CN110874875B - Door lock control method and device - Google Patents
Door lock control method and device Download PDFInfo
- Publication number
- CN110874875B CN110874875B CN201810916855.7A CN201810916855A CN110874875B CN 110874875 B CN110874875 B CN 110874875B CN 201810916855 A CN201810916855 A CN 201810916855A CN 110874875 B CN110874875 B CN 110874875B
- Authority
- CN
- China
- Prior art keywords
- action
- executed
- behavior data
- user
- sample
- 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
- 238000000034 method Methods 0.000 title claims abstract description 48
- 230000009471 action Effects 0.000 claims abstract description 141
- 230000006399 behavior Effects 0.000 claims abstract description 103
- 230000004044 response Effects 0.000 claims abstract description 31
- 238000010801 machine learning Methods 0.000 claims abstract description 10
- 238000012549 training Methods 0.000 claims description 76
- 238000007781 pre-processing Methods 0.000 claims description 51
- 238000013528 artificial neural network Methods 0.000 claims description 37
- 210000002569 neuron Anatomy 0.000 claims description 36
- 230000006870 function Effects 0.000 claims description 22
- 230000008921 facial expression Effects 0.000 claims description 14
- 238000010586 diagram Methods 0.000 claims description 13
- 238000012545 processing Methods 0.000 claims description 13
- 239000011159 matrix material Substances 0.000 claims description 8
- 230000003542 behavioural effect Effects 0.000 claims description 7
- 230000004424 eye movement Effects 0.000 claims description 7
- 230000000306 recurrent effect Effects 0.000 claims description 6
- 230000001815 facial effect Effects 0.000 claims description 3
- 238000005265 energy consumption Methods 0.000 abstract description 5
- 230000000875 corresponding effect Effects 0.000 description 29
- 230000001276 controlling effect Effects 0.000 description 11
- 230000008569 process Effects 0.000 description 9
- 230000008878 coupling Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 238000004891 communication Methods 0.000 description 2
- 238000005034 decoration Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000013507 mapping Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000035945 sensitivity Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000035699 permeability Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G07—CHECKING-DEVICES
- G07C—TIME OR ATTENDANCE REGISTERS; REGISTERING OR INDICATING THE WORKING OF MACHINES; GENERATING RANDOM NUMBERS; VOTING OR LOTTERY APPARATUS; ARRANGEMENTS, SYSTEMS OR APPARATUS FOR CHECKING NOT PROVIDED FOR ELSEWHERE
- G07C9/00—Individual registration on entry or exit
- G07C9/00174—Electronically operated locks; Circuits therefor; Nonmechanical keys therefor, e.g. passive or active electrical keys or other data carriers without mechanical keys
- G07C9/00563—Electronically operated locks; Circuits therefor; Nonmechanical keys therefor, e.g. passive or active electrical keys or other data carriers without mechanical keys using personal physical data of the operator, e.g. finger prints, retinal images, voicepatterns
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Lock And Its Accessories (AREA)
Abstract
The invention discloses a door lock control method and device. Wherein, the method comprises the following steps: acquiring behavior data of a user; analyzing the behavior data by using a first model, and determining the action to be executed of the user, wherein the first model is trained by machine learning by using a plurality of groups of data, and each group of data in the plurality of groups of data comprises the behavior data and the action to be executed obtained from the behavior data; and controlling the door lock to execute corresponding response operation according to the action to be executed. The invention solves the technical problem of higher energy consumption of the door lock caused by the simultaneous awakening of the password input module and the fingerprint identification module of the door lock.
Description
Technical Field
The invention relates to the field of door lock control, in particular to a door lock control method and device.
Background
At present, the permeability of the intelligent door lock in daily life of people is higher and higher, most of the intelligent door locks have a password input module and a fingerprint identification module, and the intelligence of the intelligent door locks also brings great convenience to the life of people.
In the prior art, in the aspect of awakening the password input module and the fingerprint identification module of the door lock, the following situations can occur: the hands of the user are required to be very close (the user experience of the intelligent door lock is reduced), the two modules can be awakened at the same time, however, the user only uses one module to unlock the door lock, and therefore the energy consumption of the door lock is increased; or an unfamiliar user performs illegal operation on the door lock, such as password input and fingerprint input in a mess, user password stealing and door burglary, and the like, which all reduce the safety and the durability of the intelligent door lock.
In view of the above problems, no effective solution has been proposed.
Disclosure of Invention
The embodiment of the invention provides a door lock control method and device, which at least solve the technical problem of high energy consumption of a door lock caused by the fact that a password input module and a fingerprint identification module of the door lock are simultaneously awakened.
According to an aspect of an embodiment of the present invention, there is provided a door lock control method including: acquiring behavior data of a user; analyzing the behavior data by using a first model, and determining the action to be executed of the user, wherein the first model is trained by machine learning by using a plurality of groups of data, and each group of data in the plurality of groups of data comprises the behavior data and the action to be executed obtained from the behavior data; and controlling the door lock to execute corresponding response operation according to the action to be executed.
