Power Guard

 

Model: PG108

Category: Sensor Device

Exhibitor: HARBOR TECHNOLOGY SOLUTIONS CO., LTD.

Booth No: N830

 

Characteristic

 

After years of continuous research and innovation, Harbor Technology has developed the Power Guard Early-Stage Partial Discharge Monitoring based on AI signal processing technology. This system includes an innovative active full-frequency range antenna, equipped with AI dynamic noise reduction, and a host with real-time PRPD (Phase Resolved Partial Discharge) feature recognition technology. This breakthrough surpasses traditional noise interference limitations, enabling the effective detection of corresponding electromagnetic waves in the early stages of partial discharge in high-voltage equipment. Additionally, it accurately identifies the type and severity of the partial discharge.
In the realm of AI dynamic noise reduction, the system utilizes an advanced active full-frequency range electromagnetic wave sensing antenna, which can simultaneously receive the full-frequency range electromagnetic wave frequency (300MHz~3GHz) generated by partial discharge. It then conducts real-time signal frequency division processing in a three-frequency band manner at the antenna end, providing the necessary real-time frequency division basic data for AI dynamic noise reduction. Coupled with AI deep learning algorithms, it can effectively recognize and filter dynamic noise characteristics from various frequency bands, such as those originating from nearby power equipment, mobile phones (900MHz, 1800MHz), broadcasts, low-power RF devices (e.g., remote controls, wireless headphones, IoT devices), etc.
Regarding the real-time PRPD feature recognition, the pre-processed signals, which have undergone the aforementioned AI dynamic noise reduction, undergo information standardization and normalization. Subsequently, signal processing techniques and spectrum analysis are employed to extract features, establishing PRPD data maps. The system then directly extracts the operating voltage frequency and phase of 60Hz/50Hz from the data, serving as the basis for the 0°~360° phase positioning of PRPD. It automatically selects suitable AI models, utilizes techniques such as cross-validation to automatically evaluate and optimize model performance, ensuring the correct analysis and judgment of different types of PRPD features (e.g., normal, noise, internal discharge, surface discharge, and tip discharge) and severity levels. It achieves the maximum applicability and generalization of partial discharge detection in various high-voltage equipment under different environments.
Another distinct feature of the Power Guard early-stage partial discharge monitoring is its sensor antenna connection via CAT6 network cable (including isolated ground network) to the host, providing POE (Power Over Ethernet) power supply function, with a range of up to 100 meters. This makes the installation of the antenna highly convenient and flexible, and it also helps to install the host in an appropriate location.
By combining the aforementioned AI dynamic noise reduction and AI real-time PRPD feature recognition technology, integrating information from different hosts, and further analyzing big data trends using AI algorithms, the Power Guard early-stage partial discharge diagnosis system offers efficient AI processing through its unique three-layer AI architecture.

 

Other Products

 

• 

  AI Diagnostic and Prognostic Solution

 

 

 

Back