An industrial building energy consumption working condition anomaly monitoring method, system and storage medium

Abstract

The method, system, and storage medium for monitoring abnormal energy consumption conditions of industrial buildings disclosed in this embodiment import building model information of industrial buildings in a park, calculate and obtain the design energy consumption index of each industrial building in the park according to a preset energy consumption assessment database, then calculate and obtain the actual energy consumption index of each industrial building as the benchmark energy consumption index, and screen and mark buildings with excessive energy consumption based on the degree of difference between the design energy consumption index and the benchmark energy consumption index; at the same time, obtain a number of historical load days with high similarity to the predicted day under set conditions from historical sample load days as similar days of the predicted day; query and obtain the historical energy consumption index of similar days, compare the historical energy consumption index of similar days with the benchmark energy consumption index, and mark the industrial buildings corresponding to the benchmark energy consumption index with a comparison difference exceeding a preset threshold as having abnormal energy consumption conditions, thereby enabling subsequent energy consumption optimization of the marked abnormal industrial buildings.

Description

Technical Field

[0001] This invention relates to the field of industrial control technology, and in particular to a method, system, and storage medium for monitoring abnormal energy consumption conditions in industrial buildings. Background Technology

[0002] In recent years, with economic development, my country's construction industry has also developed rapidly. Construction energy consumption has accounted for more than 30% of the total social energy consumption, and this trend is increasing year by year. Industrial buildings account for about 20% of the total building area, and their energy consumption accounts for approximately 20% to 30% of the national total energy consumption. Currently, my country is in a period of relative energy shortage, and the high energy consumption of industrial buildings further exacerbates the country's energy pressure, hindering the sustainable development of the national economy. Therefore, energy conservation in industrial buildings is urgently needed.

[0003] According to relevant national and local government regulations, energy conservation and environmental protection policies are being vigorously implemented nationwide, with energy conservation in industrial buildings being one of them. However, in the energy consumption monitoring of industrial buildings, the traditional method involves administrators setting a threshold and then judging the energy consumption level based on monitoring data. This monitoring method has several problems: the threshold is set arbitrarily, and being too high or too low will affect the judgment of energy consumption; it does not consider the impact of geographical conditions, seasonal changes, holidays, and other factors on the energy consumption of industrial buildings; and when excessive energy consumption occurs in industrial buildings that have already been completed, human intervention is the only option, but the results of such intervention are obviously not optimistic. Summary of the Invention

[0004] This invention addresses the shortcomings of existing technologies by providing a method for monitoring abnormal energy consumption conditions in industrial buildings, comprising the following steps:

[0005] S1, import the building model information of industrial buildings in the park, and calculate and obtain the design energy consumption index of each industrial building in the park based on the preset energy consumption assessment database.

[0006] S2 collects the total energy consumption data of each industrial building through the sensor information platform, obtains the total usable area of ​​the corresponding industrial building based on the building model information, and calculates the actual energy consumption index of each industrial building as the benchmark energy consumption index.

[0007] S3, based on the degree of difference between the design energy consumption index and the benchmark energy consumption index, screen buildings with excessive energy consumption and mark the corresponding buildings as Class I.

[0008] S4, obtain several historical load days with high similarity to the predicted day under the set conditions from the historical load days as similar days to the predicted day;

[0009] S5, query and obtain historical energy consumption indicators for similar days, compare the historical energy consumption indicators for similar days with the benchmark energy consumption indicators, and mark the industrial buildings corresponding to the benchmark energy consumption indicators whose comparison difference exceeds the preset threshold as having abnormal energy consumption conditions in the second category.

[0010] Preferably, the benchmark energy consumption index is energy consumption per unit area, which is the ratio of the building's total energy consumption to its total usable area. The total energy consumption is collected through a sensor information platform, and the total usable area of ​​the building is obtained through building model information identification.

[0011] Preferably, step S3 specifically includes:

[0012] The design performance consumption index and the benchmark performance consumption index are uniformly normalized to a range of [0, 100], and the above process is repeated for each industrial building in the park.

[0013] Let the input value of the design performance consumption index be denoted as X1, and the input value of the benchmark performance consumption index be denoted as X2. Then, the element x has a Y output of [0,100] according to the following formula, which forms Y1 and Y2 corresponding to X1 and X2, where Y = 100*(X-min) / (max-min).

[0014] Using the quartile method, the outer contour color of the building model was set to four different colors according to the average value of Y1 from low to high.

[0015] Establish a two-dimensional input data space of Y1 and Y2, and perform linear regression on Y1 and Y2 such that each point has a least squares distance D. y The value above the regression line is positive, and the value below it is negative. (This refers to the value of D.) y Outlier extraction is performed using the interquartile range method. If there is D... y If the interquartile range is greater than the upper bound of k=3, then the energy consumption is considered to be excessive and a first-class label is applied to the corresponding building model.

[0016] Preferably, step S4 specifically includes:

[0017] S41, the meteorological factor matching coefficient is calculated through grey relational analysis;

[0018] S42, calculate and obtain the time factor matching coefficient, which represents the degree of similarity between the predicted day and the historical day in time. The time factor is the number of days between the historical day and the day to be predicted. Where mod is the modulo function; t is the number of days between the i-th historical day and the predicted day; int is the integer operation; s i It is a 0 / 1 variable; when the i-th historical day and the predicted day fall on the same holiday, s iThe value is 1 otherwise; β1, β2 and β3 are attenuation coefficients, usually ranging from 0.9 to 0.98, representing the similarity reduction ratio for each additional day, week and year of distance between historical days and similar days, respectively; N1, N2 and N3 are constants, with N1 and N2 being the number of days in a week (7) and N3 being 340.

[0019] S43, calculate and obtain the weekday factor matching coefficient, which represents the similarity between the predicted day and the historical day in terms of weekday type. The weekday type is quantified, with Monday mapped to 0.1, Tuesday through Thursday to 0.2, Friday to 0.3, Saturday to 0.7, and Sunday to 1. The following function expression is used to calculate the weekday factor matching coefficient γ between the i-th historical day and the predicted day. i γ i =1-|f(X) i )-f(X0)|, where (X i Xi and X0 are the weekday types of the i-th historical day and similar day, respectively, with the possible values ​​1, 2, 3, 4, 5, 6, 7; f(Xi) i f(X0) is (X i The value after mapping X0 and X0;

[0020] S44. Multiply the meteorological factor matching coefficient, the time factor matching coefficient, and the weekday factor matching coefficient to obtain the comprehensive matching coefficient. The larger the comprehensive matching coefficient, the closer the characteristics of the selected similar day and the predicted day are.

[0021] Preferably, step S4 further includes:

[0022] Energy consumption data of each industrial building during a defined period are obtained and arranged chronologically to form a time series. The Fbprophet prediction method is used to obtain the prediction results for the time series. The prediction model is in the form of a superposition function: y(t)=g(t)+s(t)+h(t)+∈ t g(t) is the trend function, used to fit and model non-periodic changes in the time series; s(t) is the periodic term, used to reflect the periodic changes in the model; h(t) is the holiday term, used to reflect the impact of holiday effects that may occur within one or more days; ∈ t This is the error term, used to reflect abnormal changes in the model, assumed to be ∈ t It follows a normal distribution, so the trend function is obtained: Where C is the expected capacity of the system; k is the growth rate; and m is the compensation parameter.

[0023] This invention also discloses an industrial building energy consumption anomaly monitoring system, comprising: a design energy consumption acquisition module, used to import building model information of industrial buildings in the park, and calculate and obtain the design energy consumption index of each industrial building in the park according to a preset energy consumption assessment database; a benchmark energy consumption acquisition module, used to collect the total energy consumption data of each industrial building through a sensor information platform, and obtain the total usable area of ​​the corresponding industrial building according to the building model information, and calculate and obtain the actual energy consumption index of each industrial building as the benchmark energy consumption index; a first marking module, used to screen buildings with excessive energy consumption according to the degree of difference between the design energy consumption index and the benchmark energy consumption index and mark the corresponding buildings in the first category; a similar day acquisition module, used to obtain a number of historical load days with high similarity to the predicted day under set conditions from historical sample load days as similar days of the predicted day; and a second marking module, used to query and obtain the historical energy consumption index of similar days, compare the historical energy consumption index of similar days with the benchmark energy consumption index, and mark the industrial buildings corresponding to the benchmark energy consumption index with a comparison difference exceeding a preset threshold as having abnormal energy consumption conditions in the second category.

[0024] Preferably, the benchmark energy consumption index is energy consumption per unit area, which is the ratio of the building's total energy consumption to its total usable area. The total energy consumption is collected through a sensor information platform, and the total usable area of ​​the building is obtained through building model information identification.

[0025] Preferably, the first identification module is specifically configured as follows:

[0026] The design performance consumption index and the benchmark performance consumption index are uniformly normalized to a range of [0, 100], and the above process is repeated for each industrial building in the park.

[0027] Let the input value of the design performance consumption index be denoted as X1, and the input value of the benchmark performance consumption index be denoted as X2. Then, the element x has a Y output of [0,100] according to the following formula, which forms Y1 and Y2 corresponding to X1 and X2, where Y = 100*(X-min) / (max-min).

[0028] Using the quartile method, the outer contour color of the building model was set to four different colors according to the average value of Y1 from low to high.

[0029] Establish a two-dimensional input data space of Y1 and Y2, and perform linear regression on Y1 and Y2 such that each point has a least squares distance D. y The value above the regression line is positive, and the value below it is negative. (This refers to the value of D.) y Outlier extraction is performed using the interquartile range method. If there is D... y If the interquartile range is greater than the upper bound of k=3, then the energy consumption is considered to be excessive and a first-class label is applied to the corresponding building model.

[0030] The present invention also discloses an industrial building energy consumption condition abnormal monitoring device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of any of the methods described above.

[0031] The present invention also discloses a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of any of the methods described above.

[0032] The method, system, and storage medium for monitoring abnormal energy consumption conditions of industrial buildings disclosed in this embodiment first import the building model information of industrial buildings in the park, and calculate the design energy consumption index of each industrial building in the park according to a preset energy consumption assessment database. Then, the total energy consumption data of each industrial building is collected through a sensor information platform, and the total usable area of ​​the corresponding industrial building is obtained according to the building model information. The actual energy consumption index of each industrial building is calculated as the benchmark energy consumption index. Buildings with excessive energy consumption are screened and marked according to the degree of difference between the design energy consumption index and the benchmark energy consumption index. At the same time, several historical load days with high similarity to the predicted day under set conditions are obtained from the historical sample load days as similar days of the predicted day. The historical energy consumption index of similar days is retrieved and compared with the benchmark energy consumption index. The industrial buildings corresponding to the benchmark energy consumption index with a comparison difference exceeding a preset threshold are marked as having abnormal energy consumption conditions, so that the energy consumption of the marked abnormal industrial buildings can be optimized in the later stage.

[0033] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0034] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings:

[0035] Figure 1 This is a schematic diagram of the steps of an abnormal monitoring method for energy consumption conditions in industrial buildings disclosed in an embodiment of the present invention.

[0036] Figure 2 This is a schematic diagram of the Kalman filtering method disclosed in an embodiment of the present invention for estimating g(t). Detailed Implementation

[0037] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the described embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0038] In this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," "linking," and "fixing," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0039] In this invention, unless otherwise explicitly specified and limited, "above" or "below" the second feature can include direct contact between the first and second features, or contact between the first and second features through another feature between them. Furthermore, "above," "over," and "on top" of the second feature includes the first feature directly above or diagonally above the second feature, or simply indicates that the first feature is at a higher horizontal level than the second feature. "Below," "below," and "under" the second feature includes the first feature directly below or diagonally below the second feature, or simply indicates that the first feature is at a lower horizontal level than the second feature.

[0040] Unless otherwise defined, the technical or scientific terms used herein shall have the ordinary meaning as understood by one of ordinary skill in the art to which this invention pertains. The terms “first,” “second,” and similar terms used in the specification and claims of this patent application do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Similarly, the terms “an” or “a” and similar terms do not indicate a limitation of quantity, but rather indicate the presence of at least one.

[0041] As attached Figure 1 As shown, this invention discloses a method for monitoring abnormal energy consumption conditions in industrial buildings, comprising the following steps:

[0042] Step S1: Import the building model information of the industrial buildings in the park, and calculate and obtain the design energy consumption index of each industrial building in the park based on the preset energy consumption assessment database.

[0043] The design energy consumption index is determined based on relevant design specifications. In this embodiment, the preset energy consumption assessment is obtained from the database calculation.

[0044] Importing BIM information for each individual industrial building, the first step is to define a design energy consumption index. This index is calculated using a formula based on relevant design specifications. The corresponding parameters are directly read from the BIM model using the formula stored in the database. For example, the design energy consumption index value can be calculated from the following information directly obtained from the BIM model:

[0045]

[0046] Table 1 Green Building Scoring Table (Building and Envelope Section)

[0047]

[0048] Table 2 Green Building Scoring Table: Heating, Ventilation and Air Conditioning Section

[0049]

[0050] Table 3 Green Building Rating Table - Lighting and Electrical Section

[0051] Step S2: Collect the total energy consumption data of each industrial building through the sensor information platform, obtain the total usable area of ​​the corresponding industrial building based on the building model information, and calculate the actual energy consumption index of each industrial building as the benchmark energy consumption index.

[0052] The benchmark energy consumption index is measured using a series of unit energy consumption indicators based on the deployment of IoT sensors, including but not limited to energy consumption per unit area, energy consumption per capita, energy consumption per production line, and energy consumption per production unit. In this embodiment, the calculated value of the unit area energy consumption index is used as the calculated value of the benchmark energy consumption index. If there are multiple unit energy consumption index values, different weights can be set according to industrial needs. The benchmark energy consumption index is unit area energy consumption, which is the ratio of the building's total energy consumption to its total usable area. The total energy consumption is collected through a sensor information platform, and the total usable area is obtained through building model information identification; that is: unit area energy consumption = total energy consumption / total usable area. The total energy consumption can be directly collected through the sensor IoT platform, and the total usable area can be obtained through BIM information.

[0053] Step S3 involves screening buildings that exceed energy consumption standards based on the degree of difference between the designed energy consumption index and the benchmark energy consumption index, and then labeling the corresponding buildings as Category I buildings. In this embodiment, step S3 may specifically include the following:

[0054] Step S31: Normalize the design performance consumption index and the benchmark performance consumption index to a range of [0, 100], and repeat the above process for each industrial building in the park.

[0055] Step S32: The input value of the design performance consumption index is denoted as X1, and the input value of the benchmark performance consumption index is denoted as X2. Then, the element x has a y output of [0,100] according to the following formula, which forms Y1 and Y2 corresponding to X1 and X2, where Y = 100*(X-min) / (max-min).

[0056] Step S33: Using the quartile method, the outer contour color of the building model is set to four different colors according to the average value of Y1 from low to high.

[0057] Since the BIM information is already imported into the IoT platform and can be displayed interactively, Y1 can be color-coded using methods including, but not limited to, quartiles to show the natural differences in energy consumption among different industrial buildings within the park. Alternatively, unsupervised learning methods, such as k-means, can achieve the same effect. For example, this example uses quartiles to set the BIM outline color of the buildings to dark red, orange, gray, and green, respectively, based on the average value of Y1 from low to high. Through these operations, the design performance energy consumption index is correlated with the benchmark performance energy consumption index, i.e., Y1 and Y2 correspond one-to-one and are divided into four segments (four colors).

[0058] Step S34: Establish a two-dimensional input data space of Y1 and Y2, and perform linear regression on Y1 and Y2, so that each point has a least squares distance D. y The value above the regression line is positive, and the value below it is negative. (This refers to the value of D.) y Outlier extraction is performed using the interquartile range method. If there is D... y If the interquartile range is greater than the upper bound of k=3, then the energy consumption is considered to be excessive and a first-class label is applied to the corresponding building model.

[0059] In this embodiment, the first type of indicator can be a flashing upward arrow on the building. Outliers are extracted using the interquartile range (INR) method for Dy. If any Dy exceeds the upper INR bound of k=3, the energy consumption is considered high, and a flashing upward arrow is displayed on the building. Clicking this arrow allows users to zoom in on the building based on BIM information to view the conditions of different floors and areas.

[0060] The interquartile range (IQR) is used to detect outliers. Outliers are typically defined as being less than QL - kIQR or greater than QU + kIQR. QL is the lower quartile, representing one-quarter of all observations with values ​​smaller than it. QU is the upper quartile, representing one-quarter of all observations with values ​​larger than it. IQR is the interquartile range, the difference between the upper quartile (QU) and the lower quartile (QL), encompassing half of all observations. k = 1.5 is the threshold for moderate outliers, and k = 3 is the threshold for severe outliers.

[0061] In this embodiment, the first type of labeled data is removed from the benchmark energy consumption indicators to obtain the normal energy consumption indicators in this screening, which are then used as the data for the predicted day and similar day analysis is performed on them.

[0062] Step S4: Obtain several historical load days with high similarity to the predicted day under the set conditions from the historical sample load days as similar days to the predicted day.

[0063] The preceding steps describe the situation within a single time slice in a spatial context. Since industrial buildings operate continuously, IoT platform sensors continuously upload data. To eliminate inherent energy consumption differences in the time domain reflected by normal activities, the obtained P' time series is analyzed to establish a trend-based energy consumption index. Trend analysis of the P' time series requires similarity day analysis, which involves identifying several historical load days with high similarity to the predicted day in terms of weekday type, environment, meteorology, lunar calendar, and holidays. This embodiment uses a method to construct a daily feature vector = [meteorological factors, date difference, weekday type, major holidays]. The similarity calculation formula is as follows: the daily feature vector is divided into three categories: meteorological factors, time factors, and weekday factors (major holidays are manually adjusted as special cases). Then, the matching coefficients of the meteorological factors, time factors, and weekday factors are calculated through correlation, and the total similarity is obtained by multiplying them.

[0064] In this embodiment, step S4 may specifically include the following:

[0065] Step S41: The meteorological factor matching coefficient is calculated using grey relational analysis. Specifically, the grey relational analysis method (GRA) is used to calculate the correlation coefficient ξ(Xi) between a reference sequence X0 and several comparison sequences X1, X2, ..., Xn at each time point (i.e., each point on the curve) of each comparison sequence and the reference sequence. ξ(Xi) is calculated using the following formula: where ρ is the resolution coefficient, generally between 0 and 1, usually taken as 0.5. Δmin is the second-level minimum difference, and Δmax is the two-level maximum difference. Δoi(k) is the absolute difference between each point on the curve of each comparison sequence Xi and each point on the curve of the reference sequence X0.

[0066] Step S42: Calculate and obtain the time factor matching coefficient. The time factor matching coefficient represents the degree of similarity between the predicted day and the historical day in time. The time factor is the number of days between the historical day and the day to be predicted.

[0067]

[0068] Where mod is the modulo function; t is the number of days between the i-th historical day and the predicted day; int is the integer operation; s i It is a 0 / 1 variable; when the i-th historical day and the predicted day fall on the same holiday, s i The value is 1 otherwise; β1, β2 and β3 are attenuation coefficients, generally ranging from 0.9 to 0.98, representing the similarity reduction ratio for each additional day, week and year of distance between historical days and similar days, respectively; N1, N2 and N3 are constants, with N1 and N2 being the number of days in a week (7). Considering that the distance between some major holidays (such as the Spring Festival) is less than 365 days, the value of N3 is 340.

[0069] Step S43: Calculate and obtain the weekday factor matching coefficient, which represents the similarity between the day to be predicted and historical days in terms of weekday type. The weekday type can be quantified: Monday's mapping is 0.1, Tuesday through Thursday's mapping is 0.2, Friday's mapping is 0.3, Saturday's mapping is 0.7, and Sunday's mapping is 1.

[0070] Specifically, the following function expression is used to calculate the weekday factor matching coefficient γ between the i-th historical day and the predicted day. i γ i =1-|f(X) i )-f(X0)|;where: (X i Xi and X0 are the weekday types of the i-th historical day and similar day, respectively, with the possible values ​​1, 2, 3, 4, 5, 6, 7; f(Xi) i f(X0) is (X i The value after mapping X0 and X0.

[0071] Step S44: Multiply the meteorological factor matching coefficient, the time factor matching coefficient, and the weekday factor matching coefficient to obtain the comprehensive matching coefficient. The larger the comprehensive matching coefficient, the closer the characteristics of the selected similar day and the predicted day are.

[0072] In this embodiment, step S4 further includes the following: acquiring energy consumption data of each industrial building in a defined period and arranging them according to time sequence to form a time series. Since the aforementioned embodiment uses a preprocessing operation based on similar days, the resulting time series should be trend-stable. Therefore, for simplicity, SARIMA or Fbprophet can be preferentially used to obtain the time series prediction results. In this embodiment, the Fbprophet prediction method is used for prediction, employing a prediction model in the form of a superposition function: y(t)=g(t)+s(t)+h(t)+∈ t , where g(t) is the trend function, which can fit and model non-periodic changes in the time series; s(t) is the periodic term, which can reflect the periodic changes in the model (e.g., annual seasonal variations); h(t) is the holiday term, which reflects the impact of holiday effects that may occur within one or more days; ∈ t This is the error term, reflecting abnormal changes in the model. Assume ∈ t It follows a normal distribution, so the trend function is obtained: Where C is the expected capacity of the system; k is the growth rate; and m is the compensation parameter.

[0073] In this embodiment, an improved prediction method combining ACKF and Fbprophet is used, and its specific steps are as follows: Initialize parameters Sum the error covariance P0; update the time term using a nonlinear state equation. Calculate X j,k-1|k-1 And propagation volume point; calculate predicted state and error matrix The covariance. Replace the volume point X. j,k|k-1 Calculate the propagation value Z of the volume point measurement equation. j,k|k-1 And estimate the predicted value of the measurement.

[0074] Using the above-described fusion adaptive method, the variance of the estimated measurement noise and process noise is obtained. and Determine whether iteration is necessary; use fitness to determine the degree of difference between the sampled points and the true target estimate to make a decision on whether iteration is needed. This step may specifically include the following:

[0075] Define a fitness function; the fitness of the predicted value and the observed value is f1, and the fitness of the volumetric point transition value and the actual observed value is f2. Calculate the fitness function ρ. When ρ < 1, it indicates that the sampled point has a valid approximate true estimate, and the iteration ends; otherwise, the iteration loop continues.

[0076] Measurement update, calculation of new information P zz Covariance P xz , return Kk ,system State update covariance P k When ρ≥1, the difference between the sampled points and the estimated values ​​is relatively large, requiring iteration. Returning to the previous steps and re-initializing, let N be the number of iterations. Then, the state estimate and error covariance estimate at time k are respectively: Using the state estimate obtained from the improved ACKF method as the trend term of Fbprophet, the improved superposition function constitutes the prediction model: The detailed flowchart is attached. Figure 2 As shown.

[0077] in To initialize the parameters, p0 is the error covariance, and X... j,k-1|k-1 Let K-1 be the state variables at discrete steps. To calculate the nonlinear state at k steps using the state at k-1 steps, the predicted state is obtained. and error matrix Volume point X j,k|k-1 Calculate the propagation value Z of the volume point measurement equation. j,k|k-1 And estimate the predicted value of the measurement. To measure the variance of the noise estimate, Let be the variance of the process noise estimate; where ρ is the fitness factor, used to determine whether to exit the iteration; Pzz is the innovation, Pzx is the covariance, and Kk is the profit. For the system's update status, P k For covariance updates. In this embodiment, an upper limit N is set for the number of iterations, where and If the output is the result of the Nth iteration, then the state estimate at time k is: Furthermore, the trend term g(t) used to replace the model is used to obtain the historical samples of the predicted day, i.e., the predicted time series values, based on the superposition function s(t) of the prediction model, to form the historical energy consumption index.

[0078] The predicted time-series values ​​obtained can be continuously compared with the time-series values ​​measured by IoT sensors. With the accumulation of historical data, these differences can be used to identify anomalies using methods including, but not limited to, DBSCAN. Confirmed anomalies will cause color markers in the BIM visualization model to flash, indicating problems with the energy consumption design of industrial buildings in that area, thus providing quantitative feedback on potential issues in industrial buildings.

[0079] Step S5: Query and obtain historical energy consumption indicators for similar days, compare the historical energy consumption indicators for similar days with the benchmark energy consumption indicators, and mark industrial buildings corresponding to benchmark energy consumption indicators whose comparison difference exceeds a preset threshold as having abnormal energy consumption conditions in the second category.

[0080] In another embodiment, historical energy consumption indicators are compared with real-time benchmark energy consumption indicators, and industrial buildings corresponding to benchmark energy consumption indicators whose comparison difference exceeds a preset threshold are marked as having abnormal energy consumption conditions in the second category.

[0081] In this embodiment, the method may further include using the first and second categories of industrial building labels as feedback to develop an optimization scheme for low-energy consumption in industrial buildings. The first and second category label areas represent problem areas of the industrial building and are used to quantitatively reflect potential problems, guiding relevant units in low-carbon design optimization.

[0082] The method for monitoring abnormal energy consumption conditions of industrial buildings disclosed in this embodiment first imports the building model information of industrial buildings in the park and calculates the design energy consumption index of each industrial building in the park according to a preset energy consumption assessment database. Then, it collects the total energy consumption data of each industrial building through a sensor information platform, obtains the total usable area of ​​the corresponding industrial building according to the building model information, calculates the actual energy consumption index of each industrial building as the benchmark energy consumption index, and filters and marks buildings with excessive energy consumption based on the degree of difference between the design energy consumption index and the benchmark energy consumption index. At the same time, it obtains a number of historical load days with high similarity to the predicted day under set conditions from the historical sample load days as similar days of the predicted day. It queries and obtains the historical energy consumption index of similar days, compares the historical energy consumption index of similar days with the benchmark energy consumption index, and marks the industrial buildings corresponding to the benchmark energy consumption index with a comparison difference exceeding a preset threshold as having abnormal energy consumption conditions, so that the energy consumption of the marked abnormal industrial buildings can be optimized in the later stage.

[0083] In another embodiment, an industrial building energy consumption anomaly monitoring system is also disclosed, comprising: a design energy consumption acquisition module, used to import building model information of industrial buildings in the park, and calculate and acquire the design energy consumption index of each industrial building in the park according to a preset energy consumption assessment database; a benchmark energy consumption acquisition module, used to collect and acquire the total energy consumption data of each industrial building through a sensor information platform, and acquire the total usable area of ​​the corresponding industrial building according to the building model information, and calculate and acquire the actual energy consumption index of each industrial building as the benchmark energy consumption index; a first marking module, used to screen buildings with excessive energy consumption according to the degree of difference between the design energy consumption index and the benchmark energy consumption index and mark the corresponding buildings in the first category; a similar day acquisition module, used to acquire a number of historical load days with high similarity to the predicted day under set conditions from historical sample load days as similar days of the predicted day; and a second marking module, used to query and acquire the historical energy consumption index of similar days, compare the historical energy consumption index of similar days with the benchmark energy consumption index, and mark the industrial buildings corresponding to the benchmark energy consumption index with a comparison difference exceeding a preset threshold as having abnormal energy consumption conditions in the second category.

[0084] In this embodiment, the benchmark energy consumption index is energy consumption per unit area, which is the ratio of the building's total energy consumption to its total usable area. The total energy consumption is collected through a sensor information platform, and the total usable area of ​​the building is obtained through building model information identification.

[0085] In this embodiment, the first labeling module is specifically configured to: normalize the design performance consumption index and the benchmark performance consumption index values ​​to a range of [0, 100], and repeat the above processing for each industrial building in the park; denot the input value of the design performance consumption index as X1, and the input value of the benchmark performance consumption index as X2, then the element x has a Y output of [0, 100] according to the following formula, corresponding to X1 and X2 to form Y1 and Y2, where Y = 100 * (X - min) / (max - min); use the quartile method to set the outer contour color of the building model to four different colors according to the average value of Y1 from low to high; establish a two-dimensional input data space of Y1 and Y2, and perform linear regression on Y1 and Y2, so that each point has a least squares distance D. y The value above the regression line is positive, and the value below it is negative. (This refers to the value of D.) y Outlier extraction is performed using the interquartile range method. If there is D... y If the interquartile range is greater than the upper bound of k=3, then the energy consumption is considered to be excessive and a first-class label is applied to the corresponding building model.

[0086] In this embodiment, the similarity day acquisition module specifically includes: a meteorological factor acquisition module, used to calculate the meteorological factor matching coefficient through grey relational analysis; and a time factor acquisition module, used to calculate and acquire the time factor matching coefficient, wherein the time factor matching coefficient represents the degree of similarity between the predicted day and the historical day in time, and the time factor is the number of days between the historical day and the day to be predicted.

[0087] Where mod is the modulo function; t is the number of days between the i-th historical day and the predicted day; int is the integer operation; s i It is a 0 / 1 variable; when the i-th historical day and the predicted day fall on the same holiday, s iThe value is 1 otherwise; β1, β2, and β3 are attenuation coefficients, typically ranging from 0.9 to 0.98, representing the similarity reduction ratio for each additional day, week, and year between historical and similar days, respectively; N1, N2, and N3 are constants, with N1 and N2 being the number of days in a week (7) and N3 being 340; the weekday factor acquisition module is used to calculate the weekday factor matching coefficient, which represents the similarity between the predicted day and the historical day in terms of weekday type. The weekday type is quantified, with Monday mapped to 0.1, Tuesday through Thursday to 0.2, Friday to 0.3, Saturday to 0.7, and Sunday to 1; the following function expression is used to calculate the weekday factor matching coefficient γ between the i-th historical day and the predicted day. i γ i =1-|f(X) i )-f(X0)|, where (X i Xi and X0 are the weekday types of the i-th historical day and similar day, respectively, with the possible values ​​1, 2, 3, 4, 5, 6, 7; f(Xi) i f(X0) is (X i The value after mapping X0 and X0; the comprehensive matching module is used to multiply the meteorological factor matching coefficient, the time factor matching coefficient and the week factor matching coefficient to obtain the comprehensive matching coefficient. The larger the comprehensive matching coefficient is, the closer the characteristics of the selected similar day and the predicted day are.

[0088] The design energy consumption index of industrial buildings is obtained by evaluating preset energy consumption indicators. The longitudinal energy consumption data of industrial buildings is obtained through sensors, and the actual energy consumption index of industrial buildings is obtained by obtaining the total usable area of ​​industrial buildings from the building model information. This is used as the benchmark energy consumption index. By comparing the benchmark energy consumption index with the design energy consumption index, industrial buildings with excessive energy consumption are marked as Category I buildings. Furthermore, the energy consumption index of the remaining buildings is obtained by deleting the Category I marked data from the benchmark energy consumption index data. The energy consumption data of the remaining buildings is obtained by repeatedly acquiring their energy consumption data within a certain period to obtain the predicted day time series. Similar day analysis is performed on the predicted day time series to obtain historical samples of the predicted days to form their historical energy consumption index. The historical energy consumption index is compared with the benchmark energy consumption index to mark industrial buildings with excessive energy consumption as Category II buildings. The Category I and Category II markings are fed back to guide relevant units in the low-energy-consumption optimization design of industrial buildings.

[0089] The specific functions of the above-described industrial building energy consumption anomaly monitoring system correspond one-to-one with the industrial building energy consumption anomaly monitoring methods disclosed in the previous embodiments. Therefore, they will not be described in detail here. For details, please refer to the embodiments of the previously disclosed monitoring methods. It should be noted that the various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to mutually.

[0090] In other embodiments, an industrial building energy consumption anomaly monitoring device is also provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the various steps of the industrial building energy consumption anomaly monitoring method as described in the above embodiments.

[0091] The industrial building energy consumption anomaly monitoring device may include, but is not limited to, a processor and a memory. Those skilled in the art will understand that the schematic diagram is merely an example of an industrial building energy consumption anomaly monitoring device and does not constitute a limitation on the device itself. It may include more or fewer components than illustrated, or combine certain components, or use different components. For example, the industrial building energy consumption anomaly monitoring device may also include input / output devices, network access devices, buses, etc.

[0092] The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor. This processor is the control center of the industrial building energy consumption anomaly monitoring device, connecting all parts of the device via various interfaces and lines.

[0093] The memory can be used to store the computer program and / or modules. The processor, by running or executing the computer program and / or modules stored in the memory, and by calling the data stored in the memory, realizes various functions of the device for monitoring abnormal energy consumption conditions in industrial buildings. The memory may mainly include a program storage area and a data storage area. The program storage area may store the operating system, at least one application program required for a function, etc. In addition, the memory may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0094] If the industrial building energy consumption anomaly monitoring device is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the above embodiments of the industrial building energy consumption anomaly monitoring method. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc. It should be noted that the content contained in the computer-readable medium may be appropriately added to or subtracted from the content as required by the legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, the computer-readable medium may not include electrical carrier signals and telecommunication signals.

[0095] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

[0096] In summary, the above description is only a preferred embodiment of the present invention. All equivalent changes and modifications made within the scope of the claims of the present invention should be covered by the present invention.

Claims

1. A method for monitoring abnormal energy consumption conditions in industrial buildings, characterized in that, Includes the following steps: S1, import the building model information of industrial buildings in the park, and calculate and obtain the design energy consumption index of each industrial building in the park based on the preset energy consumption assessment database. S2 collects the total energy consumption data of each industrial building through the sensor information platform, obtains the total usable area of ​​the corresponding industrial building based on the building model information, and calculates the actual energy consumption index of each industrial building as the benchmark energy consumption index. The benchmark energy consumption index is energy consumption per unit area, which is the ratio of the building's total energy consumption to its total usable area. The total energy consumption is collected through a sensor information platform, and the total usable area of ​​the building is obtained through building model information identification. S3. Based on the difference between the design energy consumption index and the benchmark energy consumption index, buildings exceeding the energy consumption standard are screened and marked as Class I buildings; the design energy consumption index and the benchmark energy consumption index values ​​are unified and normalized to a range of [0, 100], and the above process is repeated for each industrial building in the park; the input value of the design energy consumption index is denoted as X1, and the input value of the benchmark energy consumption index is denoted as X2. Then, the element x has a Y output of [0, 100] according to the following formula, corresponding to X1 and X2 to form Y1 and Y2. Where Y = 100 * (X - min) / (max - min); using the quartile method, the outer contour color of the building model is set to four different colors according to the average value of Y1 from low to high; a two-dimensional input data space of Y1 and Y2 is established, and linear regression is performed on Y1 and Y2, so that each point has a least squares distance D. y The value above the regression line is positive, and the value below it is negative. (This refers to the value of D.) y Outlier extraction is performed using the interquartile range method. If there is D... y If the interquartile range is greater than the upper bound of k=3, the energy consumption is considered to be excessive and a first-class label is applied to the corresponding building model; S4. Obtain several historical load days with high similarity to the predicted day under meteorological factors, time factors, and weekday factors from the historical sample load days as similar days to the predicted day. S5, query and obtain historical energy consumption indicators for similar days, compare the historical energy consumption indicators for similar days with the benchmark energy consumption indicators, and mark the industrial buildings corresponding to the benchmark energy consumption indicators whose comparison difference exceeds the preset threshold as having abnormal energy consumption conditions in the second category.

2. The method for monitoring abnormal energy consumption conditions in industrial buildings according to claim 1, characterized in that, Step S4 specifically includes: S41, the meteorological factor matching coefficient is calculated through grey relational analysis; S42, calculate and obtain the time factor matching coefficient, which represents the degree of similarity between the predicted day and the historical day in time; Where mod is the modulo function; t is the number of days from the i-th historical day to the predicted day; int is the integer operation; s i It is a 0 / 1 variable; when the i-th historical day and the predicted day fall on the same holiday, s i The value is 1 if it is set to 1, otherwise the value is 0. , and is the attenuation coefficient, with a value of 0.9~0.98, representing the similarity reduction ratio for each additional day, week, and year of distance between historical days and similar days, respectively; N1, N2, and N3 are constants, with N1 and N2 taking the value of 7 (the number of days in a week), and N3 taking the value of 340. S43, calculate and obtain the weekday factor matching coefficient, which represents the similarity between the predicted day and the historical day in terms of weekday type. The weekday type is quantified: Monday's mapping is 0.1, Tuesday through Thursday's mapping is 0.2, Friday's mapping is 0.3, Saturday's mapping is 0.7, and Sunday's mapping is 1. The following function expression is used to calculate the weekday factor matching coefficient between the i-th historical day and the predicted day. : ,in , These are the weekday types for the i-th historical day and similar days, with the possible values ​​being 1, 2, 3, 4, 5, 6, and 7. , yes and The mapped value; S44. Multiply the meteorological factor matching coefficient, the time factor matching coefficient, and the weekday factor matching coefficient to obtain the comprehensive matching coefficient. The larger the comprehensive matching coefficient, the closer the characteristics of the selected similar day and the predicted day are.

3. The method for monitoring abnormal energy consumption conditions in industrial buildings according to claim 2, characterized in that, Step S4 further includes: Energy consumption data of each industrial building in a defined period are obtained and arranged chronologically to form a time series. The Fbprophet prediction method is used to predict the time series results, with a prediction model in the form of a superposition function. ; It is a trend function used to fit and model non-periodic changes in a time series. It is a periodic term used to reflect the periodic changes in the model; This is a holiday item, used to reflect the impact of holiday effects that may occur within one or more days; This is the error term, used to reflect abnormal changes in the model, assuming... It follows a normal distribution, so the trend function is obtained: ,in, The expected capacity of the system; The growth rate; For compensation parameters.

4. An industrial building energy consumption anomaly monitoring system, characterized in that, include: The design energy consumption acquisition module is used to import the building model information of industrial buildings in the park and calculate and obtain the design energy consumption index of each industrial building in the park based on the preset energy consumption assessment database. The benchmarking energy consumption acquisition module is used to collect the total energy consumption data of each industrial building through the sensor information platform, obtain the total usable area of ​​the corresponding industrial building based on the building model information, and calculate the actual energy consumption index of each industrial building as the benchmarking energy consumption index. The benchmark energy consumption index is energy consumption per unit area, which is the ratio of the building's total energy consumption to its total usable area. The total energy consumption is collected through a sensor information platform, and the total usable area of ​​the building is obtained through building model information identification. The first labeling module is used to screen buildings with excessive energy consumption based on the difference between the design energy consumption index and the benchmark energy consumption index, and to label the corresponding buildings in the first category. The design energy consumption index and the benchmark energy consumption index values ​​are normalized to a range of [0, 100], and this process is repeated for each industrial building in the park. The input value of the design energy consumption index is denoted as X1, and the input value of the benchmark energy consumption index is denoted as X2. Each element x has a Y output of [0, 100] according to the following formula, forming Y1 and Y2 corresponding to X1 and X2, where Y = 100 * (X - min) / (max - min). Using the quartile method, the outer contour color of the building model is set to four different colors according to the average value of Y1 from low to high. A two-dimensional input data space of Y1 and Y2 is established, and linear regression is performed on Y1 and Y2, ensuring that each point has a least squares distance D. y The value above the regression line is positive, and the value below it is negative. (This refers to the value of D.) y Outlier extraction is performed using the interquartile range method. If there is D... y If the interquartile range is greater than the upper bound of k=3, the energy consumption is considered to be excessive and a first-class label is applied to the corresponding building model; The similar day acquisition module is used to acquire several historical load days with high similarity to the predicted day under meteorological factors, time factors, and weekday factors from the historical sample load days as similar days to the predicted day. The second labeling module is used to query and obtain historical energy consumption indicators for similar days, compare the historical energy consumption indicators for similar days with benchmark energy consumption indicators, and label industrial buildings corresponding to benchmark energy consumption indicators whose comparison difference exceeds a preset threshold as having abnormal energy consumption conditions.

5. An industrial building energy consumption anomaly monitoring device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, it implements the steps of the method as described in any one of claims 1-3.

6. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1-3.

Images