CN118670390A - Multifunctional sensor data processing method and system - Google Patents

Abstract

The present invention discloses a multifunctional sensor data processing method and system, which uses quaternions to represent the attitude parameters of electronic equipment during collision, and combines gradient algorithm and quaternary element change rate to update and optimize the attitude parameters under quaternion representation in real time, so that the optimized quaternion attitude parameters can be more effectively used for subsequent collision analysis. In addition, the acceleration data collected by the collision is used for the dynamic adjustment of the learning rate in the gradient algorithm, and the angular velocity data collected by the collision is used for the quaternion change rate calculation, so as to achieve more real-time and effective parameter adjustment.

Description

A multifunctional sensor data processing method and system

Technical Field

The invention belongs to the field of computer data recognition and analysis, and in particular relates to the optimization adjustment of quaternion attitude parameters of an electronic device during a collision.

Background Art

Electronic devices are at risk of accidental drops and collisions in daily use, which may cause device damage, data loss, and even safety hazards. In order to improve the reliability and safety of electronic devices, it is of great significance to study the electronic device drop monitoring system. Attitude detection is to monitor the acceleration, angular velocity, magnetic field changes and other parameters of electronic devices in real time during the collision process through sensors such as accelerometers, gyroscopes and magnetometers.

In the field of posture detection data analysis, traditional algorithms such as filtering algorithms, peak detection algorithms and time-frequency analysis algorithms have been widely used. The traditional posture detection method mainly records acceleration, angular velocity and other data at the moment of collision, and uses Euler angles to represent the motion status at the moment of collision. Traditional methods also use other filters such as Kalman filters to improve the accuracy of posture estimation, which can suppress noise and interference.

However, in complex collision scenarios, the accuracy and robustness of these algorithms still need to be improved. Faced with key issues such as data inaccuracy, algorithm complexity, real-time data processing, model building and verification, and multi-scenario adaptability, the application of machine learning algorithms has become one of the solutions. In addition, traditional algorithms are constrained by many factors such as collision angle, collision speed, and type of electronic equipment. They lack sufficient versatility and adaptability, and one algorithm can only be applied to one model. Other algorithm designs are needed to adapt to different models, which invisibly increases cost expenditures.

Therefore, based on the complex parameter conditions in the electronic device collision scenario, it is crucial to obtain effective parameters for collision analysis in real time to form subsequent instructions for posture adjustment of the electronic device.

Summary of the invention

In order to solve the problems of the prior art, the present invention proposes a method for representing the posture parameters of an electronic device during a collision by quaternions, and combining a gradient algorithm and a quaternary element change rate to update and optimize the posture parameters under the quaternion representation in real time. The optimized quaternion posture parameters can be more effectively used for subsequent collision analysis. Specifically, the present invention relates to a multifunctional sensor data processing method, characterized in that it includes:

Parameter collection stage: The acceleration of the electronic device during the collision is collected through sensors Angular velocity And attitude parameter q, the acceleration is normalized to obtain The angular velocity is normalized to obtain The posture parameter q is represented by a quaternion.

Normalization can eliminate the scale differences between data and improve the stability and convergence speed of the algorithm.

Parameter optimization stage: Use the gradient descent algorithm to calculate the gradient of the objective function J(q) with respect to the quaternion, and adjust the quaternion value according to the update direction of the parameters during gradient iteration:

Where destangle _i is the desired angle, curangle _i is the current angle, n is the total number of samples, q ^′ is the quaternion value after iteration, q _cur is the current quaternion value, is the gradient function, a is the learning rate, which is determined by Where a _r is the preset current learning rate, and _gr is the preset normalized acceleration parameter.

The gradient descent algorithm can iteratively optimize the posture parameters, gradually improve the accuracy of posture estimation, and converge to the optimal solution more quickly.

Calculate the rate of change of a quaternion

in is the Hamilton product of quaternions, The real part is 0 and the imaginary part is Vector.

After integrating the quaternion, continue to update the value of the quaternion according to the time interval Δt between two attitude updates:

Where q _new is the updated quaternion value, and then the updated quaternion value is normalized to obtain the processed posture parameters.

Parameter application stage: performing collision analysis according to the processed posture parameters, forming a collision response instruction for the electronic device, and feeding back the collision response instruction to the electronic device.

Furthermore, the acceleration and the angular velocity are normalized, and the updated quaternion is normalized, including: taking the ratio of the corresponding acceleration, angular velocity or quaternion parameter to the modulus thereof as the result of the normalization.

Furthermore, the multifunctional sensor is a nine-axis sensor.

