CN107491799B - A kind of switch gate condition detection method and device based on RFID - Google Patents
A kind of switch gate condition detection method and device based on RFID Download PDFInfo
- Publication number
- CN107491799B CN107491799B CN201710693758.1A CN201710693758A CN107491799B CN 107491799 B CN107491799 B CN 107491799B CN 201710693758 A CN201710693758 A CN 201710693758A CN 107491799 B CN107491799 B CN 107491799B
- Authority
- CN
- China
- Prior art keywords
- inductive signal
- door
- data
- state
- intensity data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06K—GRAPHICAL DATA READING; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
- G06K17/00—Methods or arrangements for effecting co-operative working between equipments covered by two or more of main groups G06K1/00 - G06K15/00, e.g. automatic card files incorporating conveying and reading operations
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Geophysics And Detection Of Objects (AREA)
- Burglar Alarm Systems (AREA)
Abstract
The invention discloses a kind of switch gate condition detection method and device based on RFID, comprising: pass through at least three RFID tags and emit at least three inductive signals；At least one set of inductive signal is received by least one antenna, every group of inductive signal includes at least three inductive signals that at least three RFID tags are emitted to the same antenna；Determine at least one set inductive signal intensity data, the intensity of each inductive signal is determined by the distance between antenna for emitting the RFID tag with reception of the inductive signal inductive signal and angle；The switch state of door is determined according to the classifier that at least one set of inductive signal intensity data and preparatory training obtain.The present invention is avoided being existed using electromagnetic induction device and be substituted by other magnetic devices, and coded lock is made to be mistakenly considered the possibility that door is in real closed state.
Description
Technical field
The invention belongs to state inspection fields, are based on radio frequency identification (Radio more particularly, to one kind
Frequency Identification, RFID) switch gate condition detection method and device.
Background technique
In life, people are always concerned about the switch state of door and window, the case where to prevent door and window to be left unlocked or unlatched, cause important
Article and property are stolen, and reminding alarm device would generally be arranged on novel intelligent door and window.In Chinese invention patent
A kind of intelligent door and window switch state monitoring system is disclosed in CN201310434279, system is by incuding magnetic when door and window is opened and closed
Far and near different, the principle of tongue tube on-off affected by magnetic fields of iron and tongue tube distance, converts to control centre's transmission level and believes
Number detect Switch for door and window state.Also disclose that one kind makes in another utility model patent specification CN201521022394
With the method for cabinet door switching-state information collector monitoring door opening and closing state, cabinet door switching-state information collector used is several
A microswitch passes through photoelectric isolation module once detecting cabinet door switching information and is sent to microprocessor processes.
In both Switch for door and window condition detecting systems, electromagnetic induction device is all employed, electromagnetic induction device needs
Network provides technical support, and there are application limitation.In addition, electromagnetic induction device is that detection magnetic field exists in a certain range
And judging whether door really closes, this magnetic device does not have uniqueness, and criminal may be deceived using artificial false magnetic field
Electromagnetic induction device is crossed, allows electromagnetic induction device to be mistakenly considered door and is in close state, there are some potential safety problemss.
Summary of the invention
In view of the drawbacks of the prior art, it is an object of the invention to solve existing Switch for door and window condition detecting system to need net
Network is supported, and its electromagnetic induction device used may be unable to judge accurately Switch for door and window state by artificial false magnetic interference
Technical problem.
To achieve the above object, in a first aspect, the present invention provides a kind of switch gate condition checkout gear based on RFID,
It is characterized in that, comprising: at least three RFID label tags, at least one antenna, radio-frequency identification reader/writer and processing unit；At least three
A RFID tag is mounted on the different location on the door, for emitting at least three inductive signals；At least one antenna
It is mounted on space in door, for receiving at least one set of inductive signal, every group of inductive signal includes at least three radio frequency identification
Label is emitted at least three inductive signals of the same antenna；Radio-frequency identification reader/writer is for determining at least one set of inductive signal
Intensity data, every group of inductive signal intensity data is determining according to the intensity of each inductive signal in every group of inductive signal, often
The intensity of a inductive signal by between the antenna that emits the RFID tag of the inductive signal and receive the inductive signal away from
Walk-off angle degree determines；Processing unit, point for being obtained according at least one set of inductive signal intensity data and training in advance
Class device determines that the switch state of the door, the classifier that the training obtains are in two kinds of opening and closing by the door in advance
The multiple groups inductive signal intensity data training for classification received in advance of at least one described antenna obtains under state.
