CN108964270A - A kind of intelligent appliance load detecting and control system and its method - Google Patents
A kind of intelligent appliance load detecting and control system and its method Download PDFInfo
- Publication number
- CN108964270A CN108964270A CN201810725906.8A CN201810725906A CN108964270A CN 108964270 A CN108964270 A CN 108964270A CN 201810725906 A CN201810725906 A CN 201810725906A CN 108964270 A CN108964270 A CN 108964270A
- Authority
- CN
- China
- Prior art keywords
- module
- load detecting
- control
- current
- intelligent
- 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.)
- Pending
Links
Classifications
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J13/00—Circuit arrangements for providing remote indication of network conditions, e.g. an instantaneous record of the open or closed condition of each circuitbreaker in the network; Circuit arrangements for providing remote control of switching means in a power distribution network, e.g. switching in and out of current consumers by using a pulse code signal carried by the network
- H02J13/00006—Circuit arrangements for providing remote indication of network conditions, e.g. an instantaneous record of the open or closed condition of each circuitbreaker in the network; Circuit arrangements for providing remote control of switching means in a power distribution network, e.g. switching in and out of current consumers by using a pulse code signal carried by the network characterised by information or instructions transport means between the monitoring, controlling or managing units and monitored, controlled or operated power network element or electrical equipment
- H02J13/00019—Circuit arrangements for providing remote indication of network conditions, e.g. an instantaneous record of the open or closed condition of each circuitbreaker in the network; Circuit arrangements for providing remote control of switching means in a power distribution network, e.g. switching in and out of current consumers by using a pulse code signal carried by the network characterised by information or instructions transport means between the monitoring, controlling or managing units and monitored, controlled or operated power network element or electrical equipment using optical means
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02B—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO BUILDINGS, e.g. HOUSING, HOUSE APPLIANCES OR RELATED END-USER APPLICATIONS
- Y02B70/00—Technologies for an efficient end-user side electric power management and consumption
- Y02B70/30—Systems integrating technologies related to power network operation and communication or information technologies for improving the carbon footprint of the management of residential or tertiary loads, i.e. smart grids as climate change mitigation technology in the buildings sector, including also the last stages of power distribution and the control, monitoring or operating management systems at local level
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02B—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO BUILDINGS, e.g. HOUSING, HOUSE APPLIANCES OR RELATED END-USER APPLICATIONS
- Y02B90/00—Enabling technologies or technologies with a potential or indirect contribution to GHG emissions mitigation
- Y02B90/20—Smart grids as enabling technology in buildings sector
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S20/00—Management or operation of end-user stationary applications or the last stages of power distribution; Controlling, monitoring or operating thereof
- Y04S20/20—End-user application control systems
- Y04S20/242—Home appliances
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S40/00—Systems for electrical power generation, transmission, distribution or end-user application management characterised by the use of communication or information technologies, or communication or information technology specific aspects supporting them
- Y04S40/12—Systems for electrical power generation, transmission, distribution or end-user application management characterised by the use of communication or information technologies, or communication or information technology specific aspects supporting them characterised by data transport means between the monitoring, controlling or managing units and monitored, controlled or operated electrical equipment
- Y04S40/126—Systems for electrical power generation, transmission, distribution or end-user application management characterised by the use of communication or information technologies, or communication or information technology specific aspects supporting them characterised by data transport means between the monitoring, controlling or managing units and monitored, controlled or operated electrical equipment using wireless data transmission
Landscapes
- Engineering & Computer Science (AREA)
- Power Engineering (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Selective Calling Equipment (AREA)
Abstract
The invention discloses a kind of intelligent appliance load detectings and control system and its method, load detecting and control system include load detecting and control device, server, the end APP or the end Web, the end APP or the end Web are communicated by wireless network and server with load detecting and control device, realize that the end APP or the end Web send instruction and reach load detecting and control device, realization passes through the infrared control to household electrical appliances；The load detecting and control device accurately identify device for switching type by convolutional neural networks algorithm and calculate the power consumption of electric appliance in the period in time, prevent the situation of electrical equipment control mistake, reach the function of accurate electrical equipment control, and possesses higher accuracy rate.
Description
Technical field
The invention belongs to intelligent power grid technology field, in particular to a kind of intelligent appliance load detecting and control system and its
Method.
Background technique
It is always carved in the modern life from electricity must be opened, electric energy is even more a lifeline to develop national economy.In order to constantly full
The sufficient mankind demand growing to electric energy, and more rationally, effectively utilize electric energy, the understanding to household electricity situation
Also just become particularly important.
Current home user smart grid is formed not yet, and ammeter also only can calculate total electricity in user's period,
The power consumption of some accurate electric appliance can not be provided.For the power consumption for identifying some electrical appliance, first have to carry out essence to power load
Really identification.
