Multi-MEMS IMU chip module, performance optimization method and computer system

Abstract

The invention discloses a multi-MEMS IMU chip module and a performance optimization method, and the method comprises the steps: carrying out the performance evaluation of each IMU in different temperature intervals, calculating the weight of each IMU in each temperature interval, judging whether the current IMU is in a static state or not through the acceleration and angular velocity information of the IMU, judging the current acceleration quality according to the acceleration modulus length of the static state, and carrying out the calculation of the current acceleration quality. The weight ratio is adjusted in real time; and in an IMU motion state, fusing high-frequency dynamic information of an IMU chip gyroscope and low-frequency steady-state information of an accelerometer by using an extended Kalman filter model, screening out and reducing the weight of an abnormal IMU, and applying data fusion processing in a final output result. The method has the advantages that the precision problem caused by consistency difference in the manufacturing process of a single IMU is solved, and the overall measurement precision and reliability of the IMU module are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of inertial navigation of intelligent systems, and particularly to a multi-MEMS IMU chip module, a performance optimization method, and a computer system. Background Art

[0002] An important problem faced by inertial measurement units (IMUs) based on microelectromechanical systems (MEMS) in wide applications is the poor consistency of chip performance. Even for IMU chips produced in the same batch, there may be significant differences in their performance. This consistency difference has an adverse impact on the measurement accuracy and reliability of the IMU.

[0003] To overcome this problem, multi-MEMS IMU chip modules have been proposed. By using multiple IMU chips simultaneously, it is hoped to improve the overall measurement performance through redundant design. However, each IMU chip is affected by various factors such as materials and manufacturing processes during production, resulting in performance differences between different chips and ultimately affecting the accuracy of measurement results. The performance of different chips varies in different temperature ranges, which is also an important factor affecting measurement accuracy. This is where this application needs to focus on improvement. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a multi-MEMS IMU chip module and a performance optimization method to improve the overall measurement accuracy and reliability of the IMU module.

[0005] To solve the above technical problems, the present invention provides a performance optimization method for a multi-MEMS IMU chip module, including: 1) Evaluate the performance of each IMU chip in different temperature ranges to obtain its respective performance indicators; based on the performance indicators in each temperature range, calculate the weight of each IMU chip in different temperature ranges; The performance evaluation includes measuring the zero-bias fluctuation amplitude, zero-bias instability, random walk, and scale factor of the current temperature range of the IMU chip; First, standardize each performance indicator to a unified range [0, 1] for comparison; the standardization method is min-max normalization, and the obtained indicator value after standardization is ; Next, assign a weight factor Q to each performance indicator, which represents the importance of the indicator for chip performance evaluation; Finally, according to the standardized indicator value and the weight factor Q of each indicator, calculate the weight W of the IMU chip by the weighted average method; according to the weight size of the IMU chip, determine the influence of different IMU chips on the final output data of the multi-IMU chip module; 2) Differentiate different temperature ranges and different axes, and calculate a set of weights respectively; The normal operating temperature range of each IMU chip is divided into n intervals. T1 represents from T min1 °C to T max1 °C, T2 represents from T min2 °C to T max2 °C, and so on. T n-1 represents from T min（n-1） °C to T max（n-1） °C, T n represents from T minn °C to T maxn °C, where T max（n-1） = T minn , and evaluate the performance of the IMU chip within each temperature interval; Furthermore, for the performance evaluation of the IMU chip, the performance of each IMU chip is further refined to different axes (x, y, z), that is, the weights of the same IMU chip on different axes are also different. Based on the performance indicators of different chips on different axes, calculate the weights of each IMU chip on different axes; 3) Judge the working state of the current IMU chip: Use the acceleration and angular velocity information of the IMU chip to judge whether the current IMU chip is in a stationary state; When the IMU chip is in a stationary state, compare the current acceleration magnitude with the acceleration magnitude in the stationary state, reduce the weight of the abnormal IMU chip, dynamically adjust the weight ratio, and make the attitude estimation in the stationary state more stable; according to the adjusted weights, perform data fusion processing on the outputs of multiple IMU chips in each temperature interval to obtain the measurement results of each temperature interval, and synthesize the measurement results of each temperature interval to obtain the final measurement result; When the IMU chip is in a moving state, use the extended Kalman filter model to fuse the high-frequency dynamic information of the gyroscope of the IMU chip with the low-frequency steady-state information of the accelerometer to obtain a more accurate and stable attitude; reduce the weight of the abnormal IMU chip, dynamically adjust the weight ratio, perform data fusion processing on the adjusted weights and the output of the IMU chip, perform data fusion processing on the outputs of multiple IMU chips in each temperature interval to obtain the measurement results of each temperature interval, and synthesize the measurement results of each temperature interval to obtain the final measurement result.

[0006] The weight calculation adopts a data fusion processing method, which combines the outputs of different IMU chips according to the weights. The weights of multiple IMU chips are as follows: ; Among them: , The actual number of IMU chips is [[ID=]], T is the current temperature, and n is the temperature range.

[0007] The present invention also provides a multi-MEMS IMU chip module, including a plurality of IMU chips and a data processing module, and the data processing module is used to implement the performance optimization method of the multi-MEMS IMU chip module.

[0008] The present invention also provides a computer system, including a memory, a processor, and a computer program stored on the memory and executable on the processor, and the processor executes the computer program to implement the steps of the performance optimization method of the multi-MEMS IMU chip module.

[0009] The beneficial effects of the present invention are as follows: 1) The multi-IMU redundancy design improves the accuracy: By adopting a plurality of MEMS IMU chips and combining the output method of data fusion processing, the present invention effectively solves the accuracy problem brought by the consistency difference in the manufacturing process of a single IMU chip; The multi-IMU chip module can minimize the error influence of individual chips and improve the measurement accuracy by performing data fusion processing on the outputs of multiple chips; 2) Dynamic weight allocation based on the temperature range: The performance of MEMS IMU chips fluctuates with temperature changes. Under different temperature conditions, key performance indicators such as the zero-bias fluctuation amplitude, zero-bias instability, random walk, and scale factor in the current temperature segment of IMU chips will be different. The present invention divides the entire temperature range into multiple intervals, respectively evaluates the performance of IMU chips in each temperature interval, and dynamically adjusts the weights of each IMU chip under different temperature conditions according to the actual measurement situation. This method ensures that the system can maintain a high measurement accuracy in different temperature environments, thereby enhancing the adaptability and environmental robustness of the system; 3) The present invention has different processing methods for IMU chips in different motion states: In the stationary state, the acceleration magnitude information is used to evaluate the acceleration output quality of each IMU chip, and the weight ratio is dynamically adjusted to make the attitude estimation in the stationary state more stable; In the motion state, the gyroscope and accelerometer data of the IMU are affected by high-frequency vibration and low-frequency drift. The present invention effectively fuses the high-frequency dynamic information of the gyroscope of the IMU chip with the low-frequency steady-state information of the accelerometer by using the extended Kalman filter EKF model, significantly improving the accuracy of attitude calculation in a dynamic environment; 4) Real-time elimination of abnormal IMU chip outputs to enhance system robustness: The advantage of the redundant design of multiple IMU chips is that when multiple IMU chips work simultaneously, it is possible to screen out abnormal chips that deviate significantly from the output results of most IMU chips by comparing the attitude estimation results of different IMU chips. In the process of data fusion processing of the present invention, the weights of these abnormal-performing IMU chips are dynamically reduced to prevent their inaccurate data from affecting the final measurement results, thereby improving the anti-interference ability and overall robustness of the system; 5) Adaptive weight calculation for precise output control: The present invention adopts a data fusion processing method. By evaluating the performance of each IMU chip in different temperature ranges, its weight is obtained, and then the weight ratio is further adjusted according to the working state of the IMU chip, so that the finally output attitude and measurement results can accurately reflect the actual state; especially in the stationary state, the present invention can adjust the weight in real time through the acceleration magnitude and dynamically adjust the output quality; 6) Comprehensive temperature range output to ensure full-temperature range accuracy: The final measurement result is obtained by comprehensively considering the measurement results in each temperature range. This method ensures that the performance differences of IMU chips in different temperature environments can be effectively managed. Especially in application scenarios with drastic temperature changes, the design of the present invention can ensure that the measurement results always maintain high accuracy throughout the temperature range and are applicable to a wide range of application environments; 7) Improvement of the overall system reliability: The redundant design of multiple IMU chips combined with dynamic weight adjustment enables the entire system to still maintain high reliability in a harsh working environment. Even if some IMU chips perform poorly under certain circumstances, their influence can be reduced through weight adjustment, thereby ensuring the accuracy and reliability of the final measurement results. Description of the Drawings

[0010] The specification drawings forming a part of this application are used to provide a further understanding of the present invention. The schematic embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings: Figure 1 is a flowchart of a specific embodiment of the present invention; Figure 2 is a schematic diagram of the function of a single IMU chip in a specific embodiment of the present invention; Figure 3 is a schematic diagram of a multi-MEMS IMU chip module in a specific embodiment of the present invention; Figure 4 is a partition diagram of the normal operating temperature range of the IMU chip of the present invention; Figure 5 is a temperature partition diagram of a specific embodiment of the present invention; Figure 6It is the flow chart of the EKF fusion model in a specific embodiment of the present invention. Detailed implementation manners

[0011] To better understand the present invention, a specific example is now used for illustration. It should be noted that the example given is only provided for easy understanding and is not used to limit the protection scope of the present invention. The present invention can be implemented and applied through other different examples and ways.

[0012] It should be noted that the illustrations in this embodiment only schematically illustrate the basic concept of the present invention. Only the components related to the present invention are shown in the figures, and they are not drawn according to the number, shape, and size of the components in actual implementation. During the actual implementation process, the form, quantity, and ratio of each component can be changed according to actual needs, and the component layout may be more complex.

[0013] As Figure 3 shown, a complete MEMS IMU chip module includes multiple IMU chips, which are evenly distributed according to a certain arrangement rule, the spacing between each IMU chip is equal, and the arrangements are parallel and aligned. As Figure 2 shown, each IMU chip outputs the acceleration and angular velocity in the current motion state through an accelerometer and a gyroscope. Since there are performance differences between each IMU chip, which affects the accuracy of the final measurement result, to solve this technical problem, the present invention provides a performance optimization method for a multi-MEMS IMU chip module, as Figure 1 shown, including: Step S1: Based on the performance of different chips in different temperature segments, evaluate the performance of each IMU chip in different temperature ranges to obtain their respective performance indicators; based on the performance indicators in each temperature range, calculate the weight of each IMU chip in different temperature ranges; The performance evaluation includes measuring the zero-bias fluctuation amplitude, zero-bias instability, random walk, and scale factor of the IMU chip in the current temperature segment; First, standardize each performance indicator to a unified range [0, 1] for comparison; the standardization method is min-max normalization; For a certain index value Xi, such as the zero-bias fluctuation amplitude, zero-bias instability, random walk, and scale factor of the current temperature segment, the standardization formula is as follows: ; Where: Xmin is the minimum value of this index; Xmax is the maximum value of this index; is the value after standardization, and the range is between [0, 1]; According to the standardized index value and the weight factor Qi of each index, calculate the weight of this IMU chip by the weighted average method , and its calculation formula is as follows: ; Where: n is the total number of performance indicators; Qi is the weight factor of the i-th performance indicator, satisfying ; is the normalized value of the i-th performance indicator, and its range is [0, 1]; As Figure 4 shown, the normal operating temperature range of each IMU chip is divided into n intervals. T1 represents from T min1 °C to T max1 °C, T2 represents from T min2 °C to T max2 °C, and so on. T n-1 represents from T min（n-1） °C to T max（n-1） °C, T n represents from T minn °C to T maxn °C, where T max（n-1） = T minn , and the performance of the IMU chip is evaluated within each temperature interval; As Figure 5 shown, in a specific embodiment of the present invention, the normal operating temperature range of the IMU chip is divided into five intervals, namely T1 (-40°C to -18°C), T2 (-18°C to 4°C), T3 (4°C to 26°C), T4 (26°C to 48°C), and T5 (48°C to 70°C). Based on the performance indicators within each temperature interval, the weights of each IMU chip in different temperature intervals are calculated; Perform performance evaluation on the IMU chip, and further refine the performance of each IMU chip to different axes (x, y, z). Based on the different performances of different chips on different axes, the weight sizes of different axes of the same chip are also different. Specifically, for the same chip, the zero-bias fluctuation amplitude, zero-bias instability, random walk, and scale factor of the current temperature segment of the x-axis, y-axis, and z-axis are all different. Finally, according to the weight calculation method, the weight values of this axis are also different; For the same axis, the weighted average method is used to calculate the weights of different chips; if the current temperature is T, the weight calculations of multiple IMU chips in five temperature intervals are as follows: ; Where: , is the actual number of IMU chips.

[0014] Step S2, determine whether the current IMU is in a stationary state according to the acceleration and angular velocity information of the IMU chip: The acceleration components output by multiple three-axis IMU chips are respectively , calculate the magnitude of the acceleration measured by each IMU chip, as follows: ; By comparing the difference between the acceleration magnitude and the gravitational acceleration , to determine whether the acceleration change is close to zero; ; Among them: is the magnitude of the gravitational acceleration, is a threshold; When the acceleration change and the angular velocity are close to zero at this moment, it is determined to be in a stationary state; According to the acceleration magnitude in the stationary state, compare the acceleration magnitudes of all IMU chips, screen out some IMU chips whose estimated attitudes deviate significantly from the attitudes estimated by most IMU chips, reduce the weights of abnormal IMU chips, adjust the weights of each IMU chip according to the acceleration magnitude, perform data fusion processing on the outputs of multiple IMU chips in each temperature range, and obtain the measurement results for each temperature range; comprehensively combine the measurement results of each temperature range to obtain the final measurement result.

[0015] The high-frequency part of the gyroscope accurately reflects the dynamic changes of the attitude, while the low-frequency part of the accelerometer needs to estimate the attitude average value of this period through the acceleration average value over a period of time. In the motion state of the IMU chip, the extended Kalman filter EKF model is adopted to fuse the high-frequency dynamic information of the gyroscope of the IMU chip with the low-frequency steady-state information of the accelerometer, so as to obtain a more accurate and stable attitude estimation; the EKF fusion process is as Figure 6 shown: 1) First is the prediction step. Define the state vector as in Equation (6), which includes the attitude quaternion and the bias of the gyroscope ; ; The state prediction equation is used to predict the state at the next moment, based on the current state and the angular velocity measurement value of the gyroscope , as in Equation (7); ; Among them: is the control input, that is, the measurement value of the gyroscope, is the process noise; The state transition function is expressed as in Equation (8): ; Among them: represents quaternion multiplication, is the time step; For predicting the state at the next moment, it is also necessary to evaluate the uncertainty of this state, and this uncertainty is the covariance. Therefore, the covariance is updated through the covariance prediction formula, as shown in Equation (9): ; The observation equation is used to associate the measurement value of the accelerometer with the state vector, and its form is shown in Equation (10); ; where: is the measurement value of the accelerometer, is the measurement noise; Observation function is expressed as Equation (11): ; where: is the rotation matrix represented by the quaternion ; 2) Enter the update step, dynamically balance the weights of prediction and measurement, and adjust the influence of the measurement residual on the state update, that is, calculate the Kalman gain, as shown in Equation (12): ; 3) Fuse the predicted value and the measured value, reduce the influence of the measurement noise, and obtain the best estimate of the current state, that is, update the state estimate, as shown in Equation (13): ; 4) Quantify the uncertainty of the updated state estimate, provide a basis for the prediction step at the next moment, that is, update the covariance estimate, as shown in Equation (14): ; Similarly, filter out some IMU chips whose estimated attitudes deviate significantly from the attitudes estimated by most IMU chips, and adjust the weights of these abnormal IMU chips in real time; Perform a weighted average of the adjusted weights and the outputs of the IMU chips to calculate the final acceleration and angular velocity, as shown in Equation (15): ; where: is the latest weight obtained by real-time adjustment through the above steps, and are the angular velocity and acceleration measured by the IMU chips or obtained after EKF fusion, respectively; 6) According to the adjusted weights, perform data fusion processing on the outputs of multiple IMU chips in each temperature range to obtain the measurement results for each temperature range; synthesize the measurement results of each temperature range to obtain the final measurement result.

[0016] The present invention also provides a multi-MEMS IMU chip module, including multiple IMU chips and a data processing module. The data processing module evaluates the performance of each IMU chip in different temperature ranges, calculates their respective weights in each temperature range, uses the acceleration and angular velocity information of the IMU chip to determine whether the current IMU is in a stationary state, determines the current acceleration quality based on the magnitude of the acceleration in the stationary state, and adjusts the weight ratio in real time; in the case of the IMU chip being in a motion state, uses the extended Kalman filter model to fuse the high-frequency dynamic information of the gyroscope of the IMU chip with the low-frequency steady-state information of the accelerometer, screens out and reduces the weights of abnormal IMU chips; according to the adjusted weights, and applies data fusion processing to the final output result to obtain the measurement results for each temperature range; synthesize the measurement results of each temperature range to obtain the final measurement result.

