CN104865929A - Humanoid-behavior-based intelligent household power supply control system - Google Patents

Abstract

The invention, which particularly relates to the humanoid intelligent household informatization field, provides a humanoid-behavior-based intelligent household power supply control system. A household gateway provides a function of humanoid intelligent household information management. A humanoid behavior algorithm module is characterized in that a hidden-conditional-random-field-based human behavior classification algorithm is embedded into the household gateway, thereby realizing the intelligent operation of the household power supply. A wireless communication module employs wifi, Bluetooth, zigbee technologies to realize communication with all modules. A relay plug board module uses a relay to control on/off of a household power supply according to operation commands of the user and the humanoid intelligent algorithm, thereby controlling the household equipment. An Android mobile phone terminal uses a third-party application program (App) intelligent plug board of a smart phone to realize communication with the household gateway, thereby controlling the household power supply. With the system, real-time and intelligent control of the household equipment power supply can be realized; the energy-saving efficiency is improved; the cost and power consumption are low; the operation is convenient; the applicability is high; and the system has the great popularization and application values.

Description

A kind of Smart Home power control system based on apery behavior

Technical field

The invention provides a kind of Smart Home power control system based on apery behavior, be specifically related to Android phone application program mobile Internet APP development technique, radio communication ZigBee technology, Wifi technology and BlueTooth technology, relay patch panel design, apery behavior prediction research etc., belong to intelligent home information field.

Background technology

Along with the high speed development of society, people's living standard improves constantly, work rhythm is also more and more faster, time and efforts for aspects such as family's nurses is fewer and feweri, constantly perfect along with Smart Home market, the continuous progress of science and technology, realizes household's IT application, networking is the Main way that following household develops, also just for the Smart Home power control system based on apery behavior provides wide market outlook.

Smart Home concept is just suggested as far back as eighties of last century late nineteen eighties, but its development be subject to expensive, user dispose installation cost high, not easily expand, the factor restriction such as complex operation interactivity difference, it is made to be difficult to step into ordinary people family, and the appearance of smart mobile phone and mobile Internet, cultivate the brand-new interactive mode of user and use habit, reduce user to the fear of difficulty of similar electronic product and learning cost.Smart mobile phone is as a powerful terminal processes platform, and mobile Internet, as the stable communication network of a high speed, promotes the hardware platform construction of Smart Home development.The Smart Home combined with smart mobile phone has been not only independent single product, but a solution, it is effectively combined the various lifes staying idle at home such as control, information and security protection, solves and monitor the various communications, household electrical appliances, security facilities etc. in house, control and manage etc.Allow the usage behavior of intelligent domestic system analysis user, thus make it possess the ability of autonomous learning user operation behavior gradually, and then can on one's own initiative for user provides the service wishing to obtain.

Human behavior has certain rule, and namely human individual's behavior usually can fall within a kind of scope, and in this range, some behaviors are common, and some behaviors think unusual, utilizes computer technology very ripe to study human behavior at present.

The multiple power supplies distribution member of property based on plate, be connected with the attaching plug of various household electrical appliance, can realize controlling by controlling socket the object of household electrical appliance, therefore this project proposes the Smart Home power control system based on apery behavior, user can control intelligent board by the Android application program developed specially, have and use the features such as simple, deployment facilitates, can move everywhere, the life that energy is convenient daily greatly, promotes the popularization of Smart Home and popularizes.

Summary of the invention

In view of this, the object of this invention is to provide a kind of Smart Home power control system based on apery behavior, realize controlling the real-time intelligentization of home equipment power supply, raising energy-saving efficiency, and have that cost is low, power consumption is little, the feature such as simple operation, adaptation are strong.

The present invention adopts following scheme to realize: a kind of Smart Home power control system based on apery behavior, it is characterized in that: realize, with the communication of home gateway, realizing the real-time control of user to Domestic electric appliances power source with this by Internet internet with " intelligent board " in Android phone; The usage behavior of apery behavior algorithm realization to user is analyzed, and possesses independent learning ability gradually, finally reaches the autonomous type service of household power supply; Wireless communication module comprehensively adopts the technological means such as BlueTooth, Wifi and ZigBee to be transferred in relay plate the operation result of apery behavior algorithm, and the status information of plate appliances power source break-make also will transfer in home gateway by wireless communication module by relay plate simultaneously; Relay plate module, according to the information received, adopts relay element to control the break-make of household electric appliances power supply, records the power supply status of household electric appliances simultaneously.