The present invention RFID emit inductive signal by way of, avoid switch state detector need network support and
The problem of electromagnetic induction device may be by artificial false magnetic interference, can be with accurate judgement Switch for door and window state.
It should be noted that the inductive signal of usually each RFID label tag has oneself unique electronic code, therefore day
Line judges that the switch state of door is not easy to be interfered by other jam signals by way of receiving RFID label tag inductive signal,
The more existing switch gate condition checkout gear of its safe coefficient is much higher.
Optionally, the installation site of at least three RFID tags be Men Zhongxin, door close to shaft side, door lean on nearside
Along side, door along and door lower edge at least three positions.
Optionally, antenna is also used to emit radiofrequency signal, so that the RFID tag swashing in the radiofrequency signal
Give the transmitting inductive signal.
The frequency range of RFID reader and RFID label tag that the present invention selects can be hyperfrequency (902MHz~928MHz) range,
The distance of reaction of radiofrequency signal is 4 to 8 meters under hyperfrequency, and penetrability is strong.In addition, can by using the RFID label tag of anti-metal,
It is interfered to avoid the radiofrequency signal of hyperfrequency from being shielded by iron door and window.
In addition, antenna is also an option that multiple, such as 4, when space is tetragonal body in door, four antennas can be distinguished
It is mounted on any four sides in six faces of tetragonal body, to guarantee all standing of radiofrequency signal, and guarantees comprehensive full angle
The inductive signal of multiple RFID label tags on door is detected, the precision of door state judgement is improved.
Optionally, the processing unit is used for according to the inductive signal intensity data gathered in advance and support vector machines
SVM algorithm constructs door opening and closing state classifier, and the door opening and closing state classifier is used to incude the multiple groups for being used to classify
Signal strength data carries out two classification, and the intensity data of multidimensional inductive signal is transformed into door opening and closing state and corresponds to 2-D data,
By the multiple groups inductive signal intensity data that is used to classify of the door under in the open state in advance labeled as first state data, by door
Be in close state the lower multiple groups inductive signal intensity data for classification in advance and be labeled as the second status data, when it is described extremely
When few one group of inductive signal intensity data is fallen into the first state data area, the door is in the open state, when described
When at least one set of inductive signal intensity data is fallen within the scope of second status data, the door is in close state.
It is understood that two classification will sort out two states: first state and the second state, two states are right respectively
Two kinds of situations of the on and off answered the door.
It should be noted that it is strong to record multiple groups inductive signal by way of being gradually incremented by opening angle on the door in advance
Degree evidence, such as: it is every to open 2 °, 5 ° or 10 ° ... just to acquire at least one set of inductive signal intensity data, the doors such as 5000 groups are acquired altogether
Inductive signal intensity data under opening, and when the door is closed, persistently read the signal strength datas conducts such as 5000 groups
Door is in close state lower inductive signal intensity data, to classify for subsequent to door state.
In addition, first state can be referred to by data " -1 ", and the second state can pass through data " 1 " for convenience of understanding
It refers to.
Optionally, processing unit is used to according to the multiple groups inductive signal intensity data training classifier for classification, first
It needs to pre-process the multiple groups inductive signal intensity data for classification, completes data screening and zoom operations so that data
Suitable for processing；Then, suitable training parameter is chosen according to data type and distribution situation, training data and parameter is input to
In machine learning algorithm, training obtains the door opening and closing state classifier.
It should be noted that first state can be referred to by data " -1 ", the second state can be referred to by data " 1 "
Generation.
Specifically, dimension a is any dimension in the intensity data of multidimensional inductive signal, i.e., the data of dimension a include dimension
Spend multiple inductive signal intensity datas of the corresponding antenna acquisition of a.
Second aspect, the present invention provide a kind of switch gate condition detection method based on RFID, comprising: pass through at least three
RFID tag emits at least three inductive signals, the different positions that at least three RFID tag is mounted on a door
It sets；At least one set of inductive signal is received by least one antenna, every group of inductive signal includes at least three radio frequency identification
Label is emitted at least three inductive signals of the same antenna, at least one described antenna is mounted on space in door；It determines extremely
Few one group of inductive signal intensity data, every group of inductive signal intensity data is according to each inductive signal in every group of inductive signal
Intensity determine that the intensity of each inductive signal is by emitting the RFID tag of the inductive signal and receiving the inductive signal
The distance between antenna and angle determine；Train according at least one set of inductive signal intensity data and in advance obtained classification
Device determines that the switch state of the door, the classifier that the preparatory training obtains are in opening and closing two by the door in advance
At least one antenna multiple groups inductive signal intensity data for classification received in advance is trained under kind state obtains.
