CN108627153A - A kind of rigid motion tracing system and its working method based on inertial sensor - Google Patents

Abstract

The invention discloses a rigid body motion tracking system based on an inertial sensor and a working method thereof. The system includes an inertial sensor node and a data collection and processing terminal which can be attached to the rigid body. The inertial sensor node includes inertial sensors, computing modules and wireless communication modules. The data acquisition and processing terminal includes a wireless communication module and a computer system. The user arranges the inertial sensor node on the rigid body, and the node collects sensor data in real time and wirelessly transmits it to the data acquisition and processing terminal. When the motion is over, the computer system executes motion tracking and motion analysis algorithms, preprocesses the data, obtains the acceleration of the inertial sensor node, and calculates its velocity and motion trajectory, and then based on the attitude of the inertial sensor node and other points on the rigid body and sensor nodes The relative positional relationship of these points is restored to the trajectory of these points, and finally the trajectory of the selected point of the rigid body can be presented and the translation and rotation that occur during the movement can be analyzed.

Description

A rigid body motion tracking system based on inertial sensors and its working method

technical field

The invention relates to inertial sensor perception recognition, motion tracking and motion analysis, in particular to an inertial sensor-based rigid body motion tracking system and a working method thereof.

Background technique

Motion tracking refers to tracking the motion state of the target object, mainly including measuring, tracking, and recording the motion trajectory of the target object in three-dimensional space. Current motion tracking solutions are mainly based on computer graphics and image processing technologies, ultrasonic, radar, laser and other positioning technologies, and inertial sensors. The most mainstream solutions are generally based on computer graphics and image processing technology. Through multiple cameras deployed in the three-dimensional space, the motion status of the moving object (usually with marker nodes) is recorded in the form of images, and then the computer system processes these image data, and finally obtains the motion of the moving object at different times. Space coordinates (X,Y,Z). Solutions based on computer vision can achieve high accuracy, but they require large-scale deployment and are easily disturbed by visual factors. When moving objects are occluded, the accuracy will deviate greatly.

With the development of inertial sensors and wearable computing technology, motion tracking technology based on wearable inertial sensors and smart terminals has also attracted much attention. With the help of wearable inertial sensor nodes, users can obtain inertial sensor data in real time, and then processed by the motion tracking and motion analysis algorithms on the terminal device, which can track and analyze the motion process of the target rigid body. At present, some motion tracking work based on inertial sensors mainly focuses on restoring the motion posture or motion trajectory of the target object, and lacks the analysis of the transition process between different states during the motion process. How to balance motion tracking and motion analysis is a topic worth exploring. .

Contents of the invention

In order to solve the above technical problems, the present invention proposes a rigid body motion tracking system and working method based on inertial sensors. The system and method only need one inertial sensor node and a data acquisition and processing terminal, and are easy to deploy, easy to operate, and can relatively accurately capture The motion process of the rigid body attached to the inertial sensor node obtains the motion acceleration and velocity information of the rigid body at each moment, restores the motion trajectory of multiple selected points on the rigid body, and establishes a mathematical model for the motion process of the rigid body. Spin for reliable analysis.

For realizing the above-mentioned purpose of the invention, the technical scheme of the present invention is as follows:

A kind of rigid body motion tracking system based on inertial sensor of the present invention, this system comprises:

Inertial sensor node: Built-in a variety of inertial sensors, with certain computing, storage and communication capabilities, can be easily attached to the surface of rigid bodies.

Data acquisition and processing terminal: a computer system with computing, storage and communication capabilities, receiving various sensor data transmitted from inertial sensor nodes, storing and processing, and finally generating the trajectory of rigid body motion and the translation and rotation during rigid body motion Analyze the results and present them to the user.

A rigid body refers to an object whose shape and size are basically unchanged during the motion process and after the force is applied, and the relative position of each internal point is unchanged or the degree of change is negligible. The rigid body motion refers to the rotation and translation motion of a rigid body in three-dimensional space, which can be understood as the mapping from one set of three-dimensional coordinates to another set of three-dimensional coordinates in three-dimensional space. In addition, the motion tracking refers to tracking the motion state of the target object, which mainly includes measuring, tracking, and recording the motion trajectory of the target object in three-dimensional space.

