Realizing perimeter intrusion alarm system based on MEMS chip

    Abstract: In recent years, with the continuous advancement of security alarm technology, various electronic perimeter detection systems attached to solid protective structures have been continuously innovated, from the initial active infrared detection technology and electronic pulse discharge detection technology, to the current development of various specific perimeter protection systems for different application needs, including vibration detection systems attached to metal braided or welded wire fences. There are many types of perimeter intrusion systems based on vibration detection: vibration cable, fiber optic interference, inertial vibration, etc. This article focuses on the vibration detection perimeter intrusion alarm system realized by MEMS chip sensing technology.

　　Micro-Electro Mechanical Systems (MEMS) is a key technology to solve the miniaturization of sensors. Simply put, its working principle is that the physical, chemical, biological and other signal inputs from the external environment are converted into electrical signals through microsensors. After signal processing (analog signal or digital signal), the microactuator executes the action to achieve the function of "interaction" with the external environment. It is the core technology of the Internet of Things.

　　Perimeter intrusion alarm system usually refers to monitoring a linear boundary (such as the fence of an important area). When an illegal act of passing through the boundary is found, an alarm is generated to achieve monitoring and protection of the target area. The perimeter intrusion alarm system will make the user's security management very simple. It forms an electronic defense line on the fence and works 24 hours a day. Once an intrusion occurs, the system will immediately give an alarm signal to notify the duty personnel. This can save manpower, material and financial resources. The most important thing is that the system can reduce the labor intensity of the duty personnel, so that the security personnel can always maintain a high level of attention and effectively shield the perimeter loopholes caused by the negligence and fatigue of the patrol duty personnel. In recent years, with the continuous advancement of security alarm technology, various electronic perimeter detection systems attached to solid protective structures have been continuously innovated, from the initial active infrared detection technology and electronic pulse discharge detection technology, to now there are various specific perimeter protection systems for different application needs, including vibration detection systems attached to metal braided or welded wire fences. There are many types of perimeter intrusion systems based on vibration detection: vibration cable, fiber optic interference, inertial vibration, etc. This article focuses on the vibration detection perimeter intrusion alarm system realized by MEMS chip sensing technology.

　　Composition of perimeter detection system based on MEMS chip

　　The basic structure of general perimeter detection systems and Internet of Things systems is similar: sensing, networking, and control management.

　　Vibration sensor based on MEMS chip-ViSEN

　　The vibration sensor ViSEN is based on a 3-axis acceleration sensor chip. External environmental input (vibration) causes the acceleration values ​​of the MEMS chip's XYZ axes to change. The sensor processor chip on the vibration sensor ViSEN collects acceleration change information, filters out interference through multiple algorithms, and transmits the alarm information through digital communication components. Since the MEMS chip can provide 3-axis continuous, stable and reliable vibration data, we can implement multi-dimensional algorithm comprehensive analysis to ensure that the system works stably and accurately, and the on-duty personnel will not be overwhelmed by false alarms such as "the wolf is coming".

　　The principle structure of the sensor is shown in Figure 1. The sensor processor is the core component, which is responsible for collecting acceleration change state information, forming a preliminary judgment on the intrusion alarm through the 5D algorithm, and then sending the alarm information through the digital communication component. The sensor processor can also receive the configuration information of the background software through the digital communication component, set different parameters for fences with different hardness characteristics, and improve the adaptability of the sensor.

　　In this way, the sensor will sense the fence it is attached to, and effectively detect the continuous vibration caused by climbing, the high-frequency vibration caused by cutting the fence, the strong vibration caused by destroying the fence, the fence being pushed down, and the sensor being removed.

　　Building systems to form sensor networks

　　After completing the sensor that can effectively sense the vibration of the fence, we need to form a system to collect the alarm information from many sensors. This system can be a wireless transmission network based on ZigBee or a wired transmission network connected by cables.

　　In the ZigBee wireless transmission network, the vibration sensor ViSEN is powered by batteries. The sensor only works when vibration occurs, and is usually in a dormant state, which ensures that the sensor can work for more than one year without changing the battery. In this network, all sensors are terminal devices, and the system will set up routers at appropriate locations, and finally connect them to the coordination processing device VGM (Coordinator).

