阅读说明：本技术 一种基于人工智能的能源监管和预测系统及其方法 (Energy supervision and prediction system and method based on artificial intelligence ) 是由 朱海 徐恒舟 许蒙蒙 宋炯炯 王祺 王洪峰 张少辉 于 2019-09-20 设计创作，主要内容包括：本发明公开的属于能源监管系统技术领域,具体为一种基于人工智能的能源监管和预测系统及其监管和预测方法,其包括：数据采集单元、智能监控管理中心、能源监测单元和服务器单元,所述数据采集单元电性输入连接所述智能监控管理中心,所述智能监控管理中心电性输出连接所述能源监测单元,所述智能监控管理中心电性双向连接所述服务器单元,所述服务器单元通过无线双向连接移动终端,该发明实现了智能的数据监测和数据的共享,提高工作效率,及时了解异常情况的综合效果。同时,本发明基于能源消耗预测,为用户提供合理的能源消耗方案。(The invention discloses an energy supervision and prediction system based on artificial intelligence and a supervision and prediction method thereof, belonging to the technical field of energy supervision systems, and comprising the following steps: the intelligent monitoring system comprises a data acquisition unit, an intelligent monitoring management center, an energy monitoring unit and a server unit, wherein the data acquisition unit is electrically connected with the intelligent monitoring management center in an input mode, the intelligent monitoring management center is electrically connected with the energy monitoring unit in an output mode, the intelligent monitoring management center is electrically connected with the server unit in a two-way mode, and the server unit is connected with a mobile terminal in a two-way mode through wireless. Meanwhile, the invention provides a reasonable energy consumption scheme for the user based on the energy consumption prediction.)

1. An artificial intelligence based energy management and prediction system, comprising: the intelligent energy monitoring system comprises a data acquisition unit (100), an intelligent monitoring management center (200), an energy monitoring unit (300) and a server unit (400), wherein the data acquisition unit (100) is electrically connected with the intelligent monitoring management center (200) in an input mode, the intelligent monitoring management center (200) is electrically connected with the energy monitoring unit (300) in an output mode, the intelligent monitoring management center (200) is electrically connected with the server unit (400) in a bidirectional mode, and the server unit (400) is connected with a mobile terminal in a wireless bidirectional mode;

the data acquisition unit (100) comprises a water consumption management module (110), an electricity meter management module (120), a gas management module (130) and a combustible gas detector (140), wherein the water consumption management module (110), the electricity meter management module (120), the gas management module (130) and the combustible gas detector (140) are all electrically connected with the intelligent monitoring management center (200) in an output mode;

the intelligent monitoring management center (200) comprises a central control module (210), a data storage module (220), an analog-to-digital converter (230), a power supply module (240) and an ALU module (250), wherein the water consumption management module (110), the electricity meter management module (120) and the gas management module (130) are all electrically connected with the analog-to-digital converter (230) in an output and output mode, the analog-to-digital converter (230) is electrically connected with the central control module (210), the central control module (210) is electrically connected with the power supply module (240) in an input and output mode, and the central control module (210) is electrically connected with the data storage module (220) and the ALU module (250) in a bidirectional mode;

the environment monitoring system is used for collecting environmental parameters in a past period of time and monitoring the current environmental parameters in real time; the intelligent monitoring management center (200) predicts the energy load of a future period of time according to the predicted environmental parameter information of the future period of time, and then performs energy management control according to the predicted energy load of the future period of time.

2. The system of claim 1, wherein the intelligent monitoring and management center (200) collects the outdoor environmental parameter information such as indoor or outdoor temperature, humidity, wind power, sunlight intensity, etc. in real time through the environmental monitoring system to adjust the energy load in real time.

3. An artificial intelligence based energy management and prediction system according to claim 1 or 2, it is characterized in that the deep features of the energy use statistical data are obtained by collecting the energy use comprehensive statistical data of the current user and utilizing the deep learning convolutional neural network, dynamically acquiring indoor or outdoor temperature, humidity, wind power, sunlight intensity and other environmental parameter information according to indoor and outdoor sensors connected with an environmental monitoring system, inputting the acquired indoor and outdoor environmental conditions of each user and the energy use comprehensive statistical data of each stage of the user into a designed neural network for associated training, and searching the optimal comprehensive statistical data of the user energy use by using the twin neural network, forming a statistical table by using the relevant parameters of the comprehensive statistical data of the user energy use in a daily average or monthly average form, and recommending the statistical table to other users.

4. An artificial intelligence based energy supervision and prediction system according to claim 1 or 3 characterized in that the energy monitoring unit (300) comprises an alarm module (310), a printer (320) and a computer (330), the central control module (210) is electrically connected with the alarm module (310), the printer (320) and the computer (330) in output.

