Three-axis difference implementation method based on cross symmetry of double inertial sensors

Abstract

The invention discloses a three-axis differential implementation method based on cross symmetry of double inertial sensors, and the method comprises the following steps: S1, a sensor is an MEMS inertial measurement unit, i.e., an IMU, the IMU uses a right-hand Cartesian coordinate system to define the directions of three sensitive axes of the IMU, and the IMU is used for defining the positive direction of an angular velocity and also used for defining the positive direction of a linear acceleration; s2, in order to realize three-axis difference, rotating the sensor around the y axis by 180 degrees; s3, a result obtained in the step S2 is rotated by 90 degrees around the z axis; and S4, after the step S3, realizing symmetry of the x axis and the y axis through crossing, and realizing three-axis difference through double IMUs by applying a three-axis symmetry principle of an inertial sensor.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of sensors, and particularly relates to a three-axis differential implementation method based on cross symmetry of double inertial sensors. BACKGROUND

[0002] It is known that, in order to improve the precision and reliability of an inertial measurement system, a multi-sensor differential measurement technology has become a mainstream method for suppressing common-mode interference and eliminating zero-bias drift. In particular, in the field of high-precision data acquisition, parallel and differential operation of multiple inertial measurement units has become a necessary means to improve motion perception performance.

[0003] However, the existing multi-IMU differential scheme generally adopts a coaxial alignment arrangement, that is, the corresponding sensitive axes (X and X, Y and Y, Z and Z) of the two IMUs are strictly parallel or anti-parallel arranged in the physical space to realize axial differential. This mode can basically meet the demand in the ideal case and simple application scenarios, but its inherent defects are increasingly prominent in modern industrial applications that pursue high precision, high stability and complex environment adaptation.

[0004] Firstly, the core problem of the coaxial differential scheme is that the "like-named axes" of the two IMUs have strong correlation in inherent zero bias, temperature drift, scale factor error and the like due to the high consistency of their manufacturing process, material and production batch. Such correlated errors, as "pseudo common-mode signals", cannot be effectively suppressed in differential operation, resulting in significant residual errors after differential, which limits the upper limit of precision improvement.

[0005] Secondly, to realize three-axis full differential, the traditional method requires multiple IMUs to be precisely aligned or reversely installed in three orthogonal axes. This not only puts high requirements on the machining precision and assembly process of the mechanical structure, increasing the manufacturing cost, but also introduces significant inter-axis coupling errors in the differential output due to slight installation deviation, further deteriorating the measurement precision.

[0006] Furthermore, in application scenarios with severe temperature changes, the response characteristics of the sensors arranged coaxially to the environmental temperature field are extremely similar, and the output changes caused by temperature drift are highly consistent, making it difficult for the traditional differential technology to effectively isolate and eliminate the measurement interference caused by temperature.

[0007] The existing inertial sensor differential scheme mainly includes two types: 1) a multi-sensor array scheme, which deploys a large number of sensors (such as 32 consumer-grade MEMS-IMUs) to form an array, and fuses the multi-channel output data to improve the overall measurement precision; and 2) a coaxial reverse differential scheme, which strictly arranges the corresponding axes (X and X, Y and Y, Z and Z) of multiple inertial sensors in parallel or anti-parallel in the physical space, and suppresses common-mode errors through direct differential operation.

[0008] Although the above scheme is effective in certain scenarios, there are still the following limitations:

[0009] (1) Large number of sensors, high cost: The array scheme requires a large number of sensors (such as 32) to work together, which significantly increases the hardware complexity and power consumption.

[0010] (2) Strict installation accuracy requirement: The coaxial differential scheme requires the corresponding axes of the two sensors to be strictly aligned, and small installation deviation will introduce inter-axis coupling error, reducing the differential effect. For example, traditional back-to-back installation requires more than three times of accurate rotation positioning, which has very high requirements for mechanical processing and assembly process.

[0011] (3) Cannot realize absolute position symmetry: The existing method requires at least three sensors to realize three-axis full differential, but multiple sensors are difficult to arrange symmetrically in physical space, which limits the common mode rejection effect.

