A Comparative Study of All-Accelerometer Strapdowns for UAV INS
Abstract
We analyze the possibility of using arrays of accelerometers as Inertial Navigation Systems (INS) for Unmanned Aerial Vehicles (UAV). Benefiting from the fabrication processes of MEMS technologies, accelerometers now offer several advantages over gyroscopes, such as low weight, compactness, high reliability and low cost, for example. An algorithm is introduced which allows the computation of the angular acceleration and the angular velocity of a UAV from measurements of n uniaxial accelerometers located at n arbitrary points on the UAV. By uniaxial accelerometer we mean a sensor that allows the measurement of only one point-acceleration component. It is shown that the necessary conditions for the algorithm to work are that n be greater than or equal to nine, that the points be non-collinear, and that the vectors pointing in the sensitive directions of all the sensors span the whole 3D space. Notice that such analyses have been performed before, but only for specific strapdowns and never for a completely general case, in which the positions and the orientations of the sensors are arbitrary, except for the conditions stated above. In the linear algebraic system of 3n equations in 2n + 9 unknowns that has to be solved to compute the angular acceleration and the angular velocity, the coefficients of the unknowns depend only on the geometry of the accelerometer strapdown, and not on the measurements. Therefore, the condition number of the algebraic linear system gives an upper bound of the error amplification incurred when solving the linear system for any accelerometer measurements, which makes the condition number a suitable performance index for the design of UAV accelerometer strapdowns. With regard to this new performance index, a comparison of the accuracies of existing accelerometer strapdowns in conducted. Moreover, an error analysis is provided, taking into account stochastic errors-noise-and deterministic errors-bias-in the sensor outputs.
Document Details
- Document Type
- Technical Report
- Publication Date
- May 01, 2005
- Accession Number
- ADA471970
Entities
People
- Jorge Angeles
- Philippe Cardou
Organizations
- McGill University