Further, a kind of based in the Smart Home power control system implementation of apery behavior, the present invention is made up of five major parts such as Android mobile phone terminal module, home gateway module, apery behavior algoritic module, wireless communication module and relay plate modules;

Further, a kind of based on the Android mobile phone terminal module in the Smart Home power control system ingredient of apery behavior, its core is application program " intelligent board " APP, is the main mutual approach that Android phone communicates with home gateway, adopts unified communication protocol;

Wherein, application program " intelligent board " App has and opens, closes Domestic electric appliances power source, time set, receives home equipment power information, and shows with the form of Visual Chart;

Further, a kind of based on the home gateway module in the Smart Home power control system ingredient of apery behavior, be divided into outer net and Intranet two parts, outer net adopts Ethernet to communicate with Android mobile phone terminal, Intranet adopts communication to communicate with relay plate, adopts unified communication protocol;

Further, a kind of based on the apery behavior algoritic module in the Smart Home power control system ingredient of apery behavior, the human behavior sorting algorithm based on hidden conditional random fields is adopted to carry out.

Wherein, the human behavior sorting algorithm based on hidden conditional random fields adopts the model foundation of hidden conditional random fields, model training, the large step of behaviour classification three to carry out.Model based on hidden conditional random fields is set up and following step can be adopted to carry out:

Step one: set up the condition random field.Article one, length is the action trail x=(x of T ₁, x ₂..., x _t), its corresponding class label is Y=(y ₁, y ₂..., y _t); Introduce a non-directed graph model (UGM), suppose that non-directed graph G=(node V, limit E) represents that UGM, a Y represent point wherein; If condition is X, Y meet Markov, then (Y, X) is a condition random field (RCF);

Step 2: set up hidden conditional random fields.Hidden variable h=(h is introduced in condition random field ₁, h ₂..., h _t) obtaining hidden conditional random fields, hidden variable h corresponds to the point in G, and works as h _i, h _jbetween exist link time, (i, j) ∈ E is a limit in G.

Step 3: the conditional distribution function P (y|x asking for action trail x; θ), parameter to be learned is θ={ θ _a, θ _b}: normalized factor Z (x; θ) be:

$Z(x;\mathrm{\θ})=\underset{y;h}{\mathrm{\Σ}}(\exp \left(\underset{t=1}{\overset{T}{\mathrm{\Σ}}}F(y^{\′},h_{t-1},h_t,x;\mathrm{\θ})\right)$ Formula (1)

By fundamental function F (y, h _t-1, h _t, x; θ) be defined as:

$F(y,h_{t-1},h_t,x;\mathrm{\θ})=\underset{a\∈A}{\mathrm{\Σ}}{\mathrm{\θ}}_a{f}_a(y,h_{t-1},h_t,x)+\underset{b\∈B}{\mathrm{\Σ}}{\mathrm{\θ}}_b{f}_b(y,h_t,x)$ Formula (2)

Wherein, A is the set of the feature on limit, and B is the characteristic set of point, f _atransfer characteristic function, f _bstatus flag function, (h _t-1, h _t) refer to two hidden variables adjacent in non-directed graph.Conditional distribution function P (the y|x of action trail x can be obtained thus; θ) be:

$P\left(y\right|x;\mathrm{\θ})=\underset{h}{\mathrm{\Σ}}P(y,h|x;\mathrm{\θ})=\frac{1}{Z(x;\mathrm{\θ})}\underset{h}{\mathrm{\Σ}}\left(\exp \left(\underset{t=1}{\overset{T}{\mathrm{\Σ}}}F(y,h_{t-1},h_t,x;\mathrm{\θ})\right)\right)$ Formula (3)

Model training based on hidden conditional random fields can carry out in the following ways: use T ^y=(x ₁, x ₂..., x _n) representing the training data of y kind behavior, parameter to be learned can be obtained by the log-likelihood optimizing training data:

$L\left(\mathrm{\θ}\right)=\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{L}_i\left(\mathrm{\θ}\right)=\underset{i=1}{\overset{N}{\mathrm{\Σ}}}\log P\left(y_i\right|x_i;\mathrm{\θ}),i=\mathrm{1,2},...,N$ Formula (4)

Being translated into one thus can adopt gradient ascent algorithm to ask for the optimization problem of parameter to be learned.

