CN114689901B - Accelerometer field calibration method and device - Google Patents

Abstract

The application provides a method and device for on-site calibration of an accelerometer. The method includes: detecting whether the UAV is currently in a static state after rotation according to the preset multi-criteria zero-speed interval detection method, and if so, obtaining the acceleration of the UAV Calculate the initial value of the parameters of the error parameter group currently to be calibrated; based on the initial value of the parameters of the error parameter group, calibrate each error of the inertial sensor corresponding to the error parameter group, and automatically determine whether the corresponding calibration result is completed . The application can effectively improve the efficiency and accuracy of the zero-speed interval detection of the accelerometer, thereby effectively improving the efficiency and accuracy of the on-site calibration error parameters of the accelerometer, and can effectively improve the automation and reliability of the on-site calibration error parameters of the accelerometer.

Description

Accelerometer field calibration method and device

Technical Field

The present application relates to the technical field of MEMS acceleration calibration, and in particular to an accelerometer on-site calibration method and device.

Background Art

With the development of Micro-Electro-Mechanical Systems (MEMS), low-cost MEMS inertial sensors IMU have the advantages of small size, low power consumption, and light weight, which makes them gradually develop in positioning technology. The main errors of low-cost inertial sensor IMU include systematic errors and random errors, among which the systematic errors mainly include installation errors, scaling factors, and zero bias errors, so how to accurately and quickly calibrate the various error parameters of inertial sensor IMU has become an important research direction.

The on-site calibration method of the accelerometer requires the collection of static acceleration values under multiple different static postures. The collection method determines the length of the calibration time of the accelerometer. At present, the acceleration data collection methods can be divided into two methods: manual and zero-speed interval detection. Among them, the manual method has problems such as cumbersome operation and long collection time, and cannot be applied to the calibration of accelerometers of multiple drones; the zero-speed interval detection method can automatically distinguish between dynamic and static interval data, and has the advantage of high efficiency. However, the existing zero-speed interval detection methods have problems such as poor accuracy and low efficiency. Therefore, the existing methods of on-site calibration of inertial sensors IMU by accelerometers also have problems such as poor accuracy and low efficiency.

Summary of the invention

In view of this, embodiments of the present application provide an accelerometer on-site calibration method and apparatus to eliminate or improve one or more defects existing in the prior art.

One aspect of the present application provides an accelerometer field calibration method, comprising:

Detect whether the drone is currently in a stationary state after rotation according to a preset multi-criteria zero-speed interval detection method, and if so, obtain the initial value of the error parameter group of the accelerometer of the drone to be calibrated;

Based on the initial parameter values of the error parameter group, various errors of the inertial sensor corresponding to the error parameter group are calibrated respectively, and it is automatically determined whether the corresponding calibration result is in a completed state.

In some embodiments of the present application, the zero-speed interval detection method according to the preset multi-criteria detects whether the drone is currently in a stationary state after rotation, and if so, obtains the initial value of the error parameter group of the accelerometer of the drone to be calibrated, including:

Obtain the current zero-speed interval detection result of the drone according to the current acceleration differential modulus, angular velocity differential modulus and acceleration variance of the drone;

Screening out burr data in the zero-speed interval detection result based on a finite state machine to obtain a corresponding target detection result;

If the target detection result shows that the UAV is currently in a stationary state after rotation, the initial parameter values of the error parameter group to be calibrated of the accelerometer of the UAV are obtained.

In some embodiments of the present application, the step of obtaining the current zero-speed interval detection result of the drone according to the current acceleration differential modulus, angular velocity differential modulus, and acceleration variance of the drone includes:

Obtaining the current acceleration differential modulus and angular velocity differential modulus of the drone;

Determine whether the acceleration differential modulus and the angular velocity differential modulus are both less than their respective corresponding thresholds, and if so, obtain the current acceleration variance of the drone;

It is determined whether the acceleration variance is less than the corresponding threshold value. If so, the current zero-speed interval detection result of the drone is obtained.

In some embodiments of the present application, the error parameter group includes: non-orthogonal axis error rotation angle, zero point offset and scale parameter of the inertial sensor of the drone.

In some embodiments of the present application, before detecting whether the drone is currently in a stationary state after rotation according to the zero-speed interval detection method of the preset multi-criteria, and if so, obtaining the initial parameter value of the error parameter group to be calibrated of the accelerometer of the drone, it also includes:

Determine the local gravity acceleration of the drone according to the local latitude value and the local altitude value of the drone location;

and, establishing an error model of the accelerometer of the drone;

Correspondingly, the various errors of the inertial sensor corresponding to the error parameter group are calibrated based on the initial parameter values of the error parameter group, including:

The initial value of the non-orthogonal axis error rotation angle of the inertial sensor of the UAV is set to 0;

Simplifying the error model of the accelerometer, and generating an initial value of the zero point offset of the inertial sensor based on the simplified processing result;

and, generating an initial value of a scale parameter of the inertial sensor according to the local gravitational acceleration;

Generate initial values of parameters of an error parameter group based on the initial value of the non-orthogonal axis error rotation angle, the initial value of the zero point offset and the initial value of the scale parameter;

The initial values of the parameters of the error parameter group are used to calibrate the non-orthogonal axis error rotation angle, the zero point offset and the scale parameter respectively.

In some embodiments of the present application, the use of the initial values of the parameters of the error parameter group to calibrate the non-orthogonal axis error rotation angle, the zero point offset and the scale parameter respectively includes:

Obtaining a nonlinear least squares regression fitting optimization function of the error parameter group;

The nonlinear least squares regression fitting optimization function is optimized based on the trust region Dogleg algorithm, and the optimized nonlinear least squares regression fitting optimization function is iterated based on the initial parameter values of the error parameter group to obtain the calibration results corresponding to the non-orthogonal axis error rotation angle, the zero point offset and the scale parameter.

In some embodiments of the present application, the automatically determining whether the corresponding calibration result is in a completed state includes:

Based on a preset calibration accuracy factor, it is automatically determined whether the calibration results corresponding to the various errors of the inertial sensor are in a completed state.