Optionally, the method of establishing the first model includes: acquiring behavior data and actions to be executed in a training sample; and modeling the behavior data and the action to be executed in the training sample through a back propagation algorithm to obtain the first model.
Optionally, the first model comprises an input layer, a hidden layer and an output layer; wherein, the modeling the behavior data and the action to be executed in the training sample by the back propagation algorithm to obtain the first model comprises: preprocessing the behavior data in the training sample to obtain a sample preprocessing result, wherein the sample preprocessing result comprises at least one of the following: inputting a vector, an input parameter matrix and a vector diagram; transmitting the sample preprocessing result to the hidden layer through the input layer, wherein the sample preprocessing result is coded in each neuron of the hidden layer to obtain a sample action, and the sample action is output through the output layer; comparing the sample action with the action to be performed in the training sample; if the error between the sample action and the action to be executed in the training sample does not meet a preset condition, calculating the square sum of the error, and taking the square sum as a target function; calculating a partial derivative of the target function to the weight of each neuron to obtain a gradient amount corresponding to the weight; and updating the weight by adopting a gradient descent algorithm according to the gradient amount.
Optionally, the hidden layer comprises a tree graph structure including a child node and a parent node, each neuron in the child node being connected to each neuron in the parent node.
Optionally, the analyzing the behavior data by using the first model, and the determining the action to be performed by the user includes: the behavior data of the user is preprocessed to obtain a first preprocessing result; and transmitting the first preprocessing result to the hidden layer through the input layer, wherein the first preprocessing result is coded in each neuron of the hidden layer to obtain the action to be executed, and the action to be executed is output through the output layer.
Optionally, the controlling the door lock to perform the corresponding response operation according to the action to be performed includes at least one of: the method comprises the steps of starting the biological characteristic identification module, starting the password input module, closing the biological characteristic identification module and closing the password input module.
Optionally, the biometric characteristic comprises at least one of: fingerprint features, palm print features, voice print features, facial features, limb movements, iris features.
Optionally, after controlling the door lock to perform a corresponding response operation according to the action to be performed, the method further includes: in a learning mode, receiving input scores corresponding to the response operations; outputting a training result of the first model according to the score, wherein the training result is used for indicating the matching degree between the action to be executed and the actual action executed by the user; marking corresponding behavior data according to the training result; and inputting the marked behavior data into the training sample.
Optionally, the acquiring the behavior data of the user includes: collecting the behavioral data of the user by a sensor, the behavioral data including at least one of: limb movements, facial expressions, voice, eye movements.
Optionally, the sensor comprises at least one of: camera, distance sensor, iris recognition device.
Optionally, the first model is a recurrent neural network RNN.
According to another aspect of the embodiments of the present invention, there is also provided a door lock control device including: the acquiring unit is used for acquiring behavior data of a user; the processing unit is used for analyzing the behavior data by using a first model and determining the action to be executed of the user, wherein the first model is trained by machine learning by using a plurality of groups of data, and each group of data in the plurality of groups of data comprises the behavior data and the action to be executed obtained from the behavior data; and the control unit is used for controlling the door lock to execute corresponding response operation according to the action to be executed.
Optionally, the obtaining unit is further configured to obtain behavior data and an action to be performed in the training sample; the processing unit is further configured to model the behavior data and the action to be executed in the training sample through a back propagation algorithm to obtain the first model.
Optionally, the first model comprises an input layer, a hidden layer and an output layer; the processing unit is used for performing the following steps to model the behavior data and the action to be executed in the training sample through a back propagation algorithm to obtain the first model: preprocessing the behavior data in the training sample to obtain a sample preprocessing result, wherein the sample preprocessing result comprises at least one of the following: inputting a vector, an input parameter matrix and a vector diagram; transmitting the sample preprocessing result to the hidden layer through the input layer, wherein the sample preprocessing result is coded in each neuron of the hidden layer to obtain a sample action, and the sample action is output through the output layer; comparing the sample action with the action to be performed in the training sample; if the error between the sample action and the action to be executed in the training sample does not meet a preset condition, calculating the square sum of the error, and taking the square sum as a target function; calculating a partial derivative of the target function to the weight of each neuron to obtain a gradient amount corresponding to the weight; and updating the weight by adopting a gradient descent algorithm according to the gradient amount.
Optionally, the hidden layer comprises a tree graph structure including a child node and a parent node, each neuron in the child node being connected to each neuron in the parent node.
Optionally, the processing unit is configured to perform the following steps of analyzing the behavior data by using a first model, and determining an action to be performed by the user: the behavior data of the user is preprocessed to obtain a first preprocessing result; and transmitting the first preprocessing result to the hidden layer through the input layer, wherein the first preprocessing result is coded in each neuron of the hidden layer to obtain the action to be executed, and the action to be executed is output through the output layer.
Optionally, the control unit is configured to execute at least one of the following steps to control the door lock to execute a corresponding response operation according to the action to be executed: the method comprises the steps of starting the biological characteristic identification module, starting the password input module, closing the biological characteristic identification module and closing the password input module.
