Leg shaking reminding method and system and seat

Abstract

The invention relates to the technical field of intelligent furniture, and discloses a leg shaking reminding method and system and a seat, and the method comprises the steps: obtaining piezoelectric detection data measured by a deformation detection assembly in real time, extracting a frequency spectrum signal through a preset information preprocessing strategy, and converting the frequency spectrum signal into a power spectrum; accurately extracting leg shaking power from the power spectrum by using an action rule screening strategy; and finally, generating reminding information based on the leg shaking power. According to the invention, non-contact monitoring without a wearable device is realized, and the use convenience is improved; the requirements on hardware computing power and cost are reduced through a simplified signal processing algorithm, and good economical efficiency is achieved; meanwhile, interference of breathing, heart rate and the like can be effectively filtered, the accuracy of leg shaking behavior recognition is remarkably improved, and a reliable technical basis is provided for achieving intelligent and differentiated reminding.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the technical field of smart furniture, in particular to a leg shaking reminding method and system and a seat. BACKGROUND

[0002] Leg shaking is a common habitual body movement, which often occurs unconsciously when sitting for a long time, concentrating or being nervous. On the one hand, frequent or excessive leg shaking is often regarded as an inappropriate behavior in social occasions, especially for minors, which is not conducive to the formation of good behavior habits and should be avoided as much as possible. On the other hand, the significance of leg shaking in the health aspect is more complex. For the general population, leg shaking may be a manifestation of psychological states such as anxiety and distraction, and even related to nervous system diseases such as restless leg syndrome. However, for the elderly who sit for a long time, moderate leg shaking is helpful to promote blood circulation and prevent thrombosis, and has certain health value. It is this completely opposite significance in different situations that makes it a key to realize effective health management and behavior correction to accurately monitor and intelligently remind leg shaking behavior.

[0003] A variety of leg shaking monitoring schemes have been proposed in the prior art, which can be mainly divided into three categories: wearable type, pressure foot pad type and high-precision sensor type. The wearable type scheme requires the user to wear special accessories, which not only has a cumbersome wearing process, but also small-sized accessories are easy to fall off, forget or lose, and cannot realize long-term stable monitoring. The pressure foot pad type sets the sensing area as a foot pad, and the sensing range is completely limited by the size of the foot pad, which is only suitable for fixed scenes where the user's feet completely contact the foot pad, and cannot adapt to scenes without foot pads or with a large range of user's feet movement, which has strong scene limitation. The high-precision sensor scheme (such as millimeter wave radar) has high accuracy, but the cost is high and the power consumption is large, which is difficult to popularize.

[0004] Therefore, how to design a leg shaking monitoring and reminding scheme without additional wearing, strong scene adaptability and low cost has become a technical problem to be solved in the current technical field. SUMMARY

[0005] In order to overcome the problems of poor convenience, limited applicable scenes and high cost of existing leg shaking monitoring technology, the present application provides a leg shaking reminding method, system and seat.

[0006] According to a first aspect of the embodiments of the present disclosure, the present application provides a leg shaking reminding method, which adopts the following technical scheme:

[0007] A leg shaking reminding method, characterized in that it comprises:

[0008] obtaining piezoelectric detection data measured by the deformation detection component in real time;

[0009] extracting a frequency spectrum signal from the piezoelectric detection data according to a preset information preprocessing strategy;

[0010] transforming the frequency spectrum signal into a power spectrum;

[0011] extracting a leg shaking power corresponding to a leg shaking action in the power spectrum according to a preset action rule screening strategy;

[0012] generating a reminder information based on the leg shaking power.

[0013] Optionally, the extracting a frequency spectrum signal from the piezoelectric detection data according to a preset information preprocessing strategy comprises:

[0014] performing a fast Fourier transform on the piezoelectric detection data to obtain frequency spectrum data containing imaginary numbers;

[0015] performing a modulo operation on the frequency spectrum data to obtain the frequency spectrum signal.

[0016] Optionally, the performing a fast Fourier transform on the piezoelectric detection data to obtain frequency spectrum data containing imaginary numbers comprises:

[0017] decomposing a function called in the FFT into trigonometric functions = ; ;

[0018] dividing by an interval of T, and calculating a corresponding sine value ;

[0019] calculating a more fine-grained corresponding sine value based on a lookup table method and a polynomial approximation method;

[0020] based on the odd-even symmetry of the trigonometric functions, compressing the operations of cosine and sine in the trigonometric functions in the interval of to calculate; wherein, .

