SUMMARY OF INFRARED FALL DETECTOR PROJECTOR

The whole project was done within five weeks, by Tianqiong Du, Guotao Dong, Hanyi Fei and Yuxiao Fang, and supervised by Dr.Yan. we met different problems during this period, such as connection problem, design problem and signal process problems. By brainstorm and dedication, we solved various problems and finally finish this project satisfying result.

Through correcting the design circuit for several times, the final system contains one sensor RE200B-P, two operational amplifiers used to amplify the signal voltage, one operational amplifier used as analog to digital converter and an Arduino board used to idealize the signal, a PC used to monitor the motion situation.

The final result is shown in the video below: when there is no fall motion occurs, the buzzer stay silent, when there is a fall motion happens, the buzzer beats. and the effective detection range of this detector is within 1 meter, the video can be found in our week 5 blog.

There are also some limitations of this infrared fall detector. if it is turned on too long, it goes out of order; and its detection range is not wide enough. this limitations might be solved in further research.

28th FEB - The Sixth Lab - Test different movement form

This lab time, we combined the 'Arduino' board to control the signal. As I mentioned in the last blog , the signal is not stable when sensor detects movement. 'Arduino' board can completely solve this problem. This is code.

Figure 1

'a' and 'b' control the upper circuit and lower circuit delay time respectively. Meanwhile, 'm' and 'n' control the voltage of upper circuit and lower circuit respectively. When delay time is greater than 30, the generated low voltage is in the steady state which means no movement pass that sensor. Conversely, if delay time is less than 30, this low voltage is in the unstable high voltage range which means it can be ignored. Figure 2 shows the relevant range. 

Figure 2

After that, fall detection can be detected.

26th FEB - The Fifth Lab - Test different movement form

This week,we had two chances to do the experiment. On Wednesday, we check the circuits again and finally we found the problems. First,we ignored that LED would interfere the waveform. Second,it is quite embarrassed to mention that. After we change the 'coupling' button from AC to DC, the perfect square waves show on the oscilloscope. But another problem comes out.The signal is very unstable when we connected the output signal to buzzer which shows like that. Yellow line represents the lower sensor.Blue line represents the upper sensor.

Figure 1 Fall down

Figure 2 Fall down and then stand up

Figure 3 Fall down and then stand up and then pass by

The problem is that the buzzer will alarm on two conditions when people pass by and fall down because this two circuits cannot synchronize completely. Therefore,on Friday we planed to add 'Arduino' Board to control the signal to make it more stable.

21st FEB - The Forth Lab - Modify Actual Circuit & Test

Last week, we have completed two complex circuits. Then, modifying and simplifying circuits is becoming increasingly necessary and urgent.

 However, there are a few issues during the regulation. For example, third part circuit did not work after beautifying the circuits with shorter wires. LED light kept on and the sensing system, which was capable of detecting moving objects, had no reaction shown on the oscilloscope. After brainstorm, members agree that amplifier circuits had lower magnification times than required. Before conducting the modification, our group’s plan was made to adjust the value of spiral rheostats to generate better waveform as expected, including additional detail works.

 In addition, the fourth circuit could not generate the digital square wave. Depending on the design, after the successful configuration of both systems, the next step is the actual synchronization. However, the second flash was not stable. As a result, the combination wave did not get the perfect square wave. The reason might be systems had poor contact between components and circuit boards. Then, we used another circuit board to supplement the circuit. Luckily, it worked then.

This week, most of the work had been done. All that’s left is to make certain that there are no potential issues.

14th FEB - The Third Lab - Modify Actual Circuit & Test

Last week, we didn't get any signal from the output terminal, which upsets us a lot. this week, we used one entire morning on the actual circuit. this time, we changed our strategy, we build the circuit in the sequence of its function and tested it one by one. The final circuit in figure 1. The first two operational amplifiers are used to amplify the signal, the third operational amplifier is used to turn analog signal into binary form, the fourth operational amplifier is used as compactor. 

