CN105632101B - A kind of anti-tumble method for early warning of human body and system - Google Patents

A kind of anti-tumble method for early warning of human body and system Download PDF

Info

Publication number
CN105632101B
CN105632101B CN201511028550.5A CN201511028550A CN105632101B CN 105632101 B CN105632101 B CN 105632101B CN 201511028550 A CN201511028550 A CN 201511028550A CN 105632101 B CN105632101 B CN 105632101B
Authority
CN
China
Prior art keywords
tumble
human
early warning
monitored target
axis
Prior art date
Application number
CN201511028550.5A
Other languages
Chinese (zh)
Other versions
CN105632101A (en
Inventor
赵国如
全永奇
李慧奇
宁运琨
谢高生
王磊
Original Assignee
深圳先进技术研究院
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 深圳先进技术研究院 filed Critical 深圳先进技术研究院
Priority to CN201511028550.5A priority Critical patent/CN105632101B/en
Publication of CN105632101A publication Critical patent/CN105632101A/en
Application granted granted Critical
Publication of CN105632101B publication Critical patent/CN105632101B/en

Links

Classifications

    • GPHYSICS
    • G08SIGNALLING
    • G08BSIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
    • G08B21/00Alarms responsive to a single specified undesired or abnormal operating condition and not elsewhere provided for
    • G08B21/02Alarms for ensuring the safety of persons
    • G08B21/04Alarms for ensuring the safety of persons responsive to non-activity, e.g. of elderly persons
    • G08B21/0438Sensor means for detecting
    • G08B21/0446Sensor means for detecting worn on the body to detect changes of posture, e.g. a fall, inclination, acceleration, gait

Abstract

The present invention provides a kind of anti-tumble method for early warning of human body and system, the anti-tumble method for early warning of human body includes:The 3-axis acceleration information, three axis angular rate information and three axis geomagnetic field informations of the monitored target of multiple sampled points are acquired in real time;According to 3-axis acceleration information calculate monitored target each sampled point resultant acceleration;Eulerian angles sequence of differences is generated according to the 3-axis acceleration information, three axis angular rate information and three axis geomagnetic field informations of each sampled point;Whether it is doubtful tumble state according to the resultant acceleration of each sampled point and the motion state of preset resultant acceleration threshold decision monitored target;If so, the resultant acceleration mean value and preset resultant acceleration mean value threshold value that are obtained according to the resultant acceleration by all sampled points, judge whether monitored target has tumble to be inclined in conjunction with the Eulerian angles sequence of differences and preset Eulerian angles difference threshold;Warning message is generated when monitored target has tumble to be inclined to, to alarm before monitored target is fallen.