Furthermore, the electronic device and a computing device connected to the electronic device via near field communication; the method in the parameter application phase may be implemented by: the electronic device, a computing device connected to the electronic device via near field communication, and a remote server.

Furthermore, the near field communication method includes: Bluetooth connection and Wi-Fi connection.

Furthermore, the time interval Δt adopts a preset fixed time interval, or adopts a dynamic time interval.

Furthermore, the use of a dynamic time interval includes: on the basis of a fixed time interval, multiplying the fixed time interval value by an adjustment factor to obtain a dynamic time interval; wherein the value of the adjustment factor is based on Sure.

The present invention also relates to a multifunctional sensor data processing system, characterized in that it includes a parameter acquisition unit, a parameter optimization unit and a parameter application unit, wherein the parameter acquisition unit pre-processes the collected electronic device collision information and passes it to the parameter optimization unit, the parameter optimization unit optimizes the collision posture parameters under quaternion representation, and passes the optimized posture parameters to the parameter application unit for subsequent collision analysis; wherein each unit is specifically used to implement the steps of the method described in any one of claims 1-7.

The present invention also relates to a computer-readable storage medium having a computer program stored thereon, wherein the computer program implements the steps of the method according to any one of claims 1 to 7 when executed by a processor.

The present invention also relates to a computer program product, comprising a computer program, characterized in that when the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

The beneficial technical effects of this patent include:

(1) Quaternions are used to represent the attitude parameters of electronic devices during collision. Quaternions can be used to accurately describe rotations in any direction, unlike Euler angles, which may have singularity problems and have good properties. The attitude parameters represented by quaternions are updated and optimized in real time by combining the gradient algorithm and the rate of change of the four elements, so as to achieve more effective subsequent collision analysis.

(2) The acceleration data collected by the collision is used for the dynamic adjustment of the learning rate in the gradient algorithm, and the angular velocity data collected by the collision is used for the calculation of the quaternion change rate, so as to achieve more real-time and effective parameter adjustment;

(3) The data sources involved in the main framework of parameter optimization calculation are simple, and the calculation method can achieve effective attitude parameter adjustment without other auxiliary calculations, which is of great practical significance, especially for the collision analysis of electronic equipment equipped with nine-axis sensors.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG1 is a framework diagram of a multifunctional sensor data processing method according to an embodiment of the present invention.

FIG. 2 is a framework diagram of a multifunctional sensor data processing system according to an embodiment of the present invention.

DETAILED DESCRIPTION

In order to make the purpose, technical scheme and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

FIG1 shows a framework diagram of a multifunctional sensor data processing method under an embodiment. Accordingly, a multifunctional sensor data processing method involves a parameter acquisition phase, a parameter optimization phase and a parameter application phase.

Preferably, the multifunctional sensor is a nine-axis sensor. The nine-axis sensor integrates an accelerometer, a gyroscope and a magnetometer in three directions, and can fully obtain the motion state of the electronic device in space. The application of this sensor in posture detection, especially in collision scenarios, provides important data support for real-time monitoring of the force conditions of electronic devices.

In the parameter collection stage, the collision parameters of the electronic device are mainly collected, including normalizing the acceleration and angular velocity, and using quaternions to represent the posture parameters.

The output units of gyroscopes and accelerometers are different (gyroscope output is usually in angular velocity, and accelerometer output is usually in acceleration), and the outputs of different sensors may have different precisions. Normalization can map the data to the same scale range and reduce the scale differences between different sensor data. By normalizing the data to the same scale range, the range of data value changes can be reduced, the stability and convergence speed of the algorithm can be improved, and the accuracy of attitude estimation can be improved, as well as the error caused by the environment can be reduced.

Specifically, the acceleration of the electronic device during the collision is collected by sensors. Angular velocity And attitude parameter q, the acceleration is normalized to obtain The angular velocity is normalized to obtain The posture parameter q is represented by quaternion.

In one embodiment, the acceleration data is normalized and converted into a unit vector to eliminate the influence of the acceleration magnitude and retain its direction information: ||g|| represents acceleration The angular velocity data is normalized and the angular velocity unit is converted to radians per second for a unified representation:

Quaternion is a mathematical tool used to represent rotation in three-dimensional space. This embodiment uses quaternion to represent the posture parameters:

q＝q ₀ +q ₁ i +q ₂ j +q ₃ k

Where q ₀ is the real component, q ₁ , q ₂ , q ₃ are the three-dimensional imaginary components, and i, j, k are the basic units of the corresponding imaginary parts.