Another aspect of the present application provides an accelerometer field calibration device, comprising:

The zero-speed interval detection module is used to detect whether the drone is currently in a stationary state after rotation according to a preset multi-criteria zero-speed interval detection method. If so, the initial value of the error parameter group of the accelerometer of the drone to be calibrated is obtained;

The error calibration and result determination module is used to calibrate the various errors of the inertial sensor corresponding to the error parameter group based on the initial parameter values of the error parameter group, and automatically determine whether the corresponding calibration result is in a completed state.

Another aspect of the present application provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the accelerometer field calibration method when executing the computer program.

Another aspect of the present application provides a computer-readable storage medium having a computer program stored thereon, and the computer program implements the accelerometer field calibration method when executed by a processor.

The accelerometer field calibration method of the present application detects whether the drone is currently in a stationary state after rotation according to a preset multi-criteria zero-speed interval detection method, which can realize accurate and rapid judgment of the zero-speed interval detection of the accelerometer, and can effectively improve the efficiency and accuracy of the zero-speed interval detection of the accelerometer, thereby effectively improving the efficiency and accuracy of the error parameters of the accelerometer field calibration; and because there is no need to adopt machine learning and other methods, there is no need to perform complex calculations, nor is there a need to use a large amount of data training, so the efficiency of the accelerometer field calibration error parameters can be further improved; by automatically determining whether the error calibration result of the inertial sensor is in a completed state, it can effectively solve the problem of low calibration efficiency and long calibration time caused by the need for manual determination of the calibration completion state, thereby further improving the efficiency of the accelerometer field calibration error parameters, and can effectively improve the degree of automation and reliability of the accelerometer field calibration error parameters.

Additional advantages, purposes, and features of the present application will be partially described in the following description, and will become partially apparent to those skilled in the art after studying the following, or may be learned from the practice of the present application. The purposes and other advantages of the present application can be achieved and obtained by the structures specifically pointed out in the specification and the drawings.

Those skilled in the art will understand that the purposes and advantages that can be achieved by the present application are not limited to the above specific description, and the above and other purposes that can be achieved by the present application will be more clearly understood based on the following detailed description.

BRIEF DESCRIPTION OF THE DRAWINGS

The drawings described herein are used to provide a further understanding of the present application, constitute a part of the present application, and do not constitute a limitation of the present application. The components in the drawings are not drawn to scale, but are only for the purpose of illustrating the principles of the present application. In order to facilitate the illustration and description of some parts of the present application, the corresponding parts in the drawings may be enlarged, that is, they may become larger relative to other components in the exemplary device actually manufactured according to the present application. In the drawings:

FIG1 is a schematic diagram of the overall flow of an accelerometer field calibration method in an embodiment of the present application.

FIG. 2 is a schematic diagram of a specific flow chart of an accelerometer field calibration method in an embodiment of the present application.

FIG3 is a schematic structural diagram of an accelerometer field calibration device in another embodiment of the present application.

FIG4 is a schematic diagram of the overall calibration process of the accelerometer field calibration method based on the finite state machine and precision factor provided in the application example of the present application.

FIG5 is a schematic diagram of an algorithm flow of zero-speed interval detection based on multiple criteria provided in an application example of the present application.

FIG6 is a schematic diagram of the burr phenomenon in the multi-criteria zero-speed interval detection provided in the application example of the present application.

FIG. 7 is a schematic diagram of the zero-speed interval detection state transition of a finite state machine based on multiple criteria provided in an application example of the present application.

DETAILED DESCRIPTION

In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the implementation modes and the accompanying drawings. Here, the illustrative implementation modes and descriptions of the present application are used to explain the present application, but are not intended to limit the present application.

It should also be noted here that in order to avoid obscuring the present application due to unnecessary details, only the structures and/or processing steps closely related to the scheme according to the present application are shown in the accompanying drawings, while other details that are not very relevant to the present application are omitted.

It should be emphasized that the term “include/comprises” when used herein refers to the presence of features, elements, steps or components, but does not exclude the presence or addition of one or more other features, elements, steps or components.

It should also be noted that, unless otherwise specified, the term “connection” herein may refer not only to a direct connection but also to an indirect connection with an intermediate.

Hereinafter, embodiments of the present application will be described with reference to the accompanying drawings. In the accompanying drawings, the same reference numerals represent the same or similar components, or the same or similar steps.

One of the methods for on-site calibration of accelerometers is to use the least squares fitting method. By repeatedly flipping the system output inertial data, the least squares fitting method is applied to solve the various error coefficients of the inertial sensor IMU. However, this method takes a long time to collect inertial data, is time-consuming, and has cumbersome operations. It is easy to cause errors caused by human intervention, which affects the calibration accuracy.

The second method of accelerometer on-site calibration is to consider using deep learning methods, taking the measured output information of the MEMS accelerometer as input, using deep learning algorithms for error compensation, and being able to predict the key error parameters of the MEMS inertial navigation. However, the deep learning method has a large amount of calculation, a complex network, and a large amount of data requirements. The quality of the deep learning network will also affect the calibration results.

That is to say, the existing error detection methods of inertial sensors IMU have problems such as poor accuracy and low efficiency.

Based on this, in order to solve the problems of poor detection accuracy in existing accelerometer field calibration methods, resulting in poor calibration parameter accuracy, and low detection efficiency due to complex or time-consuming detection process, the present application provides an accelerometer field calibration method, an accelerometer field calibration device for executing the accelerometer field calibration method, an electronic device as an entity of the accelerometer field calibration device, and a storage medium. The accelerometer field calibration method detects whether the drone is currently in a stationary state after rotation according to a preset multi-criteria zero-speed interval detection method, and can achieve accurate and rapid judgment of the accelerometer's zero-speed interval detection, which can effectively improve the accelerometer's zero-speed interval detection efficiency. The efficiency and accuracy of time detection can be improved, thereby effectively improving the efficiency and accuracy of on-site calibration error parameters of the accelerometer; and since there is no need to adopt machine learning and other methods, there is no need to perform complex calculations or use a large amount of data training, thereby further improving the efficiency of on-site calibration error parameters of the accelerometer; by automatically determining whether the error calibration result of the inertial sensor is in a completed state, it can effectively solve the problem of low calibration efficiency and long calibration time caused by the need for manual determination of the calibration completion state, thereby further improving the efficiency of on-site calibration error parameters of the accelerometer, and effectively improving the degree of automation and reliability of on-site calibration error parameters of the accelerometer.

