A signal selection and processing method for multiple IMU sensor devices

Abstract

The present invention discloses a signal selection and processing method for multiple IMU sensor devices, which relates to the technical field of IMU processing. Aiming at the demand for high-precision and high-reliability IMU data in algorithms such as digital twin mapping, positioning, and three-dimensional reconstruction, the method preferentially selects and processes signals of redundant sensor devices of multiple IMUs, and improves the anti-interference ability and reliability of the system through the redundancy and reasonable arrangement of multiple IMUs. At the same time, through the unified processing of multiple IMU sensor signals, more accurate signal data is finally output.

Description

Technical Field

[0001] The invention discloses a method, and relates to the technical field of IMU signal processing, in particular to a signal selection and processing method of multiple IMU sensor devices. Background Art

[0002] In digital twin applications, IMU is a very important sensor used to obtain basic motion information of actual devices, such as acceleration, angular velocity, and direction. IMU data is indispensable in core mapping and positioning algorithms. In positioning, IMU can provide information such as acceleration and angular velocity to help estimate the pose of the main body; in mapping, IMU can collaborate with other sensors such as lidar and cameras to achieve map construction.

[0003] However, there are some problems with IMU sensors in practical applications. For example, IMU measurement results are greatly affected by temperature, mechanical vibration, electromagnetic interference, etc., and the signal is prone to strong noise, which leads to poor stability of IMU signals and easy drift and jitter. At the same time, since its output signal accumulates over a period of time, it will produce a large cumulative error. The above problems will seriously affect the accuracy of mapping and positioning. At the same time, when using a single IMU, it is difficult to identify sensor failures in a timely manner because the validity of the data cannot be verified, which will also have adverse effects on applications such as mapping and positioning. Summary of the invention

[0004] In view of the problems of the prior art, the present invention provides a signal selection and processing method for multiple IMU sensor devices, which provides more accurate IMU sensor signals for digital twin mapping, positioning and other operations, thereby improving the accuracy of mapping, positioning and other results.

[0005] The specific scheme proposed by the present invention is:

[0006] The present invention provides a signal selection and processing method for multiple IMU sensor devices, comprising the following steps:

[0007] Step 1: IMU sensors are arranged in a redundant manner, wherein an upper isolation plate and a lower isolation plate are arranged, the upper isolation plate and the lower isolation plate are connected by a shock absorbing structure, IMU sensors are respectively deployed at both ends of a diagonal direction of the upper isolation plate, and the same number of IMU sensors as the upper isolation plate are respectively deployed on the lower isolation plate corresponding to the other diagonal direction of the upper isolation plate, and a data processing chip is deployed on the upper isolation plate or the lower isolation plate;

[0008] Step 2: Calibrate and filter the raw data of the IMU sensor;

[0009] Step 3: Perform quality control on the IMU sensor, compare the signal data of the IMU sensor after calibration and filtering preprocessing with the prior knowledge, and determine whether the signal data meets the expected performance indicators of the IMU sensor. If so, proceed to step 4, otherwise do not select the signal data;

[0010] Step 4: Select IMU sensors for sorting, including:

[0011] Step 41: First, sort the IMU sensors in descending order of accuracy priority according to their nominal accuracy.

[0012] Step 42: Sort the signal data of the real-time IMU sensor, wherein the signal data of the IMU sensor is initially sorted in descending order according to the accuracy priority of step 41, and the timestamp and signal data of the last update of each IMU sensor are obtained. The timestamp of the last update of the IMU sensor with the best accuracy priority is used as a reference, and the timestamps of the last update of other IMU sensors are compared to obtain the most recently updated IMU sensor. The update time interval between the most recently updated IMU sensor and the IMU sensor with the best accuracy priority is compared, and the IMU sensor closer to the current moment is selected as the optimal sensor. The remaining IMU sensors are sorted according to the proximity of the timestamp of the last update to the current moment, and then go to step 5.

[0013] If the last updated timestamps of each IMU sensor are the same or the difference is less than the set threshold, go to step 43.

[0014] Step 43: Obtain the confidence of each IMU sensor, sort the IMU sensors in descending order of priority according to the confidence, and proceed to step 5;

[0015] Step 5: Select the IMU sensor with the best priority as the main IMU sensor and output the signal data of the main IMU sensor, or fuse the signal data of each IMU sensor in proportion and weighted manner to obtain the fused signal data and output it.