In the parameter optimization stage, the gradient descent algorithm is first used to calculate the gradient of the objective function J(q) with respect to the quaternion, and the quaternion value is adjusted according to the update direction of the parameters during gradient iteration:

Where destangle _i is the desired angle, curangle _i is the current angle, n is the total number of samples, q ^′ is the quaternion value after iteration, q _cur is the current quaternion value, is the gradient function and a is the learning rate.

The gradient descent algorithm optimizes the parameters of the posture estimation model, iteratively updates the parameters through the objective function, optimizes the estimated value of the quaternion, and makes the quaternion more accurate. The gradient descent algorithm has strong adaptability, can improve the robustness of the system, can apply the system to different collision situations, and improve the universality of the present invention.

Once we have the objective function, we need to find a direction to update the posture parameter q so that J(q) decreases. This direction is determined by the objective function J(q) about The gradient of is given by . The gradient is a vector whose components are the partial derivatives of the objective function with respect to the corresponding parameters, pointing in the direction of the fastest growth of the function. In order to minimize the objective function, we need to update the parameters in the opposite direction of the gradient. Therefore, the objective function J(q) provides a way to measure the quality of the current pose estimate, and the parameter update rule provides a way to optimize the pose parameters based on this objective function. By iteratively applying this update rule, we can gradually improve the pose estimate until we find the parameter value that minimizes the objective function.

Preferably, the learning rate is determined as Where a _r is the preset current learning rate, and _gr is the preset normalized acceleration parameter. Therefore, It is equivalent to an adjustment factor. By considering the significance of acceleration on the current state of the electronic device, it is applied to gradient learning and shows better learning effects in practical applications.

Further calculation of the rate of change of the quaternion

in is the Hamilton product of quaternions, The real part is 0 and the imaginary part is A vector of

After integrating the quaternion, continue to update the value of the quaternion according to the time interval Δt between two attitude updates:

Where q _new is the updated quaternion value, and then the updated quaternion value is normalized to obtain the processed posture parameters.

Preferably, the time interval Δt adopts a preset fixed time interval or a dynamic time interval. When the dynamic time interval is adopted, on the basis of the fixed time interval, the fixed time interval value is multiplied by the calculation result of the adjustment factor as the dynamic time interval, wherein the value of the adjustment factor is based on Those skilled in the art can understand that the collision transient represented by the acceleration parameter can reflect the intensity of the current state change of the electronic device, and determining the adjustment factor based on this can reduce the calculation overhead in unnecessary update cycles.

In one embodiment, the parameter optimization algorithm may be implemented by a computing device installed in the electronic device or other computing devices connected to the electronic device via near field communication, wherein the near field communication method includes: Bluetooth connection and Wi-Fi connection.

In the parameter application stage, a collision analysis is performed according to the processed posture parameters to form a collision response instruction for the electronic device, and the collision response instruction is fed back to the electronic device.

In one embodiment, the method functions in the parameter application phase may be implemented by an electronic device, another computing device connected to the electronic device by near field communication, or a remote server.

FIG2 shows a framework diagram of a multifunctional sensor data processing system under another embodiment, including a parameter acquisition unit, a parameter optimization unit and a parameter application unit, wherein the parameter acquisition unit pre-processes the collected electronic device collision information and transmits it to the parameter optimization unit, the parameter optimization unit optimizes the collision posture parameters under the quaternion representation, and transmits the optimized posture parameters to the parameter application unit for subsequent collision analysis. In addition, the above-mentioned units are also specifically used to implement the steps of the corresponding method under the embodiment described above.

The embodiments of the present disclosure further provide a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, it is also used to execute the steps of the corresponding method under the embodiments described above.

In addition, the embodiments of the present disclosure further provide a computer program product, including a computer program, which, when executed by a processor, is also used to execute the steps of the corresponding method under the embodiments described above.

The computer-readable storage medium may be an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or component, or any combination thereof. More specific examples of computer-readable storage media may include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in combination with an instruction execution system, device or device. In the present disclosure, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, wherein a computer-readable program code is carried. Such propagated data signals may take a variety of forms, including electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium, which may send, propagate, or transmit a program used by or in combination with an instruction execution system, device, or device. The program code contained on the computer-readable medium can be transmitted by any suitable medium, including: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above. The above-mentioned computer-readable medium can be contained in the above-mentioned electronic device; it can also exist alone without being assembled into the electronic device. The computer program code for performing the operation of the present disclosure can be written in one or more programming languages or a combination thereof, and the above-mentioned programming languages include object-oriented programming languages-such as Java, Smalltalk, C++, and also include conventional procedural programming languages-such as "C" language or similar programming languages. The program code can be executed completely on the user's computer, partially on the user's computer, as an independent software package, partially on the user's computer and partially on the remote computer, or completely on the remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any kind of network including a local area network (AN) or a wide area network (WAN), or can be connected to an external computer.