[0021] Optionally, the transforming the frequency spectrum signal into a power spectrum comprises:

[0022] the power spectrum is Pv=k* ; wherein, is the frequency spectrum signal, and k is a normalization coefficient.

[0023] Optionally, the action rule screening strategy comprises:

[0024] A numerical interval of the leg shaking frequency is set, referred to as a first frequency interval;

[0025] A numerical interval of the breathing frequency is set, referred to as a second frequency interval;

[0026] A maximum power value in the first frequency interval is identified as a first power peak value;

[0027] A maximum power value in the second frequency interval is identified as a second power peak value;

[0028] The power corresponding to the leg shaking motion is extracted from the power spectrum according to a preset motion rule screening strategy, including:

[0029] When the following conditions are met simultaneously, the power corresponding to the first power peak value is determined as the leg shaking power:

[0030] (a) The first power peak value is greater than a preset power threshold value;

[0031] (b) The first power peak value is greater than the second power peak value;

[0032] (c) The power difference between the first power peak value and its adjacent trough is greater than a set threshold value.

[0033] Optionally, the reminder information is generated based on the leg shaking power, including:

[0034] Obtaining body data of a user; the body data includes one or more of age, gender, weight, height, and medical history;

[0035] Generating personalized reminder information according to the body data of the user and the corresponding leg shaking power.

[0036] According to a second aspect of the embodiments of the present disclosure, the present application provides a leg shaking reminder system, which adopts the following technical solution:

[0037] A leg shaking reminder system, the system comprising: a deformation detection component and a processor, a reminder component, the deformation detection component and the reminder component are electrically connected with the processor;

[0038] The deformation detection component is used for real-time measurement to obtain piezoelectric detection data, and sends to the processor;

[0039] The processor is configured to acquire the piezoelectric detection data, extract a frequency spectrum signal from the piezoelectric detection data according to a preset information preprocessing strategy, perform relative power spectrum conversion on the frequency spectrum signal to obtain a power spectrum, extract a leg shaking power corresponding to a leg shaking action in the power spectrum according to a preset action rule screening strategy, and generate a reminding information according to the leg shaking frequency; and the processor is further configured to send the reminding information to the reminding component.

[0040] The reminding component is configured to collect the reminding information and remind a user by means of sound and / or display and / or vibration.

[0041] Optionally, the deformation detection component is one or more of a piezoelectric sheet, a piezoelectric film strip, an acceleration sensor and a gyroscope.

[0042] According to a third aspect of the embodiments of the present disclosure, the present application provides a backrest, which comprises a backrest body and a leg shaking reminding system as described above, and the deformation detection component is arranged below a side surface of the backrest body that is close to a human body.

[0043] According to a fourth aspect of the embodiments of the present disclosure, the present application provides a seat, which comprises a backrest as described above.

[0044] In summary, the present application provides a leg shaking reminding method, system and seat. The leg shaking reminding method provided by the present application acquires piezoelectric signals by using a deformation detection component to detect user body behavior in a non-contact manner, without the need to wear a sensing device, thereby improving the convenience and scene applicability. The frequency spectrum signal is extracted by using a preset information preprocessing strategy, the relative power spectrum is converted, and the leg shaking power is extracted in combination with an action rule screening strategy. The preset information preprocessing strategy can simplify the calculation logic, reduce the demand for processor computing power, and adapt to low-cost hardware, thereby solving the problems of high cost and high power consumption of a high-precision sensor scheme and having good economic efficiency. The action rule screening strategy can effectively filter out interference signals such as breathing and heart rate, accurately extract the leg shaking power in the relative power spectrum, and greatly improve the accuracy of leg shaking behavior recognition. Finally, the reminding information is generated according to the recognition result, and different users and scenes can be reminded differently.

[0045] Other features and advantages of the present application will be described in detail in the following specific embodiments. BRIEF DESCRIPTION OF DRAWINGS

[0046] The accompanying drawings are included to provide a further understanding of the present disclosure and constitute a part of the specification, and are used together with the following specific embodiments to explain the present disclosure, but do not constitute a limitation of the present disclosure. In the drawings:

[0047] Figure 1is a flow chart of a leg shaking reminding method according to an exemplary embodiment.

[0048] Figure 2 is a flow chart of a method for extracting a frequency spectrum signal from piezoelectric detection data according to a preset information preprocessing strategy according to an exemplary embodiment.

