EP3108462A1

EP3108462A1 - Technology for detecting a fall of a person

Info

Publication number: EP3108462A1
Application number: EP15700460.7A
Authority: EP
Inventors: Thomas VON CHOSSY; Henning MARCHFELD
Original assignee: Individual
Current assignee: MARCHFELD, HENNING; VON CHOSSY, THOMAS
Priority date: 2014-02-17
Filing date: 2015-01-15
Publication date: 2016-12-28
Anticipated expiration: 2035-01-15
Also published as: EP3108462B1; WO2015121018A1; US9898914B2; US20170061762A1; DE102014002124A1

Abstract

The invention relates to a technology for detecting a fall of a person. A corresponding device (100) comprises an interface (102), which is designed to capture a time-dependent air-pressure signal (600) determined by means of at least one air-pressure sensor (104, 106) worn on the body of the person. The device (100) also comprises an evaluating unit (108), which is designed to determine a fall height (λ) with respect to an evaluation time (t_e) by means of a window-based signal analysis of the time-dependent air-pressure signal. The window-based signal analysis comprises a first time window (702) before the evaluation time and a second time window (704) after the evaluation time, which second time window does not overlap with the first time window. The fall height is determined from a difference between a first filter value calculated on the basis of the time-dependent air-pressure signal in the first time window and a second filter value calculated on the basis of the time-dependent air-pressure signal in the second time window.

Description

Technique for detecting a person falling Technical area

The present invention relates to a technique for detecting a lintel. In particular, devices and methods for fall detection by means of at least one air pressure sensor are apparent.

Technical background

Groups of persons with an increased risk of falls, for example craftsmen on treads or ladders, firefighters, forestry workers, miners, hospital and care patients, people suffering from epilepsy, or

Senior citizens living alone can receive help in the event of a fall if they are monitored to see if they have fallen. The fall can injure a person so much that an exemption from an emergency situation or the discontinuation of an emergency call from their own resources is no longer possible. An injury-free fall can also be an indication of an acute need for treatment.

for example, unconsciousness.

Document EP 1 642 248 A1 describes a system for detecting falls. A system component with pressure sensor and motion sensor is worn on the wrist. A second pressure sensor for measuring the pressure of the air in the home is arranged in a base station which stores the air pressure data of the portable

System component receives.

The document DE 10 2008 049 750 AI describes a method for detecting a fall of a person based on at least one air pressure sensor. Alternative to

Adjustment of the air pressure sensor worn on the person's body with a

Comparison pressure sensor is called the possibility of fall detection by evaluating the measured value of the air pressure sensor as a function of time. The expected time course of the measured values defines a "pattern". By conventional

Pattern recognition method, measured values of the air pressure sensor are evaluated. In conventional methods, falls with a not preprogrammed pattern pattern can go undetected. Conversely, weather conditions have a considerable influence on the temporal course of the local air pressure, independent of any movement of the person. Also, opening windows or doors can cause numerous pressure gradients that coincidentally resemble one another

preprogrammed pattern trigger a false alarm.

Summary of the Invention Thus, it is an object of the present invention to provide a technique for reliable fall detection independent of details of the time course of a fall.

In one aspect, an apparatus for detecting a person's fall is provided. The device comprises an interface which is designed to detect a time-dependent air pressure signal determined by means of at least one air pressure sensor carried on the body of the person, and an evaluation unit which is designed to do so by means of a window-based signal analysis of the

time-dependent air pressure signal a lintel height with respect to a

Wherein the window-based signal analysis comprises a first time window before the evaluation time and a second time window not overlapping the first time window after the evaluation time, and wherein the fall height of a difference of a calculated based on the time-dependent air pressure signal in the first time window first filter value and a Basis of the time-dependent air pressure signal in the second time window calculated second filter value is determined.

Embodiments of the device can by juxtaposing the

Filter values from the disjoint first and second time windows numerically determine the lintel height, and optionally an error of estimation of the lintel height, of a lintel occurring between the time windows. The determination can be made, for example, without resorting to specific patterns of the fall. The particular

Falling height may be the basis for an alarm signal, possibly in conjunction with the estimation error. The estimation error may be a standard deviation.

The air pressure signal may include a difference between the at least one body-worn air pressure sensor and at least one reference air pressure sensor. The reference air pressure sensor may be carried by the person at a body position other than the supported air pressure sensor. Alternatively, or in combination, the interface may detect signals from one or more stationary reference air pressure sensors. The interface can do this one

Provide radio interface to the stationary air pressure sensor.

