CN104915562A - Building energy efficiency diagnosis method and system - Google Patents

Building energy efficiency diagnosis method and system Download PDF

Info

Publication number
CN104915562A
CN104915562A CN201510320498.4A CN201510320498A CN104915562A CN 104915562 A CN104915562 A CN 104915562A CN 201510320498 A CN201510320498 A CN 201510320498A CN 104915562 A CN104915562 A CN 104915562A
Authority
CN
China
Prior art keywords
data
electricity
history
forecast
subitem
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
Application number
CN201510320498.4A
Other languages
Chinese (zh)
Other versions
CN104915562B (en
Inventor
张皓
苑登阔
熊真真
陈莹
屠盛春
郭杰
郁松龄
葛燕龙
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
SHANGHAI TWENTY-FIRST ENERGY SAVING TECHNOLOGY CO., LTD.
Original Assignee
Shanghai Twenty-First Energy Saving Technology Co Ltd
Shanghai New Changning Low-Carbon Investment Management Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Shanghai Twenty-First Energy Saving Technology Co Ltd, Shanghai New Changning Low-Carbon Investment Management Co Ltd filed Critical Shanghai Twenty-First Energy Saving Technology Co Ltd
Priority to CN201510320498.4A priority Critical patent/CN104915562B/en
Publication of CN104915562A publication Critical patent/CN104915562A/en
Application granted granted Critical
Publication of CN104915562B publication Critical patent/CN104915562B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Abstract

The invention provides a building energy efficiency diagnosis method and system. The method includes: determining a prediction model and precision thereof based on model training according to historical electric breakdown data, corresponding historical external parameters and searched historical specific rule information; selecting diagnosis thresholds for the predication data and the precision thereof; acquiring corresponding predicated electric breakdown data through the predication model according to current external parameters and the searched specific rule information; comparing the predicated electric breakdown data and current electric breakdown data so as to determine energy efficiency status on the basis of the diagnosis thresholds. Therefore, accuracy in energy consumption abnormity determination is improved in building energy efficiency monitoring and management, energy consumption abnormities can be determined in time, and efficiency of energy efficiency monitoring and management is increased. Due to the fact that abnormal energy consumption is located timely and accurately, building managers can be practically helped to take action timely, and energy waste is reduced.