In one or more embodiments of the present application, IMU refers to an inertial measurement unit or an inertial sensor.

Based on the above content, the present application also provides an accelerometer field calibration device for implementing the accelerometer field calibration method provided in one or more embodiments of the present application. The accelerometer field calibration device can be a server. The accelerometer field calibration device can communicate with the drone and the control station by itself or through a third-party server, etc. to obtain various sensor data corresponding to the drone, and execute the accelerometer field calibration method mentioned in the embodiments of the present application based on these data. After obtaining the final calibration result, the calibration result can be sent to the control station or the client device held by the operation and maintenance personnel for display, so that the operation and maintenance personnel can promptly know and analyze the calibration result.

In addition, the part of the accelerometer field calibration device that performs the accelerometer field calibration can be executed in the server as described above, and in another practical application scenario, all operations can also be completed in the client device. The specific selection can be based on the processing capability of the client device and the limitations of the user's usage scenario. This application does not limit this. If all operations are completed in the client device, the client device may also include a processor for specific processing of the accelerometer field calibration.

The client device may have a communication module (i.e., a communication unit) that can communicate with a remote server to achieve data transmission with the server. The server may include a server on the task scheduling center side, and other implementation scenarios may also include a server on an intermediate platform, such as a server on a third-party server platform that has a communication link with the task scheduling center server. The server may include a single computer device, or a server cluster consisting of multiple servers, or a server structure of a distributed device.

The server and the client device may communicate with each other using any suitable network protocol, including network protocols that have not been developed on the date of filing this application. The network protocols may include, for example, TCP/IP protocol, UDP/IP protocol, HTTP protocol, HTTPS protocol, etc. Of course, the network protocols may also include, for example, RPC protocol (Remote Procedure Call Protocol) and REST protocol (Representational State Transfer) used on top of the above protocols.

The details are described in detail through the following embodiments and application examples.

In order to solve the problems of poor accuracy and low efficiency in the existing zero-speed interval detection methods, which also lead to the problems of poor accuracy and low efficiency in the existing methods of calibrating the error parameters of the inertial sensor IMU on-site by the accelerometer, the present application provides an embodiment of an accelerometer on-site calibration method, referring to FIG1, the accelerometer on-site calibration method performed by the accelerometer on-site calibration device specifically includes the following contents:

Step 100: Detect whether the drone is currently in a stationary state after rotation according to a preset multi-criteria zero-speed interval detection method. If so, obtain the initial parameter values of the error parameter group to be calibrated of the accelerometer of the drone.

In step 100, if it is detected that the drone is not currently in a stationary state after rotation according to the preset multi-criteria zero-speed interval detection method, no subsequent execution is performed, and step 100 can be re-executed after a preset time interval until it is determined that the drone is currently in a stationary state after rotation.

It can be understood that the post-rotation static state mentioned in step 100 refers to the transition of the drone from a rotating state to a static state, which is used to distinguish the initial static state of the drone when it is not in operation.

In addition, the multi-criteria zero-speed interval detection method refers to a method of using multiple detection standards or criteria to determine whether the drone is currently in a zero-speed interval (ie, a stationary state after rotation).

In one or more embodiments of the present application, the error parameter group includes multiple error parameters of the inertial sensor. It can be understood that the initial value of the parameter of the error parameter group refers to the initial value corresponding to each error parameter in the error parameter group.

Step 200: Based on the initial parameter values of the error parameter group, calibrate the errors of the inertial sensor corresponding to the error parameter group respectively, and automatically determine whether the corresponding calibration result is in a completed state.

From the above description, it can be seen that the accelerometer field calibration method provided in the embodiment of the present application can realize accurate and rapid judgment of the zero-speed interval detection of the accelerometer by detecting whether the drone is currently in a stationary state after rotation according to a preset multi-criteria zero-speed interval detection method, and can effectively improve the efficiency and accuracy of the zero-speed interval detection of the accelerometer, thereby effectively improving the efficiency and accuracy of the error parameters of the accelerometer field calibration; and since there is no need to adopt machine learning and other methods, there is no need to perform complex calculations, nor is there a need to use a large amount of data training, so the efficiency of the accelerometer field calibration error parameters can be further improved; by automatically determining whether the error calibration result of the inertial sensor is in a completed state, it can effectively solve the problem of low calibration efficiency and long calibration time caused by the need for manual determination of the calibration completion state, thereby further improving the efficiency of the accelerometer field calibration error parameters, and can effectively improve the degree of automation and reliability of the accelerometer field calibration error parameters.

In order to further effectively improve the efficiency and accuracy of the zero-speed interval detection of the accelerometer, in an embodiment of an accelerometer field calibration method provided in the present application, referring to FIG. 2 , step 100 of the accelerometer field calibration method specifically includes the following contents:

Step 110: Obtain the current zero-speed interval detection result of the drone according to the current acceleration differential modulus, angular velocity differential modulus and acceleration variance of the drone.

Step 120: Screening out burr data in the zero-speed interval detection result based on a finite state machine to obtain a corresponding target detection result.

It is understandable that in actual situations, due to the intense data jumps of the accelerometer and gyroscope during the switching between motion and static states, there are misjudgment problems in the static interval, and there are some glitches in the multi-criteria zero-speed interval judgment. For this reason, a zero-speed interval detection state machine is proposed to improve the accuracy of zero-speed interval detection by removing data with a zero-speed interval time length of less than 1.5 seconds (i.e., glitch data).