The air pressure signal can be sampled periodically. The first time window may be separated in time from the second time window by at least one sampling period. A time interval between the first time window and the second time window may be shorter than a total duration of the first time window and the second time window.

The determination of the lintel height, and possibly of the estimation error, can disregard the time-dependent air pressure signal between the first time window and the second time window.

There may be a third time window between the first time window and the second time window. The fall can occur completely in the third time window. The first time window may include a person's movement before the fall. The second time window may include a rest position of the person after the fall.

The first filter value and / or the second filter value may be an average value of the time-dependent air pressure signal. The mean may include an unweighted arithmetic mean, a weighted arithmetic mean and / or a median.

The calculation of the first filter value and / or the second filter value can compensate for a time dependence of the time-dependent air pressure signal. The

Air pressure signal can be adjusted for a trend, such as a weather trend. The course of the trend may be longer than the fall or the temporal

Distance between the first time window and the second time window. The

Compensation may be upstream of the determination of the lintel height or the compensation may take place in the course of the determination, for example in the calculation of the first filter value and / or the second filter value.

The time dependence for the first time window, Fi (t), and for the second time window, F ₂ (t), can each be compensated independently of each other. In the first Time window, the first time dependence, Fi (t), be fitted to the time-dependent air pressure signal. In the second time window, the second

Time dependence, F ₂ (t) is fitted to the time-dependent air pressure signal, wherein the lintel height of the difference of the first time dependence, F ^ t ^ _x ), the end time t _{1 / max of} the first time window and the second time dependence, F ₂ (t _{2 / min} ) to

Start time t _{2 / min of} the second time window is determined.

The first time dependence, Fi (t), may be a zero, first, or second degree polynomial in time. The second time dependence, F ₂ (t), may be a zero, first, or second degree polynomial in time.

Alternatively, the same time dependence, F (t), can be compensated for the first time window and for the second time window. In the first time window, the

Time dependence, F (t), with a first time independent offset value Q as

Fitparameter be fit to the time-dependent air pressure signal. In the second

Time window, the same time dependence, F (t), is fitted with a second time-independent offset value C ₂ as Fit parameters to the time-dependent air pressure signal. The lintel height can be determined from the difference of the first offset value and the second offset value.

The time dependence, F (t), may be a linear and / or quadratic function of time. The time dependence, F (t), may correspond to a proportion in the time-dependent air pressure signal (caused, for example, by the weather conditions of the air pressure). The time dependence, F (t), can be independent of a fall. A uniform time dependence, F (t), can be present before and after falling in atmospheric pressure. Of the

Definition domain of the function F (t) may include the first time window and the second time window.

The time-dependent air pressure signal may comprise a sequence of measured values in each case in association with a measuring time. The lintel height can with respect to a

Variety of evaluation times are determined. The evaluation time, the first time window and / or the second time window can be shifted step by step by one measurement time of the sequence at later times. For each of the evaluation times, the lintel height, and if necessary the estimation error of the lintel height, can be determined. The first time window may be completely in time before the second time window. The end time of the first time window may be before the start time of the second time window by a predetermined time interval for all evaluation steps. The first time window can only include measurement times of the sequence which lie before the times of the second time window.

An initial time of the second time window and / or an end time of the first time window may be shifted by one time of the sequence. The evaluation time may be equal to the end time of the first time window or between the end time of the first time window and the start time of the second time window. The length of the second time window may be the same during the signal analysis at each of the evaluation times.

For each of the evaluation times can be checked whether the specific

Fall height exceeds a first significance level, e.g. according to A> Si. The first significance measure may be twice or three times the estimation error.

For each of the evaluation times it can be checked whether the difference between the determined lintel height and a height threshold exceeds a second degree of significance, eg, according to λ - A _S w> s ₂ . The second significance measure may be twice or three times the estimation error.

An alarm condition can be set if the second significance measure is exceeded.

The device may further include a trip unit configured to output an alarm signal. The alarm signal can be output if, in the alarm state, a fit of the time-dependent air pressure signal in the second time window exceeds a predetermined quality measure. After an audible pre-alarm, and optionally after a user acknowledgment has failed, the trip unit will transmit

Alarm signal, e.g. by calling a preprogrammed emergency number via landline or mobile network.