Optionally, the installation site of at least three RFID tags be Men Zhongxin, door close to shaft side, door lean on nearside
Along side, door along and door lower edge at least three positions.
Optionally, this method further include: radiofrequency signal is emitted by the antenna, so that the RFID tag is in institute
It states and emits the inductive signal under the excitation of radiofrequency signal.
Optionally, it is opened according to the inductive signal intensity data gathered in advance and support vector machines algorithm building door
Off status classifier, the door opening and closing state classifier be used to for door being in advance open and close under two states it is described at least
One antenna multiple groups inductive signal intensity data for classification received in advance carries out two classification, by the strong of multidimensional inductive signal
Degree evidence is transformed into door opening and closing state and corresponds to 2-D data, and the multiple groups that are used to classify of the door under in the open state in advance are incuded
Signal strength data is labeled as first state data, the multiple groups inductive signal for being used to classify under door is in close state in advance
Intensity data is labeled as the second status data；When at least one set of inductive signal intensity data falls into the first state data
When in range, the door is in the open state, when at least one set of inductive signal intensity data falls into second status number
When according in range, the door is in close state.
Optionally, according to the multiple groups inductive signal intensity data training classifier for classification, it is necessary first to for dividing
The multiple groups inductive signal intensity data of class is pre-processed, and completes data screening and zoom operations so that data are suitable for processing；So
Afterwards, suitable training parameter is chosen according to data type and distribution situation, training data and parameter is input to machine learning and calculated
In method, training obtains the door opening and closing state classifier.
In general, through the invention it is contemplated above technical scheme is compared with the prior art, have below beneficial to effect
Fruit:
(1) present invention avoids switch state detector and needs network branch in such a way that RFID emits inductive signal
May be by artificial false magnetic interference with electromagnetic induction device the problem of is held, it can be with accurate judgement Switch for door and window state.
(2) frequency range of the RFID reader of the invention selected and RFID label tag can be hyperfrequency (902MHz~928MHz) model
It encloses, the distance of reaction of radiofrequency signal is 4 to 8 meters under hyperfrequency, and penetrability is strong.In addition, can be marked by using the RFID of anti-metal
Label interfere so that the radiofrequency signal of hyperfrequency be avoided to be shielded by iron door and window.
(3) the door opening and closing state detection method proposed in the present invention has low cost, high-precision, strong interference immunity and reading
The features such as speed is fast.This method is applied to not need addition additional modules when box for material circulation switch state detects, it is convenient to facilitate.
For different door and windows, our initial training be would also vary from, and can obtain optimum door opening and closing state under specific condition
Detection criteria.
Detailed description of the invention
Fig. 1 is the operation principle schematic diagram that RFID reader of the invention is connect with RFID label tag；
Fig. 2 is the RFID reader of an example of the present invention and the specific location schematic diagram of label；
Fig. 3 is the switch gate condition detection method flow diagram provided by the invention based on RFID.
Specific embodiment
In order to make the objectives, technical solutions, and advantages of the present invention clearer, with reference to the accompanying drawings and embodiments, right
The present invention is further elaborated.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, and
It is not used in the restriction present invention.As long as in addition, technical characteristic involved in the various embodiments of the present invention described below
Not constituting a conflict with each other can be combined with each other.
RFID radio frequency identification is a kind of communication technology, can be identified in the case where contactless by radio signals specific
Target reads and modifies target data, and without mechanical or optical contact, with strong anti-interference performance, reading rate is fast, reads
The strong feature of reading ability.RFID system is made of RFID reader and RFID label tag, and it is any to store user in RFID label tag
Specified information, has sole electron coding, and the working principle of whole system is as shown in Figure 1.It include: RFID label tag 100, antenna
200, RFID reader 300 and processing unit 400.Wherein, RFID label tag 100 is at least three, and antenna 200 is at least one.
It is understood that can by rationally utilize RFID label tag quantity and degree of scatter and the quantity of antenna and
Degree of scatter, to reduce detection error.
The different location that RFID label tag 100 is mounted on a door, for emitting at least multiple inductive signals；Antenna 200 is installed
The space in door, for receiving at least one set of inductive signal, every group of inductive signal includes that RFID label tag 100 is emitted to the same day
Multiple inductive signals of line 200.Radio-frequency identification reader/writer 300 is for determining at least one set of inductive signal intensity data, every group of sense
Induction signal intensity data is determining according to the intensity of each inductive signal in every group of inductive signal, the intensity of each inductive signal
It is determined by emitting the distance between the RFID label tag 100 of the inductive signal and the antenna 200 that receives the inductive signal and angle.Place
Unit 400 is managed, the classifier for obtaining according at least one set of inductive signal intensity data and preparatory training determines door
Switch state, the classifier that the preparatory training obtains be in advance by the door open and close under two states it is described extremely
The multiple groups inductive signal intensity data training for classification received in advance of a few antenna obtains.