figure 1

The reaction of our first stage operational amplifier is shown in the video below, it can be seen that when a palm is swept above, there is a signal from its terminal, the output is too tiny to be used, and the reason is about the gain of this amplifier, which can be changed by the variable resistor.
video of stage 1:

the reaction of the second stage operational amplifier is shown in the video, it can be seen that its reaction towards a sweeping palm is larger in amplitude compared to the first stage amplifier.
video of stage 2:

the reaction of the third stage operational amplifier is shown in the video below, it converts the analog signal to digital signal, it can be seen from the video that when a palm sweep out, there is a '1' in the output, otherwise it is '0'. moreover, the output signal is also related to the velocity of the sweeping palm, it a palm swept quicker, there is more '1' in output with shorter time period.
video of stage 3: 

the reaction of fourth amplifier is a DC voltage about 2-3 V, and its function is to stablize the system and make the output signal more smooth.
the video below shows the result when a LED is connected to the output of third stage, it proves to be sensitive and stable.
After this, we began to do some test about the sensor's working range, the result is that , at about two meters away, and within the range of 60 degrees, the sensor can detects people's motion sensitively. And this result is really satisfying.
next week, we will build another PIR sensor circuit, and calibrate the two circuit with same amplification effect and response time, and put them vertically so that we can detect the motion of fall, hopefully we can finish our job next week.  

7th FEB - The Second Lab - Build Actual Circuit & Test

As we had simulated the circuit we are going to use in the experiment in the first lab, this week, we are going to do something on the breadboard to build our actual circuit.

According to the supervisor's suggestion, we were supposed to divide the circuit into different parts which have different functions, so, our strategy is that, building the circuit firstly and test different parts of it step by step.

Before building the circuit, we firstly tested the PIR sensor according to its specification sheet and we build a simple amplification circuit to test the function of LM358 operational amplifier, according to the table below, again it proves the importance of the amplification circuit, without it, the response of the PIR sensor is too tiny to be detected. After mastering how to connect and drive them, we started to build the circuit.

Source Voltage when no heat movement in front of sensor (V)	Source Voltage when heat movement in front of sensor (V)
0.79	0.72

As can be seen in figure 1, we built the complete circuit and checked it carefully to ensure there is no mistakes in the connection.

After building the circuit, we started to test it step by step, first of all, we tested the first amplification circuit, with a sine wave signal as input, however, the output is not as our expectation, we tried to do some modifications to our circuit for several times but still, no satisfying result.

The procedures above took us an entire day, and obviously, we have already been after our schedule, we have to speed up and find out the problems in next week's lab time. 

Our strategy is to build the circuit step by step again and this time, we will pay specific attention to the connection of LM 358, the value of resistors and capacitors to make sure all these elements are correct. Hopefully we can finish the circuit in the morning section of next lab and we can do some more tests in the afternoon, for example sensing angel, sensing distance, response time.

31th JAN - The First Lab - Research & Simulation

On the first week, we mainly did two things including searching relevant circuit information on the internet and simulating the waveform using Multisim. Finally, we decided using following circuit diagram to connect our circuit.

figure 1

For simulation part, we tested every part respectively and using sine wave as input signal.The frequency and amplitude were adjusted to 1 KHz and 4V. We found that the first amplifier on the left hand side is used to amplify the signal which shows in figure 2.

figure 2

The red line represents the output waveform of the first amplifier and green line represents the input waveform. The second amplifier on the left hand side is used to clip and amplify the signal that was amplified by the first amplifier which shows in figure 3.

figure 3

The green line represents output signal of the first amplifier and red line represents output signal of the second amplifier. 

figure 4

After passing through the third amplifier, the waveform nearly become a DC signal but it is not stable which is shown in figure 4 that there is some glitches on the purple line. After passing through the last amplifier, it is much more stable which is at -19.6V which is shown in figure 4 by blue line.

For second week, we will check the operating voltage of PIR Sensor before and if it is within the normal range we will connect the circuit in the lab and replace the function generator by PIR sensor. According to the simulation results of the first week, it is much easier to check the real circuit. If output signal corresponds to our simulation, we will check the following part: sensitive angle, sensitive distance,etc.  

Welcome to the project of a multil Infrared sensor system for fall detection

Welcome

This mini project is a multil Infrared sensor system for fall detection produced by four year 2 undergraduate students in University of Liverpool named Tianqiong Du, Guotao Dong, Hanyi Fei and Yuxiao Fang.

This system is mainly designed for the elderly and disabled people. Unlike young man, these kinds of people are easily get hurt. Meanwhile, many elderly people will spend much time living alone in the home. Therefore, it is quite necessary to build a system to detect their falling at the first moment. 

It will take nearly five weeks to complete this mini project and we will upload the process week by week. If there is any questions, please contact me the email 18036393480@163.com.