[0016] Further, in the signal selection and processing method of a multi-IMU sensor device, a total of 4 IMU sensors are deployed in step 1, one IMU sensor is deployed at both ends of a diagonal direction of the upper isolation plate, and one IMU sensor is deployed at both ends of another diagonal direction of the lower isolation plate corresponding to the upper isolation plate.

[0017] The data processing chip is placed on the lower isolation board, and the upper isolation board transmits signal data to the lower board and the data processing chip through pins.

[0018] Furthermore, in step 43 of the signal selection and processing method for multiple IMU sensor devices, the mean and variance of the difference between the instantaneous value and the steady-state value of the signal data of each IMU sensor at the most recent moment are counted through a sliding window with a fixed time interval, and the confidence of each IMU sensor is calculated based on the probability density function of the normal distribution.

[0019] Furthermore, in the signal selection and processing method for multiple IMU sensor devices, the signal data of each IMU sensor is weighted and fused proportionally in step 5, including:

[0020] Using the formula Get the fused signal data, A represents the fused signal data, n is the number of IMU sensors, K i and A i are the weight coefficient and signal data of the i-th IMU sensor respectively. Ki is equal to the square of (1 / error_density(i)), and error_density(i) represents the error density of the i-th IMU sensor.

[0021] The present invention also provides a signal selection and processing device for multiple IMU sensor devices, including a deployment module, a preprocessing module, a quality control module, a sorting module and an output module.

[0022] The deployment module arranges the IMU sensors in a redundant form, wherein the deployment module arranges an upper isolation plate and a lower isolation plate, the upper isolation plate and the lower isolation plate are connected by a shock absorbing structure, IMU sensors are respectively deployed at both ends of a diagonal direction of the upper isolation plate, and the same number of IMU sensors as the upper isolation plate are respectively deployed on the lower isolation plate corresponding to the other diagonal direction of the upper isolation plate, and a data processing chip is deployed on the upper isolation plate or the lower isolation plate;

[0023] The preprocessing module calibrates and pre-filters the raw data of the IMU sensor;

[0024] The quality control module performs quality control on the IMU sensor, compares the signal data of the IMU sensor after calibration and filtering preprocessing with the prior knowledge, and determines whether the signal data meets the expected performance indicators of the IMU sensor. If so, proceed to step 4, otherwise, do not select the signal data;

[0025] The sorting module selects IMU sensors for sorting, including:

[0026] Step 41: First, sort the IMU sensors in descending order of accuracy priority according to their nominal accuracy.

[0027] Step 42: Sort the signal data of the real-time IMU sensor, wherein the signal data of the IMU sensor is initially sorted in descending order according to the accuracy priority of step 41, and the timestamp and signal data of the last update of each IMU sensor are obtained. The timestamp of the last update of the IMU sensor with the best accuracy priority is used as a reference, and the timestamps of the last update of other IMU sensors are compared to obtain the most recently updated IMU sensor. The update time interval between the most recently updated IMU sensor and the IMU sensor with the best accuracy priority is compared, and the IMU sensor closer to the current moment is selected as the optimal sensor. The remaining IMU sensors are sorted according to the proximity of the timestamp of the last update to the current moment, and sent to the output module.

[0028] If the last updated timestamps of each IMU sensor are the same or the difference is less than the set threshold, go to step 43.

[0029] Step 43: Obtain the confidence of each IMU sensor, sort the IMU sensors in descending order of priority according to the confidence, and send them to the output module;

[0030] The output module selects the IMU sensor with the best priority as the main IMU sensor and outputs the signal data of the main IMU sensor, or fuses the signal data of each IMU sensor in proportion and weighted manner to obtain the fused signal data and output it.

[0031] Furthermore, in the signal selection and processing device for a multi-IMU sensor device, a deployment module deploys a total of 4 IMU sensors, one IMU sensor is deployed at both ends of a diagonal direction of the upper isolation plate, and one IMU sensor is deployed at both ends of another diagonal direction of the lower isolation plate corresponding to the upper isolation plate.