Step 130: If the target detection result shows that the UAV is currently in a stationary state after rotation, the initial parameter values of the error parameter group to be calibrated of the accelerometer of the UAV are obtained.

From the above description, it can be seen that the accelerometer field calibration method provided in the embodiment of the present application can further realize the accurate and rapid judgment of the zero-speed interval detection of the accelerometer by obtaining the current zero-speed interval detection result of the drone according to the current acceleration differential modulus, angular velocity differential modulus and acceleration variance of the drone, and can further effectively improve the efficiency and accuracy of the zero-speed interval detection of the accelerometer, thereby effectively improving the efficiency and accuracy of the accelerometer field calibration error parameters; by performing data screening on the zero-speed interval detection result based on a finite state machine to obtain the corresponding target detection result, the burr data of the zero-speed interval (for example: data with a time length of less than 1.5 seconds) can be effectively removed, further improving the accuracy of the zero-speed interval detection.

In order to further reduce the computational complexity of the zero-speed interval detection process and improve the detection efficiency, in an embodiment of an accelerometer field calibration method provided in the present application, step 110 of the accelerometer field calibration method specifically includes the following contents:

Step 111: Obtain the current acceleration differential modulus and angular velocity differential modulus of the UAV.

Step 112: Determine whether the acceleration differential modulus and the angular velocity differential modulus are both smaller than their corresponding thresholds; if so, execute step 113.

Step 113: Obtain the current acceleration variance of the drone.

Step 114: Determine whether the acceleration variance is less than a corresponding threshold value, and if so, execute step 115.

Step 115: Obtain the current zero-speed interval detection result of the drone.

Specifically, in response to the problems of poor zero-speed interval detection accuracy and complex calculation in traditional on-site calibration methods, a zero-speed interval detection method based on a finite state machine with multiple criteria was proposed. The method judged whether the vehicle was in a zero-speed state based on three criteria: the differential modulus of acceleration and angular velocity and the acceleration variance.

In order to reduce unnecessary calculations, when the differential modulus value is judged to be a non-stationary state, there is no need to further calculate the variance; when the differential modulus value is judged to be a stationary state, whether it is a stationary state is further determined based on the variance dimension.

From the above description, it can be seen that the accelerometer on-site calibration method provided in the embodiment of the present application first determines whether the acceleration differential modulus and the angular velocity differential modulus are both less than their respective corresponding thresholds. If so, the current acceleration variance of the drone is obtained, which can reduce unnecessary calculations. When the differential modulus is judged to be a non-stationary state, there is no need to further calculate the variance; when the differential modulus is judged to be a stationary state, it is further determined whether it is a stationary state based on the variance dimension, thereby further reducing the computational complexity of the zero-speed interval detection process and improving the detection efficiency.

In order to further improve the comprehensiveness and effectiveness of the calibration error, in an embodiment of an accelerometer field calibration method provided in the present application, the error parameter group in the accelerometer field calibration method includes: non-orthogonal axis error rotation angle, zero point offset and scale parameter of the inertial sensor of the drone.

From the above description, it can be seen that the accelerometer on-site calibration method provided in the embodiment of the present application, the error parameter group includes: the non-orthogonal axis error rotation angle, zero point offset and scale parameters of the inertial sensor of the drone, which can further improve the comprehensiveness and effectiveness of the calibration error, and thus can effectively improve the MEMS positioning accuracy.

In order to improve the efficiency and reliability of calibrating the errors of the inertial sensor corresponding to the error parameter group, in an embodiment of an accelerometer field calibration method provided in the present application, referring to FIG. 2 , the accelerometer field calibration method further specifically includes the following contents before step 110:

Step 010: Determine the local gravity acceleration of the UAV according to the local latitude value and the local altitude value of the location of the UAV.

Step 020: Establish an error model of the accelerometer of the drone.

Correspondingly, referring to FIG. 2 , step 200 of the accelerometer on-site calibration method specifically includes the following contents:

Step 210: setting the initial value of the non-orthogonal axis error rotation angle of the inertial sensor of the drone to 0.

Step 220: Simplify the error model of the accelerometer, and generate an initial value of the zero offset of the inertial sensor based on the simplified processing result.

Step 230: Generate an initial value of a scale parameter of the inertial sensor according to the local gravity acceleration.

Step 240: Generate initial values of parameters of the error parameter group based on the initial value of the non-orthogonal axis error rotation angle, the initial value of the zero point offset and the initial value of the scale parameter.

Step 250: using the initial parameter values of the error parameter group to calibrate the non-orthogonal axis error rotation angle, the zero point offset and the scale parameter respectively.

From the above description, it can be seen that the accelerometer field calibration method provided in the embodiment of the present application, by pre-acquiring the gravity acceleration and the error model, and simplifying the error model of the accelerometer, and generating the initial value of the zero point offset of the inertial sensor based on the simplified processing result; and, generating the initial value of the scale parameter of the inertial sensor according to the local gravity acceleration, can effectively improve the efficiency and reliability of calibrating each error of the inertial sensor corresponding to the error parameter group, so as to further improve the efficiency and reliability of the accelerometer field calibration.

In order to solve the problem that the Hessian matrix (Hessian) is non-positive definite and irreversible in the Gauss-Newton calculation step process during the calculation process, in an embodiment of an accelerometer field calibration method provided by the present application, step 250 of the accelerometer field calibration method specifically includes the following content:

Step 251: Obtain a nonlinear least squares regression fitting optimization function of the error parameter group.

Step 252: Optimize the nonlinear least squares regression fitting optimization function based on the trust region Dogleg algorithm, and iterate the optimized nonlinear least squares regression fitting optimization function based on the initial parameter values of the error parameter group to obtain the calibration results corresponding to the non-orthogonal axis error rotation angle, the zero point offset and the scale parameter.

