Research on Multi-Channel Signal Acquisition System of Mechanical Manpower Sensor Based on Dynamic Calibration Matrix and Temperature Algorithm Compensation

This paper proposes an embedded solution based on dynamic calibration matrix and real-time temperature compensation algorithm to solve the signal coupling and temperature drift problems of multi-dimensional force/torque sensors in complex environments. Traditional linear calibration methods perform poorly under complex working conditions such as dynamic loading and multi-axis coupling, and are difficult to cope with environmental changes such as temperature fluctuations. To address these issues, this paper designs a high-precision, low-cost, low-power, and compact six-dimensional force/torque sensor data acquisition system. The system uses the AD7124 analog-to-digital converter for signal acquisition and employs the STM32 controller for real-time processing, combined with a dynamic calibration matrix model and temperature compensation algorithm, to achieve real-time signal processing and high-precision decoupling. Experimental results show that the system's dynamic decoupling error is less than 1%, and the measurement error after temperature compensation is controlled within 1.6%. While balancing precision, speed, power efficiency, and miniaturization, the system provides reliable support for humanoid robot control and intelligent manufacturing.