CN116859761A - Cross-platform intelligent home linkage system - Google Patents
Cross-platform intelligent home linkage system Download PDFInfo
- Publication number
- CN116859761A CN116859761A CN202310777702.XA CN202310777702A CN116859761A CN 116859761 A CN116859761 A CN 116859761A CN 202310777702 A CN202310777702 A CN 202310777702A CN 116859761 A CN116859761 A CN 116859761A
- Authority
- CN
- China
- Prior art keywords
- blood oxygen
- signal
- skin
- feature
- sensing unit
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 239000008280 blood Substances 0.000 claims abstract description 112
- 210000004369 blood Anatomy 0.000 claims abstract description 112
- QVGXLLKOCUKJST-UHFFFAOYSA-N atomic oxygen Chemical compound [O] QVGXLLKOCUKJST-UHFFFAOYSA-N 0.000 claims abstract description 111
- 229910052760 oxygen Inorganic materials 0.000 claims abstract description 111
- 239000001301 oxygen Substances 0.000 claims abstract description 111
- 230000008451 emotion Effects 0.000 claims abstract description 71
- 238000001514 detection method Methods 0.000 claims abstract description 29
- 238000001914 filtration Methods 0.000 claims abstract description 18
- 230000001960 triggered effect Effects 0.000 claims abstract description 11
- 238000000034 method Methods 0.000 claims description 45
- 230000035790 physiological processes and functions Effects 0.000 claims description 28
- 238000012549 training Methods 0.000 claims description 27
- 238000012544 monitoring process Methods 0.000 claims description 16
- 238000004458 analytical method Methods 0.000 claims description 13
- 238000005286 illumination Methods 0.000 claims description 13
- 238000012790 confirmation Methods 0.000 claims description 10
- 230000007613 environmental effect Effects 0.000 claims description 9
- 230000002452 interceptive effect Effects 0.000 claims description 9
- 230000005611 electricity Effects 0.000 claims description 8
- 238000004364 calculation method Methods 0.000 claims description 7
- 230000002996 emotional effect Effects 0.000 claims description 7
- 238000012216 screening Methods 0.000 claims description 7
- 230000009191 jumping Effects 0.000 claims description 6
- 230000002093 peripheral effect Effects 0.000 claims description 6
- 238000000605 extraction Methods 0.000 claims description 5
- 238000007781 pre-processing Methods 0.000 claims description 5
- 238000012545 processing Methods 0.000 claims description 5
- 230000004044 response Effects 0.000 claims description 5
- 238000013507 mapping Methods 0.000 claims description 4
- 238000001228 spectrum Methods 0.000 claims description 4
- 238000004891 communication Methods 0.000 claims description 3
- 238000010276 construction Methods 0.000 claims description 3
- 238000002790 cross-validation Methods 0.000 claims description 3
- 230000004962 physiological condition Effects 0.000 claims description 3
- 238000012163 sequencing technique Methods 0.000 claims description 3
- 230000008909 emotion recognition Effects 0.000 abstract description 14
- 230000007246 mechanism Effects 0.000 abstract description 9
- 238000012706 support-vector machine Methods 0.000 abstract description 7
- 230000006870 function Effects 0.000 description 20
- 230000000875 corresponding effect Effects 0.000 description 10
- 230000000694 effects Effects 0.000 description 8
- 238000013186 photoplethysmography Methods 0.000 description 7
- 238000006243 chemical reaction Methods 0.000 description 6
- 206010016256 fatigue Diseases 0.000 description 6
- 230000008569 process Effects 0.000 description 6
- 238000009434 installation Methods 0.000 description 5
- 230000031700 light absorption Effects 0.000 description 5
- 238000010606 normalization Methods 0.000 description 5
- 108010054147 Hemoglobins Proteins 0.000 description 4
- 102000001554 Hemoglobins Human genes 0.000 description 4
- 108010064719 Oxyhemoglobins Proteins 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 238000010521 absorption reaction Methods 0.000 description 2
- 210000001367 artery Anatomy 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 230000017531 blood circulation Effects 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- 230000009467 reduction Effects 0.000 description 2
- 231100000430 skin reaction Toxicity 0.000 description 2
- 210000000106 sweat gland Anatomy 0.000 description 2
- 230000035900 sweating Effects 0.000 description 2
- 210000002820 sympathetic nervous system Anatomy 0.000 description 2
- 230000003867 tiredness Effects 0.000 description 2
- 208000016255 tiredness Diseases 0.000 description 2
- 210000001519 tissue Anatomy 0.000 description 2
- 208000019901 Anxiety disease Diseases 0.000 description 1
- 238000012935 Averaging Methods 0.000 description 1
- 229910003798 SPO2 Inorganic materials 0.000 description 1
- 101100478210 Schizosaccharomyces pombe (strain 972 / ATCC 24843) spo2 gene Proteins 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 230000036506 anxiety Effects 0.000 description 1
- 230000002238 attenuated effect Effects 0.000 description 1
- 230000004888 barrier function Effects 0.000 description 1
- 210000000988 bone and bone Anatomy 0.000 description 1
- 238000013145 classification model Methods 0.000 description 1
- 210000002808 connective tissue Anatomy 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 238000012217 deletion Methods 0.000 description 1
- 230000037430 deletion Effects 0.000 description 1
- 230000005059 dormancy Effects 0.000 description 1
- 230000008030 elimination Effects 0.000 description 1
- 238000003379 elimination reaction Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000006996 mental state Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 210000003205 muscle Anatomy 0.000 description 1
- 230000009914 physiological arousal Effects 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
- 230000008844 regulatory mechanism Effects 0.000 description 1
- 230000028327 secretion Effects 0.000 description 1
- 230000003595 spectral effect Effects 0.000 description 1
- 238000011425 standardization method Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 210000003462 vein Anatomy 0.000 description 1
- 238000002759 z-score normalization Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B15/00—Systems controlled by a computer
- G05B15/02—Systems controlled by a computer electric
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B19/00—Programme-control systems
- G05B19/02—Programme-control systems electric
- G05B19/418—Total factory control, i.e. centrally controlling a plurality of machines, e.g. direct or distributed numerical control [DNC], flexible manufacturing systems [FMS], integrated manufacturing systems [IMS] or computer integrated manufacturing [CIM]
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B2219/00—Program-control systems
- G05B2219/20—Pc systems
- G05B2219/26—Pc applications
- G05B2219/2642—Domotique, domestic, home control, automation, smart house
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/02—Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]
Landscapes
- Engineering & Computer Science (AREA)
- General Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Manufacturing & Machinery (AREA)
- Quality & Reliability (AREA)
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
The application provides a cross-platform intelligent home linkage system. According to the application, the handle sleeve is fixedly arranged at the armrest or leaning position of the furniture, the surface of the handle sleeve is provided with the plurality of spring pins, and the skin electric sensing unit and the sensing signals acquired by the blood oxygen sensing unit which are embedded in the handle sleeve are input to the control system by utilizing the spring pins. According to the application, the control system is used for detecting the signal states of the skin electric sensing signal and the blood oxygen concentration sensing signal, filtering the sensing signals outside the effective range, and then utilizing the SVM emotion classifier to identify the tired state and the emotion state of the user corresponding to the sensing signals, and when the user is in tired or negative emotion, the detection data is stored and the linkage system is triggered to carry out corresponding. The application can accurately monitor the fatigue state and the emotion state of the user in real time through the optimized support vector machine and emotion recognition algorithm and trigger the linkage system to execute a corresponding adjustment mechanism.
