CN114786122A

CN114786122A - Criminal positioning and alarming system based on UWB

Info

Publication number: CN114786122A
Application number: CN202210409700.0A
Authority: CN
Inventors: 史永; 储兆玲; 仲崇锦; 瞿燕; 蔡婷婷; 袁登晨
Original assignee: Changzhou Institute of Technology
Current assignee: Changzhou Institute of Technology
Priority date: 2022-04-19
Filing date: 2022-04-19
Publication date: 2022-07-22

Abstract

The invention relates to the application of an indoor positioning technology, in particular to a criminal positioning and alarming system based on UWB (ultra wide band). Therefore, vicious events such as jail crossing, fighting Europe and the like can be effectively prevented, and a complete and flexible intelligent management system for prisoners is established and comprises a position information acquisition module: the module collects positioning information by using a UWB-based department positioning system to obtain an original positioning data packet; a position calculating module: after filtering the original data of the position information acquisition module, resolving the position information through a positioning algorithm to obtain TDOA positioning data, and resolving to obtain two-dimensional coordinate positioning data; a positioning judgment module: calculating the Loc positioning data of the resolving module through an algorithm to obtain a positioning error; an alarm judging module: and (5) formulating an alarm rule, and realizing abnormal point alarm through an algorithm.

Description

Criminal personnel positioning and alarming system based on UWB

Technical Field

The invention relates to the application of indoor positioning technology, in particular to a criminal positioning and alarming system based on UWB.

Background

With the increasing application of internet of things, smart cities, disaster relief and epidemic prevention and accurate services thereof, the information service industry based on the position is rapidly developed and is paid more and more attention to and paid attention by more and more people. Global Positioning System (GPS) and Beidou positioning system (Beidou) cover and provide outdoor positioning information service, but cannot be effectively applied to indoor environment due to reasons such as shielding. The industry has proposed many indoor positioning schemes; the indoor positioning technology based on UWB has the advantages of high positioning accuracy, strong penetration capability and the like, and draws wide attention of researchers.

The current outdoor positioning mainly relies on GPS and the beidou system, which is based on satellite positioning. In the open air, its positioning accuracy can reach the meter level, can satisfy the demand of outdoor location basically. However, compared with an open outdoor environment, the satellite positioning system has a narrow indoor space and obstacles such as walls, furniture and electric appliances, satellite position signals are easily interfered by the obstacles to be attenuated, the signals are reflected by the obstacles or penetrate through the obstacles while being transmitted to be weakened, and the problems of multipath fading and the like are caused, so that the positioning error is large and even the positioning cannot be performed. In addition, the indoor environment also needs to consider the problems of indoor space layout, circuit layout, equipment cost and the like. Therefore, positioning systems such as GPS and beidou, which are popular outdoors, cannot be completely applied to indoor environments.

In recent years, indoor positioning technology has been greatly improved under the diligence of numerous academic researchers, and particularly, the precision is greatly improved. Through years of research and experiments, there are several typical technical schemes, mainly including the following seven indoor positioning technologies: infrared technology, ultrasonic technology, radio frequency tag technology, ultra wideband radio technology, WiFi technology, ZigBee and iBeacon (bluetooth low energy) technology.

Through the reference of literature data, the seven types of technologies are compared in advantages and disadvantages, as shown in the table 1-1.

TABLE 1-1 comparison of advantages and disadvantages of seven indoor positioning techniques

Unlike other wireless communication technologies, UWB technology, although also employing wireless communication, has data transmission rates of up to several hundred megabits per second. The Federal Communications Commission (FCC) defines: with a relative bandwidth η greater than 0.2, UWB signals are obtained, so that UWB technology has a very wide transmission bandwidth.

Currently, UWB-based personnel location technology has been widely applied in many fields such as prison management, factory prisoners, underground tunnel location, hospital patient accompanying, warehouse management, logistics tracking, and convention and exhibition navigation. As shown in fig. 1-1, in a factory, a UWB-based personnel location system can help a conventional factory to implement digital management, and by installing a location base station in a work area and allowing an employee to wear a location card, the location, working hours, and movement trajectories of the employee can be displayed in real time, thereby improving inspection efficiency. The position of the stored goods can be monitored, the positions of the goods and the warehouse can be checked, and loss of materials and equipment is prevented. The use of UWB personnel positioning systems in prisons by law is gradually popularized, base stations are installed in all areas of prisons, prisoners wear wristbands with positioning cards, and the positioning systems monitor the movement tracks of the prisoners in real time to prevent illegal behaviors such as prison crossing and the like. Therefore, the UWB positioning technology has considerable prospect and advantages in the aspect of indoor personnel positioning.