Description

A kind of anti-tumble method for early warning of human body and system
Technical field
The present invention relates to wearable technology field more particularly to a kind of anti-tumble method for early warning of human body and systems.
Background technology
Tumble refer to burst, involuntary, unintentional Body Position Change --- fall on the ground or lower plane on.According to system Meter, has about 424000 death directly related with tumble behavior, tumble is to cause unexpected injury in the world every year in the world The dead second largest reason (being only second to traffic accident).Studies have shown that in the elderly of over-65s, have every year fall 1 time or The ratio for experience of repeatedly falling is up to 1/3, wherein 20%~30% old man can cause scratch, hip bone in tumble event Folding, head injury etc., and increase with the increase at age.In the U.S., the annual medical total cost for tumble is more than 20,000,000,000 Dollar.There is about 1.3 hundred million the elderly in China at present, and there are about 20,000,000 the elderlys at least to occur to amount to 25,000,000 tumble things every year Therefore direct medical cost is more than 5,000,000,000 RMB.It threatens the elderly's life security it can be seen that falling and becoming and increases social An important factor for burden.
Judge to fall mainly there are following 3 kinds of methods at present:1) analysis based on video image, the real time kinematics of object are by taking the photograph As head monitoring, deficiency is the personal secrets it cannot be guaranteed that user, and usage scenario is limited, expensive;2) audio frequency is based on to believe Number analysis, cause the frequency-portions of vibration to judge tumble event by analysis impact, but more complicated, the fund of such equipment installation Input is also bigger;3) it is detected based on wearable device.Consider the right of privacy of user and reduces interference user's life as far as possible Mode, wearable device are most suitable.Wearable fall detection/monitoring and warning device are typically based on inertia Sensing unit.The sensing module of current common " fall alarm " generally uses single three axis accelerometer, measurement accuracy It is limited with data modality or used thresholding algorithm is not excellent enough, maximum defect is can only to realize tumble alarm, that is, is occurred Alarm after tumble is sought help, and cannot be accomplished anticipation and be dropped to risk and remind user.
Chinese invention patent 201210472813.1 discloses a kind of fall detection automatic alarm waistband, including waistband and inspection Device is surveyed, detection device uses acceleration detection module, by 3-axis acceleration information (the used module for detecting user For Freescale MMA7455 3-axis acceleration sensors, processor is microcontroller) to judge whether user is in tumble state, Customer position information can be obtained automatically by GPS module and is sought advice in short message to guardian's mobile phone by gsm module transmission, be convenient for Carry out treatment in time after human body tumble.
Chinese invention patent 201410134724.5 discloses a kind of fall monitoring alarm system, including earphone portion and movement Communication terminal portion, earphone portion can continue to obtain the information such as acceleration information, the body angle data of human body, mobile communication terminal Portion receives earphone portion data, judges whether to meet tumble condition, is judged as that tumble state then can first be sent to server end and falls Alarm and location information.This mode communication terminal design comparison is flexible, but earphone end needs to be worn on ear portion for a long time Position easily causes human body discomfort, and real-time data collection power consumption is high, and earphone is small, and battery capacity is low, is unsuitable for wearing for a long time It uses.
Utility model patent 201520186743.2 discloses a kind of tumble alarm waistcoat, including emergency alarm switchs, adds Fast sensor, signal projector and several parts of locating module, fall when detecting, can be sent and be succoured by signal projector Information asks relief to signal receiver.The vest is easy to wear, has the function of to communicate and position etc., but be a lack of necessary Preventing mechanism and warning function.
Above-mentioned technology has following defect:1) early warning before falling is not implemented.The sensing of current common " fall alarm " Module generally uses single three axis accelerometer, and measurement accuracy and data modality are limited or used thresholding algorithm is inadequate It is excellent, therefore be only capable of the warning function after realization human body is fallen and can not achieve the early warning before falling.Although special in Chinese invention Profit 201210561342.1 has been related to " early warning ", but it is achieved in that based on video image analysis, as previously described, no Conducive to guarantee privacy of user, and only mainly for house monitoring data, cause its application scenarios significantly limited, to be less likely And then realize instant anti-tumble intervening measure and real-time protection.
2) have a single function, interface it is simple:Published precaution device only realizes simple early warning function mostly, and early warning Thing of the device as the personal wearing of user, actually a critically important data acquisition platform.If the certain interface of configuration And hardware device, you can be connected, can also be interconnected with the mobile device of oneself by the guardian of network and user, or even also Social function may be implemented.
3) it is not convenient enough to wear use:A part of tumble alarm device uses accelerometer module, attitude transducer mould The multi-module combinations such as block and pressure sensor module are more although being theoretically conducive to improve the accuracy fallen and judged A sensing module necessarily can also increase the volume and power consumption of device, the convenience for affecting wearing and using.In addition, multiple sensings More stringent requirements are proposed in terms of precision especially speed to microprocessor and thresholding algorithm for the Data Fusion of module.Separately Outside, theoretically for sensor assembly closer to human body, the sensing data measured is more accurate, in practical applications, multimode Can combination really effectively improve the accuracy fallen and judged and need further to verify.
4) tumble accuracy judges undesirable:Current tumble judgment means mainly only rely on acceleration to judge mostly The generation of tumble event, but the movement of human body schedule is extremely complex, and it is typically the accident behavior (1~2 occurred in a flash to fall Second), predict not high (the multisensor block combiner side of reliability fallen by less or single human motion posture information The analysis of formula is shown in upper one, and details are not described herein again), when human body has tumble to be inclined to and fails to judge in time, just do not have pre- Alert effect is necessarily made troubles to wearer when human body does not fall tendency when making erroneous judgement.
5) safeguard is not portable enough, wearing is inconvenient:Protective garment energy disclosed in utility model patent 201220626477.7 The protection of large area is enough provided, but handicapped people's wearings such as wearing is particularly troublesome, is unfavorable for the elderly, and without corresponding Fall detection algorithm, only simple physical protection;Chinese invention patent 201410141280.8,201510038329.1 etc. Disclosed long-range tumble monitoring intelligent crutch, anti-tumble seat etc., can only partly realize tumble safeguard function, usage scenario by Limit, can not spread to entire the elderly group on a large scale.
Invention content
In order to solve the above technical problems, the present invention provides a kind of anti-tumble method for early warning of human body and systems.
One aspect of the present invention provides a kind of anti-tumble method for early warning of human body, and the anti-tumble method for early warning of human body includes:
With acquiring the 3-axis acceleration information, three axis angular rate information and three axis of the monitored target of multiple sampled points in real time Magnetic Field;
According to the 3-axis acceleration information calculate monitored target each sampled point resultant acceleration;
Europe is generated according to the 3-axis acceleration information, three axis angular rate information and three axis geomagnetic field informations of each sampled point Draw angular difference value sequence;
Motion state according to the resultant acceleration of each sampled point and preset resultant acceleration threshold decision monitored target is No is doubtful tumble state;
If so, the resultant acceleration mean value and preset resultant acceleration that are obtained according to the resultant acceleration by all sampled points are equal It is worth threshold value, judges whether monitored target has tumble to incline in conjunction with the Eulerian angles sequence of differences and preset Eulerian angles difference threshold To;
Warning message is generated when monitored target has tumble to be inclined to, to alarm before monitored target is fallen.
In one embodiment, the anti-tumble method for early warning of the human body further includes:Air bag executing agency is according to the alarm signal Breath punctures gas cylinder therein to be inflated to air bag therein, to provide buffering when human body is fallen.
In one embodiment, the anti-tumble method for early warning of the human body further includes:Start to count after generating the warning message When, judge whether the monitored target within the preset alarm time cancels the warning message, if it is not, then full described in meter The warning message is sent to guardian after the preset alarm time.
In one embodiment, the anti-tumble method for early warning of the human body further includes:To described the three of collected each sampled point Axis acceleration information, three axis angular rate information and three axis geomagnetic field informations carry out Kalman filtering.
In one embodiment, monitored target is calculated according to the 3-axis acceleration information in the conjunction of each sampled point to accelerate Degree, including:
According to the 3-axis acceleration information, conjunction of the monitored target in each sampled point is calculated using acceleration modulus algorithm Acceleration ASVM
Wherein, Ax、Ay、AzThe respectively acceleration of three axial direction of the monitored target at same sampled point.
In one embodiment, according to the 3-axis acceleration information of each sampled point, three axis angular rate information and three axis Magnetic Field generates Eulerian angles sequence of differences:
The 3-axis acceleration information, three axis angular rate information and three axis geomagnetic field informations are fused into Eulerian angles, it is described Eulerian angles include course angle, pitch angle and roll angle;
Difference is carried out respectively to the course angle, pitch angle and roll angle, is obtained comprising the course angle difference, pitch angle The sequence of difference and roll angle difference:
Wherein, yaw is course angle, and pitch is pitch angle, and roll is roll angle, and k is current sample time, and k-1 is upper one Sampling instant.
In one embodiment, according to the resultant acceleration of each sampled point and preset resultant acceleration threshold decision monitored target Motion state whether be doubtful tumble state, including:
When the resultant acceleration of an at least sampled point is more than the resultant acceleration threshold value, the fortune of the monitored target is judged Dynamic state is doubtful tumble state.
In one embodiment, the resultant acceleration mean value and preset close obtained according to the resultant acceleration by all sampled points adds Speed mean value threshold value judges whether monitored target has in conjunction with the Eulerian angles sequence of differences and preset Eulerian angles difference threshold It falls and is inclined to, including:
According to the resultant acceleration of each sampled point, the conjunction for calculating monitored target accelerates mean value;
When the resultant acceleration mean value is more than preset resultant acceleration mean value threshold value, and at least one Euler's angular difference When the absolute value of value is more than the Eulerian angles difference threshold, judges that the monitored target has to fall and be inclined to.
Another aspect of the present invention additionally provides a kind of anti-tumble early warning system of human body, the anti-tumble early warning system packet of human body It includes:
Sensor unit, the 3-axis acceleration information of the monitored target for acquiring multiple sampled points in real time, three shaft angles Velocity information and three axis geomagnetic field informations;
