Design and implementation of wireless alarm system for drunk driving

With the popularization of automobiles, the harm of drunk driving has become increasingly prominent. In order to effectively control the occurrence of drunk driving accidents, the public security traffic management department uses portable alcohol detectors to quickly check the drunk driving of car drivers, and the measurement results can be used as an objective evaluation standard. However, these works are mainly used for fair law enforcement after traffic accidents occur, and it is difficult to prevent them before they happen. This project combines traditional alcohol detection technology with mobile communication technology to design and implement a wireless early warning system, so that car owners and passengers can know whether the driver is driving under the influence of alcohol at the first time, providing technical guarantee for effectively eliminating traffic accidents.
1 Detection principle
There are five basic types of sensors for detecting alcohol content in gas, including fuel cell type (electrochemical), semiconductor type, infrared type, gas chromatography type, and colorimetric type. Considering factors such as price and ease of use, fuel cell type (electrochemical type) and semiconductor type sensors are currently commonly used. Using these two sensors, a portable breath alcohol tester can be manufactured, which is suitable for on-site detection [1].
The semiconductor sensor uses tin oxide as a semiconductor. When the concentration of the sensitive gas in the gas it contacts increases, the resistance value it presents to the outside decreases. At different operating temperatures, the sensitivity of this semiconductor to different gases is different. In order to make the sensor have the highest sensitivity to alcohol, a heating element is used in the semiconductor breath alcohol tester to heat the sensor to the required temperature.
The detection principle diagram is shown in Figure 1. The WS-2 ethanol gas-sensitive semiconductor sensor is used, which has high sensitivity to ethanol gas and high resolution for other gases. Its response time is less than 10 s, recovery time is less than 30 s, and conductivity change is large. Considering that the DC voltage provided in the car [2] is mostly 12 V/14 V, the input voltage is 12 V, and an LED is used to indicate the power supply status. The 5 V voltage output by LM7805 is used as the heating voltage of the sensor and provides the driving voltage for the subsequent chip. The load resistor RP1 of the sensor is an adjustable resistor of 0.5 kΩ-5 kΩ to achieve the best effect.

2 System Design
When the alcohol sensor is exposed to ethanol molecules in the breath of the driver or passenger, the output voltage of the alcohol sensor will change, completing the conversion of "gas" to "electrical" signal.
The output voltage of the sensor is judged into three levels by the comparison circuit, corresponding to "no drinking", "trace drinking" and "excessive drinking". Trace drinking is the standard for drunk driving, corresponding to a blood alcohol concentration of 0.2 mg/ml, which is converted into a breath concentration of 44 ppm; excessive drinking is the standard for drunk driving[3], corresponding to a blood alcohol concentration of 0.8 mg/ml, which is converted into a breath concentration of 176 ppm. At this time, the output of the comparison circuit controls the switch circuit to drive the voice chip to alarm; the output voltage of the sensor is amplified and added to the square wave generating circuit. When alcohol is detected, the circuit will drive the light-emitting diode to control the automatic dialing of the mobile phone, so that the remote car owner can receive the voice alarm signal on the spot through the public network, so that the passengers and the car owner can monitor the drunk driver at the same time. The system principle is shown in Figure 2.

2.1 Comparison circuit design
The comparison circuit is shown in Figure 3. The three single-limit comparators of the LM339 chip are used to compare and judge the sensor output signal in three voltage limits. The input signal Uin is added to the non-inverting input terminal of the comparator, and the corresponding threshold voltage Ur is connected to the inverting input terminal. When the input voltage is greater than the threshold voltage, the output is a high level UOH. The threshold voltage of each voltage comparator can be set by adjusting the potentiometer.

2.2 Voice alarm circuit design
Use the comparator outputs S2 and S3 to control the on and off of channels M1 and M2 of the voice chip APR9600 through CD4052, and control the voice chip alarm sound. The circuit principle is shown in Figure 4. The first alarm sentence entered is: "Driver friend, you have drunk alcohol, please drive carefully!" The second paragraph is: "Driver friend, you have drunk too much alcohol, please don't drive!"

The low-power audio signal amplifier LM386 is used to amplify the alarm voice to obtain a clear and moderate voice signal.
2.3 Design of mobile phone automatic dialing circuit
After testing, the output voltage of the sensor is 1.7 V when it detects alcohol and reaches the S2 level drunk standard, and the output voltage is 3.9 V when the alcohol concentration reaches the designed S3 level drunk standard, and the output voltage does not exceed 0.39 V when there is no alcohol. Therefore, the output voltage of the sensor is amplified three times by LM358 and used as the working voltage of NE555, which can ensure that NE555 can work reliably and not be damaged when alcohol is detected. The mobile phone dialing circuit is shown in Figure 5.