[0032] The data processing chip is placed on the lower isolation board, and the upper isolation board transmits signal data to the lower board and the data processing chip through pins.

[0033] Furthermore, in the signal selection and processing device for multiple IMU sensors, the sorting module in step 43 counts the mean and variance of the difference between the instantaneous value and the steady-state value of the signal data of each IMU sensor at the most recent moment through a sliding window with a fixed time interval, and obtains the confidence of each IMU sensor based on the probability density function of the normal distribution.

[0034] Furthermore, the output module in the signal selection and processing device of the multi-IMU sensor device performs proportional weighted fusion of the signal data of each IMU sensor, including:

[0035] Using the formula Get the fused signal data, A represents the fused signal data, n is the number of IMU sensors, K i and A i are the weight coefficient and signal data of the i-th IMU sensor respectively. Ki is equal to the square of (1 / error_density(i)), and error_density(i) represents the error density of the i-th IMU sensor.

[0036] The benefits of the present invention are:

[0037] The present invention provides a signal selection and processing method for multiple IMU sensor devices. Aiming at the demand for high-precision and high-reliability IMU data in algorithms such as digital twin mapping, positioning, and three-dimensional reconstruction, the present invention performs preferential selection and signal processing on redundant sensor devices of multiple IMUs, and improves the anti-interference ability and reliability of the system through the redundancy and reasonable arrangement of multiple IMUs. At the same time, through the unified processing of multiple IMU sensor signals, more accurate signal data is finally output. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 It is a schematic diagram of the layout of the IMU involved in the method of the present invention.

[0039] Figure 2 It is a schematic flow chart of the method of the present invention. DETAILED DESCRIPTION

[0040] The present invention is further described below in conjunction with the accompanying drawings and specific embodiments so that those skilled in the art can better understand the present invention and implement it, but the embodiments are not intended to limit the present invention.

[0041] The present invention provides a signal selection and processing method for multiple IMU sensor devices, comprising the following steps:

[0042] Step 1: IMU sensors are arranged in a redundant manner, wherein an upper isolation plate and a lower isolation plate are arranged, the upper isolation plate and the lower isolation plate are connected by a shock absorbing structure, IMU sensors are respectively deployed at both ends of a diagonal direction of the upper isolation plate, and the same number of IMU sensors as the upper isolation plate are respectively deployed on the lower isolation plate corresponding to the other diagonal direction of the upper isolation plate, and a data processing chip is deployed on the upper isolation plate or the lower isolation plate;

[0043] Step 2: Calibrate and filter the raw data of the IMU sensor;

[0044] Step 3: Perform quality control on the IMU sensor, compare the signal data of the IMU sensor after calibration and filtering preprocessing with the prior knowledge, and determine whether the signal data meets the expected performance indicators of the IMU sensor. If so, proceed to step 4, otherwise do not select the signal data;

[0045] Step 4: Select IMU sensors for sorting, including:

[0046] Step 41: First, sort the IMU sensors in descending order of accuracy priority according to their nominal accuracy.

[0047] Step 42: Sort the signal data of the real-time IMU sensor, wherein the signal data of the IMU sensor is initially sorted in descending order according to the accuracy priority of step 41, and the timestamp and signal data of the last update of each IMU sensor are obtained. The timestamp of the last update of the IMU sensor with the best accuracy priority is used as a reference, and the timestamps of the last update of other IMU sensors are compared to obtain the most recently updated IMU sensor. The update time interval between the most recently updated IMU sensor and the IMU sensor with the best accuracy priority is compared, and the IMU sensor closer to the current moment is selected as the optimal sensor. The remaining IMU sensors are sorted according to the proximity of the timestamp of the last update to the current moment, and then go to step 5.

[0048] If the last updated timestamps of each IMU sensor are the same or the difference is less than the set threshold, go to step 43.

[0049] Step 43: Obtain the confidence of each IMU sensor, sort the IMU sensors in descending order of priority according to the confidence, and proceed to step 5;

[0050] Step 5: Select the IMU sensor with the best priority as the main IMU sensor and output the signal data of the main IMU sensor, or fuse the signal data of each IMU sensor in proportion and weighted manner to obtain the fused signal data and output it.