Description
Technical Field
The application relates to the technical field of intelligent furniture, in particular to a cross-platform intelligent home linkage system.
Background
Home appliances such as furniture, lighting devices, temperature control devices, audio devices, entertainment systems, and interactive voice systems are commonly installed in home environments. The existing lighting equipment, temperature control equipment, sound equipment, entertainment systems and interactive voice systems are not communicated with each other, and cannot be adjusted in a linkage mode according to the real-time physiological conditions of users.
In order to realize intelligent home linkage, in the prior art, a single general control module is used for triggering each lighting device, temperature control device, sound device, entertainment system and interactive voice system to switch respective working states in a linkage manner according to a user trigger signal or according to execution time preset by a user. However, the linkage interaction mode is too single, cannot actively respond to the demands of clients in real time, and is complex in operation.
Disclosure of Invention
Aiming at the defects of the prior art, the application provides a cross-platform intelligent home linkage system, wherein skin electric signals and blood sample saturation sensors are embedded in a handle sleeve in the arrangement of a furniture armrest or leaning position, and the fatigue state and the emotion state of a user are monitored in real time through an optimized support vector machine and an emotion recognition algorithm, and the running state of home equipment is correspondingly triggered and regulated. The application adopts the following technical scheme.
Firstly, in order to achieve the above purpose, a cross-platform intelligent home linkage system is provided, which comprises: the furniture body is fixedly provided with a handle sleeve in the circumferential direction of the armrest or leaning position, and a plurality of spring pins are further arranged on the surface covered by the handle sleeve; the device comprises a handle sleeve, a control system, a spring pin connection control system, a blood oxygen sensing unit, a skin electric sensing unit, a signal connection wire and a power supply connection wire, wherein the skin electric sensing unit and the blood oxygen sensing unit are embedded in the handle sleeve, the blood oxygen sensing unit is arranged on the outer peripheral surface of the handle sleeve, the skin electric sensing unit is arranged on the inner peripheral surface of the handle sleeve, the signal connection wire and the power supply connection wire of the skin electric sensing unit and the blood oxygen sensing unit are respectively connected with the control system through the spring pin connection control system, and a skin electric sensing signal acquired by the skin electric sensing unit and a blood oxygen concentration sensing signal acquired by the blood oxygen sensing unit are output to the control system; the button is arranged on the linkage module of the furniture body, is in communication connection with the control system and is used for responding to the trigger of a user and feeding back a confirmation signal to the control system; the control system is provided with: the signal state detection module is used for receiving the skin electric sensing signals and the blood oxygen concentration sensing signals through signal connecting wires of the skin electric sensing unit and the blood oxygen sensing unit respectively, judging whether the sensing signals are in an effective range or not, and filtering sensing signals outside the effective range; the physiological state analysis module is connected with the signal state detection module, receives sensing signals in an effective range, and recognizes the tired state and the emotional state of a user corresponding to the sensing signals through the SVM emotion classifier, and when the user is in tired or negative emotion, the physiological state analysis module stores detection data and triggers the multi-platform interface of the linkage module to switch the running state of household equipment; the control system also responds to the confirmation signal of the button to cancel the trigger signal for switching the running state of the household equipment to the multi-platform interface.
Optionally, the cross-platform intelligent home linkage system according to any one of the preceding claims, wherein the multi-platform interface is connected with a lighting device, a temperature control device, an audio device, an entertainment system, and an interactive voice system in a home environment.
Optionally, the cross-platform smart home linkage system according to any one of the preceding claims, wherein the handle sleeve comprises: the rear sleeve body is fixedly arranged at the rear side of the furniture body, a blood oxygen sensing unit mounting groove is formed in the rear side wall of the rear sleeve body, and a blood oxygen sensing unit is embedded in the blood oxygen sensing unit mounting groove; the front sleeve body is fixedly arranged on the front side of the furniture body and fixedly connected with the rear sleeve body, and the front sleeve body and the rear sleeve body are surrounded and fixed on the periphery of the furniture body; a split mounting groove is further formed between the front sleeve body and the rear sleeve body, and the skin electric sensor is embedded into the split mounting groove; the skin electric sensor and the blood oxygen sensing unit are at least partially exposed on the surface of the handle sleeve.
Meanwhile, in order to achieve the above purpose, the application also provides a physiological state monitoring method based on skin electricity and blood oxygen signals, which is used for the cross-platform intelligent home linkage system, and comprises the following steps: after power-on, respectively receiving skin electric sensing signals acquired by the skin electric sensing unit and blood oxygen concentration sensing signals acquired by the blood oxygen sensing unit, judging whether each sensing signal is in an effective range, and filtering sensing signals outside the effective range; and extracting features of the skin electric sensing signals in the effective range and the blood oxygen concentration sensing signals in the effective range, inputting the extracted features into an SVM emotion classifier for physiological state analysis, identifying the tired state and the emotion state of a user, and storing detection data and triggering a multi-platform interface of a linkage module to switch the running state of household equipment when the user is in tired or negative emotion.
Optionally, the method for monitoring physiological states based on galvanic skin and blood oxygen signals according to any one of the above, wherein, when extracting features of the galvanic skin sensing signal in the effective range and the blood oxygen concentration sensing signal in the effective range, each extracted feature for inputting into the SVM emotion classifier for physiological state analysis is determined by the following steps: step s1, acquiring a skin electric sample signal and an blood oxygen concentration sample signal under tired or negative emotion, performing low-pass filtering on the skin electric sample signal in an effective range and the blood oxygen concentration sensing signal in the effective range, and removing baseline drift to obtain a preprocessing signal; step s2, extracting various types of features from the preprocessed signals to construct a feature set M; step s3, respectively carrying out standardization processing on each feature in the feature set M, eliminating individual differences of each feature, obtaining data samples with normal distribution, and constructing a training set according to the data samplesWherein x is E R D ,y∈{-1,+1}，x i For the ith data sample, N is the total data sample amount, D is the data sample feature number, R D For feature space, y ε { 1, +1} is a sample label indicating whether the data sample corresponds to a user being tired or in a negative emotion; step s4, constructing an SVM emotion classifier, and training a data sample: to bring the effective characteristics into a set F IN Initializing to include all features in the training set, initializing the ordered set P to null, and setting the effective feature set F IN And the ordered set P iterates according to the following steps until the effective feature set F IN Is empty: computing an active feature set F IN The cost function of each feature in the list is shifted out to the sorting set P, and the effective feature set F IN Updated to the remaining features, continuing with the updated active feature set F IN Performing iterative calculation on the sorting set P; and step s5, adopting a stepping method for each feature in the sequencing set P, and screening out the feature which minimizes the overall Weibull Lambda in each step as the feature input into the SVM emotion classifier.
Optionally, the method for monitoring physiological states based on galvanic skin and blood oxygen signals according to any one of the preceding claims, wherein the effective feature set F is further compared with the cost function in step s4 each time IN After iterative updating is carried out on the sorting set P, the SVM-RFE-CBR method is further adopted to screen out the features highly related to the marking features Q from the updated sorting set P, and the screened features are adjusted to an effective feature set F IN To continue to screen out the adjusted effective feature set F IN Performing iterative calculation on the sorting set P; the marking feature Q is the feature with the highest score in the sorting set P obtained by updating according to the cost function.