Optionally, the apparatus further comprises: the receiving unit is used for receiving the input scores corresponding to the response operations in the learning mode; the output unit is used for outputting a training result of the first model according to the score, wherein the training result is used for indicating the matching degree between the action to be executed and the actual action executed by the user; marking corresponding behavior data according to the training result; and inputting the marked behavior data into the training sample.
In the embodiment of the invention, behavior data of a user is acquired; analyzing the behavior data by using a first model, and determining the action to be executed of the user, wherein the first model is trained by machine learning by using a plurality of groups of data, and each group of data in the plurality of groups of data comprises the behavior data and the action to be executed obtained from the behavior data; according to the action to be executed, controlling the door lock to execute a corresponding response operation mode, training a first model by using collected behavior data through machine learning, analyzing the current behavior data of a user based on the first model, identifying the action to be executed by the user, and further controlling the door lock to perform corresponding operation control, thereby achieving the purpose of realizing that the door lock predicts the action of the user in advance and makes a response, further achieving the technical effects of improving the operation sensitivity of the door lock and reducing the power consumption of the door lock, and further solving the technical problem that the energy consumption of the door lock is large due to the fact that a password input module and a fingerprint identification module of the door lock can be awakened at the same time.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the invention and together with the description serve to explain the invention without limiting the invention. In the drawings:
FIG. 1 is a schematic flow diagram of an alternative door lock control method according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of an alternative neural network according to an embodiment of the present invention;
FIG. 3 is a schematic diagram of an alternative neural network according to an embodiment of the present invention;
FIG. 4 is a schematic flow diagram of an alternative door lock control method according to an embodiment of the present invention;
fig. 5 is a schematic structural diagram of an alternative door lock control device according to an embodiment of the present invention.
Detailed Description
In order to make the technical solutions of the present invention better understood, 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.
It should be noted that the terms "first," "second," and the like in the description and claims of the present invention and in the drawings described above are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used is interchangeable under appropriate circumstances such that the embodiments of the invention described herein are capable of operation in sequences other than those illustrated or described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
Example 1
In accordance with an embodiment of the present invention, there is provided a method embodiment of a door lock control method, it should be noted that the steps illustrated in the flowchart of the accompanying drawings may be performed in a computer system such as a set of computer executable instructions, and that while a logical order is illustrated in the flowchart, in some cases the steps illustrated or described may be performed in an order different than that herein.
Fig. 1 is a door lock control method according to an embodiment of the present invention, as shown in fig. 1, the method including the steps of:
step S102, behavior data of the user is obtained.
Optionally, the acquiring the behavior data of the user includes: collecting the behavioral data of the user by a sensor, the behavioral data including at least one of: limb movements, facial expressions, voice, eye movements.
Wherein the sensor comprises at least one of: camera, distance sensor, iris recognition device.
In this embodiment, one or more of the body behavior, the eye behavior, and the facial expression (i.e., the behavior data described above) of the user are acquired by using a camera, a distance sensor, iris recognition, and other devices, and the data that cannot be recognized is not acquired.
And step S104, analyzing the behavior data by using the first model, and determining the action to be executed of the user.
The first model is trained through machine learning by using a plurality of groups of data, and each group of data in the plurality of groups of data comprises behavior data and an action to be executed obtained from the behavior data. The first model is the recurrent neural network RNN.
In this embodiment, the effective raw data of the user's body behavior, eye movement, facial expression, etc. is subjected to parameterization preprocessing, and converted into an input vector, an input parameter matrix, or a vector diagram, to obtain an input data set and data distribution, which are used as data inputs of a neural network algorithm function set (i.e., the first model).
Optionally, the method of establishing the first model includes: acquiring behavior data and actions to be executed in a training sample; and modeling the behavior data and the action to be executed in the training sample through a back propagation algorithm to obtain the first model.
Optionally, the first model comprises an input layer, a hidden layer and an output layer.
As shown in fig. 2, the network structure of the first model is divided into an input layer, an output layer and a hidden layer, wherein the hidden layer may be one or more layers. The input layer is the neural network input preprocessing data. The output layer is related control parameters of the door lock, such as parameters for independently opening (or closing) the password input module (fingerprint identification module) and the password touch module (fingerprint identification module) without feedback. The hidden layer is formed by training and analyzing various input parameters (behavior action 1, behavior action 2, … … and behavior action 6) obtained from the input layer by adopting various related neural network algorithms through each node (hidden node 1, hidden node 2, … … and hidden node n), and finally transmitting the input parameters to the output layer to obtain various output parameters (opening a fingerprint identification module, opening a password input module, closing the fingerprint identification module, closing the password input module and other door lock control modes).
Optionally, the hidden layer comprises a tree graph structure including a child node and a parent node, each neuron in the child node being connected to each neuron in the parent node.