[0051] The method of the present invention improves the reliability and accuracy of measurement results through the combined application and reasonable arrangement of multiple IMUs and the unified processing of multiple IMU signals, thereby providing more accurate signal data for upper-level mapping, positioning, 3D reconstruction and other algorithms.

[0052] In specific applications, in some embodiments of the method of the present invention, based on the technical solution of the method of the present invention, when performing signal selection and processing of multiple IMU sensor devices, the reference process is as follows:

[0053] Step 1: IMU sensors are arranged in a redundant manner, wherein an upper isolation plate and a lower isolation plate are arranged, and the upper isolation plate and the lower isolation plate are connected by a shock-absorbing structure. IMU sensors are deployed at both ends of a diagonal direction of the upper isolation plate, and the same number of IMU sensors as the upper isolation plate are deployed on the lower isolation plate at both ends of another diagonal direction of the upper isolation plate. A data processing chip is deployed on the upper isolation plate or the lower isolation plate.

[0054] Further, refer to Figure 1 In step 1, a total of 4 IMU sensors are deployed, one at each end of a diagonal line of the upper isolation plate, and one at each end of another diagonal line of the lower isolation plate. This can minimize the impact caused by position, electromagnetic interference, temperature, etc. Between the upper and lower isolation plates, shock-absorbing materials are used for support isolation. They can be made of metal studs, rubber rods, memory foam, and other materials. According to the vibration frequency of the actual use scenario, the isolation support material with corresponding elasticity and damping characteristics can be reasonably selected to achieve the best vibration reduction effect.

[0055] The data processing chip is deployed on the lower isolation board, and the upper isolation board transmits signal data to the lower board and the data processing chip through the pins. The IMU connects the electrical signal it measures to the processing chip through the data line on the printed circuit board, such as spi / i2c, and the upper isolation board transmits the electrical signal to the lower board and the processing chip through the pins. According to the signal processing algorithm designed later, the chip uses the measurement data of 1-4 IMUs to calculate the optimal output result in the current situation as the final result of the measurement value at that moment.

[0056] Step 2: Calibrate and filter the raw data of the IMU sensor. For example, for the accelerometer, due to factors such as errors and noise, zero bias calibration and ratio calibration are performed. At the same time, in order to eliminate high-frequency noise, low-pass filtering is also performed to eliminate high-frequency noise while retaining local changes in low-frequency components to improve data processability.

[0057] Step 3: Perform quality control on the IMU sensor, compare the signal data of the calibrated and filtered IMU sensor with the prior knowledge, and determine whether the signal data meets the performance indicators expected by the IMU sensor. If it does, proceed to step 4, otherwise the signal data is not selected. After calibrating and denoising the data of each IMU sensor, compare the signal data of the calibrated sensor with the prior knowledge to see whether it meets the performance indicators expected by the sensor. If there is any data that does not meet the requirements, it will be considered as abnormal data and will be excluded from the data set for state estimation, thereby improving the accuracy of state estimation.

[0058] Step 4: Select IMU sensors for sorting, including:

[0059] Step 41: First, sort the IMU sensors in descending order of accuracy priority according to their nominal accuracy. The four IMUs used are sensors of different brands and specifications. In the initial stage, they are prioritized according to their nominal accuracy and other performance, such as IMU1>IMU2>IMU3>IMU4. If there is no output or they cannot be used in steps 42 and 43, the data of the sensors with higher priority can be used in the order of nominal accuracy priority. If the difference between the high-priority IMU sensor data and the low-priority IMU sensor data is greater than a certain ratio, the low-priority data will be discarded; if the errors are within a certain range, the final result can be obtained by weighted fusion.

[0060] Step 42: Sort the signal data of the real-time IMU sensor, wherein the signal data of the IMU sensor is initially sorted in descending order according to the accuracy priority of step 41, and the timestamp and signal data of the last update of each IMU sensor are obtained. The timestamp of the last update of the IMU sensor with the best accuracy priority is used as a reference, and the timestamps of the last update of other IMU sensors are compared to obtain the most recently updated IMU sensor. The update time interval between the most recently updated IMU sensor and the IMU sensor with the best accuracy priority is compared, and the IMU sensor closer to the current moment is selected as the optimal sensor. The remaining IMU sensors are sorted according to the proximity of the timestamp of the last update to the current moment, and then go to step 5.