Currently, remained capacity algorithm it is main there is several methods that, including based on steady-state characteristic remained capacity algorithm and based on temporary
Step response remained capacity algorithm etc..Referred to based on steady-state characteristic remained capacity algorithm and is acquired under stable state after electric switch
The features such as voltage, electric current, power, energy identify.Traditional load identification based on transient characterisitics carries out after an event occurs
Monitoring, the transient characterisitics such as calculating current, power, transient admittance identify load.It is either special based on steady-state characteristic or transient state
Property, electric current, power, harmonic wave, current envelops, instantaneous admittance etc. when Main Analysis electric equipment operation.Although accuracy of identification
Increase, but it is many kinds of to be still limited to electric appliance, working environment is complicated, and there are generalization abilities not enough and cannot be complete
The problem of accurately identifying.
Meanwhile with the continuous development of technology of Internet of things, people begin trying to operate control by tele-control system
Electric appliance in family, the method for home wiring control includes appliance control system based on blueteeth network and based on Internet's at present
Intelligent control system of domestic electric appliances.Appliance control system based on blueteeth network refers to the spy using Bluetooth technology towards mobile environment
Property, to realize the addition and deletion and control of household appliance.Intelligent control system of domestic electric appliances based on Internet, which refers to, to be devised
The Household information processing run in one Windows system and electrical appliances intelligent control platform, realize household appliance on the platform
Intelligent control, Internet remotely control, the functions such as information management.
But the household appliances such as air-conditioning, television set, DVD, set-top box traditional in family all use IR remote controller control
System, and the coding/decoding mode that the infrared coding/decoding chip of each company production uses and pulse width period difference, result in market
On all kinds of remote controlers function it is incompatible, this brings inconvenience to the use of people.Moreover, existing control system
Remote controlled electrical household appliances in same room, and operational issue as the necessary reception device against electric appliance can not be solved, therefore is caused general
And degree is lower；Once user forgets to close some heating electric appliances, such as electric heater, immersion heater, the danger for generating fire is had
Danger.
Summary of the invention
In view of the deficienciess of the prior art, the present invention provides a kind of intelligent appliance load detecting and control system and its side
Method accurately identifies device for switching type based on convolutional neural networks and calculates the power consumption of electric appliance in the period in time, passes through
Infrared transmitter realizes the closing and function switch of electric appliance, prevents the situation of electrical equipment control mistake, reaches accurate electric appliance control
The function of system, and possess higher accuracy rate.
In order to solve the above-mentioned technical problem, the technical solution adopted by the present invention is that: a kind of intelligent appliance load detecting and control
System processed, including load detecting and control device, server, the end APP or the end Web, the load detecting and control device packet
Include data acquisition module, data preprocessing module, analog-to-digital conversion module, consumption detection module, memory module, power module, control
Molding block, enabling dog module, WIFI communication module and intelligent processor module, the data acquisition module is in household electricity
Electric signal acquire in real time, and after data preprocessing module carries out noise reduction, filtering and signal enhanced processing, signal flow point
For two-way, enter analog-to-digital conversion module all the way, enter consumption detection module all the way, measures electricity for intelligent processor module；
The intelligent processor module is dual core processor, and core one realizes Intelligent treatment, identifies electricity by convolutional neural networks algorithm
Device, core two realize home wiring control, by parsing WIFI data and protocol data, realize the infrared timed power on/off function to household electrical appliances
It can switching.
Further, the end APP or the end Web are connect by wireless network with server communication, so with load detecting and control
Device communication processed realizes that the end APP or the end Web send the intelligent processor module that instruction reaches load detecting and control device, institute
It states intelligent processor module and parsing and infrared coding is carried out to received instruction, and infrared coding is transmitted to control module, institute
It states control module transmitting infrared signal and realizes the control to household electrical appliances after the infrared receiver end of household electrical appliances receives.
Further, the data acquisition module includes that the current sensor being installed in household electricity bus and voltage pass
Sensor, the electric signal of the intelligent processor module real-time reception current sensor and voltage sensor acquisition, and generate electric current
Waveform diagram, and carry out convolutional neural networks algorithm output electric appliance classification results, is sent to server, user by the end APP or
The end Web obtains electric appliance classification results and electricity consumption situation.
Further, multiplexing electric abnormality information is also passed through the APP of server transport to user by the intelligent processor module
End or the end Web.
The present invention also provides a kind of intelligent appliance load detecting and control methods, comprising the following steps:
Step 1, collection voltages, current signal, and it is transmitted to dual core processor；
Step 2 identifies load switch；
Step 3 generates waveform image；
Step 4, remained capacity: being handled waveform image by convolutional neural networks algorithm, is accurately identified load
Type simultaneously calculates power consumption, and result is then sent to server；
Step 5, user check each household electrical appliances electricity consumption situation by the end APP or the end Web；
Step 6, electrical equipment control: the end APP or the end Web send instruction and reach dual core processor, solve to received instruction
Analysis and infrared coding emit infrared signal and realize after the infrared receiver termination of household electrical appliances receives infrared signal to remotely to household electrical appliances
Control；
Step 7, abnormality processing: when monitoring multiplexing electric abnormality information, dual core processor sends prompt information extremely by network
User, user can be closed or be continued to run by the end APP or the selection of the end Web.