[0012] Therefore, there is an urgent need for an innovative method with low cost, which can fundamentally break through the technical limitations of traditional coaxial differential, solve the problems of high error correlation, complex installation, and poor environmental adaptability, and fully utilize the theoretical advantages of double-sensor differential measurement, to meet the increasing performance and reliability requirements of modern industrial systems. SUMMARY

[0013] The purpose of the present application is to solve the above problems, and to provide a three-axis differential implementation method based on double-inertial sensor cross-symmetry, which is low-cost, simple, and can overcome the strong error correlation in traditional coaxial differential, eliminate the dependence on high-precision installation, and reduce the process requirements.

[0014] To solve the above technical problems, the technical scheme of the present application is as follows: a three-axis differential implementation method based on double-inertial sensor cross-symmetry, comprising the following steps:

[0015] S1, the sensor is a MEMS inertial measurement unit, IMU, the IMU uses a right-handed Cartesian coordinate system to define the direction of its three sensitive axes, both for defining the positive direction of angular velocity and for defining the positive direction of linear acceleration;

[0016] S2, in order to realize three-axis differential, the sensor is rotated 180° around the y-axis;

[0017] S3, the result obtained in step S2 is rotated 90° around the z-axis;

[0018] S4, after step S3, the x-axis and y-axis are symmetrically achieved by crossing, and the three-axis symmetry principle of inertial sensor is applied, and three-axis differential is realized by double IMU.

[0019] Further, the IMU in S1 comprises a three-axis accelerometer or / and a three-axis gyroscope.

[0020] Further, after the rotation in S2, the x-axis and the z-axis are reversed compared with the x-axis and the z-axis before the rotation, forming a differential pair, while the y-axis is in the same direction as the y-axis before the rotation.

[0021] Further, the z-axis after the rotation in S3 forms a differential pair with the z-axis in S1, and the x-axis forms a differential pair with the y-axis in S1.

[0022] Further, when the three-axis differential implementation method based on the cross symmetry of the two inertial sensors is applied, the two sensors are arranged on the top layer and the bottom layer of the same PCB, i.e. the front side and the back side, so as to realize the absolute symmetric architecture with the centers of the front and back sensors coinciding, thereby realizing the function of three-axis differential through the two sensors.

[0023] Further, the three-axis differential implementation method based on the cross symmetry of the two inertial sensors can realize the function of attitude measurement or inertial navigation positioning in combination with the processor, and when applied, the bottom IMU communicates with the top processor through a through hole, the processor controls the working state of the top and bottom IMUs through the control line, and receives the output data of the top and bottom IMUs through the data line.

[0024] Further, the processor performs double-IMU three-axis differential processing on the received IMU data, and the processing method is shown in formula (1):

[0025] (1) ;

[0026] wherein DIFF_IMU represents a virtual IMU obtained according to the differential of the top IMU and the bottom IMU, DIFF_IMU.X, DIFF_IMU.Y and DIFF_IMU.Z represent the X, Y and Z axis data of the virtual IMU respectively, TOP_IMU.X, TOP_IMU.Y and TOP_IMU.Z represent the X, Y and Z axis data of the top IMU respectively, and BOTTOM_IMU.X, BOTTOM_IMU.Y and BOTTOM_IMU.Z represent the X, Y and Z axis data of the bottom IMU respectively; the processor directly uses DIFF_IMU.X, DIFF_IMU.Y and DIFF_IMU.Z data for subsequent attitude and inertial navigation applications.

[0027] The beneficial effects of the present application are:

[0028] 1. The three-axis differential implementation method based on the cross symmetry of double inertial sensors provided by the application significantly reduces the system cost and hardware complexity. Only two traditional MEMS inertial sensors are needed to realize complete three-axis differential measurement, compared with the traditional array scheme (32 sensors are needed) or the three-sensor scheme, the hardware cost is significantly reduced, without the need for complex multi-channel synchronous acquisition circuit and precise clock distribution network, significantly reducing the PCB area and layer requirements. The double-sensor scheme also reduces the system power consumption, especially suitable for battery-powered portable devices and Internet of Things terminals.

[0029] 2. The application breaks through the measurement accuracy and anti-interference ability, and through the cross-axis differential technology, the highly correlated sensor error (zero offset, temperature drift) in the traditional coaxial scheme is converted into statistically independent random error, the residual error after differential is significantly reduced, the common mode rejection ratio (CMRR) is better than that of the traditional coaxial differential scheme, and the attitude solution accuracy in strong vibration environment can be significantly improved. The cross-symmetry layout effectively balances the temperature gradient effect, and can significantly reduce the temperature drift error.