[0061] If the last updated timestamps of each IMU sensor are the same or the difference is less than the set threshold, go to step 43.

[0062] Step 43: Obtain the confidence of each IMU sensor, sort the IMU sensors in descending order of priority according to the confidence, and proceed to step 5.

[0063] Furthermore, in step 43, the general method provided by the IMU sensor can count the mean and variance of the difference between the instantaneous value and the steady-state value of the signal data of each IMU sensor at the most recent moment through a sliding window with a fixed time interval, and obtain the confidence of each IMU sensor according to the probability density function of the normal distribution.

[0064] Step 5: Select the IMU sensor with the best priority as the main IMU sensor, output the signal data of the main IMU sensor, or fuse the signal data of each IMU sensor in proportion and weighted, obtain the fused signal data and output it. The sensor with the largest confidence value is set as the optimal sensor. The data output by the sensor with a high confidence value is more credible and can better represent the posture or motion state in the physical sense. The above method can be used to continuously update the selection of the optimal IMU device so that the results of using sensors with higher data accuracy can be ensured at all times.

[0065] Or if the errors between the high-priority IMU sensor data and the low-priority IMU sensor data are within a certain range, the final result can be obtained by weighted fusion. Further, in step 5, the signal data of each IMU sensor is weighted fused in proportion, including:

[0066] Using the formula Get the fused signal data, A represents the fused signal data, n is the number of IMU sensors, K i and A i They are the weight coefficient and signal data of the ith IMU sensor respectively. Ki is equal to the square of (1 / error_density(i)). Error_density(i) represents the error density of the ith IMU sensor, that is, the noise level of the IMU output data. Generally, the error density value is a basic parameter to describe the sensor performance and is provided by the sensor equipment supplier.

[0067] The present invention also provides a signal selection and processing device for multiple IMU sensor devices, including a deployment module, a preprocessing module, a quality control module, a sorting module and an output module.

[0068] The deployment module arranges the IMU sensors in a redundant form, wherein the deployment module arranges an upper isolation plate and a lower isolation plate, the upper isolation plate and the lower isolation plate are connected by a shock absorbing structure, IMU sensors are respectively deployed at both ends of a diagonal direction of the upper isolation plate, and the same number of IMU sensors as the upper isolation plate are respectively deployed on the lower isolation plate corresponding to the other diagonal direction of the upper isolation plate, and a data processing chip is deployed on the upper isolation plate or the lower isolation plate;

[0069] The preprocessing module calibrates and pre-filters the raw data of the IMU sensor;

[0070] The quality control module performs quality control on the IMU sensor, compares the signal data of the IMU sensor after calibration and filtering preprocessing with the prior knowledge, and determines whether the signal data meets the expected performance indicators of the IMU sensor. If so, proceed to step 4, otherwise, do not select the signal data;

[0071] The sorting module selects IMU sensors for sorting, including:

[0072] Step 41: First, sort the IMU sensors in descending order of accuracy priority according to their nominal accuracy.

[0073] Step 42: Sort the signal data of the real-time IMU sensor, wherein the signal data of the IMU sensor is initially sorted in descending order according to the accuracy priority of step 41, and the timestamp and signal data of the last update of each IMU sensor are obtained. The timestamp of the last update of the IMU sensor with the best accuracy priority is used as a reference, and the timestamps of the last update of other IMU sensors are compared to obtain the most recently updated IMU sensor. The update time interval between the most recently updated IMU sensor and the IMU sensor with the best accuracy priority is compared, and the IMU sensor closer to the current moment is selected as the optimal sensor. The remaining IMU sensors are sorted according to the proximity of the timestamp of the last update to the current moment, and sent to the output module.

[0074] If the last updated timestamps of each IMU sensor are the same or the difference is less than the set threshold, go to step 43.

[0075] Step 43: Obtain the confidence of each IMU sensor, sort the IMU sensors in descending order of priority according to the confidence, and send them to the output module;

[0076] The output module selects the IMU sensor with the best priority as the main IMU sensor and outputs the signal data of the main IMU sensor, or fuses the signal data of each IMU sensor in proportion and weighted manner to obtain the fused signal data and output it.