Optionally, the analyzing the behavior data by using the first model, and the determining the action to be performed by the user includes: the behavior data of the user is preprocessed to obtain a first preprocessing result; and transmitting the first preprocessing result to the hidden layer through the input layer, wherein the first preprocessing result is coded in each neuron of the hidden layer to obtain the action to be executed, and the action to be executed is output through the output layer.
The first model of the present embodiment may use a recurrent neural network (also called a recurrent neural network). The recurrent neural network solves the problem of combinability, and each meaningful information in the input layer is represented into a meaningful vector after being processed by the hidden layer, and the vector is transmitted to the output layer. A tree graph structure is constructed in a hidden layer, the input in the network is a child node, the output after the child node is coded is a father node, the child node and the father node form a fully-connected neural network, namely, each neuron of the child node is connected with each neuron of the father node in pairs, and in the artificial neural network algorithm network structure, as shown in fig. 3, in the network formed by C1, C2 and P1, C1 and C2 are child nodes, and P1 is a father node; in the network formed by P1 and P2, P1 is a child node, P2 is a parent node, and so on. Therefore, data such as user behavior and action, eye action, facial expression and the like can be coded, different syntax parse trees correspond to different information, and the air conditioner can make corresponding actions through coding of the parse trees, such as: and independently starting (or closing) the password input module (fingerprint identification module) and identifying the user identity so as to realize actions such as whether the password touch module (fingerprint identification module) feeds back or not.
Optionally, the modeling the behavior data and the action to be executed in the training sample by using a back propagation algorithm to obtain the first model includes: preprocessing the behavior data in the training sample to obtain a sample preprocessing result, wherein the sample preprocessing result comprises at least one of the following: inputting a vector, an input parameter matrix and a vector diagram; transmitting the sample preprocessing result to the hidden layer through the input layer, wherein the sample preprocessing result is coded in each neuron of the hidden layer to obtain a sample action, and the sample action is output through the output layer; comparing the sample action with the action to be performed in the training sample; if the error between the sample action and the action to be executed in the training sample does not meet a preset condition, calculating the square sum of the error, and taking the square sum as a target function; calculating a partial derivative of the target function to the weight of each neuron to obtain a gradient amount corresponding to the weight; and updating the weight by adopting a gradient descent algorithm according to the gradient amount.
For the first model, the present embodiment employs a dynamic training algorithm to train the neural network, and specifically may employ a back propagation algorithm, where:
and (3) a back propagation algorithm: calculating the output value of each neuron in a forward direction; calculating the error term value for each neuron back, which is the partial derivative of the error function to the weighted input of the neuron, and back-propagating the residual in the error from the current time Tn to the initial time T1; calculating a gradient of each weight; and finally, updating the weight by using a random gradient descent algorithm. And then, through repeated training and testing of a large amount of data, the optimal weight is selected, and the neural network training is completed.
The input-output relationship of the back propagation algorithm is essentially a mapping relationship: an n-input m-output back-propagation neural network performs the function of continuously mapping from n-dimensional euclidean space to a finite field in m-dimensional euclidean space, which is highly non-linear.
The learning process of the back propagation algorithm consists of a forward propagation process and a back propagation process. In the forward propagation process, input information passes through the hidden layer through the input layer, is processed layer by layer and is transmitted to the output layer. If the expected output value cannot be obtained in the output layer, taking the square sum of the output and the expected error as a target function, switching the target function into a back propagation process, calculating the partial derivative of the target function to each neuron weight layer by layer in the back propagation process to form the gradient of the target function to the weight vector, and taking the gradient as the basis for modifying the weight, wherein the training of the network is completed in the weight modifying process. And ending the neural network training until the error reaches the expected value.
And step S106, controlling the door lock to execute corresponding response operation according to the action to be executed.
Optionally, the controlling the door lock to perform the corresponding response operation according to the action to be performed includes at least one of: the method comprises the steps of starting the biological characteristic identification module, starting the password input module, closing the biological characteristic identification module and closing the password input module.
Wherein the biometric characteristic comprises at least one of: fingerprint features, palm print features, voice print features, facial features, limb movements, iris features.
The unlocking operation of the user is identified by the neural network, one or more kinds of information such as limb actions, eye actions and facial expressions of the user are collected through various sensors, and after the neural network identification, each working module of the door lock makes a correct response aiming at the further unlocking operation of the user, so that the advance foreknowledge of the door lock is realized.
Through the steps, the collected behavior data are used for training the first model through machine learning, the current behavior data of the user are analyzed based on the first model, the action to be executed by the user is identified, and then the door lock is controlled to perform corresponding operation control, so that the aim of predicting the action of the user in advance and responding by the door lock is fulfilled, the technical effects of improving the operation sensitivity of the door lock and reducing the power consumption of the door lock are achieved, and the technical problem that the energy consumption of the door lock is large due to the fact that the password input module and the fingerprint identification module of the door lock can be awakened at the same time is solved.