[0049] Figure 3 is a flow chart of a method for performing fast Fourier transform on piezoelectric detection data to obtain frequency spectrum data containing imaginary numbers according to an exemplary embodiment.

[0050] Figure 4 is a structural composition diagram of a leg shaking reminding system according to an exemplary embodiment.

[0051] Figure 5 is a power spectrum diagram of piezoelectric detection data according to an exemplary embodiment. DETAILED DESCRIPTION

[0052] The technical solutions in the present application will be described in detail below with reference to the drawings in the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. The components of the present application described and shown in the drawings herein can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the claimed present application, but only represents selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of the present application. It should be noted that: similar reference numerals and letters represent similar items in the following drawings, therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings. Meanwhile, in the description of the present application, the terms "first", "second", etc. are only used to distinguish the description, and cannot be understood as indicating or implying relative importance.

[0053] In the exemplary embodiments disclosed in the present application, a leg shaking reminding method is provided, which aims to solve the problems of poor convenience, limited scene universality and insufficient economy in existing leg shaking monitoring technologies.

[0054] Specifically, as shown in Figure 1 The leg shaking reminding method disclosed in the embodiments of the present application includes the following steps:

[0055] Step S101: acquiring piezoelectric detection data measured by the deformation detection component in real time.

[0056] In step S101, the piezoelectric detection data is collected in real time by the deformation detection component. When the user sits on the seat and performs the leg shaking action, the body will produce slight but regular mechanical vibrations. These vibrations are transmitted to the deformation detection component through the seat structure, causing it to generate an electrical signal change corresponding to the vibration intensity and frequency, i.e., piezoelectric detection data. In this process, non-contact and non-sensing monitoring is achieved, and the user does not need to wear any equipment, greatly improving the convenience and scene universality of use.

[0057] Step S102: Extract the frequency spectrum signal from the piezoelectric detection data according to the preset information preprocessing strategy.

[0058] In step S102, the original piezoelectric detection data is a typical time domain signal that mixes leg shaking, breathing, heartbeat, body micro-motion, and environmental noise, etc. Direct analysis is extremely difficult. This step converts the signal from the time domain to the frequency domain through the preset information preprocessing strategy, thereby more clearly revealing the internal frequency composition.

[0059] Step S103: Perform relative power spectrum transformation on the frequency spectrum signal to obtain the power spectrum.

[0060] In step S103, although the frequency spectrum signal reveals the frequency components, its amplitude is easily affected by various factors such as user weight, sitting posture, sensor sensitivity, etc., resulting in a large difference in signal intensity for different users or different scenes, lacking comparability. This step performs relative power spectrum transformation on the frequency spectrum signal to obtain the power spectrum, which reflects the relative strength relationship of each frequency component rather than the absolute energy value. In this way, the signal intensity fluctuations caused by individual differences and external condition changes are eliminated, making the method applicable to different users and different seats, improving the stability and generalization ability of the method.

[0061] Step S104: Extract the leg shaking power corresponding to the leg shaking action in the power spectrum according to the preset action rule screening strategy.

[0062] The leg shaking action has its unique physical law, which is usually characterized by a relatively stable main frequency, and the energy is concentrated around the main frequency. According to research, the frequency of a general person's leg shaking is usually 240-420 times / minute, i.e., 4-7 Hz, which is relaxed to 3-8 Hz in the exemplary embodiment of the present disclosure; the breathing frequency is about 0.2-0.3 Hz, and the heartbeat frequency is about 1-2 Hz. As can be seen, the frequencies of interference signals such as breathing and heartbeat are obviously different from leg shaking. The "action rule screening strategy" in step S104 is designed based on this prior knowledge, thereby extracting the leg shaking power corresponding to the leg shaking action in the power spectrum.

[0063] Step S105: Generate a reminder message based on the leg shaking power.

[0064] In step S105, the reminding information is generated according to the size and variation trend of the leg shaking power, so as to effectively distinguish and intervene the leg shaking behavior.

[0065] In the exemplary embodiments disclosed in the present application, the leg shaking reminding method provided by the present application acquires the piezoelectric signal by using the deformation detection component, detects the user's body behavior in a non-contact manner, and does not need to wear a sensing device, thereby improving the convenience and scene applicability; the frequency spectrum signal is extracted by using the preset information preprocessing strategy, the relative power spectrum is converted, and the leg shaking power is extracted by combining the motion rule screening strategy; the preset information preprocessing strategy can simplify the calculation logic, reduce the demand for processor computing power, adapt to low-cost hardware, solve the problems of high cost and high power consumption of the high-precision sensor scheme, and have good economy; the motion rule screening strategy can effectively filter out interference signals such as breathing and heart rate, accurately extract the leg shaking power in the relative power spectrum, and greatly improve the accuracy of leg shaking behavior recognition; finally, the reminding information is generated according to the recognition result, and can be reminded differently for different users and scenes.