[0030] 3. The application greatly reduces the manufacturing process requirements and installation cost, only two standard rotation operations are needed to complete the installation, compared with the traditional three-axis precise alignment scheme, the installation time is greatly shortened. Through the software coordinate mapping algorithm to compensate for the small installation deviation, eliminate the time-consuming optical alignment and laser calibration link in the traditional scheme, the production efficiency is improved significantly. It can be realized in the standard SMT mounting process, which is convenient for large-scale automatic production, and the product consistency and reliability are fundamentally guaranteed.

[0031] 4. The application can realize the absolute position symmetry layout, two sensors are installed in the center of the front and back of the PCB, which eliminates the spatial asymmetry in the traditional multi-sensor layout from the root. Ensure that the vibration, impact and other environmental disturbances in any direction act on the two sensors at the same time and with the same amplitude, providing ideal input conditions for differential processing. The compact symmetric layout improves the overall structural strength of the module, which is suitable for high dynamic application scenarios.

[0032] 5. The application can provide a low-cost, high-precision and high-reliability inertial measurement solution for unmanned aerial vehicle flight control, robot navigation, platform stabilization, intelligent agriculture, industrial detection and other fields, which has broad market application prospect and important technical popularization value. BRIEF DESCRIPTION OF DRAWINGS

[0033] Figure 1 The sensor rotation method in the three-axis differential implementation method based on the cross symmetry of double inertial sensors is provided by the application;

[0034] Figure 2 The sensor rotation method in the three-axis differential implementation method based on the cross symmetry of double inertial sensors is provided by the application;

[0035] Figure 3 is a schematic diagram of a dual-IMU three-axis differential application of the present application;

[0036] Figure 4 is a schematic diagram of a dual-IMU three-axis differential application example of the present application. DETAILED DESCRIPTION

[0037] The present application is further illustrated below in conjunction with the accompanying drawings and specific embodiments:

[0038] As shown in Figure 1 , the present application provides a three-axis differential implementation method based on the cross symmetry of dual inertial sensors, comprising the following steps:

[0039] S1, the sensor is a MEMS inertial measurement unit, namely IMU, the IMU uses a right-hand Cartesian coordinate system to define the direction of its three sensitive axes, both for defining the positive direction of angular velocity and for defining the positive direction of linear acceleration.

[0040] In step S1, the IMU contains a three-axis accelerometer or / and a three-axis gyroscope. As shown in Figure 1 (a), this rule is used to define both the positive direction of angular velocity and the positive direction of linear acceleration. Figure 1 In (a) of the figure, (a) represents the sensor rotating 180° around the Y axis, (b) represents the sensor rotating 90° around the Z axis, and (c) represents the result after rotation.

[0041] S2, in order to realize three-axis differential, the sensor is rotated 180° around the y axis.

[0042] The rotation direction is shown by the rotation arrow of the y axis in Figure 1 (a), after rotation in step S2, the three-axis coordinates are shown in Figure 1 (b), Figure 1 The x and z axes of (b) are opposite to the x and z axes before rotation, forming a differential pair, while the y axis is the same as the y axis before rotation.

[0043] S3, the result obtained in step S2 is rotated 90° around the z axis.

[0044] The rotation direction is shown by the rotation arrow of the z axis in Figure 1 (b), after rotation, the IMU three-axis direction shown in Figure 1 (c) is obtained.

[0045] The z axis after rotation in step S3 and the z axis in S1 form a differential pair, and the x axis and the y axis in S1 form a differential pair. Specifically, the z axis of Figure 1 (c) and the z axis of Figure 1 (a) form a differential pair, and the x of Figure 1 (c) and the y of Figure 1The y-axis composition difference pair of (a) is symmetrical by crossing x and y, which ingeniously applies the three-axis symmetry principle of inertial sensor, and three-axis difference is realized by double IMU. The invention can realize three-axis difference by two sensors through only two rotations, greatly reducing the design complexity and cost of IMU sensor, and improving the noise suppression performance.

[0046] S4, after step S3, the x-axis and the y-axis are symmetrical by crossing, which applies the three-axis symmetry principle of inertial sensor, and three-axis difference is realized by double IMU.