Behaviour classification based on hidden conditional random fields carrys out test trails x by maximizing conditional probability ^*behaviour classification y ^*:

$y^{*}=\arg \underset{y\∈Y}{\max }P\left(y\right|x^{*},{\mathrm{\θ}}^{*})$ Formula (5)

Wherein, θ ^*value be from model training study.

Further, a kind of based on the wireless communication module in the Smart Home power control system ingredient of apery behavior, the technological means such as BlueTooth, Wifi and ZigBee are adopted to realize the radio communication with home equipment;

Further, relay plate module in a kind of ingredient of the Smart Home power control system based on apery behavior adopts wireless communication technology to realize data interaction, adopt relay to realize the break-make of plate power supply according to interactive information, radio communication adopts unified communication protocol.

Prominent feature of the present invention is embedded in intelligent home device by apery behavior prediction algorithm, realizes the autonomous type break-make of home equipment power supply, improves energy-saving efficiency, meets the energy-saving and emission-reduction that country advocates; Adopt modular design, user can carry out personalization according to self-demand selection function peripheral hardware and install, cost-saving, has features such as using simple, deployment aspect, can move everywhere, can greatly convenient daily life, promotes the popularization of Smart Home and popularizes.

Accompanying drawing explanation

For making object of the present invention, technical scheme and outstanding effect plays ground more clear, the invention provides the following drawings and being described:

Fig. 1 is system principle diagram of the present invention;

Fig. 2 is workflow diagram of the present invention;

Fig. 3 is Android mobile phone terminal module core application program " intelligent board " App functional block diagram of the present invention;

Fig. 4 is the partial function schematic diagram of the home gateway module in the present invention;

Fig. 5 is apery behavior algorithm principle block diagram of the present invention;

Fig. 6 is relay plate theory of constitution figure of the present invention.

Embodiment

Describe a kind of Smart Home power control system based on apery behavior of the present invention in detail below in conjunction with accompanying drawing, Fig. 1 is system principle diagram.

As shown in Figure 1, a kind of Smart Home power control system based on apery behavior, embodiment is: realize the communication with home gateway with " intelligent board " App in Android phone by Internet internet; The usage behavior of apery behavior algorithm realization to user is analyzed, and possesses independent learning ability gradually, finally reaches the autonomous type service of household power supply; Wireless communication module comprehensively adopts the technological means such as BlueTooth, Wifi and ZigBee to be transferred in relay plate the operation result of apery behavior prediction algorithm, and the status information of plate appliances power source break-make also will transfer in home gateway by wireless communication module by relay plate simultaneously; Relay plate module, according to the information received, adopts relay element to control the break-make of household electric appliances power supply, records the power supply status of household electric appliances simultaneously.

Fig. 2 gives a kind of workflow diagram of the Smart Home power control system based on apery behavior.Below be described in more detail:

Further, a kind of based in the Smart Home power control system implementation of apery behavior, the present invention is made up of five major parts such as Android mobile phone terminal module, home gateway module, apery behavior prediction algoritic module, wireless communication module and relay plate modules;

Wherein, Android mobile phone terminal module and home gateway module carry out the mutual of user's operation information by Internet each other; User behavior is carried out being instantiated as apery behavior algorithm by home gateway module provides algorithm interface, simultaneously apery behavior algorithm by result of calculation and intermediate data storage in home gateway; Operation result is transferred to wireless communication module by apery behavior algoritic module; Wireless communication module has two parts function: be apery behavior algoritic module result resolved first, pilot relay plate module, is second by on/off information, comprises the information such as switching-on and switching-off state, feed back to home gateway; Relay plate module is then the information interaction of carrying out Intranet with wireless communication module, adopts the mode of relay to control the break-make of household power supply simultaneously, and voltage measurement.