[0066] For example, as shown in the figure, Figure 2 In the exemplary embodiments disclosed in the present application, step S102: the frequency spectrum signal is extracted from the piezoelectric detection data according to the preset information preprocessing strategy, specifically including the following steps:

[0067] Step S201: performing fast Fourier transform on the piezoelectric detection data to obtain frequency spectrum data containing imaginary numbers.

[0068] In this step, the original data output by the piezoelectric detection component is a voltage sequence in the time domain, which mixes the main vibration signal generated by leg shaking, low-frequency physiological signals such as breathing and heartbeat, and environmental noise. By applying fast Fourier transform (FFT), the time domain signal is decomposed into its frequency components, and the output result is frequency spectrum data in the form of complex numbers. The frequency spectrum data contains the amplitude and phase information (represented by the real part and the imaginary part) of each frequency component. This transformation highlights the periodic leg shaking signal hidden in the time domain noise in the frequency domain, so as to facilitate identification.

[0069] Specifically, as shown in the figure, Figure 3 Step S201: performing fast Fourier transform on the piezoelectric detection data to obtain frequency spectrum data containing imaginary numbers, including the following steps:

[0070] Step S301: decomposing the function called in FFT into trigonometric functions = ;

[0071] Step S302: dividing by T intervals, and calculating The corresponding sine value ;

[0072] Step S303: Calculate finer-grained values ​​based on lookup table method and polynomial approximation method. The corresponding sine value ;

[0073] Step S304: Based on the parity symmetry of trigonometric functions, convert the cosine of the trigonometric functions... and sine The operation is compressed into The calculation is performed within the interval; where, .

[0074] In traditional FFT implementations, trigonometric function calculations typically fall into two extremes: one is a pure lookup table method, where all potentially used sine / cosine values ​​are pre-stored in ROM (Read-Only Memory) and read directly during calculation; the other is a pure real-time calculation method, such as using Taylor series expansion, calculating from scratch each time it is needed. To achieve high precision, the pure lookup table method requires extremely large step resolution, resulting in a massive lookup table that severely consumes on-chip storage resources. While the pure real-time calculation method does not require additional storage, each calculation involves a large number of multiplication and addition operations, leading to long computational latency and high power consumption, failing to meet real-time requirements. Steps S301 to S303 avoid these two extremes. First, the calculation is divided into large intervals T, and the sine values ​​of these "coarse-grained" points are stored (using a lookup table). For example, the value of T is set to 1. This yields 91 sine values, which are stored as a lookup table (LUT). , can be used =sin( The angle is obtained from the sin table; there is no need to store the cos table separately. (For the mapped angle...) ∈[0°,90°], if If it is not an integer (i.e., it has a decimal part), then decompose it into... =A+B, where A is the integer part and B is the decimal part, then use the trigonometric identity:

[0075] sinθ=sin(A+B)=sinAcosB+cosAsinB;

[0076] This identity allows for the use of pre-calculated values ​​(stored in a lookup table) to handle arbitrary decimal angles, avoiding the direct calculation of trigonometric functions for non-integer angles and improving accuracy and efficiency.

[0077] Specifically, for the integer part A, based on cosA = sin(90° - A), sinA and cosA can be directly retrieved from the lookup table. For the decimal part B, a polynomial approximation method is used for calculation; first, the decimal part B is converted into radians. Because of B , The size is very small, so it can be approximated using Taylor expansion: , Then, substituting into the above trigonometric identities, we get:

[0078] sinθ sinA ( ) + cosA ( ).

[0079] For small angles, polynomial expansion requires only a few terms to achieve high accuracy. Thus, the method of this invention, combining table lookup and polynomial approximation to calculate sinθ, achieves near real-time accuracy with minimal memory overhead, resolving the inherent trade-off between accuracy and resources in embedded systems. Step S304 utilizes the parity and periodicity of trigonometric functions to compress the calculation range to the 0-90° interval, significantly reducing the required pre-stored data and redundant calculations; this reduction in computation directly translates to shorter FFT execution time. In real-time systems, this means lower latency and faster response times, better capturing user actions and improving user experience. In digital circuits, computation is a major source of power consumption. Reducing the computational load of trigonometric functions significantly shortens the active time of the logic units performing operations in the processor, directly reducing the overall dynamic power consumption of the chip. For battery-powered wearable devices, this directly extends their battery life.