The inertial sensor node is fixedly installed on the surface of the rigid body, and its interior includes:

Inertial sensor module: contains accelerometer, gyroscope and magnetometer for real-time measurement of corresponding inertial sensor data.

Calculation module: used to process the raw data of the inertial sensor, and obtain the attitude information of the inertial sensor node in real time through the built-in signal filtering algorithm, thereby estimating the components of the acceleration of gravity on the three axes of the inertial sensor's own reference coordinate system, which are compared with the original acceleration data Subtract to get the linear acceleration.

Wireless communication module: used for information interaction with terminal equipment, mainly used to receive instructions from data acquisition and processing terminals, and periodically send sensor data to terminal equipment.

Further, the inside of the data collection and processing terminal mainly includes:

Wireless communication module: used for information interaction with inertial sensor nodes.

Computer system: refers to a computer system with computing, storage and communication capabilities. It receives inertial sensor data as input through a wireless communication module, executes motion tracking and motion analysis algorithms, obtains the acceleration and velocity information of the rigid body at each moment, and calculates the rigid body motion. trajectory, and analyze the translation and rotation of the rigid body during motion.

Based on the above-mentioned tracking system, the present invention also introduces a working method, comprising the following steps:

1) The user arranges a single inertial sensor node on the rigid body, ready to trigger the rigid body motion.

2) The user establishes a wireless connection between the inertial sensor node and the data acquisition and processing terminal.

3) The user starts data collection at the data collection and processing terminal, and the inertial sensor node starts to collect various sensor data in real time and periodically sends them to the data collection and processing terminal, and the rigid body motion starts.

4) After the rigid body motion ends, the user terminates the data collection at the data collection and processing terminal.

5) The data acquisition and processing terminal executes the motion tracking and motion analysis algorithm through the computer system, obtains the acceleration and velocity information of the rigid body at each moment, generates the trajectory of the rigid body motion and the analysis results of the translation and rotation that occur during the motion of the rigid body, and reports to the user exhibit.

Further, wherein, the motion tracking and motion analysis algorithm in the step 5) includes the following steps:

Step 5.1 Data preprocessing, first smoothing the original sensor data, then calculating the three-axis linear acceleration data and the amplitude of the gyroscope data at each moment, and using the threshold judgment method to determine the start and end moments of the movement. Specifically, at the beginning of the movement, the amplitudes of the two types of sensor data will respectively exceed a threshold, and at the moment of the end of the movement, the amplitudes of the two types of sensor data will return below the corresponding threshold. By selecting a reasonable threshold through a large amount of experimental data, the data corresponding to the motion process can be accurately extracted. Then, the extracted linear acceleration data is converted to the reference coordinate system. The conversion of the reference coordinate system here refers to the use of the gravitational acceleration vector and the magnetometer data vector To construct a geographic coordinate system, specifically, firstly due to points north, so a vector pointing geographically east on the horizon It can be solved by the following formula:

And because of It is not strictly horizontal, so the vector pointing to the geographical north on the horizontal plane can be obtained according to the following formula

then These three orthogonal vectors can form a geographic coordinate system. After the three vectors are normalized, the direction cosine matrix DCM can be constructed, so that the transformation from the inertial sensor node's own coordinate system to the geographic coordinate system can be realized:

(X _g ,Y _g ,Z _g )'=DCM×(X,Y,Z)'

Where (X _g , Y _g , Z _g ) is the sensor data in the geographic coordinate system, and (X, Y, Z) is the original data based on the inertial sensor node's own coordinate system.

According to the needs of the specific application scenario, the geographic coordinate system can be transformed again using the same method to obtain the reference coordinate system corresponding to the application scenario. Finally, we can obtain the linear acceleration data relative to the fixed reference coordinate system during the motion.