Preferably, in step 1, voltage, the current signal loaded using voltage sensor and current sensor acquisition,
Then noise reduction, filtering and signal amplification are carried out, and carries out analog-to-digital conversion, 24 high accuracy numbers are provided for dual core processor
Signal；The calculating with electrical power consumed is carried out by consumption detection module simultaneously.
Further, in step 2, in order to reduce the noise jamming to electric current, in calculating current intensity, the period is subtracted
The average value of the current cycle of interior each sampled point, then takes absolute value；Then it is followed average value as the electric current of current strength
Ring, such as formula (1)；After the current strength for calculating each period, keep current strength different, that is to say, that subtract adjacent week
Interim current strength；When difference is greater than preset threshold, it is judged as ON；On the contrary, sentencing if result is less than preset threshold
Break to close, such as formula (3)；
ΔIi=Ii+1-Ii(i=1,2.....n-1) (3)
Wherein, IijFor the current strength of each sampled point, Ii-meanFor current strength average value in the sampling period, △ IiFor
The current strength difference of adjacent periods.
Preferably, in step 3, dual core processor is integrated into current waveform figure after receiving electric signal, then carries out ash
Conversion process is spent, the image of 64*64 size is made；Each pixel is finally sought into mean value, directly as convolutional Neural net
The input picture of network.
Further, in step 4, convolutional neural networks algorithm the following steps are included:
(1) CONV1: the size of convolution input picture is 32 convolution kernels of 3 × 3 pixels, followed by ReLU operator, so
It is the max pooling and local acknowledgement normalization layer LRN in the region 2*2 afterwards；
(2) CONV2: then, handling the Feature Mapping previously exported by the second convolutional layer, wherein including 64 sizes
For the filter of 3 × 3 pixels；Followed by ReLU, be maximized for 2 × 2 regions maximum pond layer, and have with as before
Hyper parameter local acknowledgement normalize layer；
(3) CONV3: the last layer is identical as the second layer；The output on upper layer is the filter that 64 sizes are 3 × 3 pixels
Convolution, followed by pooling layers of ReLU and max；
(4) full articulamentum FC1: the output of third layer convolutional layer is converted into one-dimensional array, and is connected entirely as first layer
The input of layer；It includes 4096 neurons, followed by one Relu and one dropout layers；
(5) full articulamentum FC2: receiving 4096 neurons of full articulamentum FC1, and again include 4096 neurons,
Followed by Relu and one dropout layers；
(6) full articulamentum FC3: this layer is output layer, according to actual classification collection class, there is 12 minds in this layer
Through member；Output category result.
Further, in step 7, when user forgets powered-down device, leads to electric appliance long-play, dual core processor root
Load switch, type and power consumption are identified according to step 1 to step 4, and transmits the result to server, are passed through simultaneously
WIFI network sends prompt information to user, and user checks each household electrical appliances electricity consumption situation by the end APP or the end Web；User Ke Tong
It crosses the end APP or the end Web to send out code and close electric appliance, or continues to run.
Compared with prior art, the invention has the advantages that:
1. the intelligent real-time monitoring that the present invention realizes household electrical appliance: using electric current and voltage sensor in household electricity
Electric signal acquired in real time, to analog signal carry out low-pass filtering, reduce noise processed, using analog signal to count
The conversion of word signal, transmits a signal to dual core processor at this time, and processor carries out data processing to the signal received.It is logical
Specific data processing and analysis are crossed, the electricity consumption situation in family can be monitored in real time, load type, statistics is recognized accurately
The specific power consumption condition of household electricity in a period of time, and intellectual analysis can be carried out to it, to help to reduce household electricity energy
Consumption.
2. electric appliance recognition accuracy improves: using electric transient characterisitics when electric appliance identifies, acquiring electric transient current waveform, make
Be converted into gray level image, recycle image processing method as the input terminal of convolutional neural networks and export last classification
As a result, can with high-accuracy identify load switch and type in this way；And then the situation of electrical equipment control mistake is prevented, it reaches
To the function of accurate electrical equipment control, and possess higher accuracy rate.
3. the present invention realizes electric appliance intelligent control: the present invention proposes to be based on the infrared appliance control system of WIFI, pass through before this
After convolutional neural networks algorithm precisely identifies load type, long-range closing and the function that electric appliance is realized by infrared transmitter
It can switching；Also remote controlled electrical household appliances and the necessary reception against electric appliance in same room existing for traditional home appliance remote control mode are solved
Operational issue as device, practical value with higher；
4. the present invention can also play suggesting effect: when monitoring multiplexing electric abnormality, such as when user forgets powered-down device, causing
Some electric appliances such as electric blanket, when the long-plays such as immersion heater, load detecting and control device can be by networks to client APP
End or the end Web send prompt information, to avoid occurring the danger of fire.
Detailed description of the invention
Fig. 1 is the block diagram of load detecting and control device of the invention；
Fig. 2 is the data flow diagram of intelligent appliance load detecting and control of the present invention；
Fig. 3 is remained capacity flow chart of the invention.
Specific embodiment
With reference to the accompanying drawing and specific embodiment the present invention is further illustrated.
As shown in Figs. 1-2, the intelligent appliance load detecting and control system of the present embodiment, including load detecting and control dress
It sets, server, the end APP or the end Web.It is analyzed by the basic electricity signal to electric appliance, devises intelligent power load inspection
Survey and control device, load detecting and control device include data acquisition module, data preprocessing module, analog-to-digital conversion module,
Consumption detection module, memory module, power module, control module, enabling dog module, WIFI communication module and intelligent processor
Module, the end APP or the end Web are connect by wireless network with server communication, and then are communicated with load detecting and control device.
Data acquisition module: including current sensor and voltage sensor, the electric signal in household electricity is acquired in real time.
Current sensor is socketed in the bus of household electricity, which is able to detect that current signal on home bus, and will
Its signal is converted into the AC signal of 0-50mA；Voltage sensor is connected in parallel in the total line of household electricity, which can
Using by 220V AC conversion as the AC signal of maximum non-aliased voltage 2.5V.
Data preprocessing module: the main function of this module is handled signal, mainly include noise reduction, filtering with
And signal amplification；Voltage collected, current signal are amplified, later, signal stream is divided into two-way, carries out modulus all the way
Sampling is completed in conversion, and another way enters consumption detection module, carries out the calculating with electrical power consumed.
Analog-to-digital conversion module: the module uses high-precision modulus conversion chip, realizes the analog-to-digital conversion of voltage and current,
24 high-accuracy digital signals are provided, for intelligent processor module to meet in follow-up work to the processing of current signal, deposit
Storage and control etc. require.
Consumption detection module: after the product between the voltage and current signal of input is accumulated to certain electricity consumption, the core
Sector-meeting inputs a pulse signal to intelligent processor module.Intelligent processor module is passed through after calculation processing, will realize electricity
The accurate calculation of energy, and these values are stored in memory module, to realize electric quantity accumulation.
Memory module: memory module selects the voltage and current data of external SRAM storage analog-to-digital conversion module acquisition.
Intelligent processor module: for dual core processor.
Core one realizes Intelligent treatment function, intelligent processor module real-time reception current sensor and voltage sensor acquisition
Electric signal, obtain data, after data prediction, progress event detection first, that is, judge the switch of electric appliance is judged as
When the state that electric appliance is opened, current waveform figure is generated, and carries out convolutional neural networks algorithm (CNN) identification electric appliance, output electric appliance point
For class as a result, being sent to server, user obtains electric appliance classification results and electricity consumption situation by the end APP or the end Web.
Core two realizes home wiring control function, and the instruction that the end APP or the end Web are sent reaches load detecting and control by router
The intelligent processor module of device processed after two data receiver of core, passes through data processing, verification, parsing WIFI data and agreement
Data carry out corresponding infrared coding, and infrared coding are transmitted to control module.
Control module: this module is the execution module of intelligent processor module control function, and control module is set by infrared
Standby (infrared transmitter) emits infrared signal, after the infrared receiver end of electric appliance receives, realizes the control to household electrical appliances, realizes
The infrared timed startup/shutdown function to household electrical appliances switches.
Power module: realizing 220V to 5v and the conversion of 3.3V voltage, is each module for power supply.
Enabling dog module: the partial function is that anti-locking system enters endless loop or crash, plays the role of system reset.
WIFI communication module: this module realizes the data and instruction between intelligent processor module and the end APP or the end Web
Transmission.
Load detecting and control device further include LED light and DEBUG interface module.
The electricity consumption characteristic information that server: receiving load detecting and control device transmits realizes appliance type judgement
The functions such as display, electricity consumption status monitoring and human-computer interaction.
To avoid forgetting turning off the generation of the dangerous electric appliance bring fire hazard such as electric appliance, especially electric blanket, immersion heater,
Load detecting and control device are in daily power monitoring, if monitoring electric appliance long-play, are determined as that electricity consumption is different at this time
Multiplexing electric abnormality information can be passed through server transport to the end APP or the end Web of user, user by normal information, intelligent processor module
It may be selected to close or continue to run.
As shown in figure 3, the intelligent appliance load inspection implemented based on intelligent appliance load detecting above-mentioned and control system
Survey and control method, comprising the following steps:
Step 1, collection voltages, current signal, and it is transmitted to dual core processor；
Using voltage sensor and current sensor acquisition load voltage, current signal, then carry out noise reduction, filtering and
Signal amplification, and analog-to-digital conversion is carried out, 24 high-accuracy digital signals are provided for dual core processor；It is examined simultaneously by power consumption
Survey the calculating that module use electrical power consumed.
Step 2 identifies load switch；
In order to reduce the noise jamming to electric current, in calculating current intensity, the electric current of each sampled point in the period is subtracted
The average value in period, then takes absolute value；Then using average value as the current cycle of current strength, such as formula (1).It is counting
After the current strength for calculating each period, keep current strength different, that is to say, that subtract the current strength in adjacent periods；When
When difference is greater than preset threshold, it is judged as ON；On the contrary, being judged as closing, such as formula if result is less than preset threshold
(3)。
ΔIi=Ii+1-Ii(i=1,2.....n-1) (3)
Wherein, IijFor the current strength of each sampled point, Ii-meanFor current strength average value in the sampling period, △ IiFor
The current strength difference of adjacent periods.
Step 3 generates waveform image；
Dual core processor is integrated into current waveform figure after receiving electric signal, then carries out greyscale transform process, is allowed into
For the image of 64*64 size；Each pixel is finally sought into mean value, directly as the input picture of convolutional neural networks.This
Kind processing mode, is more advantageous to cnn category of model electric appliance type.
Step 4, remained capacity: being handled waveform image by convolutional neural networks algorithm, is accurately identified load
Type simultaneously calculates power consumption, and result is then sent to server.
Convolutional neural networks algorithm the following steps are included:
(1) CONV1: the size of convolution input picture is 32 convolution kernels of 3 × 3 pixels, followed by ReLU operator,
Followed by max pooling and local acknowledgement's normalization layer (LRN) in the region 2*2；
(2) CONV2: then, handling the Feature Mapping previously exported by the second convolutional layer, wherein including 64 sizes
For the filter of 3 × 3 pixels；Followed by ReLU, be maximized for 2 × 2 regions maximum pond layer, and have with as before
Hyper parameter local acknowledgement normalize layer；
(3) CONV3: the last layer is identical as the second layer；The output on upper layer is the filter that 64 sizes are 3 × 3 pixels
Convolution, followed by pooling layers of ReLU and max；
(4) full articulamentum FC1: the output of third layer convolutional layer is converted into one-dimensional array, and is connected entirely as first layer
The input of layer；It includes 4096 neurons, followed by one Relu and one dropout layers；
(5) full articulamentum FC2: receiving 4096 neurons of full articulamentum FC1, and again include 4096 neurons,
Followed by Relu and one dropout layers；
(6) full articulamentum FC3: this layer is output layer, according to actual classification collection class, according to actual category set
Classification is closed, there are 12 kinds of experiments in this experiment, therefore there are 12 neurons in this layer；
(7) output category result: the classification results of output are sent to server.
Model is constructed using the convolutional neural networks structure of the present embodiment, training parameter has higher standard as the result is shown
True rate.
Step 5, user check each household electrical appliances electricity consumption situation by the end APP or the end Web；
Step 6, electrical equipment control: the end APP or the end Web send instruction and reach dual core processor, solve to received instruction
Analysis and infrared coding emit infrared signal and realize after the infrared receiver termination of household electrical appliances receives infrared signal to remotely to household electrical appliances
Control；
Step 7, abnormality processing: when user forgets powered-down device, leads to electric appliance long-play, load detecting and control
For device monitoring to multiplexing electric abnormality information, dual core processor identifies load switch, type and power consumption according to step 1 to step 4
Amount, and transmits the result to server, while sending prompt information to user by WIFI network, user pass through the end APP or
Web checks at end each household electrical appliances electricity consumption situation；User can send out code by the end APP or the end Web and close electric appliance, or after
Reforwarding row.
In fact, abnormality processing situation is not limited to the case where forgetting powered-down device, during electric appliance itself is up at this time.With
Electrical anomaly information can also include that electric appliance itself exception occurs, when load detecting and control device through the invention can be long
Between monitor the power information of each electric appliance, establish the database of each appliance information, before apparatus failure, often will appear hair
Situations such as heat, power consumption increase, curent change.At this point, there may be the danger that failure occurs, load detecting and control dresses for electric appliance
Setting also can send prompt information to the end client APP or the end Web by network.
In conclusion the present invention distinguishes collection voltages and current signal using voltage, current sensor, by current signal
Switch to current waveform figure, from the angle of image, then as the input terminal of CNN, starts progress neural network and successively count
It calculates, finally exports different types of probability distribution, choose maximum probability is last classification results to get final out
Electric appliance type.By Experimental Comparison, cnn (convolutional neural networks) algorithm, identification accurately can reach 99.03%, and classification results are such as
Shown in table 1.
Table 1
Certainly, the above description is not a limitation of the present invention, and the present invention is also not limited to the example above, the art
Those of ordinary skill, within the essential scope of the present invention, the variations, modifications, additions or substitutions made all should belong to this hair
Bright protection scope.
Claims (10)
1. a kind of intelligent appliance load detecting and control system, it is characterised in that: including load detecting and control device, service
Device, the end APP or the end Web, the load detecting and control device include data acquisition module, data preprocessing module, modulus turn
Change the mold block, consumption detection module, memory module, power module, control module, enabling dog module, WIFI communication module and intelligence
Processor module, the data acquisition module acquire the electric signal in household electricity in real time, and pass through data preprocessing module
After carrying out noise reduction, filtering and signal enhanced processing, signal stream is divided into two-way, enters analog-to-digital conversion module all the way, enters function all the way
Detection module is consumed, measures electricity for intelligent processor module；The intelligent processor module is dual core processor, and core one is realized
Intelligent treatment identifies that electric appliance, core two realize home wiring control by convolutional neural networks algorithm, passes through parsing WIFI data and agreement
Data realize that the infrared timed startup/shutdown function to household electrical appliances switches.
2. intelligent appliance load detecting according to claim 1 and control system, it is characterised in that: the end APP or the end Web are logical
It crosses wireless network to connect with server communication, and then is communicated with load detecting and control device, realize that the end APP or the end Web are sent
Instruction reaches the intelligent processor module of load detecting and control device, and the intelligent processor module carries out received instruction
Parsing and infrared coding, and infrared coding is transmitted to control module, the control module emits infrared signal, through the red of household electrical appliances
After outer receiving end receives, the control to household electrical appliances is realized.
3. intelligent appliance load detecting according to claim 2 and control system, it is characterised in that: the data acquisition mould
Block includes the current sensor and voltage sensor being installed in household electricity bus, the intelligent processor module real-time reception
The electric signal of current sensor and voltage sensor acquisition, and current waveform figure is generated, and it is defeated to carry out convolutional neural networks algorithm
Electric appliance classification results out, are sent to server, and user obtains electric appliance classification results and electricity consumption situation by the end APP or the end Web.
4. intelligent appliance load detecting according to claim 3 and control system, it is characterised in that: the intelligent processor
Multiplexing electric abnormality information is also passed through server transport to the end APP or the end Web of user by module.
5. a kind of intelligent appliance load detecting and control method, which comprises the following steps:
Step 1, collection voltages, current signal, and it is transmitted to dual core processor；
Step 2 identifies load switch；
Step 3 generates waveform image；
Remained capacity: step 4 handles waveform image by convolutional neural networks algorithm, is accurately identified load type
And power consumption is calculated, result is then sent to server；
Step 5, user check each household electrical appliances electricity consumption situation by the end APP or the end Web；
Step 6, electrical equipment control: the end APP or the end Web send instruction and reach dual core processor, to received instruction carry out parsing and
Infrared coding emits infrared signal and realizes after the infrared receiver termination of household electrical appliances receives infrared signal to remotely to the control of household electrical appliances
System；
Step 7, abnormality processing: when monitoring multiplexing electric abnormality information, dual core processor sends prompt information to use by network
Family, user can be closed or be continued to run by the end APP or the selection of the end Web.
6. intelligent appliance load detecting according to claim 5 and control method, it is characterised in that: in step 1, utilize
Then voltage, the current signal of voltage sensor and current sensor acquisition load carry out noise reduction, filtering and signal amplification, and
Analog-to-digital conversion is carried out, 24 high-accuracy digital signals are provided for dual core processor；It is used simultaneously by consumption detection module
The calculating of electrical power consumed.
7. intelligent appliance load detecting according to claim 5 and control method, it is characterised in that: in step 2, in order to
Reduction subtracts the average value of the current cycle of each sampled point in the period in calculating current intensity to the noise jamming of electric current,
Then it takes absolute value；Then using average value as the current cycle of current strength, such as formula (1)；In the electricity for calculating each period
After intensity of flow, keep current strength different, that is to say, that subtract the current strength in adjacent periods；When difference is greater than default threshold
When value, it is judged as ON；On the contrary, being judged as closing, such as formula (3) if result is less than preset threshold；
ΔIi=Ii+1-Ii(i=1,2.....n-1) (3)
Wherein, IijFor the current strength of each sampled point, Ii-meanFor current strength average value in the sampling period, △ IiIt is adjacent
The current strength difference in period.
8. intelligent appliance load detecting according to claim 6 and control method, it is characterised in that: in step 3, double-core
Processor is integrated into current waveform figure after receiving electric signal, then carries out greyscale transform process, makes 64*64 size
Image；Each pixel is finally sought into mean value, directly as the input picture of convolutional neural networks.
9. intelligent appliance load detecting according to claim 8 and control method, it is characterised in that: in step 4, convolution
Neural network algorithm the following steps are included:
(1) CONV1: the size of convolution input picture is 32 convolution kernels of 3 × 3 pixels, followed by ReLU operator, followed by 2*
Max pooling and local acknowledgement in 2 regions normalize layer LRN；
(2) CONV2: then, handling the Feature Mapping previously exported by the second convolutional layer, wherein being 3 × 3 comprising 64 sizes
The filter of pixel；Followed by ReLU, be maximized for 2 × 2 regions maximum pond layer, and with hyper parameter as before
Local acknowledgement normalize layer；
(3) CONV3: the last layer is identical as the second layer；The output on upper layer is the volume for the filter that 64 sizes are 3 × 3 pixels
Product, followed by pooling layers of ReLU and max；
(4) full articulamentum FC1: the output of third layer convolutional layer is converted into one-dimensional array, and as the full articulamentum of first layer
Input；It includes 4096 neurons, followed by one Relu and one dropout layers；
(5) full articulamentum FC2: 4096 neurons of full articulamentum FC1 are received, and again include 4096 neurons, then
It is Relu and one dropout layers；
(6) full articulamentum FC3: this layer is output layer, according to actual classification collection class, there is 12 nerves in this layer
Member；Output category result.
10. intelligent appliance load detecting according to claim 8 and control method, it is characterised in that: in step 7, when with
Powered-down device is forgotten at family, when leading to electric appliance long-play, dual core processor according to step 1 to step 4 identify load switch,
Type and power consumption, and server is transmitted the result to, while prompt information is sent to user by WIFI network, user passes through
Each household electrical appliances electricity consumption situation is checked at the end APP or the end Web；User can send out code by the end APP or the end Web and close electric appliance
It closes, or continues to run.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810725906.8A CN108964270A (en) | 2018-07-04 | 2018-07-04 | A kind of intelligent appliance load detecting and control system and its method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810725906.8A CN108964270A (en) | 2018-07-04 | 2018-07-04 | A kind of intelligent appliance load detecting and control system and its method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108964270A true CN108964270A (en) | 2018-12-07 |
Family
ID=64485494
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810725906.8A Pending CN108964270A (en) | 2018-07-04 | 2018-07-04 | A kind of intelligent appliance load detecting and control system and its method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108964270A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109765443A (en) * | 2019-01-17 | 2019-05-17 | 创炘源智能科技(上海)有限公司 | Detect the device and method of the electric appliance load on power supply line |
| CN110060369A (en) * | 2019-04-03 | 2019-07-26 | 国网福建省电力有限公司 | A kind of Distribution Network Communication data interaction monitoring system |
| CN110225067A (en) * | 2019-07-24 | 2019-09-10 | 上海戎磐网络科技有限公司 | A kind of Internet of Things safety pre-warning system |
| CN111103477A (en) * | 2019-10-31 | 2020-05-05 | 武汉旗云高科信息技术有限公司 | Electrical load monitoring method and system based on 3D convolutional neural network |
| CN111539275A (en) * | 2020-04-14 | 2020-08-14 | 中南民族大学 | Electrical load classification method and system based on load characteristic visualization |
| CN113947886A (en) * | 2021-12-20 | 2022-01-18 | 北京易联通达科技有限公司 | Power consumption behavior monitoring device based on non-invasive ammeter |
| CN114545787A (en) * | 2022-03-04 | 2022-05-27 | 合肥惟新半导体科技有限公司 | Household appliance control system for reducing standby power consumption |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP2639589A1 (en) * | 2008-07-17 | 2013-09-18 | Isis Innovation Limited | Utility metering |
| CN103400487A (en) * | 2013-07-26 | 2013-11-20 | 华南理工大学 | Household appliance energy consumption managing system and household appliance energy consumption managing method based on intelligent mobile phone |
| CN104468647A (en) * | 2013-09-15 | 2015-03-25 | 国家电网公司 | Smart household system |
| CN104834297A (en) * | 2015-05-08 | 2015-08-12 | 上海电机学院 | Household electrical appliance remote control system based on Internet-of-things smart meter and method thereof |
| CN106200408A (en) * | 2016-08-26 | 2016-12-07 | 特斯联（北京）科技有限公司 | A kind of Smart Home based on cloud computing is equipped with method and system |
| CN106302059A (en) * | 2016-10-21 | 2017-01-04 | 北京中天智汉科技有限责任公司 | A kind of long-range control method for Smart Home and device |
| CN106908657A (en) * | 2017-03-06 | 2017-06-30 | 中国海洋大学 | A kind of load switch detection means and method based on power |
| CN107478882A (en) * | 2017-09-06 | 2017-12-15 | 上海斐讯数据通信技术有限公司 | Intelligent electric meter long-distance monitorng device |
-
2018
- 2018-07-04 CN CN201810725906.8A patent/CN108964270A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP2639589A1 (en) * | 2008-07-17 | 2013-09-18 | Isis Innovation Limited | Utility metering |
| CN103400487A (en) * | 2013-07-26 | 2013-11-20 | 华南理工大学 | Household appliance energy consumption managing system and household appliance energy consumption managing method based on intelligent mobile phone |
| CN104468647A (en) * | 2013-09-15 | 2015-03-25 | 国家电网公司 | Smart household system |
| CN104834297A (en) * | 2015-05-08 | 2015-08-12 | 上海电机学院 | Household electrical appliance remote control system based on Internet-of-things smart meter and method thereof |
| CN106200408A (en) * | 2016-08-26 | 2016-12-07 | 特斯联（北京）科技有限公司 | A kind of Smart Home based on cloud computing is equipped with method and system |
| CN106302059A (en) * | 2016-10-21 | 2017-01-04 | 北京中天智汉科技有限责任公司 | A kind of long-range control method for Smart Home and device |
| CN106908657A (en) * | 2017-03-06 | 2017-06-30 | 中国海洋大学 | A kind of load switch detection means and method based on power |
| CN107478882A (en) * | 2017-09-06 | 2017-12-15 | 上海斐讯数据通信技术有限公司 | Intelligent electric meter long-distance monitorng device |
Non-Patent Citations (1)
| Title |
|---|
| ZEHUA LAN等: "A non-intrusive load identification method based on convolution neural network", 《2017 IEEE CONFERENCE ON ENERGY INTERNET AND ENERGY SYSTEM INTEGRATION》 * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109765443A (en) * | 2019-01-17 | 2019-05-17 | 创炘源智能科技(上海)有限公司 | Detect the device and method of the electric appliance load on power supply line |
| CN110060369A (en) * | 2019-04-03 | 2019-07-26 | 国网福建省电力有限公司 | A kind of Distribution Network Communication data interaction monitoring system |
| CN110060369B (en) * | 2019-04-03 | 2022-01-11 | 国网福建省电力有限公司 | Power distribution network communication data interaction monitoring system |
| CN110225067A (en) * | 2019-07-24 | 2019-09-10 | 上海戎磐网络科技有限公司 | A kind of Internet of Things safety pre-warning system |
| CN110225067B (en) * | 2019-07-24 | 2021-08-24 | 上海戎磐网络科技有限公司 | Internet of things safety early warning system |
| CN111103477A (en) * | 2019-10-31 | 2020-05-05 | 武汉旗云高科信息技术有限公司 | Electrical load monitoring method and system based on 3D convolutional neural network |
| CN111103477B (en) * | 2019-10-31 | 2020-08-25 | 武汉旗云高科信息技术有限公司 | Electrical load monitoring method and system based on 3D convolutional neural network |
| CN111539275A (en) * | 2020-04-14 | 2020-08-14 | 中南民族大学 | Electrical load classification method and system based on load characteristic visualization |
| CN113947886A (en) * | 2021-12-20 | 2022-01-18 | 北京易联通达科技有限公司 | Power consumption behavior monitoring device based on non-invasive ammeter |
| CN114545787A (en) * | 2022-03-04 | 2022-05-27 | 合肥惟新半导体科技有限公司 | Household appliance control system for reducing standby power consumption |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108964270A (en) | A kind of intelligent appliance load detecting and control system and its method | |
| CN104333736B (en) | Unmanned substation's intelligent recognition monitoring system and method | |
| CN103815912B (en) | Old solitary people based on thermal infrared sensor array falls down behavior method of real-time | |
| CN108040886A (en) | A kind of intensive live pig healthy aquaculture intelligence system and method based on Internet of Things | |
| CN104375486B (en) | Intelligent home control system | |
| CN112947127A (en) | Intelligent electricity consumption control management system | |
| CN107833416A (en) | A kind of smart home remote antitheft alarm system based on Internet of Things | |
| Paradiso et al. | ANN-based appliance recognition from low-frequency energy monitoring data | |
| CN106842356B (en) | There is nobody detection method and detection system in a kind of interior | |
| CN106654739B (en) | A kind of socket and housed device intelligence power-off implementation method of band intelligence power down function | |
| CN106908657A (en) | A kind of load switch detection means and method based on power | |
| CN108233529A (en) | Intelligent remote power monitoring modularization apparatus and system | |
| CN108983622A (en) | A kind of intelligent appliance control system based on human body gesture | |
| CN110040591A (en) | A kind of escalator monitoring device and method based on Internet of Things cloud platform | |
| CN107132826A (en) | A kind of multifunctional intellectual home security control system | |
| CN112491668A (en) | Intelligence house thing networking safety monitoring device | |
| CN112071047A (en) | Resident home leaving and returning behavior judgment method | |
| CN102333020A (en) | Sub-district terminal network oriented security system | |
| CN111077817A (en) | Security monitoring management system for power company based on Internet of things | |
| CN113328527A (en) | Intelligent power distribution automatic terminal device and intelligent power distribution fault diagnosis method | |
| CN107995251A (en) | A kind of intelligent domestic system based on Internet of Things | |
| CN106910318A (en) | One kind turnover behavior monitoring method and system | |
| CN210271162U (en) | Multi-parameter fire alarm system formed by network of gateways | |
| CN107643692A (en) | Energy efficiency management system and method | |
| CN107658976A (en) | A kind of new bus warning monitoring system |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181207 |