[0080] Step S202: Perform modulo processing on the spectral data to obtain the spectral signal.

[0081] The spectral data obtained by FFT transformation is in complex form. Although it contains complete frequency information, the complex form is not conducive to direct quantitative analysis of signal strength. This step converts the complex spectral data into a real-number spectral signal through modulo-taking. Simultaneously, the modulo-taken spectral signal is in real-number form, reducing the computational complexity and hardware resource overhead in subsequent relative power spectrum calculations, and decreasing the bandwidth requirements for data storage and transmission.

[0082] For example, such as Figure 1 As shown, in the exemplary embodiment disclosed in this application, step S103: the spectral signal is converted into a relative power spectrum to obtain a power spectrum, specifically by performing the following calculations:

[0083] The power spectrum is Pv=k* ;in, is the spectral signal, and k is the normalization coefficient.

[0084] Through logarithmic operations , can compress the dynamic range of the spectrum signal Fv. When there is a large numerical difference in the spectrum signal, the logarithmic transformation can make the distribution of signals of different orders of magnitude on the power spectrum Pv more uniform, facilitating subsequent analysis of related signals such as leg shaking. Especially in the case of obvious difference between signal strength and weakness, it can more clearly reflect the characteristics of the effective signal. Second, the introduction of the normalization coefficient k can scale and adjust the entire power spectrum, so that the power spectrum can adapt to different application scenarios and device requirements. For example, in different sensitivity piezoelectric detection devices, k can be adjusted to ensure the rationality and consistency of the power spectrum, improving the flexibility and applicability of the method in practical applications. Preferably, in the exemplary embodiments of the present disclosure, k is 20, which is an ideal balance between numerical stability and computational efficiency. It ensures that all values in the power spectrum Fv, whether strong or weak, can be within a safe and efficient numerical range for subsequent peak detection, threshold comparison and other operations, avoiding algorithm failure or precision degradation due to numerical problems, especially suitable for resource-constrained hardware platforms.

[0085] For example, as shown in Figure 1 、 Figure 5 In the exemplary embodiments of the present disclosure, the action rule screening strategy includes:

[0086] Set the numerical range of the leg shaking frequency, referred to as the first frequency range;

[0087] Set the numerical range of the breathing frequency, referred to as the second frequency range;

[0088] Identify the maximum power value in the first frequency range as the first power peak;

[0089] Identify the maximum power value in the second frequency range as the second power peak;

[0090] Step S104: Extract the leg shaking power corresponding to the leg shaking action in the power spectrum according to the preset action rule screening strategy, including:

[0091] When the following conditions are met simultaneously, the power corresponding to the first power peak is determined as the leg shaking power:

[0092] (a) The first power peak is greater than the preset power threshold;

[0093] (b) The first power peak is greater than the second power peak;

[0094] (c) The power difference between the first power peak and its adjacent trough is greater than a set threshold.

[0095] In the exemplary embodiments disclosed in the present application, by dividing the first frequency interval and the second frequency interval, the first frequency interval corresponds to the leg shaking frequency interval, and the second frequency interval corresponds to the breathing frequency interval, the inherent difference between the human leg shaking action and the breathing action in the frequency characteristics is utilized to realize the preliminary separation of the two types of signals. This interval division based on the frequency characteristics can reduce the interference of the breathing action on the leg shaking recognition from the source, and solves the signal confusion problem easily caused by the fact that both are periodic human actions.

[0096] In the power peak identification link, the first power peak of the first frequency interval and the second power peak of the second frequency interval are extracted respectively, so as to convert the abstract power spectrum distribution into a quantifiable comparison feature parameter. On this basis, the three judgment conditions set form a progressive screening logic: condition (a) can effectively filter out low-power interference signals such as environmental vibration and sensor noise by setting a preset power threshold, so as to ensure that only signals with a certain energy intensity enter the subsequent judgment, thereby reducing invalid recognition; condition (b) requires that the first power peak be greater than the second power peak, which further excludes the influence of the breathing action and avoids misjudging the slight vibration caused by the breathing as the leg shaking; condition (c) limits the power difference between the first power peak and the adjacent trough, fully utilizes the periodic characteristics of the leg shaking action, i.e., the pressure change presents obvious "peak-valley" alternation when the leg is shaken, and random interference signals usually have no such rules, so as to effectively distinguish the leg shaking signal from the aperiodic interference.