Optionally, the method for monitoring physiological states based on galvanic skin and blood oxygen signals according to any one of the preceding claims, wherein the SVM emotion classifier is trained to obtain: step t1, judging whether a feature set of a data sample is linearly separable, firstly mapping a feature space of the data sample to a high-dimensional linearly separable space through a kernel function under the condition that the feature set is linearly inseparable, and then jumping to step t2; directly jumping to the step t2 under the condition of linear separable; step t2, in the linear separable feature space R corresponding to the data sample feature set D Mid-construction hyperplane f (x) =w T ·x i +b toAs an objective function, with y i (w T x i +b) is 1 or more, i=1, 2, …, n is constraint solving +.>Obtaining a hyperplane f (x) =w that maximizes the minimum distance of the sample point to the hyperplane T ·x i +b, where a i A Lagrangian multiplier, b is a real number representing the distance between the hyperplane and the origin; and t3, randomly dividing the data sample into K parts, training the SVM emotion classifier by using the data sample in a K-fold cross validation mode until the accuracy of the SVM emotion classifier in analyzing the physiological state of the data sample reaches 90%, and obtaining the SVM emotion classifier.
Optionally, the method for monitoring physiological states based on galvanic skin and blood oxygen signals according to any one of the above, wherein the galvanic skin sensing signal and the blood oxygen concentration sensing signal in the effective range are subjected to low-pass filtering before feature extraction, and baseline drift is removed.
Optionally, the method for monitoring a physiological state based on galvanic skin and blood oxygen signals according to any one of the preceding claims, wherein the features extracted from the pre-processed signals for constructing the feature set M comprise any one or any combination of the following: the ratio of low frequency and high frequency energy FFT_ratio of the pulse count PPS per second, the FFT absolute value of the pulse count PPS per second is fully integrated with L1_Norm, the first order differential of the pulse count PPS per second is the maximum value diffPP_max, the mean value SPO2_mean of the original blood oxygen saturation signal, the standard deviation SPO2_sd of the original blood oxygen saturation signal, the mean value SPO2_ afd _mean of the first order differential absolute value of the original blood oxygen saturation signal, the mean value SPO2_asd_mean of the second order differential absolute value of the original blood oxygen saturation signal, the mean value SPO2_std_ afd _mean of the first order differential absolute value of the normalized blood oxygen saturation signal, the mean value SPO2_std_asd_mean of the second order differential absolute value of the normalized blood oxygen saturation signal, the method comprises the steps of a signal mean MSKT of a skin temperature original signal, a variance STD of the skin temperature original signal, energy data SP_L of a skin temperature signal low frequency band, energy data SP_H of a skin temperature signal high frequency band, a spectrum power ratio ER_FFT between the skin temperature signal high frequency band and the skin temperature signal low frequency band, an average dropping rate AOND of skin electric signals during decay, the number SRNOLM of local minimum values of the skin electric signals, energy data FP of the skin electric signals in a frequency band of 0-2.4 Hz, spectrum power SP of the skin electric signals in the frequency band of 0-2.4 Hz, a positive-negative zero crossing rate Pnum of a low frequency signal SR in the skin electric signals and a positive-negative zero crossing rate Nnum of an extremely low frequency signal VSR in the skin electric signals.
Optionally, the method for monitoring a physiological state based on galvanic skin and blood oxygen signals according to any one of the above, wherein the method comprises the steps of storing detection data and triggering the multi-platform interface to reduce the environmental temperature set by the temperature control device, reduce the illumination intensity of the illumination device, and reduce the environmental volume of the audio device and the entertainment system when the user is in a tired state; and when the negative emotion of the user is recognized, the detection data are stored, the multi-platform interface is triggered to adjust the environmental temperature set by the temperature control equipment to a dormant interval, only the low-illumination operation of the auxiliary lighting equipment is maintained, the entertainment system is turned off, and the interactive voice system is turned off.
Advantageous effects
The cross-platform intelligent home linkage system provided by the application is characterized in that a handle sleeve is fixedly arranged at a furniture armrest or leaning position, a plurality of spring pins are arranged on the surface of the handle sleeve, and sensing signals acquired by a skin electric sensing unit and a blood oxygen sensing unit which are embedded in the handle sleeve are input to a control system by utilizing the spring pins. According to the application, the control system is used for detecting the signal states of the skin electric sensing signal and the blood oxygen concentration sensing signal, filtering the sensing signals outside the effective range, and then utilizing the SVM emotion classifier to identify the tired state and the emotion state of the user corresponding to the sensing signals, and when the user is in tired or negative emotion, the detection data is stored and the linkage system is triggered to carry out corresponding. The application can accurately monitor the fatigue state and the emotion state of the user in real time through the optimized support vector machine and emotion recognition algorithm and trigger the linkage system to execute a corresponding adjustment mechanism. According to the application, based on the sensing signals collected by the skin electric sensing unit and the blood oxygen sensing unit, various characteristics are extracted, individual differences of each characteristic are respectively eliminated through standardized means such as Z-score, correlation deviation generated when a large number of highly correlated characteristics exist in a characteristic space is avoided by utilizing an SVM-RFE-CBR method, and false characteristic ordering caused by underestimation of importance of certain characteristics is avoided, so that an SVM emotion classifier is more accurately trained, more accurate recognition results of emotional states such as tiredness and anger are obtained, and the recognition effect of a control system under the same operation cost is effectively improved.
Additional features and advantages of the application will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by practice of the application.
Drawings
The accompanying drawings are included to provide a further understanding of the application and are incorporated in and constitute a part of this specification, illustrate the application and together with the embodiments of the application, and do not limit the application. In the drawings:
FIG. 1 is an assembled state schematic diagram of a cross-platform smart home linkage system of the present application;
FIG. 2 is a schematic view of the structure of the system of the present application in the assembled position of the handle sleeve on the furniture body;
FIG. 3 is a schematic view of the rear housing structure in the system of the present application;
FIG. 4 is a schematic view of the front ferrule configuration in the system of the present application
FIG. 5 is a flow chart of a physiological condition monitoring method employed by the system of the present application;
FIG. 6 is a schematic diagram of a process for implementing emotion recognition by training a model by a physiological state analysis module according to the present application;
FIG. 7 is a schematic diagram of effective range screening of sensing signals according to the present application;
fig. 8 is a schematic diagram of the response process of the system of the present application.
In the figure, 1 denotes a furniture body; 11 represents a pin female head of a spring pin of the positive electrode of a power supply; 12 represents a power supply negative electrode spring pin female head; 13 denotes a skin-friendly sensing spring pin female; 14 represents a blood oxygen sensing spring pin female; 15 denotes a positioning projection; 2 represents a button; 3 represents a handle sleeve; 31 represents a male pin of a spring pin of the positive electrode of a power supply; 32 denotes a power supply negative spring pin male; 33 denotes a skin-facing electrical sensing spring pin male; 34 represents the blood oxygen sensing spring pin male; 35 denotes a positioning groove; 301 represents a screw hole; 304 denotes a split mounting groove; 305 denotes a blood oxygen sensor unit mounting groove; 4 represents a galvanic skin sensing unit; and 5 denotes a blood oxygen sensing unit.
Detailed Description
In order to make the purpose and technical solutions of the embodiments of the present application more clear, the technical solutions of the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present application. It will be apparent that the described embodiments are some, but not all, embodiments of the application. All other embodiments, which can be made by a person skilled in the art without creative efforts, based on the described embodiments of the present application fall within the protection scope of the present application.
It will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the prior art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.
The meaning of "and/or" in the present application means that each exists alone or both exist.