For each evaluation time, the first time window can be extended step by step to earlier measurement times. The extension may be terminated upon reaching a maximum length of the first time window. The length of a Time window can be determined by the number of measurement times in the sequence in the relevant time window.

Alternatively or in combination, the incremental extension may be terminated if the second significance measure is exceeded. In this case, the alarm status can also be set.

Alternatively or in combination, for each of the evaluation times, and optionally for each step of the window extension, it can be checked whether the difference between the altitude threshold and the determined lintel exceeds a third degree of significance, eg according to λ - A _S w <-S3. The third significance measure may be five times the estimation error. The stepwise extension can be terminated if the third significance measure is exceeded. After completing the extension, the signal analysis can be continued at an increased evaluation time (eg at one measurement time). The incremental increase of the evaluation time may be implemented as an outer loop. The incremental extension of the first time window may be implemented as an inner loop.

In another aspect, a method for detecting a fall of a person is provided. The method comprises a step of detecting a time dependent air pressure signal determined by at least one air pressure sensor carried on the person's body, and a step of determining a fall height with respect to an evaluation time point by means of a window based signal analysis of the time dependent air pressure signal, wherein the window based signal analysis includes a first time window before the evaluation time and wherein the lintel height is determined from a difference of a first filter value calculated based on the time-dependent air pressure signal in the first time window and a second filter value calculated based on the time-dependent air pressure signal in the second time window.

The method may further include any functional feature of the device in a corresponding method step. Brief description of the figures

For a complete understanding of the technique described here, see

The following exemplary embodiments are described with reference to the accompanying drawings, wherein:

Fig. 1 shows schematically a first embodiment of a

A device for detecting a fall with at least one portable integrated air pressure sensor;

Fig. 2 shows schematically a second embodiment of a

A device for detecting a fall with at least one portable external air pressure sensor; Fig. 3 shows schematically a third embodiment of a

A device for detecting a fall with at least one stationary reference air pressure sensor;

Fig. 4 shows schematically a fourth embodiment of a

A device for detecting a fall with a plurality of mobile air pressure sensors;

FIG. 5 shows a flowchart of a method for detecting a fall which is used in the devices of FIGS. 1 to 4 can be implemented;

Figs. 6A and 6B show exemplary waveforms of an air pressure signal;

Fig. 7 shows schematically a first time window and a second one

Time window for a window-based signal analysis according to the method of FIG. 5;

FIGS. 8A to 8C show schematic time dependencies for the uniform

Compensation of a time dependence of the air pressure signal in the first time window and in the second time window of the

Fig. 7; and FIGS. 9A and 9B show schematic time dependencies for the independent ones

Compensation of a time dependence of the air pressure signal in the first time window and in the second time window of FIG. 7; and

Fig. 10 shows a system for detecting a fall with a plurality of spatially distributed stationary reference air pressure sensors.

Detailed description

FIG. 1 shows a schematic block diagram of a device for detecting a person's fall, generally designated by the reference numeral 100. The first exemplary embodiment shown in FIG. 1 comprises an interface 102 within the device 100, to which an air pressure sensor 104 is connected. The device 100 further comprises an evaluation unit 108, which analyzes the time-dependency of the air pressure signals detected by the interface.

FIG. 2 shows a schematic block diagram of a second exemplary embodiment of the device 100, in which the air pressure sensor 104 is not part of the device 100. The interface 102 comprises a plug connection for the electrical and mechanical reception of the air pressure sensor 104.

A third embodiment of the device 100 is shown in FIG. The

Interface 102 detects air pressure signals of both an integrated air pressure sensor 104 and a reference air pressure sensor 106. The evaluation unit 108 and the

Air pressure sensor 104 are galvanically coupled. The evaluation unit 108

communicates with the reference air pressure sensor 106 via a radio link.

At least the air pressure sensor 104 is carried on the body of a person in all three embodiments of the device 100. The air pressure sensor 104 allows a barometric altitude measurement taking advantage of the gravitational acceleration relationship between static

Air pressure and altitude. In the simplified case of a temperature uniform over the relevant height of a possible fall, there is an exponential relationship between absolute pressure and height, so that between small and large

Pressure differences due to small height differences a linear relationship can be assumed. For example, from a pressure difference Δρ on a lintel height λ = -ΔρΉ / ρ _{0 are} closed with the so-called scale height H = RT / (g M). Here, R is the gas constant of medium molecular weight dry air M at an absolute temperature T and a pressure p ₀ at the selected zero level (eg, the level at which the fall occurred).