Fig. 3 gives a kind of Android mobile phone terminal module core application program " intelligent board " App functional block diagram of the Smart Home power control system based on apery behavior.Below make explanation specifically:

Further; first " intelligent board " App is safety protection problem; enter login interface, only login successfully employing and enter the main interface of control center, then select the large functional module of corresponding Operation system setting, system information, local function, home wiring control, logout and terminal maintenance 6.It has opens, closes Domestic electric appliances power source, time set, receives home equipment power information, and shows with the form of Visual Chart; In addition, the historical operation information of user can also be understood, system operation instruction and help etc. conventional func.

Fig. 4 gives a kind of partial function schematic diagram of home gateway module of the Smart Home power control system based on apery behavior, below makes relevant explanation:

Further, a kind of based on the home gateway module in the Smart Home power control system ingredient of apery behavior, be divided into outer net and Intranet two parts, outer net adopts Ethernet mode to communicate with Android mobile phone terminal, Intranet adopts communication to communicate with relay plate, adopts unified communication protocol;

Wherein, the modes such as BlueTooth, Wifi, ZigBee are adopted to communicate in Intranet; The function such as security guard and man-machine interaction is also had in home gateway.

Fig. 5 gives a kind of principle schematic of apery behavior algoritic module of the Smart Home power control system based on apery behavior, and embodiment can be carried out in the following ways:

Further, a kind of based on the apery behavior algoritic module in the Smart Home power control system ingredient of apery behavior, the human behavior sorting algorithm based on hidden condition field is adopted to carry out.

Wherein, the human behavior sorting algorithm based on hidden conditional random fields adopts the model foundation of hidden conditional random fields, model training, the large step of behaviour classification three to carry out.Model based on hidden conditional random fields is set up and following step can be adopted to carry out:

Step one: set up the condition random field.Article one, length is the action trail x=(x of T ₁, x ₂..., x _t), its corresponding class label is Y=(y ₁, y ₂..., y _t); Introduce a non-directed graph model (UGM), suppose that non-directed graph G=(node V, limit E) represents that UGM, a Y represent point wherein; If condition is X, Y meet Markov, then (Y, X) is a condition random field (RCF);

Step 2: set up hidden conditional random fields.Hidden variable h=(h is introduced in condition random field ₁, h ₂..., h _t) obtaining hidden conditional random fields, hidden variable h corresponds to the point in G, and works as h _i, h _jbetween exist link time, (i, j) ∈ E is a limit in G.

Step 3: the conditional distribution function P (y|x asking for action trail x; θ), parameter to be learned is θ={ θ _a, θ _b}: normalized factor Z (x; θ) be:

$Z(x;\mathrm{\θ})=\underset{y;h}{\mathrm{\Σ}}(\exp \left(\underset{t=1}{\overset{T}{\mathrm{\Σ}}}F(y^{\′},h_{t-1},h_t,x;\mathrm{\θ})\right)$ Formula (1)

By fundamental function F (y, h _t-1, h _t, x; θ) be defined as:

$F(y,h_{t-1},h_t,x;\mathrm{\θ})=\underset{a\∈A}{\mathrm{\Σ}}{\mathrm{\θ}}_a{f}_a(y,h_{t-1},h_t,x)+\underset{b\∈B}{\mathrm{\Σ}}{\mathrm{\θ}}_b{f}_b(y,h_t,x)$ Formula (2)

Wherein, A is the set of the feature on limit, and B is the characteristic set of point, f _atransfer characteristic function, f _bstatus flag function, (h _t-1, h _t) refer to two hidden variables adjacent in non-directed graph.Conditional distribution function P (the y|x of action trail x can be obtained thus; θ) be:

$P\left(y\right|x;\mathrm{\θ})=\underset{h}{\mathrm{\Σ}}P(y,h|x;\mathrm{\θ})=\frac{1}{Z(x;\mathrm{\θ})}\underset{h}{\mathrm{\Σ}}\left(\exp \left(\underset{t=1}{\overset{T}{\mathrm{\Σ}}}F(y,h_{t-1},h_t,x;\mathrm{\θ})\right)\right)$ Formula (3)

Model training based on hidden conditional random fields can carry out in the following ways: use T ^y=(x ₁, x ₂..., x _n) representing the training data of y kind behavior, parameter to be learned can be obtained by the log-likelihood optimizing training data:

$L\left(\mathrm{\θ}\right)=\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{L}_i\left(\mathrm{\θ}\right)=\underset{i=1}{\overset{N}{\mathrm{\Σ}}}\log P\left(y_i\right|x_i;\mathrm{\θ}),i=\mathrm{1,2},...,N$ Formula (4)

Being translated into one thus can adopt gradient ascent algorithm to ask for the optimization problem of parameter to be learned.