The meaning of 'inner and outer' in the application refers to that the direction of pointing the sleeve to the inner stitch of the furniture body is inner relative to the furniture body, and vice versa; and not to a particular limitation of the mechanism of the device of the present application.
The meaning of "left and right" in the present application means that when the user is facing the furniture body, the left side of the user is left, and the right side of the user is right, but is not a specific limitation of the device mechanism of the present application.
"connected" as used herein means either a direct connection between components or an indirect connection between components via other components.
The meaning of "up and down" in the present application means that when the user is facing the furniture body, the upper side of the user is up, and the lower side of the user is down, and is not a specific limitation of the mechanism of the present application.
The meaning of "front and rear" in the present application refers to that when a user is facing the furniture body, the side of the furniture body facing the user is the front side, and the side of the furniture body facing away from the user is the rear side, which is not a specific limitation of the device mechanism of the present application.
Fig. 1 and fig. 2 are cross-platform intelligent home linkage systems according to the present application, which include:
The furniture body 1 is fixedly provided with a handle sleeve 3 in the circumferential direction of the armrest or leaning position, and the surface covered by the handle sleeve 3 of the furniture body 1 is also provided with a plurality of spring pins;
the skin electric sensing unit 4 and the blood oxygen sensing unit 5 are embedded in the handle sleeve 3, the blood oxygen sensing unit 5 is arranged on the outer peripheral surface of the handle sleeve 3, the skin electric sensing unit 4 is arranged on the inner peripheral surface of the handle sleeve 3, and a signal connecting wire and a power supply connecting wire of the skin electric sensing unit 4 and the blood oxygen sensing unit 5 are respectively connected with the control system through the spring pins, so that a skin electric sensing signal acquired by the skin electric sensing unit 4 and a blood oxygen concentration sensing signal acquired by the blood oxygen sensing unit 5 are output to the control system;
a button 2, which is arranged on the linkage module of the furniture body 1, is connected with the control system in a communication way, and is used for responding to the trigger of a user and feeding back a confirmation signal to the control system;
the control system receives the skin electricity sensing signal collected by the skin electricity sensing unit 4 and the blood oxygen concentration sensing signal collected by the blood oxygen sensing unit 5 after the power-on starting, and realizes the control through an internal program unit or an internal independent circuit gate:
The signal state detection module is used for receiving the skin electric sensing signals and the blood oxygen concentration sensing signals through the signal connecting wires of the skin electric sensing unit 4 and the blood oxygen sensing unit 5 respectively, judging whether the sensing signals are in an effective range or not, and filtering sensing signals outside the effective range;
the physiological state analysis module is connected with the signal state detection module, receives the sensing signals in the effective range, extracts characteristics of the skin electric sensing signals in the effective range and the blood oxygen concentration sensing signals in the effective range, inputs the extracted characteristics into the SVM emotion classifier, and carries out physiological state analysis through the SVM emotion classifier to identify the tired state and the emotion state of the user corresponding to the sensing signals, so that when the user is tired or in a negative emotion, detection data are stored and the multi-platform interface of the linkage module is triggered to switch the running state of the household equipment until the control system receives the confirmation signal of the user trigger button to correspondingly cancel the trigger signal for switching the running state of the household equipment on the multi-platform interface in response to the triggering action of the user, and each furniture equipment is enabled to be restored to the original running state.
In the application, the blood oxygen sensing unit detects blood oxygen saturation SPO2 through a photoplethysmography and acquires a blood oxygen concentration sensing signal. The principle is as follows: the illumination is attenuated to some extent when it passes through the skin tissue and then reflects back to the photosensitive sensor. The absorption of light is substantially unchanged (provided that the measurement site does not move significantly) like muscles, bones, veins and other connective tissue, but the absorption of light naturally also changes due to the blood flow in the arteries, unlike blood. When we convert light into an electrical signal, the resulting signal can be divided into a direct current DC signal and an alternating current AC signal, just because the absorption of light by the artery changes while the absorption of light by other tissues is substantially unchanged. The AC signal is extracted to reflect the blood flow characteristic. We call this technique photoplethysmography PPG. Because of the presence of a certain proportion of oxyhemoglobin HbO2 and hemoglobin Hb contained in blood, it is simply referred to as blood oxygen saturation (SPO 2). The absorption coefficient of the oxyhemoglobin HbO2 and the hemoglobin Hb to Hb between the wavelengths 600 and 800nm is higher, and the absorption coefficient of the oxyhemoglobin HbO2 and the hemoglobin Hb to HbO2 between the wavelengths 800 and 1000 is higher. Therefore, the PPG signals of HbO2 and Hb can be detected by utilizing the light of red light (600-800 nm) and near infrared light (800-1000 nm), and then the corresponding ratio is calculated through program processing, so that the blood oxygen saturation (SPO 2) is obtained.
In the application, the skin electric sensing unit acquires skin electric sensing signals reflecting the conductivity of the skin surface through skin electric activity, namely skin electric reaction. The principle is as follows: the galvanic skin activity varies with the state of sweat glands in the skin. Whereas human sweating is controlled by the sympathetic nervous system, it is believed that galvanic skin activity is indicative of psychological or physiological arousal. In general, when a person is highly concentrated (wakes up), perspiration secretion increases, thereby increasing skin conductance.
When the control system detects and analyzes the signal states of the skin electric sensing signal and the blood oxygen concentration sensing signal, a support vector machine (support vector machines, SVM) classification model is adopted, and the physiological signal is identified to avoid the emotion state of subjective disguise, so that the control system can be closer to the intrinsic psychological feeling of emotion. Galvanic skin response refers to the electrical changes in the skin caused by psychological stimuli with and without external electricity in the circuit. Galvanic skin response signals vary significantly from emotion to emotion and are typically tested for response to an environment or stimulus that stimulates anxiety emotion.
Specifically, referring to fig. 2, the positions of the handles in the furniture body where the handles are installed according to the present application are arranged along the circumferential direction of the handle ring: the power supply anode spring stitch female head 11, the power supply cathode spring stitch female head 12, the skin electric sensing spring stitch female head 13, the blood oxygen sensing spring stitch female head 14 and the positioning protrusion 15. And correspondingly, the handle sleeve is arranged into a front sleeve body and a rear sleeve body which are connected into a whole along the periphery of the furniture body 1. An opening is arranged between the front sleeve body and the rear sleeve body, the opening can be pulled out from the inner side, and the sleeve is sleeved inwards from the outer ring of the furniture body 1. In the installation state, the rear sleeve body is fixedly arranged at the rear side of the furniture body 1, and the rear side wall of the rear sleeve body is provided with a blood oxygen sensing unit installation groove 305 so as to enable a blood oxygen sensing unit 5 to be embedded and arranged in the blood oxygen sensing unit installation groove 305; the front sleeve body is fixedly arranged on the front side of the furniture body 1 and fixedly connected with the rear sleeve body, and the front sleeve body and the rear sleeve body are surrounded and fixed on the periphery of the furniture body 1; and a split mounting groove 304 can be arranged between the front sleeve body and the rear sleeve body at the opening position of the front sleeve body, so that the skin electric sensor 4 is embedded into the split mounting groove 304, and the sensing collection of bioelectric signals is realized by exposing the skin electric sensor 4 and the blood oxygen sensing unit 5 on the surface of the sleeve 3 at least partially.