The flowcharts and block diagrams in the accompanying drawings illustrate the possible implementation architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present disclosure. Each box in the flowchart or block diagram may represent a module, a program segment, or a part of a code, which contains one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions marked in the box may also occur in an order different from that marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they may sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flowchart, and the combination of boxes in the block diagram or flowchart, can be implemented by a dedicated hardware-based system that performs the specified function or operation, or can be implemented by a combination of dedicated hardware and computer instructions. The units involved in the embodiments described in the present disclosure can be implemented by software or by hardware. Among them, the name of the unit does not constitute a limitation on the unit itself under certain circumstances.

The above introduces the preferred embodiments of the present invention, which is intended to make the spirit of the present invention clearer and easier to understand, but is not intended to limit the present invention. All modifications, substitutions, and improvements made within the spirit and principles of the present invention should be included in the scope of protection outlined by the claims attached to the present invention.

Claims (10)

1. A method of processing data of a multi-function sensor, comprising:

parameter acquisition: collecting acceleration of electronic equipment during collision through sensor Angular velocity ofAnd attitude parameter q, normalizing the acceleration to obtainNormalizing the angular velocity to obtainRepresenting the attitude parameter q by using a quaternion;

Parameter optimization stage: calculating the gradient of the objective function J (q) about the quaternion by using a gradient descent algorithm, and adjusting the quaternion according to the updating direction of the parameter during gradient iteration:

Where destangle _i is the desired angle, curangle _i is the current angle, n is the total number of samples, q ^′ is the iterated quaternion, q _cur is the current quaternion, Is a gradient function, a is a learning rate, and the determination mode is thatWherein a _r is a preset current learning rate, and g _r is a preset normalized acceleration parameter;

Calculating the rate of change of quaternion

Wherein the method comprises the steps ofIs the Hamilton product of the quaternions,Representing the real part as 0 and the imaginary part asIs a vector of (2);

after integrating the quaternion, continuously updating the value of the quaternion according to the time interval delta t between the two gesture updates:

Wherein q _new is the updated quaternion value, and then carrying out normalization processing on the updated quaternion value to obtain a processed attitude parameter;

parameter application stage: and performing collision analysis according to the processed gesture parameters to form a collision response instruction aiming at the electronic equipment, and feeding back the collision response instruction to the electronic equipment.

2. The method for processing data of a multifunctional sensor according to claim 1, wherein the normalizing the acceleration and the angular velocity, and the normalizing the updated quaternion, comprises: and taking the ratio of the corresponding acceleration, angular speed or quaternion parameter and the modulus thereof as the result of the normalization processing.

3. A multi-function sensor data processing method according to claim 1, wherein: the multifunctional sensor is a nine-axis sensor.

4. A multi-function sensor data processing method according to claim 1, wherein: the method of the parameter optimization stage may implement a body comprising: the computing device is connected with the electronic equipment in a near field communication manner; the method of the parameter application stage can be implemented by a main body comprising: the system comprises the electronic equipment, a computing device connected with the electronic equipment in a near field communication mode and a remote server.

5. The method for processing data of a multifunctional sensor according to claim 4, wherein: the near field communication mode comprises the following steps: bluetooth connection and wifi connection.

6. A multi-function sensor data processing method according to claim 1, wherein: the time interval deltat adopts a preset fixed time interval or adopts a dynamic time interval.

7. A method of multifunctional sensor data processing according to claim 6, said employing dynamic time intervals comprising: on the basis of a fixed time interval, multiplying the fixed time interval value by a calculation result of an adjustment factor to obtain a dynamic time interval; wherein the value of the adjustment factor is based onAnd (5) determining.

8. A multi-function sensor data processing system, characterized by: the system comprises a parameter acquisition unit, a parameter optimization unit and a parameter application unit, wherein the parameter acquisition unit is used for preprocessing acquired collision information of electronic equipment and transmitting the preprocessed collision information to the parameter optimization unit, and the parameter optimization unit is used for optimizing collision gesture parameters under quaternion representation and transmitting the optimized gesture parameters to the parameter application unit for subsequent collision analysis; wherein each unit is in particular also adapted to carry out the steps of the method according to any one of claims 1-7.

9. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of the method of any of claims 1-7.

10. A computer program product comprising a computer program, characterized in that the computer program, when being executed by a processor, implements the steps of the method according to any of claims 1-7.