While the in Figs. 1 to 3 of the apparatus 100 employs a supported air pressure sensor 104 for detecting an air pressure change Δρ before and after the fall, measured values of a plurality of carried

Barometric pressure sensors 104 are detected and, for example, averaged to reduce measurement errors or reduce anisotropic pressure factors such as back pressure.

In the case of the third embodiment of FIG. 3, the detected air pressure signal is based on a difference between the air pressure measured by the supported air pressure sensor 104 and a reference air pressure measured by the reference air pressure sensor 106. The reference air pressure sensor 106 may be stationary so that the difference in the air pressures measured by the air pressure sensors 104 and 106 is large

Air pressure changes eliminated. The spatial distance between the supported air pressure sensor 104 and the stationary reference air pressure sensor 106 meanwhile determines a length scale of the eliminated large-scale air pressure fluctuations.

4 shows one embodiment of a fall detection system 400 having a supported air pressure sensor 104 and a stationary one

Reference air pressure sensor 106-1. The system 400 further includes a trip unit 402. The carried air pressure sensor 104, the reference air pressure sensor 106-1, and the trip unit 402 are in radio communication. The radio connection can be done in pairs. Alternatively, the reference air pressure sensor 106-1, together with a relay station, may form a unit that provides a radio connection between the trip unit 402 and the supported air pressure sensor 104. The apparatus 100 may be spatially associated with each of the components 104,

106-1 or 402 may be arranged. An integration of the device 100 in the

Trip unit 402 can advantageously be operated via a power grid, so that the transit times of the battery-powered components 104 and / or 106-1 are extended. For uninterruptible power supply can also

Trip unit 402 include an additional battery supply. Alternatively or in addition to the reference air pressure sensor 106-1, in an alternative embodiment of the system 400, a reference air pressure sensor 106-2 is disposed at a low body position. The low - lying body position is chosen so that in a fall on a flat surface no much deeper position of the

Reference air pressure sensor 106-2 is reached. In a fall on an im

A substantially inclined surface, such as a stairway, a rest position of the reference air pressure sensor 106-2 on the worn

Be air pressure sensor 104. The trigger unit 402 is in an institutional application (for example in a hospital or retirement home) to a home emergency call system 404

connected. In a private application, the trip unit 402 is connected to a landline 404. Alternatively or for redundancy is the

Trigger unit 402 to a mobile network or a local radio network

(For example, a W-LAN according to the IEEE 802.11 standard family) connected.

As an alternative or in addition to the reference air pressure sensors 106-1 and 106-2, a reference air pressure signal of the local atmospheric pressure (for example in real time) can also be detected by the interface 102 via the landline connection, the mobile radio network or the local radio network.

5 shows a method 500 for detecting a fall of a person. In a step 502 of the method 500, a time-dependent air pressure signal determined by means of at least one air pressure sensor carried on the person's body is detected. The detected time-dependent air pressure signal may be a difference signal between a supported air pressure sensor and a reference air pressure sensor.

In a step 504 of the method 500, by means of a window-based

Signal analysis of the time-dependent air pressure signal determines a lintel height with respect to an evaluation time. The window-based signal analysis comprises a first time window before the evaluation time and a second time window not overlapping the first time window after the evaluation time. The lintel height is determined from a difference of one based on the time-dependent one

Air pressure signal exclusively calculated in the first time window first filter value and based on the time-dependent air pressure signal exclusively in the second time window calculated second filter value. The method 500 is performed by the device 100. The time-dependent air pressure signal detected thereby may be provided by each of the supported air pressure sensors 104 or may be a difference signal based on the supported air pressure sensor 104 and the reference air pressure sensor 106.

The Fign. FIGS. 6A and 6B schematically show two examples of the time-dependent one

Air pressure signal. Due to the assumed connection between

Air pressure change and altitude change, the time-dependent barometric pressure signal 600 is plotted in height units on the vertical axis. In principle, the time-dependent air pressure signal 600 can be divided into a movement phase 602, a region 604 of the actual fall process and a rest phase 606. The window-based signal analysis by the evaluation unit 108 according to the step 504 calculates the first filter value based on the motion phase and the second filter value based on the quiet phase 606 by masking the lobe area 604 when the referenced evaluation time coincides with the camber time.