[0097] By using the relative comparison of condition (b) and the morphological feature comparison of condition (c), the dependence on the preset power threshold of condition (a) is reduced, and the robustness and generalization ability of the method are improved. The leg shaking intensity of different users and the breathing depth may be different, and the gain of different devices and the noise level of different environments may also be different. If only the preset power threshold condition is relied on, the generalization ability of the algorithm will be poor, and tedious calibration needs to be performed for each user or scene. However, in the method of the present application, condition (a) is only a rough threshold, and the core judgment relies on the relative relationship between the signals and the morphology itself. Therefore, even if the absolute value of the leg shaking power of user A is lower than that of user B, as long as the leg shaking power of user A is strong enough relative to his own breathing power and the waveform is clear enough, it can also be accurately recognized.

[0098] In addition, although the method of the present application sets three judgment conditions, the calculation of each condition is relatively simple and does not involve complex matrix operations or iterative optimization. Compared with some complex recognition algorithms based on machine learning or pattern matching, the method of the present application has small calculation overhead and low power consumption, which makes it very suitable for real-time operation on embedded platforms such as wearable devices, Internet of Things nodes and the like with limited resources.

[0099] In the exemplary embodiments disclosed in the present application, the preset power threshold can be dynamically adjusted according to the actual application environment. When the overall environmental noise power is detected to be increased, the preset power threshold can be automatically adjusted upwards to avoid false triggering, and in a quiet environment, the preset power threshold can be appropriately adjusted downwards to ensure the capture of weak leg shaking actions. In the exemplary embodiments disclosed in the present application, the threshold is initially set to 10db. The power difference of 10db can ensure that the method of the present application responds to signals with concentrated energy and periodicity, effectively avoiding false judgments caused by breathing fluctuations, heartbeats or external vibrations. It should be noted that 10db is only the initial setting value and is not fixed. In actual application, it can be dynamically adjusted based on user feedback and scene changes. If the user feedback indicates that there is a missed judgment, it means that the current threshold may be too high, which can be adjusted to 8-9db to improve sensitivity. If there are more false judgments, it can be adjusted to 11-12db to enhance the anti-interference ability.

[0100] For example, in the exemplary embodiments disclosed in the present application, step S105: generating reminder information based on leg shaking power, includes:

[0101] Obtaining body data of the user; the body data includes one or more of age, gender, weight, height, and medical history;

[0102] Generating personalized reminder information according to the body data of the user and the corresponding leg shaking power.

[0103] In the exemplary embodiments disclosed in the present application, the leg shaking reminder method disclosed in the present application first obtains multi-dimensional user body data including age, gender, weight, height and medical history, and then combines such static data with the leg shaking power signal obtained by real-time dynamic monitoring, and finally generates personalized reminder information matched with the specific state and needs of the user. Of course, it can also obtain the user's input preference habits, combine the leg shaking power signal obtained by real-time dynamic monitoring, and generate targeted reminder information.

[0104] Specifically, in the exemplary embodiments disclosed in the present application, first, the reminding information can be generated based on the age dimension of the user: when it is monitored that the minor shakes legs more frequently and for a certain period of time, the correction reminding information is triggered, and the content of the reminding information can be "classmate, you have been shaking your legs for a while, shaking your legs too much will make others feel impolite, stop it quickly, let's develop good habits together! ". When it is detected that the user is a middle-aged or elderly person, if it is monitored that he or she is in a sedentary state and shakes legs lightly (which can be judged by the first power peak), a reminder is triggered, suggesting to moderately increase the leg shaking amplitude to better promote blood circulation and prevent thrombosis. If it is monitored that the user shakes legs too hard and for a certain period of time, whether or not he or she is in a sedentary state, a reminder is triggered to avoid excessive leg shaking and burden on the joints. Second, the reminding information can be generated based on the psychological state related performance of the user: when it is monitored that the user shakes legs at a high frequency and is accompanied by a significant fluctuation in heart rate for a certain period of time, it is speculated that he or she is in a state of anxiety and distraction, and a reminder is triggered, and the content of the reminder can be "your leg shaking frequency is currently high, and your heart rate also fluctuates to some extent, are you feeling anxious? You can stop and take a few deep breaths first, relax, and focus your attention better". Third, the reminding information can also be generated based on the medical history dimension of the user: if it is monitored that a user with a restless leg syndrome history shakes legs moderately at night and for a certain period of time, a reminder is triggered to indicate that the disease symptoms may have occurred, and appropriate relief measures or timely medical treatment is recommended. For users with a history of thrombosis, if it is monitored that he or she has been sitting for a period of time and hardly shakes legs, a reminder is triggered to emphasize the importance of timely leg shaking to promote blood circulation and reduce the risk of thrombosis recurrence. Finally, the user's self-defined rules and preferences can also be received, and the real-time dynamic monitoring of the leg shaking power signal is combined to generate targeted reminding information.