When the rectangular wave output by NE555 is at a low level, the LED goes out, the photoresistor resistance on the row and column lines of the mobile phone transmitter key is about 90 kΩ, and the mobile phone transmitter key is disconnected. When the rectangular wave is at a high level, the LED lights up, the photoresistor resistance becomes about 100 Ω, and the mobile phone transmitter key is turned on. When the LED flashes twice, it is equivalent to the mobile phone transmitter key being pressed twice in a row, and the pre-stored car owner's phone number is dialed. When the car owner answers the call, he can hear a clear and appropriate warning voice.
3 Test results and analysis
3.1 Sensor module test data and analysis
When the input voltage is 12 V, according to the needs of the drive circuit, adjusting the load resistance to 2 kΩ can achieve the best working state, so that the sensor reaches the most sensitive heating time of 5 min. The output voltage signal is tested, and the output voltages of each level are S1: 0.39 V, S2: 1.7 V, S3: 3.9 V. After analyzing the test results, it can be seen that the voltage signal is normal. If the car battery voltage is too low, the load resistance can be appropriately increased to ensure that the output voltage corresponding to the three levels remains stable.
3.2 Test results and analysis of threshold voltage
The three-level threshold voltage should be lower than the output voltage of the sensor module, and leave an appropriate margin to ensure the reliability of the early warning. Taking into account the characteristic parameters of the sensor itself, as well as the requirements for air and temperature, the three 10 kΩ potentiometers in the circuit are adjusted. After repeated testing and analysis, the threshold voltage of level S1 is 0.20 V, the threshold voltage of level S2 is 1.0 V, and the threshold voltage of level S3 is 2.9 V, which meets the experimental standards.
3.3 Test and analysis of voice power amplifier circuit
Connect capacitors and resistors to pins 1 and 8 of LM386 to adjust the closed-loop voltage gain of the circuit. After repeated debugging, when the capacitor value is 10 μF and the resistor value is 1.2 kΩ, the voltage amplification factor reaches 34 dB, and a clear and moderate voice alarm can be obtained. Appropriately reducing the resistance value can increase the amplification factor.
3.4 Test and analysis of NE555 rectangular wave signal
NE555 is used to generate periodic pulses. The light-emitting period of the light-emitting diode should not exceed 2 s, otherwise the system will not be able to drive the mobile phone dial-up call function and affect the remote alarm. The actual designed square wave period T is 1.148 s, and it actually runs well.
The experiment shows that the use of wireless communication based on the existing public network can greatly improve the communication reliability of the alarm system, and the communication distance is not limited, so that the car owner can monitor the car remotely and in real time. Since the car owner is the legal person responsible, the car owner's control over the driver is much higher than that of the passenger. Such a configuration can help the car owner to eliminate drunk driving from the source. The whole system is powered by a car power supply. The main cost is the monthly rental fee of the mobile card. Using a mobile phone as a transmitter saves a lot of software and hardware costs [4-5]. The experiment used an ordinary mobile phone, and the three public networks (Telecom, Mobile, and Unicom) all received the expected warning voice (or short message). Through appropriate modification, the system can also be used for car owners' requirements for car anti-theft, and can also be used for unmanned warehouse protection, building door entry and other occasions. The system makes full use of the existing public communication resources and greatly reduces the cost, so it has good promotion value.
References
[1] Lin Jishen, Huang Wenfeng, Wu Jianshan. Development of a handheld ethanol tester [J]. Sensor Technology, 2000(2): 45-47.
[2] Qin Yun. A review of new automotive electronics technologies [J]. Light Vehicle Technology, 2008, 221-222(1/2): 38-41.
[3] National Standard of the People's Republic of China: GB19522-2004, Threshold and test of blood and breath alcohol content of vehicle drivers [S].
[4] Ma Shengqian. Design and implementation of remote intelligent monitoring system [J]. Application of Electronic Technology, 2007, 33(9): 93-96.
[5] Lv Fang, Hu Linjing, Zang Chen. Software design of automobile anti-theft alarm system based on TC35 GSM [C]. The 3rd National Joint Academic Conference on Embedded Technology and Information Processing, 2009: 125-127.