Resultant acceleration computing unit, for calculating monitored target in each sampled point according to the 3-axis acceleration information Resultant acceleration;
Eulerian angles sequence of differences generation unit, for the 3-axis acceleration information, the three shaft angles speed according to each sampled point It spends information and three axis geomagnetic field informations generates Eulerian angles sequence of differences;
First judging unit, for monitored according to the resultant acceleration of each sampled point and preset resultant acceleration threshold decision Whether the motion state of object is doubtful tumble state;
Second judgment unit, for judging the motion state of monitored target for doubtful tumble in first judging unit When state, the resultant acceleration mean value and preset resultant acceleration mean value threshold value obtained according to the resultant acceleration by all sampled points, Judge whether monitored target has tumble to be inclined in conjunction with the Eulerian angles sequence of differences and preset Eulerian angles difference threshold;
Warning message generation unit, for generating warning message when monitored target has tumble to be inclined to, with monitored Object is alarmed before falling.
In one embodiment, the anti-tumble early warning system of the human body further includes:
Timing unit, for starting timing after generating the warning message;
Third judging unit, for judging whether the monitored target within the preset alarm time cancels the alarm Information.
In one embodiment, the anti-tumble early warning system of the human body further includes:Positioning unit, it is described monitored for acquiring The geographic position data of object.
In one embodiment, the anti-tumble system of the human body further includes:Communication unit, by the full preset alarm based on The warning message is sent to guardian after time, the warning message includes the geographic position data.
In one embodiment, the anti-tumble early warning system of the human body further includes:
Kalman filtering unit, for the 3-axis acceleration information of collected each sampled point, three axis angular rates Information and three axis geomagnetic field informations carry out Kalman filtering.
In one embodiment, the resultant acceleration computing unit is specifically used for:
According to the 3-axis acceleration information, conjunction of the monitored target in each sampled point is calculated using acceleration modulus algorithm Acceleration ASVM
Wherein, Ax、Ay、AzThe respectively acceleration of three axial direction of the monitored target at same sampled point.
In one embodiment, the Eulerian angles sequence of differences generation unit includes:
Digital moving processor, for believing the 3-axis acceleration information, three axis angular rate information and three axis earth's magnetic fields Breath is fused into Eulerian angles, and the Eulerian angles include course angle, pitch angle and roll angle;
Difference calculating module is obtained for carrying out difference respectively to the course angle, pitch angle and roll angle comprising described The sequence of course angle difference, pitch angle difference and roll angle difference:
Wherein, yaw is course angle, and pitch is pitch angle, and roll is roll angle, and k is current sample time, and k-1 is upper one Sampling instant.
In one embodiment, first judging unit is specifically used for:
When the resultant acceleration of an at least sampled point is more than the resultant acceleration threshold value, the fortune of the monitored target is judged Dynamic state is doubtful tumble state.
In one embodiment, the second judgment unit includes:
Resultant acceleration mean value computation module calculates monitored target for the resultant acceleration according to each sampled point It closes and accelerates mean value;
Judging submodule, for being more than preset resultant acceleration mean value threshold value when the resultant acceleration mean value, and at least When the absolute value of one Eulerian angles difference is more than the Eulerian angles difference threshold, judges that the monitored target has and fall Tendency.
In one embodiment, the anti-tumble early warning system of the human body further includes an air bag executing agency, and the air bag executes Mechanism is connect with the warning message generation unit by steering engine interface, for puncturing gas cylinder therein according to the warning message To be inflated to air bag therein, to provide buffering when human body is fallen.
In one embodiment, the anti-tumble early warning system of the human body further includes:Alarm unit, including voice alarm module, Buzzer, vibrator and LED light, for sending out sound-light alarm according to the warning message.
In one embodiment, the anti-tumble early warning system of the human body further includes:Storage unit, for storing the sensor The report that the data of unit acquisition, the intermediate data generated in described information processing procedure and the warning message generation unit generate Alert information.
In one embodiment, the anti-tumble early warning system of the human body further includes:Power supply unit, for falling to the human body is anti- The power supply of falling early warning system.
Using the present invention, early warning can be carried out before human body is fallen, remind monitored target and surrounding people, be monitored Object has fall risk, should be noted strick precaution.
Description of the drawings
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below There is attached drawing needed in technology description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this Some embodiments of invention for those of ordinary skill in the art without creative efforts, can be with Obtain other attached drawings according to these attached drawings.
Fig. 1 is the flow diagram of the anti-tumble method for early warning of human body of the embodiment of the present invention;
Fig. 2 is the resultant acceleration change curve before and after monitored target of the embodiment of the present invention is fallen;
Fig. 3 sits down for monitored target and acts front and back resultant acceleration change curve;
Fig. 4, which hurries up/jogs for monitored target, acts the curve of front and back resultant acceleration variation;
Fig. 5 is the structural schematic diagram of the anti-tumble early warning system of human body of the embodiment of the present invention;
Fig. 6 is the frame structure and connection diagram of the anti-tumble early warning system of human body of the embodiment of the present invention;
Fig. 7 is alarm unit hardware composition frame chart of the embodiment of the present invention;
Fig. 8 is the anti-tumble early warning system workflow schematic diagram of human body of the embodiment of the present invention;
Fig. 9 is that the monitoring pattern of the anti-tumble early warning system of human body of the embodiment of the present invention converts schematic diagram;
Figure 10 is gesture recognition algorithms flow diagram of the embodiment of the present invention;
Figure 11 is tumble pre-warning time result schematic diagram of the embodiment of the present invention.
Specific implementation mode
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete Site preparation describes, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every other Embodiment shall fall within the protection scope of the present invention.
Fig. 1 is the flow diagram of the anti-tumble method for early warning of human body of the embodiment of the present invention.As shown in Figure 1, the method for early warning It mainly includes the following steps that:
Step S1,3-axis acceleration information, the three axis angular rate information of the monitored target of multiple sampled points are acquired in real time With three axis geomagnetic field informations.
Step S2, according to the 3-axis acceleration information calculate monitored target each sampled point resultant acceleration.
Step S3, believed according to the 3-axis acceleration information of each sampled point, three axis angular rate information and three axis earth's magnetic fields Breath generates Eulerian angles sequence of differences.
Step S4, according to the movement of the resultant acceleration of each sampled point and preset resultant acceleration threshold decision monitored target Whether state is doubtful tumble state.
Step S5, when the motion state of monitored target is doubtful tumble state, added according to the conjunction by all sampled points The resultant acceleration mean value and preset resultant acceleration mean value threshold value that speed obtains, in conjunction with the Eulerian angles sequence of differences and preset Eulerian angles difference threshold judges whether monitored target has tumble to be inclined to.
Step S6, warning message is generated when monitored target has tumble to be inclined to, to be carried out before monitored target is fallen Alarm.
Using the present invention, early warning can be carried out before human body is fallen, remind monitored target and surrounding people, be monitored Object has fall risk, should be noted strick precaution.
If judging the motion state of monitored target according to step S4 for non-doubtful tumble state, i.e. return to step S1 after The exercise data of continuous acquisition monitored target.
In one embodiment, after generating warning message in step s 6, i.e., when monitored target has tumble to be inclined to, air bag Executing agency punctures gas cylinder therein to be inflated to air bag therein, to be provided when human body is fallen according to the warning message Buffering.
In one embodiment, the anti-tumble method for early warning of human body provided by the invention further includes a step S7, is for judgement It is no that above-mentioned warning message is issued into guardian.When it is implemented, usually presetting a time of fire alarming and storing, so as in step Start timing after generating warning message in S6, if monitored target is thought without alarm, within the above-mentioned preset alarm time Cancel the warning message, it is full above-mentioned pre- in meter if monitored target does not cancel the warning message in predetermined time of fire alarming If warning message is sent to guardian (step S8) after time of fire alarming.The above-mentioned preset alarm time can be 20 seconds or 1 Minute.
In 3-axis acceleration information, three axis angular rate information and the three axis earth's magnetic fields for acquiring monitored target using sensor When information, due to the shake of sensor, certain noise is will produce, therefore to believe the collected 3-axis acceleration of sensor Breath, three axis angular rate information and three axis geomagnetic field informations are filtered and carry out data processing again.Kalman filtering is suitable for letter in real time Number processing, can update the estimation to state variable using the estimated value and the observation of current moment of previous moment, find out current The estimated value at moment can be used in handling in real time.
When the embodiment of the present invention is filtered sensor collected data using Kalman filtering algorithm, linear time-varying The discrete state equations and observational equation of system are:
In formula (1), X (k) and Y (k) is the state vector and measurement vector at k moment respectively, and F (k, k-1) is that state turns Matrix is moved, U (k) is k moment dynamic noises, and T (k, k-1) is that system controls matrix, and H (k) is k moment observing matrixes, and N (k) is k The observation noise at moment.
Correlated variables is initialized before filtering starts, when then estimating current using the state vector of last moment The state vector at quarter:
It is the estimated value of current time (i.e. k moment) state vector, X (k-1) is last moment (i.e. k-1 moment) State vector.
Secondly, pre-estimation covariance matrix is calculated according to the following formula
Wherein, Q (k) is intermediate variable, Q (k)=U (k) UT(k) it is intermediate variable, C (k) is the association side at current time Poor matrix, subscript T represent the transposition of matrix.
After acquiring pre-estimation covariance matrix, substitutes into and calculate kalman gain matrix in formula (4):
Wherein, R (k) is intermediate variable, R (k)=N (k) NT(k)。
Secondly, the estimated value of state vector is updated:
Again, the estimated value of updated covariance matrix is calculated:
Wherein, I is unit matrix.
Finally, state vector and measurement vector are updated:
According to formula (2) to (7) cycle calculations, until completing filtering.
Monitored target is sought in step s 2 in the resultant acceleration of each sampled point, can utilize acceleration modulus algorithm According to formula (8) calculate monitored target each sampled point resultant acceleration ASVM
Wherein, Ax、Ay、AzThe respectively acceleration of three axial direction of the monitored target at a certain sampled point.
