SUMMERY OF THE UTILITY MODEL
Not enough to prior art exists, the utility model provides a personal identification and vital sign monitoring car seat based on microbend optical fiber sensor, even the driver wears gloves, also can carry out personal identification and vital sign parameter monitoring. Different from other identification and vital sign parameter monitoring devices, the utility model discloses the scheme can provide continuous, do not have sensation identification and vital sign parameter measurement. When a driver sits in a cab and is electrified, a stable and continuous ballistocardiogram is obtained through a device on a seat, the ballistocardiogram is used for identity recognition, the safety risk that an automobile is stolen can be effectively avoided, meanwhile, the Heart Rate Variability (HRV) of the ballistocardiogram can be calculated, the mental state of the driver can be evaluated according to the HRV, whether the driver can drive to get on the road or continue driving in the current state is judged, and the dual protection of driving safety and property safety is achieved.
The utility model discloses specifically adopt following technical scheme:
an intelligent car seat, comprising: the device comprises a microbend optical fiber sensor, a processor and a seat body; the microbend fiber sensor includes: the optical fiber module comprises a light source, a multimode optical fiber, a photoelectric detector and a deformer arranged on the periphery of the multimode optical fiber; the sensing part consisting of the multimode optical fiber and the deformer is embedded and distributed on the seat body; the photoelectric detector is connected with the processor; the processor is connected with the automobile circuit and the control system.
Preferably, the light source adopts a light emitting diode or a laser light source, and the photoelectric detector is matched with the working waveband of the light source.
Preferably, two side surfaces of the deformer are provided with equidistant sawtooth structures, and the multimode optical fiber is sequentially wound along the sawtooth grooves on the two sides to form a microbend optical fiber structure.
Preferably, the processor is connected to an inertial measurement unit fixed to the vehicle.
Compared with the prior art, the utility model discloses and preferred scheme's beneficial effect is:
(1) under a set of scheme framework, the driver identity can be identified, and the mental state of the driver can be monitored in real time. The apparatus and method work effectively regardless of whether the driver's hands are gloved or not;
(2) the identity recognition scheme is accurate, reliable and easy to operate, and the reliability of the identity recognition scheme can be further improved through deep learning of the convolutional neural network.
(3) The cost is low, and the application and the popularization are easy.
Detailed Description
In order to make the features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in detail as follows:
to achieve the purpose of the present invention, the apparatus of the present embodiment includes: microbend fiber optic sensors, processors, and seat body 5.
As shown in fig. 3 and 4, the microbend fiber sensor includes: the optical fiber detection device comprises a light source 1, a multimode optical fiber 2, a photoelectric detector 3 and a deformer 4 arranged on the periphery of the multimode optical fiber 2; a sensing part consisting of the multimode optical fiber 2 and the deformer 4 is embedded and distributed on the seat body 5; the photoelectric detector 3 is connected with the processor; the processor is connected with the automobile circuit and control system and the inertia measuring unit fixed on the automobile.
The light source 1 may be a light emitting diode or a laser light source 1 or other broadband or narrowband light sources 1, and the main point is that the photodetector 3 is matched with the operating band of the light source 1.
As shown in fig. 4, in the present embodiment, both side surfaces of the deformer 4 have saw-tooth structures with equal distances, and the multimode optical fiber 2 is sequentially wound along the saw-tooth grooves on both sides to form a microbend optical fiber structure.
One end of the multimode optical fiber 2 is connected with the light source 1, and the other end is connected with the photoelectric detector 3. The device of the embodiment realizes the acquisition of the ballistocardiogram by utilizing the characteristic that the microbending optical fiber structure generates different optical fiber energy losses under different applied pressures: because the heart beats and produces the effort to the human body and makes the human body take place complicated and regular mechanical motion, when the driver sits on this device, weight can exert pressure to the microbend optical fiber structure for multimode fiber 2 in the microbend optical fiber structure takes place the extrusion, makes the energy of light in the transmission course take place the loss, thereby can receive the light signal that has driver's sign information in photoelectric detector 3. The received optical signal is demodulated to form the ballistocardiogram signal.