Prison as a national criminal law enforcement agency has important functions of maintaining social security, preventing illegal crimes, punishing persons, modifying criminals and the like. At present, the management of the escort personnel of most prisons is still the traditional patrol and monitoring mode, and the manual operation accounts for a large proportion. In addition, there are following difficulties in prison management:

(1) the failure to monitor crime locations and criminals in real time;

(2) false alarm phenomena are more, and personnel close to a peripheral alarm area cannot be identified;

(3) the behavior track of a prisoner cannot be analyzed, so that the police strength is reasonably distributed;

(4) and the criminal identity cannot be strictly and accurately identified.

Aiming at the various management difficulties, the intelligent management of specific limited personnel such as prisoners and the like is one of the important applications of indoor positioning technology. Therefore, how to meet the requirement of fine management service of prisoners realizes the basic functions of checking positions of prisoners, moving tracks and the like in real time. Furthermore, the monitoring of behaviors of prisoners is realized by counting movement distance, movement time, movement range and the like, and an alarm is given to abnormal behaviors or abnormal areas and the like, which is a technical problem to be solved.

Disclosure of Invention

Aiming at the problems mentioned in the background technology, the invention aims to provide a criminal positioning and alarming system based on UWB, which can interact with the existing monitoring system through the UWB positioning technology, thereby realizing real-time positioning of the criminal and the manager, real-time prison condition inspection and early warning of abnormal behaviors. Therefore, vicious events such as jail crossing, fighting Europe and the like can be effectively prevented, and a complete and flexible intelligent management system for prisoners is established.

The technical purpose of the invention is realized by the following technical scheme: a criminal positioning and alarming system based on UWB comprises a position information acquisition module: the module collects positioning information by using a UWB-based department positioning system to obtain an original positioning data packet;

a position calculating module: after filtering the original data of the position information acquisition module, resolving the position information through a positioning algorithm to obtain TDOA positioning data, and resolving to obtain two-dimensional coordinate positioning data;

a positioning judgment module: calculating Loc positioning data of the resolving module by an algorithm to obtain a positioning error, calibrating the positioning error according to the obtained error by combining an experimental environment, and judging whether the positioning precision of the positioning system meets the requirement that the indoor positioning error is less than 30 cm;

an alarm judging module: setting an alarm rule, and realizing abnormal point alarm through an algorithm;

the system comprises a position information acquisition module, a criminal staff information database and a system management module, wherein the position information acquisition module is used for acquiring position information of the criminal staff; each prisoner wears a positioning tag, and the prisoner system collects the original positioning data of the tag and sends the original positioning data to the position information collection module;

original data of the label, namely raw data packets, acquired by a driver system are stored in a CSV text file format and can be directly opened in Excel or a notebook, each column of data is provided, and the original data packets store original positioning data acquired by a position information acquisition module;

the prisoners wearing the labels can derive identity information of the prisoners from a management system of prisons for accurately identifying the identities of the prisoners;

the position resolving module firstly filters raw data of the information acquisition module, and then resolves the filtered data by adopting a TDOA algorithm to obtain two-dimensional coordinate data;

filtering: filtering out data below-100 for fields of fp and rx of signal quality; filtering time singular values by using a K-means algorithm;

and (3) positioning and judging: calculating positioning coordinate Loc data obtained from a position calculating module by using an error algorithm, and calibrating a positioning error by combining an experimental environment;

and (3) alarm discrimination: after receiving the positioning position of the label, the alarm module judges that the external point is out of range through alarm judgment, alarms if the internal point is overtime, and continues to display the positioning track of the point which does not accord with the alarm rule.

Preferably, the positioning algorithm may be a TDOA algorithm, that is, filtered raw data is solved to obtain TDOA data, where the TDOA data is mainly a distance difference between a tag and each base station; and then resolving the TDOA data into two-dimensional coordinate data.