In the first example of the time-dependent air pressure signal 600 shown in FIG. 6A, in the movement phase 602, a person rises from a lying position on a bed surface to a sitting position on a bed edge and loses

Trying to get out of bed the balance or even the consciousness, for example due to a temporary lack of blood flow to the brain. The camber movement ends in the resting phase 606 on a floor surface.

In the second example of the time-dependent air pressure signal 600 shown in FIG. 6B, the body slides from lying or sitting to a lower level. The

Sliding motion belongs to the fall area 604 and the end position on the lower level to the resting phase 606 of the time-dependent air pressure signal 600.

Typical seat heights are approximately 0.50 m, for example on a bed or an office chair and 0.44 m, for example, on a chair or kitchen chair. at

Attachment of the supported air pressure sensor 104 at chest level increases a fall height detected by the supported air pressure sensor 104 as a function of body size. 7 schematically shows the window-based signal analysis 504 of the time-dependent air pressure signal 600. A total time window w of the signal analysis 504 comprises the first time window 702 and the second time window 704. The time duration w can for example 120 seconds. The time windows 702 and 704 are sliding windows. The signal analysis is carried out cyclically or in packets.

The first time window 702 comprises m measuring times 706 of the detected

Air pressure signal. The second time window 704 comprises n measurement times of the detected air pressure signal. The measuring times can be at intervals of, for example, 1 second.

The first time window 702 and the second time window 704 are separated by an interval 708 with the maximum assumed camber duration τ. For example, the evaluation time t _e may be the nominal time of the fall event in the middle of

Intervals 708 are defined.

In addition to the measurement times N = m + n on which the signal analysis is based, the u measuring times in the interval 708 are not taken into account in the window-based signal analysis 504 with respect to the evaluation time te. In a continuous evaluation (which is also referred to as online evaluation), the current time stamp t _a with the upper limit t ₂ , _m ax of the second time window 704

coincide. Alternatively, there may be a delay time 710 between the second time interval 704 and the current time stamp in the signal analysis, for example due to a time for data preparation and / or

Data transfer.

In one exemplary embodiment of the signal analysis 504, there is a minimum time interval between the evaluation time t _e and the current time stamp t _a

specified as an additional condition for triggering an alarm

Minimum rest time to determine with a substantially constant air pressure signal 600, for example, a gradual increase in the first

Time window 702 and / or the second time window 704. The initial use of shorter time windows reduces the analysis effort and thereby the requirements for memory, computing power and power consumption.

The signal analysis 504 may be performed using different polynomial degrees in the first time window 702 and / or in the second time window 704. The

Signal analysis 504 may be advanced to higher polynomial degrees starting with a low degree of polynomial, eg, zero degree, eg, until a significant fall height is detected or until a maximum degree of polynomial has been reached. The Window size may depend on the degree of polynomial, for example, the window size may increase with the degree of polynomial.

Alternatively or additionally, the window size can be determined by a predetermined

Accuracy be determined according to the estimation error. For example, In the course of a fit, a covariance matrix can be determined. The estimation error can be calculated based on the covariance matrix and a theoretical accuracy value of the sensor. The latter estimation error is also called a theoretical estimation error. Alternatively or in combination, the estimation error may be based on the

Covariance matrix and the sum of the square deviations of the fit, for example as the product of a diagonal element of the covariance matrix and the root of the sum of the quadratic deviations. In this case, the sum of the square deviations of the fit can be normalized by (e.g., divided by) the number of measurement times less the number of degrees of freedom of the fit. The latter leads to an estimation error of the current signal analysis 504.

The air pressure at a certain place is not constant over time. It is mainly influenced by meteorological balancing processes. According to that

The relationship between air pressure and altitude can be this ongoing

Changes in air pressure to apparent altitude changes of up to 30 to

40 m per hour. Furthermore, the air pressure at a fixed location is also

Subject to periodic changes in the course of the day, for example, due to the daily cycle of the air temperature and by a natural vibration of the

Earth's atmosphere. The periodic changes are apparent

Height changes at a rate of about 2 to 5 meters per hour.

Since the weather situation generally overlaps the time-dependent air pressure signal 600 to be measured in the case of a barometric altitude measurement, a time dependency of the air pressure signal 600 in the first time window 702 and / or in the second time window 704 is compensated.