By the above-provided apparatus, the present embodiment correspondingly designs a set of workflow, which includes the following steps:
step S1: after the automobile is powered on, the microbend optical fiber sensor acquires ballistocardiogram information of a driver; the processor confirms the identity information of the driver through the characteristics of the ballistocardiogram information and then executes the step S2;
step S2: and when the abnormal condition of the ballistocardiogram information of the driver is detected, sending an alarm prompt and starting an emergency response. If the driver' S ballistocardiogram information is detected to be normal, executing step S3;
step S3: unlocking an ignition starting system of the automobile and acquiring ballistocardiogram information of a driver in real time;
in step S2, the HRV (heart rate variability) of the driver is calculated from the ballistocardiogram information, and the mental state of the driver is evaluated from the HRV.
In steps S1-S3, the inertia measurement unit is used to correct the vibration parameter.
In steps S1-S3, the processor uses the peaks and valleys of the ballistocardiogram as feature points for confirming driver identity information.
In steps S1-S3, the processor performs driver identification using a convolutional neural network.
In step S2, the emergency response includes turning on the warning lights and brakes of the car.
As a more specific implementation manner, fig. 1 is a schematic diagram of the working flow principle of this embodiment. After a driver gets on the vehicle and turns on a power supply, the device on the seat is activated, after activation, data detected by the optical sensor can be sent to the processor serving as the terminal in real time through the vehicle-mounted network, then the terminal processor calls vehicle owner information on the vehicle to perform identity matching and mental state assessment, and feeds the information back to the vehicle, and if the identity information is matched and the mental state of the driver meets the requirements, the driver can drive the vehicle. Moreover, as shown in fig. 1, the detection mechanism is not only used when the driver is about to drive, but also can continuously monitor in the driving process, when the driver has problems in the long journey, the driver can be reminded to have a rest at the side in time, so that the driver can effectively avoid fatigue driving, and when an emergency situation occurs, even when the driver has a situation that the vehicle cannot be controlled, the risk is reduced to the minimum automatically through an emergency response and alarm mechanism.
When using for the first time, the driver inputs own ballistocardiogram information and personal information in advance, then uses next time the utility model discloses just can carry out authentication during the device.
Fig. 5 is a schematic diagram of the waveform characteristics of a ballistocardiogram according to an embodiment of the present invention. Ballistocardiograms are generated for every living individual, and different individuals are unique due to factors such as heart position, size, age, height, weight, sex, and the like. The identity of the driver can be accurately distinguished by means of the ballistocardiogram. The ballistocardiogram reflects the heart beat versus time and therefore contains information about the heart characteristics of a particular subject. As can be seen, a typical ballistocardiogram signal contains a series of peaks
And trough of wave
Each peak to trough is of different duration and amplitude. The embodiment defines the feature points according to the characteristics, and can properly adjust the recognition result according to the range of the heart rate value.
In order to identify, the driver must record his or her ballistocardiogram into the system of the present embodiment in advance as template information. When the device is used next time, the system automatically identifies and authenticates the identity by acquiring the current ballistocardiogram and comparing the ballistocardiogram with the template library. Unlike static detection, ballistocardiogram acquisition in an automotive environment inevitably involves some other limb movements by the driver while driving. At this point, the acquired ballistocardiogram is necessarily disturbed to varying degrees by these limb movements and car vibrations. In the embodiment, a measuring mode of fusing a ballistocardiogram sensor and inertial measurement is adopted for carrying out biological identification and vital sign parameter measurement. That is to say, the signals acquired by the inertial measurement unit are used for assisting the original ballistocardiogram to carry out filtering processing, and relatively pure heart beating signals are extracted for identity recognition and vital sign parameter measurement.
As a preferred solution, the present embodiment uses a convolutional neural network to construct an identity recognition framework. Different from a general classification algorithm, the convolutional neural network has strong adaptability and learning. Even if the ballistocardiogram of the user changes along with the time, the framework of the embodiment can also learn and adjust parameters in a self-adaptive manner, and the identification precision is improved.
The present invention is not limited to the above preferred embodiments, and all other various types of intelligent car seats can be obtained by anyone who can obtain the teaching of the present invention.