Preferably, a plurality of base stations with completely synchronous clocks receive data packets from one tag at the same time, and the receiving time of the base stations at different positions is different for the broadcast packet transmitted by the same tag, so the following algorithm is adopted:

(1) the Tag sends out a broadcast packet;

(2) the two base stations receive the same packet, the time received by the base station 1 is T1, and the time received by the base station 2 is T2;

(3) calculating the time difference Td which is T2-T1;

(4) for at least four base stations, three sets of such information may be available;

(5) solving two-dimensional position coordinates of the label;

the following is the TDOA algorithm's equation for the time difference of arrival:

preferably, the K-means algorithm is an iterative clustering algorithm which takes distance as a similarity index and calculates how many classes the given data set has, the number of the classes is K, the center of the class is a clustering center which is the mean value of all numerical values in the classes; for a given data set X needing clustering, the number K of classes, and the similarity index, the Euclidean distance is selected, the clustering aims to minimize the clustering square sum of the K classes, and the formula is as follows:

the flow of the K-means algorithm:

(1) randomly selecting K centroids;

(2) calculating the Euclidean distance between each sample and the centroid;

(3) classifying each sample into a cluster where a centroid with the closest distance is located;

(4) calculating the average value of each cluster, and classifying again;

(5) and (3) finishing the algorithm when the iteration times reach, otherwise, returning to the step (2).

In summary, the invention mainly has the following beneficial effects: according to the criminal positioning and alarming system based on UWB, data filtering is achieved through a K-means clustering algorithm, TDOA position resolving is applied, and therefore monitoring of positions and tracks of criminals is achieved; on the basis, under the actual environmental conditions of enclosing walls, doors and windows and the like, the positioning error is smaller than 30cm, and furthermore, the positioning of prisoners and the behavior alarm of border crossing and non-compliance with time regulation of the prisoners are realized by setting a positioning area, alarm time and space rules and the like based on a distinguishing method of points in a polygon.

The method proves the correctness and the effectiveness of the method and the system. Under the condition of error tolerance of 30cm, abnormal behaviors under regular conditions such as rest, eating, work and the like can be correctly identified and an alarm is given.

Drawings

FIG. 1 is a system architecture diagram of the present invention;

FIG. 2 is a schematic diagram of the working principle of the position information acquisition module of the present invention;

FIG. 3 is a schematic diagram of the operating principle of the position calculation module of the present invention;

FIG. 4 is a schematic diagram of the operating principle of TDOA resolving of the present invention;

FIG. 5 is a schematic diagram of the operation of the location determination module of the present invention;

FIG. 6 is a schematic diagram of the operation of the alarm module of the present invention;

fig. 7 is a schematic diagram of the alarm flow of the present invention.

Detailed Description

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

UWB communications have evolved rapidly since the FCC approved the 3.1GHz to 10.6GHz common communications band in 2002, where UWB can operate under severe constraints. At the same time, there is a strong competition between various technical solutions established around the international standard.

Since 1999, researchers in our country have begun to attach importance to UWB technology, and UWB technology in our country has begun to develop. The national 863 project was initiated in 2001, which is a research and development project for a high-speed ultra-wideband experimental demonstration system. From Shenzhen, national communications Limited, an ultra-wideband digital fiber distribution System (UW-DDS) was introduced at the international information communication exhibition of China in 2011. The system is the latest technology for realizing multi-system and multi-service combined access. This solution has been put into use and has enjoyed great success.

Since the formation of a business model, UWB technology has found a great number of applications in both domestic and global markets, with the active development of businesses in various segments of the industry chain. Nowadays, the development of indoor positioning technology is mature, and the indoor positioning technology can be widely applied to GPS and map positioning services. However, this technology is also being developed for more mobile terminals. The current indoor positioning technologies are basically positioned based on the existing positions in the development mode, and the main functions are also huge commercial potential brought by promoting indoor position service. Currently, many merchants provide various services, and therefore, various conveniences brought by the services encourage the merchants to optimize and develop technologies, so as to bring better competition and development to indoor positioning technologies.

The mainstream positioning algorithm of UWB is based on different positioning principles, and there are two main categories:

algorithm for localization by direction finding: and (4) AOA.

Aoa (angle of arrival) is a positioning algorithm based on the angle of arrival of a signal, which requires a high level of hardware. The algorithm must ensure that the signal is direct, so this method is generally used in a barrier-free environment.

Algorithm for positioning by ranging: RSSI, TOA, TOF, TDOA, etc.

Rssi (received signal Strength indication) is a positioning method by converting signal Strength into distance information. The hardware cost of the method is low, but the advantages of high UWB time resolution and large bandwidth cannot be fully utilized, and the method is not commonly used in a UWB positioning system.

Toa (time of arrival) is a positioning algorithm with high positioning accuracy based on converting time information into distance information. This method requires time synchronization between the base station and the tag [12 ].

