CN-121977611-A - Inertial motion measurement self-rotation excitation enhancement method and system based on observability analysis

CN121977611ACN 121977611 ACN121977611 ACN 121977611ACN-121977611-A

Abstract

The present disclosure provides an inertial motion measurement self-rotation excitation enhancement method and system based on observability analysis. The invention utilizes a motor to drive an Inertial Measurement Unit (IMU) to rotate, a self-rotation excitation module formed by the motor and the IMU is bound on a motion main body, an observable degree change rule of an IMU error state is analyzed, a rotation excitation control rule is generated according to the observable degree change rule, and the IMU is actively driven to automatically rotate in a non-zero speed interval based on the rotation excitation control rule in an inertial navigation positioning process assisted by zero speed updating ZUPT, so that the IMU generates attitude change in an adjacent zero speed interval, and the observable degree of the IMU error state is actively enhanced. The invention can promote the observability of the system state by active excitation and inhibit the accumulation of errors on the premise of not depending on an external sensor, thereby improving the accuracy of motion measurement and state estimation of the system running for a long time.

Inventors

DENG ZHIHONG
REN SIZHU
LI ZHE
ZHANG PING

Assignees

北京理工大学

Dates

Publication Date: 20260505
Application Date: 20260129

Claims (9)

1. The inertial motion measurement self-rotation excitation enhancement method based on observability analysis is characterized in that a motor is used for driving an Inertial Measurement Unit (IMU) to rotate, and a self-rotation excitation module formed by the motor and the IMU is bound on a moving main body; analyzing the observability change rule of the IMU error state, and generating a rotary excitation control law according to the observability change rule; In the inertial navigation positioning process assisted by the zero-speed updating ZUPT, the IMU is actively driven to automatically rotate under control based on the rotation excitation control law in a non-zero-speed interval, so that the IMU generates attitude change in an adjacent zero-speed interval, and the observability of an error state of the IMU is actively enhanced.
2. The method of claim 1, wherein the inertial measurement unit IMU is driven to rotate by a motor, and the self-rotation excitation module formed by the motor and the IMU is bound to the moving body, and specifically comprises: Driving the IMU to rotate around the x-axis of the machine body coordinate system by utilizing a motor; The x-axis of the IMU body coordinate system is parallel to the pitch axis of the moving body.
3. The method of claim 1, wherein analyzing the observability change law of the IMU error state and generating a rotational excitation control law therefrom is: based on a piecewise linear steady system and singular value decomposition, the observability change rule of the error state of the IMU is analyzed as follows: (1) Wherein, the Is the observability of the gyro in three directions, The acceleration of the gravity is constant and the acceleration of the gravity is constant, In order to sample the time of the sample, The rotation angle of the IMU is set; The expression of the formula (1) shows that the observability degree of zero offset of the x-axis gyroscope is kept constant and is irrelevant to the rotation angle, which indicates that the error excitation of the x-axis is not coupled with kinematics in the rotation process, and the observability degrees of the y-axis and the z-axis respectively form a cosine square and sine square relation with the rotation angle, when one axis observability degree reaches a peak value, the other axis attenuates to a theoretical minimum value; According to the formula (1), the rotation excitation control law is designed, wherein the x-axis of an IMU machine body coordinate system is parallel to the pitching axis of the moving main body, and the IMU is controlled to rotate automatically within a set rotation angle range according to a set rotation speed in a non-zero speed range.
4. A method according to claim 3, wherein controlling the IMU to self-rotate within a set rotation angle range at a set rotation speed comprises: In the non-zero speed interval, the IMU automatically rotates according to the set rotating speed, namely, when the non-zero speed interval is ended but the rotating angle extreme value of the current rotating direction is not reached, the IMU stops rotating and continues to automatically rotate according to the current direction when the next non-zero speed interval arrives, and when the rotating angle extreme value of the current rotating direction is reached but the non-zero speed interval is not ended, the IMU stops rotating and waits for the next non-zero speed interval to arrive, the IMU reversely and automatically rotates.
5. The method of claim 3 or 4, wherein the set rotational speed is 10 degrees per second and the set rotational angle range is plus or minus 95 degrees.
6. The method of claim 1, wherein the actively driven IMU is controlled to self-rotate based on the rotational excitation control law using a dual closed loop control law in combination with a position loop for planning rotational angle and a speed loop for controlling the controlled rotational speed.
7. The method of claim 1, wherein the self-rotating excitation module is bound to the instep, leg or waist.
8. The inertial motion measurement self-rotation excitation enhancement system based on observability analysis is characterized by comprising an inertial measurement unit IMU, a carrier motion state detection unit, a rotation control unit, a ZUPT-assisted inertial navigation positioning unit and a communication unit; The IMU is used for acquiring triaxial angular velocity and acceleration information; the carrier motion state detection unit is used for detecting a non-zero speed interval and a non-zero speed interval according to the acquired data of the IMU; The rotation control unit comprises a motor and a rotation excitation control module; the rotation excitation control module is connected with the motor, and adopts a rotation excitation control law generated based on the observation degree change rule analysis of the IMU error state, and in a non-zero speed interval, the motor is actively controlled to move so as to drive the IMU to automatically rotate under control, so that the IMU generates gesture change in an adjacent zero speed interval, and the observability of the IMU error state is actively enhanced; The ZUPT-assisted inertial navigation positioning unit utilizes inertial measurement data including the controlled rotation period of the IMU to estimate and correct the error state of the IMU in a state estimation filter and outputs an inertial motion measurement result; The communication unit is used for realizing time synchronization and data interaction among the IMU, the rotation control unit and the ZUPT-assisted inertial navigation positioning unit.
9. The system of claim 8, wherein the motor drives the IMU to rotate about an x-axis of the body coordinate system, the x-axis of the IMU body coordinate system being parallel to a pitch axis of the moving body; The rotation excitation control law adopted by the rotation excitation control module is that in a non-zero speed interval, the IMU is controlled to rotate automatically within a set rotation angle range according to a set rotation speed, when the non-zero speed interval is ended and the rotation angle extreme value of the current rotation direction is not reached, rotation is stopped, when the next non-zero speed interval arrives, the rotation continues to rotate automatically according to the current direction, and when the rotation angle extreme value of the current rotation direction is reached and the non-zero speed interval is not ended, rotation is stopped, and when the next non-zero speed interval arrives, reverse rotation is stopped.

Description

Inertial motion measurement self-rotation excitation enhancement method and system based on observability analysis Technical Field The invention relates to the technical field of inertial motion measurement, in particular to an inertial motion measurement self-rotation excitation enhancement method and system based on observability analysis. Background The inertial measurement is an autonomous positioning and attitude measurement technology for estimating the motion state of a carrier by utilizing angular velocity and acceleration information output by an inertial measurement unit. The technology has the advantages of complete autonomy, strong anti-interference capability and capability of working under the condition of no external signal, thereby being widely applied in the fields of aerospace, vehicle navigation, robot positioning, pedestrian inertial navigation, motion monitoring and the like. However, the accuracy of the long-term operation of inertial measurement systems is limited by inertial device measurement errors. Since these errors cannot be directly observed inside the system, their estimation and compensation depends on the observability of the system. If the system lacks sufficient motion stimulus during operation, the partial error state quantities are less appreciable, so that the filter cannot effectively estimate the errors, resulting in rapid error accumulation. In many inertial measurement applications, the system uses intermittent static characteristics of the carrier to construct an observation model to estimate and compensate for navigation errors. For example, in foot robots or pedestrian motion measurement, a zero-velocity correction method is widely used, which uses the static characteristics of the foot in the support phase to suppress drift. However, such systems, which rely on static or quasi-static intervals, tend to have inadequate motion excitation. Taking wearable human body movement measurement as an example, when the inertial unit is worn on the foot, the movement in the walking process has stronger homomorphism, namely the gesture change between adjacent zero-speed intervals is single. The singleness of the motion mode causes partial error state quantity of the system, especially zero offset of a gyroscope, to be insufficiently excited, and the observability is seriously insufficient. This phenomenon can directly lead to the fact that the error state cannot be effectively estimated and compensated in the filtering process, and finally causes drift of the posture and the position, so that the measurement accuracy is obviously reduced after long-time operation. The existing technology for improving the observability mainly comprises the following improvement schemes: one is to introduce external information such as magnetometers, ultra-wideband signals, visual information, etc. to provide additional viewing information. However, the method is easily affected by magnetic field interference, signal shielding, illumination change and the like, and the system has high cost and complex deployment. Secondly, error propagation is constrained by establishing a kinematic model, such as a robot kinematics model or a human gait model, but the model depends on idealized assumptions, and consistency is difficult to maintain under different movement modes, complex terrains or changeable use conditions. And thirdly, improving a filtering algorithm, for example, adopting methods such as extended Kalman filtering, self-adaptive noise adjustment and the like, wherein the filtering algorithm is difficult to compensate the problem of excitation deficiency under the condition of insufficient observability of a system. Therefore, on the premise of not depending on an external sensor, the system state observability is improved through active excitation, and the method is a key direction for realizing high precision and long-term stability of various autonomous inertial measurement systems. At present, active excitation of an inertial system is still less studied, and a self-rotation excitation design and control method based on observability theoretical analysis is especially lacking. Disclosure of Invention In view of this, in order to overcome the problem that in the prior art, particularly in an inertial measurement system relying on intermittent static correction, the observability of part of error states is poor due to single carrier motion mode or strong repeatability, the invention provides an inertial motion measurement self-rotation excitation enhancement method and system based on observability analysis, which can improve the observability of the states of the system and inhibit error accumulation by active excitation on the premise of not relying on an external sensor, thereby improving the accuracy of motion measurement and state estimation of the system running for a long time. The method is suitable for autonomous motion measurement scenes which need long-time high-precision estima