CN113568948A

CN113568948A - Electric Internet of things system

Info

Publication number: CN113568948A
Application number: CN202110835550.5A
Authority: CN
Inventors: 徐旺; 俞泽文
Original assignee: Anhui Institute of Information Engineering
Current assignee: Anhui Institute of Information Engineering
Priority date: 2021-07-23
Filing date: 2021-07-23
Publication date: 2021-10-29

Abstract

The invention discloses an electric Internet of things system, which comprises an environment module, a switch-type module, an electric appliance parameter detection module, a data mining module and the like, and the system comprises remote control, timing control, intelligent control according to environmental change and the like, and designs and realizes an electric equipment Internet of things database, realizes the storage and management of data, and provides data support for association rule mining based on a large amount of data. According to the user requirements, mining the association rules of the use habits of the personnel and the electrical equipment facing the requirements, and establishing an association rule base for the use of the electrical equipment. The continuously updated association rule base is used for reasoning and judging the operation of the electrical equipment, so that the self-learning and adjustment of the whole Internet of things system are completed, and the optimal control of starting and stopping equipment is realized.

Description

Electric Internet of things system

Technical Field

The invention relates to the field of Internet of things, in particular to an electric Internet of things system.

Background

In China, the consumption of building electrical equipment comprises an air conditioning system, an illumination system, water pump equipment and the like, wherein the energy consumption of the air conditioning system accounts for about 40-50% of the total energy consumption of the building, the energy consumption of the illumination system accounts for about 15-25%, the energy consumption of the water pump equipment and the like accounts for about 10-15%, and the energy consumption of other equipment accounts for about 10-15%. It follows that there is a huge space for energy saving of electrical equipment.

The electric equipment internet of things refers to the fact that building equipment is constructed into an integrated management network through the internet of things technology, remote monitoring of electric equipment is achieved through network monitoring, and comprehensive sharing and coordination control of information and information are achieved. The electrical equipment internet of things brings the building electrical equipment into the network system, so that not only is the environmental data and the equipment acquisition information stored, but also data sharing is realized, and a large amount of reliable and effective data is provided for the energy conservation of the electrical equipment. The structure of the Internet of things of the electrical equipment is divided into three layers, namely a sensing layer, a network layer and an application layer. The intelligent control system has the main functions of realizing control management of electrical equipment, such as on-off control of illumination, energy-saving control of an air conditioner and the like, providing intelligent judgment, electrical parameter detection, remote control, timing control and the like for personnel, and ensuring high efficiency and energy conservation of the electrical equipment. The electrical equipment internet of things system carries out real-time monitoring on equipment and data acquisition and analysis, thereby carrying out overall optimization on energy. The user can carry out remote control, timing control and the like through modes such as internet surfing and the like, and the optimal control of the electrical equipment is realized.

In a large amount of data generated by the electrical equipment internet of things system, various user electricity utilization habits and electricity utilization rules are contained. The system analyzes and processes data by using the data generated by the Internet of things in the database, and performs association rule mining on the illumination system and the air conditioning system by using an association rule mining technology aiming at office power consumption to find different requirements of different users in the data so as to establish a system expert control knowledge base and lay a foundation for acquiring knowledge of a later expert system.

Association rule mining is an important component of data mining, namely finding valuable associations from a large amount of data. In the field of data mining, a typical example of association rule mining is known as shopping basket analysis, which is the discovery of which items may occur simultaneously in a purchase by a customer through shopping basket analysis. The significance of this analysis is that merchants can rationally arrange shelves and conduct promotional activities through this relationship.

However, the existing electrical equipment internet of things does not have the self-mining and updating functions, the operation rules of the existing electrical equipment internet of things depend on manual setting, the existing electrical equipment internet of things are rigid, the existing electrical equipment internet of things cannot adapt to requirements under different use environments, and the existing electrical equipment internet of things cannot learn and mine association rules from use habits of users, so that the existing electrical equipment internet of things cannot be intelligently evolved.

Disclosure of Invention

The technical problems to be solved by the invention are as follows: aiming at the problems existing in the existing electrical equipment Internet of things, an intelligent electrical Internet of things system is provided.

In order to solve the technical problems, the invention provides the following technical scheme:

an electric Internet of things system comprises a sensing layer, a network layer and an application layer;

the sensing layer comprises an environment module, a switch-type module and an electric appliance parameter detection module; the environment module acquires environment parameters through a sensor; the switch-type module comprises a manual control module and an automatic control module;

the network layer comprises an Internet of things node, the Internet of things node is in bidirectional communication with the modules of the sensing layer and the application layer through a wireless network, acquires the state and parameter information of the modules and sends the information to the server;

the application layer comprises an input/output control terminal and electrical equipment, wherein the input/output control terminal is a management computer;

the management computer comprises a management system and a database;

the database updates the information according to unit time according to the environmental information, the personnel information and the equipment information stored in the database, and the updating is based on the manual control behavior of the electrical equipment fed back by the manual control module in the unit time;

the management system takes the B/S system as a frame and realizes the functions of remote monitoring and real-time viewing; the management system comprises a rule mining module and a storage module, wherein the rule mining module is used for mining association rules from information data which is continuously updated in a database and uploading the obtained new association rules to the storage module so as to enrich and perfect the existing association rules;

the management computer carries out logical reasoning according to the association rule in the storage module and then outputs a signal to the switch-type module, and the output signal is converted into the switching action of the electrical equipment through the switch-type module;

preferably, the management system is additionally provided with an interface connected with the 5G module, and a user accesses the internet and downloads the APP through the 5G mobile phone to realize the remote monitoring function of the state of the electric equipment.

Preferably, the input signal of the rule mining module is the incoming/outgoing information, the rule mining module performs reasoning and matching according to the incoming/outgoing information, outputs a device start/stop control signal according to a matching result of the association rule, and the controller outputs a control signal of the electrical device, which acts on the electrical device to perform real-time control.

Preferably, the environmental parameters are indoor illumination intensity and infrared sensing personnel detection; the electrical parameter detection module measures voltage and current parameters of electrical equipment nodes and sends the voltage and current parameters to the management system, so that on-line monitoring of harmonic waves of electrical equipment and power distribution branches in a building is realized;

preferably, the mining process of the association rule is as follows:

(a) calling out sensor data and electrical equipment monitoring data within a specified time range from a database, and preparing the data;

(b) converting the data from numerical data to Boolean data;

(c) and (4) carrying out association rule mining with constraint from the Boolean type data set.

The invention has the following beneficial effects:

the electric equipment internet of things system collects various electric equipment information through the sensor, uploads the environmental information to the management computer through the network and stores the data. And the expert control system continuously updates the association rule according to the expert experience and the inference of historical data, stores the association rule into an expert system knowledge base and finally makes a control decision according to the real-time information. Along with the increase of the running time of the system, the records of the equipment information and the environmental parameters in the database and the personnel operation information are continuously expanded and perfected, and the rule base is continuously updated, so that the accuracy of the control decision of the system is continuously increased along with the time.

Drawings

FIG. 1 is a diagram of a logical inference decision process.

Detailed Description

The following examples are included to provide further detailed description of the present invention and to provide those skilled in the art with a more complete, concise, and exact understanding of the principles and spirit of the invention.

Example 1: the electrical internet of things system is structured as follows:

1. sensing layer

And the perception layer is the basis for realizing the comprehensive perception of the Internet of things of the electrical equipment. Electrical equipment thing networking perception layer divide into a plurality of modules, by sensor acquisition equipment information, environmental information etc. specifically contain:

1.1 Environment Module

The primary function of the environmental module is to collect environmental parameters, such as CO₂Concentration, temperature and humidity, illumination intensity, pyroelectric person detection and the like. The system has extremely strong expandability, smoke sensors, door magnets and the like can be added according to self needs, the environment module is like human 'nerve', building information can be collected and transmitted to the server, and subsequent rule mining is facilitated.

1.2 switch-mode module

A module that controls on-off electrical devices such as a water dispenser, a lamp, etc. is called a switch-type module. The switch type module is provided with a relay and a toggle switch, so that a user can realize field real-time control through manual control with the highest priority besides an automatic control command.

1.3 Electrical appliance parameter detection module

The electrical appliance parameter detection module is mainly used for measuring parameters such as voltage and current of electrical equipment nodes, and can send various parameters to the management system, so that on-line monitoring of harmonic waves of electrical equipment and power distribution branches in a building is realized, visual management control is realized, overrun alarm and graph real-time monitoring are realized, and data are visually and conveniently displayed to a user.

2. Network layer

And the network layer is network equipment and a platform which are used for information aggregation, transmission and primary processing in the Internet of things of the electrical equipment.

The main hardware equipment of the network layer is an internet of things node. The node of the Internet of things is in bidirectional communication with each module through the wireless network, so that the state and parameter information of each module can be obtained, and the information is sent to the server. When the power-on is started, the system is initialized firstly, and then a wireless network is established, wherein the wireless network allows a new node to join. And when a new node is added, carrying out address classification according to the node type and registering. And the second is responsible for uploading and issuing the information. It is located at the site of the master node in the wireless network in the room and can communicate with multiple end nodes, being the information hub of the whole system.

3. Application layer

The method mainly solves the problems of information processing and human-computer interface, namely, input and output control terminals such as mobile phones, computers and the like. The client is in direct contact with the application layer, and rich service functions can be provided. The management computer comprises two parts as main hardware, namely application software and a database, and the management computer is used for carrying out optimized control on the electrical equipment by utilizing environmental information, personnel information, equipment information and the like stored in the database. The management system of the Internet of things of the electrical equipment takes the B/S system as a framework, and realizes the functions of remote monitoring and real-time viewing. The system is additionally provided with an interface connected with modules such as a 5G module, and a user can realize the remote monitoring function of each state of the electric equipment by surfing the Internet and downloading the APP through a mobile phone.

Electrical equipment Internet of things data preparation

In order to obtain a data subset which is high in quality and suitable for mining, preparation work such as analyzing and selecting the data is needed, namely, a part of useful data is extracted from original data according to the requirement of a user. The data selection functions to focus subsequent mining work into a subset of data associated with the mining objective. Data selection not only ensures the efficiency of data mining, but also improves the accuracy of data mining. Through the analysis of the demand target, the interested data is selected from the database, the attribute to be analyzed and the parameter thereof are extracted, and then the extracted data is matched according to the time attribute and is stored in a new table.

For a lighting system, based on an analysis of a person's lighting needs, the need attributes of the person's operation of the lighting fixtures are analyzed with the person's on-off operation as a target. The operation of lighting fixtures by office personnel is largely dependent on time and variations in outdoor light levels. Therefore, in the data selection, the time, the illumination and the operation items are selected from the item tables of the database and stored in the new table. The following table shows some operating events of the daylighting system for a room time of 11 months and 20 months.

Table 1 Lighting System operating items table

ID	Time	Operation object	Illuminance of light	Operating items
					1	7:30	12	6782	Is opened
2	10:40	13	10923	Close off
					3	15:10	12	7673	Is opened
4	18:10	10	4572	Close off
					……	……	……	……	……

Electrical equipment thing networking data conversion

The Boolean association rule and the numerical association rule are divided according to different variable types, wherein the Boolean association rule mainly aims at Boolean data in a database, the numerical management rule needs to process the numerical data, the data of continuous and multi-valued attributes are mapped into Boolean attribute data, and the Boolean association rule mining algorithm can be used for solving the problem of the numerical association rule.

The minimum support degree and the minimum confidence degree are the core of the numerical association rule, and the determination steps are as follows:

(1) determining a numerical attribute dividing method:

(2) mapping the numerical attribute to a Boolean attribute;

(3) and finding out a frequent item set to obtain a fourth rule of the minimum confidence degree association strength.

The key of the numerical attribute association rule mining problem lies in the mapping from the numerical attributes to the corresponding Boolean attributes, and for the problem of the mapping from the numerical attributes to the Boolean attributes, the method comprises the following steps:

(1) according to the numerical attributes or the category attributes with few categories, each attribute can be mapped into a Boolean attribute;

(2) for the numerical attribute with more values, dividing the numerical value into a plurality of subintervals, and mapping each interval into a Boolean attribute.

There are two problems in the process of mapping the numerical attribute to the Boolean attribute, namely, the problem of support degree and the problem of confidence degree.

(1) The support degree problem refers to that the support degree of the region is reduced when the value range is divided into too many regions, and some attribute-containing frequent sets can be ignored;

(2) the confidence problem corresponds to the support problem, which means that the value range division interval is too small, so that the value of a single interval is too much, and the forward association rule is ignored due to the confidence condition.

Through the analysis, the confidence coefficient and the support degree are a pair of relatively contradictory factors, and a compromise processing method is adopted under the normal condition to ensure that the two mining tasks can meet the requirements. In this embodiment, the data involved is mostly of the numeric type. And for the time attribute, performing subinterval division on the time attribute by adopting predefined concept hierarchy according to the demand analysis of users and expert experience. For attributes such as temperature, humidity and illuminance, a uniform-width interval method is adopted to divide sub-intervals of numerical attributes.

In a lighting system, a subinterval partitioning method is employed, where Record ID is a transaction number; the Time attribute is a numeric attribute with a large value. The operation peak of the lighting equipment by the personnel is judged according to experience and the electricity utilization habits of the office personnel and is usually about the time period of going to work and going to work, so that the time division of the time period of going to work and going to work is more intensive. According to the morning, noon and evening work-off schedule input by the system personnel of the Internet of things of the electrical equipment, taking the work-off time point as a reference, dividing the time into the following time periods A1, wherein the time is 15min before and after the work-on time point; a2, keeping the distance from the working time point for 30-16 min; a3, within 15min before and after the time point of next shift; a4, within 16-30 min before and after the time point of next work; a5, other times. Here, since the period from the end of the first day a4 to the next day a1 is nighttime, the office worker has fewer operations on the electrical equipment and is not targeted for data mining. Outdoor illuminance also belongs to numerical data, and Xmin and Xmax represent the maximum and minimum values of the data, respectively. The operation events are divided into five sections B1-B5 by using an equal-width section method, and the attributes of the operation events are of a type and can be expressed as Cl [ Open ] and C2[ Close ]. Taking the data in table 3.3 as an example, the lighting system boolean database generated after mapping is shown in table 2:

table 2 lighting system boolean datasheet

ID	A1	A2	A3	A4	A5	B1	B2	B3	B4	B5	C1	C2
													1	1	0	0	0	0	0	0	1	0	0	1	0
2	0	0	0	0	1	1	0	0	0	0	0	1
													3	0	0	0	0	1	0	1	0	0	0	1	0
4	0	0	0	1	0	0	1	0	0	0	0	1

All strong association rules are obtained after processing by adopting a classical Apriori algorithm.

Association rule mining with constraints

After the numerical data is converted into the Boolean data, a corresponding association rule needs to be found out through a mining algorithm. In the present embodiment, association rule mining with project constraint, that is, association rule mining algorithm with "operation type" as the latter constraint, is applied.

In general, association rule mining needs to perform certain constraint control on data, and only a certain rule subset is concerned in the mining process in order to ensure mining pertinence. In this embodiment, an association rule mining algorithm with project constraints is applied. During data mining, a propulsion method is adopted, and a direct algorithm with a constrained association rule mining algorithm is adopted, wherein the direct algorithm is an improved Apriori algorithm, and the core idea is to remove the part which does not meet constraint terms when a frequent item set of each stage is generated.

Lighting system association rule mining

Data mining of the lighting equipment aims to mine the relation between the on-off operation of personnel and time and illumination intensity. Taking a certain laboratory lighting as an example, the minimum support degree in the lighting association rule is set to 0.2, the minimum confidence degree is set to 0.5, and data of 10 months and 11 months are selected as mining objects for mining during mining. An original rule table of the mining results is obtained, and a part of the mining results are listed as an original rule table of the lighting system in table 3.

Table 3 original rule table of lighting system

The above rule 7 is now exemplified for the meaning of the original rule.

In the history data, the rule "a 1, B2 → C1 (confidence 21.3%, support 80.4%) indicates that under the condition that a1 and B2 are simultaneously 1, the probability that C1 is 1 is 80.4%, and the frequency of occurrence of the rule in the history data is 21.3%.

Corresponding to the attribute meaning, the above table, rule 7, can be interpreted as:

the frequency of simultaneous occurrence of indoor illuminance of [0,2312] lx and user light turn-off in historical data within 15min before and after the working time point is 21.3%; the probability of the user turning off the light in the historical data corresponding to the indoor illuminance of [0,2312] lx within 15min before and after the working time point is 80.4%.

The system stores the association rule mining result in the storage module and realizes the automatic update of the association rule base. Considering that the association rule mining needs to consume a certain time, which affects the real-time performance of intelligent control based on the existence of personnel, a timer can be set to be 24h according to the needs of a user, namely, the rule is updated every other day, and the result obtained by the association rule mining is used as the basis for sending an instruction by a management system.

The basic structure of the management control system of the Internet of things of the electrical equipment is as follows: the system comprises an infrared sensor, a management system, a controller and a controller, wherein the infrared sensor is used for acquiring people coming/walking information, inputting the people coming/walking information into the management system, the management system conducts reasoning and matching according to the input people coming/walking information, outputting a device start-stop control signal according to a rule matching result, outputting a device control signal through the controller, and the controller acts on electrical equipment to conduct real-time management control.

Inference mechanism

The embodiment utilizes artificial intelligence to judge whether a person should turn on the electrical equipment or not and whether the person should turn off the equipment or not. The association rules are mined according to the power utilization requirements, and when reasoning is carried out, the applicable rules are found according to the existing facts, and a process of selecting a rule-executing rule is formed until the condition is met or no available rule exists.

In the reasoning process, if one rule and only one rule can be matched, the system can directly apply the current rule; however, sometimes more than one rule can be matched, at this time, conflict resolution is required, and one of the multiple executable rules is selected to be executed. Because the expert system of the present embodiment employs association rule mining to acquire knowledge, the rule includes two parameters, i.e., a confidence level and a support level, and the parameter value is between 0 and 1. Wherein, the confidence coefficient is used for characterizing the certainty of the association rule, and the higher the confidence coefficient is, the more reliable and accurate the rule is. The method of confidence degree ordering is adopted when conflict resolution is carried out. In the system, the conflict resolution is specifically implemented by arranging all successfully matched rules in sequence according to the confidence degree when the successfully matched rules have a plurality of rules, selecting the rule with the highest confidence degree, and establishing a rule object to execute the next inference process. The core content of the inference is matching-conflict resolution-execution.

The lighting system reminds people to come/walk, proposes the assumption of turning on the light or the assumption of turning off the light, finds out the rule that the conclusion part comprises the assumption, carries out the antecedent matching of the rule according to the time attribute of the change moment of people and the outdoor illuminance, and judges whether the light is to be turned on/off according to the people to come/walk. The system adopts a reasoning mode of bidirectional reasoning. As shown in fig. 1.

The above embodiments are only for illustrating the technical idea of the present invention, and the protection scope of the present invention cannot be limited thereby, and any modification made on the basis of the technical scheme according to the technical idea proposed by the present invention falls within the protection scope of the present invention; the technology not related to the invention can be realized by the prior art.

Claims

1. An electric Internet of things system is characterized by comprising a sensing layer, a network layer and an application layer;

the management computer comprises a management system and a database;

the management computer carries out logic reasoning according to the association rule in the storage module and then outputs a signal to the switch-type module, and the output signal is converted into the switching action of the electrical equipment through the switch-type module.

2. An electrical internet of things system as claimed in claim 1, wherein: the management system is additionally provided with an interface connected with the 5G module, and a user can surf the internet and download the APP through the 5G mobile phone to realize the remote monitoring function of the state of the electric equipment.

3. An electrical internet of things system as claimed in claim 1, wherein: the input signal of the rule mining module is man-coming/man-walking information, the rule mining module conducts reasoning and matching according to the input man-coming/man-walking information, a device start-stop control signal is output according to a correlation rule matching result, and the controller outputs a control signal of the electrical device, which acts on the electrical device to conduct real-time control.

4. An electrical internet of things system as claimed in claim 1, wherein: the environmental parameters are indoor illumination intensity and infrared sensing personnel detection; the electric parameter detection module measures voltage and current parameters of the electric equipment nodes and sends the voltage and current parameters to the management system, so that on-line monitoring of harmonic waves of the electric equipment and the power distribution branch in the building is realized.

5. An electrical internet of things system as claimed in claim 1, wherein: the mining process of the association rule is as follows:

(b) converting the data from numerical data to Boolean data;