Embedded sleep monitoring alarm and design

1, introduction

Normal person's pulse and heartbeat is consistent, pulse frequency by age and sex of the baby every minute 120-140 times, children 90-100 per minute, school children every minute at 80-90, adults 70-80 per minute.

In addition, movement and emotion can be increased to pulse, rest, sleep is the pulse of slowing down. Into personal rate more than 100 times per minute, called tachycardia; less than 60 times per minute, called bradycardia. Clinic has many diseases, especially heart disease can make the pulse of change. Therefore, measuring pulse on patient is an indispensable checks, TCM will slap and pulse as the primary method of diagnosis and treatment of disease. Chooses the cure is in the medicine of an infinite knowledge. Healing through chooses is many old Chinese medicine set life even generations the tireless efforts of the accumulation of extremely valuable experience. Sleep disorder is formation and development of many diseases, sleep difficulties is insomnia in one of the most common causes of the phenomenon, but its mostly because you cannot suppress long stretches constantly thought, how to effectively monitor people's sleep control sleep "thoughts", improve sleep quality, at the same time as the sudden illness of early diagnosis and prediction provides reliable protection, it is worth it to pay particular attention to. With the development of science, the pulse beat of microprocessor by sensor is accurately displayed and logged. Through comparative analysis of the normal pulse rate and pulse measurement object allows you to quickly and accurately obtain the physiological measurement of objects, in case of danger can also alert you to related personnel. The study builds on the multidisciplinary cross-penetration is, on the basis of computer technology, biomedicine and signal processing techniques for the perfect combination.

2, PVDF films pulse sensor

PVDF films while the pulse sensor for pulse measurement requirements, but there are many shortcomings: ① correlation with location.

Because of the pressure sensor, the sensor accurately positioned on a surface of pulse wave, can have the arterial pulse signal output; ② sensor sensitivity, there will be on fretting signal amplification effect, so the body of the random movement is easy to introduce noise; ③ sensor casing for metal copper, direct contact with the skin will introduce 50Hz frequency interference. Therefore the sensor output analog signal and does not properly reflect the body caused by arterial pulse wave, signal processing for certain.

Figure 1 Schematic of hydraulic capsule

In response to the above, use the "liquid delivery pressure" nature, designed a rubber quality closed fluid SAC, Figure 1 is a sketch of the hydraulic capsule.

The capsule form is flat ribbon, rubber thickness for 0.5mm, cavity thickness 2mm, figure A Department to install sensors and sensor induction surface affixed to the wall, capsule capsules down wall close to the skin, use rubber elasticity make capsules and wrist skin slightly increased sensor induction area. From both ends of the B Division into the wrist strap that can be worn on the wrist sensors, and hydraulic capsule as long as they have contact with the area and the wrist, radial pulse fluctuation will be in liquid transmission to sensor induction surface, addresses the relevance of the sensor position. Gasoline engine oil used as a pressure medium, experimental discovery: physical characteristics of viscous oil, some slight transient wave is not easy to be conductive, played a antijamming and the role of the low pass filter; oil heat, the temperature varies with rongda external changes did not significantly enhances wearing comfort. Oil sticks and inertia is large, the skin background randomly have a certain role, while blocking the pulse of delivery also has some attenuation, the actual measurement found on a wave period of "strong" and "weak" attenuation, the issue is reflected in the pulse of the form in the bottom of the graph from the larger. This problem can be corrected using software. In addition, no stimulation for skin thermal conductivity of insulating rubber manufacture hydraulic capsule, cut off the sensor metal shell and direct contact with the skin, 50Hz interference also was divided subject. Thermal conductivity at the wrist body surface temperature transmission possible, increased the measurable physiological signals.

3, sensor signal processing

Pulse sensor output signal is too weak, it must enlarge after output to ADC components.

According to the pulse of the human body in motion, the highest runout times after 240 times/min computing to design a low-pass amplifier by IC2A and C04, etc, as shown in Figure 2. Turn the frequency by R07, C04, R08 and C05, magnification by a ratio of R08 and R06.

Figure 2 low-pass filter graphs

According to the second-order low-pass filter transfer functions: the available magnification:, 0.707 times zero frequency gain calculation of transition frequency: high frequency, the highest on the pulse of the human body, low-frequency performance is satisfactory.

In order to improve the accuracy of measurement, you can use the circuit in Figure 3.

Pulse signal analysis of object is a signal, in particular non-stationary and transient signals, signal analysis method using time-frequency analysis, its main task is to describe the content of the frequency spectrum and how they change over time, so that in time and frequency at the same time represents the energy and strength of the signal, this distribution will be on various signal analysis and processing, extracting signal characteristics contained in the information, as well as comprehensive get video with expectations of the distribution of signals.

Here, using time-domain analysis.

Figure 3 secondary amplifiers and comparators

Time domain analysis is analysis of pulse wave amplitude of height and pulsating, is currently the most widely used an analysis method, in the past, most used in the study of artificial reading figure, that is, using measuring tools directly read out the pulse of the height of the wave, Gorge, the corresponding value, when pulse parameters such as area.