During installation, the handle sleeve can be positioned by clamping the positioning protrusion 15 on the surface of the furniture body 1 through the positioning groove 35 on the inner side of the handle sleeve shown in fig. 4, and then the screw hole 301 arranged at the opening position on the inner side of the handle sleeve shown in fig. 3 is fastened by a screw to realize positioning installation. At this time, the male head 31 of the power supply positive spring pin, the male head 32 of the power supply negative spring pin, the male head 33 of the skin electricity sensing spring pin and the male head 34 of the blood oxygen sensing spring pin on the sleeve are respectively connected with the female head 11 of the power supply positive spring pin, the female head 12 of the power supply negative spring pin, the female head 13 of the skin electricity sensing spring pin and the female head 14 of the blood oxygen sensing spring pin of the furniture body, so that bioelectric sensing signals are transmitted to the control unit, when a user is tired or in a negative emotion, detection data are saved and a multi-platform interface of the linkage module is triggered to switch the running state of household equipment, or a trigger signal for switching the running state of the household equipment for the multi-platform interface is cancelled in response to a confirmation signal of a button
For furniture such as seats, sofas and the like with two side armrests, the application can respectively arrange the skin electric sensing unit 4 and the blood oxygen sensing unit 5 on the left side and the right side of the furniture, and can also only arrange one sensing unit on the single side of the furniture, the anode and the cathode of each sensing unit and the signal output port can be respectively connected with a control system through a built-in line of the furniture, and a detection model is constructed according to the following steps shown in fig. 6 to realize detection analysis of biological signal states:
Step 1: collecting original signals through a skin electric sensor and an oxygen saturation sensor, wherein the original signals are generally expressed by a logic signal through a PPS sequence;
step 2: the method comprises the steps of performing simple low-pass filtering noise reduction and baseline drift removal pretreatment on original data, and then determining each characteristic which is extracted and used for inputting an SVM emotion classifier to perform physiological state analysis when performing characteristic extraction on a skin electric sensing signal in an effective range and an blood oxygen concentration sensing signal in the effective range according to the following steps s 1-s 5:
step s1, acquiring a skin electric sample signal and an blood oxygen concentration sample signal under tired or negative emotion, performing low-pass filtering on the skin electric sample signal in an effective range and the blood oxygen concentration sensing signal in the effective range, and removing baseline drift to obtain a preprocessing signal;
step s2, extracting various types of features from the preprocessed signals to construct a feature set M;
step s3, for each feature in the feature set MRespectively carrying out standardization processing to eliminate individual differences of all the features, obtaining data samples with normal distribution, and constructing a training set according to the data samplesWherein x is E R D ,y∈{-1,+1}，x i For the ith data sample, N is the total data sample amount, D is the data sample feature number, R D For feature space, y ε { 1, +1} is a sample label indicating whether the data sample corresponds to a user being tired or in a negative emotion;
step s4, constructing an SVM emotion classifier, and training a data sample: to bring the effective characteristics into a set F IN Initializing to include all features in the training set, initializing the ordered set P to null, and setting the effective feature set F IN And the ordered set P iterates according to the following steps until the effective feature set F IN Is empty: using feature sets F IN Training an SVM classifier to obtain w and calculating an effective feature set F IN Cost function DJ (i) of each feature in (2), minimum min { DJ (i) }, i ε F IN Is shifted out to the ordered set P, the effective feature set F IN Updated to the remaining features, continuing with the updated active feature set F IN Performing iterative calculation on the sorting set P;
and step s5, adopting a stepping method for each feature in the sequencing set P, and screening out the feature which minimizes the overall Weibull Lambda in each step as the feature input into the SVM emotion classifier.
Wherein, the stepping method comprises the following steps: letting the independent variables try to enter the functional formula one by one, if the independent variables entering the functional formula meet the conditions, the independent variables remain in the functional formula, otherwise, the independent variables are eliminated from the functional formula. The application uses a stepping method, and the Wilker lambda value method can be specifically selected when the independent variable is screened: the method is the ratio of the sum of squares in the group to the sum of squares in the total group, and is used for describing whether the average values of all groups have significant differences, and when the average values of all observation groups are equal, the Wilks' lambda value is 1; when the intra-group variation is small compared to the total variation, it indicates that the component variation is large, it indicates that the inter-group variation is large, and the coefficient is close to 0.
SVM-RFE (recursive feature elimination) is a method for feature ordering based on a support vector machine (support vector machine, SVM). For the linear classification problem, the ideal objective function in performing feature selection is to calculate the expected value of the error, i.e. the error rate with infinite feature combinations. In SVM-RFE, this objective function is replaced by a cost function calculated on the training set. The basic idea of SVM-RFE is therefore to calculate the variation of the cost function DJ (i) due to the deletion of the feature x, in particular:
step 3: after determining various characteristics of blood oxygen saturation and skin electric reaction signals input into the SVM emotion classifier from the original signals, training the SVM classifier by utilizing data samples according to the following steps t1 to t3 to realize identification and detection of various characteristics in skin electric reaction signals and blood oxygen concentration sensing signals, and analyzing the tired state and emotional state of a user:
step t1, judging whether a feature set of a data sample is linearly separable, firstly mapping a feature space of the data sample to a high-dimensional linearly separable space through a kernel function under the condition that the feature set is linearly inseparable, and then jumping to step t2; directly jumping to the step t2 under the condition of linear separable;
Step t2, in the linear separable feature space R corresponding to the data sample feature set D Mid-construction hyperplane f (x) =w T ·x i +b toI.e. < ->As an objective function, with y i (w T x i +b) is 1 or more, i=1, 2, …, n is constraint solving +.>Obtaining a hyperplane f (x) =w that maximizes the minimum distance of the sample point to the hyperplane T ·x i +b, where a i A Lagrangian multiplier, b is a real number representing the distance between the hyperplane and the origin;
and t3, randomly dividing the data sample into K parts, training the SVM emotion classifier by using the data sample in a K-fold cross validation mode until the accuracy of the SVM emotion classifier in analyzing the physiological state of the data sample reaches 90%, and obtaining the SVM emotion classifier.
Step 4: the SVM emotion classifier obtained by training is practically applied in the manner shown in fig. 5:
step 401, waiting for 60 seconds to start the control unit obtained by training after the system is powered on and started;
step 402: each sensor collects the skin electric reaction and blood oxygen saturation signals of the user in real time, and screens out the sensing signals outside the filtering effective range in the mode shown in fig. 7; in fig. 7, all module errors are usually calculated for the situation that the skin electric sensing unit 4 and the blood oxygen sensing unit 5 are arranged on both sides of the handle sleeve, and the signal difference between the same sensing units on both sides is calculated, and for the situation that the skin electric sensing unit 4 is arranged on one side of the handle sleeve and the blood oxygen sensing unit 5 is arranged on the other side of the handle sleeve, the difference of each sensing signal relative to the normal range of the human body can be used as left and right errors;
Step 403: extracting the characteristics finally selected in the step 2 from the filtered signals;
step 404: bringing the features into the emotion recognition classifier trained in the step 3 to obtain emotion recognition and give a final result;
step 405: and according to the emotion recognition, giving a final result, triggering the linkage system to execute a corresponding adjusting mechanism, such as: when the user is identified to be in a tired state, the detection data are stored, and the multi-platform interface is triggered to reduce the environmental temperature set by the temperature control equipment, the illumination intensity of the illumination equipment and the environmental volume of the sound equipment and the entertainment system; when the negative emotion of the user is recognized, the detection data are stored, the multi-platform interface is triggered to adjust the environmental temperature set by the temperature control equipment to a dormant interval, only the low-illumination operation of the auxiliary lighting equipment is maintained, the entertainment system is closed, and the interactive voice system is closed; and when receiving the confirmation signal fed back by the user trigger button 2, the trigger signal for switching the running state of the household equipment to the multi-platform interface is canceled, so that each household equipment is restored to the previous normal running state.
