Robot nga Autonomo para sa Pagsusuri sa 110kV Transmission Lines: Pagdisenyo ug Implementasyon sa usa ka Sistema nga Tres-Braso nga Ngaasang

Abstract

Arangin ang mga inherent nga limitasyon sa manual nga inspeksyon ug aerial survey para sa high-voltage transmission lines, kini nga proposal nagpakilala og autonomous inspection robot nga gitukod pinaagi sa 110 kV power lines. Ang robot adunay usa ka bag-ong tulo-ka braso nga suspended mechanical structure, integrado ang autonomous crawling, obstacle negotiation, online power harvesting, ug multi-fault diagnosis. Ang layo mao ang pag-automate ug intellectualize sa line inspection, kasagaran nga pag-improve sa efficiency ug safety sa grid operation ug maintenance samtang pagbawas sa cost.

I. Project Background and Objectives

1.1 Background: Challenges of Traditional Inspection Methods

Ang high-voltage transmission lines, tungod kay sila nagsulob sa outdoor environment, prone sa mga defect sama sa broken strands ug wear tungod sa mechanical tension, electrical flashover, ug material aging, kung way regular nga inspeksyon. Ang kasamtangan nga mga metodo naghahadlok sa significant nga bottlenecks:

• Manual Inspection: Labor-intensive, inefficient, high-risk, ug dako kaayo ang constraint sa weather ug terrain.
• Drone Aerial Surveying: High operational cost, limited endurance, subject sa airspace control ug adverse weather, ug challenging sa close-range defect detection.

1.2 Objectives: An Intelligent Inspection Alternative

Kini nga proyekto naglakip sa pag-develop og autonomous inspection robot para sa 110 kV high-voltage transmission lines nga makapalit sa manual labor. Ang core objectives mao ang:

• Functional Autonomy: Makamit ang autonomous crawling ug precise obstacle negotiation (e.g., crossing vibration dampers ug clamps) sa lines.
• Intelligent Detection: Integrate visual ug infrared sensors aron ma-identify ug diagnose ang typical faults sama sa broken strands.
• Energy Self-Sufficiency: Gamiton ang non-contact inductive power harvesting technology aron makamit ang online self-replenishment, enabling long-distance inspection.
• Maximized Efficiency: Dugay pa ang pag-improve sa inspection efficiency ug data accuracy, samtang pagbawas sa operational costs ug safety risks.

II. Core Technical Solutions

2.1 Innovative Mechanical Structure Design: High Mobility and Stability

• Overall Structure: Nag-adopt og tulo-ka braso nga suspended configuration nga nag-combine sa advantages sa multi-segment separated ug wheel-arm composite mechanisms, balancing ang efficiency sa wheeled movement sa stability sa inchworm-like creeping. Total weight is approximately 29 kg.
• Key Components:
• Flexible Arms: Ang front ug rear arms nag-employ og double four-bar linkage mechanism, driven sa total nga 16 motors, allowing independent or coordinated pitch motion with joint stiffness-flexibility smooth transition capability to adapt to complex line conditions.
• Drive Unit: Naggamit og high-power Swiss Maxon DC motors with center-separated drive wheels, providing strong obstacle-crossing ability (capable of passing vibration dampers) ug gradeability (routine 60°, up to 80° with braking).
• Braking Unit: Nag-employ og spiral-crank slider self-locking mechanism aron matubag ang accidental slipping or falling during slope traversal or obstacle negotiation.
• Kinematic Validation: Inverse kinematics analysis based on the CCD iterative algorithm; simulations show convergence in only 7 iterations, efficiently validating the robot’s ability to achieve complex poses such as crossing suspension clamps and 45° turn jumpers.

2.2 Hierarchical Intelligent Control System: Seamless Autonomy and Remote Control

• System Architecture: Nag-adopt og tulo-ka layer nga distributed control structure (upper ground management layer, middle robot planning layer, lower execution layer), coordinated by a PC/104 industrial computer ug ATmega128AU microcontroller for real-time decision-making ug execution.
• Hybrid Control Strategy:
• Autonomous Mode: Offline path planning based on a pre-set knowledge base, combined with real-time sensor feedback for fully autonomous crawling ug obstacle negotiation.
• Remote Control Mode: Sa extremely complex environments, ground operators can perform joint-level fine manipulation or issue macro-commands via remote intervention, supported by HD video (25–30 Hz) transmitted from the robot.
• Performance Metrics: Single inspection distance ≥ 2 km, average speed ≥ 0.9 m/h, image transmission distance ≥ 2 km.

2.3 Online Inductive Power Harvesting & Intelligent Power Management: Unlimited Endurance

• Power Harvesting Principle: Naggamit og split-core current transformer aron makakuha og energy gikan sa magnetic field around the high-voltage conductor. The CT core is made of high-permeability iron-based nanocrystalline alloy; an optimized design enables a low starting current of 32 A.
• Power System: Nag-deliver og stable rectified voltage; output power covers a line current range of 32 A to 10 kA. Equipped with a 24 V/12 A·h intelligent Li-ion battery pack using a three-stage charging algorithm, with over-temperature protection for safety, efficiency, ug long service life.

2.4 Machine Vision Obstacle Recognition: Accurate Navigation

• Recognition Targets: Accurately identifies key obstacles such as suspension clamps, straight-line jumper clamps, ug turn jumper clamps.
• Algorithm Flow:
• Positioning: Coarse positioning via sub-block grayscale analysis, precise identification of the transmission line via histogram equalization ug threshold segmentation.
• Feature Extraction: Extracts obstacle contours using morphological operations, analyzing left/right edge slopes as classification features.
• Recognition: Applies a fuzzy pattern recognition algorithm based on the maximum membership principle for fast ug accurate obstacle type identification.
• Performance: Single image processing time ≈ 108 ms; reliably identifies typical obstacles, providing real-time input for obstacle-negotiation decisions.

2.5 Broken Strand Intelligent Diagnosis: Accurate Fault Warning

• Detection Principle: Based on the phenomenon of localized resistance increase ug temperature rise due to broken strands, uses an infrared sensor to detect thermal radiation signals.
• Intelligent Diagnosis Model:
• Signal Processing: Uses the db4 wavelet base for 6-layer decomposition to filter out noise ug focus on frequency bands containing fault features.
• Feature Extraction: Introduces wavelet energy entropy to characterize signal complexity, combined with peak-to-peak values of detail components, forming a 4-dimensional feature vector.
• Diagnosis Decision: Uses a 3-layer BP neural network for diagnosis. Experimental verification shows 100 % accuracy on test samples ug a 98 % online detection success rate.

III. Solution Advantages Summary

• High Adaptability: Tulo-ka braso nga flexible structure provides excellent obstacle negotiation ug terrain adaptability.
• High Autonomy: Hybrid control system enables long-distance autonomous inspection with remote intervention capability.
• Long Endurance: Innovative online power harvesting fundamentally solves endurance limitations.
• Accurate Detection: Integration of machine vision ug infrared thermography with intelligent algorithms ensures high fault-recognition accuracy.
• Safe and Economical: Replaces high-risk manual work, reducing safety hazards ug long-term operational costs.

IV. Current Limitations and Future Prospects

4.1 Current Limitations

• Still requires minimal manual assistance in extremely complex line environments.
• Potential for further optimization of mechanism size ug obstacle-negotiation stroke for a more compact design.
• Power system starting current remains relatively high, limiting application on very low-load lines.
• Current fault detection types are mainly focused on broken strands; the range of detectable faults can be expanded.

4.2 Future Outlook

• Mechanism lightweighting ug balance optimization to improve obstacle-negotiation efficiency ug stability.
• Integration of multi-sensor navigation to enhance positioning ug environmental perception accuracy.
• Optimization of the power harvesting circuit to further reduce the starting current ug expand the application range.
• Expansion of the fault diagnosis library to include defects such as damaged insulators ug contamination.
• Enhancement of the robot’s reliability, improving industrial-grade protection (e.g., dustproof, waterproof, ug EMC capabilities).