Tof (time of flight) is a two-way ranging algorithm that converts the time of flight of a signal between a base station and a tag into a distance. This approach differs from TOA in that TOF clock synchronization requirements are low.

The TDOA algorithm is an algorithm for determining the location of a tag using the time difference of signals received by a plurality of base stations, compared to the TOA algorithm.

The intelligent prison is the mainstream of the development of a prison management technology in the future, and is a novel prison form established by emerging technologies such as the Internet of things, big data, artificial intelligence and the like. In view of the difficulties and inconveniences of the traditional supervision mode mainly based on manual management, intelligent management has become an inevitable requirement for prison modernization. Today, with the development of science and technology, a number of prison intelligent management technologies have been generated:

face recognition + smart camera: through the face recognition technology, identity recognition and face registration are carried out on managers and prisoners inside prisons, and prison-crossing behaviors are detected in real time.

Human face emotion artificial intelligence: the face recognition is upgraded, the face information of a person can be remotely captured through a camera, psychological monitoring and physiological analysis are carried out, and once physiological or psychological abnormalities of a prisoner are detected, such as abnormal behavior signs of jeer, prison crossing and the like are analyzed, an alarm is given immediately to inform a manager to check the abnormal behavior signs.

A wearable device: intelligence bracelet, intelligent shoes, intelligent dress etc. not only can monitor physiological state such as prisoner's blood pressure, rhythm of the heart, breathing, can also imbed positioner wherein, and the prisoner position is fixed a position in real time, convenient management.

The three types of technologies are mainstream technologies for intelligent management of the prisons at present, and more practical technologies will be born in the future. At present, the intelligent prison trial unit is established in China in a plurality of regions such as Jiangsu, Shenzhen, Hubei and the like. And many scientific and technological innovation enterprises have focused on creating intelligent prisons, so that an intelligent change is brought to science and technology for judicial work.

The invention discloses a criminal positioning and alarming system based on UWB, which aims to enable a criminal to wear a wrist strap type positioning tag, thereby realizing monitoring and prediction of criminal behaviors through positioning the tag, and comprises the following steps: mastering the position of a prisoner, counting labels, forcibly detaching alarm, electronic fence, tracking and replaying of motion trail, monitoring system linkage alarm and the like in real time. Therefore, intelligent management of prisoners can be realized, the law enforcement risk of prisoners is greatly reduced, and dangerous accidents are effectively prevented.

UWB is essentially a carrier-free spread spectrum technique that uses very low duty cycle impulse pulses as the information carrier, directly modulating the impulse pulses with steep rise and fall times. The UWB positioning technology is a distance measurement positioning technology based on the arrival time difference of wireless pulses, and has higher positioning precision than Bluetooth, wifi and the like, and the signal intensity and the positioning precision are much higher and reach within 30 cm.

The working flow of the UWB positioning technology is as follows:

(1) each positioning tag sends an ultra-wideband pulse data frame uninterruptedly and repeatedly;

(2) a positioning base station receives a data frame;

(3) each positioning base station measures the arrival time of each tag for transmitting a data frame to a receiving antenna through a pulse detector;

(4) the positioning machine determines the time difference of the signals received by each base station according to the calibration data, and calculates the position of the label by using an algorithm;

(5) the invention designs a two-dimensional space positioning system, generally 4 base stations are used, and the TDOA algorithm is adopted for positioning.

The UWB technology solves the propagation problem which troubles the traditional wireless communication technology for many years by the advantages of low power consumption, strong anti-jamming capability and the like. The technical characteristics are as follows:

(1) the structure of the system is simple to realize

Unlike conventional wireless communications, UWB technology does not use carriers, allowing the use of low-cost wideband transmitters by sending nanosecond-level non-sinusoidal narrow-pulse transmission signals, and receivers that do not require intermediate frequency processing.

(2) Data transmission is fast

UWB technology has a very wide frequency bandwidth, so data transmission rate is high and frequency resources are not occupied separately, which is an ideal modulation technique for personal communication and wireless local area networks.

(3) Low power consumption

The pulse duration used by the UWB system to transmit data is 0.20 to 1.5ns, and is an intermittent pulse, and the duty cycle and the system power consumption are both very low. The power of civilian UWB devices is about one hundredth, or even one-tenth, of the power of conventional mobile phones and bluetooth devices. Therefore, UWB technology has a significant power consumption advantage.

(4) Has high safety

As a physical layer technology, the UWB technology generally disperses the signal energy in a wide frequency band, and its power spectral density is lower than that of natural electronic noise, which is equivalent to a white noise signal.