It can be understood that the trust-region Dogleg algorithm may specifically refer to the Dogleg algorithm in the trust-region algorithm. The trust-region algorithm (Trust-region methods) is also called the TR method, which is an optimization method that can ensure the overall convergence of the optimization method.

From the above description, it can be seen that the accelerometer field calibration method provided in the embodiment of the present application optimizes the nonlinear least squares regression fitting optimization function based on the trust region Dogleg algorithm, and iterates the optimized nonlinear least squares regression fitting optimization function based on the initial parameter values of the error parameter group. It can effectively solve the problem of non-positive definite and irreversible Hessian matrix (Hessian) in the Gauss-Newton calculation step process during the calculation process, thereby ensuring the stability of the iterative search process of the angular velocity meter calibration parameters.

In order to improve the efficiency and reliability of calibrating the errors of the inertial sensor corresponding to the error parameter group, in an embodiment of an accelerometer field calibration method provided in the present application, referring to FIG. 2 , the accelerometer field calibration method further specifically includes the following contents after step 250 in step 200:

Step 260: Based on a preset calibration accuracy factor, automatically determine whether the calibration results corresponding to the various errors of the inertial sensor are in a completed state.

From the above description, it can be seen that the accelerometer field calibration method provided in the embodiment of the present application automatically determines whether the calibration results corresponding to the various errors of the inertial sensor are in a completed state based on a preset calibration accuracy factor. Compared with the calibration accuracy assessment method that directly adopts the residual sum of squares, the proposed calibration state judgment calibration based on the calibration accuracy factor is independent of the zero-speed interval size M, can meet the calibration accuracy assessment of zero-speed intervals of different sizes, and is more practical.

In view of the embodiment of the above-mentioned accelerometer field calibration method, the present application also provides an accelerometer field calibration device for implementing the accelerometer field calibration method. Referring to FIG. 3 , the accelerometer field calibration device specifically includes the following contents:

The zero-speed interval detection module 10 is used to detect whether the drone is currently in a stationary state after rotation according to a preset multi-criteria zero-speed interval detection method, and if so, obtain the initial value of the error parameter group of the accelerometer of the drone to be calibrated;

The error calibration and result determination module 20 is used to calibrate the various errors of the inertial sensor corresponding to the error parameter group based on the initial parameter values of the error parameter group, and automatically determine whether the corresponding calibration result is in a completed state.

From the above description, it can be seen that the accelerometer field calibration device provided in the embodiment of the present application can detect whether the drone is currently in a stationary state after rotation according to a preset multi-criteria zero-speed interval detection method, thereby realizing accurate and rapid judgment of the zero-speed interval detection of the accelerometer, and can effectively improve the efficiency and accuracy of the zero-speed interval detection of the accelerometer, thereby effectively improving the efficiency and accuracy of the accelerometer field calibration error parameters; and since there is no need to adopt machine learning and other methods, there is no need to perform complex calculations, nor is there a need to use a large amount of data training, so the efficiency of the accelerometer field calibration error parameters can be further improved; by automatically determining whether the error calibration result of the inertial sensor is in a completed state, it can effectively solve the problem of low calibration efficiency and long calibration time caused by the need for manual determination of the calibration completion state, thereby further improving the efficiency of the accelerometer field calibration error parameters, and can effectively improve the degree of automation and reliability of the accelerometer field calibration error parameters.

In addition, in order to further illustrate the accelerometer field calibration method mentioned above in this application, this application also provides a specific application example of the accelerometer field calibration method for further explanation, specifically an accelerometer field calibration method based on a finite state machine and precision factor, which is specifically described as follows:

In view of the problems existing in the field calibration methods of accelerometers, such as poor detection accuracy resulting in poor calibration parameter accuracy, and low detection efficiency due to complex or time-consuming detection process, the present application proposes a zero-speed interval detection method based on multiple criteria to achieve accurate and rapid determination of the zero-speed interval of the accelerometer; in view of the problem that in the existing field calibration methods, the calibration completion status needs to be manually determined, resulting in low calibration efficiency and long calibration time, a calibration accuracy factor is proposed to automatically determine whether the calibration status is completed.

Referring to FIG4 , the overall calibration process of the accelerometer field calibration method based on the finite state machine and precision factor is as follows:

(I) Local normal gravity acceleration g _l and accelerometer error model

The local gravity acceleration g _l can usually be approximated by using the global level surface to determine the normal gravity field formula. The local normal gravity field formula is:

g _l ＝g _e [1-0.0053sin ² φ+3.0159×10 ^-6 sin ² (2φ)]-3.0828×10 ^-6 h (1)

Among them, g _e is the gravitational acceleration value at the equator, φ is the local latitude value, and h is the local altitude value.

The error model of a low-cost accelerometer can be expressed as:

a ^I =T ^a SM ^a (a ^O + ^ba +v ^a ) (2)

Where ^aI is the acceleration in the orthogonal coordinate system IC, ^aO is the acceleration in the actual accelerometer coordinate system AC, is the accelerometer measurement noise, which is usually assumed to obey a zero-mean Gaussian distribution, ^SMa is the scale matrix, and ^Ba is the zero bias vector. The drift ^Ba usually changes slowly over time and is usually modeled as a random walk. During the on-site calibration process, ^Ba can be considered a constant due to the short calibration time.

(II) Zero-speed interval detection method based on finite state machine with multiple criteria

Aiming at the problems of poor detection accuracy and complex calculation of zero-speed interval in traditional on-site calibration methods, a zero-speed interval detection method based on a finite state machine with multiple criteria was proposed. The method judged whether the vehicle was in the zero-speed state based on three criteria: the differential modulus of acceleration and angular velocity and the acceleration variance.

In order to reduce unnecessary calculations, when the differential modulus value is judged to be a non-stationary state, there is no need to further calculate the variance; when the differential modulus value is judged to be a stationary state, whether it is a stationary state is further determined according to the variance dimension. In summary, the algorithm flow chart of zero-speed interval detection based on multiple criteria is shown in Figure 5.