The Fign. FIGS. 8A to 8C schematically show a time dependence 800, the portion 802 of which in the first time window 702 and the portion 804 in the second time window 704 of the detected time-dependent air pressure signal 600 by methods of FIGS

Adjustment calculation to be adjusted. As a result, in the course of the determination of the first and second filter values, the weather trend in the time-dependent air pressure signal 600 can be compensated. The compensation is a uniform Time dependence F (t) for both time windows 702 and 704, which differ only by an offset value ÄC corresponding to the estimated lintel height λ. For example, the time dependence 800 may include polynomial or Fourier evolution.

The function to be adopted, M (t), is adapted to the detected air pressure signal 600 taking into account the m measurement times in the first time window 702 and the n measurement times in the second time window 704: in the first time window 702 in the second time window 704

From the pressure jump ÄC = C ₂ - Q is closed to the lintel height λ. The function M (t) can also be expressed as the sum of the uniform time dependence F (t) and a

Jump function 6 (tt _e ) are represented with stage at time t _e .

The time dependence 800, which is uniform for both time windows 702 and 704, is a linear and / or quadratic function of the time t:

The case of FIG. 8A corresponds to the parameter setting ai = a ₂ = 0, ie no compensation of a time dependence. Only the offset values Q and C ₂ are

Fit parameters.

Fig. 8B shows schematically the compensation of an exclusively linear

Time dependence 800, ie a ₂ = 0 is not a fit parameter. Only the coefficient ai and the offset values Q and C ₂ (or their difference) are fit parameters.

FIG. 8C schematically shows a compensation of the time dependence 800 up to the second order in time. The compensation calculation comprises the fit parameters C _lr C ₂ , ai and a ₂ or Ci, CA = C ₂ -Ci, ai and a ₂ . The Fign. 9A and 9B show a schematic time dependency 900, which are adapted independently of each other in the first time window 702 and in the second time window 704 to the detected time-dependent air pressure signal 600. The first time dependence 902 is adjusted on the basis of the m measurement times of the first time window 702, independently of the adaptation of a second time dependence 904 on the basis of the n measurement times in the second time interval 704.

Based on the adapted to the detected air pressure signal 600 function

in the first time window 702 in the second time window 704

is closed by the pressure jump F (t _{2 / min} ) - F (t _{1 / max} ) on the lintel height λ.

The time dependencies 902 and 904 for the time windows 702 and 704 are respectively linear and / or quadratic functions of the time t:

Fi (t) = a _"0 + a '+ aW it -r.2 ₂ fo _r j = i; 2.

The compensation calculation comprises the fit parameters a ^ o, a ^, a ^ ⁾ ₂ , a ^ o, a ^{2 and a ⁽²⁾ ₂ .

The compensation shown in Fig. 9A eliminates one linear in time

Time dependence 900 in the two time windows 702 and 704, ie, the quadratic coefficients a ^ ₂ = 0 are not fit parameters.

The compensation shown in Figure 9B eliminates only a linear time dependence 904 in the second time window 704, i. it is additionally set = 0.

Together with the estimated value for the lintel height λ obtained from the compensation calculation, the compensation calculation yields a theoretical or estimated estimation error σ _{λ of} the lintel height λ. This signal analysis 504 will be for each

Evaluation time t _e executed. In a first embodiment variant, the estimation error σ _λ is calculated from the root of the sum of the quadratic deviations between the matched time dependence M (t) and the time-dependent air pressure signal. The sum comprises the N = m + n quadratic deviations in the first time window 702 and in the second time window 704.

In a second embodiment, the estimation error according to calculated from the standard deviation σ of the detected air pressure signal.

Optionally, a first significance measure τ _{λ is} determined from the estimated lintel height λ and the estimated error σ _λ :

For example, if Zi is not greater than 2, the estimated value does not reach the significance level α = 4.55%. The signal analysis then becomes this

Evaluation time ended without changing an alarm state.

If, on the other hand, the estimated lintel height λ _satisfies a second significance _measure z ₂ = (λ-λ _3νν ) / σ _λ , with z ₂ > 3 for a threshold value A _S w, the minimum _{lintel height} A _S w was exceeded with sufficient significance and an alarm state is set ,

In an extended embodiment of signal analysis 504, not only a lintel height and associated estimation error with respect to each evaluation time t _{e is} determined. In addition, for a given evaluation time starting from a minimum window size (which includes, for example, m _min = 2 measurement times), the window size of the first time window 702 is incrementally increased by adding earlier measurement times 706.

