Key Factors Affecting IMU Performance

The Inertial Measurement Unit (IMU) delivers irreplaceable core value for navigation systems. Additional sensors on each single axis can be equipped to improve reliability. To guarantee measurement accuracy, IMUs are generally mounted at the center of gravity of the measured carrier.

As autonomous driving gains wider attention, high-precision positioning and sensing technologies achieve rapid progress. IMU technology serves as a key supplement to GPS and works together with satellite navigation to provide precise positioning data for autonomous vehicles.

IMU is widely used in smartphones, automobiles, aircraft, missiles and spacecraft. The main differences among these applications lie in cost and precision. For example, when vehicles pass through tunnels where satellite signals are blocked, inertial measurement technology continues to supply speed, distance and time information for navigation.

IMU (Inertial Measurement Unit) consists of three single-axis accelerometers and three single-axis gyroscopes. Accelerometers measure linear acceleration along three orthogonal axes in the carrier coordinate system, while gyroscopes detect angular velocity relative to the navigation frame. Carrier attitude can be calculated after processing these measurement signals.

It should be noted that IMU only outputs relative positioning information and tracks motion trajectories starting from an initial point. It cannot provide absolute geographic coordinates independently, so it is commonly integrated with GPS. When satellite signals weaken or get lost, IMU enables continuous positioning to avoid navigation failure.

Inertial technology controls the attitude and trajectory of moving objects. It covers inertial instruments, inertial stabilization, inertial systems, inertial guidance and inertial measurement. This interdisciplinary subject integrates physics, mathematics, mechanics, optics, materials science, precision machinery, electronics, computer science, control engineering, measurement, simulation and manufacturing. It researches the theory, design, production, testing, application and maintenance of inertial instruments and systems. Its application scenarios include aerospace, land navigation, geodetic survey, tunnel drilling, geological exploration, robotics, vehicles, medical equipment, cameras, mobile phones and consumer toys. Any system requiring motion sensing, positioning or orientation relies on inertial technology.

As the core information source of modern precision navigation, guidance and control systems, inertial technology plays an indispensable supporting role in building multi-dimensional information systems and promoting the integration of mechanization and informatization for military equipment.

Primary Factors Affecting IMU Performance

Main Error Sources of MEMS‑IMU

IMU errors are divided into four categories as follows.

1. Accelerometer-related factors Accelerometer performance affects IMU mainly through precision and stability. High precision ensures accurate subsequent data processing, while stability determines whether the IMU works normally. The 6-position static calibration method is widely used to test accelerometer precision. After IMU assembly, data of three-axis accelerometers are collected at six predefined mounting orientations.
2. Gyroscope-related factors Gyro precision directly decides the quality of attitude calculation, namely the ability of IMU to capture carrier attitude. The gyro error model is similar to that of accelerometers, calibrated by dynamic rotation tests. The IMU is installed on a single-axis turntable. Each axis points upward and downward, rotating clockwise and counterclockwise at speeds of 50°/s, 100°/s, 150°/s, 200°/s and 250°/s for sufficient data sampling.

Besides precision, key MEMS gyro indicators include scale factor, scale factor nonlinearity, bias, bias stability and bias repeatability. These parameters comprehensively characterize gyro performance and require dedicated testing.

3. Temperature factors Temperature variation causes obvious precision fluctuation of MEMS inertial sensors, especially for gyroscopes. Working environments are rarely kept at constant temperature, so thermal effects cannot be ignored. Tests are carried out on a temperature-controlled turntable to acquire measurement data under different temperature conditions.
4. Product-level influencing factors after IMU packaging A. Low signal-to-noise ratio Low signal-to-noise ratio reduces the sensitivity of IMU-based devices, making noise suppression a critical challenge. Wavelet denoising is adopted to suppress useless noise components and strengthen valid signals. The standard wavelet denoising workflow: a. Wavelet decomposition: select proper wavelet basis and decomposition layers to decompose raw signals. b. Threshold quantization of high-frequency coefficients: apply soft threshold quantization to high-frequency coefficients at each decomposition scale. c. Wavelet reconstruction: reconstruct one-dimensional signals using low-frequency coefficients of the deepest layer and processed high-frequency coefficients. Threshold selection and quantization dominate denoising quality.

B. Severe drift and signal latency Signal latency commonly exists in MEMS devices and becomes prominent in IMU products. Inertial error rotation modulation is proposed to mitigate this issue. It is essentially an error self-compensation technology, which modulates slowly varying inertial errors by periodic IMU rotation. This method improves navigation accuracy without upgrading intrinsic device precision. Rotation-modulated INS has a modified mechanical structure. A rotary mechanism separates accelerometers and gyroscopes from the carrier, while strapdown algorithms are still used for navigation resolving. Such systems are named rotation-modulated strapdown inertial navigation systems.