(5) Strong multipath resolution

In conventional wireless communication, communication quality and data transmission rate are limited by multipath propagation effects. Because UWB radios transmit intermittent pulses with very low duty cycles and times, multipath signals can be separated in time. Therefore, the multipath resolution is strong.

(6) Accurate positioning

UWB technology is difficult to achieve with conventional radios, and integration of positioning and communication is simple to implement. GPS can only operate within the visual range of its positioning satellites. UWB technology, due to its very high penetration, can also be precisely located indoors and underground.

(7) Simple construction and low cost

Compared with other wireless technologies, the UWB technology is simple in engineering implementation, full digitalization can be achieved only by generating and modulating pulses in a mathematical mode and integrating the circuits on one chip, and equipment cost is very low.

The main technical indicators of UWB are as follows: frequency range: 3.6 GHz-10.6 GHz; system power consumption: 1 mW-4 mW; pulse width: 0.2 ns-1.5 ns; repetition period: 25 ns-1 ms; transmission power: -41.3 dBm/MHz; data rate: tens to hundreds of Mb/s; and (3) decomposing the multipath time delay: less than or equal to 1 ns; multipath fading: less than or equal to 5 dB; system capacity: the system is much higher than the 3G system; space capacity: 1000kb/m 2.

If the alarm of the indoor positioning system adopts an electronic fence technology, the electronic fence can be arranged in an forbidden zone, people without authority cannot enter and exit without permission, otherwise, the UWB positioning system can give an alarm in real time.

Electronic fences are the most advanced perimeter alarm technology at present, and the system is composed of a host and a front-end accessory. Typically, the host is installed and configured outdoors and the electronic fence is installed along the fence. The pulse host is also often installed outside, and the alarm signal is transmitted to a back-end control center by using signal transmission equipment.

The main working process of the electronic fence comprises the following steps:

after the host is electrified, the transmitting port sends out pulse voltage to the front-end fence. Completing one-time emission for about 1.5 s;

on the fence, the pulse stays for about 0.1 second, and returns to the host receiving port after a loop is formed;

a receiving port for receiving a feedback pulse signal;

the host machine is used for detecting the resistance value between the two transmitting terminals;

if the front end fence is damaged, the receiving end of the pulse host cannot receive signals;

or, the fence is short-circuited, the resistance between the transmitting ends is too small, and the host machine gives an alarm;

no matter what voltage, the rail is damaged, and the host computer receiving end can not receive the signal, and the host computer can all report to the police.

The space is divided by a polygon enclosed by the curves. Under reasonable rule conditions, such as sleeping time, a person serving a criminal should be in a polygon; the activity time should be outside the polygon. This document alarms by determining that under certain rule conditions, the person serving a criminal is inside or outside a polygon.

There are several different ideas to decide whether a point is inside a polygon, such as: ray method, angle and hair, area and method, etc. The invention adopts ray method. The principle of the ray method is that a point to be judged is used as an original point to be taken as a ray, and if the total number of intersection points of the ray and each side of a polygon is odd, the point is judged to be in the polygon. If the point is even, the point is judged to be outside the polygon. As in fig. 2-2, at points outside the polygon, the ray with its origin has 4 intersections with the polygon; at a point within the polygon, the ray with its origin has 1 intersection with the polygon.

In order to make the data acquisition system have generality and can be widely popularized, the invention adopts a department UWB positioning system. The system consists of a positioning label, a positioning base station, a POE switch, a positioning server and an Ethernet. The tag is dormant at ordinary times, and the positioning packet is sent according to the designated frequency after vibration. The positioning precision is 10-30 cm; the refresh rate is 0.5-50 HZ (adjustable); the endurance time is more than 2 months (1 HZ). The present system uses a wristband label.

And the base station is installed at a fixed position in a closed space or an open area on site and used for carrying out real-time interaction with a high-precision positioning tag worn by a production worker, so that high-precision personnel positioning is realized. And after receiving the positioning packet, the base station stores the received system time TICK (40BIT integer, the time corresponding to each TICK is 1.0/499.2e6/128.0 seconds, and the corresponding distance is about 4.69 millimeters), and sends the system time TICK to upper computer resolving software through a network. And the main base station sends the synchronous packet at regular time according to the frequency and sends the time for sending the synchronous packet to the upper computer resolving software.

POE switch, network switch who has POE function, the mode power supply that the basic station can adopt ethernet data transmission line can be adopted to the basic station. However, it needs to support the 802.3af power standard with the sensor operating at 7w of power.

Ethernet, the current system uses a wired ethernet network, and each sensor is connected to a network switch through the wired ethernet network to perform data transmission.

As shown in fig. 1, the functions of the modules of the present invention are as follows:

(1) the position information acquisition module: the module collects positioning information by using a UWB-based serendipity positioning system to obtain an original positioning data packet.

(2) A position calculating module: after filtering the original data of the location information acquisition module, the location information is resolved by a location algorithm, such as a TDOA algorithm, to obtain TDOA location data, and two-dimensional coordinate location data (Loc data) is obtained by resolution.

(3) A positioning judgment module: and calculating the Loc positioning data of the resolving module through an algorithm to obtain a positioning error. And according to the obtained error, combining the experimental environment to calibrate the positioning error, and judging whether the positioning precision of the positioning system meets the requirement that the indoor positioning error is less than 30 cm.

(4) An alarm judging module: and (5) formulating an alarm rule, and realizing abnormal point alarm through an algorithm.

The data of the position information acquisition module is provided by an authority system and an information database of criminal personnel. Each prisoner wears a positioning tag, the driver system collects the original positioning data of the tag and sends the data to the position information collection module, as shown in fig. 2, a raw data packet, the original data (raw) of the tag collected by the driver system is stored in a CSV text file, and the file can be directly opened in Excel or a notebook, and each column of data is provided with one piece of data. The raw data (raw) packet stores raw positioning data acquired by the position information acquisition module. Data descriptions are shown in the following Table 3-1:

TABLE 3-1 raw data description

Personnel: the prisoners wearing the labels can derive identity information of the prisoners from a management system of prisons for accurately identifying the identities of the prisoners.

The data format is shown in the following table 3-2:

TABLE 3-2 personnel information

Name of field	Types of	Examples of the invention	Remarks to note
				Prisoner ID	int	001001	Identity code of prisoner
Tag ID	int	258	Name of label
				Position of	string	(0.8，0.6)	Two dimensional position coordinates
Time	time	15：37：12	Positioning information acquisition time

The position calculating module firstly filters raw data of the information acquisition module, then calculates the filtered data by adopting a TDOA algorithm to obtain two-dimensional coordinate data, the operation flow is shown in figure 3,

the following error sources are mainly used in UWB positioning: clock synchronization precision; multipath propagation; non-line-of-sight propagation; multiple access interference, etc.

Due to the positioning errors caused by the above reasons, the position information errors of part of labels in raw data of the position information acquisition module are large, and the positioning accuracy of the whole system is influenced. Therefore, in order to improve the positioning accuracy, the raw data is filtered, which includes two important stages:

(1) for the signal quality fp, rx fields, data below-1 () () is filtered out.

(2) The temporal singular values are filtered using the K-means algorithm.

The position solution adopts a TDOA algorithm, and the process is shown in FIG. 4, (1) filtered raw data is solved to obtain TDOA data, wherein the data is mainly the distance difference of the tag to each base station. Data descriptions are shown in tables 3-3 below:

TABLE 3-3 TDOA data description

Data content	Data format	Explanation of the invention
			$TDOA	string	Data head
Region information	int	Region information
			time	time	Time to the order of milliseconds
tagID	int	Tag ID
			Number of anchors	int	Number of base stations of TDOA
anchrol ID，...anchorNID	int	ID of each base station
			dis1，....disN	int	Difference in distance (cm) to first base station
Signal quality 1, 1	int	Signal quality of received positioning packets for each base station
			seq	int	Serial number (1-127)

(2) The TDOA data is then resolved into two-dimensional coordinate data (Loc data).

The Loc data stores the calculated location results, derived from the TDOA data. Data descriptions are shown in tables 3-4 below:

tables 3-4 Loc data description

Data content	Data format	Explanation of the invention
			$LOC	string	Data head
Time	time	Positioning time
			Tag ID	int	ID of tag
X coordinate	int	Abscissa axis
			Y coordinate	int	Ordinate of the product

The positioning judgment module has the main function of calculating positioning coordinate Loc data obtained from the position calculation module by using an error algorithm and calibrating the positioning error by combining an experimental environment, as shown in fig. 5, the environmental data mainly refers to shielding materials such as walls, and under the condition, the loss error is measured.

And (3) alarm judgment rules: in prisons, the main alarming actions caused by prisoners are prison crossing and non-compliance with the daily work and rest time regulations of prisons. The prison crossing behavior can be understood as border crossing behavior in the positioning system, namely a movable area is defined in a positioning coordinate system, and the border crossing is judged if the positioning coordinate exceeds the defined area. For coordinate points which are not out of range, the relevant time regulation is observed in the designated area, otherwise, the coordinate points are also regarded as abnormal behavior alarm, (1) as shown in the figure 6, after the alarm module receives the positioning position of the label, the alarm module judges that the external points are out of range through alarm judgment, and the internal points are also alarmed if the internal points are overtime. And continuously displaying the positioning track at the point which does not accord with the alarm rule.

(2) Rules, it is understood that the daily time regulations for prisoners may be referenced in tables 3-5.

TABLE 3-5 work and rest table for prisoners

Get up	6：20～6：30
		Early exercise	6：40～7：00
Breakfast	7：00～7：30
		Arrange housekeeping	7：30～8：30
Work by	8：30～11：30
		Lunch	11：30～12：00
Noon break	12：00～13：30
		Work by	13：30～17：00
Eating dinner	17：30～18：00
		Activities of giving off air, watching news, etc	18：00～20：30
Sleep by sleeping	20：30～6：20

According to the work and rest arrangement of the prisoners in the above table, the activities of the prisoners can be divided into four zones, namely a rest zone, a dining zone, a working zone and an overall activity zone, the behaviors in each zone cannot cross the border firstly, and then time regulations must be observed, and if the time of stay in a certain zone is too long, the abnormal condition is considered.

The alarm rules are shown in tables 3-6 below.

TABLE 3-6 alarm rules

Unlike the TOA algorithm, the TODA, i.e., "arrival time difference", does not add a special timestamp, thereby improving the positioning accuracy. Although the base station time is strictly synchronous, the similar transmission characteristics of the mobile channel between two base stations can reduce the error caused by the multipath effect^[21]. TDOA positioning is also called hyperbolic positioning because it uses time difference for positioning^[22]. We can determine the distance of the signal source from the time of arrival of the signal at the base station. Comparing the arrival time difference of the signals, using the base station as a focus, making a hyperbola with the distance difference as a long axis, and the intersection point of the hyperbola is the position of the signal^[23]. In this mode, multiple base stations with perfectly synchronized clocks receive packets from one tag at a time. For the broadcast packet sent by the same tag, the receiving time of the base stations at different positions is different, so the following algorithm is adopted:

(1) tag sends out a broadcast packet.

(2) Two base stations receive the same packet, and the time of receiving by the base station 1 is T₁The time received by the base station 2 is T₂。

(3) Calculating the time difference T_d＝T₂-T₁。

(4) Three sets of such information are available for at least four base stations.

(5) And solving the two-dimensional position coordinates of the label.

the K-means algorithm is an iterative clustering algorithm which takes distance as a similarity index. It calculates how many classes there are for a given data set, the number of classes is K, the center of the class is the clustering center, which is the mean of all values in the class^[24]. For a given data set X needing clustering and the number K of classes, the similarity index selects Euclidean distance, the clustering aims to minimize the clustering square sum of the K classes, and the formula is as follows:

the flow of the K-means algorithm:

(1) randomly selecting K centroids;

(2) calculating the Euclidean distance between each sample and the centroid;

(4) calculating the average value of each cluster, and classifying again;

(5) and (5) the iteration times are reached, and the algorithm is ended. Otherwise, returning to the step (2).

And calculating a positioning error by acquiring 1000 groups of two-dimensional coordinate data of a label in a static state, measuring an actual coordinate of the label, and calculating a distance difference between a measured value and an actual value, wherein the difference is the positioning error.

We can calculate the mean value of the 1000 measured sets of two-dimensional coordinate data directly in the positioning data table, but because of the complex indoor environment, some "jitter" is severe, i.e. the data with large difference between the measured value and the actual value will affect the calculation of the mean value. Therefore, it is necessary to filter out points where the error is particularly large. Data screening is required here. The purpose of data screening is to improve the usability of label positioning data measured in previous experiments, so that later data analysis is facilitated.

Because loc positioning data obtained by using the positioning system is in a csv format and is convenient to open in an Excel table, the value of the positioning data information can be statistically analyzed by using a mathematical formula in the table. For example, using the trimman function.