The advantages of resultant acceleration (acceleration mould vector) mode, is that the acceleration value in three axial directions, which is combined into one, to be added The variation of velocity amplitude so that human body no matter fall by whichaway, can reflect that acceleration value must change, and setting is closed The threshold value of reason, which can be completed simply to fall, to be judged.
It falls in human body and will produce a prodigious acceleration between moment, with ground, therefore the present invention uses acceleration Information is fallen to detect, and compares individual axis acceleration information, and resultant acceleration can more reflect the impact process size that human body is fallen.Figure 2 be the front and back resultant acceleration change curve of monitored target of embodiment of the present invention tumble, and Fig. 3 is sat down for monitored target before action Resultant acceleration change curve afterwards, Fig. 4, which hurries up/jogs for monitored target, acts the curve of front and back resultant acceleration variation.This Invention to measurement data by carrying out correlation analysis, it can be found that judging the movement of monitored target using resultant acceleration instantaneous value State is easily obscured " running " and is occurred with " tumble ", generates false alarm information.Fallen using resultant acceleration to differentiate merely, meeting False positive and false negative is caused to increase.The reason is that, wink of the monitored target under " running " and " tumble " two kinds of motion states When resultant acceleration value it is possible that situations such as phase, cause " running " condition discrimination to be " tumble " state, to generate false-alarm. Comparison is found, if running action excessive velocities, just will produce the case where instantaneous resultant acceleration is more than resultant acceleration when falling, And resultant acceleration when there is a situation where falling only occurs in the very short time more than a certain given threshold, and action of running Resultant acceleration can periodically be more than threshold value.
Therefore, in step s 4, according to the resultant acceleration of each sampled point judge monitored target motion state whether be When doubtful tumble state, a resultant acceleration threshold value need to be preset and stored, when at least one sampling obtained in step S2 When the resultant acceleration of point is more than above-mentioned resultant acceleration threshold value, then judge that the motion state of monitored target is doubtful tumble state. Doubtful tumble state herein includes the motion states such as run, jog, hurrying up, standing, sitting down, jumping, rotating, and also includes falling .But it is only capable of judging whether monitored target is doubtful tumble state at this time only by resultant acceleration, to determine and be supervised Survey whether object has tumble to be inclined to, it is also necessary to make further to judge in conjunction with the size of Eulerian angles.
In order to reduce human body due to run, jump etc. significantly act caused by judge by accident and whether to start automatic alarm after falling Signal provides foundation, and the embodiment of the present invention introduces Eulerian angles as the subsidiary conditions judged of falling, the appearance of Eulerian angles, that is, human body State angle.It acquires three axis angular rates, 3-axis acceleration, three axis geomagnetic field informations in real time first, is then calculated using quaternary number mode Go out three Eulerian angles (pitch angle, roll angle, course angle) of human body, then utilizes three Euler's angle informations and resultant acceleration, root The motion state (sit down, walking, lie down, fall) of monitored target is identified according to recognizer.
When it is implemented, in step s3 to collected 3-axis acceleration information, three axis angular rate information and three axis Magnetic Field is fused into Eulerian angles, is then carrying out difference to Eulerian angles.
For referential in three dimensions, the orientation of any coordinate system can be indicated with three Eulerian angles.Europe Angle is drawn to indicate the rotational value on three axial components of x, y, z using course angle yaw, pitch angle pitch and roll angle roll, it can To judge the variation of monitored target posture according to the difference of the angle change of above three Eulerian angles.When it is implemented, right The course angle, pitch angle and roll angle carry out difference respectively, obtain including the course angle difference, pitch angle difference and roll The sequence of angular difference value:
In formula, yaw is course angle, and pitch is pitch angle, and roll is roll angle, and k is current sample time, and k-1 is upper one Sampling instant.Wherein droll(0)=dpitch(0)=dyaw(0)=0,1 < k < Kmax, KmaxFor the length of above-mentioned sequence of differences, with Sampled point quantity is identical, and when the sampling period is T, the present invention usually samples within 0.3~0.6 sampling period, with inspection Survey whether monitored target has tumble to be inclined to, to realize early warning.
The deficiency sensitive for the threshold value in instantaneous resultant acceleration discriminant approach, rate of false alarm is high, the present invention in step s 5, The posture of monitored target is further identified in conjunction with the sequence of differences obtained in step S3.
When it is implemented, when the motion state to the monitored target in doubtful tumble state is made further to judge, this Invention has preset a resultant acceleration mean value threshold value and has been locally stored, and quilt is calculated according to by the resultant acceleration of all sampled points The resultant acceleration mean value of monitoring object, when the resultant acceleration mean value acquired is more than preset resultant acceleration mean value threshold value, and extremely The absolute value of a few above-mentioned Eulerian angles difference (such as the difference at current time and last moment pitch angle) is more than the Euler When the difference threshold of angle, then judges that monitored target has to fall and be inclined to.
In conclusion used in the embodiment of the present invention is to detect tumble behavior, base based on the mode of inertia sensing technology In the monitoring principle of " based on resultant acceleration judgement, supplemented by Eulerian angles judgement ", external auxiliary equipment is not needed, internal algorithm is multiple It is miscellaneous, but have the characteristics that real-time, reliability is high.
Based on inventive concept identical with the anti-tumble method for early warning of human body shown in FIG. 1, the embodiment of the present invention additionally provides A kind of anti-tumble early warning system of human body, as described in following example.The original solved the problems, such as due to the anti-tumble early warning system of the human body Reason is similar to the anti-tumble method for early warning of human body in Fig. 1, therefore the implementation of the anti-tumble early warning system of the human body may refer to people in Fig. 1 The implementation of the anti-tumble method for early warning of body, overlaps will not be repeated.
Fig. 5 is the structural schematic diagram of the anti-tumble early warning system of human body of the embodiment of the present invention.As shown in figure 5, the anti-tumble of human body Early warning system includes:Sensor unit 1, resultant acceleration computing unit 2, Eulerian angles sequence of differences generation unit 3, judging unit 4, Judging unit 5, warning message generation unit 6.
Sensor unit 1 is used to acquire the 3-axis acceleration information of the monitored target of multiple sampled points, three shaft angles in real time Velocity information and three axis geomagnetic field informations.
Resultant acceleration computing unit 2 is used to calculate monitored target in each sampled point according to the 3-axis acceleration information Resultant acceleration.When it is implemented, resultant acceleration computing unit 2 can be believed according to 1 collected 3-axis acceleration of sensor unit Breath, using acceleration modulus algorithm calculate monitored target each sampled point resultant acceleration.
Eulerian angles sequence of differences generation unit 3 is used for the 3-axis acceleration information according to each sampled point, three shaft angles speed It spends information and three axis geomagnetic field informations generates Eulerian angles sequence of differences.
Judging unit 4 is used for the resultant acceleration according to each sampled point and preset resultant acceleration threshold decision monitored target Motion state whether be doubtful tumble state.
Judging unit 5 is used for when judging unit 4 judges the motion state of monitored target for doubtful tumble state, according to The resultant acceleration mean value and preset resultant acceleration mean value threshold value obtained by the resultant acceleration of all sampled points, in conjunction with the Euler Angular difference value sequence and preset Eulerian angles difference threshold judge whether monitored target has tumble to be inclined to.
Warning message generation unit 6 is used to generate warning message when monitored target has tumble to be inclined to, with monitored Object is alarmed before falling.
In one embodiment, the anti-tumble early warning system of above-mentioned human body further includes a timing unit 7 and a judging unit 8.Its In, timing unit 7 is used to start timing after warning message generation unit 6 generates warning message, and judging unit 8 is for judging quilt Whether monitoring object within the preset alarm time cancels above-mentioned warning message.
In one embodiment, the anti-tumble early warning system of above-mentioned human body further includes locating unit 9, for acquiring monitored pair The geographic position data of elephant.For example, when monitored target is in outdoor environment, the Big Dipper/GPS dual-mode can be used in positioning unit 9 It positions to obtain the geographic position data of monitored target;When monitored target is in indoor environment, positioning unit 9 can be used Base station location obtains the geographic position data of monitored target.
Common, the anti-tumble early warning system of human body further includes a communication unit 10, full above-mentioned pre- based in timing unit 7 If warning message is sent to guardian after time of fire alarming, and the warning message includes the geographical location that positioning unit 9 acquires Data, to guard expansion rescue behind the place for knowing monitored target tumble.
Since sensor unit 1 may be shaken when acquiring the exercise data of monitored target, collect in this way Exercise data in have noise, to filter out noise, the anti-tumble early warning system of above-mentioned human body also typically includes Kalman filter Wave unit 11 is believed for the 3-axis acceleration information, three axis angular rate information and three axis earth's magnetic fields to collected each sampled point Breath carries out Kalman filtering, so as to subsequently carry out precision higher when data processing.
Alarm unit 12 includes voice alarm module, buzzer, vibrator and LED light, for being generated according to warning message The warning message that unit 6 generates sends out sound-light alarm.
Usually, Eulerian angles sequence of differences generation unit 3 includes a digital moving processor and a difference calculating module.Number Word movement processor is used for 1 collected 3-axis acceleration information of sensor unit, three axis angular rate information and three axis earth magnetism Field information is fused into Eulerian angles, wherein above-mentioned Eulerian angles include course angle, pitch angle and roll angle.Difference calculating module is used for Difference is carried out respectively to above-mentioned course angle, pitch angle and roll angle, obtains including the course angle difference, pitch angle difference and cross The sequence of roll angle difference.
Judging unit 4 is being identified the motion state of monitored target with resultant acceleration threshold value according to resultant acceleration When, it is only necessary to when the resultant acceleration of a sampled point is more than preset resultant acceleration threshold value, you can judge the fortune of monitored target Dynamic state is doubtful tumble state.If there is the resultant acceleration of multiple sampled points is all higher than above-mentioned resultant acceleration threshold value, then by The motion state one of monitoring object is set to doubtful tumble state.
Judging unit 5 generally includes a resultant acceleration mean value computation module and a judging submodule.Wherein, resultant acceleration is equal It is worth the conjunction acceleration mean value that computing module is used to calculate monitored target according to the resultant acceleration of each sampled point.Judging submodule is used for When above-mentioned resultant acceleration mean value is more than preset resultant acceleration mean value threshold value, and at least one above-mentioned Eulerian angles difference is absolute When value is more than preset Eulerian angles difference threshold, judgement monitored target, which has to fall, to be inclined to.