In other implementations of the application, the system described above may also implement detection analysis of biological signal status in the following manner as shown in fig. 8:
Step 1: collecting original signals through a skin electric sensor and an oxygen saturation sensor;
step 2: the original data is subjected to simple low-pass filtering noise reduction and baseline drift removal pretreatment;
step 3: extracting various characteristics of PPS (pulse per second) aiming at blood oxygen saturation, skin electric reaction signals, skin temperature and skin temperature from the original signals, wherein the characteristics are shown as M as a characteristic set of an SVM classifier, and the characteristics are not limited to the characteristics of the following table in general, and other characteristics can be selected;
step 4: all features in M are standardized, and data comparability is improved. The normalization method is not limited to the following: polar error normalization, Z-score normalization, linear scale normalization, log function normalization, and arctangent function normalization.
Wherein, the calculation formula of the Z-score standardization method is as follows:wherein->For normalized data, μ and σ are the mean and standard deviation of the data, respectively. The processed data accords with standard normal distribution, the mean value is 0, and the standard deviation is 1.
Step 5: (SVM-REF) building an SVM classifier, training samples,
step 5-1 calculates a cost function for each feature, minimizing the cost functionThe features of the numbers are taken as output and put into a sorting set P, and the rest of the features are left as a valid feature set F IN With active feature set F IN The iteration is performed as input to the next ordering.
Specifically, for training setsWherein x is E R D ,y∈{-1,+1}，x i For the ith sample, N is the total sample amount, D is the sample feature number, R D The optimal classification hyperplane of the SVM is w T ·x i +b＝0
The solution for w is found as follows:
a i is a lagrange multiplier, and therefore, the cost function is:
DJ(i)＝w 2
in consideration of the correlation between various features, redundancy may occur, resulting in long model training time, overfitting, low recognition rate, and reduced model performance. Therefore, it is necessary to perform feature selection. In this embodiment, after the feature ordering by using the SVM-RFE-CBR method, in each iteration process, a step method is further adopted to perform feature selection in the following manner, so as to further reduce the correlation deviation: step 5-2, after each iteration in step 5-1 is completed, further adopting an SVM-RFE-CBR method to screen out the features highly related to the marking feature Q from the updated ordering assembly P: finding out the characteristic with the highest score in the output characteristic set generated by the iteration, and marking the characteristic as a marking characteristic Q; checking whether the left features have features with higher correlation with the marked features Q, if the features with higher correlation with the marked features Q are found, taking the marked features Q out of the ordered set P and putting the marked features Q back into the effective feature set F IN The method comprises the steps of carrying out a first treatment on the surface of the Otherwise, continue to put in P, continue to screen out the put-back adjusted effective feature set F IN And performing iterative calculation on the sorting set P.
Step 7: step 5-1 and step 5-2 are iterated until all features are put into the sorted set P.
Step 8: features are selected through a stepping method to serve as input of the emotion recognition classifier, features which minimize the overall Weirk Lambda in each step are selected to serve as input features of the emotion recognition classifier, and finally the features with greater importance to the anger and fatigue emotion recognition classifier are obtained through screening.
The SVM classifier is trained as follows.
Step 9: linearly time-sharing, creating a hyperplane in the feature space, and distinguishing samples of different emotions; the linear non-time-sharing method has the advantages that the computational complexity caused by dimension increase is reduced through a kernel function, and the characteristic space is projected to a high-dimensional space, so that the problem of non-linear separability in the original characteristic space is converted into the problem of linear separability in the high-dimensional space. Specifically, in the feature space R D Constructed hyperplane f (x) =w T ·x i +b, distinguishing samples. Prediction categoryRepresenting two emotional states of anger and fatigue, respectively.
The linear non-time-division method can specifically use one of the following kernel functions, but is not limited to, mapping samples to a higher dimensional space to make the samples linearly separable, a linear kernel function, a Gaussian kernel and a sigmoid kernel, and the rest steps are the same.
Step 10: maximizing the minimum distance from the sample point to the hyperplane, the objective function is:
s.t.y i (w T x i +b)≥1,i＝1,2,…,n
step 11: the method comprises the specific steps of randomly dividing sample data into K parts, taking K-1 parts as a training set, taking one part as a test set for detecting accuracy, repeating K times and averaging the results of the classification accuracy to obtain a single estimated value. And when the accuracy reaches 90%, finally screening after model training to obtain the features with larger importance on the anger and fatigue emotion recognition classifier, and otherwise, continuing training.
To this end, an SVM emotion classifier for ultimately analyzing the user's tired state as well as the emotional state may be trained. Using the classifier: collecting skin electric reaction and blood oxygen saturation signals of a user in real time at each sensor, screening out sensing signals outside a filtering effective range in a mode shown in fig. 7, performing simple preprocessing such as low-pass filtering, baseline drift removal and the like on the filtered signals, extracting each characteristic finally selected in the step 8 from the preprocessed signals, and then bringing each characteristic into the emotion recognition classifier trained in the step 11 to obtain emotion recognition and give a final result;
Step 406: and according to the emotion recognition, giving a final result, triggering and storing detection data, and triggering the multi-platform interface to execute an adjusting mechanism of the environmental atmosphere according to the current tired state and the emotion state. For example, when the user is identified to be in a tired state, the detection data are stored, and the multi-platform interfaces such as the intelligent gateway, the Bluetooth regulation terminal, the intelligent home total control platform and the like are triggered to reduce the environment temperature of the temperature control equipment in the home environment, reduce the illumination intensity of the illumination equipment, reduce the environment volume of the sound equipment and the entertainment system until the confirmation signal fed back by the user trigger button 2 is received to close the regulation mechanism, so that each equipment is restored to the original working state;
when the user is identified to be in a negative emotion, the detection data are stored, the multi-platform interface is triggered to adjust the environmental temperature set by the temperature control equipment in the home environment to a temperature interval suitable for dormancy of the user, only auxiliary lighting equipment such as night lamps, atmosphere lamps and the like are maintained to run in low illumination, the entertainment system is closed, the interactive voice system is closed, an environment which is easy to relax and can be less interfered by the outside is maintained for the user to rest, and the adjustment mechanism is closed until a confirmation signal fed back by the user trigger button 2 is received, so that the equipment is restored to the original working state.