By pulse amplitude and time of the analysis, you can learn about the frequency of the pulse, rhythm and pulse power of strong or weak, pulse of the potential and the pulseLike shape characteristics. It extracts some pulse curve have obvious physiological significance of points as the evaluation of the pulse wave of feature points, intuitive, and easier for clinicians.

Due to the efficient Matlab mathematics computing power and perfect function, in this study, using mathematical tools, Matlab will be collected to process the signal directly from the map shown in Figure 4 and Figure 5, abandon the process of artificial reading figure, estimated by the experience of the error.

Figure 4 health human pulse wave waveform

The healthy human pulse wave-domain map is visible, the main waves (P-wave), Bo wave (D) clear, no contraction of strings like the high-frequency, identified as healthy people of pulses.

By the time graph is visible in patients with hypertension, primary wave is flat and wide, heavy beat wave and wave fusion to the spacious main square, falling in the gap elevation, heavy beat wave flat, identified as hypertension of chord pulse, and should be a medium-high blood pressure.

Figure 5 hypertensive patients pulse wave waveform

Sleepless-domain map of pulse wave is between these two situations, between the clinical judgement on easier.

4, design ideas

In order to resolve the inhibition constant thoughts leads on the issue, we can let the computer based on the State of the brain as a way to continue to provide appropriate incentives to stimulate uncontrolled source of thought, it may be due to the continued transfer of attention and is inhibiting state until final deactivation.

Machine-brain interfaces is a computer-monitoring, control, to intervene in the brain thinking activity of the system. Its role was to breaks from sober to sleep this transition process out-of-thinking activities. These features can be "micro-embedded computer system".

Figure 6 machine-brain interface structure diagram

Machine-brain interfaces is a computer monitor, control, to intervene in the brain thinking activity of the system, its role was to breaks from sober to sleep this transition process out-of-thinking activities.

In order to achieve this function, you first need an interrupt source, here a stimulator to produce sound stimulation as an interrupt source; stimulus, the consumer's response to stimulation by sensor transforms it into electrical signals; weak signal from the sensor through amplification, filtering into a variety of embedded computer systems. Experiencing sudden disease (such as cerebral thrombosis) embedded computer systems can be remote-controlled "sound" alarm or "voice" phone alarm to inform family and emergency Center. Machine-brain interfaces system architecture diagram shown in Figure 1.

5, hardware design

S3C440X is Samsung's ARM7TDMI core based RISC reduced instruction system for 32-bit high-speed processor.

Its working voltage 3.3V, the kernel's ARM7TDMI working voltage of only 2.5V, low-power chip, and more conducive to the development of a wearable products.

S3C44B0X implementation sleep monitoring hardware connections as shown in Figure 7.

Figure 7 sleep detection hardware connection diagram

Sound alarm and telephone voice alarm circuit common, here not in here.

6, software designing

Linux is a very popular operating system, it is compatible with UNIX systems, open source.

It was originally designed for a desktop system, now widely used in the server field. And greater impact in that it was gradually applied to embedded systems, uClinux it is in this atmosphere, it is Micro-Control-Linux, literally understood as "areas for micro-control and design of Linux system". Specific system transplant do details here, here, give the system's software control block, as shown in Figure 8

Figure 8 software control block diagram

7. conclusions

The pulse sensors based on thin-film sleep monitoring alarm system research, research on sleep on the basis of the transitional stage, highlight the sleep if there is a sudden onset of heart and brain diseases in a timely manner to achieve sound alarm or telephone voice alarm functionality.

Past sleep research mainly to monitor the project's main innovation lies in the following two ways.

1, using dual sensor structure.

Measurement monitoring sleep EEG in the heart of the measured object to test the cognitive responses accomplished on mild sleep wake up, on a serious illness in patients with sleep in patients with cognitive testing.

2, sound alarm and telephone voice alarm.

In the heart of the sleep state on the basis of the EEG study, if sleep to appear when the sudden heart and brain diseases in a timely fashion show heart EEG and embedded processors processing triggered wireless alarm time of alarm.

Reference documents

[1] hcma000080 UK sleep monitoring of control systems [J] 2006 Chinese clinical rehabilitation; 10 (13): 119-121 [2] Yu mengsun Zhang Hong Kim. Sleep medicine monitoring of new model [J] Chinese therapy equipment information, 2003, 9 (3): 4-6 [3] big wave embedded system design and application of the principle, [M] Beijing: China machine press 2005: 448-449 [4] Zhang Yong sleep monitoring system [G], South China University of science and a master's degree thesis 2005: 27-28 [5] sinus zhenzhong, song-Peng, Frank Lee. embedded system design of the new development and challenges [J

]. The microcontroller and embedded system applications, 2004, (12): 15-18 [6] forest regulation Mau. Sleep transition process monitoring of new method [J]. China medical device information, 2005, 11 (3): 29-32 [7] Jiang Hu Yu-lun Tan,, Zhang Wei. Pulse measurement hardware design [J]. Medical device information, 2005, 20 (7): 1-3 [8] Yuan Kai-ming. Sleep monitoring system [J]. Healthcare devices, 2006, 3: 1-2