Step 5.2 Execute motion tracking, update the motion state of the inertial sensor node at the current moment based on the motion state of the inertial sensor node at the previous moment and the inertial sensor data at the current moment, so as to obtain the relative position and attitude of the inertial sensor node in three-dimensional space at each moment information.

Step 5.3 Execute trajectory restoration, according to the relative position of the inertial sensor node in space at the current moment and its attitude, combined with the relative position relationship between the selected point on the rigid body and the inertial sensor node, estimate the position of these selected points on the rigid body, so that Restore the trajectory of multiple points on the rigid body.

Step 5.4 performs motion analysis, selects the motion trajectory of a certain plane in the rigid body, analyzes the motion process of the selected time slice length, establishes a mathematical model, and decomposes the continuous motion equivalently into a combination of translation and rotation, and the rotation of motion The situation can be solved by using the quaternion method, and finally output the translation and rotation that occurs during the motion process. The translation is represented by a vector in three-dimensional space, and the rotation can be represented by a quaternion (including the rotation axis and rotation angle). .

Further, the motion state of the inertial sensor node in the step 5.2 mainly includes the speed, relative position and posture of the inertial sensor node in space, and the specific motion state update steps include:

Step 5.2.1 Relative position information update: The motion tracking algorithm takes the position where the inertial sensor node starts to move as the origin of the reference coordinate system corresponding to the application scene, and first integrates the linear acceleration in the time domain to obtain the speed information at each moment. Then use error correction methods such as zero speed correction to compensate the speed. Then continue to integrate the compensated velocity in the time domain to solve the displacement occurring at adjacent moments, so as to update the relative position of the inertial sensor node in space (that is, the three-dimensional coordinates) at each sampling moment.

Step 5.2.2 Attitude information update: In the initial stage of the movement, a matrix containing the attitude information of the inertial sensor nodes is constructed from the instantaneous data of the magnetometer and the acceleration of gravity as the initial attitude of the inertial sensor nodes, and the local coordinate system of the initial stage is recorded to the geographic coordinate system In the subsequent stage, the initial attitude is updated by an integration of the gyroscope data in the time domain, and the attitude measured by the magnetometer and the acceleration of gravity is calibrated at some suitable time points, so that at each sampling moment Update the attitude information of the inertial sensor node.

The invention provides a rigid body motion tracking system based on an inertial sensor and a working method thereof. Compared with a motion tracking system based on computer vision, it is lighter in weight, easy to deploy, low in cost and not limited by visual factors. Compared with other motion tracking systems based on inertial sensors of the same type, this system can restore the motion trajectories of multiple points on the rigid body and give reliable motion analysis results.

The beneficial effects of the present invention are embodied as:

1. A new method for analyzing the motion process of a rigid body in three-dimensional space is proposed: in addition to analyzing the motion trajectory, the analysis of the translation and rotation during the motion process is introduced. By establishing a geometric model to decompose the motion of a rigid body in three-dimensional space into a combination of translation and rotation, the results of motion analysis are more intuitive.

2. It can realize the restoration of the motion trajectory of multiple selected points on the rigid body: by constructing a fixed reference coordinate system corresponding to the application scene, make full use of the real-time attitude information of the inertial sensor and the relationship between the selected point on the rigid body and the inertial sensor node The relative position relationship realizes the restoration of the motion trajectory of multiple selected points on the rigid body.

3. Easy to deploy: It is not necessary to deploy a large number of cameras and cables around the target in advance, and only needs the user to arrange the inertial sensor nodes on the rigid body to be tracked.

4. The cost is relatively low: there is no need to purchase high-cost cameras, switches, cables and other equipment, only a common data acquisition and processing terminal and a relatively low-cost inertial sensor node can realize the tracking and analysis of the rigid body motion process .

Description of drawings

Figure 1 is an architecture diagram of a rigid body motion tracking system based on inertial sensors.

Fig. 2 is a flow chart of the motion tracking and motion analysis algorithm of the inertial sensor-based rigid body motion tracking system.

FIG. 3 is a schematic diagram of a reference coordinate system involved in the working method of the inertial sensor-based rigid body motion tracking system.

Figure 4 is a diagram of the relative position relationship between the selected point on the rigid body and the inertial sensor node.

Figure 5 is a schematic diagram of rigid body motion.

Figure 6 is a schematic diagram of the modeling and analysis of the rigid body motion process.

Detailed ways

In order to facilitate the understanding of those skilled in the art, the present invention will be further described below in conjunction with the accompanying drawings.

Figure 1 is an architecture diagram of a rigid body motion tracking system based on inertial sensors. It is mainly composed of two parts: an inertial sensor node attached to a rigid body and a data acquisition and processing terminal. As shown in the figure, the two mainly exchange data through Bluetooth wireless communication.

The inertial sensor node is fixedly installed on the surface of the rigid body, and can be optionally pasted on the surface of the rigid body to ensure that the relative positional relationship between the inertial sensor module and the rigid body is fixed. The inertial sensor node contains inertial sensors, computing modules and communication modules. The inertial sensor includes accelerometer, gyroscope and magnetometer, which are used to measure the corresponding inertial sensor data in real time; the calculation module is used to process the raw sensor data, and obtain the attitude information of the inertial sensor node in real time through the built-in signal filtering algorithm, so as to estimate The linear acceleration can be obtained by subtracting the components of the acceleration of gravity on the three axes of the inertial sensor's own reference coordinate system from the original acceleration data; the wireless communication module is used for information interaction with the terminal device, mainly to receive instructions from the data acquisition and processing terminal , and periodically send sensor data to the terminal device.

The data acquisition and processing terminal is mainly composed of a wireless communication module and a computer system. The wireless communication module is used for information interaction with the inertial sensor nodes; the computer system refers to a computer system with computing, storage and communication capabilities, which receives inertial sensor data as input through the wireless communication module, executes motion tracking and motion analysis algorithms, and calculates rigid body The trajectory of the movement, and analyze the translation and rotation of the rigid body during the movement.

Figure 2 is a flow chart of the motion tracking and motion analysis algorithm. The motion tracking and motion analysis algorithm mainly includes data preprocessing, motion tracking, trajectory restoration, and motion analysis.

The data preprocessing part is mainly to scan the linear acceleration and gyroscope sensor data from front to back in time order, and accurately find out the moment when the movement starts and ends. First smooth the sensor data, some high frequency noise can be filtered out with simple mean smoothing. Then calculate the amplitudes AccM and GyroM of the three-axis linear acceleration and gyroscope data respectively, that is, solve the 2-norm of the two three-dimensional vectors respectively. Here, through the analysis of a large number of experimental data in the early stage, we have thresholds thresA and thresG for the amplitudes of the two three-axis sensor data respectively. We set a fixed-length sliding window to count the sensor data in the window greater than and less than the corresponding threshold. ratio. When the ratio of the amplitudes of the two sensor data is less than the set threshold is greater than ρ (this system is set to 80%), it means that the current window motion state is not active, and when the amplitude of one sensor data is greater than the set threshold When the ratio is greater than ρ, it means that the current window motion state is active. We find out the first window where the inactive state migrates to the active state, and then set the motion start time t _s , and then search for the window that transitions from the active state to the inactive state, and set the motion end time t _e , so far , we only need to intercept the sensor data between t _s and t _e , and then convert the intercepted linear acceleration data to the reference coordinate system as the input of the later part of the algorithm. The conversion of the reference coordinate system here refers to the use of the gravitational acceleration vector and the magnetometer data vector To construct a geographic coordinate system, specifically, firstly due to points north, so a vector pointing geographically east on the horizon It can be solved by the following formula:

And because of It is not strictly horizontal, so the vector pointing to the geographical north on the horizontal plane can be obtained according to the following formula

then These three orthogonal vectors can form a geographic coordinate system. After the three vectors are normalized, the direction cosine matrix DCM can be constructed, so that the transformation from the inertial sensor node's own coordinate system to the geographic coordinate system can be realized:

(X _g ,Y _g ,Z _g )'=DCM×(X,Y,Z)'

Where (X _g , Y _g , Z _g ) is the sensor data in the geographic coordinate system, and (X, Y, Z) is the original data based on the inertial sensor node's own coordinate system.

According to the needs of the specific application scenario, the geographic coordinate system can be transformed again using the same method to obtain the reference coordinate system corresponding to the application scenario. Finally, we can obtain the linear acceleration data relative to the fixed reference coordinate system during the motion.

The motion tracking part mainly uses the linear acceleration data and gyroscope data generated in the preprocessing stage based on the reference coordinate system corresponding to the application scene to update the motion trajectory of the inertial sensor node, and at the same time generate the attitude of the inertial sensor node at each sampling moment information. Specifically, in the trajectory update part, the algorithm uses the position where the inertial sensor node starts to move as the origin of the reference coordinate system corresponding to the application scene, and first integrates the linear acceleration in the time domain to obtain the speed information at each moment. There is noise in the data, and integration will lead to error accumulation. We use error correction methods such as zero-speed correction to correct the speed. Then continue to integrate the compensated velocity in the time domain to solve the displacement occurring at adjacent moments, so as to update the relative position of the inertial sensor node in space (that is, the three-dimensional coordinates). The zero-speed correction is mainly based on the assumption that the speed of the inertial sensor node is zero at the end of the motion, so that the speed can be compensated when the linear acceleration is integrated for the first time to obtain the speed at each moment, so that Cumulative error can be reduced. In the attitude update part, in the initial stage of the movement, a matrix containing the attitude information of the inertial sensor node is constructed from the instantaneous data of the magnetometer and the acceleration of gravity as the initial attitude of the inertial sensor node, and the rotation matrix from the local coordinate system of the node to the geographic coordinate system in the initial stage is recorded. In the subsequent stage, an integration of gyroscope data in the time domain is used to update the initial attitude, and at some suitable time points, the attitude measured by the magnetometer and the acceleration of gravity is used for calibration to eliminate the cumulative error caused by the gyroscope. In this way, the attitude information of the inertial sensor node is updated. The appropriate time point mentioned here is mainly determined by these factors:

1) Real-time angular velocity information. When the value of the angular velocity is below a certain threshold, we believe that the attitude information measured by the magnetometer and the acceleration of gravity at this moment is relatively reliable and can be used for calibration.

2) The attitude change calculated by the gyroscope is similar to the attitude change obtained by using the magnetometer and the acceleration of gravity. When the attitude changes measured by the two methods are similar, we believe that the attitude change measured by the magnetometer and the acceleration of gravity Attitude information is also relatively reliable and can be used for calibration.

The trajectory restoration part uses the inertial sensor node trajectory information and attitude information generated by the motion tracking part, combined with the pre-known relative position relationship between the selected point on the rigid body and the inertial sensor node in the node's own reference coordinate system, to obtain the fixed reference corresponding to the application scene. The relative position relationship between the selected point on the rigid body and the inertial sensor node in the coordinate system, and then estimate and restore the trajectory of the selected point on the rigid body. For the relative position relationship between the selected point on the rigid body and the inertial sensor node, please refer to the description of FIG. 4 below for details.

The motion analysis part takes the trajectory of the selected point on the rigid body output by the trajectory restoration part as input, and uses the mathematical model of the rigid body motion process we have established for analysis. Specifically, we select three points on a plane on the rigid body that are not on the same straight line, and their motion trajectories are known, which means that the motion process of this plane is known, so that using the rigid body motion we established The process mathematical model can equivalently simplify the motion process of a rigid body from one moment to another into a combination of a translation and a rotation. Finally, we give the analysis result of the motion is the vector of the translation during the motion Axis of rotation for the rotation process And the angle θ _equal rotated counterclockwise around the axis of rotation. For the mathematical model of the rigid body motion process, please refer to the description of Fig. 5 and Fig. 6 below.

FIG. 3 is a schematic diagram of several reference coordinate systems involved in the working method of the inertial sensor-based rigid body motion tracking system. In the left half of the figure, the coordinate system with point N as the origin of the coordinate system and X _N Y _N Z _N as the coordinate axis is the coordinate system of the inertial sensor node itself. The original sensor data generated by the inertial sensor node is based on this coordinate system. The coordinate system with point O as the origin and XYZ as the coordinate axis is the reference coordinate system corresponding to the above-mentioned application scenario. Here is just one of the possible situations as an example. The reference coordinate system corresponding to the specific application scenario There can be different definitions. The right half of the figure shows the geographic coordinate system. We can see that the X-axis points to the geographic east, the Y-axis points to the geographic north, and the Z-axis points to the opposite direction of gravity. In the actual algorithm process, the process of converting the reference coordinate system is to convert the linear acceleration data from the coordinate system of the inertial sensor node itself to the geographic coordinate system, and then convert from the geographic coordinate system to the reference coordinate system corresponding to the application scene.

Figure 4 is a schematic diagram of the relative positional relationship between some selected points on the rigid body and the inertial sensor nodes. Point N in the figure represents the inertial sensor node. Four points O, A, B, and C that are not on the same plane are selected on the rigid body as an example. The coordinate system given in the figure is the coordinate system of the inertial sensor node itself. For simplicity, point N is taken as the origin of the coordinate system here. In this coordinate system, since the shape and size of the rigid body itself and the deployment position of the inertial sensor are known, the coordinates of points O, A, B, C, and N can be measured and obtained, that is, the points marked in the figure four vectors are all measurable. Therefore, we can estimate the relative positions of the selected points O, A, B, and C on the rigid body in the three-dimensional space through the relative position of the inertial sensor node N in the three-dimensional space. Specifically, the relative position relationship between the selected point and the inertial sensor node (that is, the four vectors ) is converted from the node's own coordinate system to the geographic coordinate system, and then converted to the reference coordinate system corresponding to the application scenario shown in FIG. 3 . The trajectory of the inertial sensor node N (that is, the relative position of the inertial sensor node in the three-dimensional space at each moment) that we solve in the motion tracking part is based on the reference coordinate system corresponding to the application scene, so that the application scene can be used. Four vectors in the reference coordinate system Solve the coordinates of points O, A, B, and C in the reference coordinate system corresponding to the application scene, and finally realize the restoration of the trajectory of the selected point on the rigid body.

Figure 5 is an example of the motion process of a rigid body. Without loss of generality, four points on the rigid body that are not on the same plane are used to represent a rigid body. We can see that the starting positions of the four vertices of the rigid body in the figure are O, A, B, and C. After the movement, the positions of the corresponding vertices are transformed to O', A', B', and C'. This is the movement of the rigid body before and after The generalized representation of state, as mentioned above, is equivalent to the mapping from one set of three-dimensional coordinates (O, A, B, C) to another set of three-dimensional coordinates (O', A', B', C').

Figure 6 is a schematic diagram of the analysis and modeling of the rigid body motion process. The motion process of the rigid body is relatively complicated, but we can perform equivalent transformations on the motion process to achieve The purpose of simplifying the exercise process. We decompose the motion process of a rigid body between adjacent time slices into translation and rotation, which requires a geometric model to describe. Figure 6(a) is the first step in decomposing the motion process. We select a point (O) on the rigid body, which points to the vector corresponding to the point (O′) after the motion That is, the vector for the translational motion of the rigid body. As shown in Figure 6(b), after the translation, the position of the rigid body is transformed into O ₁ , A ₁ , B ₁ , and C ₁ , where O ₁ coincides with O′. In the second step, we need to convert the side O ₁ A ₁ of the rigid body The axis around the point O' is rotated to coincide with O'A', the axis of rotation The calculation of is as follows:

And the rotation angle θ ₀ can be determined by the vector with vector Click to get. As shown in Figure 6(c), after the first rotation, the position of the rigid body is transformed into O ₂ , A ₂ , B ₂ , and C ₂ , and the side O ₂ A ₂ coincides with the side O'A'. At this point, we choose the third point B ₂ which is not on the same straight line as O ₂ and A ₂ , in this step, the rigid body needs to go around the axis (converted to a unit vector ) is rotated until the plane O ₂ A ₂ B ₂ is coplanar with the plane O'A'B', and the rotation angle θ ₁ can be obtained by directly calculating the angle between the two planes. Specifically, here is by first solving the normal vectors of the two planes shown in Figure 6(c) and Then solve the angle θ between the two normal vectors to obtain the angle θ ₁ between the two planes. Finally, figure (d) shows the rigid body positions O ₃ , A ₃ , B ₃ , C ₃ after the second rotation, ie positions O', A', B', C'. The rigid body motion decomposition process of one translation and two rotations shown in Figure 4 is relatively intuitive. In the actual calculation process, the transformation process of translation can be realized by adding and subtracting the translation process vector, and the transformation process of rotation can be realized by using the quaternion method to solve. In addition, we can use the quaternion method to compound the two rotations, so that the motion process of the rigid body is simplified to a translation and a rotation around a fixed axis. We take the quaternion and Representing the two rotations in the above process, is the quaternion representation of the point coordinates before rotation, is the quaternion representation of the point coordinates after the first rotation, The quaternion representation of the point coordinates after the second rotation, the specific rotation composite process is as follows:

It is the rotation quaternion after compounding, from which the rotation axis can be obtained and rotation angle θ _equal . The final motion analysis is to analyze the motion process between any two selected moments, and give the vector that simplifies the translation that occurs during the motion process rotating axis of rotation And the angle θ _equal rotated counterclockwise.

The present invention has many application ways, and the above-mentioned embodiment is only a preferred implementation of the present invention, so the present invention is not limited to the above-mentioned implementation. Without departing from the principles of the present invention, those skilled in the art may design other implementations under the inspiration of the present invention, and these should also be regarded as the protection scope of the present invention.

Claims (8)

1. a kind of rigid motion tracing system based on inertial sensor, it is characterised in that the system includes：

Inertial sensor node：Built-in a variety of inertial sensors, the sensor have calculating, storage and communication capacity, period Sensing data is sent to property to data acquisition process terminal, and the inertial sensor node is attached to rigid body surface；

Data acquisition process terminal：Have the computer system of calculating, storage and communication capacity, receives from the inertial sensor The various sensing datas that node is sent, store and handle, and ultimately generate track and the rigid motion of rigid motion The analysis result of the translation and rotation that occur in journey, and shown to user.

2. the rigid motion tracing system according to claim 1 based on inertial sensor, which is characterized in that described is rigid Body motion tracking refers to the motion state of target rigid body into line trace, including measurement, tracking, record target rigid body are in three-dimensional space Between in movement locus.

3. the rigid motion tracing system according to claim 2 based on inertial sensor, which is characterized in that the inertia Sensor node is fixedly installed in rigid body surface, inside include：

Inertial sensor module：The inertial sensor module can select to be pasted on rigid body surface, it is ensured that inertial sensor mould The relative position relation of block and rigid body is fixed, and includes three axis accelerometer, gyroscope and magnetometer, for measuring in real time accordingly Inertial sensor data；

Computing module：For handling inertial sensor initial data, inertia is obtained by built-in signal filtering algorithm in real time and is passed The posture information of sensor node, to estimate point of the acceleration of gravity on three axis of inertial sensor itself reference frame Amount and raw acceleration data, which are subtracted each other, can obtain linear acceleration；

Sensor wireless communication module：For carrying out information exchange with terminal device.

4. the rigid motion tracing system according to claim 1 based on inertial sensor, which is characterized in that the data Collecting and processing unit includes：

Processing terminal wireless communication module：For carrying out information exchange with inertial sensor node；

Computer system：Refer to the computer for having calculating, storage and communication capacity, is received by processing terminal wireless communication module Inertial sensor data executes motion tracking and motion analysis algorithms as input, know rigid body each moment acceleration and Velocity information, calculates the track of rigid motion, and analyzes the case where rigid body is translated and rotated during the motion.

5. a kind of rigid motion tracing system working method based on inertial sensor, which is characterized in that include the following steps：

Single inertial sensor inserting knot on rigid body, is prepared triggering rigid motion by user；

User is that inertial sensor node establishes wireless connection with data acquisition process terminal；

User acquires in data acquisition process terminal turn-on data, and inertial sensor node starts to acquire various kinds of sensors number in real time According to and be sent periodically to data acquisition process terminal, rigid motion starts；

Rigid motion terminates, and user terminates data acquisition in data acquisition process terminal；

Data acquisition process terminal executes motion tracking and motion analysis algorithms by computer system, knows rigid body each moment Acceleration and velocity information, the translation and rotation for generating the track of rigid motion, and occurring during modeling analysis rigid motion The case where turning, most rear line show.

6. the rigid motion tracing system working method according to claim 5 based on inertial sensor, feature

It is, the motion tracking in the step 5) and motion analysis algorithms include the following steps：

Step 5.1 data prediction first carries out smooth operation to original sensor data, then calculates three axis at each moment Property acceleration information and gyro data amplitude, using threshold decision method determine movement starting and terminate at the time of, and The data between movement starting and end time are extracted, then carry out the conversion of reference frame to this partial data, to To in motion process relative to the linear acceleration data under a fixed reference frame；

Step 5.2 executes motion tracking, the inertia of motion state and current time based on inertial sensor node last moment Sensing data updates the motion state of current time inertial sensor node, to obtain inertial sensor node each moment Relative position in three dimensions and posture information；

Step 5.3 perform track restores, according to the relative position and its appearance of inertial sensor node current time in space State, in conjunction with the relative position relation of Chosen Point and inertial sensor node on rigid body, to the position of these Chosen Points on rigid body Estimated, to restore the movement locus of multiple points on rigid body；

Step 5.4 executes motion analysis, chooses the movement locus of some plane in rigid body, the movement to seclected time leaf length Process is analyzed, founding mathematical models, will continuously move the equivalent combination for being decomposed into translation and rotation, final output movement The case where translation and rotation for occurring in the process.

7. the rigid motion tracing system working method according to claim 6 based on inertial sensor, which is characterized in that The conversion of reference frame refers to the sensing data based on accelerometer and magnetometer measures in the step 5.1, structure ground Reference frame is managed, geographic coordinate system is once converted further according to practical application scene, the corresponding ginseng of the scene that is applied Coordinate system is examined, it finally will be under linear acceleration data conversion to the corresponding reference frame of application scenarios.

8. the rigid motion tracing system working method according to claim 6 based on inertial sensor, which is characterized in that

The motion state of inertial sensor node mainly includes the speed of inertial sensor node in space in the step 5.2 Degree, relative position and its posture, motion state update step include：

Step 5.2.1 relative position informations update：Motion tracking algorithms using the position of inertial sensor node setting in motion as The origin of the corresponding reference frame of application scenarios, in the time domain first once integrates linear acceleration, when obtaining each The velocity information at quarter recycles zero-speed correction error correction means, is compensated to speed；Then to the speed after compensation when Continue to do primary integral on domain, solve the displacement that adjacent moment occurs, realizes to inertial sensor node in space opposite Position is updated；

Step 5.2.2 posture informations update：The movement starting stage is built by the transient data of magnetometer and acceleration of gravity Initial attitude of the matrix of inertial sensor node posture information as inertial sensor node, and record starting stage node Local coordinate system to geographic coordinate system spin matrix, follow-up phase using gyroscope data in the time domain it is primary integrate pair Initial attitude is updated, and is calibrated using the posture that magnetometer and acceleration of gravity measure, and reduces gyroscope as far as possible The influence of accumulated error, to be updated to the posture information of inertial sensor node.