In summary, the application integrates a physiological monitoring system in intelligent furniture, and designs a set of cross-platform intelligent home linkage system, which comprises the following components:
(1) When the skin surface temperature is measured through a blood oxygen temperature sensor, a photoplethysmography (PPG) technology is applied, according to the light absorption characteristics of oxyhemoglobin HbO2 and hemoglobin Hb on wavelengths of 600-1000 nm, PPG signals of HbO2 and Hb are respectively detected through red light (600-800 nm) and near infrared light (800-1000 nm), then corresponding ratios are calculated through program processing, so that blood oxygen saturation (SPO 2) is obtained, and the emotion state of a user can be more comprehensively reflected through characteristic values of various different types and dimensions;
(2) The system utilizes the skin electric sensor to measure the skin electric activity, can sensitively detect the sweating condition of the human body controlled by the sympathetic nervous system, and further more comprehensively and accurately realize the inference of psychological or physiological signs according to the characteristic that the skin electric activity changes along with the state of sweat glands in the skin;
(3) The system performs low-pass filtering, baseline removal and other preprocessing on various collected physiological signals, performs characteristic extraction on data in two states of anger/fatigue in the training process to obtain a PPS sequence, and extracts various characteristic types such as first-order difference, second-order difference, different frequency band energy distribution, high-low frequency band spectral power ratio and the like on preprocessed blood oxygen saturation (SPO 2), skin temperature (SKT) and skin electric signal (GSR) respectively, so that information carried in a sensing signal can be reflected more comprehensively, different individual differences can be effectively eliminated by utilizing a Z-score standardized processing mode, and more accurate and comprehensive characteristic information can be obtained;
(4) According to the application, the SVM technology is utilized to carry out training classification on the obtained multiple features, in the feature training process, the SVM-RFE-CBR method is utilized to avoid correlation deviation generated when a large number of highly-relevant features exist in the feature space, and the false feature ordering is avoided because the importance of certain features is underestimated, so that the SVM emotion classifier is more accurately trained, more accurate recognition results on emotional states such as tiredness, anger and the like are obtained, and the recognition effect of a control system under the same operation cost is effectively improved;
(5) The application can effectively monitor and record the physiological state of the user, automatically trigger each device in the home environment to provide environment atmosphere easy for the user to rest when the user is tired and tired, and regulate the emotion of the user through the home environment when the emotion of the user is negative; the data recorded by the application can also be analyzed to obtain the user emotion fluctuation statistical data so as to trigger the intelligent home system to further correspondingly adjust the operation mode according to the life cycle of the user. The intelligent home system can overcome the technical barriers that the traditional physiological monitoring system cannot recognize emotion based on skin electricity, and can overcome the defects that the existing intelligent home system needs a large amount of redundant data when detecting the physical and mental states of a user, the operation processing process is complex, the system operation efficiency is low, and an accurate detection result cannot be obtained rapidly. The application does not need to wear additional sensing devices in the aspect of structural design, has smaller equipment, is more friendly to users than the existing detection means, and can promote user experience.
The foregoing is a description of embodiments of the application, which are specific and detailed, but are not to be construed as limiting the scope of the application. It should be noted that it will be apparent to those skilled in the art that several variations and modifications can be made without departing from the spirit of the application, which are all within the scope of the application.
Claims (10)
1. A cross-platform smart home linkage system, comprising:
the furniture comprises a furniture body (1), wherein a handle sleeve (3) is fixedly arranged at the circumference of a handrail or a leaning position, and a plurality of spring pins are arranged on the surface covered by the handle sleeve (3) of the furniture body (1);
the device is characterized in that a skin electric sensing unit (4) and a blood oxygen sensing unit (5) are embedded in the handle sleeve (3), the blood oxygen sensing unit (5) is arranged on the outer peripheral surface of the handle sleeve (3), the skin electric sensing unit (4) is arranged on the inner peripheral surface of the handle sleeve (3), and a signal connecting wire and a power supply connecting wire of the skin electric sensing unit (4) and the blood oxygen sensing unit (5) are connected with a control system through a spring stitch respectively to output skin electric sensing signals collected by the skin electric sensing unit (4) and blood oxygen concentration sensing signals collected by the blood oxygen sensing unit (5) to the control system;
A button (2) arranged on the linkage module of the furniture body (1) and in communication connection with the control system, for feeding back a confirmation signal to the control system in response to a trigger of a user;
the control system is provided with:
the signal state detection module is used for receiving the skin electric sensing signals and the blood oxygen concentration sensing signals through signal connecting wires of the skin electric sensing unit (4) and the blood oxygen sensing unit (5) respectively, judging whether the sensing signals are in an effective range or not, and filtering sensing signals outside the effective range;
the physiological state analysis module is connected with the signal state detection module, receives sensing signals in an effective range, and recognizes the tired state and the emotional state of a user corresponding to the sensing signals through the SVM emotion classifier, and when the user is in tired or negative emotion, the physiological state analysis module stores detection data and triggers the multi-platform interface of the linkage module to switch the running state of household equipment;
the control system also responds to the confirmation signal of the button to cancel the trigger signal for switching the running state of the household equipment to the multi-platform interface.
2. The cross-platform smart home linkage system of claim 1, wherein the multi-platform interface connects lighting devices, temperature control devices, audio devices, entertainment systems, interactive voice systems in a home environment.
3. The cross-platform smart home linkage system of claim 1, wherein the handle sleeve (3) comprises:
the rear sleeve body is fixedly arranged at the rear side of the furniture body (1), a blood oxygen sensing unit mounting groove (305) is formed in the rear side wall of the rear sleeve body, and a blood oxygen sensing unit (5) is embedded in the blood oxygen sensing unit mounting groove (305);
the front sleeve body is fixedly arranged at the front side of the furniture body (1) and fixedly connected with the rear sleeve body, and the front sleeve body and the rear sleeve body are surrounded and fixed on the periphery of the furniture body (1);
a split mounting groove (304) is further formed between the front sleeve body and the rear sleeve body, and the skin electric sensor (4) is embedded into the split mounting groove (304);
the skin electric sensor (4) and the blood oxygen sensing unit (5) are at least partially exposed on the surface of the handle sleeve (3).
4. A physiological state monitoring method based on skin electricity and blood oxygen signals, which is used for the cross-platform intelligent home linkage system as claimed in any one of claims 1 to 4, and is characterized by comprising the following steps:
after power-on is started, respectively receiving skin electric sensing signals acquired by a skin electric sensing unit (4) and blood oxygen concentration sensing signals acquired by a blood oxygen sensing unit (5), judging whether the sensing signals are in an effective range or not, and filtering sensing signals outside the effective range;
And extracting features of the skin electric sensing signals in the effective range and the blood oxygen concentration sensing signals in the effective range, inputting the extracted features into an SVM emotion classifier for physiological state analysis, identifying the tired state and the emotion state of a user, and storing detection data and triggering a multi-platform interface of a linkage module to switch the running state of household equipment when the user is in tired or negative emotion.
5. The method for monitoring physiological states based on galvanic skin and blood oxygen signals according to claim 4, wherein when the characteristic extraction is performed on the galvanic skin sensing signals within the effective range and the blood oxygen concentration sensing signals within the effective range, the specific extracted characteristics for inputting the physiological states into the SVM emotion classifier are determined by the following steps:
step s1, acquiring a skin electric sample signal and an blood oxygen concentration sample signal under tired or negative emotion, performing low-pass filtering on the skin electric sample signal in an effective range and the blood oxygen concentration sensing signal in the effective range, and removing baseline drift to obtain a preprocessing signal;
step s2, extracting various types of features from the preprocessed signals to construct a feature set M;
Step s3, respectively carrying out standardization processing on each feature in the feature set M, eliminating individual differences of each feature, obtaining data samples with normal distribution, and constructing a training set according to the data samplesWherein x is E R D ,y∈{-1,+1}，x i For the ith data sample, N is the total data sample amount, D is the data sample feature number, R D For feature space, y ε { 1, +1} is a sample label indicating whether the data sample corresponds to a user being tired or in a negative emotion;
step s4, constructing an SVM emotion classifier, and training a data sample: to bring the effective characteristics into a set F IN Initializing to include all features in the training set, initializing the ordered set P to null, and setting the effective feature set F IN And the ordered set P iterates according to the following steps until the effective feature set F IN Is empty: computing an active feature set F IN The cost function of each feature in the list is shifted out to the sorting set P, and the effective feature set F IN Updated to the remaining features, continuing with the updated active feature set F IN Performing iterative calculation on the sorting set P;
and step s5, adopting a stepping method for each feature in the sequencing set P, and screening out the feature which minimizes the overall Weibull Lambda in each step as the feature input into the SVM emotion classifier.
