Innovative Application and Technological Breakthrough of Multi-node Pressure Sensing Array in Precision Robot Control

Abstract

Traditional robotic tactile sensing systems suffer from low node density (≤8 nodes/cm²), limited pressure resolution (≥0.05 N), slow dynamic response (≥5 ms) and poor sensing-control coordination. This paper presents a 24-node high-density pressure sensing array based on CNT/PDMS composite sensitive layer, and builds a “perception-cognition-execution” precision control system. By optimizing sensing unit parameters (50 μm sensitive layer, 100 μm electrode spacing, 3 wt% CNT doping), establishing a CNT conductive network percolation model, and designing a signal conditioning circuit (CMRR ≥140 dB@1 kHz), the pressure distribution gradient is first introduced as the third-dimensional input of fuzzy PID, constructing a pressure-position coupled 3D fuzzy decision space. This breaks the single-dimensional limitation of traditional fuzzy PID, achieving 0.008 N pressure resolution and ≤1.5 ms dynamic response. A precision assembly platform was built, and in 0402 electronic component assembly (1.0 mm×0.5 mm×0.5 mm, 8 mg), the system achieved ±0.012 mm (3σ) repeat positioning accuracy and 99.6% assembly success rate, outperforming commercial systems (±0.025 mm, 85.3%). Verified by 10⁶ cycle durability tests, -20℃~60℃ environmental tests and 30-day industrial validation, the system shows excellent stability and practicability. This research provides a high-performance tactile control solution for semiconductor packaging and MEMS assembly, with 4 authorized patents and 3 software copyrights, boasting important academic and industrial value.