Behaviour classification based on hidden conditional random fields carrys out test trails x by maximizing conditional probability ^*behaviour classification y ^*:

$y^{*}=\arg \underset{y\∈Y}{\max }P\left(y\right|x^{*},{\mathrm{\θ}}^{*})$ Formula (5)

Wherein, θ ^*value be from model training study.

Fig. 6 gives a kind of theory of constitution figure of relay plate of the Smart Home power control system based on apery behavior, and embodiment can be carried out in the following ways:

Further, relay plate module in a kind of ingredient of the Smart Home power control system based on apery behavior adopts wireless communication technology to realize data interaction, adopt the break-make of relay array realization to plate power supply according to interactive information, adopt voltage measurement array to carry out to the supervision of appliances power source.

Prominent feature of the present invention is embedded in intelligent home device by apery behavior algorithm, realizes the autonomous type break-make of home equipment power supply, improves energy-saving efficiency, meets the energy-saving and emission-reduction that country advocates; Adopt modular design, user can carry out personalization according to self-demand selection function peripheral hardware and install, cost-saving, has features such as using simple, deployment aspect, can move everywhere, can greatly convenient daily life, promotes the popularization of Smart Home and popularizes.

Finally it should be noted that, above embodiment is only in order to illustrate technical scheme of the present invention and unrestricted, although be described in further detail invention has been with reference to preferred embodiment, those of ordinary skill in the art is to be understood that, can modify to technical scheme of the present invention or equivalent replacement, and not departing from aim and the scope of technical solution of the present invention, it all should be encompassed in the middle of right of the present invention.

Claims (9)

1. the Smart Home power control system based on apery behavior, it is characterized in that: realize, with the communication of home gateway, realizing the Long-distance Control of user to Domestic electric appliances power source with this by Internet internet with " intelligent board " App in Android phone; The usage behavior of apery behavior algorithm realization to user is analyzed, and possesses independent learning ability gradually, finally reaches the autonomous type service of household power supply; Wireless communication module comprehensively adopts the technological means such as BlueTooth, Wifi and ZigBee to be transferred in relay plate the operation result of apery behavior algorithm, and the status information of plate appliances power source break-make also will transfer in home gateway by wireless communication module by relay plate simultaneously; Relay plate module, according to the information received, adopts relay element to control the break-make of household electric appliances power supply, records the power supply status of household electric appliances simultaneously.

2. a kind of Smart Home power control system based on apery behavior according to claim 1, is characterized in that: this system comprises Android mobile phone terminal module, home gateway module, apery behavior algoritic module, wireless communication module and relay plate module five major part.

3. a kind of Smart Home power control system based on apery behavior according to claim 1, it is characterized in that: Android mobile phone terminal module core is application program " intelligent board " APP, it is the main mutual approach that Android phone communicates with home gateway, adopts unified communication protocol.

4. a kind of Smart Home power control system based on apery behavior according to claim 1, it is characterized in that: home gateway module adopts embedded home gateway, be divided into ethernet module and the Intranet wireless communication module of outer net, outer net realizes the communication with Android mobile phone terminal, Intranet realizes and relay plate module communication, adopts unified communication protocol.

5. a kind of Smart Home power control system based on apery behavior according to claim 1, is characterized in that: apery behavior algoritic module adopts and carries out based on the human behavior sorting algorithm of hidden conditional random fields.

6. a kind of Smart Home power control system based on apery behavior according to claim 1, is characterized in that: wireless communication module comprehensively adopts the technological means such as BlueTooth, Wifi and ZigBee to realize the radio communication with home equipment.