Taking the accelerometer as an example, the effectiveness of the differential-based UAV motion state determination is analyzed. First, the difference between two adjacent sampling moments in formula (2) can be obtained:

When the drone is stationary, the left side of the equation should be a 0 vector, because the matrix SM ^a T ^a is not a zero matrix, so we can get:

When the drone is stationary, the above equation satisfies the inequality:

Where th ^a is the threshold for judging the stationary state of the drone, and its selection is related to the random walk and Gaussian white noise of the accelerometer.

The three criteria for detection are as follows:

(1) Accelerometer differential modulus

When the drone is stationary, the differential modulus of the accelerometer output value should be less than the set threshold value th ^a , and the differential modulus of the accelerometer output value at time t can be obtained. for:

The criterion for judging whether the drone is stationary can be expressed as:

(2) Accelerometer variance

When the drone is stationary, the variance of the accelerometer output value should be less than the set threshold By approximating the accelerometer variance with the sample variance of the accelerometer, we can obtain:

Where w is the window size, is the average value within the window, defined as:

The criterion for judging whether the drone is stationary can be expressed as:

(3) Gyroscope differential modulus

When the drone is stationary, the differential modulus of the gyroscope output value should be less than the set threshold th ^η , and the differential output modulus of the accelerometer at time t can be obtained for:

The criterion for judging whether the drone is stationary can be expressed as:

Among them, the judgment criterion of the differential modulus value in Figure 5 is and In actual situations, due to the intense data jumps of the accelerometer and gyroscope during the switching between motion and static states, there are misjudgment problems in the static interval. See Figure 6. There are some glitches in the multi-criteria zero-speed interval judgment. For this reason, a zero-speed interval detection state machine is proposed. The state transition is shown in Figure 7. By removing the data with a zero-speed interval time length of less than 1.5 seconds, the accuracy of zero-speed interval detection is improved.

(III) Zero point offset and initial value of scale parameters

The parameter estimation accuracy of the trust region Dogleg algorithm is related to the initial value setting of the error parameter. Usually, the non-orthogonal axis error of the accelerometer is small, and the non-orthogonal rotation angle is near 0, so the initial value of the non-orthogonal axis error rotation angle can be selected as 0, that is, α _yz = α _zy = α _zx = 0. Therefore, in the process of determining the initial value, the accelerometer error model of formula (2) can be simplified as:

When the accelerometer is pointing vertically upward in the same direction as the gravity acceleration, due to the scale parameter and If it is not 0, the initial value of zero offset is:

Then the scale parameter is determined according to the local gravitational acceleration and

(IV) Acceleration field calibration method based on trust region Dogleg

Accelerometer calibration requires estimation of zero bias, scale, and non-orthogonal axis deviation, a total of 9 parameters:

The nonlinear least squares regression fitting optimization function is shown in formula (17):

Where ||·|| ₂ represents the Euclidean norm, M is the number of sampling points in the zero-speed interval, represents the accelerometer sampling value at time k, and g _l is the local gravity acceleration.

Aiming at the problem that the Hessian matrix is non-positive definite and irreversible during the Gauss-Newton calculation step, an improved trust region Dogleg algorithm is proposed based on the singular value decomposition (SVD) method to ensure the stability of the iterative search process of the angular velocity meter calibration parameters.

According to the trust region Dogleg algorithm, the function F(e ^a ) is defined as:

F(e ^a )=[f ₁ (e ^a ) f ₂ (e ^a ) ...f _M (e ^a )] (18)

The function f _k (e ^a ) is defined as:

Then the optimization problem corresponding to formula (17) is equivalent to:

The iterative process of the trust region Dogleg algorithm is shown in Table 1.

Table 1

The iteration termination condition may be selected as infinitesimal thresholds th ₁ , th ₂ , th ₃ , such as 10 ⁻¹¹ . The selection of the thresholds has nothing to do with the final iteration accuracy.

(V) Calibration status determination based on calibration accuracy factor

Traditional methods often use a judgment method based on the residual sum of squares, but the size of the residual sum of squares is related to the number of zero-speed interval sampling M, and cannot correctly characterize the calibration error. In order to evaluate the calibration capability of accelerometer parameters and realize automatic discrimination of calibration status, the statistical characteristics of the residual in formula (17) are first analyzed, the expectation of the residual is derived, the relationship between the expectation and M is proved, the calibration accuracy factor is proposed, and the automatic discrimination of the calibration status is realized.

According to the accelerometer error model, it is assumed that Satisfying the independent zero mean homoscedastic Gaussian distribution, the corrected Should satisfy Gaussian distribution, that is in Defined as:

Usually, the scale is very close to 1, so It can be approximately equivalent to Since the mean values of the XYZ axes are inconsistent after correction, and the true projection of the gravity acceleration on the XYZ axes cannot be determined during the calibration process, The probability density function does not have an analytical form. For this reason, the mean of the residual sum of squares is analyzed, and based on the relationship between the mean and M, a calibration accuracy factor that is independent of the size of the zero-speed interval is proposed for calibration state judgment.

Define the calibration precision factor as r:

The calibration precision factor has nothing to do with the size of the zero-speed interval M, and can meet the calibration precision evaluation of zero-speed intervals of different sizes. Compared with directly using the residual sum of squares, it has more practical value. The smaller the calibration precision factor, the higher the calibration precision. When the calibration precision factor decreases gently, the calibration can be considered completed.

To sum up, the application example of the present application can accurately determine the zero-speed space and reduce the influence of burrs. Specifically, the zero-speed interval detection method of the finite state machine based on multiple criteria can more accurately determine the zero-speed space; and compared with the calibration accuracy assessment method that directly adopts the residual sum of squares, the calibration state judgment method based on the calibration accuracy factor proposed in the application example of the present application is independent of the zero-speed interval size M, can meet the calibration accuracy assessment of zero-speed intervals of different sizes, has more practical value, and is a more effective and practical calibration state judgment method of the calibration accuracy factor.