The principle of the trimman function is to size the data and then remove a certain percentage of each of the data from the front and back of the data set, the end result being the average of the remaining data. This filters out a significant portion of the data with errors, and the resulting average value is closer to the actual value. The syntax structure of this function is TRIMMEAN (percentage). In this experiment, the first parameter array is 1000 sets of two-dimensional coordinates locX and locY of a given tag measured using a positioning system. The second parameter percentage is a ratio of data having a particularly large error in coordinate values to be removed in calculation. For example, in the present experiment, 200 data points are removed from 1000 coordinate values, 100 data points with very small errors are removed, and 100 data points with very large errors are removed, thereby totally removing 200 data points. Then Percent 200/1000 is 0.2.

Filtering loc positioning data by using TRIMMEAN, averaging, and then comparing the average value (x, y) with the actual value (x)₀,y₀) The value of the positioning error can be calculated by substituting the following formula 4.5.

Based on the alarm rules, the system can realize the alarm of the abnormal coordinates through an algorithm, as shown in fig. 7.

In addition, the judicial department emphasizes the promotion of intelligent prison construction, and an intelligent prison management system is constructed on the basis of the existing prison construction. The system is suitable for the development of prisons in new times, so that the prison management is developed towards the intelligent direction; the prisoner transformation quality standard is trampled, and the prison safety risk is controllable; the method has the important significance of improving the modernization level of the work of the prisons, ensuring the supervision safety in the complex prison environment and the like.

In addition, the invention is based on an industrialized data acquisition system, and realizes the high-precision positioning and alarming application of prisoners by the refining means of filtering and TDOA positioning technology. The method is not only an innovation of a management mode of prisoners, but also has practical popularization value.

Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and alterations can be made in these embodiments without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.

Claims

1. The utility model provides a criminal personnel location and alarm system based on UWB which characterized in that:

the system comprises a position information acquisition module: the module collects positioning information by using a UWB-based department positioning system to obtain an original positioning data packet;

a positioning judgment module: calculating Loc positioning data of the resolving module through an algorithm to obtain positioning errors, calibrating the positioning errors according to the obtained errors by combining an experimental environment, and judging whether the positioning accuracy of the positioning system meets the requirement that the indoor positioning error is less than 30 cm;

the system comprises a position information acquisition module, a criminal personnel information database and a server, wherein the position information acquisition module is used for acquiring position information of the criminal personnel; each prisoner wears a positioning tag, and the prisoner system collects the original positioning data of the tag and sends the original positioning data to the position information collection module;

original data of the labels, namely raw data packets, acquired by a driver system are stored in a CSV text file format and can be directly opened in Excel or a notepad, each column of data is provided, and the original data packets store original positioning data acquired by a position information acquisition module;

the prisoners wearing the labels can derive identity information of the prisoners from a management system of the prison for accurately identifying the identities of the prisoners;

filtering: filtering data below-100 for fields of signal quality fp and rx; filtering time singular values by using a K-means algorithm;

and (4) alarm discrimination: after receiving the positioning position of the label, the alarm module judges that the external point is out of range through alarm judgment, alarms if the internal point is overtime, and continues to display the positioning track of the point which does not accord with the alarm rule.

2. The UWB-based criminal locating and warning system according to claim 1, wherein: the positioning algorithm can be a TDOA algorithm, namely filtered raw data is solved to obtain TDOA data, and the data mainly refers to the distance difference between a tag and each base station; and then resolving the TDOA data into two-dimensional coordinate data.

3. The UWB-based criminal locating and warning system according to claim 2, wherein: a plurality of base stations with completely synchronous clocks simultaneously receive data packets from a label, and for broadcast packets sent by the same label, the receiving time of the base stations at different positions is different, so the following algorithm is adopted:

(1) the Tag sends out a broadcast packet;

(3) calculating the time difference Td between T2 and T1;

(5) solving the two-dimensional position coordinates of the label;

the equation for the TDOA algorithm with respect to the time difference of arrival is as follows:

4. the UWB-based criminal locating and warning system according to claim 2, wherein: the K-means algorithm is an iterative clustering algorithm taking distance as a similarity index, and is used for calculating the number of classes of a given data set, wherein the number of the classes is K, and the center of each class is a clustering center which is the mean value of all numerical values in the classes; for a given data set X needing clustering and the number K of classes, the similarity index selects Euclidean distance, the clustering aims to minimize the clustering square sum of the K classes, and the formula is as follows:

the flow of the K-means algorithm:

(1) randomly selecting K centroids;

(2) calculating the Euclidean distance between each sample and the centroid;

(3) classifying each sample into a cluster where a centroid closest to the sample is located;

(4) calculating the average value of each cluster, and classifying again;