Mobile Robot State Estimation With IMU Bias Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing mobile robot navigation systems face challenges in providing accurate and complete state estimation due to measurement noises and sensor biases, requiring excessive computation time and memory, and existing methods fail to correct these issues effectively.
Innovation Solution
A state estimation method involving an inertial measurement unit, localization apparatus, field-programmable gate array, application-specific integrated circuit, and processing module, which performs vector padding, normalization, alignment error computation, filter measurement, bias estimation, velocity estimation, and orthogonalization processing to reduce measurement noises and correct sensor biases.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple long-iteration algorithms (neural network, data-driven method or extended Kalman filter) are used to combine IMUs and localizing apparatuses for state estimation, then complete state estimation is achieved, but excessive computation time and memory are required
Solution Approach 1:
The patent segments the state estimation process into distinct computational modules: a first processing unit that processes IMU data to generate preliminary estimates, and a second processing unit that processes localization apparatus data to generate correction values. This segmentation allows parallel processing and reduces overall computation time while maintaining complete state estimation accuracy.
Solution Approach 2:
The patent performs preliminary state estimation using IMU data before incorporating localization apparatus corrections. The first processing unit generates initial estimates of position, orientation, velocity, and acceleration from IMU measurements, which are then refined by the second processing unit using localization data. This preliminary action reduces the computational burden on the final estimation step.
2Ease of operation
If traditional sensors (Doppler or laser radar) are used to measure translational velocity and angular velocity for feedback control, then velocity measurements are obtained, but measurement noises and sensor biases are introduced
Solution Approach 1:
The patent merges data from multiple sensor sources (IMU and localization apparatus) to estimate velocity and correct biases. The first processing unit derives velocity estimates from IMU acceleration data, while the second processing unit uses localization apparatus measurements to compute correction values that reduce measurement noises and sensor biases in the final velocity estimates.
Solution Approach 2:
The patent implements feedback control where the second processing unit continuously compares localization apparatus measurements with IMU-derived estimates, computes bias corrections, and applies these corrections to reduce measurement noises. This feedback mechanism continuously refines the velocity measurements to maintain high precision.
3Productivity
If complementary filters or known non-linear filters are used to estimate pose and translational velocity with low computation cost, then computation time is reduced, but accurate complete state estimation cannot be provided and measurement noises and sensor biases remain
Solution Approach 1:
The patent divides the filtering process into two efficient segments: a first filter (complementary or Kalman filter) in the first processing unit that processes IMU data at high speed, and a second filter in the second processing unit that processes localization apparatus data to generate bias corrections. This segmentation achieves complete state estimation with low computation cost by using simple filters rather than complex multi-algorithm approaches.
Solution Approach 2:
The patent applies partial filtering action where the first processing unit performs sufficient filtering on IMU data to obtain usable estimates, and the second processing unit performs targeted filtering on localization data specifically for bias correction. This partial action approach avoids the excessive computation of comprehensive filtering while still achieving accurate complete state estimation.
Data Source
AI summary
A state estimation method for mobile robot navigation is provided which includes: step S1, initializing the processing module, and then controlling the field-programmable gate array to obtain measurement values of a moment k collected by the inertial measurement unit and the localization apparatus in a constant sampling period; step S2, controlling the application-specific integrated circuit to send the measurement values to the processing module; step S3, controlling the processing module to, based on the measurement values, sequentially perform vector padding, additional inertial vector acquisition, normalization processing, alignment error computation, filter measurement value computation, bias estimation value computation, velocity estimation value computation, pose estimation value computation and orthogonalization processing to obtain state estimation values of the moment k; step S4, controlling the processing module to send the state estimation values to the human-machine interface for visual display.