Optionally, after controlling the door lock to perform a corresponding response operation according to the action to be performed, the method further includes: in a learning mode, receiving input scores corresponding to the response operations; outputting a training result of the first model according to the score, wherein the training result is used for indicating the matching degree between the action to be executed and the actual action executed by the user; marking corresponding behavior data according to the training result; and inputting the marked behavior data into the training sample.
When the neural network identifies the user control, user scores can be increased, sensor data are marked, the marked data are used as a training data set to train the neural network, the identification rate of the neural network on unlocking operation of a specific user is improved, the dynamic training of the neural network is realized, the function of autonomous learning of the door lock is achieved, and the safety of the door lock is improved.
In the learning mode, there may be a scoring operation, which is divided into: the door lock system is accurate (5 scores), general (2 scores) and inaccurate (0 score), wherein the matching degree of the conclusion obtained after the neural network identifies the user behavior data and the further unlocking operation of the user in practice determines the interval selected by the response feedback result of each module of the door lock after the dynamic training data of the user is finished each time, the scores obtained after the dynamic training of each time are accumulated, the user can intuitively and clearly know the learning condition of the door lock, the neural network parameters are continuously optimized until the user does not need to make a scoring mark, and the door lock can make a correct response through certain limb action, eye movement or facial expression of the user, so that the door lock is predicted in advance.
As shown in fig. 4, a description will be given of a door lock control method of the present embodiment:
and a step a, starting.
And b, detecting a user awakening condition.
And c, whether user data is acquired or not.
If yes, executing step d;
if not, executing the step b.
And d, acquiring the behavior action of the user.
The user behavior (corresponding to the behavior data) includes a label, a gesture, a head motion, an eye motion, and the like.
Specifically, by using a camera, a distance sensor, iris recognition, or the like, one or more of the user's body behavior, eye movement, and facial expression are acquired, and the data that cannot be recognized will not be subjected to the acquisition operation.
And e, learning the neural network.
The method comprises the steps of carrying out parameterization preprocessing on effective original data such as user limb behavior action, eye action, facial expression and the like, converting the effective original data into an input vector, an input parameter matrix or a vector diagram, obtaining an input data set and data distribution, and taking the input data set and the data distribution as data input of a neural network algorithm function set.
And f, controlling the door lock.
The data of user behavior and action, eye action, facial expression and the like are coded, different syntax parse trees correspond to different information, and the corresponding action to be made by the air conditioner is obtained through coding of the parse trees, such as: the password input module (fingerprint identification module) is independently opened (or closed), the user identity is identified, and then whether the password touch module (fingerprint identification module) feeds back and other actions are realized, so that the door lock control is realized.
And g, judging whether the learning mode is started or not.
If yes, executing step h;
if not, executing the step b.
And h, judging whether the user is satisfied.
If yes, executing step b;
if not, executing the step e.
In the embodiment, when the neural network identifies the user control, the user score can be increased, the sensor data are marked, the marked data are used as the training data set to train the neural network, the identification rate of the neural network on the unlocking operation of a specific user is improved, the dynamic training of the neural network is realized, the function of autonomous learning of the door lock is achieved, and the safety of the door lock is improved.
In this mode, there is a scoring operation that is divided into: the door lock system is accurate (5 scores), general (2 scores) and inaccurate (0 score), wherein the matching degree of the conclusion obtained after the neural network identifies the user behavior data and the further unlocking operation of the user in practice determines the interval selected by the response feedback result of each module of the door lock after the dynamic training data of the user is finished each time, the scores obtained after the dynamic training of each time are accumulated, the user can intuitively and clearly know the learning condition of the door lock, the neural network parameters are continuously optimized until the user does not need to make a scoring mark, and the door lock can make a correct response through certain limb action, eye movement or facial expression of the user, so that the door lock is predicted in advance.
The invention utilizes an artificial neural network algorithm, corresponding operation control is carried out by a fingerprint identification module and a password input module in the intelligent door lock through collected user limb behavior actions, eye actions and facial expressions, the user scores the responses made by the door lock, the neural network uses the scoring marks to train through the neural network, the neural network parameters are continuously optimized, and the door lock can make correct responses through a certain limb action, eye actions or facial expressions of the user until the user does not need to make the scoring marks, so that the advance foreknowledge of the door lock is realized.
Example 2
According to an embodiment of the present invention, there is provided an embodiment of a door lock control apparatus, and fig. 5 is a door lock control apparatus according to an embodiment of the present invention, as shown in fig. 5, the apparatus including:
an acquisition unit 50 for acquiring behavior data of a user;
a processing unit 52, configured to analyze the behavior data by using a first model, and determine an action to be performed by the user, where the first model is trained through machine learning by using multiple sets of data, and each set of data in the multiple sets of data includes behavior data and an action to be performed obtained from the behavior data;
and the control unit 54 is used for controlling the door lock to execute corresponding response operation according to the action to be executed.
Optionally, the obtaining unit is further configured to obtain behavior data and an action to be performed in the training sample; the processing unit is further configured to model the behavior data and the action to be executed in the training sample through a back propagation algorithm to obtain the first model.
