An attitude estimation hardware simulation system is realized by using MEMS gyroscopes and MEMS accelerometers

which is based on error model of gyroscope and accelerometer. The system can be used to simulate three-axis strapdown inertial navigation system（ INS） under the situation of any noise intensity and installation error

that is to output data of gyroscopes and accelerometers according to given motion curve

in order to supply simulation verification platform for designing attitude estimation algorithm. Meanwhile

an attitude estimation algorithm which is based on extended Kalman filter is designed. In the algorithm

the attitude quaternion is chosen as state variable

the pitching angle and roll angle are chosen as observed variables. As a result

estimation error of pitching angle is less than 0. 04°

estimation error of roll angle is less than 0. 05°

and the drift rate of yaw angle is0. 01（ °）/s.