It should be noted that the intensity of inductive signal may also suffer from blocking between antenna 100 and RFID label tag 200
The influence of object, since the distinguishing rule of the classifier of the invention obtained using machine learning algorithm has certain generalization ability,
And the present invention can determine the influence for carrying out shelter by using multiple RFID label tags and antenna, therefore temporarily ignore screening here
Block material.It repeats no more below.
Wherein, antenna 200 emits the radiofrequency signal that coverage distance is greater than 4 meters.For example, the present invention uses RFID reader
300 and the frequency range of RFID label tag 100 can be hyperfrequency (902MHz~928MHz) range, under hyperfrequency the induction of radiofrequency signal away from
From being 4 to 8 meters, penetrability is strong.In addition, can be by using the RFID label tag of anti-metal, to avoid the radiofrequency signal of hyperfrequency
It is shielded and interferes by iron door and window.
After RFID label tag 100 receives the radiofrequency signal of the transmitting of antenna 200, the inductive signal of corresponding frequencies can be emitted, it can
Inductive signal is interpreted as to the label signal of RFID label tag.In embodiments of the present invention, RFID reader 300 passes through antenna 200
Determine received signal intensity instruction (the Received Signal Strength of the RFID label tag signal of specific electron coding
Indication, RSSI) value, to judge the switch state of door and window.
It should be noted that the RSSI value of inductive signal (label signal) is by emitting the RFID label tag 100 of the inductive signal
With the distance between antenna 200 and the angle decision for receiving the inductive signal.Therefore, the intensity value of RFID label tag signal can be passed through
Judge the switch state of door and window.
Preferably, the door in the present invention refers to logistics chamber door, and RFID label tag is common in box for material circulation, using RFID rather than is passed
Detection device of uniting can reduce installation cost to avoid new functional module is added.But this door opening and closing state detection method is applicable in
On all door and windows, it is not limited to box for material circulation.
In one example, 4 groups of antennas and 5 groups of RFID label tags can be used to increase characteristic amount for the present invention, same
Under kind test mode, 20 groups of signal strength indications can be obtained, the performance of algorithm classification device is greatly improved.But specific antenna and RFID
Number of labels can depend on the circumstances, and general no less than 3 groups.
Preferably, the present invention in RFID label tag installation site be Men Zhongxin, door close to shaft side, door close to lateral edge one
Edge and door lower edge on side, door are maximized and are spaced between label.But specific installation site can according to number of labels optionally and
It is fixed.
Preferably, the machine learning algorithm in the present invention uses support vector machines.SVM algorithm is most healthy and strong, most accurate
One of machine learning method, two classical classification problems can be completed.We are using the RSSI value of RFID label tag as data spy
Sign, is regarded as two states for the Push And Release of door, and RFID label tag signal strength and door opening and closing state can be obtained by SVM algorithm
Relationship.Certainly, it can also complete similarly to work using other algorithms, principle is similar, also belongs to the contents of the present invention.
In a specific application, switch gate detection device provided by the invention and method can be led applied to security protection
Domain, such as the switch state of detectable door or window give a warning prompt when detecting that door or window are illegally turned on or off,
To avoid bring the person and property safety hidden danger.
In one example, the switch gate condition detection method and device which provides are applicable to box for material circulation switch
Safety testing field.Correspondingly, according to an embodiment of the invention, providing a kind of for detecting the dress of logistics door switch state
It sets, as shown in Figure 2.Device includes power supply, the RFID reader in box for material circulation and antenna, the RFID mark positioned at logistics chamber door
Label.Preferably, 5 groups of RFID label tag installation sites be Men Zhongxin, door close to shaft side, door close to lateral edge side, Men Shangyan
With door lower edge.For example, selecting centre frequency is the ultra-high frequency RFID label of 915MHz, distance of reaction is 4 to 8 meters, it is ensured that signal
Cover entire logistics box body.Power supply is used to power to each equipment, passes through ether between the host and RFID reader of processing unit
Net connection, by cable connection between reader and antenna, antenna is communicated with RFID label tag by radio.RFID read-write
Device identifies 5 RFID label tags for judging door opening and closing state by antenna, so that it may obtain RSSI from data packet and receive letter
Number intensity value.Door, due to RFID reader and label relative position difference, reads mark in different switch states and angle
The RSSI value of label has very big difference.We record the corresponding door opening and closing state of different RSSI values, by SVM support to
The training of amount machine algorithm, the method that can obtain judging door opening and closing state by signal strength.
