CN105204490A - Intelligent diagnosis system and method for standby power consumption based on integration characteristic selection and classification - Google Patents
Intelligent diagnosis system and method for standby power consumption based on integration characteristic selection and classification Download PDFInfo
- Publication number
- CN105204490A CN105204490A CN201510475741.XA CN201510475741A CN105204490A CN 105204490 A CN105204490 A CN 105204490A CN 201510475741 A CN201510475741 A CN 201510475741A CN 105204490 A CN105204490 A CN 105204490A
- Authority
- CN
- China
- Prior art keywords
- power consumption
- temperature
- samples
- support vector
- vector machine
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 24
- 238000003745 diagnosis Methods 0.000 title claims abstract description 23
- 230000010354 integration Effects 0.000 title description 3
- 238000012706 support-vector machine Methods 0.000 claims abstract description 30
- 230000035772 mutation Effects 0.000 claims description 15
- 230000005540 biological transmission Effects 0.000 claims description 14
- 238000005265 energy consumption Methods 0.000 claims description 13
- 230000003044 adaptive effect Effects 0.000 claims description 12
- 238000012360 testing method Methods 0.000 claims description 12
- 238000012549 training Methods 0.000 claims description 12
- 238000004422 calculation algorithm Methods 0.000 claims description 8
- 239000003795 chemical substances by application Substances 0.000 claims description 6
- 238000002405 diagnostic procedure Methods 0.000 claims description 6
- 230000002068 genetic effect Effects 0.000 claims description 5
- 238000012216 screening Methods 0.000 claims description 4
- 206010064571 Gene mutation Diseases 0.000 claims description 3
- 239000011159 matrix material Substances 0.000 claims description 3
- 108090000623 proteins and genes Proteins 0.000 claims description 3
- 238000012795 verification Methods 0.000 claims description 3
- 238000002921 genetic algorithm search Methods 0.000 claims 1
- 238000007635 classification algorithm Methods 0.000 abstract description 2
- 230000006870 function Effects 0.000 description 10
- 238000005516 engineering process Methods 0.000 description 4
- 238000009434 installation Methods 0.000 description 3
- 238000012423 maintenance Methods 0.000 description 3
- 238000004891 communication Methods 0.000 description 2
- 239000003814 drug Substances 0.000 description 2
- 230000005611 electricity Effects 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 230000006855 networking Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000007792 addition Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B23/00—Testing or monitoring of control systems or parts thereof
- G05B23/02—Electric testing or monitoring
- G05B23/0205—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults
- G05B23/0259—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterized by the response to fault detection
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Automation & Control Theory (AREA)
- Supply And Distribution Of Alternating Current (AREA)
- Testing And Monitoring For Control Systems (AREA)
Abstract
本发明提供了一种基于集成特征选择分类的待机功耗智能诊断系统及其诊断方法，上位机对采集到的家用电器信号进行特征参数选择，利用支持向量机对特征参数进行筛选，得到最优特征子集以及训练后的支持向量机分类器，通过该集成特征选择分类算法判断家用电器是否处于待机状态，若处于待机状态，则通过终端节点控制控制开关模块将待机的家用电器关闭，达到减少待机功耗，节约电能的目的。
The invention provides a standby power consumption intelligent diagnosis system and its diagnosis method based on integrated feature selection and classification. The upper computer selects the characteristic parameters of the collected household electrical appliances signals, and uses the support vector machine to screen the characteristic parameters to obtain the optimal The feature subset and the trained support vector machine classifier judge whether the household appliance is in the standby state through the integrated feature selection classification algorithm. If it is in the standby state, the terminal node controls the control switch module to turn off the standby household appliance to reduce Standby power consumption, the purpose of saving power.
Description
技术领域technical field
本发明涉及家用电器功耗诊断技术领域，具体涉及一种基于集成特征选择分类的待机功耗智能诊断系统及其诊断方法。The invention relates to the technical field of power consumption diagnosis of household appliances, in particular to an intelligent diagnosis system of standby power consumption based on integrated feature selection and classification and a diagnosis method thereof.
背景技术Background technique
随着智能电网和阶梯、峰谷电价的推广，各种智能电表应运而生，人们节电意识也越来越强，也希望对家用电器进行智能化管理并且能够实时的了解家用电器的待机状态，然后自动的关掉处于待机状态的家用电器。智能节电系统可以帮助人们实现这一愿望。然而传统的智能设备最大的弊端就是增减设备都需要重新布线，重复投资且影响美观；另一方面，系统的可扩展性和移动性也比较差，安装和维护成本高。With the promotion of smart grids, ladders, and peak-to-valley electricity prices, various smart meters have emerged as the times require, and people's awareness of power saving is becoming stronger and stronger. They also hope to intelligently manage household appliances and be able to understand the standby status of household appliances in real time. , and then automatically turn off the household appliances in the standby state. Intelligent power saving system can help people realize this wish. However, the biggest disadvantage of traditional smart devices is that adding or removing devices requires re-wiring, duplicating investment and affecting the appearance; on the other hand, the scalability and mobility of the system are relatively poor, and the installation and maintenance costs are high.
发明内容Contents of the invention
本申请通过提供一种基于集成特征选择分类的待机功耗智能诊断统及其诊断方法，以解决现有技术中智能节电系统可扩展性和可移动性差，安装以及维护成本较高的技术问题。This application provides an intelligent diagnosis system and diagnosis method for standby power consumption based on integrated feature selection and classification to solve the technical problems of poor scalability and mobility of intelligent power-saving systems and high installation and maintenance costs in the prior art .
为解决上述技术问题，本申请采用以下技术方案予以实现：In order to solve the above-mentioned technical problems, the application adopts the following technical solutions to achieve:
一种基于集成特征选择分类的待机功耗智能诊断系统，包括上位机、协调器以及每个家用电器处设置的终端节点、信息采集模块、控制开关，其中，所述信息采集模块包括温度传感器和电能采集芯片，用以采集家用电器的温度、电压以及电流信息，由所述终端节点将采集到的信息通过所述协调器传输给所述上位机，所述上位机对采集到的信息进行分析判断，并通过所述协调器向所述终端节点发出控制指令，所述终端节点控制所述控制开关，从而控制家用电器的关闭，所述上位机处设置有温湿度传感器用以采集室内温度和湿度。An intelligent diagnostic system for standby power consumption based on integrated feature selection classification, including a host computer, a coordinator, and terminal nodes set at each household appliance, an information collection module, and a control switch, wherein the information collection module includes temperature sensors and The power collection chip is used to collect the temperature, voltage and current information of household appliances, and the terminal node transmits the collected information to the host computer through the coordinator, and the host computer analyzes the collected information Judging, and sending a control command to the terminal node through the coordinator, the terminal node controls the control switch, thereby controlling the shutdown of household appliances, and the host computer is provided with a temperature and humidity sensor to collect indoor temperature and humidity.
作为优选的技术方案，所述上位机采用ARM9TQ2440上位机，所述协调器采用系统芯片CC2430，所述终端节点为ZigBee电路板，所述控制开关模块采用SL-C电磁式继电器，所述温度传感器采用DS18B20，所述电能采集芯片采用ADE7755，所述温湿度传感器采用DHT21。As a preferred technical solution, the upper computer adopts an ARM9TQ2440 upper computer, the coordinator adopts a system chip CC2430, the terminal node is a ZigBee circuit board, the control switch module adopts an SL-C electromagnetic relay, and the temperature sensor DS18B20 is used, the power collection chip is ADE7755, and the temperature and humidity sensor is DHT21.
其中，上位机选用32位的ARM微控制器ARM9TQ2440，它的工作频率可达几百MHz。集成有许多片内外设，并有多种通信接口，体积小，功耗和成本低，可靠性高，特别适合作为嵌入式上位机。系统一般采用Flash作为程序存储器，采用SDRAM作为系统内存。可以采用VxWorks、WinCE、Linux等嵌入式操作系统。在基于ARM平台上可嵌入较完整的TCP/IP协议，实现较强的Web服务功能。并且系统中能集成多种接口部件，可以完成较多复杂的功能。为家庭网关后继功能的扩展提供了可能。Among them, the upper computer selects 32-bit ARM microcontroller ARM9TQ2440, and its operating frequency can reach several hundred MHz. It integrates many on-chip and external devices, and has a variety of communication interfaces. It is small in size, low in power consumption and cost, and high in reliability. It is especially suitable as an embedded host computer. The system generally uses Flash as the program memory and SDRAM as the system memory. Embedded operating systems such as VxWorks, WinCE, and Linux can be used. A relatively complete TCP/IP protocol can be embedded on the ARM-based platform to realize strong Web service functions. And the system can integrate a variety of interface components to complete more complex functions. It provides the possibility for the expansion of the successor function of the home gateway.
协调器采用了华诺的CC2430系统芯片，CC2430系统芯片是集成ZigBee技术、8051MCU处理核心的SOC芯片，在集成度和成本以及研发难度上，都具备相当的优势。ZigBee具有近距离、低功耗、低速率、双向传输等特点，是一种基于IEEE802.15.4无线标准研制开发的有关组网、安全和应用软件方面无线网络技术，主要适合于承载数据流量小、数据传输速率低的业务，可嵌入各种设备中，能够实现对家庭、工业以及医学等各种重要场所的监控。ZigBee网络主要由协调器、路由器和终端节点组成。ZigBee支持星状型、网状和树簇状的网络拓扑结构。每一ZigBee网络中最多可以拥有65535个节点，每个节点的地址由ZigBee的网络协调节点(NetworkCoordinator)负责分配。另外，每个节点的传输范围在30-100m之间，而且传输的距离还可以通过使用功率放大器和多跳网状结构得到延伸。The coordinator adopts Huanuo's CC2430 system chip. CC2430 system chip is an SOC chip integrating ZigBee technology and 8051MCU processing core. It has considerable advantages in terms of integration, cost and R&D difficulty. ZigBee has the characteristics of short distance, low power consumption, low speed, and two-way transmission. It is a wireless network technology related to networking, security, and application software developed based on the IEEE802.15.4 wireless standard. It is mainly suitable for carrying small data traffic. Services with low data transmission rates can be embedded in various devices, enabling monitoring of various important places such as homes, industries, and medicine. ZigBee network is mainly composed of coordinator, router and terminal nodes. ZigBee supports star, mesh and tree cluster network topologies. Each ZigBee network can have up to 65535 nodes, and the address of each node is assigned by ZigBee's network coordinator node (NetworkCoordinator). In addition, the transmission range of each node is between 30-100m, and the transmission distance can also be extended by using power amplifiers and multi-hop mesh structures.
每个终端节点都是一个小的ZigBee电路板，当上位机对采集到的家用电器信号进行判断，得出家用电器处于待机状态后，上位机通过终端节点控制SL-C电磁式继电器的闭合来达到控制插座通电与否，关掉处于待机状态的家用电器，从而减少了待机功耗，节约电能。Each terminal node is a small ZigBee circuit board. When the host computer judges the collected household electrical appliance signals and concludes that the household appliances are in the standby state, the host computer controls the closing of the SL-C electromagnetic relay through the terminal node. To control whether the socket is powered on or not, turn off the household appliances in the standby state, thereby reducing the standby power consumption and saving electric energy.
一种基于集成特征选择分类的待机功耗智能诊断系统的诊断方法，包括以下步骤：A diagnostic method for an intelligent diagnostic system for standby power consumption based on integrated feature selection and classification, comprising the following steps:
S1：家用电器信号的采集；S1: Acquisition of household electrical appliances signals;
S2：家用电器信号的传输；S2: transmission of household appliances signal;
S3：根据采集到的家用电器信号计算家用电器的功率P；S3: Calculate the power P of the household appliance according to the collected household appliance signal;
S4：提取家用电器信号的电压特征Vi(i＝1,2,…,m)、电流特征Iq(q＝1,2,…,n)、温度特征Tp(p＝1,2,…,s)以及室内温度特征T1r(r＝1,2,…,l)和湿度特征S1t(t＝1,2,…,z)，其中电压特征个数i的取值为1到m的自然数，电流特征个数q的取值为1到n的自然数，家用电器温度特征个数p的取值为1到s的自然数，室内温度特征个数r的取值为1到l的自然数，室内湿度特征个数t的取值为1到z的自然数；S4: Extract the voltage characteristics V i (i=1,2,...,m), current characteristics I q (q=1,2,...,n), temperature characteristics T p (p=1,2, …,s) and indoor temperature features T1 r (r=1,2,…,l) and humidity features S1 t (t=1,2,…,z), where the number of voltage features i ranges from 1 to The natural number of m, the value of the number of current characteristics q is a natural number from 1 to n, the value of the number of temperature characteristics of household appliances p is a natural number from 1 to s, the value of the number of indoor temperature characteristics r is from 1 to 1 A natural number, the number t of indoor humidity characteristics is a natural number from 1 to z;
S5：基于支持向量机回归算法，建立反演模型，通过反演准确性进行特征筛选，以获得最优特征子集Ffinal(1,2,…k)以及反演得到最接近功率P的相应功率P′，最优特征子集的个数final取值为1到k的自然数；S5: Based on the support vector machine regression algorithm, establish an inversion model, and perform feature screening through the inversion accuracy to obtain the optimal feature subset F final (1,2,...k) and inversion to obtain the corresponding corresponding power P Power P', the number of optimal feature subsets final is a natural number from 1 to k;
S6：将功率P′与电压特征Vi(i＝1,2,…,m)、电流特征Iq(q＝1,2,…,n)、温度特征Tp(p＝1,2,…,t)以及室内温度特征T1r(r＝1,2,…,l)和湿度特征S1t(t＝1,2,…,z)合并，构成总的待选特征集Fj(j＝1,2,…,h)，待选特征集的个数j取值为1到h的自然数；S6: Compare power P′ with voltage characteristics V i (i=1,2,…,m), current characteristics I q (q=1,2,…,n), temperature characteristics T p (p=1,2, …,t) and indoor temperature features T1 r (r=1,2,…,l) and humidity features S1 t (t=1,2,…,z) are combined to form the total candidate feature set F j (j =1,2,...,h), the number j of feature sets to be selected is a natural number from 1 to h;
S7：基于支持向量机分类器和总的待选特征集Fj(j＝1,2,…,h)进行特征选择和分类；S7: Feature selection and classification based on the support vector machine classifier and the total feature set F j (j=1,2,...,h) to be selected;
S8：获得最优特征子集Ffinal(final＝1,2,…,k)和训练后的支持向量机分类器SVM_final；S8: Obtain the optimal feature subset F final (final=1,2,...,k) and the trained support vector machine classifier SVM_final;
S9：构建基于集成特征选择分类的待机功耗智能诊断系统；S9: Build an intelligent diagnostic system for standby power consumption based on integrated feature selection classification;
S10：判断家用电器是否处于待机状态，如果是，则进入步骤S11，否则，跳转回步骤S3；S10: Determine whether the household appliance is in the standby state, if yes, go to step S11, otherwise, go back to step S3;
S11：所述上位机通过所述无线传输模块控制控制开关模块，关闭处于待机状态的家用电器。S11: The host computer controls the control switch module through the wireless transmission module to turn off the household appliances in the standby state.
通过同时优化待选信号特征和支持向量机分类器参数,可以提高信号特征选择和获取与家用电器能耗关系式的精度。采用高精度的封装式特征选择模式，评价准则为分类器的模式分类准确率。以家用电器的能耗作为分类标准，从而将获取信号特征与家用电器能耗的关系式转化为模式分类问题。By simultaneously optimizing the signal features to be selected and the parameters of the support vector machine classifier, the accuracy of signal feature selection and obtaining the relationship between household appliances energy consumption can be improved. A high-precision packaged feature selection model is adopted, and the evaluation criterion is the pattern classification accuracy of the classifier. Taking the energy consumption of household appliances as the classification standard, the relationship between the acquired signal features and the energy consumption of household appliances is transformed into a pattern classification problem.
进一步地，步骤S8中采用链式智能体遗传算法搜索最优特征子集Ffinal(final＝1,2,…,k)，种群数量选择大于基因长度，自适应交叉概率为：Further, in step S8, the chain agent genetic algorithm is used to search for the optimal feature subset F final (final=1,2,...,k), the population size is selected to be greater than the gene length, and the adaptive crossover probability is:
式中，pc1和pc2为两个待交叉的个体，初始化pc1＝0.9，pc2＝0.6，favg为每代种群的平均适应度，fmax为每代种群的最大适应度，f'为待交叉的两个个体中较大的适应度值，交叉操作采用自适应交叉概率的单点交叉法；In the formula, p c1 and p c2 are two individuals to be crossed, initialized p c1 =0.9, p c2 =0.6, f avg is the average fitness of each generation population, f max is the maximum fitness of each generation population, f ' is the larger fitness value among the two individuals to be crossed, and the crossover operation adopts the single-point crossover method of adaptive crossover probability;
基因变异同样采用自适应的变异概率：Gene mutation also adopts adaptive mutation probability:
式中，pm1、pm2分别为个体1和个体2的变异概率，初始化pm1＝0.1，pm2＝0.006，favg为每代种群的平均适应度，fmax为每代种群的最大适应度，f为待变异个体的适应度值，变异操作采用自适应变异概率的二进制变异法。In the formula, p m1 and p m2 are the mutation probabilities of individual 1 and individual 2 respectively, initializing p m1 = 0.1, p m2 = 0.006, f avg is the average fitness of each generation population, and f max is the maximum fitness of each generation population degree, f is the fitness value of the individual to be mutated, and the mutation operation adopts the binary mutation method of adaptive mutation probability.
进一步地，步骤S9中，支持向量机的核函数为径向基函数，采用五阶校验法，训练收敛准则为均方误差，将信号样本分为A、B、C、D四组，其中A组样本用于训练支持向量机分类器，B组样本用于指导链式智能体遗传算法进行搜索最优特征子集，C组样本用于实行参数反演，D组样本用于进行性能测试。Further, in step S9, the kernel function of the support vector machine is a radial basis function, the fifth-order verification method is adopted, the training convergence criterion is the mean square error, and the signal samples are divided into four groups A, B, C, and D, where Group A samples are used to train the support vector machine classifier, group B samples are used to guide the chain agent genetic algorithm to search for the optimal feature subset, group C samples are used to perform parameter inversion, and group D samples are used for performance testing .
采用留一法对A组样本和B组样本进行测试，同时输出选择后的信号样本特征和训练好的支持向量机分类器参数；样本的特征值在特征选择分类前要进行归一化处理。Use the leave-one-out method to test the samples of group A and group B, and output the selected signal sample features and the parameters of the trained support vector machine classifier at the same time; the feature values of the samples must be normalized before feature selection and classification.
采用留十法将C组样本随机分为训练样本和测试样本，按此分发，得到多组训练样本和测试样本，基于已获得的训练样本和支持向量机分类器参数，对支持向量机进行参数回归，输入向量为信号特征值，输出向量为家用电器功耗的标准值，均方误差满足要求后停止，从而获取参数矩阵，即：信号特征值与家用电器功耗的关系式；样本的特征值在参数反演前不要进行归一化处理。Group C samples are randomly divided into training samples and test samples by using the method of leaving ten, and distributed according to this, multiple groups of training samples and test samples are obtained, based on the obtained training samples and support vector machine classifier parameters, parameterize the support vector machine Regression, the input vector is the signal eigenvalue, the output vector is the standard value of the power consumption of household appliances, the mean square error meets the requirements and stops, so as to obtain the parameter matrix, that is: the relationship between the signal eigenvalue and the power consumption of household appliances; the characteristics of the sample Values are not normalized before parameter inversion.
通过信号特征值与家用电器功耗的关系式可计算出家用电器在某一时间段内的耗能，对D组样本进行测试，获取家用电器能耗分布及数字的平均值和标准差。The energy consumption of household appliances in a certain period of time can be calculated through the relationship between the signal characteristic value and the power consumption of household appliances. The samples of group D are tested to obtain the distribution of energy consumption of household appliances and the average value and standard deviation of the figures.
作为一种优选的技术方案，步骤S4中提取的电压特征包括电压分布的不均匀性、电压平均、电压均方差、电压熵，电流特征包括电流分布的不均匀性、电流平均、电流均方差、电流熵，温度特征包括温度分布的不均匀性、温度平均、温度均方差、温度熵，室内温度特征包括室内温度分布的不均匀性、室内温度平均、室内温度方差、室内温度熵，室内湿度特征包括室内湿度分布的不均匀性、室内湿度平均、室内湿度方差、室内湿度熵。As a preferred technical solution, the voltage features extracted in step S4 include the unevenness of voltage distribution, voltage average, voltage mean square error, and voltage entropy, and the current features include current distribution unevenness, current average, current mean square error, Current entropy, temperature characteristics include temperature distribution inhomogeneity, temperature average, temperature mean square deviation, temperature entropy, indoor temperature characteristics include indoor temperature distribution inhomogeneity, indoor temperature average, indoor temperature variance, indoor temperature entropy, indoor humidity characteristics Including the unevenness of indoor humidity distribution, indoor humidity average, indoor humidity variance, and indoor humidity entropy.
与现有技术相比，本申请提供的技术方案，具有的技术效果或优点是：系统可控性高，可扩展性好，且降低了系统的能耗，易于推广应用。Compared with the prior art, the technical solution provided by the present application has the following technical effects or advantages: high system controllability, good scalability, reduced system energy consumption, and easy popularization and application.
附图说明Description of drawings
图1为本发明的待机功耗智能诊断系统结构框图；Fig. 1 is a structural block diagram of the standby power consumption intelligent diagnosis system of the present invention;
图2为本发明的待机功耗智能诊断方法的流程图。FIG. 2 is a flow chart of the intelligent diagnosis method for standby power consumption of the present invention.
具体实施方式Detailed ways
本申请实施例通过提供一种基于集成特征选择分类的待机功耗智能诊断系统及其诊断方法，以解决现有技术中智能节电系统可扩展性和可移动性差，安装以及维护成本较高的技术问题。The embodiment of the present application provides an intelligent diagnosis system for standby power consumption and its diagnosis method based on integrated feature selection and classification to solve the problem of poor scalability and mobility and high installation and maintenance costs of the intelligent power saving system in the prior art. technical problem.
为了更好的理解上述技术方案，下面将结合说明书附图以及具体的实施方式，对上述技术方案进行详细的说明。In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation manners.
实施例Example
一种基于集成特征选择分类的待机功耗智能诊断系统，如图1所示，包括上位机、协调器以及每个家用电器处设置的终端节点、信息采集模块、控制开关，其中，所述信息采集模块包括温度传感器和电能采集芯片，用以采集家用电器的温度、电压以及电流信息，由所述终端节点将采集到的信息通过所述协调器传输给所述上位机，所述上位机对采集到的信息进行分析判断，并通过所述协调器向所述终端节点发出控制指令，所述终端节点控制所述控制开关，从而控制家用电器的关闭，所述上位机处设置有温湿度传感器用以采集室内温度和湿度。An intelligent diagnostic system for standby power consumption based on integrated feature selection and classification, as shown in Figure 1, includes a host computer, a coordinator, and a terminal node, an information collection module, and a control switch set at each household appliance, wherein the information The collection module includes a temperature sensor and an electric energy collection chip, which is used to collect temperature, voltage and current information of household appliances, and the terminal node transmits the collected information to the host computer through the coordinator, and the host computer The collected information is analyzed and judged, and a control instruction is sent to the terminal node through the coordinator, and the terminal node controls the control switch to control the shutdown of the household appliances. The upper computer is provided with a temperature and humidity sensor Used to collect indoor temperature and humidity.
作为优选的技术方案，所述上位机采用ARM9TQ2440上位机，所述协调器采用系统芯片CC2430，所述终端节点为ZigBee电路板，所述控制开关模块采用SL-C电磁式继电器，所述温度传感器采用DS18B20，所述电能采集芯片采用ADE7755，所述温湿度传感器采用DHT21。As a preferred technical solution, the upper computer adopts an ARM9TQ2440 upper computer, the coordinator adopts a system chip CC2430, the terminal node is a ZigBee circuit board, the control switch module adopts an SL-C electromagnetic relay, and the temperature sensor DS18B20 is used, the power collection chip is ADE7755, and the temperature and humidity sensor is DHT21.
其中，上位机选用32位的ARM微控制器ARM9TQ2440，它的工作频率可达几百MHz。集成有许多片内外设，并有多种通信接口，体积小，功耗和成本低，可靠性高，特别适合作为嵌入式上位机。系统一般采用Flash作为程序存储器，采用SDRAM作为系统内存。可以采用VxWorks、WinCE、Linux等嵌入式操作系统。在基于ARM平台上可嵌入较完整的TCP/IP协议，实现较强的Web服务功能。并且系统中能集成多种接口部件，可以完成较多复杂的功能。为家庭网关后继功能的扩展提供了可能。Among them, the upper computer selects 32-bit ARM microcontroller ARM9TQ2440, and its operating frequency can reach several hundred MHz. It integrates many on-chip and external devices, and has a variety of communication interfaces. It is small in size, low in power consumption and cost, and high in reliability. It is especially suitable as an embedded host computer. The system generally uses Flash as the program memory and SDRAM as the system memory. Embedded operating systems such as VxWorks, WinCE, and Linux can be used. A relatively complete TCP/IP protocol can be embedded on the ARM-based platform to realize strong Web service functions. And the system can integrate a variety of interface components to complete more complex functions. It provides the possibility for the expansion of the successor function of the home gateway.
协调器采用了华诺的CC2430系统芯片，CC2430系统芯片是集成ZigBee技术、8051MCU处理核心的SOC芯片，在集成度和成本以及研发难度上，都具备相当的优势。ZigBee具有近距离、低功耗、低速率、双向传输等特点，是一种基于IEEE802.15.4无线标准研制开发的有关组网、安全和应用软件方面无线网络技术，主要适合于承载数据流量小、数据传输速率低的业务，可嵌入各种设备中，能够实现对家庭、工业以及医学等各种重要场所的监控。ZigBee网络主要由协调器、路由器和终端节点组成。ZigBee支持星状型、网状和树簇状的网络拓扑结构。每一ZigBee网络中最多可以拥有65535个节点，每个节点的地址由ZigBee的网络协调节点(NetworkCoordinator)负责分配。另外，每个节点的传输范围在30-100m之间，而且传输的距离还可以通过使用功率放大器和多跳网状结构得到延伸。The coordinator adopts Huanuo's CC2430 system chip. CC2430 system chip is an SOC chip integrating ZigBee technology and 8051MCU processing core. It has considerable advantages in terms of integration, cost and R&D difficulty. ZigBee has the characteristics of short distance, low power consumption, low speed, and two-way transmission. It is a wireless network technology related to networking, security, and application software developed based on the IEEE802.15.4 wireless standard. It is mainly suitable for carrying small data traffic. Services with low data transmission rates can be embedded in various devices, enabling monitoring of various important places such as homes, industries, and medicine. ZigBee network is mainly composed of coordinator, router and terminal nodes. ZigBee supports star, mesh and tree cluster network topologies. Each ZigBee network can have up to 65535 nodes, and the address of each node is assigned by ZigBee's network coordinator node (NetworkCoordinator). In addition, the transmission range of each node is between 30-100m, and the transmission distance can also be extended by using power amplifiers and multi-hop mesh structures.
每个终端节点都是一个小的ZigBee电路板，当上位机对采集到的家用电器信号进行判断，得出家用电器处于待机状态后，上位机通过终端节点控制SL-C电磁式继电器的闭合来达到控制插座通电与否，关掉处于待机状态的家用电器，从而减少了待机功耗，节约电能。Each terminal node is a small ZigBee circuit board. When the host computer judges the collected household electrical appliance signals and concludes that the household appliances are in the standby state, the host computer controls the closing of the SL-C electromagnetic relay through the terminal node. To control whether the socket is powered on or not, turn off the household appliances in the standby state, thereby reducing the standby power consumption and saving electric energy.
一种基于集成特征选择分类的待机功耗智能诊断系统的诊断方法，如图2所示，包括以下步骤：A kind of diagnosis method of the intelligent diagnosis system of standby power consumption based on integrated feature selection classification, as shown in Figure 2, comprises the following steps:
S1：家用电器信号的采集；S1: Acquisition of household electrical appliances signals;
S2：家用电器信号的传输；S2: transmission of household appliances signal;
S3：根据采集到的家用电器信号计算家用电器的功率P；S3: Calculate the power P of the household appliance according to the collected household appliance signal;
S4：提取家用电器信号的电压特征Vi(i＝1,2,…,m)、电流特征Iq(q＝1,2,…,n)、温度特征Tp(p＝1,2,…,s)以及室内温度特征T1r(r＝1,2,…,l)和湿度特征S1t(t＝1,2,…,z)，其中电压特征个数i的取值为1到m的自然数，电流特征个数q的取值为1到n的自然数，家用电器温度特征个数p的取值为1到s的自然数，室内温度特征个数r的取值为1到l的自然数，室内湿度特征个数t的取值为1到z的自然数；S4: Extract the voltage characteristics V i (i=1,2,...,m), current characteristics I q (q=1,2,...,n), temperature characteristics T p (p=1,2, …,s) and indoor temperature features T1 r (r=1,2,…,l) and humidity features S1 t (t=1,2,…,z), where the number of voltage features i ranges from 1 to The natural number of m, the value of the number of current characteristics q is a natural number from 1 to n, the value of the number of temperature characteristics of household appliances p is a natural number from 1 to s, the value of the number of indoor temperature characteristics r is from 1 to 1 A natural number, the number t of indoor humidity characteristics is a natural number from 1 to z;
S5：基于支持向量机回归算法，建立反演模型，通过反演准确性进行特征筛选，以获得最优特征子集Ffinal(1,2,…k)以及反演得到最接近功率P的相应功率P′，最优特征子集的个数final取值为1到k的自然数；S5: Based on the support vector machine regression algorithm, establish an inversion model, and perform feature screening through the inversion accuracy to obtain the optimal feature subset F final (1,2,...k) and inversion to obtain the corresponding corresponding power P Power P', the number of optimal feature subsets final is a natural number from 1 to k;
S6：将功率P′与电压特征Vi(i＝1,2,…,m)、电流特征Iq(q＝1,2,…,n)、温度特征Tp(p＝1,2,…,t)以及室内温度特征T1r(r＝1,2,…,l)和湿度特征S1t(t＝1,2,…,z)合并，构成总的待选特征集Fj(j＝1,2,…,h)，待选特征集的个数j取值为1到h的自然数；S6: Compare power P′ with voltage characteristics V i (i=1,2,…,m), current characteristics I q (q=1,2,…,n), temperature characteristics T p (p=1,2, …,t) and indoor temperature features T1 r (r=1,2,…,l) and humidity features S1 t (t=1,2,…,z) are combined to form the total candidate feature set F j (j =1,2,...,h), the number j of feature sets to be selected is a natural number from 1 to h;
S7：基于支持向量机分类器和总的待选特征集Fj(j＝1,2,…,h)进行特征选择和分类；S7: Feature selection and classification based on the support vector machine classifier and the total feature set F j (j=1,2,...,h) to be selected;
S8：获得最优特征子集Ffinal(final＝1,2,…,k)和训练后的支持向量机分类器SVM_final；S8: Obtain the optimal feature subset F final (final=1,2,...,k) and the trained support vector machine classifier SVM_final;
S9：构建基于集成特征选择分类的待机功耗智能诊断系统；S9: Build an intelligent diagnostic system for standby power consumption based on integrated feature selection classification;
S10：判断家用电器是否处于待机状态，如果是，则进入步骤S11，否则，跳转回步骤S3；S10: Determine whether the household appliance is in the standby state, if yes, go to step S11, otherwise, go back to step S3;
S11：所述上位机通过所述无线传输模块控制控制开关模块，关闭处于待机状态的家用电器。S11: The host computer controls the control switch module through the wireless transmission module to turn off the household appliances in the standby state.
通过同时优化待选信号特征和支持向量机分类器参数,可以提高信号特征选择和获取与家用电器能耗关系式的精度。采用高精度的封装式特征选择模式，评价准则为分类器的模式分类准确率。以家用电器的能耗作为分类标准，从而将获取信号特征与家用电器能耗的关系式转化为模式分类问题。By simultaneously optimizing the signal features to be selected and the parameters of the support vector machine classifier, the accuracy of signal feature selection and obtaining the relationship between household appliances energy consumption can be improved. A high-precision packaged feature selection model is adopted, and the evaluation criterion is the pattern classification accuracy of the classifier. Taking the energy consumption of household appliances as the classification standard, the relationship between the acquired signal features and the energy consumption of household appliances is transformed into a pattern classification problem.
进一步地，步骤S8中采用链式智能体遗传算法搜索最优特征子集Ffinal(final＝1,2,…,k)，种群数量选择大于基因长度，自适应交叉概率为：Further, in step S8, the chain agent genetic algorithm is used to search for the optimal feature subset F final (final=1,2,...,k), the population size is selected to be greater than the gene length, and the adaptive crossover probability is:
式中，pc1和pc2为两个待交叉的个体，初始化pc1＝0.9，pc2＝0.6，favg为每代种群的平均适应度，fmax为每代种群的最大适应度，f'为待交叉的两个个体中较大的适应度值，交叉操作采用自适应交叉概率的单点交叉法；In the formula, p c1 and p c2 are two individuals to be crossed, initialized p c1 =0.9, p c2 =0.6, f avg is the average fitness of each generation population, f max is the maximum fitness of each generation population, f ' is the larger fitness value among the two individuals to be crossed, and the crossover operation adopts the single-point crossover method of adaptive crossover probability;
基因变异同样采用自适应的变异概率：Gene mutation also adopts adaptive mutation probability:
式中，pm1、pm2分别为个体1和个体2的变异概率，初始化pm1＝0.1，pm2＝0.006，favg为每代种群的平均适应度，fmax为每代种群的最大适应度，f为待变异个体的适应度值，变异操作采用自适应变异概率的二进制变异法。In the formula, p m1 and p m2 are the mutation probabilities of individual 1 and individual 2 respectively, initializing p m1 = 0.1, p m2 = 0.006, f avg is the average fitness of each generation population, and f max is the maximum fitness of each generation population degree, f is the fitness value of the individual to be mutated, and the mutation operation adopts the binary mutation method of adaptive mutation probability.
进一步地，步骤S9中，支持向量机的核函数为径向基函数，采用五阶校验法，训练收敛准则为均方误差，将信号样本分为A、B、C、D四组，其中A组样本用于训练支持向量机分类器，B组样本用于指导链式智能体遗传算法进行搜索最优特征子集，C组样本用于实行参数反演，D组样本用于进行性能测试。Further, in step S9, the kernel function of the support vector machine is a radial basis function, the fifth-order verification method is adopted, the training convergence criterion is the mean square error, and the signal samples are divided into four groups A, B, C, and D, where Group A samples are used to train the support vector machine classifier, group B samples are used to guide the chain agent genetic algorithm to search for the optimal feature subset, group C samples are used to perform parameter inversion, and group D samples are used for performance testing .
采用留一法对A组样本和B组样本进行测试，同时输出选择后的信号样本特征和训练好的支持向量机分类器参数；样本的特征值在特征选择分类前要进行归一化处理。Use the leave-one-out method to test the samples of group A and group B, and output the selected signal sample features and the parameters of the trained support vector machine classifier at the same time; the feature values of the samples must be normalized before feature selection and classification.
采用留十法将C组样本随机分为训练样本和测试样本，按此分发，得到多组训练样本和测试样本，基于已获得的训练样本和支持向量机分类器参数，对支持向量机进行参数回归，输入向量为信号特征值，输出向量为家用电器功耗的标准值，均方误差满足要求后停止，从而获取参数矩阵，即：信号特征值与家用电器功耗的关系式；样本的特征值在参数反演前不要进行归一化处理。Group C samples are randomly divided into training samples and test samples by using the method of leaving ten, and distributed according to this, multiple groups of training samples and test samples are obtained, based on the obtained training samples and support vector machine classifier parameters, parameterize the support vector machine Regression, the input vector is the signal eigenvalue, the output vector is the standard value of the power consumption of household appliances, the mean square error meets the requirements and stops, so as to obtain the parameter matrix, that is: the relationship between the signal eigenvalue and the power consumption of household appliances; the characteristics of the sample Values are not normalized before parameter inversion.
通过信号特征值与家用电器功耗的关系式可计算出家用电器在某一时间段内的耗能，对D组样本进行测试，获取家用电器能耗分布及数字的平均值和标准差。The energy consumption of household appliances in a certain period of time can be calculated through the relationship between the signal characteristic value and the power consumption of household appliances. The samples of group D are tested to obtain the distribution of energy consumption of household appliances and the average value and standard deviation of the figures.
作为一种优选的技术方案，步骤S4中提取的电压特征包括电压分布的不均匀性、电压平均、电压均方差、电压熵，电流特征包括电流分布的不均匀性、电流平均、电流均方差、电流熵，温度特征包括温度分布的不均匀性、温度平均、温度均方差、温度熵，室内温度特征包括室内温度分布的不均匀性、室内温度平均、室内温度方差、室内温度熵，室内湿度特征包括室内湿度分布的不均匀性、室内湿度平均、室内湿度方差、室内湿度熵。As a preferred technical solution, the voltage features extracted in step S4 include the unevenness of voltage distribution, voltage average, voltage mean square error, and voltage entropy, and the current features include current distribution unevenness, current average, current mean square error, Current entropy, temperature characteristics include temperature distribution inhomogeneity, temperature average, temperature mean square deviation, temperature entropy, indoor temperature characteristics include indoor temperature distribution inhomogeneity, indoor temperature average, indoor temperature variance, indoor temperature entropy, indoor humidity characteristics Including the unevenness of indoor humidity distribution, indoor humidity average, indoor humidity variance, and indoor humidity entropy.
本申请的上述实施例中，通过提供一种基于集成特征选择分类的待机功耗智能诊断系统及其诊断方法，上位机对采集到的家用电器信号进行特征参数选择，利用支持向量机对特征参数进行筛选，得到最优特征子集以及训练后的支持向量机分类器，通过该集成特征选择分类算法判断家用电器是否处于待机状态，若处于待机状态，则通过终端节点控制控制开关模块将待机的家用电器关闭，达到减少待机功耗，节约电能的目的。In the above-mentioned embodiments of the present application, by providing an intelligent diagnosis system and diagnosis method for standby power consumption based on integrated feature selection and classification, the upper computer selects the characteristic parameters of the collected household electrical appliance signals, and uses the support vector machine to analyze the characteristic parameters Screening is performed to obtain the optimal feature subset and the trained support vector machine classifier. Through the integrated feature selection classification algorithm, it is judged whether the household appliance is in the standby state. If it is in the standby state, the terminal node controls the control switch module to switch the standby Household appliances are turned off to reduce standby power consumption and save electricity.
应当指出的是，上述说明并非是对本发明的限制，本发明也并不仅限于上述举例，本技术领域的普通技术人员在本发明的实质范围内所做出的变化、改性、添加或替换，也应属于本发明的保护范围。It should be noted that the above description is not intended to limit the present invention, and the present invention is not limited to the above-mentioned examples. Those skilled in the art may make changes, modifications, additions or replacements within the scope of the present invention. It should also belong to the protection scope of the present invention.
Claims (7)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510475741.XA CN105204490B (en) | 2015-08-06 | 2015-08-06 | The diagnostic method for the stand-by power consumption intelligent diagnosis system classified based on Ensemble feature selection |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510475741.XA CN105204490B (en) | 2015-08-06 | 2015-08-06 | The diagnostic method for the stand-by power consumption intelligent diagnosis system classified based on Ensemble feature selection |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105204490A true CN105204490A (en) | 2015-12-30 |
| CN105204490B CN105204490B (en) | 2017-10-20 |
Family
ID=54952231
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510475741.XA Expired - Fee Related CN105204490B (en) | 2015-08-06 | 2015-08-06 | The diagnostic method for the stand-by power consumption intelligent diagnosis system classified based on Ensemble feature selection |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105204490B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105719006A (en) * | 2016-01-18 | 2016-06-29 | 合肥工业大学 | Cause-and-effect structure learning method based on flow characteristics |
| WO2018023709A1 (en) * | 2016-08-05 | 2018-02-08 | 黄新勇 | Method and system for monitoring and prompting energy consumption in broadcast network |
| CN114066071A (en) * | 2021-11-19 | 2022-02-18 | 厦门大学 | An energy consumption-based power parameter optimization method, terminal device and storage medium |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US4380243A (en) * | 1980-01-16 | 1983-04-19 | Braley Charles A | Overflow control system |
| WO1995010875A1 (en) * | 1993-10-12 | 1995-04-20 | Fischer, Helmut | Power supply for standby circuits of electrical devices |
| CN1987697A (en) * | 2006-12-22 | 2007-06-27 | 浙江大学 | Fault diagnostic system and method for under industrial producing process small sample condition |
| CN101873009A (en) * | 2010-07-12 | 2010-10-27 | 浙江昱能光伏科技集成有限公司 | Home intelligent power monitoring system |
| CN102307126A (en) * | 2011-08-15 | 2012-01-04 | 武汉安通科技产业发展有限公司 | Multi-network accessed multifunctional digital home intelligent terminal and control method thereof |
| CN103135464A (en) * | 2013-01-23 | 2013-06-05 | 山东建筑大学 | Electricity saving control system and method of standby electric equipment |
| CN103365269A (en) * | 2013-06-28 | 2013-10-23 | 南京邮电大学 | Wireless sensor network-based intelligent household electrical appliance multimode remote monitoring system and method |
| CN103760862A (en) * | 2014-01-17 | 2014-04-30 | 东莞维锐光电科技有限公司 | Smart home management method and smart home management system |
-
2015
- 2015-08-06 CN CN201510475741.XA patent/CN105204490B/en not_active Expired - Fee Related
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US4380243A (en) * | 1980-01-16 | 1983-04-19 | Braley Charles A | Overflow control system |
| WO1995010875A1 (en) * | 1993-10-12 | 1995-04-20 | Fischer, Helmut | Power supply for standby circuits of electrical devices |
| CN1987697A (en) * | 2006-12-22 | 2007-06-27 | 浙江大学 | Fault diagnostic system and method for under industrial producing process small sample condition |
| CN101873009A (en) * | 2010-07-12 | 2010-10-27 | 浙江昱能光伏科技集成有限公司 | Home intelligent power monitoring system |
| CN102307126A (en) * | 2011-08-15 | 2012-01-04 | 武汉安通科技产业发展有限公司 | Multi-network accessed multifunctional digital home intelligent terminal and control method thereof |
| CN103135464A (en) * | 2013-01-23 | 2013-06-05 | 山东建筑大学 | Electricity saving control system and method of standby electric equipment |
| CN103365269A (en) * | 2013-06-28 | 2013-10-23 | 南京邮电大学 | Wireless sensor network-based intelligent household electrical appliance multimode remote monitoring system and method |
| CN103760862A (en) * | 2014-01-17 | 2014-04-30 | 东莞维锐光电科技有限公司 | Smart home management method and smart home management system |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105719006A (en) * | 2016-01-18 | 2016-06-29 | 合肥工业大学 | Cause-and-effect structure learning method based on flow characteristics |
| WO2018023709A1 (en) * | 2016-08-05 | 2018-02-08 | 黄新勇 | Method and system for monitoring and prompting energy consumption in broadcast network |
| CN114066071A (en) * | 2021-11-19 | 2022-02-18 | 厦门大学 | An energy consumption-based power parameter optimization method, terminal device and storage medium |
| CN114066071B (en) * | 2021-11-19 | 2024-09-13 | 厦门大学 | Power parameter optimization method based on energy consumption, terminal equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105204490B (en) | 2017-10-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Xia et al. | A real-time monitoring system based on ZigBee and 4G communications for photovoltaic generation | |
| Shinde et al. | Review on-IoT based environment monitoring system | |
| CN203705893U (en) | Intelligent greenhouse-management system based on Internet of Things | |
| CN107728680A (en) | A kind of remote pig house environment multiparameter measurement and control system and its method based on LoRa | |
| CN102736599A (en) | Real-time interaction intelligent terminal control system based on internet of things technology | |
| CN102523272B (en) | Electric energy measuring, controlling and protecting system | |
| CN108593990A (en) | A kind of stealing detection method and application based on electric power users electricity consumption behavior pattern | |
| CN202995383U (en) | Real-time interactive intelligent terminal control system based on internet-of-things technology | |
| CN201955886U (en) | Agricultural environment information collection system based on wireless sensor network | |
| WO2013061300A1 (en) | A distributed monitoring method for achieving energy efficiency in buildings and apparatus therefor | |
| CN101715242A (en) | Intensive aquaculture wireless transmission convergent node device and information fusion method | |
| CN105204490B (en) | The diagnostic method for the stand-by power consumption intelligent diagnosis system classified based on Ensemble feature selection | |
| CN107681661A (en) | A kind of instant inserted and instant using type energy interacted system | |
| CN113220751A (en) | Metering system and evaluation method for multi-source data state quantity | |
| CN110706471A (en) | Intelligent electricity information acquisition method based on capacitive isolation multichannel redundancy technology by adopting edge algorithm | |
| CN104994536B (en) | A kind of configuration method and portable configuration device towards WSCN nodes | |
| CN104807498A (en) | Farmland environment wireless monitoring system based on Zigbee and 3G technology | |
| Luo et al. | The implementation of wireless sensor and control system in greenhouse based on ZigBee | |
| CN110191481A (en) | A Method and System for Evaluating Link Quality Using Overlimit Fast Decision Tree | |
| CN107948262A (en) | Intelligent greenhouse management system based on Internet of Things | |
| Tong et al. | Surrogate model-based energy-efficient scheduling for LPWA-based environmental monitoring systems | |
| CN203376857U (en) | Remote meter reading system based on wireless sensor network | |
| CN110824384A (en) | An online transformer fault diagnosis system based on artificial intelligence extreme learning machine | |
| CN104458340B (en) | Many intake weighting water intake systems and method based on wireless sensor network | |
| Moayedi et al. | Energy Optimization Technologies in Smart Homes |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20171020 Termination date: 20180806 |