For each enlarged first time window 702, the lintel height λ and the associated estimation error σ _{λ are} determined. The gradual enlargement of the first Time window 702 takes place until a maximum window size is reached

(For example, m _max = 320), optionally until the second significance measure z ₂ a

Lower limit (for example, according to z ₂ <-5), or until a fall is detected (for example by z ₂ > 3).

Alternatively or in addition to the incremental enlargement of the first time window 702, in the alarm state, the second time window 704 can be shifted in time independently of the first time window 702 as a sliding second time window 704. For example, when an alarm condition occurs, an alarm signal is not immediately output by the device 100 or triggering unit 402. In the alarm state, the second time window 704 is first shifted to larger measuring times 706. Optionally, the end time t _{1 / max of} the first time window 702 may be shifted to larger measurement times 706 if the shift results in an increase in the altitude value corresponding to the end time of the first time window 702 (F (t _{1 / max} )). The end time t _{1 / max of} the first time window

702 can be set to the maximum altitude value. The starting time t _{1 / min} , for example, remains unchanged.

For each shifted second time window, a quality Q of the fit of the

Time dependence 804 or 904 in the second time window 704 determined. Alternatively, a chow test can be performed in which a first

Embodiment variant, the division into the first time window 702 and the second time window 704 is maintained, or in a second embodiment only the second time window 704 is examined with a suitable breakdown on a so-called structural break out.

If the quality Q satisfies a threshold value Q _S w when the predetermined value is reached

Minimum time interval between the current time stamp t _a and the evaluation time t _e , an alarm signal is output.

The alarm can be triggered, for example within 60 to 90 seconds, after the actual fall at time t _e . This can be a

alarm-preventing reaction of the person. The trip unit 402 may allow a pre-alarm with abort capability prior to automatic emergency call forwarding, for example, to preclude a deliberate lie down. Signal analysis 504 using uniform time dependence 800 is also referred to as template analysis (TA) and abbreviated to TAO, TAI and TA2 (for the cases shown in Figures 8A, 8B and 8C, respectively), indicating the degree of polynomial. The signal analysis 504 by means of two time dependencies 902 and 904 is also referred to as Gap Analysis (GA) and with the addition of the corresponding

Polynomial degree abbreviated to GA1 (in the case of FIG. 9A) and GA-0-1 (in the case of FIG. 9B). To achieve a given theoretical estimation error (theoretical

Accuracy), for example, of 2.5 cm, each of the first time window 702 and the second time window 704 may comprise 32 measurement times for the signal analysis 504 according to TAO, respectively 128 measurement times for the signal analysis 504 according to TAI and TA2, and 227 measurement times for each Signal analysis 504 according to TA3. The minimum time interval between the evaluation time t _e and the current time stamp t _a can be evaluated according to the maximum

Polynomial degrees are chosen.

Extensive series of measurements provided reliable fall detection

Implementation of the signal analyzes 504 marked "x" below:

In the column headers, the abbreviation of the implementation of the signal analysis 504 is given. An air pressure sensor 104 was used with the process accuracy θχ specified in centimeters. The reliably detected fall height and the

Threshold values are given in the table above on the left.

In an extended embodiment of the device 100, a temperature sensor carried on the body is further provided. The output of an alarm signal may be coupled to the additional condition that the temperature sensor has a Recorded temperature drop. The temperature sensor may be thermally decoupled from the body, for example by a heat insulation layer.

10 shows an enhanced system 1000 including at least one person fall detection apparatus 100 and a plurality of reference air pressure sensors. The system 1000 extends over several floors accessible through a staircase 1002 and / or elevator. There is one on each floor

Reference air pressure sensor 106-1, 106-2 and 106-3 arranged. Depending on a signal strength of the air pressure signal detected by the carried air pressure sensor 104-1 or 104-2, the device 100 results in a spatially associated one

Tripping unit of the trip units 402-1 and 402-2, the method 500 from.

For example, in the trip unit 402-1, the device 100 executes the method 500 based on the spatially associated carried air pressure sensor 104-1 in consideration of the reference air pressure sensor 106-1. If a person carrying the air pressure sensor 104-1 rises from the first floor to the second floor

Floor, the device 100 integrated in the trip unit 402-2 undertakes the execution of the method 500 based on the supported air pressure sensor 104-1 and taking into account the reference air pressure sensor 106-2. At least in individual embodiments, the technique described for detecting a fall allows reliable fall detection without specifying a particular fall pattern. External influences, such as large-scale pressure fluctuations due to weather changes, can be compensated by a

appropriate time dependence and / or taking into account a

Reference air pressure sensor can be eliminated by the signal analysis. Thus, false alarms can be prevented at low thresholds for reliable

Detection of falls.

Claims

Claims 1. A device (100) for detecting a person's fall, comprising:

an interface (102) adapted to detect a time-dependent air pressure signal (600) determined by at least one air pressure sensor (104, 104, 106; 104, 106-1, 106-2) carried on the body of the person, and

an evaluation unit (108), which is designed by means of a

Window-based signal analysis of the time-dependent air pressure signal to determine a lintel height (λ) with respect to an evaluation time point (t _e ), wherein the

window-based signal analysis a first time window (702) before

And wherein the lintel height is calculated from a difference of a first filter value calculated on the basis of the time-dependent air pressure signal in the first time window and a second calculated on the basis of the time-dependent air pressure signal in the second time window

Filter value is determined.

The apparatus of claim 1, wherein the air pressure signal (600) is sampled periodically and the first time window (702) is time separated from the second time window (704) by at least one sampling period.

3. Apparatus according to claim 1 or 2, wherein the determination of the lintel height disregards the time-dependent air pressure signal (600) between the first time window (702) and the second time window (704).

4. Device according to one of claims 1 to 3, wherein the first filter value and / or the second filter value is an average value of the time-dependent air pressure signal.

The apparatus of any one of claims 1 to 4, wherein the calculation of the first filter value and / or the second filter value compensates for a time dependence (800; 900) of the time-dependent air pressure signal.

The apparatus of claim 5, wherein the time dependence (900) for the first time window (702) and for the second time window (704) respectively from each other

is compensated independently.

7. Apparatus according to claim 5 or 6, wherein in the first time window (702) a first time dependence (902; F ^ t)) is fitted to the time-dependent air pressure signal and in the second time window a second time dependence (904; F ₂ (t)) the time-dependent air pressure signal is fitted, wherein the lintel height of the difference of the first time dependence (F ^ t ^ x)) at the end time of the first time window and the second time dependence (F ₂ (t _{2 / min} )) at the start time of the second

Time window is determined.

The apparatus of claim 5, wherein the same time dependence (800) is compensated for the first time window (702) and the second time window (704).

9. Apparatus according to claim 5 or 8, wherein in the first time window the

Time dependence (802; F (t)) is fitted with a first time-independent offset value (CO as fit parameter to the time-dependent air pressure signal, and in the second

Time window (804; F (t)) is fitted with a second time independent offset value (C ₂ ) as fit parameter to the time dependent air pressure signal, the fall height being determined from the difference of the first offset value and the second offset value.

10. Device according to one of claims 1 to 9, wherein the time-dependent air pressure signal (600) comprises a sequence of values each associated with a measurement time (706), and wherein the lintel height (λ) is determined with respect to a plurality of evaluation times by the Evaluation time (t _e ), the first time window (702) and the second time window (704) by one step

Measuring time of the sequence are moved to later times and for each of the evaluation times the lintel height is determined.

11. The device according to claim 10, wherein it is checked for each of the evaluation times whether the determined lintel height is a first significance measure (z0

exceeds.

12. The apparatus of claim 10 or 11, wherein for each of

Evaluation time is checked whether the difference between the specific lintel height and a Höhenschwel Iwert a second degree of significance (z ₂ )

exceeds.

13. The apparatus of claim 12, wherein an alarm condition is set if the second significance measure (z ₂ ) is exceeded.

14. The apparatus of claim 13, further comprising a trip unit (402) configured to output an alarm signal if, in the alarm state, a fit of the time-dependent air pressure signal in the second time window is a predetermined one

Quality measure (Q) exceeds.

15. Device according to one of claims 12 to 14, wherein for each

lo evaluation time (t _e ) the first time window (702) gradually to earlier

Measuring points (706) is extended until a maximum length of the first time window (702) is reached until the second significance measure (z ₂ ) is exceeded or until the determined lintel height of the height threshold by a third

Significance (z ₃ ) falls below.

15

16. Device according to one of claims 1 to 15, wherein the device further comprises a body worn and thermally decoupled from the body

Includes temperature sensor, and wherein an alarm signal is issued only if the temperature sensor detects a temperature drop.

20