[0105] FIG. 4 is a block diagram of a leg shaking reminding system according to an exemplary embodiment, as shown, the system comprises a deformation detection component 401 and a processor 403, a reminding component 402, the deformation detection component 401 and the reminding component 402 are electrically connected with the processor 403; Figure 4

[0106] The deformation detection component 401 is used to measure the piezoelectric detection data in real time and send it to the processor 403;

[0107] The processor 403 is used to acquire the piezoelectric detection data, extract the frequency spectrum signal from the piezoelectric detection data according to the preset information preprocessing strategy, perform relative power spectrum transformation on the frequency spectrum signal to obtain the power spectrum, extract the leg shaking power corresponding to the leg shaking action in the power spectrum according to the preset action rule screening strategy, and generate the reminding information according to the leg shaking frequency; the processor 403 is also used to send the reminding information to the reminding component 402;

[0108] ​The reminding component 402 is used to collect reminding information and remind the user through sound and / or display and / or vibration.

[0109] The leg shaking reminding system provided by the application does not need to wear additional equipment, realizes non-contact monitoring through the deformation detection component integrated in the seat back, adopts low-cost sensing elements, cooperates with a simplified algorithm, reduces the cost of hardware and computing power, is beneficial to popularization and does not lose recognition accuracy, can effectively reduce the influence of interference signals such as breathing and heart rate, and provides differentiated reminders for different groups of people.

[0110] For example, in the exemplary embodiments disclosed in the application, the deformation detection component is one or more of a piezoelectric sheet, a piezoelectric film strip, an acceleration sensor and a gyroscope. The piezoelectric sheet and the piezoelectric film strip have very low cost advantages and are very suitable for cost-sensitive mass consumer products such as office chairs and study chairs. The acceleration sensor and the gyroscope are mature integrated circuit (IC) sensors, can directly output digital signals, simplify the design of the signal conditioning circuit, improve the integration and consistency of the system, and are suitable for scenes with higher requirements for reliability and ease of use. In actual application, the most suitable sensor can be selected according to the specific structure of the seat and the expected use scene, or multiple sensors can be used in combination to further improve the recognition accuracy through data complementation.

[0111] For example, in the exemplary embodiments disclosed in the application, the application also discloses a backrest, which comprises a backrest body and the leg shaking reminding system described above, and the deformation detection component is arranged below the side surface of the backrest body that is close to the human body. The backrest is a support point for the natural leaning of the human body, and actions such as leg shaking are often transmitted to the backrest structure through the spine. Placing the deformation detection component at this position can more directly and stably capture the action signals, reduce signal attenuation or interference caused by clothing, sitting posture changes and the like, and thus improve the accuracy and robustness of detection. Therefore, the leg shaking reminding system can capture the slight deformation or vibration caused by the leg shaking action in real time without the user's awareness. This design avoids the restraint and discomfort caused by traditional external devices, and improves the user's acceptance and long-term use willingness.

[0112] For example, in the exemplary embodiments disclosed in the application, the application also discloses a seat, which comprises the backrest described above. The seat itself is the core bearing platform of the human body sitting behavior, and the backrest and the seat cushion form a cooperative support structure. Actions such as leg shaking are often transmitted to the backrest through the seat cushion, and the leg shaking reminding system can realize multi-dimensional data acquisition through multiple sensing nodes, so as to more accurately recognize the leg shaking behavior, reduce the false alarm rate, and improve the overall robustness of the system.

[0113] The terms "comprises", "comprising", or any other variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus.