[0077] The information interaction, execution process and other contents between the modules in the above-mentioned device are based on the same concept as the embodiment of the method of the present invention. For the specific contents, please refer to the description in the embodiment of the method of the present invention, and no further description is given here.

[0078] Similarly, the device of the present invention is aimed at the demand for high-precision and high-reliability IMU data in digital twin mapping, positioning, three-dimensional reconstruction and other algorithms, and selectively selects and processes signals for multiple IMU redundant sensor devices. Through the redundancy and reasonable arrangement of multiple IMUs, the anti-interference ability and reliability of the system are improved. At the same time, through the unified processing of multiple IMU sensor signals, more accurate signal data is ultimately output.

[0079] It should be noted that not all steps and modules in the above-mentioned processes and device structure diagrams are necessary, and some steps or modules can be ignored according to actual needs. The execution order of each step is not fixed and can be adjusted as needed. The system structure described in the above-mentioned embodiments can be a physical structure or a logical structure, that is, some modules may be implemented by the same physical entity, or some modules may be implemented by multiple physical entities, or some components in multiple independent devices may be implemented together.

[0080] The above-described embodiments are only preferred embodiments for fully illustrating the present invention, and the protection scope of the present invention is not limited thereto. Equivalent substitutions or changes made by those skilled in the art based on the present invention are within the protection scope of the present invention. The protection scope of the present invention shall be subject to the claims.

Claims

1. A signal selection and processing method for multiple IMU sensor devices, Its characteristics are The following steps are involved: Step 1: IMU sensors are arranged in a redundant manner, wherein an upper isolation plate and a lower isolation plate are arranged, the upper isolation plate and the lower isolation plate are connected by a shock absorbing structure, IMU sensors are respectively deployed at both ends of a diagonal direction of the upper isolation plate, and the same number of IMU sensors as the upper isolation plate are respectively deployed on the lower isolation plate corresponding to the other diagonal direction of the upper isolation plate, and a data processing chip is deployed on the upper isolation plate or the lower isolation plate; Step 2: Calibrate and filter the raw data of the IMU sensor; Step 3: Perform quality control on the IMU sensor, compare the signal data of the IMU sensor after calibration and filtering preprocessing with the prior knowledge, and determine whether the signal data meets the expected performance indicators of the IMU sensor. If so, proceed to step 4, otherwise do not select the signal data; Step 4: Select IMU sensors for sorting, including: Step 41: First, sort the IMU sensors in descending order of accuracy priority according to their nominal accuracy. Step 42: Sort the signal data of the real-time IMU sensor, wherein the signal data of the IMU sensor is initially sorted in descending order according to the accuracy priority of step 41, and the timestamp and signal data of the last update of each IMU sensor are obtained. The timestamp of the last update of the IMU sensor with the best accuracy priority is used as a reference, and the timestamps of the last update of other IMU sensors are compared to obtain the most recently updated IMU sensor. The update time interval between the most recently updated IMU sensor and the IMU sensor with the best accuracy priority is compared, and the IMU sensor closer to the current moment is selected as the optimal sensor. The remaining IMU sensors are sorted according to the proximity of the timestamp of the last update to the current moment, and then go to step 5. If the last updated timestamps of each IMU sensor are the same or the difference is less than the set threshold, go to step 43. Step 43: Obtain the confidence of each IMU sensor, sort the IMU sensors in descending order of priority according to the confidence, and proceed to step 5; Step 5: Select the IMU sensor with the best priority as the main IMU sensor and output the signal data of the main IMU sensor, or fuse the signal data of each IMU sensor in proportion and weighted manner to obtain the fused signal data and output it.

2. The signal selection and processing method of a multi-IMU sensor device according to claim 1, Its characteristic is the steps 1 A total of 4 IMU sensors are deployed, one IMU sensor is deployed at both ends of a diagonal direction of the upper isolation plate, and one IMU sensor is deployed at both ends of the other diagonal direction of the lower isolation plate corresponding to the upper isolation plate. The data processing chip is placed on the lower isolation board, and the upper isolation board transmits signal data to the lower board and the data processing chip through pins.

3. The signal selection and processing method of a multi-IMU sensor device according to claim 1, Its characteristics are In step 43, the mean and variance of the difference between the instantaneous value and the steady-state value of the signal data of each IMU sensor at the most recent moment are counted through a sliding window with a fixed time interval, and the confidence of each IMU sensor is calculated according to the probability density function of the normal distribution.

4. The signal selection and processing method of a multi-IMU sensor device according to claim 1, Its characteristics are In step 5, the signal data of each IMU sensor is weighted and fused proportionally, including: Using the formula Get the fused signal data, A represents the fused signal data, n is the number of IMU sensors, K i and A i are the weight coefficient and signal data of the i-th IMU sensor respectively. Ki is equal to the square of (1 / error_density(i)), and error_density(i) represents the error density of the i-th IMU sensor.

5. A signal selection and processing device for multiple IMU sensor devices, Its characteristics are It includes deployment module, preprocessing module, quality control module, sorting module and output module. The deployment module arranges the IMU sensors in a redundant form, wherein the deployment module arranges an upper isolation plate and a lower isolation plate, the upper isolation plate and the lower isolation plate are connected by a shock absorbing structure, IMU sensors are respectively deployed at both ends of a diagonal direction of the upper isolation plate, and the same number of IMU sensors as the upper isolation plate are respectively deployed on the lower isolation plate corresponding to the other diagonal direction of the upper isolation plate, and a data processing chip is deployed on the upper isolation plate or the lower isolation plate; The preprocessing module calibrates and pre-filters the raw data of the IMU sensor; The quality control module performs quality control on the IMU sensor, compares the signal data of the IMU sensor after calibration and filtering preprocessing with the prior knowledge, and determines whether the signal data meets the expected performance indicators of the IMU sensor. If so, proceed to step 4, otherwise, do not select the signal data; The sorting module selects IMU sensors for sorting, including: Step 41: First, sort the IMU sensors in descending order of accuracy priority according to their nominal accuracy. Step 42: Sort the signal data of the real-time IMU sensor, wherein the signal data of the IMU sensor is initially sorted in descending order according to the accuracy priority of step 41, and the timestamp and signal data of the last update of each IMU sensor are obtained. The timestamp of the last update of the IMU sensor with the best accuracy priority is used as a reference, and the timestamps of the last update of other IMU sensors are compared to obtain the most recently updated IMU sensor. The update time interval between the most recently updated IMU sensor and the IMU sensor with the best accuracy priority is compared, and the IMU sensor closer to the current moment is selected as the optimal sensor. The remaining IMU sensors are sorted according to the proximity of the timestamp of the last update to the current moment, and sent to the output module. If the last updated timestamps of each IMU sensor are the same or the difference is less than the set threshold, go to step 43. Step 43: Obtain the confidence of each IMU sensor, sort the IMU sensors in descending order of priority according to the confidence, and send them to the output module; The output module selects the IMU sensor with the best priority as the main IMU sensor and outputs the signal data of the main IMU sensor, or fuses the signal data of each IMU sensor in proportion and weighted manner to obtain the fused signal data and output it.

6. The signal selection and processing device of a multi-IMU sensor device according to claim 5, Its characteristics are The deployment module deploys a total of 4 IMU sensors, one IMU sensor is deployed at both ends of a diagonal direction of the upper isolation plate, and one IMU sensor is deployed at both ends of the other diagonal direction of the lower isolation plate corresponding to the upper isolation plate. The data processing chip is placed on the lower isolation board, and the upper isolation board transmits signal data to the lower board and the data processing chip through pins.

7. The signal selection and processing device of a multi-IMU sensor device according to claim 5, Its characteristics are In step 43, the sorting module counts the mean and variance of the difference between the instantaneous value and the steady-state value of the signal data of each IMU sensor at the most recent moment through a sliding window with a fixed time interval, and obtains the confidence of each IMU sensor according to the probability density function of the normal distribution.

8. The signal selection and processing device of a multi-IMU sensor device according to claim 5, Its characteristics are The output module fuses the signal data of each IMU sensor in a proportional weighted manner, including: Using the formula Get the fused signal data, A represents the fused signal data, n is the number of IMU sensors, K i and A i are the weight coefficient and signal data of the i-th IMU sensor respectively. Ki is equal to the square of (1 / error_density(i)), and error_density(i) represents the error density of the i-th IMU sensor.

Images