When monitored target has tumble to be inclined to, to reduce the monitored target probability injured when falling, above-mentioned human body Anti- tumble system further includes an air bag executing agency 13.Air bag executing agency 13 generally includes a compressed gas cylinder and an air bag, gas Capsule executing agency 13 can be connect with warning message generation unit 6 by steering engine interface, for being punctured according to above-mentioned warning message , to provide buffering when human body is fallen, personal injury's probability is reduced to be inflated to air bag in the air port of compressed gas cylinder.It is above-mentioned Air bag can make the form of waistband or vest, but the present invention is not limited thereto.
Air bag executing agency 13 in the embodiment of the present invention can be designed to the purse shape being made of dark lattice and storage space Formula, for dark lattice for fixing control circuit board and gas cylinder, executing agency, purse bandage includes air bag, have purse, tumble early warning, Tumble protection, lamp multiple functions of seeking help of falling.In addition, air bag executing agency 13 is also designed to other shapes such as waistband, vest Formula, flexible design are various.
Triggering air bag executing agency behaviour weight wants joint to provide protection, communication unit when monitored target has tumble to be inclined to 10 after monitored target tumble timely alert, guardian can receive tumble message by network, short message, with The people of family at one's side can hear the voice help information that alarm unit 12 is sent out, and provide help in time.
The anti-tumble early warning system of human body further includes a storage unit 14, data, letter for the 1 yuan of acquisition of storage sensor list The warning message etc. that the intermediate data and warning message generation unit 6 generated in breath processing processing procedure generates.
Currently, most of similar products do not have memory module 14 or use fixed storage core on circuit boards in the market Piece.Memory module 14 can select built-in TF card in the anti-tumble early warning system of human body in the embodiment of the present invention, store inertia The parameters such as sensing data, tumble pre-warning time facilitate carry out off-line analysis, and facilitate plug, help to carry out user a Propertyization analysis, modeling, provide more accurately personalized service.In another embodiment, memory module can also use SDHC cards.
The above-mentioned anti-tumble early warning system of human body further includes power supply unit 15, for each in the anti-tumble early warning system of human body Unit is powered.
The anti-tumble early warning system of human body for a better understanding of the present invention and advantage, with reference to specific example Son illustrates.
Fig. 6 is the frame structure and connection diagram of the anti-tumble early warning system of human body of the embodiment of the present invention.The present embodiment carries The controller mainboard of the anti-tumble early warning system of human body of confession is by sensor unit, micro controller module, power supply module, communication mould Block, alarm unit, rudder control interface module and other interfaces composition, connection type are shown in Fig. 6.
When carrying out PCB layout to the anti-tumble early warning system of human body shown in Fig. 6, power cord and signal wire can be layered cloth It sets, avoids parallel long apart from cabling, to improve Electro Magnetic Compatibility, while sensor unit and micro controller module are integrated in together On one piece of circuit board, the anti-tumble early warning system of human body of more high integration and specificity is made.
As shown in fig. 6, the sensor unit in the present embodiment uses the combination of MPU6050 chips and AK8975 chips, MPU6050 built-in chip types three axis accelerometer and three-axis gyroscope, AK8975 built-in chip types three axle magnetometer, are respectively used to Acquire 3-axis acceleration information, the three axis geomagnetic field informations of three axis angular rate information and monitored target of monitored target.
The main control chip of micro controller module uses the STM32F407VGTE microcontrol processors of ST companies, has 32 Position ARM Contex M4 kernels, and there is 1MB Flash, 192KB RAM, it is possible to provide the up to clock of 168MHz ensure that Algorithm can be fast and reliable execution.Micro controller module passes through spi bus and three-axis gyroscope, three axis accelerometer, three axis Magnetometer (for sake of convenience, three sensors are referred to as nine axis inertial sensors by this by the present invention) communicates, Ke Yijie The 3-axis acceleration, three axis angular rates, three axis geomagnetic field informations for receiving human body, to provide original for the identification of human motion and posture Beginning exercise data source.The exercise data of nine axis inertial sensor of main control chip pair acquisition carries out complicated data fusion and appearance Gesture identifies, to judge the athletic posture of monitored target.The dangerous play to people is realized according to the posture of monitored target It reminds and seeks help when falling.In addition, abundant serial port resource cooperation communication module that main control chip additionally provides (such as Simcom companies release SIM900) it may be implemented to communicate with the data of server, it can also realize and guard by short message mode The short message interacting of people.
In another embodiment, the communication between sensor unit and micro controller module can also use I2C mode.
In the present embodiment, alarm unit group will be made of audio alert, buzzer, vibrator and LED light, wherein language Sound alarm portion is made of a piece of speech chip and loudspeaker, can also be by voice around when monitored target is fallen Crowd seeks help, and hardware composition frame chart is as shown in Figure 7.
Speech chip receives the alarm signal from microcontrol processor, selects preset voice segments order, the output phase The voice signal answered to loudspeaker, driving loudspeaker make a sound.In addition, vibrator receives signal, vibration is generated, is prompted monitored Object danger action, to prevent fall.Mode of vibration has the elderly being hard of hearing and the scene to need peace and quiet more preferable Effect.After alarm unit receives warning message, LED light therein can constantly flicker, and prompt dangerous play behavior.
The UM220-III-N produced with Xin Xingtong companies and antenna and its peripheral circuit composition may be used in locating module, It is the third generation BDS/GPS bimodulus locating modules newly released with Xin Xingtong companies, is smallest size of complete currently on the market The BDS/GPS modules of production domesticization have the characteristics that low-power consumption, low cost, high performance-price ratio and high integration, support that single system is independent Positioning and multisystem combined positioning, under bimodulus positioning scenarios, outdoor positioning precision can reach 3-5 meters, and it is independent to be much higher than GPS 10-15 meters of the precision of positioning, positioning accuracy is outstanding.
In the present embodiment, the anti-tumble early warning system of human body can select battery powered mode, USB can also be selected to power Mode.When choosing battery power supply mode, the 3.7V lithium battery power supplies of standard may be used, final product form can be can not Detachable battery.If selecting USB power supply modes, the Mini USB interfaces of standard, filling in combined charge circuit can be used Electric managing chip TP4056 completes the charging to lithium battery.USB power source passes through linear regulator chip PAM3310 (not shown)s Can directly be that whole circuit board is powered after voltage stabilizing rectification.It is filled in addition, being usually also configured with two in the anti-tumble early warning system of human body Electric indicator light, red colored lamp are used to indicate power supply connection, and blue lamp after continuously flickering, being full of in battery underfill for being always on.
Why the anti-tumble early warning system of human body provided in an embodiment of the present invention, which can realize that human body is fallen, is protected, and is to pass through inspection Control steering engine punctures compressed gas cylinder after measuring tumble, quickly protects to the important joint of human body, avoids to realize to airbag aeration Human body hard landing.Due to using " control-execution " separation design, completing exercise data by controller mainboard and adopting After collection, the work of human body attitude analysis and identification, then tumble protection is completed by air bag executing agency and is acted, passes through controller master between the two The steering engine interface module reserved on plate realizes interaction.Steering engine interface module is by power supply and signal line group at signal wire can connect rudder The control signal wire of machine completes the control rotation of steering engine;Power cord provides the voltage of 7.8V, with for steering engine power supply power supply, in addition Power cord also provides charging interface for steering engine battery.In the present invention, it controls, the separate design of executing agency, facilitates production The design of product and the wearing of user, increase product shows form.
Also there are jtag interfaces on controller mainboard, can be by STM32F407 serial port debuggers under in host computer Program is carried to main control chip, reserved 20 extension mouths are to be connect with other systems, for example connection can be with adjust automatically foot Anti- tumble intervening measures such as " the anti-tumble rehabilitative shoes " of bottom pressure branch, realizes more functions later.
The anti-main applicable object of tumble early warning system of the human body is elderly population or easily falls risk population, by being worn on It needs on the user's body monitored, realization acquires user's posture in real time, when to make corresponding early warning and tumble generation Protection act, main workflow include initialization, monitoring pattern adjustment, tumble early warning and tumble processing and etc., As shown in Figure 8.
Initialization step:In user's wearing after the anti-tumble system of human body, the power supply of the system is opened, into initialization shape State, including micro controller module, communication module, locating module and sensor unit will all carry out initialization operation.It should be noted that , the initialization time of nine axis data pick-ups is longer, needs 30 seconds or so, and user needs to wait for the corresponding time, during which Standing state need to be kept, so as to pick up calibration.
Monitoring pattern adjusts:After the completion of initialization, automatically into real-time monitoring pattern, the anti-tumble system of the present embodiment human body In communication module it is in running order, fall once being monitored to monitored target, communication module will be from trend service Device sends warning message, and the location information of monitored target is included in the warning message, and communication module also can short massage notice monitoring People.The shortcomings that real-time monitoring pattern is exactly that hardware device all in the anti-tumble system of human body is in state of activation, power consumption compared with It is high.For this purpose, the present embodiment also sets the mode of mode adjustment, when nine axle sensors differentiate that human body remains static for a long time When, remote tracing pattern can be automatically switched to, in this mode, micro controller module is by discontinuously opening and closing communication mould Block come achieve the purpose that reduce power consumption.Can be with manual setting routine tracking and monitoring pattern, communication module is closed in this mode It closes, greatly reduces power consumption, the disadvantage is that monitored target can not notify guardian in time after falling, be suitble to ring indoors It is used under the scene that border, household accompany.In addition, can mutually switch between above-mentioned Three models, intelligent mode can be selected, i.e., Micro controller module recognizes switching automatically according to the data of acquisition, can also be by user by way of switching, sending command set It switches over.
The hardware resource used under different monitoring patterns is different, causes hardware power consumption different.In order to adapt to different fields Scape achievees the purpose that reduce power consumption, extends usage time, intelligence may be implemented also for the power consumption of the anti-tumble early warning system of human body is made It can adjust, the embodiment of the present invention additionally provides a kind of monitoring pattern method for transformation of the anti-tumble early warning system of human body, such as Fig. 9 institutes Show.
After system initialization, acquiescence enters real-time monitoring pattern, is monitored in real time to the motion state of monitored target. Monitored target can manually control, and send corresponding instruction and be switched to other patterns, to adapt to specifically used scene.Conversely, Under default situations, when the monitored target remains stationary regular hour, the anti-fall monitoring system of human body then thinks monitored target In non-athletic state, remote tracing monitoring pattern will be automatically switched to, communication module is primary by separated in time suspend mode, with Reduce power consumption.If fallen during suspend mode, micro controller module will actively wake up communication module, and rapid initialization is concurrent Go out warning message.The embodiment of the present invention provides short message, GPRS data communication and three kinds of command interfaces of serial ports, realizes and prevents falling with human body Monitor the interaction of system.
Multiple-working mode expands the usage scenario of the present invention, and the intelligence switching of monitoring pattern greatly reduces product work( Consumption reduces the operation of user, improves its ease for use as far as possible, and realizing wearer can be with intelligent selection pattern without intervening Function.
Tumble early warning:Which kind of no matter under monitoring pattern, it is inclined to once monitored target has to fall, micro controller module Monitored target will be reminded to pay attention to taking precautions against fall risk by alarm unit, specific dangerous play may refer to table 1.Work as detection After falling to monitored target, it may be possible to which monitored target injury relatively gently (is fallen in monitored target maloperation The air bag of while falling can want joint to provide protection for people's weight), therefore the present embodiment additionally provides a kind of mechanism for cancelling alarm, such as Fruit monitored target independently cancels alarm before the deadline, then warning message is not transmitted to guardian;If monitored Object can not independently cancel alarm before the deadline, i.e. injury situation is than more serious, then after timing reaches the defined time, Warning message will be sent to server and guardian by communication module.
1 dangerous play list of table
In addition, by the distinguishing rule in table 1, in conjunction with human body attitude, the present embodiment, which has also been distinguished, lies down, turns over, rising It several actions such as stands, sit down, falling, and can accurately distinguish and give corresponding prompt message, referring to table 2 and table 3.
2 common action list of table
3 doubtful tumble state of table and judgement of falling
Wherein, " inclination angle " in table 1 above~table 3 refers to the absolute value of Eulerian angles difference.
Tumble is handled:According to the distinguishing rule in table 2 and table 3, the microcontroller mould in the anti-tumble early warning system of human body When block detects that monitored target has tumble tendency, human body does not land at this time, can pass through rudder control interface and trigger compression Gas cylinder, moment are airbag aeration, air bag expansion before human body lands, to realize the protection to body key position.If Monitored target does not press cancellation alarm keys within the preset alarm time (such as 20 seconds), and micro controller module will pass through channel radio Believe that module sends warning message to server and guardian.Meanwhile alarm unit opens audio alert mould according to warning message Block, cycle report relief information, and surrounding population is reminded to give the wounded's relief.And communication module provides three kinds of interface operation sides Formula:Short message, GPRS data and USB (serial ports), the effect that these three interfaces are realized are the same, can realize to human body Configuration, operation and the data interactive function of anti-tumble early warning system, different scenes can flexibly select operating and controlling interface.To service The warning message that device is sent is the data for having specific format, includes mainly:Synchronous head " BSN ", telephone number, base station location letter It breath, GPS/BD location informations (longitude and latitude), the anti-tumble early warning system voltage of human body and verification and waits, the size of data frame is solid It is fixed, facilitate resolve packet.Wherein, BSN indicates Body Sensor Network, means human body sensor network, such as one Complete data content is following " BSN130119511******** ... ".The short message content sent to guardian includes that relief is believed Breath, longitude and latitude, content are customizable.
Figure 10 is gesture recognition algorithms flow diagram of the embodiment of the present invention.According to the reference standard provided in table 1, table 2, In conjunction with algorithm flow shown in Fig. 10, the current posture of human body can be identified by the exercise data that sensor acquires, in order to The posture for more accurately indicating human body, can also further segment the classification situation in table 1 and table 2, with more complete The real-time attitude of human body is described.It should be noted that low-power consumption monitoring is for the monitoring pattern addition before coordinating, determine Justice is:When acceleration maintains near 0 within a certain period of time, and the collected exercise data of sensor delays in prescribed limit When slow wave moves, then it is assumed that user is in non-athletic state.
By experimental verification, as a result prove that the anti-tumble method for early warning of human body provided by the invention and system are feasible.This Inventive embodiments choose adult human's (10 male 5 female, age of 15 young healthies altogether:24 ± 3.5 years old, weight:62± 14.5kg height:170 ± 12cm) take part in experiment, experimental arrangement include calibrate, sit down, standing up, lying down, left turn, the right side are turned over Body, it is preceding fall, it is right fall and it is rear fall, everyone respectively does 3 times in each action, has done 360 times (being free of calibration actions) in total and has fallen experiment, together When with a high-speed camera record actual tumble pre-warning time (can also being obtained by the data stored in memory module).
The results show, when judging that the acceleration rate threshold of tumble early warning is set as 4.5m/s2When, detection is sensitive Degree is 98.61%, and specificity 98.61, results contrast is ideal, and averagely tumble pre-warning time is 300ms, if appropriate reduce spirit Sensitivity (such as 80%~90%), pre-warning time can reach 400ms~500ms.Figure 11 is to be fallen to test sample according to above-mentioned 360 times Scheme made by this result.In fig. 11 ,+number expression outlier, piston-like figure share 5 horizontal lines from top to bottom, respectively generation Table maximum value, 75% quantile, median, 25% quantile, minimum value, so the box in the middle part of piston-like figure indicates to concentrate In 25%~75% experiment sample.The present invention is only illustrated by taking Right deviation as an example, and volunteer is when making Right deviation tumble experiment, greatly The pre-warning time of part Experiment sample concentrates between 0.25s~0.31s, and only there are one the pre-warning times of outlier is about 0.45s illustrates that the early warning sensitivity of the anti-tumble method for early warning of human body provided by the invention and system is higher.
The present invention is more remarkable to the real-time and accuracy of judgement of falling, realizes and stands up, sits down, lying down, turning over Various motion detects, significant increase tumble discrimination.
It is found according to documents and materials and actual test, the positions such as arm, wrist and leg of human body are because in daily life Act that randomness is very big, and activity is also relatively frequent in behavior, and below neck, the position of waist area above is in physical activity In the process activity infrequently and acceleration and angular speed can find violent variation, and the center of gravity position of human body during tumble It sets and also concentrates on this subregion, nine axle sensor devices, which are worn on the position among waist, can preferably reflect entire people The variation of the motion state and posture of body trunk, more conducively fall detection.
In the embodiment shown in fig. 6, sensor unit can also be replaced directly by MPU9150 chips and its peripheral circuit The combination of MPU6050 chips and AK8975 chips, MPU9150 chips are INVENSENSE companies new generation product, one single chip Inside it is integrated with accelerometer, gyroscope and magnetometer, and built-in DMP (Digital Motion Processor, digital moving Processor) it is merged for posture, 16 bit A/D converter of built-in chip type, 16 data outputs (magnetic field data is 13 outputs), tool There is self-calibration function.
Sensing unit in the embodiment of the present invention separates independent design with microcontroller mainboard, convenient for upgrading or replacing other Sensor unit is conducive to improve the Electro Magnetic Compatibility of whole circuit board;Multiple I/O mouthfuls reserved (Input/Output, it is defeated Enter/export) it ensure that the maintainability and scalability of product, also it is conducive to later stage HardwareUpgring.
The present invention easily falls crowd for the elderly etc. and designs, and has the function of tumble early warning, protection and communication etc., in human body Have the function of " early warning " before falling, breach similar products on Vehicles Collected from Market only have fall after Alarm function or only There is early warning and without tumble safeguard function.The overall construction design of the anti-tumble early warning system of human body provided by the invention is ingenious, can be single It solely wears, is used as warning function;Air bag executing agency can also be coordinated to use, realize tumble rear protecting and relief;In addition, Reserved serial ports can conveniently realize IAP (In Application Programing, in-service units), be conducive to firmware liter Grade.
Usually only has the sensing data of acceleration mode or multiple mode for similar products in the market by different circuits Module provides, and is unfavorable for real-time and accurate the problem of merging of multi-modal sensing data, the present invention is using in high precision and multi-modal The sensor chip of data, and nine axis inertia sensing design datas are on the same circuit module, collect data precision and Mode quantity is relatively more, can much sooner and accurately to the judgement of tumble.
The present invention realizes " early warning " before human body is fallen, while triggering air bag executing agency behaviour weight will close when falling Section provides protection, and timely alert after tumble, guardian can receive tumble message by network, short message, in user People at one's side can hear the voice help information that precaution device is sent out, and provide help in time.
The present invention is a kind of anti-tumble precaution device of intelligent human-body.For the serious problem of current social aging, the elderly Coordination of body ability gradually declines with the age, falls in daily life inevitable, and fall would generally cause it is prodigious Life security and economic problems.And the prior art mostly has above-mentioned " alarming after tumble ", " it is not convenient enough to wear use ", " falls The judgement of falling accuracy is undesirable " the shortcomings of or deficiency, therefore the present invention devises a kind of anti-tumble precaution device of intelligent human-body.This is pre- Whether alert device can inevitably occur in 300ms~500ms anticipations tumble behavior before human body tumble, if can keep away Exempt from, can be avoided risk by voice, buzzer and vibrator call user's attention;It, can trigger protection dress if inevitable Offer protection is provided, while notifying guardian, voice reminder surrounding population to provide relief by wireless network.
Specific embodiment is applied in the present invention, and principle and implementation of the present invention are described, above example Explanation be merely used to help understand the present invention method and its core concept;Meanwhile for those of ordinary skill in the art, According to the thought of the present invention, there will be changes in the specific implementation manner and application range, in conclusion in this specification Appearance should not be construed as limiting the invention.