[0017] The present invention also provides a computer system, including a memory, a processor, and a computer program stored on the memory and executable on the processor. The processor executes the computer program to implement the steps of the performance optimization method of the multi-MEMS IMU chip module.

[0018] The above are only the preferred embodiments of the present invention and are not used to limit the present invention. For those skilled in the art, the present invention can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A performance optimization method for a multi-MEMS IMU chip module, characterized in that: include: 1) Evaluate the performance of each IMU chip in different temperature ranges, obtain their respective performance indicators, and calculate the weight of each IMU chip in different temperature ranges based on the performance indicators in each temperature range; 2) Determine the current working status of the IMU chip: Use the acceleration and angular velocity information of the IMU chip to determine whether the current IMU chip is in a stationary state; The IMU chip is in a stationary state. The weight of the abnormal IMU chip is reduced and the weight ratio is adjusted dynamically based on the acceleration modulus in the stationary state compared with the current acceleration modulus. When the IMU chip is in motion, the extended Kalman filter model is used to fuse the high-frequency dynamic information of the IMU chip's gyroscope with the low-frequency steady-state information of the accelerometer, reduce the weight of the abnormal IMU chip, and dynamically adjust the weight ratio; 3) According to the adjusted weights, the outputs of multiple IMU chips in each temperature range are fused to obtain the measurement results of each temperature range. The measurement results of each temperature range are combined to obtain the final measurement result.

2. The performance optimization method of the multi-MEMS IMU chip module according to claim 1, characterized in that: The performance evaluation includes measuring the zero bias fluctuation amplitude, zero bias instability, random walk, and scale factor of the current temperature segment of the IMU chip as the performance indicators.

3. The performance optimization method of the multi-MEMS IMU chip module according to claim 1, characterized in that: The weight calculation adopts the weighted average method, and the weights of multiple IMU chips on the same axis are as follows: ； in: , is the actual number of IMU chips, T is the current temperature, and n is the temperature range.

4. The performance optimization method of the multi-MEMS IMU chip module according to claim 1, characterized in that: In the static state, the acceleration components output by the three-axis IMU chip are , calculate the acceleration modulus measured by each IMU chip as follows: ； By comparing the acceleration modulus with the gravitational acceleration The difference between the acceleration and the acceleration is used to determine whether the acceleration change is close to zero. ； in, is the magnitude of the acceleration due to gravity, is a threshold.

5. The performance optimization method of the multi-MEMS IMU chip module according to claim 1, characterized in that: The extended Kalman filter EKF model is used to fuse the high-frequency dynamic information of the gyroscope of the IMU chip with the low-frequency steady-state information of the accelerometer. The specific steps are as follows: 1) The first step is the prediction step, which defines the state vector, which includes the attitude quaternion and the deviation of the gyroscope ; ； The state prediction equation is used to predict the state at the next moment, based on the current state and the angular velocity measurement of the gyroscope. , as shown below; ； in: is the control input, i.e. the gyroscope measurement, is the process noise; State transfer function It is expressed as follows: ； in: represents quaternion multiplication, is the time step; The covariance is updated through the agreement difference prediction formula, as shown in the following formula: ； The observation equation is used to relate the accelerometer measurements to the state vector and is shown in the following form: ； in: is the accelerometer measurement, is the measurement noise; Observation function It is expressed as follows: ； in: It is composed of quaternions represents the rotation matrix; 2) Enter the update step, dynamically balance the weights of prediction and measurement, and adjust the impact of the measurement residual on the state update, that is, calculate the Kalman gain, as shown in the following formula: ； 3) Fusion of predicted values ​​and measured values ​​to obtain the best estimate of the current state, that is, update the state estimate, as shown in the following formula: ； 4) Update the covariance estimate as shown below: 。 6. The performance optimization method of the multi-MEMS IMU chip module according to claim 1, characterized in that: According to the adjusted weights, the adjusted weights are weighted averaged with the output of the IMU chip to calculate the final acceleration and angular velocity, as shown in the following formula: ； in: is the adjusted weight, and They are the angular velocity and acceleration measured by the IMU chip or obtained after EKF fusion.

7. A multi-MEMS IMU chip module, comprising a plurality of IMU chips and a data processing module, wherein the data processing module is used to implement the performance optimization method described in any one of claims 1 to 6.

8. A computer system comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: The processor executes the computer program to implement the steps of the performance optimization method according to any one of claims 1 to 6.

Images