The embodiment of the present application also provides a computer device (i.e., an electronic device), which may include a processor, a memory, a receiver, and a transmitter, wherein the processor is used to execute the accelerometer field calibration method mentioned in the above embodiment, wherein the processor and the memory may be connected via a bus or other means, with the bus connection being taken as an example. The receiver may be connected to the processor and the memory via a wired or wireless manner. The computer device is communicatively connected to the accelerometer field calibration device to receive real-time motion data from sensors in the wireless multimedia sensor network, and to receive original video sequences from the video acquisition device.

The processor may be a central processing unit (CPU). The processor may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or a combination of the above chips.

As a non-transient computer-readable storage medium, the memory can be used to store non-transient software programs, non-transient computer executable programs and modules, such as the program instructions/modules corresponding to the accelerometer field calibration method in the embodiment of the present application. The processor executes various functional applications and data processing of the processor by running the non-transient software programs, instructions and modules stored in the memory, that is, the accelerometer field calibration method in the above method embodiment is implemented.

The memory may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required by at least one function; the data storage area may store data created by the processor, etc. In addition, the memory may include a high-speed random access memory, and may also include a non-transitory memory, such as at least one disk storage device, a flash memory device, or other non-transitory solid-state storage device. In some embodiments, the memory may optionally include a memory remotely arranged relative to the processor, and these remote memories may be connected to the processor via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

The one or more modules are stored in the memory, and when executed by the processor, the accelerometer field calibration method in the embodiment is executed.

In some embodiments of the present application, the user equipment may include a processor, a memory, and a transceiver unit, which may include a receiver and a transmitter. The processor, memory, receiver, and transmitter may be connected through a bus system. The memory is used to store computer instructions, and the processor is used to execute the computer instructions stored in the memory to control the transceiver unit to send and receive signals.

As an implementation method, the functions of the receiver and the transmitter in the present application can be considered to be implemented through a transceiver circuit or a dedicated chip for transceiver, and the processor can be considered to be implemented through a dedicated processing chip, a processing circuit or a general chip.

As another implementation method, it is possible to use a general-purpose computer to implement the server provided in the embodiment of the present application, that is, to store the program code for implementing the functions of the processor, receiver, and transmitter in a memory, and the general-purpose processor implements the functions of the processor, receiver, and transmitter by executing the code in the memory.

The embodiment of the present application also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the above-mentioned accelerometer field calibration method are implemented. The computer-readable storage medium can be a tangible storage medium, such as a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a floppy disk, a hard disk, a removable storage disk, a CD-ROM, or any other form of storage medium known in the technical field.

It should be understood by those skilled in the art that the exemplary components, systems and methods described in conjunction with the embodiments disclosed herein can be implemented in hardware, software or a combination of the two. Whether it is specifically performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but this implementation should not be considered to be beyond the scope of this application. When implemented in hardware, it can be, for example, an electronic circuit, an application specific integrated circuit (ASIC), appropriate firmware, a plug-in, a function card, etc. When implemented in software, the elements of this application are programs or code segments used to perform the required tasks. The program or code segment can be stored in a machine-readable medium, or transmitted on a transmission medium or a communication link by a data signal carried in a carrier.

It should be clear that the present application is not limited to the specific configuration and processing described above and shown in the figures. For the sake of simplicity, a detailed description of the known method is omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present application is not limited to the specific steps described and shown, and those skilled in the art can make various changes, modifications and additions, or change the order between the steps after understanding the spirit of the present application.

In the present application, features described and/or illustrated for one embodiment may be used in the same manner or in a similar manner in one or more other embodiments, and/or combined with features of other embodiments or replace features of other embodiments.

The above description is only the preferred embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the embodiments of the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims (5)

1. A method for accelerometer field calibration, comprising: Determine the local gravity acceleration of the drone according to the local latitude value and the local altitude value of the drone location; And, establish an error model of the accelerometer of the drone; the error model of the accelerometer is shown in the following formula: a ^I =T ^a SM ^a (a ^O + ^ba + ^va ) Where ^aI is the acceleration in the orthogonal coordinate system IC, ^aO is the acceleration in the actual accelerometer coordinate system AC, is the accelerometer measurement noise, which is assumed to obey a zero-mean Gaussian distribution, SM ^a is the scale matrix, and is the scale parameter; ^ba is the zero bias vector; and is the zero point offset parameter; α _yz , α _zy , α _zx are the non-orthogonal axis error rotation angles; According to a preset multi-criteria zero-speed interval detection method, it is detected whether the drone is currently in a stationary state after rotation. If so, the initial value of the parameter of the error parameter group to be calibrated of the accelerometer of the drone is obtained; the error parameter group includes: the non-orthogonal axis error rotation angle, zero point offset and scale parameter of the inertial sensor of the drone; Based on the initial values of the parameters of the error parameter group, calibrate the errors of the inertial sensor corresponding to the error parameter group respectively, and automatically determine whether the corresponding calibration result is in a completed state; Among them, the zero-speed interval detection method based on the preset multi-criteria detects whether the drone is currently in a stationary state after rotation, and if so, obtains the initial value of the error parameter group of the accelerometer of the drone to be calibrated, including: obtaining the current acceleration differential modulus and angular velocity differential modulus of the drone; Determine whether the acceleration differential modulus and the angular velocity differential modulus are both less than their respective corresponding thresholds, and if so, obtain the current acceleration variance of the drone; Determine whether the acceleration variance is less than the corresponding threshold value, and if so, obtain the current zero-speed interval detection result of the drone; Screening out burr data in the zero-speed interval detection result based on a finite state machine to obtain a corresponding target detection result; If the target detection result shows that the UAV is currently in a stationary state after rotation, then obtaining the initial value of the error parameter group of the accelerometer of the UAV to be calibrated; Correspondingly, the various errors of the inertial sensor corresponding to the error parameter group are calibrated based on the initial parameter values of the error parameter group, including: Since the parameter estimation accuracy of the trust region Dogleg algorithm is related to the initial value setting of the error parameter, the initial value of the non-orthogonal axis error rotation angle of the inertial sensor of the UAV is set to 0, that is, α _yz =α _zy =α _zx =0; The error model of the accelerometer is simplified, and the initial value of the zero offset of the inertial sensor is generated based on the corresponding simplified processing result; the simplified error model of the accelerometer is shown in the following formula: is the acceleration in the orthogonal coordinate system IC at the previous sampling moment between two adjacent sampling moments; is the acceleration in the orthogonal coordinate system IC at the previous sampling moment when the drone is stationary; and, generating the initial value of the scale parameter of the inertial sensor according to the local gravity acceleration; Generate initial values of parameters of an error parameter group based on the initial value of the non-orthogonal axis error rotation angle, the initial value of the zero point offset and the initial value of the scale parameter; Using the initial values of the parameters of the error parameter group to calibrate the non-orthogonal axis error rotation angle, the zero point offset and the scale parameter respectively; The adopting the initial values of the parameters of the error parameter group to calibrate the non-orthogonal axis error rotation angle, the zero point offset and the scale parameter respectively comprises: Obtaining a nonlinear least squares regression fitting optimization function of the error parameter group; The nonlinear least squares regression fitting optimization function is optimized based on the trust region Dogleg algorithm, and the optimized nonlinear least squares regression fitting optimization function is iterated based on the initial parameter values of the error parameter group to obtain the calibration results corresponding to the non-orthogonal axis error rotation angle, the zero point offset and the scale parameter. 2. The accelerometer on-site calibration method according to claim 1, characterized in that the automatically determining whether the corresponding calibration result is in a completed state comprises: Based on a preset calibration accuracy factor, it is automatically determined whether the calibration results corresponding to the various errors of the inertial sensor are in a completed state. 3. An accelerometer field calibration device, characterized in that the accelerometer field calibration device is used to perform the following steps: Determine the local gravity acceleration of the drone according to the local latitude value and the local altitude value of the drone location; And, establish an error model of the accelerometer of the drone; the error model of the accelerometer is shown in the following formula: a ^I =T ^a SM ^a (a ^O + ^ba + ^va ) Where ^aI is the acceleration in the orthogonal coordinate system IC, ^aO is the acceleration in the actual accelerometer coordinate system AC, is the accelerometer measurement noise, which is assumed to obey a zero-mean Gaussian distribution, SM ^a is the scale matrix, and is the scale parameter; ^ba is the zero bias vector; and is the zero point offset parameter; α _yz , α _zy , α _zx are the non-orthogonal axis error rotation angles; The accelerometer on-site calibration device also includes: The zero-speed interval detection module is used to detect whether the drone is currently in a stationary state after rotation according to a preset multi-criteria zero-speed interval detection method. If so, the initial value of the parameter of the error parameter group to be calibrated of the accelerometer of the drone is obtained; the error parameter group includes: the non-orthogonal axis error rotation angle, zero point offset and scale parameter of the inertial sensor of the drone; An error calibration and result determination module, used to calibrate the various errors of the inertial sensor corresponding to the error parameter group based on the initial parameter values of the error parameter group, and automatically determine whether the corresponding calibration result is in a completed state; Among them, the zero-speed interval detection method based on the preset multi-criteria detects whether the drone is currently in a stationary state after rotation, and if so, obtains the initial value of the error parameter group of the accelerometer of the drone to be calibrated, including: obtaining the current acceleration differential modulus and angular velocity differential modulus of the drone; Determine whether the acceleration differential modulus and the angular velocity differential modulus are both less than their respective corresponding thresholds, and if so, obtain the current acceleration variance of the drone; Determine whether the acceleration variance is less than the corresponding threshold value, and if so, obtain the current zero-speed interval detection result of the drone; Screening out burr data in the zero-speed interval detection result based on a finite state machine to obtain a corresponding target detection result; If the target detection result shows that the UAV is currently in a stationary state after rotation, then obtaining the initial value of the error parameter group of the accelerometer of the UAV to be calibrated; Correspondingly, the various errors of the inertial sensor corresponding to the error parameter group are calibrated based on the initial parameter values of the error parameter group, including: The initial value of the non-orthogonal axis error rotation angle of the inertial sensor of the UAV is set to 0; Since the parameter estimation accuracy of the trust region Dogleg algorithm is related to the initial value setting of the error parameter, the initial value of the non-orthogonal axis error rotation angle is selected to be 0, that is, α _yz =α _zy =α _zx =0, the error model of the accelerometer is simplified, and the initial value of the zero offset of the inertial sensor is generated based on the corresponding simplified processing result; the simplified accelerometer error model is shown in the following formula: is the acceleration in the orthogonal coordinate system IC at the previous sampling moment between two adjacent sampling moments; is the acceleration in the orthogonal coordinate system IC at the previous sampling moment when the drone is stationary; and, generating the initial value of the scale parameter of the inertial sensor according to the local gravity acceleration; Generate initial values of parameters of an error parameter group based on the initial value of the non-orthogonal axis error rotation angle, the initial value of the zero point offset and the initial value of the scale parameter; Using the initial values of the parameters of the error parameter group to calibrate the non-orthogonal axis error rotation angle, the zero point offset and the scale parameter respectively; The adopting the initial values of the parameters of the error parameter group to calibrate the non-orthogonal axis error rotation angle, the zero point offset and the scale parameter respectively comprises: Obtaining a nonlinear least squares regression fitting optimization function of the error parameter group; The nonlinear least squares regression fitting optimization function is optimized based on the trust region Dogleg algorithm, and the optimized nonlinear least squares regression fitting optimization function is iterated based on the initial parameter values of the error parameter group to obtain the calibration results corresponding to the non-orthogonal axis error rotation angle, the zero point offset and the scale parameter. 4. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the accelerometer field calibration method according to claim 1 or 2 when executing the computer program. 5. A computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the accelerometer field calibration method according to claim 1 or 2.