[0047] According to the symmetry of the IMU sensor, three-axis difference is also realized by double inertial sensor, which includes Figure 2 The other two rotation methods are shown. Figure 2 (a) represents that the sensor rotates 90° around the Z-axis, (b) represents that the sensor rotates 180° around the X-axis, (C) represents the rotation result of the sensor, (d) represents that the sensor rotates 180° around the X-axis, (e) represents that the sensor rotates 90° around the Z-axis, and (f) represents the rotation result of the sensor.

[0048] As shown in Figure 3 A three-axis difference implementation method based on double inertial sensor cross symmetry can realize absolute symmetry architecture of the center of the front and back sensors coinciding by arranging two sensors on the top and bottom layers of the same PCB, i.e. the front and back, which cannot be realized by traditional methods, so that three-axis difference function is realized by two sensors.

[0049] A three-axis difference implementation method based on double inertial sensor cross symmetry can realize the function of attitude measurement or inertial navigation positioning in combination with a processor. In application, the bottom IMU communicates with the top processor through a through hole, the processor controls the working state of the top and bottom IMUs through control lines, and receives the output data of the top and bottom IMUs through data lines. The processor includes MCU, DSP, ARM, FPGA or CPU, and application examples are shown as Figure 4 .

[0050] In order to realize double IMU three-axis difference, in addition to the rotation methods shown in Figure 1 and Figure 2 , double IMU three-axis difference processing of the received IMU data is also needed in the processor, and the processing method is shown as formula (1):

[0051] (1).

[0052] Wherein, DIFF_IMU represents a virtual IMU obtained according to top-layer IMU and bottom-layer IMU difference, DIFF_IMU.X, DIFF_IMU.Y and DIFF_IMU.Z represent X, Y and Z axis data of the virtual IMU respectively, TOP_IMU.X, TOP_IMU.Y and TOP_IMU.Z represent X, Y and Z axis data of the top-layer IMU respectively, and BOTTOM_IMU.X, BOTTOM_IMU.Y and BOTTOM_IMU.Z represent X, Y and Z axis data of the bottom-layer IMU respectively; DIFF_IMU.X, DIFF_IMU.Y and DIFF_IMU.Z data are directly used by the processor for subsequent attitude and inertial navigation applications.

[0053] In summary, the three-axis differential implementation method based on the cross symmetry of the dual inertial sensor provided by the application can realize the virtual IMU of three-axis differential by using dual sensors through twice rotation and processor differential operation, so that the system common mode interference, system design complexity and cost can be greatly reduced.

[0054] The application can solve many problems of the traditional method, and has the advantages that:

[0055] (1) The problems of high cost and complex system of the traditional multi-sensor scheme are solved;

[0056] In the prior art, multiple inertial sensors (such as 32 sensor arrays) or a scheme of at least three sensor combinations are usually used to realize three-axis differential measurement, resulting in high system hardware cost, high power consumption and complex circuit structure. The application aims to realize three-axis differential function by using only two traditional MEMS inertial sensors, thereby significantly reducing system cost and complexity.

[0057] (2) The defect of strong error correlation in the traditional coaxial differential is overcome;

[0058] In the traditional coaxial differential scheme, the homonymous axes (X-X, Y-Y and Z-Z) of the two sensors are highly correlated in error characteristics such as zero offset and temperature drift due to consistent manufacturing process, and the residual error after differential is still significant. The application solves the problem of strong error correlation by cross symmetric installation, so that the error characteristics of different physical axes are independent.

[0059] (3) High-precision installation dependence is eliminated, and process requirements are reduced;

[0060] The traditional back-to-back installation method requires multiple sensors to be strictly aligned on three axes, and needs multiple accurate rotation positioning, so that the mechanical processing and assembly precision requirements are extremely high. The application greatly reduces the installation complexity and improves the process tolerance by specific twice rotation installation (180° rotation around the Y axis and 90° rotation around the Z axis).

[0061] (4) Achieve absolute position symmetry, improve common mode rejection effect;

[0062] The traditional multi-sensor scheme cannot realize absolute position symmetrical arrangement of sensors due to physical layout limitation. The application realizes true absolute position symmetry by centering and installing two sensors on the front and back of the PCB, thereby improving the inhibition ability to common mode interference at the root.