Optionally, the first model comprises an input layer, a hidden layer and an output layer; the processing unit is used for performing the following steps to model the behavior data and the action to be executed in the training sample through a back propagation algorithm to obtain the first model: preprocessing the behavior data in the training sample to obtain a sample preprocessing result, wherein the sample preprocessing result comprises at least one of the following: inputting a vector, an input parameter matrix and a vector diagram; transmitting the sample preprocessing result to the hidden layer through the input layer, wherein the sample preprocessing result is coded in each neuron of the hidden layer to obtain a sample action, and the sample action is output through the output layer; comparing the sample action with the action to be performed in the training sample; if the error between the sample action and the action to be executed in the training sample does not meet a preset condition, calculating the square sum of the error, and taking the square sum as a target function; calculating a partial derivative of the target function to the weight of each neuron to obtain a gradient amount corresponding to the weight; and updating the weight by adopting a gradient descent algorithm according to the gradient amount.
Optionally, the hidden layer comprises a tree graph structure including a child node and a parent node, each neuron in the child node being connected to each neuron in the parent node.
Optionally, the processing unit is configured to perform the following steps of analyzing the behavior data by using a first model, and determining an action to be performed by the user: the behavior data of the user is preprocessed to obtain a first preprocessing result; and transmitting the first preprocessing result to the hidden layer through the input layer, wherein the first preprocessing result is coded in each neuron of the hidden layer to obtain the action to be executed, and the action to be executed is output through the output layer.
Optionally, the control unit is configured to execute at least one of the following steps to control the door lock to execute a corresponding response operation according to the action to be executed: the method comprises the steps of starting the biological characteristic identification module, starting the password input module, closing the biological characteristic identification module and closing the password input module.
Optionally, the apparatus further comprises: the receiving unit is used for receiving the input scores corresponding to the response operations in the learning mode; the output unit is used for outputting a training result of the first model according to the score, wherein the training result is used for indicating the matching degree between the action to be executed and the actual action executed by the user; marking corresponding behavior data according to the training result; and inputting the marked behavior data into the training sample.
The above-mentioned serial numbers of the embodiments of the present invention are merely for description and do not represent the merits of the embodiments.
In the above embodiments of the present invention, the descriptions of the respective embodiments have respective emphasis, and for parts that are not described in detail in a certain embodiment, reference may be made to related descriptions of other embodiments.
In the embodiments provided in the present application, it should be understood that the disclosed technology can be implemented in other ways. The above-described embodiments of the apparatus are merely illustrative, and for example, the division of the units may be a logical division, and in actual implementation, there may be another division, for example, multiple units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, units or modules, and may be in an electrical or other form.
The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.
In addition, functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, and can also be realized in a form of a software functional unit.
The integrated unit, if implemented in the form of a software functional unit and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a Read-only Memory (ROM), a Random Access Memory (RAM), a removable hard disk, a magnetic or optical disk, and other various media capable of storing program codes.
The foregoing is only a preferred embodiment of the present invention, and it should be noted that, for those skilled in the art, various modifications and decorations can be made without departing from the principle of the present invention, and these modifications and decorations should also be regarded as the protection scope of the present invention.
Claims (14)
1. A door lock control method, comprising:
acquiring behavior data of a user;
analyzing the behavior data by using a first model, and determining the action to be executed of the user, wherein the first model is trained by machine learning by using a plurality of groups of data, and each group of data in the plurality of groups of data comprises the behavior data and the action to be executed obtained from the behavior data;
identifying an unlocking mode which a user wants to use according to the action to be executed so as to control the door lock to execute corresponding response operation;
after controlling the door lock to execute corresponding response operation according to the action to be executed, the method further comprises the following steps: in a learning mode, receiving input scores corresponding to the response operations; outputting a training result of the first model according to the score, wherein the training result is used for indicating the matching degree between the action to be executed and the actual action executed by the user; marking corresponding behavior data according to the training result; inputting the marked behavior data into a training sample;
the identifying the unlocking mode which the user wants to use according to the action to be executed so as to control the door lock to execute the corresponding response operation comprises at least one of the following operations: the method comprises the steps of starting the biological characteristic identification module, starting the password input module, closing the biological characteristic identification module and closing the password input module.
2. The method of claim 1, wherein the method of building the first model comprises:
acquiring behavior data and actions to be executed in a training sample;
and modeling the behavior data and the action to be executed in the training sample through a back propagation algorithm to obtain the first model.
3. The method of claim 2, wherein the first model comprises an input layer, a hidden layer, and an output layer; wherein, the modeling the behavior data and the action to be executed in the training sample by the back propagation algorithm to obtain the first model comprises:
preprocessing the behavior data in the training sample to obtain a sample preprocessing result, wherein the sample preprocessing result comprises at least one of the following: inputting a vector, an input parameter matrix and a vector diagram;
transmitting the sample preprocessing result to the hidden layer through the input layer, wherein the sample preprocessing result is coded in each neuron of the hidden layer to obtain a sample action, and the sample action is output through the output layer;
comparing the sample action with the action to be performed in the training sample;
if the error between the sample action and the action to be executed in the training sample does not meet a preset condition, calculating the square sum of the error, and taking the square sum as a target function;
calculating a partial derivative of the target function to the weight of each neuron to obtain a gradient amount corresponding to the weight;
and updating the weight by adopting a gradient descent algorithm according to the gradient amount.