Description

Building efficiency diagnostic method and system
Technical field
The application relates to data monitoring diagnostic field, particularly relates to building efficiency diagnostic method and system.
Background technology
Along with the trend of energy-saving and emission-reduction, building energy consumption measures day by day specification, and along with the development of Internet technology, corresponding all kinds of building energy efficiency monitoring control and management platform is also progressively set up.These platforms can by carrying out the on-the-spot real-time process such as monitoring, analysis to energy for building, energy dissipation point in Timeliness coverage energy for building, instruct and assist operation maintenance personnel to implement administration of energy conservation work, and find and determine energy dissipation point, need in real time/diagnose in time, exactly the technology of this building efficiency (such as: various building energy consumptions etc.).In actual motion, there is the problems such as building energy consumption metering kind is single, common management person analysis ability finite sum professional diagnosis is undermanned when adopting manual analysis, a large amount of power consumption data are caused to pile up and from data, accurately and timely cannot excavate implicit information, and energy-conservation constructive path cannot be provided, and then data monitoring (collection) system that waste initial cost builds, and still accurately and timely cannot find the energy dissipation point of energy for building, efficiency monitoring cannot be met and assist energy-conservation effect.
Thus, need the existing efficiency monitoring management platform that improvement is perfect, by gathering power consumption data to energy for building metering and carrying out automatic diagnosis (without the need to people's Analysis of Division of Labor) and locate obtaining with the abnormal position of energy and time.Build this automatic efficiency diagnostic techniques without the need to manual analysis, more accurately and timely really to consume exception surely, improve the efficiency of efficiency monitoring management, reduce energy dissipation unnecessary in building, thus save efficiency, avoid waste.
Summary of the invention
The fundamental purpose of the application is to provide a kind of building efficiency diagnostic method and system, promotes the abnormal accuracy problems determined of energy consumption, and then improve the problem determining the efficiency supervisory efficiency of energy consumption exception in time to solve in the monitoring management of building efficiency.Make can locate abnormal energy consumption timely and accurately without the need to manual analysis, conscientiously help the timely action of architectural control personnel, thus reduce energy dissipation.
The application provides a kind of building efficiency diagnostic method on the one hand, comprise: according to history electricity subitem data, gather described electricity subitem data time the corresponding historical external parameter that obtains and the history ad hoc rules information of collection, based on model training, to determine forecast model and precision thereof; Threshold selection is carried out, to draw diagnostic threshold to forecast model and precision thereof; The corresponding external parameter obtained during electric subitem data current according to collection and the ad hoc rules information of collection, obtain the electric data of itemizing of corresponding prediction by described forecast model; Based on described diagnostic threshold, contrast described prediction electricity subitem data and the corresponding current electricity subitem data gathered, to determine efficiency situation.
Wherein, according to history electricity subitem data, gather described electricity subitem data time the corresponding historical external parameter that obtains and the history ad hoc rules information of collection, based on model training, to determine forecast model and precision thereof, also comprise: according to the analysis to every history electricity subitem data, and in conjunction with corresponding historical external parameter and/or history ad hoc rules information, set forecast model to be trained; Using the history ad hoc rules information of history electricity subitem data, historical external parameter and collection accordingly as the sample of model training, the forecast model of setting is trained, determines model parameter and the model accuracy thereof of forecast model.
Wherein, according to the analysis to every history electricity subitem data, and in conjunction with corresponding historical external parameter and/or history ad hoc rules information, set forecast model to be trained, comprise: to its power consumption feature of history electricity subitem data analysis of input, according to dispersion and/or the degree of correlation, in conjunction with corresponding historical external parameter and/or history ad hoc rules information, setting uses regression model and/or feature averaging model as forecast model to be trained.
Wherein, described external parameter comprises the temperature in meteorologic parameter; Described ad hoc rules information comprises the job specification information in social information; Electricity subitem data, by the frequency of data collector according to 15 minutes/time, the electrisity consumption of surveyed electricity consumption branch road, electric current and/or voltage are gathered, each electric branch road is divided to the power consumption data obtaining each branch road in specific electricity consumption subitem, and utilize data aggregate to obtain each branch road power consumption data of every day, described each branch road power consumption data to be itemized data as electricity; Meteorologic parameter, directly obtained the temperature data of outdoor dry-bulb temperature by the frequency collection of 15 minutes/time by the Temperature sampler of meteorological data collection end, and by data average treatment, the multiple temperature datas gathered in special time period are averaged, calculate this temperature, as meteorologic parameter; Or, indirectly searched for by weather station, the building location database carrying out efficiency diagnosis and obtain this temperature; Social information, by social information's data acquisition end, arranges to give job specification information the same day to corresponding electricity subitem data place according to country's legal festivals and holidays.
Wherein, based on described diagnostic threshold, contrast described prediction electricity subitem data and the corresponding current electricity subitem data gathered, to determine efficiency situation, comprise: the error of computational prediction electricity subitem data and corresponding current electricity subitem data, according to described diagnostic threshold scope, draw efficiency diagnostic result, and export described efficiency situation.
Wherein, the error of computational prediction electricity subitem data and corresponding current electricity subitem data, according to described diagnostic threshold scope, draw efficiency diagnostic result, and export described efficiency situation, comprising: the relative error of computational prediction electricity subitem data and corresponding current electricity subitem data and absolute error, when exceeding the relative threshold upper limit in relative error or comparing absolute error more in limited time lower than under relative threshold, and set up grade separately according to the size of absolute error, the alarm of showed different.
The application provides a kind of building efficiency diagnostic system on the other hand, comprise: model determining device, for according to history electricity subitem data, gather described electricity subitem data time the corresponding historical external parameter that obtains and the history ad hoc rules information of collection, based on model training, to determine forecast model and precision thereof; Diagnostic threshold device, for carrying out Threshold selection, to draw diagnostic threshold to forecast model and precision thereof; Prediction unit, for the ad hoc rules information according to the corresponding external parameter obtained when gathering current electricity subitem data and collection, obtains corresponding prediction electricity subitem data by described forecast model; Contrast diagnostic device, for based on described diagnostic threshold, contrasts described prediction electricity subitem data and the corresponding current electricity subitem data gathered, to determine efficiency situation; Wherein, model determining device, comprising: training pattern device is treated in setting, for the analysis according to data of itemizing to every history electricity, and in conjunction with corresponding historical external parameter and/or history ad hoc rules information, sets forecast model to be trained; And model training apparatus, for using the sample of the history ad hoc rules information of history electricity subitem data, corresponding historical external parameter and collection as model training, the forecast model of setting is trained, determines model parameter and the model accuracy thereof of forecast model.
Wherein, described model determining device, comprise: to its power consumption feature of history electricity subitem data analysis of input, according to dispersion and/or the degree of correlation, in conjunction with corresponding historical external parameter and/or history ad hoc rules information, setting uses regression model and/or feature averaging model as forecast model to be trained.
Wherein, described external parameter comprises the temperature in meteorologic parameter; Described ad hoc rules information comprises the job specification information in social information; Electricity subitem data, by the frequency of data collector according to 15 minutes/time, the electrisity consumption of surveyed electricity consumption branch road, electric current and/or voltage are gathered, each electric branch road is divided to the power consumption data obtaining each branch road in specific electricity consumption subitem, and utilize data aggregate to obtain each branch road power consumption data of every day, described each branch road power consumption data to be itemized data as electricity; Meteorologic parameter, directly obtained the temperature data of outdoor dry-bulb temperature by the frequency collection of 15 minutes/time by the Temperature sampler of meteorological data collection end, and by data average treatment, the multiple temperature datas gathered in special time period are averaged, calculate this temperature, as meteorologic parameter; Or, indirectly searched for by weather station, the building location database carrying out efficiency diagnosis and obtain this temperature; Social information, by social information's data acquisition end, arranges to give job specification information the same day to corresponding electricity subitem data place according to country's legal festivals and holidays.
Wherein, contrast diagnostic device, also comprise: the relative error of computational prediction electricity subitem data and corresponding current electricity subitem data and absolute error, according to described diagnostic threshold scope, draw efficiency diagnostic result, and exporting described efficiency situation, the size according to absolute error sets up grade separately, the alarm of showed different.
Compared with prior art, according to the efficiency diagnostic techniques scheme of the application, can in time by the energy consumption data in the energy for building environment of monitoring monitoring collection in real time in other words, and find and determine the abnormal energy consumption that occurs in energy for building, carried to existing building energy efficiency management platform, according to diagnosis algorithm, automatic analysis diagnosis is carried out to mass data, Timeliness coverage and the efficiency monitoring management efficiency determining energy consumption exception can be improved, promote the accuracy of diagnosis, thus conscientiously help the timely action of architectural control personnel, reduce energy dissipation, save efficiency, avoid the waste of manpower financial capacity.
Accompanying drawing explanation
Accompanying drawing described herein is used to provide further understanding of the present application, and form a application's part, the schematic description and description of the application, for explaining the application, does not form the improper restriction to the application.In the accompanying drawings:
Fig. 1 is the process flow diagram of an embodiment of the building efficiency diagnostic method of the application;
Fig. 2 is the process flow diagram of the embodiment setting forecast model in the method for the application;
Fig. 3 is the schematic diagram of the embodiment setting forecast model in the application's method;
Fig. 4 is the schematic diagram of an embodiment of efficiency prediction and diagnosis in the application's method;
Fig. 5 is the structured flowchart of an embodiment of the building efficiency diagnostic system of the application.
Embodiment
The main thought of the application is, obtain power consumption data based on metering separate, be aided with the external parameter of impact building efficiency and set up corresponding energy consumption forecast model, Real-Time Monitoring and the actual state drawing building subitem electricity consumption data, the i.e. efficiency of diagnosis building effectively, to determine abnormal energy consumption exactly, thus adjust in time and get rid of abnormal, save the energy, reduce the wasting of resources.
For making the object of the application, technical scheme and advantage clearly, below in conjunction with the application's specific embodiment and corresponding accompanying drawing, technical scheme is clearly and completely described.Obviously, described embodiment is only some embodiments of the present application, instead of whole embodiments.Based on the embodiment in the application, those of ordinary skill in the art are not making the every other embodiment obtained under creative work prerequisite, all belong to the scope of the application's protection.
According to the embodiment of the application, provide a kind of building efficiency diagnostic method.
Flow process Figure 100 of an embodiment of the building efficiency diagnostic method of the application with reference to figure 1, Fig. 1.
In step S110, according to history electricity subitem data, gather described electricity subitem data time the corresponding historical external parameter that obtains and the history ad hoc rules information of collection, based on model training, to determine forecast model and precision thereof.
Wherein, electricity subitem data, external parameter, ad hoc rules information are all basic datas.
In one embodiment, electricity subitem data at least comprise energy use (mainly comprising electric energy use and electricity consumption) in various building (as: public building etc.), carry out the power consumption data that metering separate obtains, such as: illumination and socket electricity consumption, air conditioning electricity, power electricity consumption, the special power consumption data with other electricity consumptions etc.Wherein, public building can include but not limited to office building, market building, comprehensive office building etc.Wherein, illumination and socket electricity consumption as: public domain throw light on, the electricity consumption of several subitems such as corridor and emergency lighting, outdoor landscape throw light on, vault light, light socket; Power electricity consumption is as the electricity consumption of the several subitem of elevator, water pump, ventilation blower and Dynamic Synthesis; Special with other electricity consumptions as the electricity consumption of: information center, laundry, kitchen and dining room, swimming pool bathroom, gymnasium and other several subitems.
Owing to building the relative microcosmic of diagnosis of efficiency, be concerned about the situation (normal, exception, out-of-the way position etc.) of single the interior energy of building, and building in energy consumption system various, complicated by energy feature, the air-conditioning system of such as large-scale public construction, device category is many, it is different to form, and causes it higher to data dependency degree, and the kind of data, quality and quantity can affect the diagnosis degree of depth and precision that even determine efficiency situation.Utilize the electricity subitem data of metering separate mode Real-time Collection in the present embodiment, compared to existing energy depot investigation method, (generaI investigation mode obtains, a kind of method that update cycle is long) and data resolution method (by the active power of analysis to measure summary table, reactive power and harmonic wave change etc., be aided with common device electrical feature and draw equipment component power consumption, calculation of complex and a kind of not blanket method), more clearly can know the real-time status of all kinds of energy equipment in building, contribute to the efficiency of efficiency diagnosis and the raising of accuracy.
In one embodiment, the collection of electricity subitem data can utilize power consumption data acquisition end, namely comprises conventional data collector (data acquisition unit etc.), and the mode in conjunction with building electricity consumption subitem model and data aggregate is carried out.Such as, data collector can gather according to the electrisity consumption (even comprising corresponding electric current, voltage etc.) of frequency to surveyed electricity consumption branch road of 15 minutes/time, according to subitem model for building, each electric branch road is divided to the power consumption data obtaining each branch road in specific electricity consumption subitem, further, data aggregate can be utilized to obtain by each branch road power consumption data of sky/every day.The power consumption data gathered are metering separates, and thus these power consumption data to be itemized data as electricity, namely each branch road electricity subitem data, every day each branch road electricity subitem data.The electricity subitem data in the past gathered preserve, and as historical data and the history electricity subitem data of electricity subitem data, the needs of Real-time Collection are then current electricity subitem data according to the electricity subitem data that it carries out efficiency diagnosis.
In one embodiment, the corresponding external parameter obtained when gathering described electricity subitem data and the ad hoc rules information of collection can be the while of gathering electricity subitem data successively, the out door climatic parameter collected and the social information of collection.Out door climatic parameter at least can comprise the outer dry-bulb temperature of building respective compartments, and social information at least can comprise the job specification information of giving every day according to country's legal festivals and holidays arrangement, etc.
Further, the collection of out door climatic parameter can utilize meteorological data collection end, namely comprises Temperature sampler and carries out in conjunction with the processing mode that data are average.Such as: Temperature sampler directly obtains the temperature data of outdoor dry-bulb temperature by the frequency collection of 15 minutes/time, by data average treatment, the multiple temperature datas gathered in special time period are averaged, calculate the representation temperature by sky, as out door climatic parameter (external parameter); Also this temperature can indirectly be obtained by building weather station, location database.The itemize acquisition time section of data (power consumption data) of itself and electricity is corresponding, equally, gathered in the past also as history out door climatic parameter (historical external parameter) preserve, Real-time Collection then as current outdoor meteorologic parameter (current external parameter) for current efficiency diagnostic process.
Further, the collection of social information can utilize social information's data acquisition end/search (collection) to hold, gave job specification information according to country's legal festivals and holidays arrangement the same day to the corresponding history electricity subitem data place same day or current electricity subitem data place, and namely can ensure that this social information (specifically can be job specification information etc.) and electric data of itemizing are corresponding within the time period at the same time.Such as: the job specification information of corresponding history electricity subitem data or say that social information preserves as history social information preservations i.e. history ad hoc rules information, and corresponding current electricity itemize data job specification information or say that social information is as the preservation of current social information and the preservation of current ad hoc rules information.
In one embodiment, based on model training, to determine forecast model and precision thereof, can first according to the history electricity subitem data of preserving, gather described electricity subitem data time the corresponding historical external parameter that obtains and the history ad hoc rules information of collection carry out Model Selection process, then complete model training and determine corresponding forecast model.Below in conjunction with setting the schematic diagram setting an embodiment of forecast model in the application's method shown in flow process Figure 200 of an embodiment of forecast model and Fig. 3 in the method for the application shown in Fig. 2, descriptive model select and model training to determine an example of forecast model.
Step S210, according to the analysis to every history electricity subitem data, and in conjunction with corresponding historical external parameter and/or history ad hoc rules information, sets forecast model to be trained.
Wherein, to every history electricity subitem data analysis of input, corresponding power consumption feature can be obtained.Such as power consumption is high or low.In conjunction with corresponding historical external parameter (meteorologic parameter as at that time: temperature), history ad hoc rules information (as social information: job specification information), select the power consumption forecast model of determining can carry out accordingly training.
Such as: the electrisity consumption of history electricity subitem data representation is on weekdays high, the electrisity consumption of the history electricity subitem data representation of nonworkdays is low, time time external temperature is high or low, high, the external temperature of electrisity consumption is suitable, electrisity consumption is low, etc., can choice for use regression model or feature averaging model as forecast model to be trained.Such as, analysis of history electricity subitem data find working day and nonworkdays difference not quite, and feature averaging model can be selected as forecast model; Analysis of history electricity subitem data find that temperature impact is very large, and data discrete is serious, can select regression model; Even can according to the power consumption feature of the history electricity subitem data of analysis, the data of overall use regression model, some part are not discrete and adopt two class models combinations of feature averaging model to predict to obtain model parameter and precision thereof respectively to it; Etc..
In one embodiment, two model selection algorithm are preferably adopted to determine forecast model.
(1) with the related coefficient of WW
WW is one of the social parameter on the same day, such as: whether the same day is working day, if working day then value be 1, nonworkdays then value is 0 (division of working day and nonworkdays was as the criterion with the legal festivals and holidays).Such as: can, according to the related coefficient with WW, judge whether relevant with job specification, the data of itemizing of the electricity about certainly working day and nonworkdays will distinguish, irrelevant then can directly select feature averaging model, etc.A concrete example, the Calculation of correlation factor formula of data group X and data group Y can be established as follows:
ρ XY = Σ ( x - x ‾ ) ( y - y ‾ ) Σ ( x - x ‾ ) 2 ( y - y ‾ ) 2
Algorithm shown in this example, mainly can be used for judging whether this item number is relevant with job specification according to (as electricity subitem data), the selection of the result meeting impact prediction model of judgement.Such as: data group X be every day electricity subitem data x, and i.e. mean value; Data group Y is the job specification y on the same day, gets 1 time on weekdays, gets 0 when nonworkdays, and for mean value; And ρ xYrepresent the related coefficient of electricity subitem data and job specification.Thus, if result of calculation ρ xYenough large, illustrate that this component item power consumption and job specification correlativity are comparatively large, then separately will select forecast model to working day and nonworkdays, if result of calculation ρ xYnot quite, illustrate this component item power consumption and job specification correlativity less, then separately need not select forecast model.
(2) the standard deviation coefficient of variation
The standard deviation coefficient of variation is used to the dispersion degree of evaluating data, and the impact that the height eliminating variate-value level is different with measurement unit, be equivalent to the dispersion degree evaluation number unitized.Such as: whether discretely see by standard variation difference coefficient CV, if discrete, each discrete portions (be generally each season be a part) is distinguished, continue to judge dispersion with CV again after splitting into different piece, if still very discrete just with regression model, such as temperature is on the impact of air conditioning energy consumption, in addition, can also to single part feature averaging model if single part is not discrete, etc.A concrete example, the computing formula of the standard deviation coefficient of variation can be established as follows:
CV σ = σ x ‾ × 100 % = Σ i = 1 n ( x i - x ‾ ) 2 n - 1 x ‾ × 100 %
Algorithm shown in this example, wherein σ represents the standard deviation of these group data, represent the mean value of these group data, x irepresent each value of these group data, n represents the number of these group data.CV σrepresent the standard deviation coefficient of variation of these group data.Algorithm shown in this example, x irepresent electricity subitem data, CV σcan be used for judging the dispersion degree of this electricity subitem energy consumption, judged result can the selection of impact prediction model.
Above-mentioned algorithm is that preferred, that basis is auxiliary information show that the power consumption feature of the electricity subitem data that the combination such as the degree of correlation, dispersion is corresponding selects the example of suitable forecast model as required, not as the restriction to the application's scheme.
Step S220, using the history ad hoc rules information of history electricity subitem data, historical external parameter and collection accordingly as the sample of model training, trains the forecast model of setting, determines model parameter and the model accuracy thereof of forecast model.
In one embodiment, the electrisity consumption of history electricity subitem data representative, corresponding meteorologic parameter (temperature) and social information's (job specification information) are input in the forecast model (as: regression model or feature averaging model etc.) of setting as sample characteristics, after training, obtain the model parameter of this forecast model and corresponding precision, thus this forecast model can be utilized to predict.Wherein, the precision of forecast model can select the coefficient of corresponding model, carrys out judgment models and whether meets accuracy requirement, and for regression model, coefficient of determination R2 and the predicted root mean square error coefficient of variation CVRMSE that can get regression model judge.
In step S120, Threshold selection is carried out, to draw diagnostic threshold to forecast model and precision thereof.
In one embodiment, by the mode of Threshold selection, according to the forecast model determined after training and precise manner thereof, the required threshold value used of efficiency diagnosis can be drawn, relative threshold and/or absolute threshold can be comprised.
Be regression model for forecast model, select the relative threshold determining in other words to diagnose according to the parameter R2 of model and CVRMSE: R2 >=0.8 and CVRMSE≤10%, relative threshold is ± 20%; R2 >=0.8 and 10%<CVRMSE≤20%, relative threshold is ± 25%; R2 >=0.75 and CVRMSE≤5%, relative threshold is ± 30%, and namely the higher then corresponding threshold range of precision of forecasting model is just less.That is, if relative threshold is ± 20%, then the estimated value obtained after us and the relative error of measured value just can not exceed ± and 20%, otherwise just think there is exception.
In step S130, the corresponding external parameter obtained during electric subitem data current according to collection and the ad hoc rules information of collection, obtain the electric data of itemizing of corresponding prediction by described forecast model.
Wherein, current or Real-time Collection or collection are obtained, real-time external parameter and ad hoc rules information is input in trained forecast model and carries out efficiency prediction, namely the predicted value obtained predicts electricity subitem data (representing the power consumption of prediction).Wherein, described the content gathering (current) external parameter and real-time (current) ad hoc rules information of collection in real time in step s 110, do not repeated them here.
In one embodiment, the schematic diagram of an embodiment of efficiency prediction and diagnosis is carried out as shown in Figure 4 in the application's method, the external parameter newly gathering, collected (temperature in meteorologic parameter), ad hoc rules information (the job specification information in social information) are input in trained forecast model as sample characteristics, obtain corresponding predicted value, i.e. prediction electricity subitem data.Further, can also judge whether to meet precision, as the precision with reference to forecast model, the data that error then gathers very greatly may be wrong.
In step S140, based on described diagnostic threshold, contrast described prediction electricity subitem data and the corresponding current electricity subitem data gathered, to determine efficiency situation.
In one embodiment, data of being itemized by the prediction obtained electricity (power consumption prediction data/prediction power consumption value of namely itemizing) and current electricity subitem data (i.e. the electricity consumption the recorded in real time data/actual measurement power consumption value of this subitem) collected accordingly contrast, according to the diagnostic threshold obtained before, draw diagnostic result.Such as: whether the two difference number percent has exceeded this diagnostic threshold of setting, by exceeding the situation of setting threshold value, the efficiency situation (schematic diagram shown in Figure 4) of power consumption exception can be considered as.
Further, efficiency situation can be exported, as: the diagnostic result numerical value display translation of certain branch road subitem efficiency situation also identifies it in threshold range with green, represents normally; The diagnostic result numerical value display translation of certain branch road subitem efficiency situation also identifies it outside threshold range with red, represents extremely; Etc..In addition, can also according to the difference degree between actual measurement to prediction power consumption and corresponding diagnostic threshold, diagnostic result can be divided into slight and obvious efficiency abnormal, and the alarm of showed different.
Wherein, all right computational prediction power consumption value and actual measurement power consumption value relative error and absolute error, when exceeding the relative threshold upper limit in relative error or absolute error can being compared further in limited time lower than under relative threshold, and set up grade separately according to the size of absolute error, higher and be on the low sidely all divided into three, thus export result more intuitively, make staff more can understand fault degree intuitively.
Fig. 5 schematically shows the structured flowchart of an embodiment of the building efficiency diagnostic system according to the application.According to an embodiment of the application, this system 500 can comprise:
Model determining device 510, for according to history electricity subitem data, gather described electricity subitem data time the corresponding historical external parameter that obtains and the history ad hoc rules information of collection, based on model training, to determine forecast model and precision thereof.This device concrete function and process are see step S110.
Wherein, also comprise in model determining device 510: training pattern device 511 is treated in setting, for the analysis according to data of itemizing to every history electricity, and in conjunction with corresponding historical external parameter and/or history ad hoc rules information, set forecast model to be trained, this device concrete function and process are see step S210; And model training apparatus 512, for using the sample of the history ad hoc rules information of history electricity subitem data, corresponding historical external parameter and collection as model training, the forecast model of setting is trained, determine model parameter and the model accuracy thereof of forecast model, this device concrete function and process are see step S220.
Diagnostic threshold device 520, for carrying out Threshold selection, to draw diagnostic threshold to forecast model and precision thereof.The concrete function of this device and process are see step S120.
Prediction unit 530, for the ad hoc rules information according to the corresponding external parameter obtained when gathering current electricity subitem data and collection, obtains corresponding prediction electricity subitem data by described forecast model.The concrete function of this device and process are see step S130.
Contrast diagnostic device 540, for based on described diagnostic threshold, contrasts described prediction electricity subitem data and the corresponding current electricity subitem data gathered, to determine efficiency situation.The concrete function of this device and process are see step S140.
The process realized due to the system of the present embodiment and function are substantially corresponding to the embodiment of the method shown in earlier figures 1 ~ Fig. 4, therefore not detailed part in the description of the present embodiment, see the related description in previous embodiment, can not repeat at this.
Advantage and the good effect of the scheme of the application comprise: 1) diagnose desired data kind less, and be frequently-used data, being easy to obtain, without the need to renovating to existing building energy consumption metering system, can directly applying to actual building energy efficiency management platform.2) measure without the need to choosing specific operation in diagnostic procedure, also the use of energy for building system would not be affected, that can summarize each subitem from history itemizes power consumption data uses energy feature, thus carries out efficiency diagnosis to real-time actual measurement subitem electricity consumption data.
Further, the scheme of the application, using the subitem power consumption historical data that gathers and corresponding meteorologic parameter and social information as training data, by improve efficiency diagnostic method and system automatically select power consumption forecast model, and computation model undetermined parameter, model accuracy and diagnostic threshold, the actual measurement power consumption that need diagnose and accordingly meteorologic parameter and social information bring model computational prediction power consumption into, by the power consumption data that contrast is surveyed and predicted, provide the diagnostic result of each subitem efficiency height in special time, to realize the efficiency diagnosis to the electricity consumption of public building subitem, and provide corresponding information support for the timely eliminating of low-energy-efficiency point.The application is by the analysis to public building subitem power consumption historical data, summarize the power consumption characteristics of different subitem and characterized with corresponding forecast model, for diagnosing the efficiency height of subitem electricity consumption, the multiplexing electric abnormality value of quantitative simultaneously, there is provided tutorial message for building maintenance managerial personnel investigate fault, thus effectively reduce energy dissipation.In addition, utilize diagnosis algorithm can carry out automatic diagnosis to mass data, significantly save manpower, there is significant practice and be worth.
In one typically configuration, computing equipment comprises one or more processor (CPU), input/output interface, network interface and internal memory.
Internal memory may comprise the volatile memory in computer-readable medium, and the forms such as random access memory (RAM) and/or Nonvolatile memory, as ROM (read-only memory) (ROM) or flash memory (flashRAM).Internal memory is the example of computer-readable medium.
Computer-readable medium comprises permanent and impermanency, removable and non-removable media can be stored to realize information by any method or technology.Information can be computer-readable instruction, data structure, the module of program or other data.The example of the storage medium of computing machine comprises, but be not limited to phase transition internal memory (PRAM), static RAM (SRAM), dynamic RAM (DRAM), the random access memory (RAM) of other types, ROM (read-only memory) (ROM), Electrically Erasable Read Only Memory (EEPROM), fast flash memory bank or other memory techniques, read-only optical disc ROM (read-only memory) (CD-ROM), digital versatile disc (DVD) or other optical memory, magnetic magnetic tape cassette, tape magnetic rigid disk stores or other magnetic storage apparatus or any other non-transmitting medium, can be used for storing the information can accessed by computing equipment.According to defining herein, computer-readable medium does not comprise non-temporary computer readable media (transitory media), as data-signal and the carrier wave of modulation.
Also it should be noted that, term " comprises ", " comprising " or its any other variant are intended to contain comprising of nonexcludability, thus make to comprise the process of a series of key element, method, commodity or equipment and not only comprise those key elements, but also comprise other key elements clearly do not listed, or also comprise by the intrinsic key element of this process, method, commodity or equipment.When not more restrictions, the key element limited by statement " comprising ... ", and be not precluded within process, method, commodity or the equipment comprising described key element and also there is other identical element.
Those skilled in the art should understand, the embodiment of the application can be provided as method, system or computer program.Therefore, the application can adopt the form of complete hardware embodiment, completely software implementation or the embodiment in conjunction with software and hardware aspect.And the application can adopt in one or more form wherein including the upper computer program implemented of computer-usable storage medium (including but not limited to magnetic disk memory, CD-ROM, optical memory etc.) of computer usable program code.
The foregoing is only the embodiment of the application, be not limited to the application, for a person skilled in the art, the application can have various modifications and variations.Within all spirit in the application and principle, any amendment done, equivalent replacement, improvement etc., within the right that all should be included in the application.

Claims (10)

1. build an efficiency diagnostic method, it is characterized in that, comprising:
According to history electricity subitem data, gather described electricity subitem data time the corresponding historical external parameter that obtains and the history ad hoc rules information of collection, based on model training, to determine forecast model and precision thereof;
Threshold selection is carried out, to draw diagnostic threshold to forecast model and precision thereof;
The corresponding external parameter obtained during electric subitem data current according to collection and the ad hoc rules information of collection, obtain the electric data of itemizing of corresponding prediction by described forecast model;
Based on described diagnostic threshold, contrast described prediction electricity subitem data and the corresponding current electricity subitem data gathered, to determine efficiency situation.
2. the method for claim 1, it is characterized in that, according to history electricity subitem data, gather described electricity subitem data time the corresponding historical external parameter that obtains and the history ad hoc rules information of collection, based on model training, to determine forecast model and precision thereof, also comprise:
According to the analysis to every history electricity subitem data, and in conjunction with corresponding historical external parameter and/or history ad hoc rules information, set forecast model to be trained;
Using the history ad hoc rules information of history electricity subitem data, historical external parameter and collection accordingly as the sample of model training, the forecast model of setting is trained, determines model parameter and the model accuracy thereof of forecast model.
3. method as claimed in claim 2, is characterized in that, according to the analysis to every history electricity subitem data, and in conjunction with corresponding historical external parameter and/or history ad hoc rules information, sets forecast model to be trained, comprising:
To its power consumption feature of history electricity subitem data analysis of input, according to dispersion and/or the degree of correlation, in conjunction with corresponding historical external parameter and/or history ad hoc rules information, setting uses regression model and/or feature averaging model as forecast model to be trained.
4. the method as described in one of claims 1 to 3, is characterized in that, comprising:
Described external parameter comprises the temperature in meteorologic parameter;
Described ad hoc rules information comprises the job specification information in social information;
Electricity subitem data, by the frequency of data collector according to 15 minutes/time, the electrisity consumption of surveyed electricity consumption branch road, electric current and/or voltage are gathered, each electric branch road is divided to the power consumption data obtaining each branch road in specific electricity consumption subitem, and utilize data aggregate to obtain each branch road power consumption data of every day, described each branch road power consumption data to be itemized data as electricity;
Meteorologic parameter, directly obtained the temperature data of outdoor dry-bulb temperature by the frequency collection of 15 minutes/time by the Temperature sampler of meteorological data collection end, and by data average treatment, the multiple temperature datas gathered in special time period are averaged, calculate this temperature, as meteorologic parameter; Or, indirectly searched for by weather station, the building location database carrying out efficiency diagnosis and obtain this temperature;
Social information, by social information's data acquisition end, arranges to give job specification information the same day to corresponding electricity subitem data place according to country's legal festivals and holidays.
5. the method as described in one of claim 1-3, is characterized in that, based on described diagnostic threshold, contrasts described prediction electricity subitem data and the corresponding current electricity subitem data gathered, to determine efficiency situation, comprising:
The error of computational prediction electricity subitem data and corresponding current electricity subitem data, according to described diagnostic threshold scope, draws efficiency diagnostic result, and exports described efficiency situation.
6. method as claimed in claim 5, is characterized in that, the error of computational prediction electricity subitem data and corresponding current electricity subitem data, according to described diagnostic threshold scope, draws efficiency diagnostic result, and export described efficiency situation, comprising:
The relative error of computational prediction electricity subitem data and corresponding current electricity subitem data and absolute error, when exceeding the relative threshold upper limit in relative error or comparing absolute error more in limited time lower than under relative threshold, and set up grade separately according to the size of absolute error, the alarm of showed different.
7. build an efficiency diagnostic system, it is characterized in that, comprising:
Model determining device, for according to history electricity subitem data, gather described electricity subitem data time the corresponding historical external parameter that obtains and the history ad hoc rules information of collection, based on model training, to determine forecast model and precision thereof;
Diagnostic threshold device, for carrying out Threshold selection, to draw diagnostic threshold to forecast model and precision thereof;
Prediction unit, for the ad hoc rules information according to the corresponding external parameter obtained when gathering current electricity subitem data and collection, obtains corresponding prediction electricity subitem data by described forecast model;
Contrast diagnostic device, for based on described diagnostic threshold, contrasts described prediction electricity subitem data and the corresponding current electricity subitem data gathered, to determine efficiency situation; Wherein,
Model determining device, also comprises:
Training pattern device is treated in setting, for the analysis according to data of itemizing to every history electricity, and in conjunction with corresponding historical external parameter and/or history ad hoc rules information, sets forecast model to be trained; And model training apparatus, for using the sample of the history ad hoc rules information of history electricity subitem data, corresponding historical external parameter and collection as model training, the forecast model of setting is trained, determines model parameter and the model accuracy thereof of forecast model.
8. system as claimed in claim 7, it is characterized in that, described model determining device, comprising:
To its power consumption feature of history electricity subitem data analysis of input, according to dispersion and/or the degree of correlation, in conjunction with corresponding historical external parameter and/or history ad hoc rules information, setting uses regression model and/or feature averaging model as forecast model to be trained.
9. system as claimed in claim 7 or 8, is characterized in that, comprising:
Described external parameter comprises the temperature in meteorologic parameter;
Described ad hoc rules information comprises the job specification information in social information;
Electricity subitem data, by the frequency of data collector according to 15 minutes/time, the electrisity consumption of surveyed electricity consumption branch road, electric current and/or voltage are gathered, each electric branch road is divided to the power consumption data obtaining each branch road in specific electricity consumption subitem, and utilize data aggregate to obtain each branch road power consumption data of every day, described each branch road power consumption data to be itemized data as electricity;
Meteorologic parameter, directly obtained the temperature data of outdoor dry-bulb temperature by the frequency collection of 15 minutes/time by the Temperature sampler of meteorological data collection end, and by data average treatment, the multiple temperature datas gathered in special time period are averaged, calculate this temperature, as meteorologic parameter; Or, indirectly searched for by weather station, the building location database carrying out efficiency diagnosis and obtain this temperature;
Social information, by social information's data acquisition end, arranges to give job specification information the same day to corresponding electricity subitem data place according to country's legal festivals and holidays.
10. the system as described in one of claim 7 ~ 9, is characterized in that, contrast diagnostic device, also comprises:
The relative error of computational prediction electricity subitem data and corresponding current electricity subitem data and absolute error, according to described diagnostic threshold scope, draw efficiency diagnostic result, and export described efficiency situation, size according to absolute error sets up grade separately, the alarm of showed different.
CN201510320498.4A 2015-06-11 2015-06-11 Build efficiency diagnostic method and system Active CN104915562B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201510320498.4A CN104915562B (en) 2015-06-11 2015-06-11 Build efficiency diagnostic method and system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201510320498.4A CN104915562B (en) 2015-06-11 2015-06-11 Build efficiency diagnostic method and system

Publications (2)

Publication Number Publication Date
CN104915562A true CN104915562A (en) 2015-09-16
CN104915562B CN104915562B (en) 2019-03-22

Family

ID=54084623

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201510320498.4A Active CN104915562B (en) 2015-06-11 2015-06-11 Build efficiency diagnostic method and system

Country Status (1)

Country Link
CN (1) CN104915562B (en)

Cited By (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105488302A (en) * 2015-12-24 2016-04-13 曹岭 Energy saving amount measuring method for energy saving reconstruction of public building
CN105759608A (en) * 2016-02-26 2016-07-13 北京邮电大学 Energy-saving method and system
CN106897494A (en) * 2017-01-20 2017-06-27 重庆酷能节能科技有限公司 A kind of Internet of Things energy efficiency management system
CN107590022A (en) * 2016-07-08 2018-01-16 上海东方延华节能技术服务股份有限公司 A kind of instrument to collect data recovery method for building energy consumption metering separate
CN108764626A (en) * 2018-04-19 2018-11-06 武汉钢铁有限公司 A kind of diagnosis of energy saving method and apparatus
CN109298270A (en) * 2018-11-21 2019-02-01 珠海格力电器股份有限公司 One kind determining equipment fault detection method and its detection system based on electricity consumption
CN110456690A (en) * 2019-07-23 2019-11-15 安徽集弘物联科技有限公司 A kind of intelligent type comprehensive managing power consumption control system
CN110516847A (en) * 2019-07-27 2019-11-29 中建科技有限公司 A kind of building energy consumption exception feedback method and device
CN111968356A (en) * 2020-08-11 2020-11-20 重庆电子工程职业学院 Intelligent building energy consumption monitoring system and method

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102938092A (en) * 2012-10-08 2013-02-20 珠海派诺科技股份有限公司 Prediction method of building energy consumption in festivals and holidays based on neural network
CN103745279A (en) * 2014-01-24 2014-04-23 广东工业大学 Method and device for monitoring energy consumption abnormity
CN103853106A (en) * 2012-11-28 2014-06-11 同济大学 Energy consumption prediction parameter optimization method of building energy supply device
CN104239982A (en) * 2014-10-12 2014-12-24 刘岩 Method for predicting energy consumption of buildings during holidays and festivals on basis of time series and neural networks
CN104636999A (en) * 2015-01-04 2015-05-20 江苏联宏自动化系统工程有限公司 Detection method for building abnormal energy consumption data

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102938092A (en) * 2012-10-08 2013-02-20 珠海派诺科技股份有限公司 Prediction method of building energy consumption in festivals and holidays based on neural network
CN103853106A (en) * 2012-11-28 2014-06-11 同济大学 Energy consumption prediction parameter optimization method of building energy supply device
CN103745279A (en) * 2014-01-24 2014-04-23 广东工业大学 Method and device for monitoring energy consumption abnormity
CN104239982A (en) * 2014-10-12 2014-12-24 刘岩 Method for predicting energy consumption of buildings during holidays and festivals on basis of time series and neural networks
CN104636999A (en) * 2015-01-04 2015-05-20 江苏联宏自动化系统工程有限公司 Detection method for building abnormal energy consumption data

Cited By (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105488302A (en) * 2015-12-24 2016-04-13 曹岭 Energy saving amount measuring method for energy saving reconstruction of public building
CN105759608A (en) * 2016-02-26 2016-07-13 北京邮电大学 Energy-saving method and system
CN105759608B (en) * 2016-02-26 2018-12-28 北京邮电大学 A kind of power-economizing method and system
CN107590022A (en) * 2016-07-08 2018-01-16 上海东方延华节能技术服务股份有限公司 A kind of instrument to collect data recovery method for building energy consumption metering separate
CN107590022B (en) * 2016-07-08 2021-06-25 上海东方延华节能技术服务股份有限公司 Instrument collected data restoration method for building energy consumption subentry measurement
CN106897494A (en) * 2017-01-20 2017-06-27 重庆酷能节能科技有限公司 A kind of Internet of Things energy efficiency management system
CN108764626A (en) * 2018-04-19 2018-11-06 武汉钢铁有限公司 A kind of diagnosis of energy saving method and apparatus
CN109298270A (en) * 2018-11-21 2019-02-01 珠海格力电器股份有限公司 One kind determining equipment fault detection method and its detection system based on electricity consumption
CN110456690A (en) * 2019-07-23 2019-11-15 安徽集弘物联科技有限公司 A kind of intelligent type comprehensive managing power consumption control system
CN110516847A (en) * 2019-07-27 2019-11-29 中建科技有限公司 A kind of building energy consumption exception feedback method and device
CN111968356A (en) * 2020-08-11 2020-11-20 重庆电子工程职业学院 Intelligent building energy consumption monitoring system and method

Also Published As

Publication number Publication date
CN104915562B (en) 2019-03-22

Similar Documents

Publication Publication Date Title
CN104915562A (en) Building energy efficiency diagnosis method and system
Reis et al. Spatial metrics to study urban patterns in growing and shrinking cities
CN103903408B (en) Method for early warning and system are investigated in equipment fault
CN103853106B (en) A kind of energy consumption Prediction Parameters optimization method of building energy supplied equipment
CN106447210B (en) A kind of distribution net equipment health degree dynamic diagnosis method of meter and trust evaluation
Meng et al. Degree-day based non-domestic building energy analytics and modelling should use building and type specific base temperatures
CN102928720A (en) Defect rate detecting method of oil immersed type main transformer
CN105548949A (en) Method and system for fault remote determination of intelligent ammeter
CN101853290A (en) Meteorological service performance evaluation method based on geographical information system (GIS)
CN105868886B (en) Rain and sewage storage tank operation efficiency evaluation method
CN105678481A (en) Pipeline health state assessment method based on random forest model
Sikorska et al. Effective precipitation duration for runoff peaks based on catchment modelling
CN105182122A (en) Fault early warning method of random power supply access equipment
CN103699668A (en) Power distribution network electric equipment combination state evaluation method based on data section consistency
CN105956752A (en) Energy efficiency evaluation and diagnosis analysis method used for green public building operation phase
CN106570779A (en) DC power distribution network reliability analysis method and system
Wang et al. Environmental impacts of short building lifespans in China considering time value
Elgar et al. Simulations of firm location decisions: Replicating office location choices in the Greater Toronto Area
D’Alpaos et al. Prioritization of energy retrofit strategies in public housing: An AHP model
CN206312210U (en) A kind of status assessing system of Distribution Network Equipment
CN107292484A (en) The recognition methods of urban redevelopment soil and system based on city space big data
CN104751292A (en) Integrated management method for electromechanical equipment operation
Kazemi-Beydokhti et al. Determination of the physical domain for air quality monitoring stations using the ANP-OWA method in GIS
CN108700851A (en) System, method and platform based on cloud for predicting energy expenditure
CN104618133B (en) A kind of grid voltage quality Monitoring Data acquisition method and system

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
TA01 Transfer of patent application right
TA01 Transfer of patent application right

Effective date of registration: 20190202

Address after: Room 2106, 1027 Changning Road, Changning District, Shanghai 200050

Applicant after: SHANGHAI TWENTY-FIRST ENERGY SAVING TECHNOLOGY CO., LTD.

Address before: Room 2107, 1027 Changning Road, Changning District, Shanghai 200050

Applicant before: SHANGHAI NEW CHANGNING LOW-CARBON INVESTMENT MANAGEMENT CO., LTD.

Applicant before: SHANGHAI TWENTY-FIRST ENERGY SAVING TECHNOLOGY CO., LTD.

GR01 Patent grant
GR01 Patent grant