6. The method of physiological state monitoring based on galvanic skin and blood oxygen signals according to claim 5, wherein the effective feature set F is further determined in step s4 according to a cost function each time, respectively IN After iterative updating is carried out on the sorting set P, the SVM-RFE-CBR method is further adopted to screen out the features highly related to the marking features Q from the updated sorting set P, and the screened features are adjusted to an effective feature set F IN To continue to screen out the adjusted effective feature set F IN Performing iterative calculation on the sorting set P; wherein the marking characteristic Q is according toThe cost function updates the highest scoring feature in the resulting sorted set P.
7. The method for monitoring physiological states based on galvanic skin and blood oxygen signals according to claim 4 or 5, wherein the SVM emotion classifier is trained to obtain:
step t1, judging whether a feature set of a data sample is linearly separable, firstly mapping a feature space of the data sample to a high-dimensional linearly separable space through a kernel function under the condition that the feature set is linearly inseparable, and then jumping to step t2; directly jumping to the step t2 under the condition of linear separable;
step t2, in the linear separable feature space R corresponding to the data sample feature set D Mid-construction hyperplane f (x) =w T ·x i +b toAs an objective function, with y i (w T x i +b) is 1 or more, i=1, 2, n is constraint solving ∈1>Obtaining a hyperplane f (x) =w that maximizes the minimum distance of the sample point to the hyperplane T ·x i +b, where a i A Lagrangian multiplier, b is a real number representing the distance between the hyperplane and the origin;
and t3, randomly dividing the data sample into K parts, training the SVM emotion classifier by using the data sample in a K-fold cross validation mode until the accuracy of the SVM emotion classifier in analyzing the physiological state of the data sample reaches 90%, and obtaining the SVM emotion classifier.
8. The method of claim 4, wherein the step of low pass filtering the galvanic skin sensor signal and the blood oxygen concentration sensor signal to remove baseline wander is performed before the step of feature extraction of the galvanic skin sensor signal and the blood oxygen concentration sensor signal.
9. The method for physiological state monitoring based on galvanic skin and blood oxygen signals according to claim 5, wherein the features extracted from the pre-processed signals for constructing feature set M comprise any one or any combination of the following:
The ratio of low frequency and high frequency energy FFT_ratio of the pulse count PPS per second, the FFT absolute value of the pulse count PPS per second is fully integrated with L1_Norm, the first order differential of the pulse count PPS per second is the maximum value diffPP_max, the mean value SPO2_mean of the original blood oxygen saturation signal, the standard deviation SPO2_sd of the original blood oxygen saturation signal, the mean value SPO2_ afd _mean of the first order differential absolute value of the original blood oxygen saturation signal, the mean value SPO2_asd_mean of the second order differential absolute value of the original blood oxygen saturation signal, the mean value SPO2_std_ afd _mean of the first order differential absolute value of the normalized blood oxygen saturation signal, the mean value SPO2_std_asd_mean of the second order differential absolute value of the normalized blood oxygen saturation signal,
the signal mean MSKT of the skin temperature original signal, the variance STD of the skin temperature original signal, the energy data SP_L of the skin temperature signal low frequency band, the energy data SP_H of the skin temperature signal high frequency band, the ratio ER_FFT of the spectrum power between the skin temperature signal high frequency band and the skin temperature signal low frequency band,
the average dropping rate AOND of the skin electric signal during decay, the number SRNOLM of local minimum values of the skin electric signal, the energy data FP of the skin electric signal in the frequency band of 0-2.4 Hz, the spectrum power SP of the skin electric signal in the frequency band of 0-2.4 Hz, the positive and negative zero-crossing rate Pnum of the low-frequency signal SR in the skin electric signal and the positive and negative zero-crossing rate Nnum of the very low-frequency signal VSR in the skin electric signal.
10. The method for monitoring physiological conditions based on galvanic skin and blood oxygen signals according to any one of claims 1 to 9, wherein the detection data is stored and triggers the multi-platform interface to decrease the ambient temperature set by the temperature control device, decrease the illumination intensity of the illumination device, decrease the ambient volume of the audio device and entertainment system when the user is identified as being in a tired state;
and when the negative emotion of the user is recognized, the detection data are stored, the multi-platform interface is triggered to adjust the environmental temperature set by the temperature control equipment to a dormant interval, only the low-illumination operation of the auxiliary lighting equipment is maintained, the entertainment system is turned off, and the interactive voice system is turned off.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310777702.XA CN116859761A (en) | 2023-06-28 | 2023-06-28 | Cross-platform intelligent home linkage system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310777702.XA CN116859761A (en) | 2023-06-28 | 2023-06-28 | Cross-platform intelligent home linkage system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116859761A true CN116859761A (en) | 2023-10-10 |
Family
ID=88226073
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310777702.XA Pending CN116859761A (en) | 2023-06-28 | 2023-06-28 | Cross-platform intelligent home linkage system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116859761A (en) |
-
2023
- 2023-06-28 CN CN202310777702.XA patent/CN116859761A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US12004879B2 (en) | Systems and methods for detecting photoplethysmographic device usage | |
| Khan et al. | Biometric systems utilising health data from wearable devices: applications and future challenges in computer security | |
| WO2020119245A1 (en) | Wearable bracelet-based emotion recognition system and method | |
| US9380982B2 (en) | Adaptive alarm system and method | |
| CN108577830A (en) | A kind of user oriented sign information dynamic monitor method and dynamic monitor system | |
| CN115881305B (en) | Method, system and device for detecting, quantifying and assisting in intervention of sleep stability | |
| CN109859843A (en) | Intelligent health all-in-one machine | |
| CN110074765A (en) | A kind of method of discrimination of SAHS screening and sleep stage | |
| CN116564561A (en) | Intelligent voice nursing system and nursing method based on physiological and emotion characteristics | |
| CN105808959A (en) | Motion detection system, motion detection terminal and cloud platform | |
| CN116440425B (en) | Intelligent adjusting method and system of LED photodynamic therapeutic instrument | |
| CN115191961B (en) | Cardiopulmonary health detection method and device, wearable equipment and storage medium | |
| CN115670419A (en) | Data processing method and system based on motion information predicted by smart watch | |
| KR20070087787A (en) | Emotion classification system of real time bio signals and emotion classification method | |
| CN109394203A (en) | The monitoring of phrenoblabia convalescence mood and interference method | |
| CN109998497A (en) | System and plane of illumination illumination testing apparatus are sentenced in inspection of falling asleep in luminous environment | |
| CN115952460A (en) | Emotion recognition method and device, computer equipment and storage medium | |
| CN109893113A (en) | Information processing method based on BCG and HRV technology | |
| CN117814766B (en) | Near infrared spectrum physical and mental pressure and sleep quality monitoring system | |
| US20230263400A1 (en) | System and method for filtering time-varying data for physiological signal prediction | |
| CN116859761A (en) | Cross-platform intelligent home linkage system | |
| CN117116480A (en) | Health data monitoring and analyzing system for aged people | |
| CN115317304A (en) | Intelligent massage method and system based on physiological characteristic detection | |
| CN110197186B (en) | PPG-based illumination comfort level measurement method and system | |
| US20200126402A1 (en) | Individualized control system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |