WO2014022955A1

WO2014022955A1 - Method for managing energy efficiency under periodic load in building

Info

Publication number: WO2014022955A1
Application number: PCT/CN2012/079700
Authority: WO
Inventors: 刘岩
Original assignee: 珠海派诺科技股份有限公司
Priority date: 2012-08-05
Filing date: 2012-08-05
Publication date: 2014-02-13
Also published as: CN103890806B; CN103890806A

Abstract

By analyzing the energy use of a periodically energy-consuming device in a building, the solution determines the energy consumption characteristics of the periodic device, and describes the energy consumption of the periodic device in the building in the form of a Fourier series. By analyzing the coefficient of a series expansion, the solution evaluates the energy use of the device, finds out promptly if the energy use of the device is abnormal or unreasonable, and at the same time can identify whether the reason for periodic load fluctuation belongs to normal energy use.

Description

Description of a building periodic load energy efficiency management method

^ ^·Field

The present invention relates to energy efficiency management, and more particularly to energy efficiency management in building energy use. Background

With the urbanization process in China, the proportion of building energy consumption has been increasing. The rationality of construction equipment has become the focus of attention. The proportion of building energy is increasing, and its impact on the overall energy use of society is becoming more and more obvious. At present, the high-energy buildings in buildings are mainly concentrated in large urban public buildings, of which large public buildings are defined as: 20,000 building area Office buildings, commercial buildings, tourist buildings, science and education buildings, communication buildings and transportation houses of more than square meters. In order to control the energy use of buildings, domestic and foreign experts and scholars analyze the rationality of building energy, and use some methods to make discriminative methods. These methods include: total amount control method, homogeneous energy method for similar buildings; time division ^ ^ standard method, use Energy density method, instantaneous value method, regional comparison method, year-on-year method, ring method, etc.

These methods play a role in controlling energy use and finding that the energy consumption is unreasonable, but these methods have a series of deficiencies. For example, the total amount control method cannot consider the energy use characteristics and energy use characteristics of buildings. The set energy management objectives are often changed due to the choice of samples, which is not conducive to the management of energy use in buildings. For regional comparison methods, Factors such as lighting, floors, walls, window sizes, curtain walls, etc. in different areas of the same building will have an impact on the regional energy use, and the actual energy efficiency level of the area cannot be well reflected by regional comparison. At present, the above method evaluates the energy use of buildings with the amount of energy used as the evaluation parameter, and does not fully analyze the information contained in the energy cycle of the building.

In fact, most of the large-scale public buildings have their own functions. For example, office buildings show daily rules according to their working hours. Tourism buildings alternate with the seasons of the tourist season and the off-season. Wenwei follows the nature of the student's school and holiday schedule. Therefore, it is necessary to propose a method for building energy efficiency management using building energy cycle information. Summary of the invention The invention proposes a building periodic load energy efficiency management method, and uses the Fourier transform to extract the periodic load data of the building, extracts the frequency information in the building energy consumption data, and analyzes the frequency i to find the building. Unreasonable, and then find the loopholes in energy waste during the use of the building.

The method mainly includes the following steps:

In the first step, from the monitoring system, the energy of the device to be judged is collected, and the energy consumption data may be the power consumption of the device for a period of time, and the device may be a certain electrical device in the building, or For a certain type of electrical equipment in the building;

In the second step, an appropriate judging period is selected, and the judging period may be days, weeks, months, years, etc., and the collected energy consumption data is divided into a plurality of data groups according to the judging period, and each data group includes the same judging period. Can be ^3⁄4 according to;

The third step is to define the attributes of each data group, such as working hours, non-working hours, special holidays, time, etc.

In the fourth step, the Ken group with the same attribute is assumed to be N, and the N data groups are sorted in chronological order. Each M consecutive data groups form a data group set, which constitutes N-M+1. Data set collection; for example, there are 10 data groups with the same property, namely {1, 2, 3, 4, 5, 6, 7, 8, 9, 10}, and each 4 consecutive data groups form a data. The set consists of {1,2,3,4}, {2,3,4,5}, {3,4,5,6}, {4,5,6,7}, {5,6,7 , 8}, {6, 7, 8, 9}, {7, 8, 9, 10}, a total of 10-4+1 = 7 data sets;

In the fifth step, the N-M+1 ^3⁄4 data set is separately normalized; in the sixth step, the normalized N-M+1 data sets are respectively subjected to Fourier Transforming, transforming the time domain signal of the data into a frequency domain signal;

In the seventh step, the corresponding amplitudes of the frequency domain signals of the N-M+1 energy groups of the Fourier transform are summed and averaged to obtain the mean value of the amplitude values of the frequency domain signals. ;

In the eighth step, the amplitudes of the frequency domain signals of the N-M+1 capable groups are respectively compared with the corresponding mean values, and it is determined that the Kens data group of the same attribute is used in an unreasonable data group. The collection and the ^^ data group. DRAWINGS

Figure 1 shows the data collected per unit area of the air conditioning unit;

Figure 2 is a set of energy data sets of the same attribute after Fourier transform, all frequency pairs The distribution of the mean value of the magnitude of Fn[k]. Implementation

This analysis uses the Discrete Fourier Transform (DFT) method to analyze the energy efficiency management of a building's periodic load for an air-conditioning electricity in a building. The electricity consumption data of the air-conditioning distribution box is taken out from the database of the energy consumption monitoring system, and the hourly energy consumption data of the air-conditioning distribution box in April 2011 is retrieved. The judging period is selected as one day, that is, starting from 1 o'clock on the day and 24 o'clock as the end point, dividing the air-conditioning distribution box data into energy consumption data groups, and each data group has 24 hour energy consumption data. By querying the calendar, in April 2011, April 2, 3, 9, 10, 16, 17, 23, 24, 24, 30 are non-working days, so the above dates correspond. The attribute of the energy data group is a non-working day, and the attribute of the data group corresponding to the remaining date is a work 曰.

Take the energy data group with the attribute as the working day and sort it by time. They are 1 曰, 4 曰, 5 曰, 6 曰, 7 曰, 8 曰, 11 曰, 12 曰, 13 曰, 14 曰, 15 曰, 18 曰, 19 曰, 20 曰, 21 曰, 22 曰, 25 曰, 26 曰, 27 曰, 28 曰, 29 3 3⁄4 groups, a total of 21 can ^ 3⁄4 groups.

For the 21 energy consumption data groups, four sequentially connected data groups are combined to form 18 data group sets. The data of each data group in each data set is normalized, that is, the energy consumption data of 24 hours per day is divided by the maximum value of the current energy consumption data to obtain 24 hours of normalized data.

The four energy consumption data sets normalized in one energy consumption data set are expressed as: {ml, m2, m3,..., m24}; {pl,p2,p3,...,p24} ; {al,a2,a3,... ,a24}; {bl,b2,b3,b4,...,b24}. Then, the 24 hours data normalized in each energy consumption data group is expanded to 32 data (filled with 0, supplemented by 8 0s), and the 4 energy consumption data groups become {ml, respectively. M2,m3,...,m24,0,0,0,0,0,0,0,0}; {pl,p2,p3,... ,p24,0,0,0,0,0, 0,0,0 }; { al ,a2,a3, · .. ,a24,0,0,0,0,0,0,0,0 }; {bl,b2,b3,b4,..., B24,0,0,0,0,0,0,0,0}. The data of the four energy consumption data sets are connected end to end to obtain the data of the energy consumption data set, that is, x[n]={ml, m2, ..., m24, 0, 0, 0, 0, 0, 0, 0,0,p 1 ,p2, ... ,p24,0,0,0,0,0,0,0,0,al ,a2 •••,a24,0,0,0,0,0, 0,0,0,bl,b2 ,· · .,1)24,0,0,0,0,0,0,0,0 } (where n=0,l,2,...,127) , the data set set x[n] is subjected to Fourier transform, and the obtained Fu The Lie series expansion, and find the magnitude En[k] of the kth item in the expansion.

Since the N-point sequence {X[M] }0 ≤ W < N, its discrete Fourier transform (DFT) is

_ _; 3⁄4L _ni ,

x[k] = ^ e ^N x[«], fc = 0,1,2,..., N -l. where e is the base of the natural logarithm and i is the imaginary unit. Then, for the above-mentioned 128 point sequence of the present invention, { ^Χ ["] } 0 ^≤ "^{< 128} , its discrete Fourier transform ( ) is x[k] = ^e ¹²⁸ x[n], k = 0,l,2,...,m.

En[k]: x[k] fc = 0,1,2,...,127.

The amplitude of the kth term, that is, the modulus of ^x [ ^fc ], where En[0] is the DC component amplitude, and Fourier transform is performed on the above 18 sets of energy consumption data sets, respectively, and 18 Fourier series can be obtained. In the expansion mode, the amplitudes of the corresponding kth items in each expansion, that is, the moduli of ^x [ ^fc ] are respectively summed and averaged to obtain 128 average values Fn[k].

Figure 2 shows the distribution of the energy value data set of the weekday attribute after the Fourier transform mean Fn[k]. Since the value of Fn[k] is extremely small as k increases, it is only shown in Figure 2. The data of Fn[0] to Fn[32].

We select the DC component amplitude En[0] in each data set corresponding to the mean value of the DC component amplitude, and the AC component in each data set corresponding to the maximum value of the average of the AC component amplitudes. The amplitude (hereinafter referred to as the target octave component amplitude) is used as a research object to judge the unreasonable energy consumption data in the energy consumption data group of the same attribute.

As can be seen from Fig. 2, in this embodiment, among the above 18 sets of energy consumption data sets having the weekday attribute, the maximum value of the mean values of the AC component amplitudes is Fn [5].

Through analysis, we can obtain that when the set of data sets is subjected to Fourier transform and the DC component amplitude En[0] is greater than Fn[0], if the target octave component amplitude En[5] is also greater than its corresponding mean Fn [5] indicates that the overall energy consumption of the equipment is increased, which is caused by the increase in overall load energy consumption. This problem is not within the scope of load cyclical research. If the DC component amplitude En[0] is greater than Fn[0], the amplitude En[5] of the target frequency component is less than its corresponding value. The mean value of Fn[5] indicates that the energy consumption is periodically abnormal, which is the target of the energy consumption periodicity we analyzed. We compare the DC component amplitude, the target octave component amplitude and its corresponding mean value to determine the abnormal energy consumption in the energy consumption data set.

Then, in the set of 18 energy consumption data sets, the DC component amplitudes En[0], En[5] and the corresponding Fn[0] and En[5, respectively, of the set of energy consumption data sets are respectively subjected to Fourier transform. ] to compare, judge the use of unreasonable:

Criterion 1: If the DC component amplitude En[0] exceeds 1.1 times Fn[0], when the target multiplication component amplitude En[5] is equal to or greater than the mean Fn[5], then the energy consumption data set is No abnormal energy consumption occurred.

Criterion 2: If the DC component En[0] exceeds 1.1 times Fn[0], the target octave component amplitude En[5] is less than the mean Fn[5] and satisfies En[5] < 0.98*Fn[5] In the fourth energy consumption data group included in the energy consumption data group set, the last energy consumption data group exhibits abnormal energy consumption, and the abnormal energy consumption performance is increased in the energy consumption period of the day. There are loads in the electrical equipment that should be turned off and not turned off.

Criterion 3: If the DC component En[0] exceeds 1.1 times Fn[0], the target octave component amplitude En[5] is less than the mean Fn[5] and satisfies En[5] < 0.96*Fn[5] In the fourth energy consumption data group included in the energy consumption data group set, the last energy consumption data group exhibits abnormal energy consumption, and the abnormal energy consumption performance is that the energy consumption exceeds the standard during the valley period of the current energy consumption. There are a large number of electrical equipment that should be turned off without being turned off.

Criterion 4: If the DC component En[0] exceeds 1.1 times Fn[0], the target octave component amplitude En[5] is less than the mean Fn[5] and satisfies En[5] < 0.92*Fn[5] In the fourth energy consumption data group included in the energy consumption data group set, the last energy consumption data group exhibits abnormal energy consumption, and the abnormal energy consumption performance is that the energy consumption in the valley period of the day energy consumption exceeds the standard, almost The load on all devices should be turned off and not turned off.

For the energy consumption data group having the non-working day attribute and the special holiday attribute, the above-mentioned judgment is respectively performed, and the unreasonable energy consumption in the energy consumption data of the non-working day attribute and the special holiday attribute can be judged.

The invention solves the previous influence on the energy consumption, is not affected by the change caused by the energy growth, fully exerts the periodicity of the energy fluctuation, and uses the mathematical transformation combined with the building energy characteristics to judge the rationality of the building energy.

This technology can effectively manage the cyclical load of the office building and prevent it. The use of energy can be wasted. Furthermore, the data of the same acquisition unit is analyzed to prevent interference with the judgment result due to changes in the environment, nature, and region of the analysis data.

Claims

Claim

A construction cycle energy efficiency management method, comprising the following steps: First, from the monitoring system, the data of the i-leveling device is retrieved;

In the second step, an appropriate judging period is selected, and the collected energy consumption data is divided into a plurality of groups according to the judging period, and each data group includes the credits in the same judging period;

The third step is to define the attributes of each data group;

In the fourth step, the groups of the same attribute are assumed to be N, and the N data groups are sorted in chronological order, and each M consecutively connected data groups form a data group set, which constitutes N-M+1. Data set collection;

In the seventh step, the Fourier-transformed N-M+1 frequency domain signals of the set of frequency domain signals are averaged and summed to obtain the N-M+1 The mean of the amplitudes of the frequency domain signals of the set of data sets;

In the eighth step, the amplitudes of the frequency domain signals of the N-M+1^3⁄4 group are respectively compared with the corresponding mean values, and the energy consumption data groups with unreasonable energy consumption in the energy consumption data group of the same attribute are determined. The collection and the Ken 3⁄4 group.

2. The method of claim 1 wherein said teachings are electrical power usage of said device.

3. The method of claim 1 wherein the evaluation period is day, week, month or year.

4. The method of claim 1, wherein the attributes of the data set are work, non-work, or holiday.