　　For linear boundaries that are often used in perimeter detection systems, a wired transmission network connected by cables is more suitable. The system connects the vibration sensor ViSEN in series to the field control host VGM through an induction cable, and the latter provides power and communication media to the sensor through the cable. A CAN bus connection is formed between the sensor and the field control host to ensure that the alarm on the sensor can be reported to the host in a timely and proactive manner, and almost zero-delay alarm can be achieved. If the VGM host uses RS485 to poll the sensor, once the number of sensors connected to the bus is too large, the alarm delay that may occur will be very unacceptable.

Figure 1. Sensor principle structure diagram

    Sensor control and management

　　The system controls and manages the sensor through the integrated management platform software and the field control host. First, whether it is a wireless network or a wired network, the sensor ViSEN is connected to the field control host VGM, and the field control host implements the underlying communication protocols such as sensor configuration, parameter changes, and alarm information collection; secondly, the integrated management platform software manages the vibration sensor by reading and writing to the field control host; finally, the alarm is presented on the interface in the form of an event.

Figure 2. ZigBee wireless transmission network diagram

　　Features of perimeter detection system based on MEMS chip

　　After the above introduction to the perimeter alarm system based on MEMS vibration sensors and comparing it with the existing vibration detection perimeter alarm system, we have discovered some characteristics of the perimeter alarm system based on MEMS chips.

　　Fully digital system

　　Micro-electro-mechanical systems directly realize the conversion of external stimuli to outputs within the chip, and the chip does not require high external conditions when working normally. Therefore, vibration sensing using MEMS chips is easier to control and implement than traditional detection systems such as optical cable vibration.

　　The external interface of the 3-axis acceleration sensor chip is IIC, and the processor on the sensor directly obtains the digital signal, so the entire sensor has low requirements for power supply, communication and algorithm redundancy, and the sensor has strong anti-interference ability. The traditional vibration cable and optical cable vibration detection system needs to go through small signal amplification, analog signal acquisition and other steps before the detection algorithm can be implemented, and these processes must strictly control the noise, otherwise it will directly lead to a decrease in the accuracy of the algorithm and affect the most important indicators of the perimeter system: false alarm rate and missed alarm rate.

Figure 3. Cable-connected wired transmission network diagram

　　Achieve precise positioning

　　The sensor is a fully digital system based on MEMS chips. We can assign an ID to each sensor. If a sensor detects an intrusion alarm, this ID can be uploaded to the management platform along with the alarm information, so that the entire system has the ability to locate the alarm to each sensor.

　　Generally, if one sensor is installed on one mesh, the system can locate the alarm on each fence. Accurate alarm location is very conducive to the precise linkage of multiple systems. For example, when doing video linkage of perimeter alarm, we can correspond the position of each fence to the camera preset position. When there is an alarm, the camera preset position is directly called according to the alarm sensor address, and the camera is quickly adjusted to the alarm position.

　　The positioning accuracy of traditional optical cable vibration detection systems is generally around 100 meters, and the positioning accuracy of vibration cables is also around 10 meters. The camera can only view the overall image in a linked manner, and the actual point of intrusion may be missed, and the actual situation at the intrusion site cannot be seen in time.

    Realize multi-angle sensing

　　The sensor is based on a 3-axis acceleration MEMS chip. The vibrations caused by shearing the fence and climbing the fence can be easily distinguished through intensity and frequency distribution algorithms. In addition, the sensor can also detect the overturning of the fence through the orientation detection function of the MEMS chip.

　　Traditional vibration detection systems using vibrating cables and optical cables have very low accuracy in distinguishing between shearing and climbing, and cannot tell if a fence has capsized.

　　Combining multiple algorithms

　　The FIRST 5D algorithm is an algorithm that comprehensively analyzes vibration from five different domains.

　　F-Frequency Domain, analyze vibration from the frequency domain

　　Different vibrations have different frequency distributions. The vibrations caused by wind blowing the fence are mostly low-frequency vibrations; the vibrations caused by shearing the fence are mostly concentrated in high frequencies; and the continuous vibrations caused by climbing are just between the first two. Please see Figures 4-1, 4-2, and 4-3 for examples.

　　I-Intensity Domain, analyze vibration from the intensity domain

　　The vibration frequency distribution caused by hail and raindrops hitting the fence is similar to shear, and is also reflected in obvious high-frequency changes. To shield these interferences caused by weather reasons, another dimension must be introduced for analysis, that is, vibration intensity. Reasonable selection of intensity thresholds or adding intensity elements to the algorithm can effectively shield the interference of hail and raindrops.

　　R-Relationship Domain, analyze the correlation of vibration

　　The vibration characteristics analyzed in the frequency domain, intensity domain, space domain and time domain are analyzed in correlation to achieve full-network-level integrated perception and detection, thereby improving the stability and reliability of the entire system.

　　S-Space Domain, analyzes the spatial correlation of vibration from the spatial domain

　　The sensors are distributed along the linear perimeter, and the field control host uses the information of multiple sensors distributed at different positions of the perimeter to perform joint analysis to achieve fusion in the spatial domain.

　　T-Time Domain, analyze the time correlation of vibration from the time domain

　　The temporal relationship of the perceived information is used to establish a fusion in the time domain and calculate the vibration characteristics within the time window.

　　Highly sensitive and accurate

　　For sensors using micro-electromechanical system chips, the stability of the sensing input is firmly grasped by the MEMS chip. The chip can achieve sensitive perception and obtain accurate low-noise data. Combined with the subsequent 5D algorithm for screening and control, the sensor can easily achieve extremely low false alarm and missed alarm rates.

　　BG-VGS fence vibration detection system based on MEMS chip

　　BG-VGS fence vibration detection system is one of the BG series perimeter detection systems produced by Beijing Shenlan Tongchuang Technology Co., Ltd. It consists of MEMS vibration sensor ViSEN, field control host VGM, dry contact expansion module RPM (optional) and alarm control software. BG-VGS fence vibration detection system is suitable for installation on various metal wire welding or braided fences, iron fences, walls and other carriers, and detects intruders climbing and destroying the carriers. The detection cable can be designed and made into a protective box sensor type or a concealed cable type according to the site. The alarm positioning accuracy can reach a minimum of 1 meter (depending on the sensor deployment spacing). The vibration detection sensor ViSEN can be deployed in the middle of the physical fence of the defense zone, on the inside of the column or wall, with an average of one every 1 to 6 meters, to detect vibrations caused by various intrusion behaviors. Each field controller VGM can control no more than 126 vibration detection sensors on the left and right, that is, each VGM can theoretically control a perimeter of up to 756 meters (standard 3-meter spacing).

Figure 4-1. Shear frequency distribution

Figure 4-2. Climbing frequency distribution

Figure 4-3. Frequency distribution of wind disturbance

　　Through software settings, all detection sensors can be arbitrarily combined and divided on each field controller VGM to form dry contact sub-defense zones of any different lengths. The minimum defense zone can be set for each detection sensor.

　　The BG-VGS fence vibration detection system has strong environmental adaptability and low false alarms. It can work stably and reliably around the clock in harsh environments such as wind, frost, rain, snow, dense fog, dust, high and low temperatures.

　　BG-VGS fence vibration detection system is very suitable for the protection of infrastructure with fence carriers, such as airports, military bases, energy facilities, nuclear power facilities, key cultural relics protection, as well as enterprises, factories, warehouses, substations, water plants, power plants, schools, detention centers, prisons and other important places, and the system is very suitable for relatively harsh environmental conditions. The system can also be used for perimeter security protection of ordinary residential areas and villa residential areas.

　　BG-VGS Fence Vibration Detection System has the following core features:

　　Original FIRST 5D algorithm;

　　Effectively detects continuous vibrations caused by climbing, pulling, stepping on and cutting the fence;

　　Users can set detection parameters to adapt to fences, fences and walls of different types and hardness;

　　All ViSEN connected to the control host VGM can be divided into small defense zones by software. If each small defense zone has a separate dry contact output, it can be divided into up to 12 small defense zones; if dry contact output is not required, the software can divide the small defense zones arbitrarily without restriction;

　　The alarm can be located at every detection sensor and every fence;

　　The zone control host VGM has anti-dismantle alarm, low voltage alarm and sensor cable interruption alarm functions;

　　The sensor ViSEN can output 4 types of alarms: shear, climb, tilt and tamper;

　　The system provides serial port and dry contact external communication methods, which is easy to integrate with third-party systems;

　　Low power consumption and low maintenance cost.

　　BG-VGS fence vibration detection system can effectively filter out the interference caused by environmental changes, and can reliably and accurately judge shearing, climbing and overturning of the fence, truly allowing users to experience a sensitive, accurate and fully controlled feeling.