5. The system according to claim 1 or 2, wherein the server unit (400) comprises a short message server (410), a database server (420), a web server (430), and a GPRS mobile network (440), the central control module (210) is electrically connected to the short message server (410), the database server (420), and the web server (430), the short message server (410) and the web server (430) are both electrically connected to the GPRS mobile network (440), the GPRS mobile network (440) is electrically connected to the wireless transceiver module (450), and the wireless transceiver module (450) is wirelessly connected to the mobile terminal.

6. The artificial intelligence based energy management and prediction method performed by the system according to any one of claims 1-5, wherein the energy management method of the internet of things and cloud computing comprises the following steps:

the method comprises the following steps: data acquisition: firstly, monitoring the water consumption condition by using a water consumption management module (110), monitoring the power consumption condition by using an electricity meter management module (120), monitoring the gas consumption condition by using a gas management module (130), realizing the centralized collection of various data, and simultaneously installing a combustible gas detector (140) in a gas use place;

step two: monitoring and managing: transmitting the data acquired in the step one to an intelligent monitoring management center (200), converting the data by using an analog-to-digital converter (230) and transmitting the data to a central control module (210), then temporarily storing the data by using a data storage module (220), and executing logic operation by using an ALU (250) to calculate the related cost of water, electricity and gas;

step three: energy monitoring: the central control module (210) is used for transmitting the monitored data to the computer (330), when monitoring output faults, the alarm module (310) can be used for alarming, and abnormal data can be printed out through the printer (320);

step four: server management: the short message server (410) is connected with the GPRS mobile network (440) to achieve the purpose of sending short message energy consumption and related expense reminding to the user, meanwhile, the network server (430) is connected with the GPRS mobile network (440) and wirelessly connected with the remote mobile terminal through the wireless transceiving module (450) to transmit the monitoring data condition to the remote terminal.

7. The method according to claim 6, wherein the prediction is made for user energy usage, in particular comprising,

step 1, collecting user energy consumption data;

step 2, accumulating the user energy consumption data and the environmental parameter information as data samples of the training network model, and storing the optimal user energy consumption data and the optimal environmental parameter information in a database so as to search for reference;

step 3, training the convolutional neural network by the collected user energy consumption data to obtain the characteristics of the user energy consumption, and putting the characteristics trained by the convolutional neural network and the environmental parameter information corresponding to each point into an LSTM recurrent neural network as input data for training to obtain the energy consumption vector of the user;

step 4, putting the acquired new user energy consumption data characteristics and the user energy consumption data of the existing database into a twin network model for model training, and searching the user energy consumption data with the highest similarity;

and 5, searching the user energy consumption data with the highest similarity in the data, and then outputting the data.

8. The method of claim 7,

in the step 1, counting and recording time average, day average and month average of user energy consumption as training data samples;

in step 2, at the time of counting the energy consumption of the user, the environmental monitoring system records parameters of sensors related to the environment, and collecting environmental parameters includes: indoor or outdoor temperature, humidity, wind power and sunlight intensity, and recording the somatosensory state and user experience of a user at any time as the energy consumption benefit of the user; expressing by using one vector to be called a user energy consumption benefit vector; each value of the vector is represented by using an interval of 0-1, the larger the value is, the more ideal the benefit of the energy consumption of the user is represented, and the environment condition vector sequence is defined as the optimal user energy consumption scheme of the user;

in step 2, the data sample comprises three types of data, wherein the first type is comprehensive statistics of user energy consumption data, the second type is an environmental parameter vector, the third type is a benefit vector of user energy consumption, the third type is used for predicting user energy consumption benefits under corresponding schemes, and user energy consumption habits, corresponding optimal energy consumption schemes and user energy consumption benefit vector data are stored in a database; the database is constantly updated as the practice of consumption proceeds.

9. The method of claim 8,

in step 3, training the convolutional neural network with the collected user energy consumption comprehensive statistical data to obtain the characteristics of the user energy consumption by putting the characteristics trained by the convolutional neural network and the environmental parameter vector corresponding to each point as input data into an LSTM recurrent neural network for training, wherein the training result label is the optimal benefit vector of the user energy consumption;

in step 3, dividing the user energy consumption comprehensive statistical data into a training data set and a testing data set, training a convolutional neural network by using the training data, and then extracting features of the data, wherein the features are features of convolutional operation, the convolutional neural network is a VGGNet deep convolutional neural network, and comprises 5 sections of convolutions, 2-3 convolutional layers are arranged in each section, meanwhile, the tail of each section is connected with a maximum pooling layer for reducing the data size, the number of convolutional kernels in each section is the same, and the number of convolutional kernels of the sections which are positioned later is more: 64-128-256-512-two 3-by-3 convolutional layers connected in series are equivalent to 1 5-by-5 convolutional layer, the size of the sensing field is considered to be 5-by-5, the sensing field is the effective area of the convolution operation, the series effect of the 3-by-3 convolutional layers is equivalent to 1 7-by-7 convolutional layer, and the 3-series 3-by-3 convolutional layers have less parameter quantity than 1 7-by-7 convolutional layer;

in step 4, the method for obtaining the comprehensive statistical data of the energy consumption characteristics of the current user and the energy consumption of the existing users is put into a twin network model for model training, and the habitual behavior with the highest similarity is found by the method comprising the steps of providing a mapping function Gw (X) of a given set of data, wherein the parameter is W, and aiming at finding a set of parameters W, so that when X1 and X2 belong to the same class of data, X1 and X2 are the energy consumption characteristics of the new user and the energy consumption data of the existing users, the similarity EW (X1, X2) is a smaller value, when X1 and X2 belong to different classes of data, the similarity measure EW (X1, X2) is the highest, training is carried out by using paired samples in a training set, when X1 and X2 are from the same class, the loss functions EW (X1, X2) are minimized, when X1 and X2 are from different classes, maximizing Ew (X1, X2), two identical functions G, having one identical parameter W, namely structural symmetry;

in step 5, the user energy consumption data with the highest similarity is obtained by putting the data of the best user energy consumption and the previous user energy consumption data into a twin neural network, comparing the similarity to obtain the most similar user energy consumption data, then decomposing the user energy consumption by using the user energy consumption scheme, mapping the user energy consumption data into a target space through inputting, comparing the similarity, namely the Euclidean distance, in the target space by using a simple distance, minimizing the loss function value of a pair of samples from the same category in a training stage, and maximizing the loss function value of a stack of samples from different categories.

10. The method of claim 9, wherein through the prediction process, the energy consumption level prediction for the user is predicted for several days or even a long period of time; meanwhile, an optimal energy consumption mode can be recommended to the user based on the training result so as to guide the user to reasonably consume energy; and prompting whether the indoor environment needs to be adjusted currently or not to the user based on the optimal energy consumption mode according to the current real-time environment condition.

Technical Field

The invention relates to the technical field of energy supervision systems, in particular to an energy supervision and prediction system and method based on artificial intelligence.

Background

The cloud computing technology, the big data analysis and the Internet of things technology are proposed and developed rapidly, and a brand new thought is provided for solving the problems of the energy management system. In summary, cloud computing is a public facility that is based on information technology support capabilities (hardware and software), is extremely flexible (add and drop), and is provided to customers in the form of services over a network. The mode realizes high concentration of computing resources (mainly comprising three resources of a server, a storage and a network), namely, the computing resources distributed in various places are integrated into a virtual uniform resource, and the purpose of acquiring according to needs and paying according to quantity is realized, so that the mode is convenient as electricity and tap water. Therefore, the cloud computing era will drastically change the ecological environment of the entire information technology industry, and great changes will occur to infrastructure, computer and network device manufacturing, and software.

"big data" is a particularly large number of data sets with particularly large data categories. Big data technology refers to technology that quickly obtains valuable information from a large amount of data of various types. The term "big data" as used herein refers not only to the size of the data itself, but also includes tools, platforms, and data analysis systems for collecting the data. The purpose of big data research and development is to develop big data technology and apply the big data technology to related fields, and breakthrough development of the big data technology is promoted by solving the problem of huge data processing.

Disclosure of Invention

This section is for the purpose of summarizing some aspects of embodiments of the invention and to briefly introduce some preferred embodiments. In this section, as well as in the abstract and the title of the invention of this application, simplifications or omissions may be made to avoid obscuring the purpose of the section, the abstract and the title, and such simplifications or omissions are not intended to limit the scope of the invention.

The invention is provided in view of the above and/or the problems existing in the energy supervision system and energy use of the existing internet of things and cloud computing.

Therefore, the invention aims to provide an energy monitoring system of internet of things and cloud computing and a monitoring method thereof, which can realize intelligent data monitoring and data sharing, improve the working efficiency and know abnormal conditions in time. Meanwhile, the invention provides a reasonable energy consumption scheme for the user based on the energy consumption prediction.

To solve the above technical problem, according to an aspect of the present invention, the present invention provides the following technical solutions:

an artificial intelligence based energy management and prediction system, comprising: the intelligent energy monitoring system comprises a data acquisition unit (100), an intelligent monitoring management center (200), an energy monitoring unit (300) and a server unit (400), wherein the data acquisition unit (100) is electrically connected with the intelligent monitoring management center (200) in an input mode, the intelligent monitoring management center (200) is electrically connected with the energy monitoring unit (300) in an output mode, the intelligent monitoring management center (200) is electrically connected with the server unit (400) in a bidirectional mode, and the server unit (400) is connected with a mobile terminal in a wireless bidirectional mode;

the data acquisition unit (100) comprises a water consumption management module (110), an electricity meter management module (120), a gas management module (130) and a combustible gas detector (140), wherein the water consumption management module (110), the electricity meter management module (120), the gas management module (130) and the combustible gas detector (140) are all electrically connected with the intelligent monitoring management center (200) in an output mode;

the intelligent monitoring management center (200) comprises a central control module (210), a data storage module (220), an analog-to-digital converter (230), a power supply module (240) and an ALU module (250), wherein the water consumption management module (110), the electricity meter management module (120) and the gas management module (130) are all electrically connected with the analog-to-digital converter (230) in an output and output mode, the analog-to-digital converter (230) is electrically connected with the central control module (210), the central control module (210) is electrically connected with the power supply module (240) in an input and output mode, and the central control module (210) is electrically connected with the data storage module (220) and the ALU module (250) in a bidirectional mode;

the environment monitoring system is used for collecting environmental parameters in a past period of time and monitoring the current environmental parameters in real time;

the intelligent monitoring management center (200) predicts the energy load of a future period of time according to the predicted environmental parameter information of the future period of time, and then performs energy management control according to the predicted energy load of the future period of time.

Furthermore, the intelligent monitoring management center (200) collects outdoor environment parameter information such as indoor or outdoor temperature, humidity, wind power and sunlight intensity in real time through an environment monitoring system, and adjusts energy load in real time.

Furthermore, the depth characteristics of the energy use statistical data are obtained by collecting the energy use comprehensive statistical data of the current user and utilizing a deep learning convolutional neural network, the collected indoor and outdoor environmental conditions of each user and the energy use comprehensive statistical data of each stage of the user are input into a designed neural network for correlation training according to indoor and outdoor sensors connected with an environment monitoring system, the optimal user energy use comprehensive statistical data are searched by using a twin neural network, and the relevant parameters of the user energy use comprehensive statistical data form a statistical table in a day-average or month-average mode to be recommended to other users.

Further, the energy monitoring unit (300) comprises an alarm module (310), a printer (320) and a computer (330), and the central control module (210) is electrically connected with the alarm module (310), the printer (320) and the computer (330).

Further, the server unit (400) includes a short message server (410), a database server (420), a web server (430), and a GPRS mobile network (440), the central control module (210) is electrically connected to the short message server (410), the database server (420), and the web server (430), the short message server (410) and the web server (430) are both electrically connected to the GPRS mobile network (440) in a bidirectional manner, the GPRS mobile network (440) is electrically connected to the wireless transceiver module (450) in a bidirectional manner, and the wireless transceiver module (450) is wirelessly connected to the mobile terminal.

Further, the energy supervision method for the Internet of things and cloud computing comprises the following steps:

the method comprises the following steps: data acquisition: firstly, monitoring the water consumption condition by using a water consumption management module (110), monitoring the power consumption condition by using an electricity meter management module (120), monitoring the gas consumption condition by using a gas management module (130), realizing the centralized collection of various data, and simultaneously installing a combustible gas detector (140) in a gas use place;

step two: monitoring and managing: transmitting the data acquired in the step one to an intelligent monitoring management center (200), converting the data by using an analog-to-digital converter (230) and transmitting the data to a central control module (210), then temporarily storing the data by using a data storage module (220), and executing logic operation by using an ALU (250) to calculate the related cost of water, electricity and gas;

step three: energy monitoring: the central control module (210) is used for transmitting the monitored data to the computer (330), when monitoring output faults, the alarm module (310) can be used for alarming, and abnormal data can be printed out through the printer (320);

step four: server management: the short message server (410) is connected with the GPRS mobile network (440) to achieve the purpose of sending short message energy consumption and related expense reminding to the user, meanwhile, the network server (430) is connected with the GPRS mobile network (440) and wirelessly connected with the remote mobile terminal through the wireless transceiving module (450) to transmit the monitoring data condition to the remote terminal.

Further, the prediction is made with respect to energy usage by the user, specifically including,

step 1, collecting user energy consumption data;

step 2, accumulating the user energy consumption data and the environmental parameter information as data samples of the training network model, and storing the optimal user energy consumption data and the optimal environmental parameter information in a database so as to search for reference;

step 3, training the convolutional neural network by the collected user energy consumption data to obtain the characteristics of the user energy consumption, and putting the characteristics trained by the convolutional neural network and the environmental parameter information corresponding to each point into an LSTM recurrent neural network as input data for training to obtain the energy consumption vector of the user;

step 4, putting the acquired new user energy consumption data characteristics and the user energy consumption data of the existing database into a twin network model for model training, and searching the user energy consumption data with the highest similarity;

and 5, searching the user energy consumption data with the highest similarity in the data, and then outputting the data.

Further, the air conditioner is provided with a fan,

in the step 1, counting and recording time average, day average and month average of user energy consumption as training data samples;

in step 2, at the time of counting the energy consumption of the user, the environmental monitoring system records parameters of sensors related to the environment, and collecting environmental parameters includes: indoor or outdoor temperature, humidity, wind power and sunlight intensity, and recording the somatosensory state and user experience of a user at any time as the energy consumption benefit of the user; expressing by using one vector to be called a user energy consumption benefit vector; each value of the vector is represented by using an interval of 0-1, the larger the value is, the more ideal the benefit of the energy consumption of the user is represented, and the environment condition vector sequence is defined as the optimal user energy consumption scheme of the user;

in step 2, the data sample comprises three types of data, wherein the first type is comprehensive statistics of user energy consumption data, the second type is an environmental parameter vector, the third type is a benefit vector of user energy consumption, the third type is used for predicting user energy consumption benefits under corresponding schemes, and user energy consumption habits, corresponding optimal energy consumption schemes and user energy consumption benefit vector data are stored in a database; the database is constantly updated as the practice of consumption proceeds.

Further, the air conditioner is provided with a fan,

in step 3, training the convolutional neural network with the collected user energy consumption comprehensive statistical data to obtain the characteristics of the user energy consumption by putting the characteristics trained by the convolutional neural network and the environmental parameter vector corresponding to each point as input data into an LSTM recurrent neural network for training, wherein the training result label is the optimal benefit vector of the user energy consumption;

in step 3, dividing the user energy consumption comprehensive statistical data into a training data set and a testing data set, training a convolutional neural network by using the training data, and then extracting features of the data, wherein the features are features of convolutional operation, the convolutional neural network is a VGGNet deep convolutional neural network, and comprises 5 sections of convolutions, 2-3 convolutional layers are arranged in each section, meanwhile, the tail of each section is connected with a maximum pooling layer for reducing the data size, the number of convolutional kernels in each section is the same, and the number of convolutional kernels of the sections which are positioned later is more: 64-128-256-512-two 3-by-3 convolutional layers connected in series are equivalent to 1 5-by-5 convolutional layer, the size of the sensing field is considered to be 5-by-5, the sensing field is the effective area of the convolution operation, the series effect of the 3-by-3 convolutional layers is equivalent to 1 7-by-7 convolutional layer, and the 3-series 3-by-3 convolutional layers have less parameter quantity than 1 7-by-7 convolutional layer;

in step 4, the method for obtaining the comprehensive statistical data of the energy consumption characteristics of the current user and the energy consumption of the existing users is put into a twin network model for model training, and the habitual behavior with the highest similarity is found by the method comprising the steps of providing a mapping function Gw (X) of a given set of data, wherein the parameter is W, and aiming at finding a set of parameters W, so that when X1 and X2 belong to the same class of data, X1 and X2 are the energy consumption characteristics of the new user and the energy consumption data of the existing users, the similarity EW (X1, X2) is a smaller value, when X1 and X2 belong to different classes of data, the similarity measure EW (X1, X2) is the highest, training is carried out by using paired samples in a training set, when X1 and X2 are from the same class, the loss functions EW (X1, X2) are minimized, when X1 and X2 are from different classes, maximizing Ew (X1, X2), two identical functions G, having one identical parameter W, namely structural symmetry;

in step 5, the user energy consumption data with the highest similarity is obtained by putting the data of the best user energy consumption and the previous user energy consumption data into a twin neural network, comparing the similarity to obtain the most similar user energy consumption data, then decomposing the user energy consumption by using the user energy consumption scheme, mapping the user energy consumption data into a target space through inputting, comparing the similarity, namely the Euclidean distance, in the target space by using a simple distance, minimizing the loss function value of a pair of samples from the same category in a training stage, and maximizing the loss function value of a stack of samples from different categories.

Furthermore, through the prediction process, the energy consumption level prediction in several days or even a long period of time in the future is predicted for the user; meanwhile, an optimal energy consumption mode can be recommended to the user based on the training result so as to guide the user to reasonably consume energy; and prompting whether the indoor environment needs to be adjusted currently or not to the user based on the optimal energy consumption mode according to the current real-time environment condition.

Compared with the prior art: the intelligent meter reading is realized by the data acquisition unit, the working efficiency is improved, the data sharing is realized by the server unit, the short message server is used for realizing the timely consumption condition of the client energy meter, and secondly, when the abnormal condition occurs, the alarm module on the energy monitoring unit can be used for realizing the timely alarm and the data can be printed out timely by the printer. Finally, the optimal energy consumption scheme can be provided for the user according to the current environmental condition.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the present invention will be described in detail with reference to the accompanying drawings and detailed embodiments, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings without inventive exercise. Wherein:

FIG. 1 is a system block diagram of a first embodiment of the present invention;

FIG. 2 is a system block diagram of a second embodiment of the present invention.

Detailed Description

In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in detail below.

In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention, but the present invention may be practiced in other ways than those specifically described herein, and it will be apparent to those of ordinary skill in the art that the present invention may be practiced without departing from the spirit and scope of the present invention, and therefore the present invention is not limited to the specific embodiments disclosed below.

Next, the present invention will be described in detail with reference to the drawings, wherein for convenience of illustration, the cross-sectional view of the device structure is not enlarged partially according to the general scale, and the drawings are only examples, which should not limit the scope of the present invention. In addition, the three-dimensional dimensions of length, width and depth should be included in the actual fabrication.

In order to make the objects, technical solutions and advantages of the present invention more apparent, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

The invention provides an energy supervision system for internet of things and cloud computing, which is used for realizing intelligent data monitoring and data sharing, improving working efficiency and timely knowing about abnormal conditions, and please refer to fig. 1, and the energy supervision system comprises: the system comprises a data acquisition unit 100, an intelligent monitoring management center 200, an energy monitoring unit 300 and a server unit 400;

referring to fig. 1 again, the data acquisition unit 100 is electrically connected to the intelligent monitoring management center 200, the data acquisition unit 100 includes a water consumption management module 110, an electricity meter management module 120, a gas management module 130 and a combustible gas detector 140, the water consumption management module 110, the electricity meter management module 120, the gas management module 130 and the combustible gas detector 140 are all electrically connected to the intelligent monitoring management center 200, the water consumption management module 110 is a water control management system, the electricity meter management module 120 is an intelligent remote electricity meter, the gas management module 130 is a gas intelligent monitoring system, the intelligent monitoring management center 200 includes a central control module 210, a data storage module 220, an analog-to-digital converter 230, a power supply module 240 and an ALU module 250, the water consumption management module 110, the electricity meter management module 120 and the gas management module 130 are all electrically connected to the analog-to-digital converter 230, the analog-to-digital converter 230 is electrically connected to the central control module 210, the central control module 210 is electrically connected to the power module 240, the central control module 210 is electrically connected to the data storage module 220 and the ALU module 250 in a bidirectional manner, the intelligent monitoring management center 200 is electrically connected to the energy monitoring unit 300 in an output manner, the energy monitoring unit 300 includes an alarm module 310, a printer 320 and a computer 330, the central control module 210 is electrically connected to the alarm module 310, the printer 320 and the computer 330 in an output manner, the alarm module 310 is an audible and visual alarm, the printer 320 is a thermal printer, the intelligent monitoring management center 200 is electrically connected to the server unit 400 in a bidirectional manner, the server unit 400 is connected to the mobile terminal in a wireless bidirectional manner, and the server unit 400 includes a short message server 410, a short message server, The mobile terminal comprises a database server 420, a network server 430 and a GPRS mobile network 440, wherein the central control module 210 is electrically connected with the short message server 410, the database server 420 and the network server 430 in an output mode, the short message server 410 and the network server 430 are both electrically and bidirectionally connected with the GPRS mobile network 440, the GPRS mobile network 440 is electrically and bidirectionally connected with the wireless transceiver module 450, the wireless transceiver module 450 is wirelessly connected with the mobile terminal, and the mobile terminal comprises a computer and a mobile phone.

The water consumption management module 110, the electricity meter management module 120 and the gas management module 130 are not limited to the specific quantities described in the embodiment, and those skilled in the art can increase or decrease the quantities as required on the premise that the device can complete its supervision function.

An energy supervision method for Internet of things and cloud computing comprises the following steps:

the method comprises the following steps: data acquisition: firstly, the water consumption condition is monitored by using the water consumption management module 110, the power consumption condition is monitored by using the meter management module 120, and the gas consumption condition is monitored by using the gas management module 130, so that the centralized collection of various data is realized;

step two: monitoring and managing: transmitting the data acquired in the step one to an intelligent monitoring management center 200, converting the data by using an analog-to-digital converter 230 and transmitting the data to a central control module 210, then temporarily storing the data by using a data storage module 220, and performing logic operation by using an ALU module 250 to calculate the related cost of water, electricity and gas;

step three: energy monitoring: the central control module 210 is used for transmitting the monitored data to the computer 330, when monitoring and outputting faults, the alarm module 310 can be used for alarming, and abnormal data can be printed out through the printer 320;

step four: server management: the short message server 410 is connected with the GPRS mobile network 440 to send the short message energy consumption and the related fee reminding to the user, and the network server 430 is connected with the GPRS mobile network 440 and wirelessly connected with the remote mobile terminal through the wireless transceiving module 450 to transmit the monitoring data condition to the remote terminal.

Furthermore, in the technical scheme of the invention, an environment monitoring system is added, the energy load of a future period of time is predicted by collecting environmental parameters such as temperature and humidity of a past period of time and according to the predicted information of the environmental parameters of the future period of time, and then the energy management control is carried out according to the predicted energy load of the future period of time.

Furthermore, the system can also simply and conveniently adjust the energy load in real time by acquiring the indoor or outdoor environment parameter information such as indoor or outdoor temperature, humidity, wind power, sunlight intensity and the like in real time, thereby simplifying the complexity of energy management control and reducing the cost of the energy management control system.

Specifically, the depth characteristics of the energy use statistical data are acquired by collecting the energy use comprehensive statistical data of the current user and utilizing a deep learning convolutional neural network, the collected indoor and outdoor environmental conditions of each user and the energy use comprehensive statistical data of each stage of the user are input into a designed neural network for correlation training according to indoor and outdoor sensors connected with an environment monitoring system and environmental parameter information such as indoor or outdoor temperature, humidity, wind power and sunlight intensity, the optimal user energy use comprehensive statistical data are searched by using a twin neural network, and relevant parameters of the user energy use comprehensive statistical data form a statistical table in a day-to-month or month-to-month mode and are recommended to other users.

To achieve the above object, according to an aspect of the present disclosure, there is provided a user energy usage prediction method including the steps of:

step 1, collecting user energy consumption data;

step 2, accumulating the user energy consumption data and the environmental parameter information as data samples of the training network model, and storing the optimal user energy consumption data and the optimal environmental parameter information in a database so as to search for reference;

step 3, training the convolutional neural network by the collected user energy consumption data to obtain the characteristics of the user energy consumption, and putting the characteristics trained by the convolutional neural network and the environmental parameter information corresponding to each point into an LSTM recurrent neural network as input data for training to obtain the energy consumption vector of the user;

step 4, putting the acquired new user energy consumption data characteristics and the user energy consumption data of the existing database into a twin network model for model training, and searching the user energy consumption data with the highest similarity;

and 5, searching the user energy consumption data with the highest similarity in the data, and then outputting the data.

Further, in step 1, statistical records are recorded for the user energy consumption in time average, day average and month average, and the statistical records are used as training data samples.

Further, in step 2, at the time of counting the energy consumption of the user, the environmental monitoring system records parameters of sensors related to the environment, and collecting the environmental parameters includes: indoor or outdoor temperature, humidity, wind power, sunlight intensity to record the somatosensory state of the user, user experience and the like at any time as user energy consumption benefits. The expression using one vector is referred to herein as a user energy consumption benefit vector. This vector is used to express the somatosensory state of the user at the current energy consumption. Each value of the vector is represented by an interval of 0-1, the larger the value is, the more ideal the benefit of the energy consumption of the user is, and the sequence of the environmental condition vectors is defined as the optimal energy consumption scheme of the user.

Further, in step 2, the data sample includes three types of data, the first type is comprehensive statistics of the user energy consumption data, the second type is an environmental parameter vector, the third type is a benefit vector of the user energy consumption, the third type is used for predicting the user energy consumption benefit under a corresponding scheme, and the data such as the user energy consumption habit, the corresponding optimal energy consumption scheme, the user energy consumption benefit vector and the like are stored in the database. This database is constantly updated as the practice of consumption proceeds.

Further, in step 3, the method for training the convolutional neural network to obtain the features of the user energy consumption by using the collected user energy consumption comprehensive statistical data is that the features trained by the convolutional neural network and the environmental parameter vector corresponding to each point are used as input data and put into the LSTM recurrent neural network for training, and the training result label is the optimal benefit vector of the user energy consumption.

Further, in step 3, dividing the user energy consumption comprehensive statistical data into a training data set and a testing data set, training a convolutional neural network by using the training data, and then extracting features of the data, wherein the features are features of convolutional operation, the convolutional neural network is VGGNet (deep convolutional neural network), and comprises 5 sections of convolution, 2-3 convolutional layers are arranged in each section, meanwhile, the tail of each section is connected with a maximum pooling layer for reducing the data size, the number of convolutional cores in each section is the same, and the number of convolutional cores in the later section is more: 64-128-256-512-3 convolutional layers, two 3 × 3 convolutional layers are connected in series to correspond to 1 convolutional layer of 5 × 5, it can be considered that the size of the sensing field is 5 × 5, the sensing field is the effective region of the convolution operation, the effect of the series connection of 3 × 3 convolutional layers is equivalent to 1 convolutional layer of 7 × 7, and the 3 series connected convolutional layers of 3 × 3 have less parameters than 1 convolutional layer of 7 × 7.

Further, in step 4, the method for obtaining the comprehensive statistical data of the energy consumption characteristics of the current user and the energy consumption of the existing users to train the model includes putting the obtained comprehensive statistical data into a twin network model for training, and finding the habitual behavior with the highest similarity includes, for a given set of data, a mapping function gw (X), where the parameter is W, in order to find a set of parameters W, such that when X1 and X2 belong to the same category of data, the X1 and X2 are the data of the energy consumption characteristics of the new user and the energy consumption of the existing users, the similarity Ew (X1, X2) is a smaller value, when X1 and X2 belong to different categories of data, the similarity measure Ew (X1, X2) is the highest, training is performed by using paired samples in a training set, when X1 and X2 belong to the same category, the loss function Ew (X1, X2) is minimized, and when X1 and X2 are from different categories, maximizing Ew (X1, X2), two identical functions G, having one identical parameter W, i.e. structural symmetry.

Further, in step 5, the user energy consumption data with the highest similarity is obtained by putting the data of the best user energy consumption and the previous user energy consumption data into a twin neural network, comparing the similarity to obtain the most similar user energy consumption data, then decomposing the user energy consumption by using the user energy consumption scheme, mapping the user energy consumption data into a target space by inputting, comparing the similarity, namely the Euclidean distance, in the target space by using a simple distance, minimizing the loss function value of a pair of samples from the same category in the training stage, and maximizing the loss function value of a pair of samples from different categories.

Through the prediction process, the energy consumption level prediction in the next several days or even a long period of time is predicted for the user. Meanwhile, the optimal energy consumption mode can be recommended to the user based on the training result so as to guide the user to reasonably consume energy. And whether the indoor environment needs to be adjusted currently or not can be prompted to the user according to the current real-time environment condition based on the optimal energy consumption mode, specifically, for example, the user is prompted to open a window for ventilation, or the user is prompted to increase or decrease the temperature of the air conditioner. The specific reminding mode can be realized by means of short messages or specific APP and the like.

Furthermore, an artificial intelligence learning mode is introduced, and the user can gradually adjust the energy consumption based on the recommended comprehensive statistical data of the energy consumption and personal habits, so that reasonable energy conservation is achieved, and the best user experience can be obtained.

In order to achieve the purpose, after the optimal user energy consumption scheme is recommended to the user, deep learning correction is continuously carried out according to the current energy consumption mode of the user. The specific mode comprises that the first stage is a primary stage of learning, the energy consumption mode strategy level generated by the learning device is far lower than the recommended scheme execution level, the system correction can be continuously prompted only by the user experience, and the learning device continuously learns and optimizes the strategy by learning the current energy consumption data of the user. The second stage is a mature stage of learning, and at this time, the strategy generated by the learning device through continuous learning is close to the experience level adapted by the user at ordinary times. At this stage, the strategy generated by the learning device needs to be evaluated continuously, and when the strategy generated by the learning device reaches the limit level of the energy consumption of the user, the learning of the current energy consumption data of the user can be stopped. The third stage is a continuous optimization stage of learning, at which the learning device has reached or exceeded the current energy consumption level of the user, and the learning device continuously tries to research the influence of different energy consumption combinations on the user experience in different states, so that the strategy can be continuously optimized through continuous trial and learning.

After the deep learning, the energy consumption scheme used by the user is completely matched with the expected use mode of the user, so that the user can save energy consumption and achieve the best physical and mental environment experience. The method has the advantages that the method achieves the maximum sharing for the maintenance and the rise of the life quality of the users.

Furthermore, the monitoring system can return the energy consumption same ratio/ring ratio change condition of the user to the user regularly, so that the user can feel that the user can not save energy any time and any time, and the monitoring system contributes to the environmental protection.

While the invention has been described above with reference to an embodiment, various modifications may be made and equivalents may be substituted for elements thereof without departing from the scope of the invention. In particular, the various features of the disclosed embodiments of the invention may be used in any combination, provided that no structural conflict exists, and the combinations are not exhaustively described in this specification merely for the sake of brevity and resource conservation. Therefore, it is intended that the invention not be limited to the particular embodiments disclosed, but that the invention will include all embodiments falling within the scope of the appended claims.

It will be understood by those skilled in the art that all or part of the steps of the above methods may be implemented by a program instructing associated hardware (e.g., a processor) to perform the steps, and the program may be stored in a computer readable storage medium, such as a read only memory, a magnetic or optical disk, and the like. Alternatively, all or part of the steps of the above embodiments may be implemented using one or more integrated circuits. Accordingly, each module/unit in the above embodiments may be implemented in hardware, for example, by an integrated circuit to implement its corresponding function, or in software, for example, by a processor executing a program/instruction stored in a memory to implement its corresponding function. The present invention is not limited to any specific form of combination of hardware and software.

Although the embodiments disclosed in the present application are described above, the descriptions are only used for facilitating the understanding of the embodiments of the present application, and are not intended to limit the present application, such as the specific implementation methods in the embodiments of the present invention. It will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the disclosure as defined by the appended claims.

While the foregoing is directed to the preferred embodiment of the present invention, it will be understood by those skilled in the art that various changes and modifications may be made without departing from the spirit and scope of the invention as defined in the appended claims.