Claims (19)

1. a kind of anti-tumble method for early warning of human body, which is characterized in that the anti-tumble method for early warning of human body includes:
The 3-axis acceleration information, three axis angular rate information and three axis earth's magnetic fields of the monitored target of multiple sampled points are acquired in real time Information;
According to the 3-axis acceleration information calculate monitored target each sampled point resultant acceleration;
Eulerian angles are generated according to the 3-axis acceleration information, three axis angular rate information and three axis geomagnetic field informations of each sampled point Sequence of differences;
According to the resultant acceleration of each sampled point and the motion state of preset resultant acceleration threshold decision monitored target whether be Doubtful tumble state;
If so, the resultant acceleration mean value obtained according to the resultant acceleration by all sampled points and preset resultant acceleration mean value threshold Value judges whether monitored target has tumble to be inclined in conjunction with the Eulerian angles sequence of differences and preset Eulerian angles difference threshold;
Warning message is generated when monitored target has tumble to be inclined to, to alarm before monitored target is fallen;
The anti-tumble method for early warning of human body further includes:
The 3-axis acceleration information, three axis angular rate information and three axis geomagnetic field informations of collected each sampled point are carried out Kalman filtering;
When being filtered to collected data using Kalman filtering algorithm, the discrete state equations of linear time varying system and sight Surveying equation is:
In formula (1), X (k) and Y (k) is the state vector and measurement vector at k moment respectively, and F (k, k-1) is that state shifts square Battle array, U (k) are k moment dynamic noises, and T (k, k-1) is that system controls matrix, and H (k) is k moment observing matrixes, and N (k) is the k moment Observation noise;
Correlated variables is initialized before filtering starts, then estimates current time using the state vector of last moment State vector:
It is the estimated value of current time (i.e. k moment) state vector, X (k-1) is the shape of last moment (i.e. k-1 moment) State vector;
Secondly, pre-estimation covariance matrix is calculated according to the following formula
Wherein, Q (k) is intermediate variable, Q (k)=U (k) UT(k) it is intermediate variable, C (k) is the covariance square at current time Battle array, subscript T represent the transposition of matrix;
After acquiring pre-estimation covariance matrix, substitutes into and calculate kalman gain matrix in formula (4):
Wherein, R (k) is intermediate variable, R (k)=N (k) NT(k);
Secondly, the estimated value of state vector is updated:
Again, the estimated value of updated covariance matrix is calculated:
Wherein, I is unit matrix;
Finally, state vector and measurement vector are updated:
According to formula (2) to (7) cycle calculations, until completing filtering.
2. the anti-tumble method for early warning of human body according to claim 1, which is characterized in that the anti-tumble method for early warning of human body Further include:
Air bag executing agency punctures gas cylinder therein to be inflated to air bag therein, in human body according to the warning message Buffering is provided when tumble.
3. the anti-tumble method for early warning of human body according to claim 1, which is characterized in that the anti-tumble method for early warning of human body Further include:
Start timing after generating the warning message, judges whether the monitored target is cancelled within the preset alarm time The warning message, if it is not, the warning message is then sent to guardian after the meter full preset alarm time.
4. the anti-tumble method for early warning of human body according to claim 1, which is characterized in that according to the 3-axis acceleration information Calculate monitored target each sampled point resultant acceleration, including:
According to the 3-axis acceleration information, calculates monitored target using acceleration modulus algorithm and accelerate in the conjunction of each sampled point Spend ASVM
Wherein, Ax、Ay、AzThe respectively acceleration of three axial direction of the monitored target at same sampled point.
5. the anti-tumble method for early warning of human body according to claim 1, which is characterized in that according to three axis of each sampled point Acceleration information, three axis angular rate information and three axis geomagnetic field informations generate Eulerian angles sequence of differences:
The 3-axis acceleration information, three axis angular rate information and three axis geomagnetic field informations are fused into Eulerian angles, the Euler Angle includes course angle, pitch angle and roll angle;
Difference is carried out respectively to the course angle, pitch angle and roll angle, is obtained comprising the course angle difference, pitch angle difference And the sequence of roll angle difference:
Wherein, yaw is course angle, and pitch is pitch angle, and roll is roll angle, and k is current sample time, and k-1 is a upper sampling Moment.
6. the anti-tumble method for early warning of human body according to claim 1, which is characterized in that according to the resultant acceleration of each sampled point And whether the motion state of preset resultant acceleration threshold decision monitored target is doubtful tumble state, including:
When the resultant acceleration of an at least sampled point is more than the resultant acceleration threshold value, the movement shape of the monitored target is judged State is doubtful tumble state.
7. the anti-tumble method for early warning of human body according to claim 1, which is characterized in that added according to the conjunction by all sampled points The resultant acceleration mean value and preset resultant acceleration mean value threshold value that speed obtains, in conjunction with the Eulerian angles sequence of differences and preset Eulerian angles difference threshold judges whether monitored target has tumble to be inclined to, including:
According to the resultant acceleration of each sampled point, the conjunction for calculating monitored target accelerates mean value;
When the resultant acceleration mean value is more than preset resultant acceleration mean value threshold value, and at least one Eulerian angles difference When absolute value is more than the Eulerian angles difference threshold, judges that the monitored target has to fall and be inclined to.
8. a kind of anti-tumble early warning system of human body, which is characterized in that the anti-tumble early warning system of human body includes:
Sensor unit, the 3-axis acceleration information of the monitored target for acquiring multiple sampled points in real time, three axis angular rates Information and three axis geomagnetic field informations;
Resultant acceleration computing unit adds for calculating monitored target according to the 3-axis acceleration information in the conjunction of each sampled point Speed;
Eulerian angles sequence of differences generation unit, for being believed according to the 3-axis acceleration information of each sampled point, three axis angular rates Breath and three axis geomagnetic field informations generate Eulerian angles sequence of differences;
First judging unit is used for the resultant acceleration according to each sampled point and preset resultant acceleration threshold decision monitored target Motion state whether be doubtful tumble state;
Second judgment unit, for judging the motion state of monitored target for doubtful tumble state in first judging unit When, the resultant acceleration mean value and preset resultant acceleration mean value threshold value obtained according to the resultant acceleration by all sampled points, in conjunction with The Eulerian angles sequence of differences and preset Eulerian angles difference threshold judge whether monitored target has tumble to be inclined to;
Warning message generation unit, for generating warning message when monitored target has tumble to be inclined to, in monitored target It alarms before tumble;
The anti-tumble early warning system of human body further includes:
Kalman filtering unit, for the 3-axis acceleration information, the three axis angular rate information to collected each sampled point Kalman filtering is carried out with three axis geomagnetic field informations;
When being filtered to collected data using Kalman filtering algorithm, the discrete state equations of linear time varying system and sight Surveying equation is:
In formula (1), X (k) and Y (k) is the state vector and measurement vector at k moment respectively, and F (k, k-1) is that state shifts square Battle array, U (k) are k moment dynamic noises, and T (k, k-1) is that system controls matrix, and H (k) is k moment observing matrixes, and N (k) is the k moment Observation noise;
Correlated variables is initialized before filtering starts, then estimates current time using the state vector of last moment State vector:
It is the estimated value of current time (i.e. k moment) state vector, X (k-1) is the state of last moment (i.e. k-1 moment) Vector;
Secondly, pre-estimation covariance matrix is calculated according to the following formula
Wherein, Q (k) is intermediate variable, Q (k)=U (k) UT(k) it is intermediate variable, C (k) is the covariance square at current time Battle array, subscript T represent the transposition of matrix;
After acquiring pre-estimation covariance matrix, substitutes into and calculate kalman gain matrix in formula (4):
Wherein, R (k) is intermediate variable, R (k)=N (k) NT(k);
Secondly, the estimated value of state vector is updated:
Again, the estimated value of updated covariance matrix is calculated:
Wherein, I is unit matrix;
Finally, state vector and measurement vector are updated:
According to formula (2) to (7) cycle calculations, until completing filtering.
9. the anti-tumble early warning system of human body according to claim 8, which is characterized in that the anti-tumble early warning system of human body Further include:
Timing unit, for starting timing after generating the warning message;
Third judging unit, for judging whether the monitored target within the preset alarm time cancels the alarm signal Breath.
10. the anti-tumble early warning system of human body according to claim 9, which is characterized in that human body anti-tumble early warning system System further includes:Positioning unit, the geographic position data for acquiring the monitored target.
11. the anti-tumble early warning system of human body according to claim 10, which is characterized in that the anti-tumble system of human body is also Including:The warning message is sent to guardian, the alarm by communication unit after the full preset alarm time based on Information includes the geographic position data.
12. the anti-tumble early warning system of human body according to claim 8, which is characterized in that the resultant acceleration computing unit It is specifically used for:
According to the 3-axis acceleration information, calculates monitored target using acceleration modulus algorithm and accelerate in the conjunction of each sampled point Spend ASVM
Wherein, Ax、Ay、AzThe respectively acceleration of three axial direction of the monitored target at same sampled point.
13. the anti-tumble early warning system of human body according to claim 8, which is characterized in that the Eulerian angles sequence of differences life Include at unit:
Digital moving processor, for melting the 3-axis acceleration information, three axis angular rate information and three axis geomagnetic field informations Eulerian angles are synthesized, the Eulerian angles include course angle, pitch angle and roll angle;
Difference calculating module obtains including the course for carrying out difference respectively to the course angle, pitch angle and roll angle The sequence of angular difference value, pitch angle difference and roll angle difference:
Wherein, yaw is course angle, and pitch is pitch angle, and roll is roll angle, and k is current sample time, and k-1 is a upper sampling Moment.
14. the anti-tumble early warning system of human body according to claim 8, which is characterized in that first judging unit is specific For:
When the resultant acceleration of an at least sampled point is more than the resultant acceleration threshold value, the movement shape of the monitored target is judged State is doubtful tumble state.
15. the anti-tumble early warning system of human body according to claim 8, which is characterized in that the second judgment unit includes:
Resultant acceleration mean value computation module, for the resultant acceleration according to each sampled point, the conjunction for calculating monitored target adds Fast mean value;
Judging submodule, for being more than preset resultant acceleration mean value threshold value when the resultant acceleration mean value, and it is at least one When the absolute value of the Eulerian angles difference is more than the Eulerian angles difference threshold, judges that the monitored target has to fall and incline To.
16. the anti-tumble early warning system of human body according to claim 8, which is characterized in that human body anti-tumble early warning system System further includes an air bag executing agency, and the air bag executing agency is connected with the warning message generation unit by steering engine interface It connects, for puncturing gas cylinder therein to be inflated to air bag therein, to when human body is fallen according to the warning message Buffering is provided.
17. the anti-tumble early warning system of human body according to claim 8, which is characterized in that human body anti-tumble early warning system System further includes:Alarm unit, including voice alarm module, buzzer, vibrator and LED light, for according to the warning message Send out sound-light alarm.
18. according to the anti-tumble early warning system of claim 8-17 any one of them human bodies, which is characterized in that the human body is anti-to fall Early warning system further includes:Storage unit is generated for storing in the data of sensor unit acquisition, information process Intermediate data and the warning message generation unit generate warning message.
19. according to the anti-tumble early warning system of claim 8-17 any one of them human bodies, which is characterized in that the human body is anti-to fall Early warning system further includes:Power supply unit, for powering to the anti-tumble early warning system of the human body.
CN201511028550.5A 2015-12-31 2015-12-31 A kind of anti-tumble method for early warning of human body and system CN105632101B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201511028550.5A CN105632101B (en) 2015-12-31 2015-12-31 A kind of anti-tumble method for early warning of human body and system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201511028550.5A CN105632101B (en) 2015-12-31 2015-12-31 A kind of anti-tumble method for early warning of human body and system