[0114] The description above merely illustrates preferred embodiments of the present application and the principles of the technology employed. It should be understood by those skilled in the art that the scope of the application involved in the present application is not limited to the technical solutions formed by the specific combinations of the above technical features, and should also cover other technical solutions formed by the combinations of the above technical features or their equivalent features without departing from the above application concept. For example, the above technical features can be replaced with the technical features applied in the present application (but not limited to) having similar functions to form technical solutions.

Claims

1. A fidgeting reminder method characterized by, The method comprises: obtaining piezoelectric detection data measured by a deformation detection component in real time; extracting a frequency spectrum signal from the piezoelectric detection data according to a preset information preprocessing strategy; performing relative power spectrum conversion on the frequency spectrum signal to obtain a power spectrum; extracting a leg shaking power corresponding to a leg shaking action from the power spectrum according to a preset action rule screening strategy; and generating a reminder information based on the leg shaking power.

2. The method of claim 1, wherein, The extracting of the frequency spectrum signal from the piezoelectric detection data according to the preset information preprocessing strategy comprises: performing fast Fourier transform on the piezoelectric detection data to obtain frequency spectrum data containing imaginary numbers; and performing modulo processing on the frequency spectrum data to obtain the frequency spectrum signal.

3. The method of claim 2, wherein, The performing of the fast Fourier transform on the piezoelectric detection data to obtain the frequency spectrum data containing imaginary numbers comprises: The functions called in the FFT are broken down into trigonometric functions The functions called in the FFT are broken down into trigonometric functions = ; Divide by intervals of T, calculate Divide by intervals of T, calculate Corresponding sine values ; Based on the table lookup method and polynomial approximation method to calculate more fine-grained Corresponding sine values ; Based on the parity symmetry of the trigonometric functions, the operation of cosine and sine in the trigonometric functions is compressed to be calculated in the interval of ; wherein .

4. The method of claim 3, wherein the leg shaking reminder is provided by a device selected from the group consisting of a smartphone, a smartwatch, a computer, a tablet, a wearable device, and a smart home device. The performing of the relative power spectrum conversion on the frequency spectrum signal to obtain the power spectrum comprises: The power spectrum is Pv=k ; wherein, is the frequency spectrum signal, and k is a normalization coefficient.

5. The method of claim 4, wherein, The action rule screening strategy comprises: setting a numerical interval of a leg shaking frequency, referred to as a first frequency interval; setting a numerical interval of a breathing frequency, referred to as a second frequency interval; identifying a maximum power value in the first frequency interval as a first power peak value; identifying a maximum power value in the second frequency interval as a second power peak value; The extracting of the leg shaking power corresponding to the leg shaking action from the power spectrum according to the preset action rule screening strategy comprises: when the following conditions are simultaneously satisfied, determining a power corresponding to the first power peak value as the leg shaking power: (a) the first power peak value is greater than a preset power threshold value; (b) the first power peak value is greater than the second power peak value; (c) a power difference between the first power peak value and an adjacent valley thereof is greater than a set threshold value.

6. The method of claim 1, wherein, The generating of the reminder information based on the leg shaking power comprises: obtaining body data of a user; the body data comprises one or more of age, gender, weight, height, and medical history; generating personalized reminder information according to the body data of the user and the corresponding leg shaking power.

7. A fidget reminder system, characterized in that The system comprises a deformation detection component, a processor, and a reminder component; the deformation detection component and the reminder component are electrically connected to the processor; the deformation detection component is configured to measure piezoelectric detection data in real time and send the piezoelectric detection data to the processor; the processor is configured to obtain the piezoelectric detection data, extract a frequency spectrum signal from the piezoelectric detection data according to a preset information preprocessing strategy, perform relative power spectrum conversion on the frequency spectrum signal to obtain a power spectrum, extract a leg shaking power corresponding to a leg shaking action from the power spectrum according to a preset action rule screening strategy, and generate a reminder information based on the leg shaking power; the processor is further configured to send the reminder information to the reminder component; the reminder component is configured to receive the reminder information and remind a user through sound and / or display and / or vibration.

8. A fidget reminder system as claimed in claim 7, wherein, The deformation detection component is one or more of a piezoelectric sheet, a piezoelectric film strip, an acceleration sensor, and a gyroscope.

9. A backrest, characterized in that The backrest comprises a backrest body, and comprises the leg shaking reminding system as claimed in any one of claims 7-8, and the deformation detection assembly is arranged below the side surface of the backrest body which is close to the human body.

10. A seat furniture, characterized in that The seat comprises the backrest as claimed in claim 9.

Images