[0063] (5) Break through the technical bottleneck of heterogeneous sensor cooperative work;

[0064] The data fusion of different types of inertial sensors in the prior art has problems such as time sequence synchronization difficulty and complex coordinate system. The application realizes seamless cooperative work of two sensors through a unified shaft system mapping algorithm, effectively improving the overall performance of the system.

[0065] By solving the above technical problems, the application realizes the goal of obtaining optimal measurement accuracy at the lowest hardware cost on the basis of maintaining the advantages of traditional differential measurement, and provides an innovative solution for high-precision inertial measurement systems.

[0066] Those skilled in the art will realize that the embodiments described herein are for the purpose of helping the reader to understand the principles of the application and should be understood as not limiting the scope of protection of the application to such specific statements and embodiments. Those skilled in the art can make various other specific modifications and combinations according to the technical inspiration disclosed in the application without departing from the essence of the application, and these modifications and combinations are still within the scope of protection of the application.

Claims

1. A method for implementing triaxial differential based on cross-symmetry of dual inertial sensors, characterized in that, Includes the following steps: S1. The sensor is a MEMS inertial measurement unit, or IMU. The IMU uses a right-handed Cartesian coordinate system to define the directions of its three sensing axes, which are used to define both the positive direction of angular velocity and the positive direction of linear acceleration. S2. To achieve triaxial differential, rotate the sensor 180° around the y-axis; S3. Rotate the result obtained in step S2 around the z-axis by 90°. S4. After step S3, the x-axis and y-axis achieve symmetry through intersection, applying the principle of three-axis symmetry of inertial sensors, and three-axis differential is realized through dual IMUs.

2. The method for implementing triaxial differential based on cross-symmetry of dual inertial sensors according to claim 1, characterized in that: The IMU in S1 includes a three-axis accelerometer and / or a three-axis gyroscope.

3. The method for implementing triaxial differential based on cross-symmetry of dual inertial sensors according to claim 1, characterized in that: After the rotation in S2, the x-axis and z-axis are opposite to the x-axis and z-axis before the rotation, forming a difference pair, while the y-axis is in the same direction as the y-axis before the rotation.

4. The method for implementing triaxial differential based on cross-symmetry of dual inertial sensors according to claim 1, characterized in that: The rotated z-axis in S3 forms a difference pair with the z-axis in S1, and the x-axis forms a difference pair with the y-axis in S1.

5. The method for implementing triaxial differential based on cross-symmetry of dual inertial sensors according to claim 1, characterized in that: In the application of the triaxial differential implementation method based on the cross-symmetry of dual inertial sensors, the two sensors are arranged on the top and bottom layers of the same PCB, i.e. the front and back sides, which can realize an absolutely symmetrical architecture with the center of the front and back sensors coinciding, thus realizing the triaxial differential function with only two sensors.

6. The method for implementing triaxial differential based on cross-symmetry of dual inertial sensors according to claim 1, characterized in that: The aforementioned triaxial differential implementation method based on dual inertial sensors with cross-symmetry, combined with a processor, can realize attitude measurement or inertial navigation and positioning functions. In application, the bottom IMU communicates with the top processor through a through-hole. The processor controls the working state of the top and bottom IMUs through control lines and receives the output data of the top and bottom IMUs through data lines.

7. The method for implementing triaxial differential based on cross-symmetry of dual inertial sensors according to claim 6, characterized in that: The processor performs dual-IMU triaxial differential processing on the received IMU data, as shown in equation (1): （1）； Wherein, DIFF_IMU represents the virtual IMU obtained by differential calculation of the top-level IMU and the bottom-level IMU, DIFF_IMU.X, DIFF_IMU.Y, and DIFF_IMU.Z represent the X, Y, and Z axis data of the virtual IMU, respectively, TOP_IMU.X, TOP_IMU.Y, and TOP_IMU.Z represent the X, Y, and Z axis data of the top-level IMU, respectively, and BOTTOM_IMU.X, BOTTOM_IMU.Y, and BOTTOM_IMU.Z represent the X, Y, and Z axis data of the bottom-level IMU, respectively. The processor directly uses the DIFF_IMU.X, DIFF_IMU.Y, and DIFF_IMU.Z data for subsequent attitude and inertial navigation applications.

Images