4. The method of claim 3, wherein the hidden layer comprises a tree graph structure including a child node and a parent node, wherein each neuron in the child node is connected to each neuron in the parent node.
5. The method of claim 3, wherein the analyzing the behavioral data using the first model, and wherein determining the action to be performed by the user comprises:
the behavior data of the user is preprocessed to obtain a first preprocessing result;
and transmitting the first preprocessing result to the hidden layer through the input layer, wherein the first preprocessing result is coded in each neuron of the hidden layer to obtain the action to be executed, and the action to be executed is output through the output layer.
6. The method of claim 1, wherein the biometric characteristic comprises at least one of: fingerprint features, palm print features, voice print features, facial features, limb movements, iris features.
7. The method of claim 1, wherein the obtaining behavior data of the user comprises:
collecting the behavioral data of the user by a sensor, the behavioral data including at least one of: limb movements, facial expressions, voice, eye movements.
8. The method of claim 7, wherein the sensor comprises at least one of: camera, distance sensor, iris recognition device.
9. The method according to any one of claims 1 to 8, wherein the first model is a Recurrent Neural Network (RNN).
10. A door lock control apparatus, characterized by comprising:
the acquiring unit is used for acquiring behavior data of a user;
the processing unit is used for analyzing the behavior data by using a first model and determining the action to be executed of the user, wherein the first model is trained by machine learning by using a plurality of groups of data, and each group of data in the plurality of groups of data comprises the behavior data and the action to be executed obtained from the behavior data;
the control unit is used for identifying an unlocking mode which a user wants to use according to the action to be executed so as to control the door lock to execute corresponding response operation;
the device further comprises: the receiving unit is used for receiving the input scores corresponding to the response operations in the learning mode; the output unit is used for outputting a training result of the first model according to the score, wherein the training result is used for indicating the matching degree between the action to be executed and the actual action executed by the user; marking corresponding behavior data according to the training result; inputting the marked behavior data into a training sample;
the control unit is used for executing at least one of the following steps to identify an unlocking mode which a user wants to use according to the action to be executed so as to control the door lock to execute corresponding response operation: the method comprises the steps of starting the biological characteristic identification module, starting the password input module, closing the biological characteristic identification module and closing the password input module.
11. The apparatus of claim 10,
the acquisition unit is also used for acquiring behavior data and actions to be executed in the training samples;
the processing unit is further configured to model the behavior data and the action to be executed in the training sample through a back propagation algorithm to obtain the first model.
12. The apparatus of claim 11, wherein the first model comprises an input layer, a hidden layer, and an output layer; the processing unit is used for performing the following steps to model the behavior data and the action to be executed in the training sample through a back propagation algorithm to obtain the first model:
preprocessing the behavior data in the training sample to obtain a sample preprocessing result, wherein the sample preprocessing result comprises at least one of the following: inputting a vector, an input parameter matrix and a vector diagram;
transmitting the sample preprocessing result to the hidden layer through the input layer, wherein the sample preprocessing result is coded in each neuron of the hidden layer to obtain a sample action, and the sample action is output through the output layer;
comparing the sample action with the action to be performed in the training sample;
if the error between the sample action and the action to be executed in the training sample does not meet a preset condition, calculating the square sum of the error, and taking the square sum as a target function;
calculating a partial derivative of the target function to the weight of each neuron to obtain a gradient amount corresponding to the weight;
and updating the weight by adopting a gradient descent algorithm according to the gradient amount.
13. The apparatus of claim 12, wherein the hidden layer comprises a tree graph structure including a child node and a parent node, wherein each neuron in the child node is connected to each neuron in the parent node.
14. The apparatus of claim 12, wherein the processing unit is configured to perform the following steps of analyzing the behavior data using a first model to determine the action to be performed by the user:
the behavior data of the user is preprocessed to obtain a first preprocessing result;
and transmitting the first preprocessing result to the hidden layer through the input layer, wherein the first preprocessing result is coded in each neuron of the hidden layer to obtain the action to be executed, and the action to be executed is output through the output layer.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810916855.7A CN110874875B (en) | 2018-08-13 | 2018-08-13 | Door lock control method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810916855.7A CN110874875B (en) | 2018-08-13 | 2018-08-13 | Door lock control method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110874875A CN110874875A (en) | 2020-03-10 |
| CN110874875B true CN110874875B (en) | 2021-01-29 |
Family
ID=69714153
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810916855.7A Active CN110874875B (en) | 2018-08-13 | 2018-08-13 | Door lock control method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110874875B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US11157734B1 (en) * | 2020-12-01 | 2021-10-26 | Motorola Solutions, Inc. | Autonomous self-learning artificial intelligence intent system for access control |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| GB2472295B (en) * | 2009-07-27 | 2011-11-09 | Chin-Lun Lai | Security system with power saving feature and method thereof |
| CN103425239A (en) * | 2012-05-21 | 2013-12-04 | 刘鸿达 | Control system with facial expressions as input |
| CN107564144A (en) * | 2017-08-20 | 2018-01-09 | 聚鑫智能科技（武汉）股份有限公司 | A kind of intelligent robot gate control system and control method |
| CN107682536A (en) * | 2017-09-25 | 2018-02-09 | 努比亚技术有限公司 | A kind of sound control method, terminal and computer-readable recording medium |
| CN107726540A (en) * | 2017-09-27 | 2018-02-23 | 珠海格力电器股份有限公司 | Air conditioner control method and device |
| CN108140386A (en) * | 2016-07-15 | 2018-06-08 | 谷歌有限责任公司 | Speaker verification |
| CN108363999A (en) * | 2018-03-22 | 2018-08-03 | 百度在线网络技术（北京）有限公司 | Operation based on recognition of face executes method and apparatus |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR101603572B1 (en) * | 2015-07-20 | 2016-03-15 | 최환희 | Security controller with reading function of security card |
| CN206003184U (en) * | 2016-08-11 | 2017-03-08 | 蒋宇涵 | Door control identity recognition system |
| CN207663490U (en) * | 2017-11-27 | 2018-07-27 | 浙江理工大学 | A kind of mixing recognition access control system management system based on neural calculation rod |
-
2018
- 2018-08-13 CN CN201810916855.7A patent/CN110874875B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| GB2472295B (en) * | 2009-07-27 | 2011-11-09 | Chin-Lun Lai | Security system with power saving feature and method thereof |
| CN103425239A (en) * | 2012-05-21 | 2013-12-04 | 刘鸿达 | Control system with facial expressions as input |
| CN108140386A (en) * | 2016-07-15 | 2018-06-08 | 谷歌有限责任公司 | Speaker verification |
| CN107564144A (en) * | 2017-08-20 | 2018-01-09 | 聚鑫智能科技（武汉）股份有限公司 | A kind of intelligent robot gate control system and control method |
| CN107682536A (en) * | 2017-09-25 | 2018-02-09 | 努比亚技术有限公司 | A kind of sound control method, terminal and computer-readable recording medium |
| CN107726540A (en) * | 2017-09-27 | 2018-02-23 | 珠海格力电器股份有限公司 | Air conditioner control method and device |
| CN108363999A (en) * | 2018-03-22 | 2018-08-03 | 百度在线网络技术（北京）有限公司 | Operation based on recognition of face executes method and apparatus |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110874875A (en) | 2020-03-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111652066B (en) | Medical behavior identification method based on multi-self-attention mechanism deep learning | |
| CN115345278A (en) | Structural learning of convolutional neural networks | |
| Coyle et al. | Faster self-organizing fuzzy neural network training and a hyperparameter analysis for a brain–computer interface | |
| CN108388876A (en) | A kind of image-recognizing method, device and relevant device | |
| CN110333783B (en) | Irrelevant gesture processing method and system for robust electromyography control | |
| CN109101876A (en) | Human bodys' response method based on long memory network in short-term | |
| CN109993102A (en) | Similar face retrieval method, apparatus and storage medium | |
| CN109117817B (en) | Face recognition method and device | |
| CN110516537B (en) | Face age estimation method based on self-learning | |
| Alvarez-Alvarez et al. | Body posture recognition by means of a genetic fuzzy finite state machine | |
| CN112069916B (en) | Face beauty prediction method, device and system and readable storage medium | |
| CN106909938A (en) | Viewing angle independence Activity recognition method based on deep learning network | |
| CN108960171B (en) | Method for converting gesture recognition into identity recognition based on feature transfer learning | |
| CN109726662A (en) | Multi-class human posture recognition method based on convolution sum circulation combination neural net | |
| CN111543988B (en) | Adaptive cognitive activity recognition method and device and storage medium | |
| KR20200080418A (en) | Terminla and operating method thereof | |
| CN112329974A (en) | LSTM-RNN-based civil aviation security event behavior subject identification and prediction method and system | |
| CN110874875B (en) | Door lock control method and device | |
| Savio et al. | Image processing for face recognition using HAAR, HOG, and SVM algorithms | |
| Ntalampiras et al. | An incremental learning mechanism for human activity recognition | |
| Gill et al. | Human action detection using EfficientNetB3 model | |
| CN111160161A (en) | Self-learning face age estimation method based on noise elimination | |
| CN113643283A (en) | Method, device, equipment and storage medium for detecting aging condition of human body | |
| Cao et al. | Advancing the incremental fusion of robotic sensory features using online multi-kernel extreme learning machine | |
| Li et al. | One-shot learning gesture recognition based on evolution of discrimination with successive memory |
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 |