SVM (Support Vector Machine), i.e. support vector machines are a kind of two sorting algorithms of classics.It is intended to
It finds the maximum hyperplane in interval to separate two class samples, interval refers to the space size or super between two categories
The separation degree that plane determines, that is, the geometric distance between data point and hyperplane, largest interval can ensure that the hyperplane has
There is best generalization ability, the new data with same distribution also can be predicted accurately.We pass through the distribution of given data
Situation finds the plane that can separate data category, two classifiers can be obtained, for the new data to distribution of the same race
Carry out classification judgement.
In some cases, some classifying rules can correctly classify most of data point, only a few data
The reason of putting there are mistake, causing such case is likely to be artificial error, we cannot deny that the classifying rules is wrong
Accidentally.SVM algorithm extends rigid space requirement then to allow a small amount of such noise to exist, and introduces slack variable ξ and punishment
Factor C carrys out the unlawful practice of quantitative classification device.Slack variable ξ indicates some distance of mistake point example to hyperplane, penalty factor
For the complexity of balancing machine and the quantity of inseparable data point；In other cases, even if joined this soft margin
Optimization, still cannot separate two class data points well.It is proposed in Cover mode separability theorem, complicated pattern classification is asked
Topic is embedded into after higher dimensional space by Nonlinear Mapping than being more likely to linearly be separated in lower dimensional space.Then, we can be with
Data-oriented can appropriately be calculated to the kernel function of inner product by defining one, and data are empty from the input space nonlinear transformation to feature
Between, so that data are converted into linear separability within this space.
The process of the door opening and closing state detection method proposed using this example is described below:
First, it would be desirable to obtain test data and be used to train sorter model.Preferably, using 4 groups of antennas and 5 groups
The collocation of RFID label tag, every group of antenna can read the signal strength indication of all 5 groups of labels, it is strong can to obtain 20 groups of signals every time
Degree evidence.We first close chamber door, persistently read 5000 groups or more signal strength datas, the state of these data is labeled as
1, it refers to door and correctly closes.Then chamber door is opened, degree records the signal strength data under every 10 degree of different angles from 0 to 90,
Totally 5000 groups, the state of these data is labeled as -1, door is referred to and opens.By the RSSI value of the lower five groups of RFID label tags of every kind of state
As the data point of five dimensions, the building of SVM algorithm progress classifier is inputed to together with the label of corresponding states.
It should be noted that the demand about the signal strength data gathered in advance for training classifier can basis
Depending on actual conditions, data volume is bigger, and obtained classifier is more accurate, above to be only illustrated for 5000 groups, and does not have to
It does to any restriction of the invention.
Then, we use obtained test data train classification models.Major Software Development Platform has for engineering
The development kit of habit, preferably, we are trained data using the SVM function in LibSVM software package.For original number
According to, it should be first by data zooming to a certain range, to prevent certain features excessive or too small, role is uneven in training
Weighing apparatus, while promoting whole calculating speed.A kind of preferred scheme is to use the SVM-Scale function pair in LibSVM software package
Initial data zooms in and out, computation rule are as follows:
Wherein y is data before scaling, and y ' is the data after scaling；Lower is the data lower bound of scaling, and upper is scaling
The data upper bound；Min is the data minimum value of dimension a, and max is the data maximums of dimension a.Choosing lower herein is -1,
Upper is 1, and data are carried out linear transformation.
After the completion of scaling, we are preferred using the grid.py function progress parameter in LibSVM software package.Due to characteristic
Divide according to unobvious, then chooses Gauss radial direction kernel optimization.It needs to be determined that parameter mainly have Gauss radial direction kernel functionFactor Gamma value and above-mentioned penalty factor.
It can start train classification models after parameter is selected, use the svm_train function in LibSVM software package, input
Parameters numerical value can train two classifiers obtained under the situation.The host in classifier deposit case that training is obtained
Among, complete entire door opening and closing state detection device.
Finally, we also need to test obtained door opening and closing state detection device.Any switch gate is to different angles
Degree and position, the correctness of comparator-sorter judging result, and with amendment sorter model, raising classifier performance.
In use, host and RFID reader keep connecting by Ethernet, and reader is obtained in real time by antenna
The signal strength RSSI value of RFID label tag is taken to be transferred to host as new data characteristics, host is according to two classifiers pair of foundation
The data point is predicted that, if data point drops into the region of the instruction of classifier 1, we are considered as door at this time and correctly close；If
Data point drops into region shown in classifier -1, we are considered as door at this time and open.It can also be preset in host
Further operating under door different conditions, completes the functions such as open alarm.
Certainly, an only specific example described herein, the present invention includes but is not limited to this example.For example,
RFID label tag installation site, which can choose, to be mounted on four angles of chamber door；The particular number of label and antenna is no less than 3
It can arbitrarily arrange in pairs or groups under the premise of group；The sorting algorithm of use can be replaced CART algorithm etc..Can according to specific implementation situation depending on,
But the detection method and device for this door opening and closing state that this patent is proposed are not influenced.
Fig. 3 is the switch gate condition detection method flow diagram provided by the invention based on RFID, as shown in figure 3, packet
Step S10 is included to step S40.
S10 emits at least three inductive signals, at least three radio frequency identification by least three RFID tags
The different location that label is mounted on a door.
S20 receives at least one set of inductive signal by least one antenna, and every group of inductive signal includes described at least three
RFID tag is emitted at least three inductive signals of the same antenna, at least one described antenna is mounted in door empty
Between.
S30 determines at least one set inductive signal intensity data, and every group of inductive signal intensity data is according to every group of induction
The intensity of each inductive signal determines in signal, and the intensity of each inductive signal is by emitting the RFID tag of the inductive signal
With the distance between antenna and the angle decision for receiving the inductive signal.
S40 determines institute according to the classifier data that at least one set of inductive signal intensity data and preparatory training obtain
The switch state of door is stated, the classifier that the preparatory training obtains is in advance by the door and is opened and closed under two states
The multiple groups inductive signal intensity data training for classification received in advance of at least one described antenna obtains.
Specifically, the refinement process of each step of the above method can be found in earlier figures 1, the introduction in Fig. 2, not do herein superfluous
It states.
In view of popularizing for Internet of Things, RFID frequency read/write and label are housed in existing box for material circulation, generally for identifying object
The cargo in case is flowed, the switch state for detecting door can be considered through received RFID label tag signal strength indication in we.RFID
Label has unique electronic code, to improve the safety of system.The switch state of logistics chamber door is detected using RFID
It does not need that new module is additionally installed, the cost of device can be reduced, and do not need the support of network, reading rate is fast, reads
Reading ability is strong.
The present invention uses the RFID reader and label of hyperfrequency, to increase distance of reaction range；Increase number of tags,
According to the switch of multiple label comprehensive descision doors, anti-interference is improved；The data being collected into are analyzed using the algorithm of machine learning,
Rather than simple basis judges whether detect label, promotes accuracy.
As it will be easily appreciated by one skilled in the art that the foregoing is merely illustrative of the preferred embodiments of the present invention, not to
The limitation present invention, any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should all include
Within protection scope of the present invention.
Claims (6)
1. a kind of switch gate condition checkout gear based on RFID characterized by comprising at least three radio frequency discrimination RFID marks
Label, at least one antenna, radio-frequency identification reader/writer and processing unit；
The different location that at least three RFID tag is mounted on a door, for emitting at least three inductive signals；Institute
State at least three RFID tags installation site be Men Zhongxin, door close to shaft side, door close to lateral edge side, Men Shangyan
And at least three positions in door lower edge；
At least one described antenna is mounted on space in door, and for receiving at least one set of inductive signal, every group of inductive signal includes
At least three RFID tag is emitted at least three inductive signals of the same antenna；
The radio-frequency identification reader/writer is for determining at least one set of inductive signal intensity data, every group of inductive signal intensity data root
Determine that the intensity of each inductive signal is by emitting the inductive signal according to the intensity of each inductive signal in every group of inductive signal
RFID tag and receive the distance between antenna and the angle decision of the inductive signal；
The processing unit, the classifier for being obtained according at least one set of inductive signal intensity data and preparatory training are true
The switch state of the fixed door, the classifier that the preparatory training obtains are in advance by the door and open and close two kinds of shapes
The multiple groups inductive signal intensity data training for classification received in advance of at least one described antenna obtains under state；The processing
Unit is used for according to the inductive signal intensity data gathered in advance and support vector machines algorithm building door opening and closing state point
Class device, the door opening and closing state classifier are used to the multiple groups inductive signal intensity data for being used to classify carrying out two classification,
The intensity data of multidimensional inductive signal is transformed into door opening and closing state and corresponds to 2-D data, by door in advance it is in the open state under
Multiple groups inductive signal intensity data for classification is labeled as first state data, is used under door is in close state in advance
The multiple groups inductive signal intensity data of classification is labeled as the second status data, when at least one set of inductive signal intensity data is fallen
When entering in the first state data area, the door is in the open state, when at least one set of inductive signal intensity data
When falling within the scope of second status data, the door is in close state.
2. switch gate condition checkout gear according to claim 1, which is characterized in that the antenna is also used to emit radio frequency
Signal, so that the RFID tag emits the inductive signal under the excitation of the radiofrequency signal.
3. switch gate condition checkout gear according to claim 1, which is characterized in that the processing unit is according to for dividing
Class multiple groups inductive signal intensity data training classifier, it is necessary first to the multiple groups inductive signal intensity data for classification into
Row pretreatment completes data screening and zoom operations so that data are suitable for processing；Then, it is selected according to data type and distribution situation
Suitable training parameter is taken, training data and parameter are input in machine learning algorithm, training obtains the door opening and closing state
Classifier.
4. a kind of switch gate condition detection method based on RFID characterized by comprising
Emit at least three inductive signals, at least three RFID tags installation by least three RFID tags
Different location on door；The installation site of at least three RFID tag is Men Zhongxin, door close to shaft side, door
Close to edge on lateral edge side, door and at least three positions in door lower edge；
At least one set of inductive signal is received by least one antenna, every group of inductive signal includes at least three radio frequency identification
Label is emitted at least three inductive signals of the same antenna, at least one described antenna is mounted on space in door；
Determine at least one set inductive signal intensity data, every group of inductive signal intensity data is according to every in every group of inductive signal
The intensity of a inductive signal determines that the intensity of each inductive signal is somebody's turn to do by emitting the RFID tag of the inductive signal and receiving
The distance between antenna of inductive signal and angle determine；
The switch shape of the door is determined according to the classifier that at least one set of inductive signal intensity data and preparatory training obtain
State, the classifier that the preparatory training obtains be in advance by the door open and close under two states it is described at least one
The antenna multiple groups inductive signal intensity data training for classification received in advance obtains；Believed according to the induction gathered in advance
Number intensity data and support vector machines algorithm construct door opening and closing state classifier, and the door opening and closing state classifier is used for will
Door is in the multiple groups induction letter for classifying received in advance of at least one described antenna under opening and closing two states in advance
Number intensity data carries out two classification, and the intensity data of multidimensional inductive signal is transformed into door opening and closing state and correspond to 2-D data, general
The multiple groups inductive signal intensity data for classification under door is in the open state in advance is labeled as first state data, and door is pre-
The multiple groups inductive signal intensity data for classification under being first in close state is labeled as the second status data；When it is described at least
When one group of inductive signal intensity data is fallen into the first state data area, the door is in the open state, when it is described extremely
When few one group of inductive signal intensity data is fallen within the scope of second status data, the door is in close state.
5. switch gate condition detection method according to claim 4, which is characterized in that further include:
Emit radiofrequency signal by the antenna, so that the RFID tag emits institute under the excitation of the radiofrequency signal
State inductive signal.
6. switch gate condition detection method according to claim 4, which is characterized in that according to the multiple groups induction for classification
Signal strength data trains classifier, it is necessary first to the multiple groups inductive signal intensity data for classification pre-processed, it is complete
At data screening and zoom operations so that data are suitable for processing；Then, suitable instruction is chosen according to data type and distribution situation
Practice parameter, training data and parameter are input in machine learning algorithm, training obtains the door opening and closing state classifier.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710693758.1A CN107491799B (en) | 2017-08-14 | 2017-08-14 | A kind of switch gate condition detection method and device based on RFID |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710693758.1A CN107491799B (en) | 2017-08-14 | 2017-08-14 | A kind of switch gate condition detection method and device based on RFID |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107491799A CN107491799A (en) | 2017-12-19 |
| CN107491799B true CN107491799B (en) | 2019-11-12 |
Family
ID=60645602
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710693758.1A Active CN107491799B (en) | 2017-08-14 | 2017-08-14 | A kind of switch gate condition detection method and device based on RFID |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107491799B (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108222674B (en) * | 2018-01-02 | 2020-06-09 | 深圳绿米联创科技有限公司 | Door lock state detection method and device and door lock |
| CN111882702B (en) * | 2020-07-02 | 2022-11-15 | 合肥美的智能科技有限公司 | Door locking method, door locking control device, container and computer readable storage medium |
| CN112097433B (en) * | 2020-09-29 | 2022-08-19 | 青岛海尔电冰箱有限公司 | Refrigerator and control method thereof |
| CN113960943B (en) * | 2021-09-23 | 2024-06-21 | 德迅创新(深圳)有限公司 | Door state detection method based on multi-bin public service intelligent device |
| CN117114031A (en) * | 2023-07-17 | 2023-11-24 | 广东金朋科技有限公司 | Door opening state detection method and system and intelligent home system |
| CN118036634B (en) * | 2024-04-11 | 2024-07-05 | 南京大全电气研究院有限公司 | RFID cabinet door state detection method, system, electronic equipment and storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101782652A (en) * | 2010-02-08 | 2010-07-21 | 上海和为科技有限公司 | Indoor positioning system based on RFID technology |
| CN202244899U (en) * | 2011-10-12 | 2012-05-30 | 四川科泰智能电子有限公司 | Safe intelligent detection device for satellite positioning container |
| KR20130082969A (en) * | 2011-12-26 | 2013-07-22 | 에스에이치 주식회사 | Desiccator for liquid crystal, method and system for administrating liquid crystal by using rfid |
| CN103793978A (en) * | 2012-10-30 | 2014-05-14 | 大陆汽车系统公司 | Determining the state of a garage door using vehicle sensors |
| CN105629324A (en) * | 2014-12-01 | 2016-06-01 | 丰唐物联技术（深圳）有限公司 | Method and device for detecting opening and closed state of door |
-
2017
- 2017-08-14 CN CN201710693758.1A patent/CN107491799B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101782652A (en) * | 2010-02-08 | 2010-07-21 | 上海和为科技有限公司 | Indoor positioning system based on RFID technology |
| CN202244899U (en) * | 2011-10-12 | 2012-05-30 | 四川科泰智能电子有限公司 | Safe intelligent detection device for satellite positioning container |
| KR20130082969A (en) * | 2011-12-26 | 2013-07-22 | 에스에이치 주식회사 | Desiccator for liquid crystal, method and system for administrating liquid crystal by using rfid |
| CN103793978A (en) * | 2012-10-30 | 2014-05-14 | 大陆汽车系统公司 | Determining the state of a garage door using vehicle sensors |
| CN105629324A (en) * | 2014-12-01 | 2016-06-01 | 丰唐物联技术（深圳）有限公司 | Method and device for detecting opening and closed state of door |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107491799A (en) | 2017-12-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107491799B (en) | A kind of switch gate condition detection method and device based on RFID | |
| RU2519467C2 (en) | System and method for detection of electronic article surveillance marker shielding | |
| EP3373035A1 (en) | Systems and methods for object localization and path identification based on rfid sensing | |
| EP0689161A2 (en) | Identification system | |
| WO2016086135A2 (en) | Intelligent security management system | |
| WO2018157101A1 (en) | Rfid checkout station using virtual shielding | |
| CN206863759U (en) | RFID formula logistics doors | |
| CN209149424U (en) | A kind of latch system of unmanned vending machine | |
| US20060255127A1 (en) | System, method, and computer program product for biometric radiofrequency id | |
| CN103268532A (en) | Intelligent campus bicycle anti-theft and information management system and method based on RFID | |
| CN111599119A (en) | Self-service police article borrowing and returning system | |
| Hauser et al. | Pushing the limits of RFID: Empowering RFID-based electronic article surveillance with data analytics techniques | |
| CN103106376A (en) | Method for identifying radio frequency tag | |
| RU2352995C1 (en) | Radio-wave detection method of moving objects | |
| CN208314799U (en) | The noninductive control device of RFID and system based on range sensor | |
| Tellbach et al. | Wireless material identification in the recycling chain using chipless RFID Tags | |
| CN104169983A (en) | Method and system for adaptive sliding door pattern cancellation in metal detection | |
| CN106056339B (en) | A kind of article identification method using electromagnetic wave penetrability | |
| CN112216000A (en) | Gate machine | |
| KR102553207B1 (en) | Multi-use beverage cup collection apparatus | |
| CN208538157U (en) | A kind of label apparatus and object translocation identifying system based on RFID technique | |
| Sumi et al. | Application of RFID technology in libraries | |
| Zhu et al. | Environmentally adaptive real-time detection of RFID false readings in a new practical scenario | |
| Choong et al. | Investigation of Features for Classification RFID Reading Between Two RFID Reader in Various Support Vector Machine Kernel Function | |
| CN118036634B (en) | RFID cabinet door state detection method, system, electronic equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |