CN113517613A

CN113517613A - Intelligent disaster prevention energy-saving socket based on load identification algorithm and management system thereof

Info

Publication number: CN113517613A
Application number: CN202110388273.8A
Authority: CN
Inventors: 黄智栋; 胡尚; 高佳彬; 龙胜春; 吕诗; 姚懿; 曹迪; 周越瀚; 王海鹏; 李爱萍; 高诚裕; 余金隆; 程昊哲
Original assignee: Zhejiang University of Technology ZJUT
Current assignee: Zhejiang University of Technology ZJUT
Priority date: 2021-04-12
Filing date: 2021-04-12
Publication date: 2021-10-19

Abstract

An intelligent disaster-prevention energy-saving socket based on a load identification algorithm and a management system thereof comprise a socket upper cover plate, a plug guide groove, a retainer, jack reeds, power-on indicator lamps, a socket bottom side cover plate and a PCB circuit board, wherein each jack corresponds to one power-on indicator lamp, the PCB circuit board comprises an MCU controller, a WIFI module, a USB serial port, an NFC module, an electric energy metering module, a relay, a wiring terminal, a fuse, an AC-DC converter and a power supply module, the invention provides an intelligent disaster prevention energy-saving socket based on a load identification algorithm and a management system thereof, a WeChat small program is used for on-line management of electric appliances, the NFC module is embedded into an independently designed socket circuit, the intelligent household appliance management system can provide services of identifying the inserted electric appliance, converting the common household appliance into intelligent household appliance management by pasting an NFC label, refusing access of the forbidden electric appliance in a specific scene, remotely managing the household appliance and the like.

Description

Intelligent disaster prevention energy-saving socket based on load identification algorithm and management system thereof

Technical Field

The invention relates to the technical field of intelligent sockets, in particular to an intelligent disaster prevention energy-saving socket based on a load identification algorithm and a management system thereof.

Background

The electrical appliance becomes necessary household equipment in each family, the socket is used as equipment which is necessary to be used when the electrical appliance is connected with a power supply, if the electrical appliance is subjected to intelligent control management, intelligent control over common electrical appliances can be realized to a great extent by means of management over the socket, so that the simple management of the working state of the electrical appliance and the safety of power utilization are achieved, and the waste of power resources caused when the electrical appliance is in a standby state is reduced.

Li Miao et al teach a WiFi type smart jack based on WeChat platform control, Zhouqing et al propose a smart jack outage delay control method under the environment of Internet of things, utilize AD converter to detect smart jack power, feed back the power consumption information of electrical equipment connected with the smart jack to the user in real time, the user can switch on and off the jack at any time, realize the control of the smart jack, thus reach the energy-conserving purpose. Few students relate to the intelligent socket based on load identification, and the intelligent socket function mainly researched by the students at present is a power-off function, but the realization of functions such as identification management of miscellaneous electric appliances is ignored.

The existing intelligent socket scheme generally has the functions of remotely controlling the on-off of a socket power supply, counting the total power consumption, timing a switch and the like, but cannot identify specific electric appliances connected to specific sockets and the power consumption management of independent electric appliances, and only intelligent home furnishing can be replaced by common independent electric appliance management, for example, a common electric fan is replaced by a series of intelligent products such as a rice household appliance fan, and the like, namely incomplete intelligence is achieved. In the aspect of safety, the common intelligent socket only has the functions of simple overload protection and the like, and cannot meet special safety requirements in special scenes. At present, fire disasters caused by lithium battery charging in campuses and communities are all the same, existing load identification functional equipment such as overseas Sense and other company products adopt an active learning marking technology, small furniture electric appliances can be identified for months or even years, and the defects that the identification accuracy of the electric appliances of the same type is poor, and articles without background label libraries cannot be subjected to real-time learning prediction by using the active learning technology in real time exist.

Disclosure of Invention

In order to overcome the defects that the existing intelligent socket cannot identify inserted electric appliances, is inconvenient to manage the access of high-risk electric appliances according to scenes, cannot intelligently manage common electric appliances, has insufficient energy-saving display functions such as power consumption and the like, the invention provides an intelligent disaster-prevention energy-saving socket based on a load identification algorithm and a management system thereof, a WeChat small program is used for carrying out on-line management on the electric appliances, an NFC module is embedded into an independently designed socket circuit, and a double verification type electric appliance intelligent marking technology based on an NFC label and the load identification algorithm is designed to solve the problems that the identification speed of small electric appliances is extremely slow, new data cannot be predicted in real time, the identification of electric appliances of the same model is difficult to solve by an active learning algorithm, the intelligent household appliance management system can provide services of identifying the inserted electric appliance, converting the common household appliance into intelligent household appliance management by pasting an NFC label, refusing access of the forbidden electric appliance in a specific scene, remotely managing the household appliance and the like.

The technical scheme adopted by the invention for solving the technical problems is as follows:

an intelligent disaster prevention energy-saving socket based on a load identification algorithm comprises a socket upper cover plate, a plug guide groove, a retainer, jack reeds, energizing indicator lamps, a socket bottom side cover plate and a PCB circuit board, wherein each jack corresponds to one energizing indicator lamp, the PCB circuit board comprises an MCU controller, a WIFI module, a USB serial port, an NFC module, an electric energy metering module, a relay, a wiring terminal, a fuse, an AC-DC converter and a power supply module, the MCU controller adopts an ESP32-S module, the ESP32-S module integrates an antenna switch, a radio frequency Balun, a power amplifier, a low noise amplifier, a filter and a power supply management module, the NFC module comprises an NFC read-write chip and an NFC antenna, the NFC antenna is connected with the NFC read-write chip, the NFC read-write chip is connected with the ESP 801 32-S module, the electric energy metering module adopts an HLW 2 chip, pins 2 of the HLW8012 chip are current differential signal input ends, pin 4 is a voltage sampling signal input end,

pins

6 and 7 of the HLW8012 chip are respectively connected with

pins

13 and 16 of the ESP32-S module, and the ESP32-S module calculates power, current and voltage values by measuring the period of high-frequency pulses output by

pins

6 and 7 of the HLW8012 chip;

the power supply module is connected with 220v alternating current through the wiring terminal, and voltage conversion is carried out through the AC-DC converter after the current flows through the fuse, so that power is supplied to other circuit modules; the ESP32-S module is connected with a user terminal through a relay, is connected with the cloud end through a WIFI module, an NFC label is pasted on a plug end of the electrical equipment, and data information in the NFC label is matched with relevant parameters of the household appliance.

Furthermore, the NFC read-write chip adopts a PN5321A3HN chip,

pins

27 and 28 of the PN5321A3HN chip are respectively connected with

pins

36 and 33 of an ESP32-S module, and the PN5321A3HN chip and the ESP32-S module are communicated through I2C; the USB serial port adopts a CH340C chip, and

pins

2 and 3 of the CH340C chip are respectively connected to pins U0TXD and U0RXD of the ESP32-S module.

An intelligent disaster-prevention energy-saving socket management system based on a load identification algorithm comprises a data acquisition and calculation framework, a cloud and a terminal, wherein the data acquisition and calculation framework comprises a sensor layer and an edge gateway layer, the cloud comprises a cloud server and a GPU server, the terminal comprises the intelligent disaster-prevention energy-saving socket and a user terminal according to claim 1 or 2, and a user interacts with the sensor layer through the user terminal;

the method comprises the steps that an MCU (micro control unit) controller and a WIFI (wireless fidelity) module form an edge gateway layer, an electric energy metering module forms a sensor layer, the sensor layer collects data and uploads the data to the edge gateway, the edge gateway runs an activation monitoring algorithm to obtain effective data and transmits the effective data to a GPU (graphics processing unit) server interface layer, the server interface layer stores the effective data of the electric appliance to a MySQL (structured query language) database and a Redis database, load identification is carried out by using a label consistency convolution transformation neural network model, the effective data are uploaded to a cloud server, the cloud server is responsible for label consistency convolution transformation neural network model deployment and various functional interface deployment, electric appliance identification, real-time early warning, intelligent analysis electric data and electric data viewing function services are provided for a user, and meanwhile, an App (application), a WeChat small program and a user terminal are provided to finish interaction.

The invention has the following beneficial effects: the data acquisition and calculation framework adopts a sensor layer, an edge gateway layer and a cloud layer, so that the quality of transmitted data is ensured, and the bandwidth is saved; the convolution transformation network is adopted to carry out load identification on the socket access electric appliance, so that the inserted electric appliance can be intelligently identified and the common electric appliance can be intelligently managed; the socket is embedded with the NFC module which can read and write the information of the electric appliance and store the information on the NFC label of the plug, so that the problems that the small electric appliance identification speed is extremely slow, new data cannot be predicted in real time, the electric appliances of the same type are difficult to identify and the like which are difficult to solve by an algorithm are solved; compared with a common intelligent socket, the intelligent disaster prevention energy-saving socket based on the load identification algorithm additionally provides an electric appliance load identification function, a special electric appliance customized access rejection service under a specific scene and a series of services derived from the service.

Drawings

Fig. 1 is an exploded view of the intelligent disaster prevention energy-saving socket of the present invention.

Fig. 2 is a view of the general household appliance plug attached with NFC.

Fig. 3 is a hardware architecture diagram of a PCB circuit board.

FIG. 4 is a schematic circuit diagram of the Power Supply module of the present invention.

FIG. 5 is a schematic circuit diagram of a USB module according to the present invention.

Fig. 6 is a schematic circuit diagram of the WIFI & MCU module of the present invention.

FIG. 7 is a schematic diagram of a Power Monitoring circuit according to the present invention.

Fig. 8 is a schematic diagram of an NFC circuit in the present invention.

Fig. 9 is a schematic circuit diagram of a RELAY module according to the present invention.

Fig. 10 is a diagram of a smart jack system architecture.

Fig. 11 is a flow chart of the intelligent marking technique of the present invention.

Detailed Description

The invention is further described below with reference to the accompanying drawings.

Referring to fig. 1 to 11, an intelligent disaster prevention energy-saving socket based on a load identification algorithm comprises a socket upper cover plate 1, a plug guide groove and holder 2, jack reeds 3, energizing indicator lamps 4, a socket bottom side cover plate 5 and a PCB circuit board 6, wherein each jack corresponds to one energizing indicator lamp, the PCB circuit board 6 comprises an MCU controller, a WIFI module, a USB serial port, an NFC module, an electric energy metering module, a relay, a connecting terminal, a fuse, an AC-DC converter and a power module, the MCU controller adopts an ESP32-S module, the ESP32-S module integrates an antenna switch, a radio frequency Balun, a power amplifier, a low noise amplifier, a filter and a power management module, the NFC module comprises an NFC read-write chip and an NFC antenna, the NFC antenna is connected with the NFC read-write chip, the NFC read-write chip is connected with the ESP32-S module, the electric energy metering module adopts an HLW8012 chip,

pins

2 and 3 of the HLW8012 chip are current differential signal input ends, pin 4 is a voltage sampling signal input end,

pins

6 and 7 of the HLW8012 chip are respectively connected with

pins

13 and 16 of an ESP32-S module, and the ESP32-S module calculates power, current and voltage values by measuring the period of high-frequency pulses output by

pins

6 and 7 of the HLW8012 chip;

Furthermore, the NFC read-write chip adopts a PN5321A3HN chip,

pins

27 and 28 of the PN5321A3HN chip are respectively connected with

pins

As shown in fig. 1, the intelligent disaster prevention energy-saving socket based on the load recognition algorithm is composed of six parts, namely a socket upper cover plate 1 (including a plurality of jacks and wire plugging holes), a plug guide groove and holder 2, jack reeds 3, an energizing indicator lamp 4, a socket bottom side cover plate 5 and a PCB 6.

As shown in fig. 2, both the three plugs and the two plugs can be pasted with ready readable and writable NFC tags. Paste the NFC label at electrical equipment's plug end, data information and the relevant parameter phase-match of household electrical appliances (can set up at the removal end) in the NFC label, the user is before inserting household electrical appliances, is close to smart jack with the NFC label, and the embedded NFC read-write module of socket reads the information in the coil, uploads to the high in the clouds through the WIFI module with corresponding data. Household electrical appliances are in the use, and electric energy metering module can read the effective value of electric current and voltage to provide the temperature and detect, corresponding data is also transmitted to the high in the clouds in real time through the WIFI module, with the in service behavior that detects household electrical appliances. When the cloud detects that the current curve is abnormal, an instruction is sent to the MCU controller to control the relay to be disconnected, and danger is avoided. The user can also control the on-off of the relay at the mobile terminal so as to realize the control of the household appliance. The NFC tag attach location is as in fig. 2.

The PCB circuit board shown in FIG. 3 is composed of six parts, namely a Power Supply module, a USB module, a WIFI & MCU module, an NFC module, a RELAY RELAY module and an electric energy metering module, namely a Power Monitoring module.

As shown in fig. 4, the power module is connected to 220V AC through the connection terminal, and the 220V AC is converted into 5V DC through the AC-DC converter after the current flows through the ceramic fuse, so as to be used by the relay circuit and the HLW8012 chip. And then, the 5V direct current is converted into 3.3V direct current by an LOD (low drop regulator) voltage stabilizer to be used by the NFC circuit, the USB download circuit and the ESP32-S module.

As shown in fig. 5, the USB circuit mainly implements a program downloading function, and after a download line is connected, the USB bus is converted into a serial port through a CH340C chip, and

pins

2 and 3 of the CH340C chip are connected to pins U0TXD and U0RXD of the ESP32-S module, so as to implement convenient burning of the program.

As shown in fig. 6, the WIFI & MCU part is mainly composed of an ESP32-S module, the ESP32-S module integrates an antenna switch, a radio frequency Balun, a power amplifier, a low noise amplifier, a filter, and a power management module, and Wi-Fi supports an extremely large-range communication connection and also supports a direct connection to the internet through a router; and bluetooth can let the user connect the cell-phone or broadcast BLE Beacon so that signal detection. The invention has excellent connectivity and usability, and also has a certain performance of CPU, integrates a great number of peripheral interfaces, SPI, IIC, IIS, AD, DA, PWM, IR, UART and CAN, and CAN completely meet the control and operation requirements of the invention.

As shown in fig. 7, a part of the electric energy metering module circuit mainly functions to realize the acquisition of electric energy related parameters. The sampling circuit of the HLW8012 chip adopts a resistance sampling mode.

Pins

2 and 3 of the HLW8012 chip are current differential signal input ends, and pin 4 is a voltage sampling signal input end.

Pins

6 and 7 of the HLW8012 chip are respectively connected with

pins

13 and 16 of the ESP32-S module, and the high-frequency pulse output by the pin 6 of the HLW8012 chip indicates power and calculates electric energy; outputting square waves with the duty ratio of 1: 1. The high-frequency pulse output by the pin 7 of the HLW8012 chip indicates the effective value of current or voltage, SEL selects; outputting square waves with the duty ratio of 1: 1. The interface of the ESP32-S module and the HLW8012 chip calculates power, current and voltage values by measuring the period of high-frequency pulses output by

pins

6 and 7 of the HLW8012 chip.

As shown in fig. 8, the NFC module is a part that performs an NFC tag read/write function. The external antenna is connected to the TX1, the TX2 and the RX pin of the PN5321A3HN chip after a series of impedance matching. The 27 and 28 pins of the PN5321A3HN chip are respectively connected with the 36 and 33 pins of the ESP32-S module, and the PN5321A3HN chip and the ESP32-S module communicate through I2C.

As shown in fig. 9, the RELAY module is used as a part for controlling the on/off of the circuit, when the smart socket is used, a user can send a command to the ESP32-S module through a mobile terminal of a mobile phone or a WeChat applet, and the on/off of the circuit is controlled through the RELAY. GPIO12 of the ESP32-S module generates small current, and the small current passes through the MOS tube to control the electromagnetic coil, so that the on-off of the circuit is controlled. When the GPIO12 is at low frequency, the circuit is turned on, and when the GPIO12 is at high frequency, the circuit is turned off.

As shown in fig. 10, the sensing layer mainly includes an HLW8012 chip, and collects power data continuously and uploads the power data to the edge gateway layer. The edge gateway layer is composed of edge computing groups with light-weight computing capability, the edge computing groups are embedded in a socket module and integrated in an ESP32-S module, and are responsible for performing activation detection and extraction on sensing data by using an activation monitoring algorithm, screening out invalid disturbance data, and outputting valid data to be uploaded to a cloud end layer. In the aspect of selecting the database, the invention selects the MySQL relational database to store the sensor data, and selects the Redis non-relational database as the buffer. The database continuously receives and stores the electric energy data transmitted by the intelligent terminal, and meanwhile, a training set is provided for the load recognition model.

Under low-frequency sampling, step-like changes caused by load state changes can be represented as abrupt changes of data point dispersion and can be identified through variation coefficients and load operation rules, the CNN sliding window algorithm can extract representative activation curve features, and the characteristics of the activation curve can be more accurately described by combining manually defined features.

The activation monitoring algorithm is described in detail below:

in practical situations, WIFI&The MCU module can continuously receive the data acquired by the Power Monitoring module in real time, so that new n-point active data is taken as a calculation window P_tMarking the window with the intermediate time t:

the intuitive embodiment that the load state is changed is that the power curve has step-like change, and the discrete degree of sample data is correspondingly increased. The local standard deviation, local average amplitude of the samples can be calculated to detect potential activation events. After the window is continuously slid, the load state change points can be screened out, the load state change points are classified into candidate switching time sets and candidate switching time sets through the positive and negative of the large difference value, the candidate class step point sets are generated, and the elements in the two sets are further paired to obtain the head and tail time of the activation event, and the following condition screening is carried out:

firstly, the load should have a small activation time omega in a normal state and should meet a small activation interval delta

② the power in the activation process should not be less than the initial background power

Thirdly, the background power before and after the activation is similar

Fourthly, the power fluctuation caused by the start and the end of the activation should be similar

Traversing the candidate class step point set and applying the rule, and recording the activation starting and ending time stamps t meeting the condition_on、t_offAnd if the effective activation is not found, continuing to wait and analyze newly uploaded data.

This example is row socket for ally oneself with, can peg graft on current wiring board to realize convenient and intelligent, row modular design has also given the more nimble selection of user. The NFC label is pasted at the plug end of the common electrical equipment, information in the NFC label is matched with relevant parameters of an electrical appliance (the information can be set at a mobile end), before the user inserts the electrical appliance, the NFC label is close to the socket, the socket reads information in a coil, and corresponding data are uploaded to the cloud end through the ESP32-S module. When the electric appliance is used, the electric energy metering module can read the effective values of current and voltage and provide temperature detection, and corresponding data are transmitted to the cloud end in real time through the ESP32-S module to detect the service condition of the electric appliance. When the cloud detects that the current curve is abnormal, an instruction is sent to the MCU controller to control the relay to be disconnected, and danger is avoided. The user can also control the on-off of the relay at the mobile terminal so as to realize the control of the electric appliance.

The electric energy metering module selects an HLW8012 chip, the HLW8012 chip is internally provided with a 2-path programmable gain amplifier PGA and an analog-to-digital converter ADC, analog-to-digital conversion is carried out on current and voltage sampling signals to obtain digital signals, an active power value, a current effective value and a voltage effective value are calculated inside the chip, the HLW8012 chip converts the active power value, the current effective value and the voltage effective value into square wave pulses to be output (duty ratio 1:1) after passing through a frequency conversion module, and the magnitude of each numerical value is in direct proportion to the magnitude of frequency and in inverse proportion to the magnitude of period. Indexes such as active power, electric quantity, voltage effective value, current effective value and total power consumption acquired by the HLW8012 chip are used as main classification bases, and human behavior mode data such as multi-mode data auxiliary classification of time intervals (morning, noon, afternoon and evening, lamp electrical appliances and battery chargers are started at night) are used for improving the identification precision.

Convolution Transform Learning (CTL) is a recently proposed unsupervised learning convolution-based feature generation framework. CTL can be viewed as another emerging analytical counterpart that characterizes the learning paradigm, Convolutional Dictionary Learning (CDL) in which the main idea is to learn convolutional filters that operate on data to produce features that can account for data reliability constraints. In the project, a supervised convolution transformation formula is provided to solve the problem of multi-label classification. In multi-class classification, there is a one-to-one mapping between samples and classes, whereas in multi-label classification, one sample may correspond to multiple classes simultaneously (one-to-many mapping).

CTL learning a set of filters (t) operating on an observed sample (s (k)) < 1 ≦ k_m) M is 1 or less to form a set of features

Applying sparsity punishment to the features according to initial research transformation learning; consistent with the CNN study, non-negative constraints (rectangular linear cells) were also imposed on the features. The convolution filters and coefficients learned from the data are then trained. This represents a minimization problem, and the present invention employs a minimization method after adding tag consistency. By means of an approximation operator of the shrinkage property, it can be demonstrated that the generated sequence converges to a local solution. The training phase ends here. The test phase includes testing dataThe update rule of the feature matrix X is applied. Thus, the generated features are projected onto the label space. In practice, the generated label mapping may not be binary, but it is real-valued. Therefore, the present invention will use a threshold of 0.5 to determine the activity class.

The electric load identification system can analyze and return the result obtained by electric load identification, accurately provide the type of the electric appliance, and feed back the type to a user to carry out double-verification type intelligent marking on the electric appliance. In addition, the algorithm has the characteristics of high identification precision, high identification reaction speed, strong stability and the like.

The relay module is used as a control module, and the relay has an interaction relation between a control system (an input circuit) and a controlled system (an output circuit). When the intelligent socket is used, a user can send an instruction to the embedded MCU controller through the mobile terminal of the mobile phone or the WeChat small program, so that the MCU controller generates a small current, and then the small current controls a large current on the socket through the relay. The relay is an electric control device for controlling large current through small current, and is well embodied on an intelligent socket.

The NFC module is composed of an NFC read-write chip and an NFC antenna which are embedded in a PCB. NFC is a safe and reliable close-range private communication mode, the NFC communication setting procedure is simple, the communication establishment time is short and only needs about 0.1s, and therefore the NFC reading mode is selected to be used on the intelligent socket to read the electrical appliance information.

The invention selects to use an ESP32-S module on the MCU controller. The ESP32-S module is a general WiFi-BT-BLE MCU module, the ESP32-S module is designed based on an ESP32-S0WD single-core chip, an 802.11b/g/nWi-Fi + BT SoC module with an ultra-small size and a low-power dual-core 32-bit CPU are adopted, the ESP module can be used as an application processor, the dominant frequency is up to 240MHz, and the operation capability is up to 600 DMIPS. The antenna, the RF balun, the power amplifier, the receiving low-noise amplifier, the filter, the power management module and other functions are integrated, and strong processing performance, reliable safety performance and Wi-Fi & Bluetooth functions can be realized only by few peripheral devices.

The dual verification type electrical appliance intelligent marking technology based on the NFC label and the load identification algorithm has the following use flows: a user uses a special read-write NFC label to paste the NFC label in the middle of an electric appliance plug (the three-head seat and the two-head seat are both universal). When the new electric appliance is used, a user clicks the new electric appliance in the small program, the electric appliance is opened, so that the socket acquires useful power change and transmits the useful power change to the ESP32 chip for activation processing, data are uploaded to the cloud server through the WIFI module after an activation function is met, Mysql records the electrical characteristics of the electric appliance, the background uses a convolution transformation neural network to predict the type of the electric appliance, and the small program end reminds the user to enter confirmation information. After a user inputs basic information and parameters of a new electric appliance at the cloud, the information is written into a Mysql table to establish a user tag library, meanwhile, the information is fed back to an ESP32 chip, the ESP32 chip exchanges returned ground data to a miniature NFC read-write chip built in a socket, and the NFC read-write chip is responsible for writing electric appliance data into an NFC tag. The built-in miniature NFC antenna of socket reads out label data when old electrical apparatus inserts the socket to pass through WIFI with consumer electrical information and label data by ESP32-S module and reach the high in the clouds, the result contrast that electrical apparatus label data and consumer electrical information pass through label storehouse and compare is with the consumer to the high in the clouds, if match then continue the record information, if mismatch then state label information incorrect or the label storehouse does not exist with electrical apparatus information. If no information exists, the type of the electrical appliance is predicted again through the convolutional neural network, the prediction result is returned, the user writes information again, and the label is covered with the information. And if not, the label library information is taken as the standard. After the electrical appliance image is generated at the cloud end and corresponds to the socket modeling, functions of managing electrical appliances, using the electrical appliances and checking basic information by one key at the small program end can be tested and distributed in the jacks.

The socket can recognize the charging of the lithium battery and automatically refuse to be electrified. It needs to be pointed out that a scene manager can set which electric appliance is rejected by a specific scene in an applet, so that the application flexibility is ensured. The background big data statistics and mining module can count the power consumption and other parameters of the appliance and the like, and summarize an energy-saving report to be fed back to a user, and meanwhile, the power-off of the unnecessary appliances can be remotely turned off, so that the energy-saving safety is realized.

If the type of the electric appliance is identified as the forbidden electric appliance set by the administrator, the background sends a cut-off instruction to the ESP32-S module to control the relay to realize the power-off operation, so that the operation of refusing the access of the electric appliance of the specified type is realized.

The load identification method of the convolutional conversion network based on the label consistency is used for roughly classifying the electric appliances, the power utilization conditions of the electric appliances which are easily damaged due to heating such as resistance and the like can be identified and monitored, when the accessed electric appliances are identified as high-risk electric appliances under the scene, the connection is automatically refused, and a user can prompt the user to mark and name according to the identified electric appliances when the user connects the electric appliances at a small program end for the first time, the next time, the user is inserted, the convolutional conversion network model is automatically matched with NFC, the electric appliance information stored by the label is subjected to bidirectional verification to realize the automatic identification, and the MCU controller is controlled to refuse the access after the illegal electric appliances issued by a scene manager are identified.

The embodiments described in this specification are merely illustrative of implementations of the inventive concepts, which are for purposes of illustration only. The scope of the present patent shall not be considered limited to the specific forms set forth in the examples, but rather the scope of the present patent shall be deemed to be equivalent to the technical solutions which will be suggested to one of ordinary skill in the art based on the present patent concepts.

Claims

1. An intelligent disaster prevention energy-saving socket based on a load identification algorithm is characterized in that: the socket comprises a socket upper cover plate, a plug guide groove and a retainer, jack reeds, energizing indicator lamps, a socket bottom side cover plate and a PCB circuit board, wherein each jack corresponds to one energizing indicator lamp, the PCB circuit board comprises an MCU (microprogrammed control Unit) controller, a WIFI (wireless fidelity) module, a USB (universal serial bus) serial port, an NFC (near field communication) module, an electric energy metering module, a relay, a wiring terminal, a fuse, an AC-DC (alternating current-direct current) converter and a power supply module, the MCU controller adopts an ESP32-S module, the ESP32-S module integrates an antenna switch, a radio frequency Balun, a power amplifier, a low noise amplifier, a filter and a power supply management module, the NFC module comprises an NFC read-write chip and an NFC antenna, the NFC antenna is connected with the NFC read-write chip, the NFC read-write chip is connected with the ESP32-S module, the electric energy metering module adopts an HLW8012 chip, pins 2 and pins 3 of the HLW8012 chip are current differential signal input ends, and a pin 4 is a voltage sampling signal input end, pins 6 and 7 of the HLW8012 chip are respectively connected with pins 13 and 16 of an ESP32-S module, and the ESP32-S module calculates power, current and voltage values by measuring the period of high-frequency pulses output by the pins 6 and 7 of the HLW8012 chip;

2. The intelligent disaster prevention and energy saving socket based on the load identification algorithm as claimed in claim 1, wherein: the NFC read-write chip adopts a PN5321A3HN chip, pins 27 and 28 of the PN5321A3HN chip are respectively connected with pins 36 and 33 of an ESP32-S module, and the PN5321A3HN chip is communicated with the ESP32-S module through I2C; the USB serial port adopts a CH340C chip, and pins 2 and 3 of the CH340C chip are respectively connected to pins U0TXD and U0RXD of the ESP32-S module.

3. An intelligent disaster prevention energy-saving socket management system based on a load identification algorithm is characterized in that: the system architecture comprises a data acquisition and calculation architecture, a cloud end and a terminal, wherein the data acquisition and calculation architecture comprises a sensor layer and an edge gateway layer, the cloud end comprises a cloud server and a GPU server, the terminal comprises the intelligent disaster prevention energy-saving socket and a user terminal according to claim 1 or 2, and a user interacts with the sensor layer through the user terminal;