7. a kind of Smart Home power control system based on apery behavior according to claim 1, it is characterized in that: relay plate module adopts wireless communication technology to realize data interaction, adopt the break-make of relay array realization to plate power supply according to interactive information, adopt voltage measurement array to carry out to the supervision of appliances power source.

8. a kind of Smart Home power control system based on apery behavior according to claim 3, it is characterized in that: " intelligent board " APP has and opens, closes Domestic electric appliances power source, time set, reception home equipment power information, and show with Visual Chart.

9. a kind of Smart Home power control system based on apery behavior according to claim 5, is characterized in that: the human behavior sorting algorithm based on hidden conditional random fields adopts the model foundation of hidden conditional random fields, model training, the large step of behaviour classification three to carry out.

Model based on hidden conditional random fields is set up and following step can be adopted to carry out:

Step one: set up the condition random field.Article one, length is the action trail x=(x of T ₁, x ₂..., x _t), its corresponding class label is Y=(y ₁, y ₂..., y _t); Introduce a non-directed graph model (UGM), suppose that non-directed graph G=(node V, limit E) represents that UGM, a Y represent point wherein; If condition is X, Y meet Markov, then (Y, X) is a condition random field (RCF);

Step 2: set up hidden conditional random fields.Hidden variable h=(h is introduced in condition random field ₁, h ₂..., h _t) obtaining hidden conditional random fields, hidden variable h corresponds to the point in G, and works as h _i, h _jbetween exist link time, (i, j) ∈ E is a limit in G.

Step 3: the conditional distribution function P (y|x asking for action trail x; θ), parameter to be learned is θ={ θ _a, θ _b}:

Normalized factor Z (x; θ) be:

$Z(x;\mathrm{\θ})=\underset{y;h}{\mathrm{\Σ}}(\exp \left(\underset{t=1}{\overset{T}{\mathrm{\Σ}}}F(y^{\′},h_{t-1},h_t,x;\mathrm{\θ})\right)$ Formula (1)

By fundamental function F (y, h _t-1, h _t, x; θ) be defined as:

$F(y,h_{t-1},h_t,x;\mathrm{\θ})=\underset{a\∈A}{\mathrm{\Σ}}{\mathrm{\θ}}_a{f}_a(y,h_{t-1},h_t,x)+\underset{b\∈B}{\mathrm{\Σ}}{\mathrm{\θ}}_b{f}_b(y,h_t,x)$ Formula (2)

Wherein, A is the set of the feature on limit, and B is the characteristic set of point, f _atransfer characteristic function, f _bstatus flag function, (h _t-1, h _t) refer to two hidden variables adjacent in non-directed graph.Conditional distribution function P (the y|x of action trail x can be obtained thus; θ) be:

$P\left(y\right|x;\mathrm{\θ})=\underset{h}{\mathrm{\Σ}}P(y,h|x;\mathrm{\θ})=\frac{1}{Z(x;\mathrm{\θ})}\underset{h}{\mathrm{\Σ}}\left(\exp \left(\underset{t=1}{\overset{T}{\mathrm{\Σ}}}F(y,h_{t-1},h_t,x;\mathrm{\θ})\right)\right)$ Formula (3)

Model training based on hidden conditional random fields can carry out in the following ways: use T ^y=(x ₁, x ₂..., x _n) representing the training data of y kind behavior, parameter to be learned can be obtained by the log-likelihood optimizing training data:

$L\left(\mathrm{\θ}\right)=\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{L}_i\left(\mathrm{\θ}\right)=\underset{i=1}{\overset{N}{\mathrm{\Σ}}}\log P\left(y_i\right|x_i;\mathrm{\θ}),i=\mathrm{1,2},...,N$ Formula (4)

Being translated into one thus can adopt gradient ascent algorithm to ask for the optimization problem of parameter to be learned.

Behaviour classification based on hidden conditional random fields carrys out test trails x by maximizing conditional probability ^*behaviour classification y ^*:

$y^{*}=\arg \underset{y\∈Y}{\max }P\left(y\right|x^{*},{\mathrm{\θ}}^{*})$ Formula (5)

Wherein, θ ^*value be from model training study.