpressurized aircraft circuits facilitate the initiation of electrical discharges in wiring systems, with consequent damage to related insulation materials and safety hazards, that can and have already caused severe incidents and accidents. Specific sensors and solutions must be developed to detect these types of faults at a very incipient stage, before further damage occurs. Electrical discharges in air generate the corona effect, which is characterized by emissions of bluish light, which are found in the ultraviolet (UV) and visible spectra. However, due to sunlight interference, the corona effect is very difficult to detect at the very initial stage, so the use of solar-blind sensors can be a possible solution. This work analyzes the feasibility of using inexpensive non-invasive solar-blind sensors in a range of pressures compatible with aircraft environments to detect the electrical discharges at a very incipient stage. Their behavior and sensitivity compared with other alternatives, i.e., an antenna sensor and a CMOS imaging sensor, is also assessed. Experimental results presented in this paper show that the analyzed solar-blind sensors can be applied for the on-line detection of electrical discharges in unpressurized aircraft environments at the very initial stage, thus facilitating and enabling the application of predictive maintenance strategies. They also offer the possibility to be combined with existing electrical protections to expand their capabilities and improve their sensitivity to detect very early discharges, thus allowing the timely identification of their occurrence.

show abstract

Section: Introductionmentioning

confidence: 99%

Performance Evaluation of Solar-Blind Gas-Filled Sensors to Detect Electrical Discharges for Low-Pressure Aircraft Applications

Ruiz

Moreno-Eguilaz

Boizieau

et al. 2022

Sensors

View full text Add to dashboard Cite

show abstract

“…Insulation faults are not often directly evaluated. Instead, some measurable physical effects can be analyzed [ 20 ], such as acoustic noise, electromagnetic radiation, radio interference voltage, ultraviolet or visible emissions, generation of chemical components such as ozone or others or temperature changes. There is an imperative need to evaluate the severity of insulation faults in existing installations, since it is vital for ensuring a safe, reliable and stable operation of power systems.…”

Section: Introductionmentioning

confidence: 99%

“…There is an imperative need to evaluate the severity of insulation faults in existing installations, since it is vital for ensuring a safe, reliable and stable operation of power systems. It is well known that PDs often generate optical radiation [ 20 ] in the ultraviolet (UV), visible and infrared bands [ 21 , 22 ], whereas the spectral content of the discharges depends upon the type of electrode and the nature of the discharges. Many research works focusing on the analysis of optical emissions generated by PDs analyze the UV band [ 23 , 24 ].…”

Section: Introductionmentioning

confidence: 99%

“…Many research works focusing on the analysis of optical emissions generated by PDs analyze the UV band [ 23 , 24 ]. However, the use of UV cameras is still problematic, due to their high cost, issues related to noise performance or limited image resolution, thus leading to unsatisfactory spatial resolution and PD source location accuracy [ 20 ]. Although hybrid UV-visible cameras based on the latest fusion technology merging both visible and UV images are commercially available, its industrial application becomes very difficult because of their high volume and weight and prohibitive cost.…”

Section: Introductionmentioning

confidence: 99%

“…In addition, industrial applications are increasingly applying visible cameras due to the fast evolution of visible imaging sensors [ 26 ]. With regard to the aforementioned and the fact that visible images of PDs contain rich information, it is highly appealing to study and characterize the PDs by analyzing the information contained in visible images of the PDs [ 20 ]. Nowadays, this strategy can be implemented due to the latest developments of small-size, low-cost and high-resolution visible imaging sensors, most of which are also partially sensitive to specific regions of the UV spectrum [ 4 ].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Insulation Failure Quantification Based on the Energy of Digital Images Using Low-Cost Imaging Sensors

Riba

Gómez-Pau

Moreno-Eguilaz

2020

Sensors

View full text Add to dashboard Cite

Insulation faults in high-voltage applications often generate partial discharges (PDs) accompanied by corona activity, optical radiation mainly in the ultraviolet (UV) and visible bands. Recent developments in low-cost, small-size, and high-resolution visible imaging sensors, which are also partially sensitive to the UV spectral region, are gaining attention due to their many industrial applications. This paper proposes a method for early PD detection by using digital imaging sensors, which allows the severity of insulation faults to be assessed. The electrical power dissipated by the PDs is correlated to the energy of the acquired visible images, and thus, the severity of insulation faults is determined from the energy of the corona effect. A criterion to quantify the severity of insulation faults based on the energy of the corona images is proposed. To this end, the point-to-plane gap configuration is analyzed in a low-pressure chamber, where digital image photographs of the PDs are taken and evaluated under different pressure conditions ranging from 10 to 100 kPa, which cover the typical pressure range of aeronautic applications. The use of digital imaging sensors also allows an early detection, location and quantification of the PD activity, and thus assessing the severity of insulation faults to perform predictive maintenance tasks, while enabling the cost and complexity of the instrumentation to be reduced. Although the approach proposed in this paper has been applied to detect PDs in aeronautic applications, it can be applied to many other high-voltage applications susceptible of PD occurrence.

show abstract

Quantitative measurement of NO₂^— distribution in plasma jet treated model tissue, based on image processing technology

Wang

et al. 2022

Plasma Processes & Polymers

View full text Add to dashboard Cite

One of the key issues in plasma medicine is the measurement of the spatial‐temporal distribution and total amount of reactive species in treated tissue. Here, we provide a solution based on image processing technology. A DC‐driven air plasma jet was used, and NO2— was selected as the representative for the test. The model tissue was frozen sliced, and the relationship between the color and standard concentration was established. The temporal distributions of NO2— concentration versus time and flow rate were analyzed in each slice, and the total amount of NO2— in the treated tissue was obtained. This study offers a new diagnostic method for reactive species amount into tissue, which is an effective guide for plasma medicine applications and numerical modeling research.

show abstract

State Recognition of Surface Discharges by Visible Images and Machine Learning

Cited by 15 publications

References 26 publications

Performance Evaluation of Solar-Blind Gas-Filled Sensors to Detect Electrical Discharges for Low-Pressure Aircraft Applications

Performance Evaluation of Solar-Blind Gas-Filled Sensors to Detect Electrical Discharges for Low-Pressure Aircraft Applications

Insulation Failure Quantification Based on the Energy of Digital Images Using Low-Cost Imaging Sensors

Quantitative measurement of NO₂^— distribution in plasma jet treated model tissue, based on image processing technology

Contact Info

Product

Resources

About

State Recognition of Surface Discharges by Visible Images and Machine Learning

Cited by 15 publications

References 26 publications

Performance Evaluation of Solar-Blind Gas-Filled Sensors to Detect Electrical Discharges for Low-Pressure Aircraft Applications

Performance Evaluation of Solar-Blind Gas-Filled Sensors to Detect Electrical Discharges for Low-Pressure Aircraft Applications

Insulation Failure Quantification Based on the Energy of Digital Images Using Low-Cost Imaging Sensors

Quantitative measurement of NO2— distribution in plasma jet treated model tissue, based on image processing technology

Contact Info

Product

Resources

About

Quantitative measurement of NO₂^— distribution in plasma jet treated model tissue, based on image processing technology