Publications (2)

Publication Number Publication Date
CN105632101A CN105632101A (en) 2016-06-01
CN105632101B true CN105632101B (en) 2018-10-23

Family

ID=56046975

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201511028550.5A CN105632101B (en) 2015-12-31 2015-12-31 A kind of anti-tumble method for early warning of human body and system

Country Status (1)

Country Link
CN (1) CN105632101B (en)

Families Citing this family (18)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106056849B (en) * 2016-07-15 2018-06-26 西安电子科技大学 Falls Among Old People intelligent measurement and the passive warning system of master and method of positioning
CN106023528A (en) * 2016-08-01 2016-10-12 王琰 Emergency monitoring device for old people
CN106373335A (en) * 2016-08-16 2017-02-01 中国人民解放军海军工程大学 Inertial technology based and Android phone applied detection and smart alarm method for old people's fall
CN106228751B (en) * 2016-08-26 2019-06-07 陈雷 Based on Android platform MEMS/ Magnetic Sensor/GPS tumble intelligent alarm system and method
CN106473265B (en) * 2016-12-03 2018-01-12 石家庄学院 A kind of intelligence, which is fallen, guards T-shirt and its fall detection algorithm
CN106710141A (en) * 2016-12-12 2017-05-24 东莞市广信知识产权服务有限公司 Electronic sensing system capable of monitoring falling of old person
CN106846726B (en) * 2016-12-23 2019-02-01 上海正苒智能科技有限公司 It is a kind of for obtaining the monitoring alarm method and system of indoor occupant situation
CN106803219A (en) * 2017-01-18 2017-06-06 吉林大学 A kind of work and rest identification record method towards newborn care based on Intelligent foot ring
CN108510707B (en) * 2017-02-27 2020-04-28 芜湖美的厨卫电器制造有限公司 Electric water heater and method and device for prompting falling based on electric water heater in bathroom
CN107146377A (en) * 2017-04-14 2017-09-08 深圳大学 Fall detection method and device before collision
CN107085246A (en) * 2017-05-11 2017-08-22 深圳合优科技有限公司 A kind of human motion recognition method and device based on MEMS
CN107202594A (en) * 2017-05-11 2017-09-26 深圳合优科技有限公司 A kind of data processing method and device based on MTK platforms
CN109035696A (en) * 2018-06-21 2018-12-18 福州大学 One kind falling down detection method based on acceleration transducer
CN108876822B (en) * 2018-07-09 2020-11-20 山东大学 Behavior risk degree evaluation method and home security nursing system
CN109091151A (en) * 2018-09-06 2018-12-28 中国人民解放军战略支援部队信息工程大学 A kind of pedestrian's fall detection method and device based on MIMU
CN109171751A (en) * 2018-10-24 2019-01-11 广东工业大学 A kind of human posture recognition method and device
CN109686051A (en) * 2018-12-18 2019-04-26 广东工业大学 A kind of determination method, device, equipment and the medium of target object tumble behavior
CN110269373A (en) * 2019-05-31 2019-09-24 汕头大学 Intelligence based on photovoltaic power generation is ridden knapsack, working method and storage medium

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103027687A (en) * 2012-12-05 2013-04-10 北京工业大学 Old people fall detection and alarm system based on 3D (3-dimensional) accelerometer and gyroscope
CN103308069A (en) * 2013-06-04 2013-09-18 电子科技大学 Falling-down detection device and method
CN103377542A (en) * 2013-07-16 2013-10-30 中国科学院深圳先进技术研究院 Human body fall-down preventing early warning method and human body fall-down preventing early warning device
CN104680718A (en) * 2015-02-10 2015-06-03 柳州市金旭节能科技有限公司 Intelligent bracelet
WO2015159484A1 (en) * 2014-04-15 2015-10-22 パナソニックIpマネジメント株式会社 Controller and device state determination system using same

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103027687A (en) * 2012-12-05 2013-04-10 北京工业大学 Old people fall detection and alarm system based on 3D (3-dimensional) accelerometer and gyroscope
CN103308069A (en) * 2013-06-04 2013-09-18 电子科技大学 Falling-down detection device and method
CN103377542A (en) * 2013-07-16 2013-10-30 中国科学院深圳先进技术研究院 Human body fall-down preventing early warning method and human body fall-down preventing early warning device
WO2015159484A1 (en) * 2014-04-15 2015-10-22 パナソニックIpマネジメント株式会社 Controller and device state determination system using same
CN104680718A (en) * 2015-02-10 2015-06-03 柳州市金旭节能科技有限公司 Intelligent bracelet

Also Published As

Publication number Publication date
CN105632101A (en) 2016-06-01

Similar Documents

Publication Publication Date Title
US10052034B2 (en) Wearable devices for sensing, displaying, and communicating data associated with a user
US10134256B2 (en) Portable monitoring devices and methods of operating the same
CN104799826B (en) A kind of intelligence Ankang service system and the reliable detection method of alarm
US10670621B2 (en) Fall prevention
US9681827B2 (en) Systems, methods, applications for smart sensing, motion activity monitoring, and motion activity pattern recognition
US10561341B2 (en) Device packaging for an impact detection device
US10869615B2 (en) Wearable electronic device and emotion monitoring method
US8784271B2 (en) Biometric monitoring device with contextually-or environmentally-dependent display
CN104272711B (en) Methods and apparatuses for operating a display in an electronic device
CN104522945B (en) A kind of smart motion health care bracelet
CN105981035B (en) Manage the display of privacy information
US9640057B1 (en) Personal fall detection system and method
JP5438725B2 (en) System and method for real-time physiological monitoring
CN104090659B (en) Operating pointer based on eye image and Eye-controlling focus indicates control device
JP2017512619A (en) Multifunctional smart mobility aid and method of use
KR20170066702A (en) Dynamic sampling
CN101950464B (en) Method and system for fall monitoring and warning
US10104458B2 (en) Enhanced biometric control systems for detection of emergency events system and method
Seeger et al. myHealthAssistant: a phone-based body sensor network that captures the wearer's exercises throughout the day.
CN104436597B (en) Exercise support device and exercise support method
Cao et al. E-FallD: A fall detection system using android-based smartphone
CN106662449A (en) Techniques for determining movements based on sensor measurements from a plurality of mobile devices co-located with a person
Dai et al. PerFallD: A pervasive fall detection system using mobile phones
CN204618208U (en) The wearable system of a kind of Real-Time Monitoring health, active state and environment
CN108369765A (en) Activities monitoring system and method for accident detection and response

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant